£EPA
Labo'i-
Athens GA
Development of a
Rapid Analytical
Method for
Determining
Asbestos in Water
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1 Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-6ooA-78-o66
December 1978
DEVELOPMENT OF A RAPID ANALYTICAL METHOD
FOR DETERMINING ASBESTOS IN WATER
by
Carl W. Melton, Sandra J. Anderson, Carolyn F. Dye,
W. Eugene Chase, and Richard E. Heffelfinger
BATTELLE
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-03-2199
Project Officer
Charles H. Anderson
Analytical Chemistry Branch
Environmental Research Laboratory
Athens, Georgia 30605
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30605
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory, U.S. Environ-
mental Protection Agency, Athens, Georgia, 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 en-
dorsement or recommendation for use.
11
-------
FOREWORD
Nearly every phase of environmental protection depends on a capability to identify and
measure specific pollutants in the environment. As part of this Laboratory's research on the
occurrence, movement, transformation, impact, and control of environmental contaminants,
the Analytical Chemistry Branch develops and assesses new techniques for identifying and mea-
suring chemical constituents of water and soil.
The widespread use of asbestos-containing materials gives rise to concern about exposure of
the general population to low level concentrations in air, water supplies, and food. Although
hazards associated with the inhalation of asbestos at high concentrations are recognized, the
health significance of ingested particles is not fully understood. Although electron microscopic
methods are required for the quantitative determination of asbestos levels in water, such
methods are expensive, time-consuming and require sophisticated equipment not usually avail-
able in a conventional chemical laboratory. There is a need, therefore, for a simple, rapid ana-
lytical method that can be used for the preliminary examination of a large number of water
samples. This report describes the progress that has been made toward the development of a
rapid analytical procedure for chrysotile asbestos.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
ui
-------
ABSTRACT
The development of a rapid analytical method for determining chrysotile asbestos in water
that requires substantially less time per analysis than electron microscopy methods is described.
Based on the proposition that separation of chrysotile from other waterborne particulate would
greatly simplify the task of detection, the research effort was directed toward establishing sepa-
ration and concentration techniques. This investigation led to the development of a separation
procedure whereby chrysotile is extracted from a water sample into an immiscible organic
liquid phase. The procedure is called two-phase liquid separation (TPLS).
TPLS has been combined with a light microscopic intercept counting technique, and also
with a colorimetric spot test detection technique to result in two complete rapid analytical
methods. The TPLS — light microscopic (LM) method requires more expensive equipment than
the TPLS - spot test method; however, the limit of detection of TPLS-LM method is 1.0 ng,
approximately 40,000 fibers, at the 99 percent confidence level, whereas the limit of detection
of the TPLS - spot test method is approximately 100 ng or 4.0 x 106 fibers. The TPLS-LM
method, therefore, is recommended as a first choice, and the TPLS — spot test method is
recommended to be used under conditions that require no greater detection sensitivity than
100 ng per sample.
TPLS extracts chrysotile from water into isooctane after the chrysotile surface has been
rendered hydrophobic by reaction with an anionic surfactant (dioctyl sodium sulfosuccinate).
Extraction of the chrysotile from the water phase into the isooctane phase occurs as the two
liquids are shaken in a separatory funnel. Agitation creates an emulsion that is broken by
adding sodium chloride solution. The isooctane is then filtered to deposit the chrysotile on a
filter where its concentration is analyzed by light microscopy or spot test procedures.
For light microscopic detection, the chrysotile is deposited on a Nuclepore filter, which
has a relatively smooth surface. The filtered deposit is carbon coated to make it reflective and
also to enhance the apparent fiber diameters to make more fibers detectable by light microscopy.
Intercept counts are made of fiber segment lengths equal to the spacing between concentric
circles in a calibrated eyepiece reticule. The measurement of total fiber length as the diameter
of the filter is traversed is referred to a standard curve to obtain mass of chrysotile per filter
deposit.
For spot test detection, the extracted chrysotile is deposited in a small (3-mm diameter)
area on a Millipore filter. The chrysotile is reacted with manganous ion and washed, and the
reacted manganous ion is detected by a spot test reaction between potassium periodate and
tetrabase, which is catalyzed by manganese. Varying shades of blue serve to indicate different
concentrations of chrysotile.
IV
-------
Experimental results have shown that 75 percent of the chrysotile is extracted by TPLS
procedures. This is taken into account when evaluating analytical results.
This report was submitted in fulfillment of Contract No. 68-03-2199 by Battelle's
Columbus Laboratories, under the sponsorship of the U.S. Environmental Protection Agency.
This report covers the period from June 1, 1975, to June 30, 1978, and work was completed.
as of June 1978.
-------
CONTENTS
Foreword iii
Abstract iv
Figures viii
Tables x
Acknowledgments xi
1. Introduction 1
2. Conclusions 3
3. Recommendations 5
4. Experimental Work 6
Phase I: Separation Experiments 6
Initial Separation Experiments 6
Dielectric Separation 9
Two-Phase Liquid Separation 13
Evolution of TPLS for Chrysotile 13
Factors Affecting TPLS Optimization 21
Effects of Critical TPLS Factors on Selectivity and
Percent Recovery 22
TPLS Recovery
22
TPLS Selectivity 25
Handling Large Water Sample Volumes by TPLS 27
TPLS Recovery as a Function of Particle Size and
Aspect Ratio 29
Phase II: Investigation of Detection Methods 33
Change of Flow Rate as a Function of Chrysotile
Concentration 33
Spot-Test Detection 34
X-Ray Diffraction 35
Light Microscopy 36
Phase III: Application of Analytical Procedures 39
TPLS — Light Microscopic Analysis of Real Water Samples 39
TPLS - Spot Test Analysis of Real Water Samples 45
5. Intel-laboratory Analysis of Four Water Samples 49
References 52
Appendixes
A. Instructions for Carrying Out Two-Phase Liquid Separation Coupled
With Light Microscopic Detection of Chrysotile in Water Samples 53
B. Instructions for Carrying Out Two-Phase Liquid Separation (TPLS) Coupled
With Spot Test Detection of Chrysotile in Water Samples 63
vii
-------
FIGURES
Number Page
1 CPE instrument 8
2 Electron micrographs of effluents collected in test tubes 22 through 29
in a CPE run made at pH 8.6 showing how pure chrysotile is confined
to a band position in the curtain cell 10
3 Electron micrographs of effluents collected in CPE test tubes 20 through 28
showing separation of diatoms from chrysotile in a run at pH 10.7 .... 11
4 An electron micrograph of the original mixture of diatoms and chrysotile
before CPE separation at pH 10.7 12
5 Chrysotile collected on the anode in the dielectric cell containing
isopropyl alcohol 14
6 Chrysotile separated from diatomaceous earth in the first TPLS
experiment 16
7 Particulate remaining in the water phase in the first TPLS experiment.... 16
8 Distribution of chrysotile fiber lengths in sample 96-1 before (•) and
after (A) TPLS 31
9 Distribution of chrysotile fiber aspect ratios in sample 96-1 before (•)
and after (A) TPLS 31
10 Distribution of chrysotile fiber lengths in sample 96-2 before (•) and
after (A) TPLS 31
11 Distribution of chrysotile fiber aspect ratios in sample 96-2 before (•)
and after (A) TPLS 31
12 Distribution of chrysotile fiber lengths in sample 96-3 before (•) and
after (A) TPLS 32
13 Distribution of chrysotile fiber aspect ratios in sample 96-3 before (•)
and after (A) TPLS 32
Vlll
-------
14 Distribution of chrysotile fiber lengths in sample 96-4 before (•)
and after (A) TPLS 32
15 Distribution of chrysotile fiber aspect ratios in sample 96-4 before (•)
and after (A) TPLS 32
16 Plot of pressure drop and flow rate vs chrysotile concentration 34
17 The eyepiece reticule used in fiber counting 37
18 Standard curve that relates light microscope counts to mass (/ug) of
chrysotile per filter 41
19 Light micrograph showing a heavy mat of TPLS-extracted pure
chrysotile from sample M3 0307 1500 D2 WGO 47
20 Comparison of spot test results on real water samples to spot test
results on a series of chrysotile standards 48
A-l The mechanical shaker 57
A-2 Separatory funnels attached to mechanical shaker with rubber bands .... 57
A-3 Before (right) and after (left) breaking the emulsion by "salting out"
with NaCl 58
A-4 Draining off the water phase 59
A-5 Filtering the isooctane 59
A-6 A TPLS specimen prepared for light microscopic analysis. The filtered
TPLS fraction was taped to a microscope slide and coated with vapor-
deposited carbon in a vacuum evaporator 60
A-7 Concentric circles in eyepiece reticle superimposed on image of TPLS
fraction from ORF water sample 61
A-8 Standard curve which relates light microscope measurement of total
fiber length to mass (jug) of chrysotile per filter 62
B-l Special reservoir with 3-mm-diameter hole 66
B-2 Spraying device with vial containing spot test working solution 66
B-3 A typical spot test result obtained on an 0.8-/jg chrysotile standard 67
IX
-------
TABLES
Number Page
1 Effect of Agitation by Stirring 24
2 Percent Recovery With 2.5 x ID'3 M MT-70 24
3 Percent Recovery With 1.0 x 1(T3 M MT-70 26
4 Percent TPLS Recovery of Particles Detectable by Light
Microscopy 38
5 Detection Limit Determinations 39
6 Light Microscopic Detection of Chrysotile at Twenty
Concentrations 40
7 Results of TPLS Applied to Real Water Samples 42
8 LM Analytical Results on Real Water Samples 46
9 Spot Test Results on Real Water Samples 48
10 Comparison of Analytical Results Obtained at Battelle and
ORF on Four Water Samples 51
-------
ACKNOWLEDGMENTS
The authors wish to acknowledge Mr. George Reimschussel of the Johns Manville Research
Laboratory, Denver, for supplying samples of pure chrysotile having different electrophoretic
mobilities. They also wish to express appreciation to Dr. Eric Chatfield and Ms. M. Jane Dillon
of the Ontario Research Foundation, Toronto, Canada, for their aid in the evaluation of two-
phase liquid separation coupled with light microscopic detection and to Dr. C. H. Anderson for
his helpful suggestions and interest in the development of analytical methodology for chrysotile
in water.
XI
-------
SECTION 1
INTRODUCTION
The presence of asbestos in water supplies in the United States and Canada, as well as
other parts of the world, has led to the need to rapidly analyze large numbers of natural- and
industrial-water samples for asbestos content. Streams, runoff, lakes and city water systems
are the major receptacles of asbestos contamination. Natural weathering of asbestos deposits,
eroding of asbestos-cement city water pipes, and dumping of tailings from mining operations
are some of the mechanisms by which water supplies become contaminated with asbestos. In
the past, transmission electron microscopy (TEM) has been the primary method for deter-
mining asbestos concentrations, but the time and cost required to analyze a sample becomes
prohibitive when dealing with a large number of sampling sites. Clearly, there is a need for a
more rapid and less expensive analytical method.
The objective of the research program reported here was to develop a rapid analytical
method for asbestos in water requiring substantially less time and cost and less sophisticated
equipment than a method based on TEM. Research was performed to develop a rapid monitor-
ing method that would eliminate from further examination those samples containing asbestos
concentrations below significant levels. After screening, only those samples identified as having
significant quantities of asbestos would be submitted for more definitive TEM analysis.
Asbestos analytical methodology employs TEM because the fiber sizes range well below the
resolution capabilities of the light microscope, and health effects investigators require fiber-size-
frequency data. This does not imply that the rapid method must obtain fiber sizes. It need
only identify those samples that need to be fully characterized. Therefore, the detection
method employed in the rapid procedure is based on bulk analysis.
The logic behind the approach taken in the development of a rapid method was guided by
an awareness of problems arising from gross amounts of interfering particulate prevalent in most
water samples. The sample could be rid of interfering organic matter by low-temperature
ashing, but the inorganic particulate could not easily be eliminated to leave only the asbestos.
Analysis either had to be carried out in the presence of both organic and inorganic particulate
or in the presence of inorganic particulate or the asbestos had to be separated from the inter-
fering particulate. Analysis could be performed for separated and concentrated asbestos with
much less difficulty than for asbestos in the presence of interfering particulate. Accordingly,
our research effort emphasized the development of a separation method for asbestos in order
to facilitate its subsequent detection. The separation of chrysotile was investigated first because
chrysotile is the most prevalent form of asbestos found in most sampling locations. Also, its
surface chemistry is sufficiently different than most other waterborne particulate to make its
selective separation possible.
-------
The goal has been to develop a separation method that will not only eliminate interfering
extraneous participate, but will also reduce the time required for specimen preparation. Speci-
men preparation methods employed in conjunction with electron microscopy detection often
require lengthy ashing periods to rid the sample of organic matter, but fail to eliminate masking
inorganic particles from the ash residue. The developed separation method does not include an
ashing step, but does selectively and rapidly separate and concentrate the asbestos directly from
the original water sample.
Three approaches to chrysotile separation were explored. All exploited differences in
electrical properties between chrysotile and other prevalent waterborne particles. The methods
investigated were continuous particle electrophoresis (CPE), dielectric separation, and extraction
of asbestos from aqueous suspension into an immiscible oil phase (two-phase liquid separation).
Although CPE proved promising, there was greater success with the two-phase liquid extraction
technique. Consequently, most of the research effort was directed toward the study of two-
phase liquid separation (TPLS).
Attempts to apply TPLS to amphibole separation were not successful; however, this was a
minor research effort. Early in the program it was decided to concentrate on chrysotile separation
rather than concurrently try to devise a procedure for the separation of amphiboles. It is believed
that the separation of amphiboles by modified TPLS procedures remains a possibility.
The complete analytical method for chrysotile includes TPLS coupled with light micro-
scopic analysis of the chrysotile in the separated fraction. Alternatively, the separated and
concentrated chrysotile is analyzed by a catalytic spot test procedure.
Two-phase liquid separation promotes the selective transfer of chrysotile from water to an
immiscible organic liquid after the chrysotile surface has been rendered hydrophobic by reaction
with an anionic surfactant. The organic liquid is then filtered to deposit the separated chrysotile
on a Nuclepore filter. The filtered chrysotile is carbon coated and viewed with a light micro-
scope employing bright field incident illumination at 500X, and counts are made of fiber
segments as the diameter of the effective filter area is traversed. The counts thus obtained are
translated into chrysotile mass (jug) per filter by means of a standard curve that was generated
using standard filtered chrysotile suspensions.
The spot-test analytical procedure is applied to the TPLS fraction that has been deposited
on an «3 mm diameter area on a Millipore filter. First, the deposited chrysotile is reacted with
manganous ion and washed, and then the reacted manganese is detected by a chromogenic spot
test reaction between potassium periodate and tetrabase, which is catalyzed by manganese.
The limit of detection of the TPLS - light microscopic method is 1 nanogram at the
99 percent confidence level. The smallest quantity of chrysotile detectable by the spot test is
100 nanograms. Although the spot test is much less sensitive, it can be used effectively in
certain monitoring situations. The relationship between mass and total number of chrysotile
fibers is variable and strongly dependent upon fiber dimensions. All water samples show a
broad range of fiber size distribution; for simplification in this work it has been assumed that
the average fiber is 1 jum long and 0.1 jum wide. When this assumption is made 1 ng is
equivalent to approximately 40,000 fibers.
-------
SECTION 2
CONCLUSIONS
(1) Continuous particle electrophoresis (CPE) will separate chrysotile, but CPE will not
handle sufficiently large volumes of particulate to be used in an analytical method.
(2) Dielectric separation as investigated is ineffective for chrysotile.
(3) Two-phase liquid separation of chrysotile from water samples coupled with light
microscopic analysis of the separated chrysotile by intercept counting will detect
chrysotile quantities as low as 1.0 ng at a 99 percent confidence level. Assuming an
average chrysotile fiber size of 1 /xm length and 0.1 jum width, 1 ng is the equivalent
to approximately 40,000 fibers.
(4) TPLS extracts 75 percent of those chrysotile particle sizes detectable by light
microscopy.
(5) The sensitivity of the light microscopic detection procedure is high over the concentra-
tion range above 1.0 ng; an average of nine counts is registered at 1.0 ng and 15,000
counts at 1000 ng.
(6) Few interferences occur in TPLS; it is surprisingly selective. Organic filaments and platy
mineral particles, however, are occasionally extracted along with chrysotile. The fila-
ments can be distinguished from chrysotile with a little practice. The presence of the
platy mineral can result in some of the chrysotile being obscured from view.
(7) The TPLS — spot test procedure requires relatively inexpensive equipment and can be
run with a minimum of training by an individual with no specific background.
(8) The detection limit of the spot test procedure is approximately 100 ng.
(9) The spot test is selective for chrysotile; no interferences were encountered while running
real water samples.
(10) TPLS can separate as large as 0.4-mg quantities of chrysotile from 100-ml water aliquots.
(11) Degree of agitation, pH, concentration of surfactant, the ratio of the volumes of oil/
water, and the number of extractions from a single water aliquot are the most critical
TPLS factors.
-------
(12) Percent chrysotile recovery by TPLS is optimized:
• at pH 3.5 to 5.0
• when the oil/water volume ratio is 1:2
• when shaking is the mode of agitation
• when 2.5 x 10~3 M anionic surfactant concentration is used
• when five extractions are made from the same water aliquot.
(13) TPLS favors the extraction of longer fibers with larger aspect ratios.
(14) Filtration and resuspension is a practical means of concentrating particulate from large
water volumes for TPLS.
(15) It is not as feasible to scale up TPLS to handle large water volumes as it is to filter and
resuspend particulate in smaller, more manageable water volumes for TPLS.
(16) The results of the interlaboratory analyses at Battelle and ORF were in agreement
within a factor of 4, except on Sample No. 4 in which Battelle found only 1/10 of the
chrysotile reported by ORF. Battelle's TPLS — light microscopic result in this case
agreed with the electron microscopic result obtained at ORF.
(17) Approximately eight samples can be prepared and analyzed by the TPLS — light
microscopic method by two people in one day.
-------
SECTION 3
RECOMMENDATIONS
As with any new analytical technique, one must proceed with caution and become
thoroughly familiar with the application of two-phase liquid separation, light microscopic
detection and spot test application before relying on the analytical results. It is recommended,
therefore, that the results of these procedures be compared initially to results obtained on the
same samples by electron microscopy. Afterwards, when there is sufficient confidence in the
method's effectiveness, one can progress toward the application of TPLS coupled with either of
the detection procedures as requirements dictate without comparison to electron microscopic
analytical results.
Although water samples were analyzed using both TPLS coupled with light microscopic
counting and the TPLS — spot test procedure, it was not possible to investigate all varieties of
water samples. Thus, unanticipated interferences may occur to give high or low results under
sample conditions not investigated.
It is recommended that the developed methodology be applied first to monitoring varia-
tions in chrysotile concentrations from well characterized sources, and then, as experience
develops, be applied to the analysis of chrysotile from different sources.
-------
SECTION 4
EXPERIMENTAL WORK
This section of the report discusses the experimental work performed to develop separation
and detection methods for chrysotile. Separation experiments explored dielectric separation,
continuous particle electrophoresis, and two-phase liquid separation. When promising results
were obtained using TPLS to separate chrysotile, emphasis was placed on the optimization of
this process. No further search was made for other separation methods for chrysotile. Also,
when initial experiments to separate amphiboles by various modifications of TPLS were not
completely successful, it was decided to concentrate the entire effort on improvement of the
separation of chrysotile.
Several detection methods were investigated with the objective of developing a method
requiring less time and less expensive equipment than electron microscopy. Spot-test reactions,
changes in flow rate through filtered chrysotile deposits, X-ray diffraction, and light microscopy
were evaluated. Light microscopy of the separated and filtered chrysotile was the most success-
ful detection method found, and it has been written into the detailed instructions for the final
method. A spot-test procedure was also developed for chrysotile detection and is recommended
as an alternative.
PHASE I: SEPARATION EXPERIMENTS
Initial Separation Experiments
Continuous Particle Electrophoresis —
The Beckman Continuous Particle Electrophoresis (CPE) System was used in experiments
designed to separate and concentrate chrysotile from synthetic mixtures of chrysotile and
diatomaceous earth. The operating principle of CPE and related experiments are described in
the following sections.
Operating Principle of CPE —
Simply stated, the operating principle of Beckman CPE is as follows. A suspension of
sample material is introduced as a fine, steady stream into a vertical, flowing "curtain" of
electrolyte. A horizontal dc electrical field is applied to the curtain. Each particle has two
components of motion: a vertical component (Vv) that is the same as the curtain velocity, and
a horizontal component (V^) that is proportional to the electrophoretic mobility or zeta
potential of the particle. Each different particulate component is therefore deviated a different
angle, a, from the vertical, given by
tan a = ~
vv
-------
The net result at the lower end of the cell is a particle band position that is proportional to the
zeta potential of the particles. Where a mixture of particles having different electrophoretic
mobilities is analyzed, each type of particle will assume a different band position.
The possibility of fractionating different types of particles by CPE assumes a certain mini-
mum difference in their electrophoretic mobilities. The relative mobilities in a mixture can
frequently be manipulated. One useful procedure is to change the pH. Others include the use
of polyvalent cations, and use of surfactants and various agents that alter the surface char-
acteristics of the particles.
The CPE instrument is shown in Figure 1. This system can be separated into five sub-
sections: (1) curtain flow system, (2) sample flow system, (3) electrode rinse systems,
(4) cooling system, and (5) power supply. The curtain flow system provides a filtered source
of the electrolyte at a constant flow rate to the cell. The sample flow system provides a means
of continuously metering a sample into the flowing curtain medium. The electrode rinse system
is designed to remove electrolysis products from the electrodes and thereby maintain a high
degree of pH stability within the curtain. The power supply is designed to provide a constant
voltage gradient within the cell regardless of changes in internal impedance or electrolyte
conductivity.
The electrodes, located in two chambers positioned at the sides of the curtain cell, are
isolated from the curtain by semipermeable membranes. This separation prevents mixing of the
electrode rinse with the curtain buffer. The sample is introduced at the top of the curtain and
is thereby diluted by the curtain buffer solution as it flows downward between the electrodes.
The sample particles are influenced by the voltage gradient to migrate laterally, and finally to be
discharged through drainage tubes positioned at uniform (1 mm) intervals across the bottom of
the curtain cell. The effluent is thus divided into 48 fractions and collected in test tubes.
According to the instructions in the use of the CPE system, most particles possess a higher
zeta potential at higher pH conditions. Consequently, they recommend the utilization of buffer
media at relatively high pH values. The instructions also suggest the use of a 30 V/cm voltage
gradient for most applications. Therefore, most of the initial CPE experiments were carried out
at pH values greater than 7.0 — employing a voltage gradient of 30 V/cm.
CPE Experiments —
CPE experiments were conducted at pH 8.6, 3.6, and 10.7. Runs were made at pH 8.6
and 10.7 because the CPE instructions stated that better separation usually was effected at
higher pH conditions. The run at pH 3.6 was made because literature^) indicated that the
greatest positive surface charge was developed on chrysotile at this pH.
The first attempt to separate chrysotile from diatomaceous earth was made in a veronal
buffer at pH 8.6. An initial run was made with pure chrysotile dispersed in filtered, deionized
water containing 1.9% aerosol OT to determine into which tubes the chrysotile would be dis-
charged. It was found that the bulk of the chrysotile was collected in the test tubes numbered
20 through 29. This was determined by filtering 5-ml aliquots from each tube onto 25-mm
HAWP Millipore filters and examining prepared specimens by electron microscopy. The speci-
mens were prepared by cutting a piece from the center of the filter and dissolving the filter
material away in an acetone bath while the deposit side was in contact with a carbon support
-------
Cell F il ling Control
Magnetic Sample Holder
Sample Inlet Tube
Electrophoresis Cell
Curtain Flow Meter
Fractionation Viewing Area
Collecting Tubes
48-Tube Fraction Collector
Cell Current Monitor
Cell Voltage Monitor
Voltage Gradient Monitor
Vcltage Gradient Adjust
Voltage Gradient Runye Switch
Curtain Pump Speed Control
Curtain Pump Switch
— Rinse Pump Switch
Cell Voltage Switch
Power Swi tch
Figure 1. CPE instrument.
-------
film held by a mesh electron microscope grid. Electron micrographs were taken of preparations
containing asbestos, and those micrographs showing the bulk of the asbestos were contact
printed. The resulting contact print is included in Figure 2.
A second run at pH 8.6 was made on pure diatomaceous earth also dispersed in filtered
deionized water containing 1.9% aerosol OT. The diatoms were discharged in the same tubes
as was the chrysotile; consequently, no further experiments were performed at pH 8.6.
Subsequently, separations were attempted at pH 3.6 (phosphate buffer) at which agglomer-
ation of particulate resulted in plugging of the tubes through which the effluent from the
curtain cell was discharged. The addition of the dispersant Aerosol OT did not prevent
agglomeration at pH 3.6.
Next, runs were made at pH 10.7 (in 0.001 M Na2CO3> first separately on pure chrysotile
and pure diatomaceous earth dispersed in 1.9% Aerosol OT in deionized water. These experi-
ments indicated that separation might be obtained at this pH; the chrysotile was concentrated
primarily in tubes 20 through 24 and diatoms in tubes 25 through 30. Then, a run was made
in which separation was attempted on a 10:1 by weight (7.0 mg and 0.7 mg in 100 ml H2O)
mixture of diatomaceous earth and chrysotile, respectively. This mixture also was dispersed in
1.0% Aerosol OT in filtered deionized water. As predicted by the runs on the pure materials,
a separation was effected. Figure 3 shows electron microscope preparations from test tubes
numbered 20 through 27. Although some diatoms were found in tubes 22, 23, and 24, the
degree of separation represents a marked improvement over the original mixture shown in
Figure 4.
Possibly, deviation from the 30-V/cm voltage gradient would yield even better separations.
Also, instrument instructions state that lower curtain flow rates result in better separations.
All experiments were conducted at a curtain flow rate of 20 ml/minute with a sample introduc-
tion rate of 50 /xl/minute. Of course, as the curtain flow and sample introduction rates are
decreased, longer times are required for CPE separations. Compromises must be made in
achieving the desired balance between degrees of CPE separation and run duration.
At this point, concurrent experiments with two-phase liquid separation proved more
successful, and CPE studies were discontinued.
Dielectric Separation
Dielectric separation^) is based on the relationship between the dielectric constants of
suspended particulate material and that of the suspending liquid. When the suspending liquid
has a lower dielectric constant than the particle, the particle is attracted to immersed electrodes
when either an ac or dc voltage is imposed. Theoretically, when the difference is the least
between the dielectric constants of the liquid and the particle type to be separated, the most
effective separations are obtained.
First experiments were performed on pure chrysotile suspended in liquids having relatively
low dielectric constants compared to chrysotile (dielectric constant = 33.7) in order to study the
attraction of chrysotile to the electrodes at various voltages under both ac and dc conditions.
Most of these trials were carried out in Freon MF having a dielectric constant of 2.4. It was
-------
,-
>
x
-'
; -v -5
"%/
/;'
,
p .. »ffl&
4d
22
23
24
25
26
27
29
Figure 2. Electron micrographs of effluents collected in test tubes 22 through 29 in a CPE run made
at pH 8.6 showing how pure chrysotile is confined to a band position in the curtain cell.
5000X
-------
20
22
23
25
26
27
28 5000X
Figure 3. Electron micrographs of effluents collected in CPE test tubes 20 through 28 showing separation
of diatoms from chrysotile in a run at pH 10.7.
-------
found that chrysotile was best attracted and held at 200 Vdc to electrodes separated by a
distance of approximately 1 cm. Below 200 V the chrysotile did not collect as rapidly on the
electrodes, and above 200 V a turbulence developed that dislodged much of the deposit.
Consequently, 200 Vdc was employed in subsequent dielectric separation experiments.
5000X
Figure 4. An electron micrograph of the original mixture of diatoms and
chrysotile before CPE separation at pH 10.7.
Although chrysotile was deposited on the electrodes, it was not held well enough to lift
the electrodes from the liquid and thereby remove the deposit. In order to be able to retain
the chrysotile deposit on the electrodes while undeposited particulate was removed, a flow-
through cell was designed. The cell design permitted gentle flushing of the liquid containing
undeposited suspended particulate from the cell without disturbing the chrysotile deposit. A
reservoir that contained particle-free liquid and that had a drain on the side delivered this
liquid to the dielectric cell positioned below. The dielectric cell, which held 40 ml, also had a
drain at the bottom which could be adjusted to a flow which matched the rate of introduction
of the particle-free liquid. After the deposit had formed on the electrodes, the unattached
particulate was flushed from the cell with 400 ml of liquid. Then the deposited participate
was resuspended in the remaining liquid and drained from the cell into a separate container.
Investigations of dielectric separation were carried out primarily in methyl, ethyl, and
isopropyl alcohols because their dielectric constants are below that of chrysotile but relatively
high compared to most other liquids. Also their dielectric constants can be altered upward with
the addition of water to create solutions that are even nearer to the dielectric constant of
-------
chrysotile. For example, water has a dielectric constant of 80 and ethyl alcohol a constant of
24. The change in dielectric constant is a straight line function, and thus 90% ethyl alcohol-
10% water solution has a dielectric constant of 30. Separation trials were made, therefore,
first with pure ethyl, methyl, and isopropyl alcohols and later with ethyl alcohol to which 10%
water was added to elevate the dielectric constant.
First, separate suspensions of pure chrysotile and pure diatomaceous earth were run in
various candidate liquids. In cases where it was observed that chrysotile was deposited and
diatomaceous earth was not, 10:1 mixtures by weight of diatomaceous earth and chrysotile
were run. Figure 5 illustrates how chrysotile was deposited on the anode from an isopropyl
alcohol suspension of chrysotile. Despite the fact that this was originally a suspension of
chrysotile fibrils in the colloidal size range, a visible deposit was formed on the electrode.
Always before making a run, the particulate was dispersed ultrasonically in the dielectric
liquid without the aid of additives. On the basis of previously described experiments, pure
methyl alcohol, pure isopropyl alcohol, and 90% ethyl alcohol-10% water solution were used in
dielectric separation experiments. Although in each case some enrichment of asbestos was
realized, the results were inferior to continuous particle electrophoresis and two-phase liquid
separation. It appears that interaction may occur between chrysotile and diatoms to result in
the attraction of both species to the electrode. Also, inevitably, as asbestos particles agglom-
erated, a mass was formed that trapped diatoms as they circulated in the dielectric liquid.
Experimentation with dielectric separation was curtailed in favor of the pursuit of the
more promising two-phase liquid separation.
Two-Phase Liquid Separation
Two-phase liquid separation was conceived as a possible means of chrysotile separation.
First experiments were successful, but success occurred somewhat accidentally. The following
section discusses the process of TPLS discovery.
Evolution of TPLS for Chrysotile
The first TPLS experiment was conducted in an attempt to separate chrysotile from a
10:1 by weight synthetic mixture of diatomaceous earth and chrysotile that had been used in
CPE experiments. The original mixture contained 7.0 mg of diatomaceous earth and 0.7 mg of
chrysotile dispersed ultrasonically in 100 ml of 0.001 N Na2CC»3 containing 0.1% Aerosol OT.
The mixture had been treated ultrasonically to create a dispersion that would not clog the CPE
discharge tubes.
Three ml of the original synthetic mixture was diluted by the addition of 10 ml of 0.001 N
sodium carbonate, combined with 50 ml of benzene containing 0.5 ml of oleic acid, and agitated
ultrasonically in a separatory funnel. The oleic acid addition was made because it was stated'*)
that long-chain aliphatic acids are chemisorbed by chrysotile. It was reasoned that chemisorp-
tion of oleic acid would create a hydrophobic surface on the chrysotile fiber and thus promote
its transfer to the benzene phase.
13
-------
Figure 5. Chrysotile collected on the anode in the dielectric
cell containing isopropyl alcohol.
-------
An emulsion formed upon agitation separated into a benzene layer, a cream layer, and a
water layer after standing overnight. One ml of the benzene layer was pipetted off and filtered
through a 25-mm Millipore filter having a 0.45 /xm pore size. Ten drops of the water layer were
also filtered. Both filtered residues were prepared for electron microscopy. When the prepara-
tions were examined in the electron microscope it was found that the particulate in the benzene
phase was comprised almost entirely of chrysotile (Figure 6) and that the particulate in the
water phase was primarily diatomaceous earth (Figure 7).
These experimental results set the stage for further experiments designed to determine the
critical factors involved in TPLS.
Results of the first experiment indicated that the oleic acid addition might be effecting a
separation as predicted. In additional experiments employing comparable synthetic mixtures
of 10:1 by weight of diatoms and chrysotile suspended in water containing no sodium
carbonate or Aerosol OT, however, the addition of oleic acid did not promote chrysotile
separation. Then, 0.05% of Aerosol OT was added to the water suspension, and another trial
was made in which 50 ml of the water suspension was shaken in a separatory funnel with 50 ml
of benzene containing 0.5 ml of oleic acid; again chrysotile was found to be selectively trans-
ferred to the benzene phase. Subsequent experiments in which the oleic acid was eliminated,
but in which Aerosol OT was added, effected chrysotile separation. Other experiments without
Aerosol OT or any other additions resulted in no chrysotile separation. It was concluded,
therefore, that Aerosol OT was reacting with the surface of the chrysotile to promote its
transfer to the benzene phase.
This discovery led to the investigation of TPLS mechanisms and conditions that affect
percent recovery and selectivity. The early experiments were carried out with benzene; how-
ever, because of the possible toxicity of benzene, isooctane was ultimately used as the oil phase.
The following section discusses the mechanisms identified through pertinent literature and
experimentation.
The Mechanisms of TPLS -
Essentially, two-phase liquid separation selectively extracts chrysotile from a water sample
into a water-immiscible oil phase, such as benzene, mineral spirits, or isooctane, after the sur-
face of chrysotile fibers is made hydrophobic through a reaction with an anionic surfactant.
TPLS selectivity is based on the difference between the zeta potential of chrysotile and the
zeta potentials of most other waterborne particulate. The surface charge of chrysotile is posi-
tive over a pH range from 2 to 11.4; whereas most other waterborne particulate is negative
over this pH range. The chrysotile, therefore, selectively reacts with the negatively charged
anionic surfactant, becomes hydrophobic, and is transferred from the water phase to the oil
phase.
Speil and Leinweber^^ state that chrysotile reacts with certain anionic wetting agents,
such as Aerosol OT, and that this reaction is accompanied by a strong chemisorption of the
agents with permanent modification of the surface. Because MgOH groups of the brucite layer
are presented for reaction on the surface of chrysotile(3), the Aerosol OT (dioctyl sodium
sulfosuccinate) most likely reacts with the brucite layer to create a tightly bound organic coat-
ing covering the fiber surface.
15
-------
t _ I
7500X
Figure 6. Chrysotile separated from diatomaceous earth in the first TPLS experiment.
J22542
J22545 7500X
Figure 7. Particulate remaining in the water phase in the first TPLS experiment.
16
-------
Reinders(4) has defined interfacial tension conditions governing transfer of dispersed solids
from the aqueous phase to the oil phase. If 7WO is the interfacial tension of the oil-water
interface, 7SW is the interfacial tension of the solid-water interface, and 7SO is the interfacial
tension of the solid-oil interface, the following conditions may exist.
(1) If 7SO > 7WO + 7SW, the solid remains suspended in the water phase.
(2) If 7SW > 7WO + 7SO, the solid is transferred to the oil phase.
(3) If 7WO > 7SO + 7SW, or if none of the interfacial tensions is greater than
the sum of the other two, the solid will concentrate at the oil-water
interface.
Condition (1) exists when the solid surface is hydrophilic and condition (2) or (3) may
exist when the solid surface is hydrophobic. Experimentally, it can be determined which condi-
tion prevails by contact angle measurement.
No contact angle measurements have been made during the investigation of the chrysotile
separation. They are rather difficult to make on a fibrous mineral because the surface cannot
be polished smooth.
The sign and magnitude of the zeta potential of the mineral surface is affected by the pH.
In turn, the zeta potential determines whether and to what degree a reagent will react with the
mineral surface.
In the case of stoichiometric chrysotile, the zeta potential is positive in the pH range below
the isoelectric point (zero point charge), which occurs at pH 11.4, attaining a maximum positive
value in the range pH 3 to pH 4. Above pH 11.4 the zeta potential of chrysotile is negative.
Having a positive zeta potential at pH 7.0 and below is relatively unique among water-
borne particles. This phenomenon provides the basis for separation of chrysotile from other
particles. By making its surface hydrophobic through reaction with anionic reagents at
7WO + 7SO is created to promote selective transfer of chrysotile
to the oil phase. Aerosol OT, MT-70, and MO-70 surfactants, which are anionic analogous
alkyl sodium sulfosuccinates, have been employed successfully in the separation of chrysotile.
The idealized formula for stoichiometric chrysotile^) is Mg3Si2C>5 (OH)4- The fibrils have
a rolled sheet structure composed of an outer layer of magnesia bonded to a layer of silica.
Hydrated chrysotile presents MgOH groups for reaction with reagents. Chowdhury and
Kitchener state that the higher the magnesia content the greater is the positive zeta potential.
Also, there is evidence that a greater degree of hydration of the magnesia surface increases
electrophoretic mobility. When all magnesia is leached out under acid conditions or by
weathering, a silica gel pseudomorph that is left has a strongly negative zeta potential. This
structure, not meeting the compositional criteria, is not chrysotile.
According to Lai and Fuerstenau(6), anionic reagent (collector) adsorption by or reaction
with the mineral surface is primarily determined by the reagent concentration and the pH as it
affects zeta potential. When reagent concentration and mineral zeta potential are optimum, the
adsorption density of the reagent, as expressed in moles/cm^, on the mineral surface is at a
17
-------
maximum. As a result, the hydrophobicity also becomes the highest possible with that par-
ticular reagent.
The degree of hydrophobicity is also a function of reagent type. Those reagents having
larger hydrophobic alkyl groups render the mineral surface more hydrophobic. Lai and
Fuerstenau also found that the percent recovery of alumina from water suspension into
isooctane, using sodium akyl sulfonate reagents with C8, CIO, C12, and C14 alkyl groups,
increased markedly with larger carbon chain length.
Heeding this finding, a search was made for a sodium sulfosuccinate having larger alkyl
groups than the octyl groups of Aerosol OT. A reagent was found, Monawet 70 (MT 70),
which is ditridecyl sodium sulfosuccinate. Experiments with this compound established that
it also promoted transfer of chrysotile to the oil phase. Subsequent experiments (to be
discussed later) indicated improved recovery but also reduced selectivity when compared to
dioctyl sodium sulfosuccinate (MO-70 and Aerosol OT).
Most of the literature discusses adsorption of reagents. The reaction of the sulfo-
succinates with the chrysotile surface may be more than adsorption, however: Speil and
Leinweben') state that the highly reactive chrysotile surfaces causes many interesting surface
reactions to take place that are intermediate between simple adsorption and true chemical re-
action. They also report that the reaction of chrysotile with certain anionic wetting agents,
such as Aerosol OT, is accompanied by a strong chemisorption of the reagents with permanent
modification of the surface.
In TPLS, an organic liquid that is immiscible with water is agitated with water, and the
transfer of suspended chrysotile from the water phase to the oil phase is promoted by the
anionic surface active agent. The degree of interpenetration of phases as it affects the total
interfacial contact is of critical importance in recovery of particles. Interfacial contact is
determined by the intensity and mode of agitation, the oil/water volume ratio, and the
surfactant concentration that controls emulsion formation.
Continual agitation is more effective than intermittent agitation because the droplets are
not allowed time to form a stable emulsion. Consequently, the total interfacial contact becomes
a function of agitation time and intensity.
The anionic reagent concentration must be high enough to lower the 7WO to the point that
contact between the phases is optimized. Lai and Fuerstenau found that a 2.5 x 10~4 M con-
centration of C14 sulfonate promoted maximum alumina recoveries.
The water/oil volume ratio of 25% contains the minimum oil volume to attain close pack-
ing of ideal uniform spheres. Lai and Fuerstenau state that, under these conditions, the pos-
sible phase inversion from an oil-in-water to a water-in-oil emulsion is diminished.
As one might anticipate, the more wetting agent added, the more stable and finely
divided was the emulsion formed during agitation. When effective concentrations of surfactant
were added, the time required for the emulsion to break and separate completely into two
layers became a problem. The dispersed oil droplets first coagulated, rose to the top of the
water layer to form a cream layer; then the cream layer very slowly coalesced contributing
18
-------
to the formation of a benzene layer above the cream layer. Complete coalescence of the cream
layer sometimes required as long as overnight.
Emulsification and Demulsification - There are conflicting requirements in the TPLS sys-
tem. Ideally, a finely divided oil in water emulsion must be formed to extract the major
quantity of asbestos present, and then this emulsion must break rapidly and completely into
a water layer and an oil layer. The more finely divided the emulsion, the more stable it is.
The time required for complete separation becomes too long, without some aid to accelerate
the process.
Emulsification — As previously described, the first dispersion method employed to form
the emulsion was ultrasonic treatment. Then, merely hand shaking was used. In either case,
when «0.04% of Aerosol OT was added to the water phase or benzene phase, an emulsion was
formed that required a standing period of approximately 12 hours to break completely.
Literature^) indicates that intermittent shaking was more effective in emulsion formation
than uninterrupted shaking. Sherman suggests that it requires a small amount of time for the
surface-active reagent to be adsorbed on the surfaces of newly formed droplets and to stabilize
them fully. Experimentally he found that five intermittent shakings (2 shakes each) with a rest
period of 20 to 30 seconds between shakings would completely emulsify 60% by volume of
benzene in 1% aqueous sodium oleate. It required 3000 uninterrupted shakes in a machine,
ksting about 7 minutes, to completely emulsify the same mixture. First, intermittent shaking
was incorporated into the TPLS procedure, and appeared to be reasonably effective. However,
Sherman also states that hand shaking results in globules ranging from 50 to 100 /xm and that
to get smaller droplets one must apply more vigorous agitations to the liquids. It was speculated
that more complete extraction of chrysotile would be promoted if the surface contact were in-
creased between the water phase and benzene phase by the creation of smaller oil globules, but
ultimately it was found that more complete chrysotile separation was promoted by continual
agitation than by intermittent agitation. Continual agitation presumably increases the contact
between the water and oil phases by not allowing time for dispersed oil droplets to acquire the
stabilizing surfactant layer; the droplets form, coalesce, and reform, and thereby extract more
chrysotile from the water phase than when the first droplets formed are allowed time to
stabilize.
Demulsification — In the initial stages of development of TPLS, it was obvious that some
procedure must be devised to promote rapid and complete demulsification. Consequently, a
search was made for a simple, reproducible method of demulsification that would not interfere
with the selective extraction of chrysotile into the benzene phase.
In consulting the literature^), it was found that there are several ways to promote
demulsification. Some of the approaches to breaking an emulsion are: addition of excess
solvent, addition of a solvent miscible with each of the two phases, destruction of the emulsify-
ing agent, salting out, filtration, heating, freezing, electrolyzing, centrifuging, or changing the pH.
The approaches to demulsification considered most appropriate for application to the TPLS
system from the standpoint of simplicity and possible lack of interference were heating, salting
out, and changing the pH. Changing the pH was tried first, and it was found that the emulsion
19
-------
broke at pH 3. In view of the rather high solubility of chrysotile in acid, this approach was
abandoned.
Heating the emulsion was tried next, and it was found necessary to heat almost to the
boiling point of water to break the emulsion. Heating to such a high temperature caused the
benzene phase to boil off explosively evolving large quantities of noxious benzene vapor. How-
ever, demulsification by heating has the attraction of not changing the chemical environment;
thus successive extractions of chrysotile under similar conditions from the same water sample
are possible.
Salting out was the most successful method. G. M. Sutheim(^) states that if a strong
electrolyte salt is added, the emulsion coagulates as a result of interference with the adsorbed
emulsifying agent at the benzene-water interface. He found the coagulating effect of anions
to decrease as follows: citrate, sulfate, acetate, chloride; for cations: sodium, potassium,
ammonium, magnesium, calcium, barium. According to this listing, sodium citrate should have
the greatest coagulating effect. Therefore, sodium citrate was the first reagent to be tried in a
salting out experiment.
Ten ml of 25% sodium citrate was added to an emulsion created by shaking together 50 ml
of water containing a suspension of 0.6 mg of chrysotile and 9.2 mg of diatoms and 50 ml of
benzene to which 2.0 ml of 1.0% Aerosol OT dissolved in benzene was added. After the addi-
tion of sodium citrate the mixture was shaken once and allowed to stand. Within 5 minutes the
emulsion broke completely into two layers. The benzene layer was pipetted off and filtered,
and a preparation was made for electron microscopy. Both diatoms and chrysotile were found
to be present in the benzene fraction. The interference with selective separation was attributed
to the organic citrate ion. Consequently, sodium chloride was selected for subsequent trials.
Ten ml of 25% NaCl was added to an emulsion formed by intermittently shaking together:
(a) 50 ml of water containing a suspension of 0.2 mg of chrysotile and 5.0 mg
of diatoms
(b) 50 ml of benzene to which was added 2.0 ml of 1.0% Aerosol OT in benzene.
After adding the NaCl, the emulsion was shaken again and allowed to stand until it separated
completely into benzene and water layers. This required approximately 5 minutes. The
benzene layer was then pipetted off and filtered, and a preparation was made for electron
microscopy. Only asbestos was found in this preparation indicating that salting out with NaCl
did not interfere with TPLS selectivity.
When these developments indicated that TPLS could be speeded up, effort was directed
toward TPLS optimization. Attention was focused on the investigation of the identified
factors affecting degree of TPLS selectivity and recovery of chrysotile from the water.
The following listed factors were taken into consideration during the experimental studies
to establish the best TPLS procedure. Those factors subject to control were varied to search
for optimum conditions for TPLS selectivity and recovery.
20
-------
Factors Affecting TPLS Optimization
(1) Relative interfacial tensions in the solid-liquid-liquid system
(2) Hydrogen ion activity (pH)
(3) Concentration of the surfactant
(4) Oil/water volume ratio
(5) Alkyl chain length on the surfactant
(6) Zeta potential of the particle type
(7) Adsorption density (moles/cm^) of surfactant surface
(8) Chemistry of the chrysotile surface
(9) Effects of added ions
(10) Agitation conditions
(11) Emulsification.
Experimental Procedures in TPLS Optimization
The criteria used to judge TPLS optimization were degree of recovery and selectivity.
Experimental results were evaluated by transmission electron microscopy. In recovery studies,
aliquots from a standard chrysotile suspension were subjected to TPLS and recovery was
evaluated by comparing the quantity of extracted chrysotile to that in comparable aliquots not
subjected to TPLS. By using heavy concentrations of chrysotile (^4.0 mg/1), it was possible to
make rapid comparisons of chrysotile concentrations without resorting to fiber counting.
Although this analytical procedure was quite effective, no numbers expressing recoveries were
derived. However, when conditions were found that favored maximum recovery, atomic
absorption spectrometric analysis (AAS) for magnesium was run to obtain data expressing
percentage recovery. Again large quantities of chrysotile were employed to obtain the most
sensitive AAS measurement of recovery.
Specimens for TEM recovery analysis were prepared by depositing the chrysotile from an
aliquot of the standard suspension on a 0.45 //m, 25 mm, Millipore filter, and depositing the
chrysotile separated by TPLS from a similar aliquot on another Millipore filter. A piece was cut
from each filter and placed deposit side down on an electron microscope specimen grid bearing
a carbon support film. The Millipore filter material was dissolved by the Jaffe Wick Method
to prepare a specimen suited to viewing in the TEM where recovery was evaluated by comparing
asbestos concentrations at 5000X magnification.
A great many experiments were run before achieving recovery approaching 100 percent.
Consequently, the recovery results of most experiments could be quickly judged by electron
microscopy. Specimen preparations were saved from the various recovery experiments in order
to be able to make comparisons among results as TPLS development progressed.
After TPLS recovery was optimized and combined with light microscopic detection, re-
covery of those fibers detectable by light microscopy was assessed by comparing light
21
-------
microscopic counts on as-filtered aliquots to counts made on TPLS fractions from similar
aliquots. The results are discussed in the report section dealing with determination of limits
of detection.
Selectivity was assessed from the results of TPLS applied to synthetic mixtures of
diatomaceous earth and chrysotile as well as from results obtained when TPLS was run on
real water samples. Interferences giving both high and low results were assessed. TPLS
selectivity was not found to be as great a problem as recovery. Selectivity of the final TPLS
procedure for chrysotile in the presence of diatomaceous earth is excellent; only an occasional
diatom was found among matted pure chrysotile fibers in the electron microscope preparations.
TPLS selectivity when applied to real water samples is discussed in a later section entitled
"TPLS Spot Test Analysis of Real Water Samples". Although interferences giving both high and
low results were identified in these studies, they did not prevent chrysotile detection in these
samples.
Effects of Critical TPLS Factors on
Selectivity and Percent Recovery
The most critical factors in TPLS proved to be (1) mode of agitation, (2) type of
surfactant, (3) concentration of surfactant, (4) number of extractions from a single water
aliquot, (5) ratio of oil to water phase, (6) pH, and (7) the promotion of the reaction between
the chrysotile surface and the surfactant. Although the effects of some factors overshadowed
others, all were found to be important in the optimization of recovery and selectivity.
The problems of recovery and selectivity were dealt with separately during experimenta-
tion primarily because there was no satisfactory procedure to evaluate recovery when there
were large amounts of interfering extraneous particulate in the original water suspension. Con-
sequently, all recovery experiments were run on standard suspensions of pure chrysotile.
The general TPLS procedure in its stage of development at the beginning of these optimi-
zation studies consisted of adding anionic surfactant to the water phase, agitating the water
phase and oil phase together, salting out the emulsion that formed upon agitation, and after
phase separation, filtering the oil phase to concentrate the extracted chrysotile on a filter.
Progress toward optimization of TPLS recovery and selectivity is summarized in the fol-
lowing discussions of effects of the critical factors.
TPLS Recovery
Effects of Agitation —
The mode of agitation is perhaps the most critical of all the investigated factors. When all
other factors are optimized, if agitation is inadequate, low recoveries are inevitable.
First success was realized using ultrasonic treatment. Later, hand shaking in a separatory
funnel produced good results. After consulting the literature(6), it was decided to attempt to
obtain better control of agitation by substituting mechanical stirring for shaking by hand. Lai
and Fuerstenau reported «95 percent recovery of alumina from water suspension into isooctane
22
-------
using mechanical stirring and they were able to use it while varying other parameters to work
toward optimum recovery. The effects of stirring appeared to be quite reproducible.
After switching to stirring, varying recoveries were obtained; some were quite high but
others were low. This variability was attributed to other, possibly unidentified, parameters
when the manipulation of identified parameters did not consistently improve results.
Mechanical stirring was performed at speeds of 600, 1200 and 2200 RPM and recovery was
found to increase with stirring speed (Table 1). However, stirring speeds of 2200 RPM and
above were impractical. We settled on 1600 RPM, ran 18 experiments under identical condi-
tions, and analyzed recovery by atomic absorption analysis for magnesium. A mean recovery of
28% ± 14 was obtained. This degree of recovery was considered to be too low and to have un-
acceptable reproducibility. Consequently, a superior means of agitation was sought.
At this point, shaking by hand was tried again and better recoveries were found. A
mechanical shaker was used and consistently higher recoveries were obtained than by stirring.
Then differences were observed as factors other than agitation were varied. Ultimately, a
recovery of 74% ± 9.6 was obtained (Table 2) using shaking at 150 cycles per minute for
5 minutes. This was considered adequate for the purposes of the rapid analytical method.
Effect of pH -
Hydrogen ion activity is not extremely critical in the range from pH 3 to 7, and there is
no problem in maintaining the pH in this range during TPLS. However, to attain the highest
positive charge on the chrysotile surface to promote maximum reaction with the anionic
surfactant, the pH of the water phase was adjusted to 3.5 where the zeta potential has the
highest positive value. This is accomplished usually by addition of 0.1 N HC1. In all cases,
the pH of water samples has been found to be above 3.5.
Effects of Surfactant Compound and Surfactant Concentration —
Only one type of surfactant has been investigated to any extent, this is the anionic alkyl
sodium sulfosuccinate. However, two analogous compounds of this type have been investi-
gated. Because of initial success with these compounds, dioctyl sodium sulfosuccinate (Aerosol
OT and MO-70) and ditridecyl sodium sulfosuccinate (MT-70), no others were sought.
There was some evidence that MT-70 promoted slightly higher TPLS yields than Aerosol
OT, but later experiments indicated comparable recoveries with MT-70 and MO-70, the dioctyl
sodium sulfosuccinate supplied by Mona Industries Inc., Paterson, New Jersey 07524. Greater
TPLS selectivity was observed while using MO-70. Therefore, MO-70 has been recommended
in the finalized TPLS procedure.
Surfactant concentration is critical in TPLS, not only from the standpoint of recovery but
also in that too high concentration produces emulsion stability and phase inversion. Therefore,
concentration must be carefully adjusted to obtain acceptable TPLS recovery and to avoid
emulsification problems.
While employing MT-70, optimized recoveries were observed to occur at concentrations
ranging from 1 x 10-3 M to 2.5 x 10~3 M. No phase inversion was detected at 1 x 10-3 M, but
it was occasionally seen at the 2.5 x 10~3 M concentration. Recoveries were determined by
23
-------
TABLE 1. EFFECT OF AGITATION BY STIRRING
Run
3 1370-93 A
31370-93B
31370-85
32608-1
Stirring Speed, rpm
600
600
1600
2200
Percent Recovery Based
on Mg Analysis
7
7
25
39
TABLE 2. PERCENT RECOVERY WITH 2.5 X lO'3 M MT-70
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
Mean = 74%
% Recovery After Five
Consecutive Extractions
From Same Sample
77
59
81
64
67
54
77
72
86
81
72
69
99
Recovery
pH After Five
Consecutive
Extractions
4.5
4.7
4.7
4.3
4.4
4.5
4.5
4.3
4.4
4.5
4.6
4.3
4.4
Standard Deviation = ±9.6
24
-------
AAS analysis for magnesium at these two surfactant concentrations to determine whether there
was a significant difference in recovery. Twenty-nine determinations were made at 1 x 10"3 M
MT-70 concentration and the recovery was measured to be 63% ±11. At 2.5 x 10~3 M con-
centration, it was measured to be 74% ± 9.6 (Tables 2 and 3).
Because no further problems with phase inversion were encountered, the higher surfactant
concentration is recommended in the description of the rapid analytical method.
Number of TPLS Extractions From a Single Water Aliquot -
In initial TPLS studies, only one extraction was made from the water aliquot. Upon mak-
ing several extractions and separately filtering each extracted fraction, it was discovered that the
bulk of the chrysotile was not necessarily extracted in the first pass. The TPLS procedure was
modified to include three extractions. As further studies were made of TPLS yields, it was
again modified to include five passes.
Five extractions from the same water aliquot are recommended in the final TPLS pro-
cedure. Under these conditions, 75 percent of the chrysotile is extracted.
Ratio of Oil Phase to Water Phase —
The ratio of the volume of oil phase to the volume of water phase is important with
respect to phase inversion. The oil volume was varied from 25 percent of the total liquid
volume to 50 percent (1:1 water-oil). We finally settled on 100 ml of water phase and 50 ml
of isooctane. Under these conditions satisfactory interfacial contact occurs and phase inversion
is, in most cases, avoided.
Promotion of the Reaction Between Chrysotile and Surfactant —
It was found that TPLS recovery was improved by heating the water phase after the addi-
tion of the surfactant. Presumably, heating promotes the reaction between the chrysotile sur-
face and surfactant. Perhaps one of the mechanisms by which this occurs is the desorption of
interfering species from the chrysotile surface.
Heating at 60°C is recommended in the routine TPLS method.
TPLS Selectivity
Preliminary experiments to investigate TPLS selectivity were run on synthetic mixtures of
chrysotile and diatomaceous earth. Diatomaceous earth was chosen because of the prevalence
of diatoms in water samples. The recommended method is quite selective for chrysotile in the
presence of diatoms.
It is impossible to predict the interferences that might be encountered when applying
TPLS to water samples; however, TPLS has been run on water samples from 26 different
sources. Some interferences were observed, but they were not extremely serious. All were re-
lated to lack of TPLS selectivity. Interferences that could give both high and low results were
observed.
High results were obtained on some samples from which organic filaments were extracted
by TPLS. Nevertheless, these filaments can be identified as not asbestos with a little experience.
25
-------
TABLE 3. PERCENT RECOVERY WITH 1.0 X KT3 M MT-70
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Mean = 63%
Standard Deviation
% Recovery After Five
Consecutive Extractions
From Same Sample
71
68
61
50
75
75
82
89
82
46
55
82
65
68
71
48
48
63
55
55
55
63
68
51
40
55
55
63
63
= ±11
pH After Five
Consecutive
Extractions
4.8
4.7
4.8
4.2
4.3
4.6
4.7
4.8
4.8
4.4
4.6
4.7
4.8
4.5
4.6
4.6
4.7
4.5
4.6
4.6
4.6
4.7
4.7
4.7
4.8
4.8
4.7
4.8
4.8
26
-------
These filamentous structures are approximately 0.1 urn in diameter and have varying lengths.
One of their identifying characteristics is that along their lengths they have swellings 0.2 i*m in
diameter and approximately 0.5 fim long. These swellings can be seen in the light microscope
with vertical illumination at 500X magnification after the filtered TPLS deposit has been carbon
coated. No other interferences giving high results were discovered.
Two types of interferences produce low results. One is an extracted platy mineral that
could obscure chrysotile fibers. These plates yield a diffraction pattern the same as that of
talc. In no case was a sufficient quantity extracted to prevent the detection of chrysotile.
The other type of interference that will give low results is related to the nature of the
chrysotile surface. If magnesium has been leached from the fiber surface, the zeta potential
becomes negative, hence, it will not react with the surfactant. Also, coatings may form on the
fiber and thus prevent access of the surfactant. Observed in the TEM, these coatings have low
electron densities indicating that they are probably organic.
In the cases in which interference from organic material is a problem, heating after
surfactant addition improves recoveries of coated fibers, but removal of organic fractions by a
chemical treatment possibly would result in further improvement. Because of the time re-
quired, low-temperature ashing was not considered during the development of the rapid analyti-
cal method. Neither were "wet ashing" techniques explored. It is believed that any future work
should include the investigation of "wet ashing".
Handling Large Water Sample Volumes by TPLS
It is desirable to apply TPLS to the particulate in relatively large volumes of water in order
to realize the most sensitive detection possible. One of the approaches taken was to increase the
volume of water sample handled by TPLS. This resulted in a corresponding increase in the
volume of isooctane and an increase in the time required to filter the separated chrysotile.
Thus, it appears that scaling up TPLS is not a satisfactory solution.
A scaled-up procedure, was used to run 500-ml water aliquots. The filtration time required
to filter the 500 ml of isooctane from each of 5 extractions through a 25-mm, 0.2-jum
Nuclepore filter using the standard Millipore reservoir was approximately 2-1/2 hours. Two
anionic surfactants (MO-70 and MT-70) were used in these experiments. Scaling up to handle
the 500-ml aliquots involved the use of a 2000-ml separatory funnel equipped with a Teflon
stopcock to avoid stopcock grease. Pure chrysotile was suspended in 500 ml of filtered
deionized water; 250 ml of 0.01 M surfactant was added and the suspension was sonicated
for 15 minutes. The suspension was heated to 150°F while stirring and subsequently cooled
to room temperature. Then, 250 ml of water and 500 ml of isooctane were added. TPLS
was performed and the separated fractions were deposited on Nuclepore filters. Those separa-
tions using MO-70 (dioctyl sodium sulfosuccinate) were relatively free of NaCl crystals and
exhibited good recovery. MT-70 (ditridecyl sodium sulfosuccinate) separations contained a
great many NaCl crystals; however, chrysotile recovery was also quite good.
Even if smaller oil-to-water ratios were used, the filtration time would still pose a problem.
Filtration and resuspension was found to be a better way of handling large water volumes.
27
-------
Filtration and Resuspension —
Work has been done to determine the feasibility of filtering relatively large volumes of
water, removing the filtered deposit, and resuspending the particulate in smaller water volumes
to facilitate the application of two-phase liquid separation. Six different types of filters were
used, and removal of the deposits was tried using ultrasonic treatment and shaking.
Standard chrysotile suspensions (200- to 500-ml aliquots) were filtered using MF-Millipore,
Duralon, Polyvic, Celotate, Nuclepore, Gelman Metricel (DM 45), and Mitex filters. Fluoropore
filters were also tried, but water would not pass through them without the addition of alcohol
or other liquid to promote wetting. Ultrasonic treatment disintegrated the MF-Millipore,
Duralon, Polyvic, and Celotate filters; however, the best recoveries were obtained by ultrasonic
treatment of Nuclepore, Mitex, and Metricel DM 450 filter deposits. Recoveries from Metricel
DM 450 and Nuclepore filters were best of all and were judged by microscopic examination to
approach 100 percent. Some Teflon fibers were dislodged from Mitex filters by ultrasonic
treatment, but, despite this problem, the redeposited chrysotile was quite visible.
Shaking, although not as effective in particle removal, did not break any of the filter types.
In all cases, some of the chrysotile was removed, and was found to contain no extraneous
debris when redeposited. A means of agitation employing less energy than the particular ultra-
sonic treatment used and greater energy than shaking would likely make it feasible to remove
the major quantity of deposited particulate from any of the filter types.
TPLS did not separate the chrysotile from filter fragments; both were transferred to the
isooctane. Chrysotile removed ultrasonically from Mitex, Metricel DM 450, and Nuclepore
filters was separated from the water phase by TPLS. Although Nuclepore is one of the two
best filters from the standpoint of particle removal, it is fragile and difficult to handle in a field
operation. The procedure found most effective for the removal of the filtered deposit in
preparation for TPLS was to add 50 ml of 0.005 M MT-70 to the filter, deposit side down,
in a 250-ml flask and to sonicate for 15 minutes; 25 ml of deionized water was then used to
rinse the filter as it was removed from the beaker, bringing the volume to 75 ml. At this
point, the resulting suspension was heated to 150°F while stirring to promote the reaction
between the surfactant and the chrysotile surface. Next, the volume was increased to 100 ml
by adding deionized water to make the MT-70 concentration 2.5 x 10-3 M; then 50 ml of
isooctane was added and the TPLS procedure was carried out. Five extractions were made and
the separated fractions were combined on one 0.2-Mm-pore-size Nuclepore filter. Evaluations
of recovery were made on carbon-coated deposits with the light microscope, or on Jaffe
preparations in the electron microscope.
Twenty-four experiments were made in which filter deposits were removed ultrasonically
and subjected to TPLS. Good recoveries were obtained.
Metricel DM 450 filters were the best found. They are easily handled in the field,
particles are readily removed from the filter surface by ultrasonification, and they do not
break up in the ultrasonic bath. The DM 450 filter has a 0.45-Mm pore size and it is composed
of a PVC copolymer. The diameter of the DM 450 filters used was 47 mm. Larger diameters,
of course, could be used to filter larger quantities of water or water containing heavy concen-
trations of suspended particulate.
28
-------
If large water samples that are heavily loaded with particulate must be handled, several
47-mm-diameter filters can be used. The collected particulate can be removed ultrasonically in
a 100-ml volume of water containing the anionic surfactant.
Filtration and resuspension is recommended as an alternate procedure in the TPLS - light
microscopic analytical method. When greater sensitivity is desired than is possible by extracting
a 75-ml aliquot, then it is suggested that filtration and resuspension be employed.
Another advantage of filtration and resuspension is that samples can be collected and
filtered soon afterwards. This would minimize sample storage problems.
TPLS Recovery as a Function of Particle
Size and Aspect Ratio
Experiments were conducted to determine whether TPLS favors the extraction of certain
chrysotile fiber sizes and aspect ratios. Several groups of experiments have been run investi-
gating TPLS particle size recovery; however, the set of experiments that used the finalized
TPLS conditions are the most pertinent. This last set of experiments also was designed to
determine the effects on particle size recovery as a function of zeta potential.
Four samples of chrysotile having different zeta potentials were supplied by Mr. George
Reimschussel of Johns Mansville Research and Engineering Center. Mr. Reimschussel measured
their electrophoretic mobilities as listed below:
Sample Electrophoretic Mobility at pH 10
96-1 +3.9 MV
96-2 +0.8 MV
96-3 -1.2 MV
96-4 -2.6 MV
No electrophoretic mobilities were measured at lower pH conditions. Nevertheless, they must
all have positive zeta potentials at pH 3.5-4. Good TPLS recoveries were obtained in this
pH range. The existence of zeta potential variations among chrysotile types makes it important
to control the pH during TPLS.
Standard chrysotile suspensions were prepared from samples 96-1, 96-2, 96-3 and 96-4,
and equal (50 ml) aliquots were taken from the same suspension. One aliquot was merely
filtered onto a 25-mm, 0.2-Mm-pore-size Nuclepore filter; three other aliquots from the same
suspension were subjected to TPLS and the extracted chrysotile was also deposited on the
same type Nuclepore filters. Specimens for TEM were prepared from the filtered deposits by
carbon coating followed by dissolution of the filter material by the Jaffe Wick Method.
The following TPLS procedure was used.
1. To a 75-ml water aliquot adjusted to pH 3.5 with 0.1 N HC1, add 25 ml of
0.01 M MO-70, heat to 140°F while stirring and cool to room temperature.
29
-------
2. Combine resulting solution with 50 ml of isooctane in a 250-ml separatory
funnel.
3. Shake the isooctane and water phase together in the separatory funnel at
150 cycles per minute for 5 minutes.
4. Add 10 ml of 10 percent NaCl and shake gently to break the emulsion.
5. When the phases have separated drain off the water phase completely allow-
ing a small amount of isooctane to drain into the water.
6. Pour the isooctane into a beaker from the separatory funnel avoiding water
droplets that may remain in the bottom or on the sides of the separatory
funnel.
7. Filter the isooctane onto a 25-mm, 0.2-/um-pore-size Nuclepore filter using
the standard filter reservoir.
8. Repeat the extraction procedure four times filtering the isooctane each time
through the same filter.
9. Dry filter and coat the deposit side with carbon by vacuum evaporating 1 cm
of carbon rod 1 mm in diameter at a distance of 15 cm.
10. Prepare TEM specimens by the Jaffe Wick Method.
Specimens were prepared for electron microscopy by carbon coating the filtered chrysotile
deposits and dissolving away the Nuclepore filter by the Jaffe technique. TEM fiber counts and
measurements of fiber lengths and widths were made of at least 100 fibers in each preparation.
A computer was used to assign lengths, widths, and aspect ratios to size classes and to plot
each versus cumulative number percent and cumulative volume percent.
Upon examination of the plots, it was evident that the cumulative volume percent plots
were too greatly affected by occasional fibers having unusually large diameters. Also, the
widths of fibers extracted by TPLS did not differ significantly from those not subjected to
TPLS. Consequently, it was decided to include in this report only two sets of plots for each
sample, i.e., length versus cumulative number percent and aspect ratio versus cumulative num-
ber percent. Each plot presents the results of four determinations; one of the as-deposited
chrysotile, and three of TPLS extractions from comparable aliquots. The plots are designated
as Figures 8 through 15.
The results show that TPLS favors the extraction of the longer fibers with larger aspect
ratios. In this respect, these latest results are consistent with those previously obtained under
different TPLS conditions. This is acceptable in the sense that it is more desirable that the
more visible larger fibers are selectively extracted when light microscopic detection is used.
In addition, investigators of health effects have reported that the longer fibers are more
carcinogenic than the shorter ones.
30
-------
33.33
99.9
99
«- 98
£ 95
o
fc 90
o.
80
fc" 70
-£ 60
| 50
Z 40
„> 30
(1)
-> 20
1 '°
0 2
1
0.5
°0'.f
0.05
0.0.Q
—
—
—
• AA
• A A
• At
• AA
- *A AA
•AA
- A A
•A
•A
—
—
-
-
!i i i i l l 1 1 i i l i i i 1 1 1
2 05 1.0 2.0 5.0 10 2C
Length,
Figure 8. Distribution of chrysotile fiber
lengths in sample 96-1 before (•) and
after (A) TPLS.
33.33
99.9
99
-H- 98
£ 95
5 90
o.
80
of 70
-§ 60
§ 50
Z 40
a, 30
~ 20
0
1 '°
3 5
° 2
0.5
0.2
0§>5
o.ob
—
—
—
—
• AA A
• AA
• AA
• AA
— • At
• A A
• AA
•A AA
• AA
—
1 1 1 1 1 1 ll 1 1 | 1 1 1 1 1 1
2 0.5 1.0 2.0 5.0 10 2C
Length, fj.m
Figure 10. Distribution of chrysotile fiber
lengths in sample 96-2 before (•) and
after (A) TPLS.
c
0>
o
Iw
03
Q.
-0
E
3
3
3
33.33
99.9
99
98
95
90
80
70
60
50
40
30
20
10
5
2
1
0.5
0.2
0.05
0.01
—
—
—
• m.
• AAA
• AAA
• A A
— «AA
•AAA
•tA
•At
•A*
—
-
—
-
_
^~
-
1 1 1 1 1 1 ll 1 1 1 I 1 M ll
5.0 10 20 50
Length/Width
100 200
Figure 9. Distribution of chrysotile fiber
aspect ratios in sample 96-1 before (•)
and after (A) TPLS.
"E
a.
_
<5
JO
E
3
Z
0)
~
0
3
E
3
O
33.33
99.9
99
98
95
90
80
70
60
50
40
30
20
10
5
2
0.5
0.0 12
—
—
• A
• AA
• A
— A A^A
A\ AA
• AAA
• AA
• A
• AA
—
—
-
l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ll
.0 5.0 10 20 50 100 20
Length/ Width
Figure 11. Distribution of chrysotile fiber
aspect ratios in sample 96-2 before (•)
and after (A) TPLS.
31
-------
99.9
99
98
S 95
fc 90
0.
80
a3~ 70
-£ 60
| 50
Z 40
20
1 10
E 5
13 J
0 2
1
0.5
0.2
O.I
0.05
001
—
—
• —
• A A
• A AA
• AM
• AM
— • AAA
• AAA
• A A A
• AAA
•MA
-
—
-
-
—
-
l i i l 1 l 1 1 l 1 1 1 1 l 1 i 1
c
O)
o
k 8.
1__
co
.a
E
co
'•g
I
3
O
0.2 05 1.0 2.0 5.0 10 20
Length, ^.m
Figure 1 2. Distribution of chrysotile fiber
lengths in sample 96-3 before (•) and
after (A) TPLS.
99.99
99.9
99
*- 98
c
v 95
o
S3 90
Q.
80
of 70
-g 60
i 50
Z 40
a, 30
~ 20
"5 10
1 5
o
2
0.5
0.2
0.05
—
—
—
—
• A A
• MA
• M
• M
— • M
• M
- tMft.
• M\
•Wk
—
~
-
L
i i i i 1 1 1 1 i i i i i i 1 1 1
"c
•^
-5
E
(_j
b.2 0.5 1.0 2.0 5.0 10 20
Length, fj.m
Figure 14. Distribution of chrysotile fiber
lengths in sample 96-4 before (•) and
after (A) TPLS.
33.33
99.9
99
98
95
90
80
70
60
50
40
30
20
10
5
2
1
0.5
0.2
0.05
0.0 12
—
—
—
—
• AA,
• A M
• AA A
• AA A
- • AA A
• M A
• A A
• A AA
• 4*
—
-
—
-
-
-
i i i i 1 1 1 1 i i i i i i 1 1 !
0 5.0 10 20 50 100 20
Length/Width
Figure 13. Distribution of chrysotile fiber
aspect ratios in sample 96-3 before (•)
and after (A) TPLS.
33.33
99.9
99
98
95
90
80
70
60
50
40
30
20
10
5
2
0.5
a?
0.05
nni
—
—
—
•4A
• MA
•tA
A JM
^K ^^^V
A 4k
•i
• M
—
-
_
1 1 1 1 1 1 1 1 1 1 1 1 l 1 1 1 1
200
Length/Width
Figure 15. Distribution of chrysotile fiber
aspect ratios in sample 96-4 before (•)
and after (A) TPLS.
32
-------
PHASE II: INVESTIGATION OF DETECTION METHODS
Change of Flow Rate as a-Function of Chrysotile Concentration
A survey of detection methods was made in an effort to provide rapid semiquantitative
analysis of chrysotile separated and concentrated by TPLS. The methods investigated were
change in flow rate of a liquid through a filter as a function of the deposited mass, spot-test
procedures, X-ray diffraction, and light microscopy. The objective was to develop a simple
routine detection procedure with sufficient sensitivity to detect asbestos concentrations below
the significant level in order to eliminate insignificant samples from electron microscopic
analysis.
No attempt was made to confine detection to fiber counting. Rather, a search was made
for a bulk detector that could be used without the need for a high degree of skill to screen large
numbers of water samples rapidly. This was the reason for the investigation of a method so
simple as that which attempted to correlate changes in flow rate through a filtered chrysotile
deposit with the mass of chrysotile extracted by TPLS.
The flow rate of a liquid through a filtered deposit becomes slower as the mass of
deposited particulate becomes greater. It was believed that this phenomenon could provide
the basis for a rapid detection method.
Experiments were run to determine the flow rates through measured quantities of
chrysotile deposited on 25-mm-diameter Nuclepore (0.1 jum pore size) filters and dried.
Initial experiments were run using the standard Millipore 25-mm filtering apparatus.
The measurements were made by introducing 15 ml of isooctane into the reservoir,
applying the vacuum, and recording the time required for the isooctane to pass through the
filter. The pressure drop was noted for each point.
Quantities of chrysotile ranging from 5.0 to 40.0 jug were used. Both pressure drop across
the filter and flow rate were measured. The plot of pressure drop and flow rate versus
chrysotile concentration is shown in Figure 16. Pressure drop was not a very sensitive indi-
cation of chrysotile concentration. Flow rate was a more sensitive measurement, but not
sufficiently so under these experimental conditions.
Further experimentation to obtain more sensitive measurements by flow rate resulted in
detection of quantities as low as 0.5 /xg- This was accomplished by making a special filter
reservoir with a 3-mm hole in order to confine the filtered chrysotile to a smaller filter area,
thus increasing the change in flow rate for a given amount of chrysotile.
Attempts to extend detection below 0.5 /ug were not successful; furthermore, when analysis
of the TPLS fraction from a real water sample was tried, a relatively small amount of inter-
fering platy mineral particles gave too high results. The 0.5 Mg detection limit was not con-
sidered to be adequate. No further effort was made to use this approach.
33
-------
no
100
90
tn 80
§ 70
o>
tn
~ 60
4, and (4) detection of reacted manganese by the periodate-tetrabase spot test.
34
-------
The detailed spot-test procedure is as follows:
1. Chrysotile as extracted by TPLS is deposited on a 25-mm, 0.45-jum-pore-size
Millipore filter in an ^S-mm-diameter area and dried. This is done by using
a special reservoir with a 3-mm-diameter opening.
2. The special reservoir is then removed and the filter bearing the deposited
chrysotile is covered with a Millipore filter, and the standard 25-mm reservoir
is used for Step 3.
3. Fifteen ml of 25% MnSC^-P^O at room temperature is poured into the filter
reservoir and aspirated through the deposit over a period of 1 minute.
4. Next, the filtered deposit is washed with 15-ml quantities of room temperature
deionized water 5 times aspirating each washing completely through the filter.
5. The filter reservoir is removed while applying suction to the filter to hold it
to the frit and then the covering Millipore filter is removed with tweezers.
6. The filter bearing the chrysotile deposit is removed and dried under a heat
lamp at 145°F (63°C).
7. The filter is attached by the edge to a piece of white blotter paper deposit
side up and sprayed with the spot testing working solution and placed under
the heat lamp at 145°F until development is maximized («*2 minutes).
8. The depth of blue color that develops is compared to photographs of a series
of standards made using known quantities of chrysotile.
The working spot test solution is made up by combining equal volumes of the Solutions A
and B:
Solution A: 1.0% solution of tetrabase in 95% ethyl alcohol.
Solution B: A saturated solution of potassium periodate in 2N acetic acid.
Comparison of the depth of color of the spot test result is possible over a chrysotile con-
centration range from 0.1 ^g to «0.8 ng. At 0.8 /ug and above the color is so intense that little
change with increase in concentration can be detected. Nevertheless, when used primarily to
select samples for TEM analysis the lack of good quantitative indications above 0.8 fig poses no
problem.
The spot test procedure has the advantage that it can be carried out quickly with a mini-
mum of expensive apparatus and training; but the preparation, unlike that used for light
microscopic detection, cannot be readily used for electron microscopy when confirmation of
results is desired.
X-Ray Diffraction
The techniques of powder X-ray diffraction were explored in an effort to establish a
detection procedure. The primary difficulties were quantification and sensitivity of detection.
The X-ray powder camera was used in most of the experiments; diffractometer methods
were not considered to have sufficient sensitivity. The major problem was with specimen
35
-------
preparation when dealing with lower quantities of chrysotile (<30 ng). No way was found to
prepare a specimen so that it would be entirely irradiated by the X-ray beam.
Light microscopy was found to be far superior to X-ray diffraction. Fibers detectable by
light microscopy have always been found in water samples containing chrysotile even when the
mean fiber length is below 1.0 /um.
Light Microscopy
It was found that light microscopy employing vertical illumination as obtained with metal-
lographic optics is an extremely sensitive means of detecting fiber shapes in a carbon-coated
TPLS deposit on a Nuclepore filter. When interferences from particulate other than chrysotile
are virtually eliminated by TPLS, the separated chrysotile fibers are quite visible. The carbon
coating increases the detectability of the chrysotile by two mechanisms; it enhances image
contrast by increasing the reflectivity of the Nuclepore substrate and also increases the apparent
fiber diameter to the point that some of the fibers with diameters too small to be seen by light
microscopy become detectable if their lengths are greater than «1.0 Mm- The fibers are seen as
dark structures against a bright background because the microscope illumination is along the
axis of the objective lens. The Nuclepore filter surface is perpendicular to the objective axis
reflecting light directly back into the objective and the fiber surfaces are not perpendicular
reflecting light away from the objective.
A quick examination with the light microscope is all that is required to determine whether
a filtered chrysotile deposit is greater than 100 nanograms. Analysis of intermediate concentra-
tions is performed with an eyepiece reticule having 10 concentric circles to minimize fiber
counting time as fiber concentration varies (Figure 17). When concentrations are relatively
high, counts are made within circles near the center, and lower concentration counts are made
within circles of greater diameter — sometimes across the diameter of the filter deposit.
Light microscopic detection of chrysotile was placed on a quantitative basis by generating
a standard curve relating chrysotile mass per filter to counts made at 500X magnification while
viewing a carbon-coated Nuclepore filter with vertical illumination. Counts are made while
scanning a 250-jum-wide path across the diameter of the effective filter area. The eyepiece
reticule having 10 concentric circles spaced 12.5 p.m apart is employed (Figure 17), and one
count is registered for a fiber segment equal to the spacing between concentric circles. Compen-
sation for variations in fiber width are made by estimating the number of minimum detectable
fiber widths in larger fibers or bundles. Because the outermost circle delineates a 250-/um-
diameter field at 500X magnification, it is used to determine the path width over which counts
were made. Counts are made only on fibers falling within that field as the path is traversed.
Standard chrysotile suspensions were filtered onto Nuclepore filters using the standard
25-mm Millipore filtering apparatus. Concentrations ranging from 1.0 nanogram to 10,000
nanograms were deposited on Nuclepore filters, which were subsequently taped flat on micro-
scope slides and coated with carbon. At least six filtered deposits were prepared at each of
seven concentrations from two sources of pure chrysotile.
Two persons have made counts on several of these preparations and the reproducibility
between the two sets of results appears to be quite acceptable as demonstrated by the agreement
36
-------
Figure 17. The eyepiece reticule used in fiber counting.
-------
of data. Counting a preparation requires approximately 15 minutes when dealing with
100-nanogram concentrations where the total number of counts is «1500. When higher
concentrations are encountered, various sized fields may be selected from the set of concentric
circles in the eyepiece reticule to count smaller areas and the data may be normalized by
multiplying the results by the appropriate area factor.
The standard curve that was thus generated is presented in Figure 18. Counts versus
chrysotile mass per filter are shown on a log-log plot. The sensitivity of light microscopic
detection appears quite adequate for the purposes of the rapid analytical method. Also, the
time per analysis is sufficiently short, especially if the method is to indicate only if a sample
should be analyzed more precisely by electron microscopy. Then an upper count limit may be
set according to allowable chrysotile concentration to minimize the required counting time.
Determination of Percent TPLS Recovery of Particle
Sizes Detectable by Light Microscopy —
The same preparations used to determine the recovery of chrysotile by TPLS by TEM
methods were also used to determine the recovery of particle sizes detectable by light
microscopy. The counting procedure described in the preceding section was applied to com-
parable aliquots before and after TPLS.
Results are presented in Table 4. The percent recovery as determined from three separate
runs from each of the four samples are quite comparable to the percent recovery as determined
by atomic absorption speclrometry. When all results were combined, the mean and standard
deviation were calculated to be 75% ± 15. Consequently, when applying light microscopic
detection to TPLS fractions, it is suggested that a factor of 2 be applied to the counts made
with the light microscope to compensate for fiber loss during TPLS.
TABLE 4. PERCENT TPLS RECOVERY OF PARTICLES DETECTABLE
BY LIGHT MICROSCOPY
Sample Run 1 Run 2 Run 3 Mean ± Std. Dev.
96-1
96-2
96-3
96-4
92
76
97
74
83
92
57
64
65
72
49
81
80 ± 13
80 ± 10
67 ± 25
73 ± 9
Limits of Light Microscopic Detection —
Limits of detection have been determined at 80, 90, 95, 98 and 99 percent confidence
levels. The data used in the calculations were obtained by the light microscopic counting
technique. Counts were made on 22 different concentrations ranging from 0.8 to 1000 nano-
grams. An average of approximately six preparations at each concentration were counted.
These data were analyzed by a computer program to determine the limits of detection of
chrysotile per Nuclepore filter deposit. Table 5 presents the detection limits at 100% TPLS re-
covery and also at 75% ± 15 TPLS recovery. The latter recovery was determined by a com-
parison of light microscope counts made on as-deposited aliquots versus similar aliquots from
38
-------
the same standard suspensions subjected to TPLS. At 60% recovery (75% - 15) the limit of
detection at the 99% confidence level was computed to be 1 nanogram per filter deposit.
TABLE 5. DETECTION LIMIT DETERMINATIONS
Percent Detection Limit (/ug) Detection Limit (/xg) at
Confidence Level at 100% Recovery 75% ± IS TPLS Recovery
75%
80
90
95
98
99
2
2
3
5
5
.2 x
.7 x
.4 x
.3 x
.2 x
ID'4
JO'4
ID'4
ID'4
ID'4
2
3
4
5
6
.9 x
.6 x
.5 x
.7 x
.9 x
ID'4
lO-4
ID'4
lO-4
10-4
2.4
3.0
3.8
4.8
5.8
+15
x
x
X
X
X
lO-4
ID'4
lO-4
ID'4
10-4
-1
3.7 x
4.5 x
7.5 x
2.6 x
8.7 x
5
JO'4
ID'4
ID'4
ID'4
ID"4
The data points used in the calculation of detection limits are tabulated in Table 6. These
data were also used to generate the standard curve.
PHASE III: APPLICATION OF ANALYTICAL PROCEDURES
TPLS — Light Microscopic Analysis of Real Water Samples
Seventeen water samples were obtained from Dr. Ian Stewart of W. C. McCrone Associates
and from Dr. Eric Chatfield of Ontario Research Foundation. All were subjected to TPLS and
the extracted fractions were analyzed by the light microscopic counting technique (see Fig-
ure 18), and some of them were analyzed by the spot test procedure. Specimens from these
samples were also examined by TEM before and after TPLS to determine whether chrysotile
was present and the nature of particulate extracted by TPLS along with chrysotile.
In the as-received condition, many of the samples contained large quantities of extraneous
particulate from which it was necessary to separate the chrysotile. First, TEM comparisons
were made between two-phase liquid separations from 50-ml aliquots and filtered deposits
prepared from 10-ml aliquots of the as-received samples to determine whether there was agree-
ment between the presence and concentration of chrysotile in the as-received sample and the
corresponding TPLS fraction. The same preparations were also examined by light microscopy
to evaluate the feasibility of using the light microscope for rapid screening in a detection
procedure.
Most of the samples had a pH of 7 or higher, which is too high to obtain optimum TPLS
results. Consequently, the pH of the 75-ml aliquot was adjusted to pH 3.5 with 0.1 N HC1
where the positive zeta potential of chrysotile is maximized. Then, 25 ml of 0.01 M MO-70
was added, the sample heated to 140°C, cooled to room temperature, and 50 ml of isooctane
added. TPLS was then carried out and the extracted chrysotile was deposited over a 16-mm-
diameter filter area.
39
-------
TABLE 6. LIGHT MICROSCOPIC DETECTION OF CHRYSOTILE AT TWENTY CONCENTRATIONS
Concentration
(Mg) Avg
0.00079
0.001
0.0018
0.005
0.0079
0.009
0.01
0.018
0.02
0.025
0.036
0.045
0.079
0.1
0.18
0.2
0.36
0.4
0.72
0.79
0.8
1.0
5.67
8.8
31.7
81.8
93.3
157.3
119.2
260.3
272.4
235.5
350.5
527.8
959.7
1356.3
1679.8
2188
6908.3
5953.2
12910.4
12426.7
6
10
40
78
62
137
139
228
338
225
410
583
972
1,435
1,708
1,951
7,298
6,232
13,202
16,705
10,496
12,300
6
6
29
72
91
128
139
272
294
244
368
486
1,076
1,064
1,924
2,425
8,528
5,904
12,218
10,707
7,954
13,776
5
7
6
93
118
186
95
386
286
264
310
522
831
1,433
1,852
5,166
5,904
11,726
9,868
13,366
12,710
8
29
88
164
77
210
251
231
296
500
1,316
1,780
4,592
5,986
13,858
11,726
16,400
5
38
86
137
116
238
225
259
392
590
1,600
1,146
6,658
5,740
1 1 ,070
12,628
11,480
Counts
8 13 11 9 11
48
86 84 70 69 102 79 75
192
149
228
254 259
190
327
486
1,290
1,669
7,298 5,986 7,462 6,806 8,118 8,364 6,624
15,744 12,628 12,464 13,284
10,414 9,184 13,694 9,184 13,038 11,168.4
12,054 13,448 16,646 15,990 17,056 15,006 14,260.5
-------
1,000,000 p
100,000 —
10,000 —
§ 1,000
o
0.0001 QOOI 0.01 O.I
Pure Chrysotile,
1.0
10
Figure 18. Standard curve that relates light microscope
counts to mass (f/g) of chrysotile per filter.
Results are presented in Table 7. In general, the results obtained were as follows.
1. In no case was chrysotile found in the as-received preparation and not found in
the TPLS fraction by TEM and light microscopy.
2. In five cases, chrysotile was found in the TPLS fraction and not in the as-
received preparation; this was attributed to masking by extraneous particulate.
3. Chrysotile was detected in only two of the as-received preparations by light
microscopy. This was also the result of masking by other particulate.
4. Filamentous structures, which appear to be organic when seen in the TEM, are
present in four of the TPLS fractions. To the inexperienced observer using
light microscopy, these filaments might be identified as chrysotile to give false
positive results.
5. In two cases, some interference was encountered from other particulate being
transferred along with chrysotile to the isooctane; however, chrysotile was
detected more readily by TEM and LM in TPLS preparations than in the filter
deposits of the corresponding as-received samples.
6. Detection of chrysotile in TPLS fractions from real water samples was found
feasible using the light microscope.
7. Light scattering, as seen with the unaided eye, from carbon-coated deposits of
the TPLS fractions were observed to correlate fairly well with the presence of
chrysotile. False positive results are more likely than false negative results when
relying on light scattering.
41
-------
TABLE 7. RESULTS OF TPLS APPLIED TO REAL WATER SAMPLES
Water Sample Number
Observation in TEM of
TPLS Fraction
TEM Observations of
As-Received Sample
Debris Chrysotile
LM Observations of
TPLS Fraction
Remarks
to
C3 0514 1530 U3WGO 11/17 I
(filter sample WG3)
C3 0514 1530 D3 WGO 11/17 II
(filter sample WG3)
C9 0305 1215 U4 WGO 11/17 III
(filter sample WG3)
C9 0305 1350 U5 WGO 11/22 I
(filter sample WG1)
C9 0305 1340 D5 WGO 11/17 IV
(filter sample WG3)
P4 0212 0920 D5 WGO 11 /15 II
(filter sample WG2)
Long filamentous organic structures Heavy
present; very little chrysotile seen
Many bacteria-like structures plus
long organic filaments. Medium
amount of chrysotile
No chrysotile, a few organic fila-
ments, and very little extraneous
particulate
Many clumps of chrysotile, a fairly
clean separation containing a few
bacteria
A clean separation containing many
short chrysotile fibers
Very few chrysotile fibers; some
extraneous organic particulate
Little
Medium None
Little
None
Medium None
Little
None
Very
heavy
None
Long filaments seen,
but no chrysotile
Medium number of
fiber-shared struc-
tures seen
Very few fiber
shapes
Many individual
fibers and several
clumps of fibers
detected
Many short fibers
seen
Very few fibers
seen
With some experience, these
filaments would not be con-
strued as chrysotile when
seen with the light
microscope
Both chrysotile fibers and
organic filaments are visible
in the light microscope
Not a significant population
of fiber shapes
Relatively high fiber popula-
tion detected by light
microscopy indicating pos-
sibly significant quantity of
chrysotile
The medium to high con-
centration of short fibers
was unique to this particular
sample; despite this char-
acteristic, many fibers were
detected by light microscopy
No significant number of
fibers shapes were seen by
light microscopy
-------
TABLE?. (Continued)
Water Sample Number
Observation in TEM of
TPLS Fraction
TEM Observations of
As-Received Sample
Debris Chrysotile
LM Observations of
TPLS Fraction
Remarks
M3 0307 0730 Dl WIO 11/151
(filter sample WIe)
M30307 1500D2WGO11/15IV
(filter sample WG2)
PI 0130 0555 D2 WIO 11/15 III
(filter sample WI2)
R2 0305 0615 Dl WIO 11/10 III
(filter sample WI3)
SI 0326 0700 D4 WIO 11/10 IV
A very clean heavy mat of chrysotile Heavy Heavy
fibers too thick to penetrate with
electron beam except at edge of
filtered deposit
A heavy mat of clean chrysotile Medium Very
fibers too thick to penetrate with to heavy heavy
electron beam except at edge of
deposit
A little extraneous organic par-
ticulate; very few chrysotile fibers
Medium None
A clean separation containing a Very
few long chrysotile fibers; very heavy
little extraneous particulate
A heavy mat of chrysotile coated Very
with a structureless organic film heavy
Very few
(other fiber
types
present)
None
visible
Mat of fibers
Heavy mat of fibers
seen
Almost no fibers
seen
A few long fibers
Heavy mat of fibers
plus heavy extraneous
particles
The TPLS fraction when de-
posited on the Nuclepore
filter contained so many
fibers that individual fibers
could be discerned only at
the edge of the deposit
A very dense chrysotile mat
from the TPLS fraction in
which individual fibers were
evident only at the edge
of the filtered deposit
No significant numbers of
fiber shapes seen in the
light microscope
Although the as-received
sample contained other fiber
types than chrysotile, only
chrysotile was detected in
the TPLS fraction
This sample is one of two
in the sample set that con-
tained a large quantity of
interfering material in the
TPLS fraction
-------
TABLE 7. (Continued)
Water Sample Number
Observation in TEM of
TPLS Fraction
TEM Observations of
As-Received Sample
Debris Chrysotile
LM Observations of
TPLS Fraction
Remarks
Gl 0505 0450 D4WIO 11/8 II
11/22 II (filter sample WI3)
N2 0302 1805 D3 WGO 11 /10 II
(filter sample WG1)
N8 0403 1730 D6 WGO 11/101
(filter sample WG2)
Nl 0606 1335 D2 WGO 11/29 I
11/8 I (filter sample WG2)
Chatfield No. 1
(US sample)
Chatfield No. 2 & 3
(Canadian sample)
No chrysotile and very few
extraneous particles
No chrysotile found, but organic
filaments were present
No chrysotile found, but organic
filaments were present
A medium to heavy amount of
chrysotile, also some platy
mineral particles
Heavy
None
Medium None
Heavy
None
Heavy
None
Gross numbers of chrysotile fibers
and bundles of fibers; a clean
separation
Many chrysotile fibers plus several
diatoms and other unidentified
structures transferred along with
chrysotile
Heavy
Very
heavy
Medium
Heavy
A few fiber shapes
seen
A few fiber shapes
seen
A few fiber shapes
seen, but not con-
strued as chrysotile
Many fibers detected
Dense population of
chrysotile
Heavy fiber con-
centration plus
diatoms and other
particles
The few fiber shapes de-
tected in the light micro-
scope gave a weak false posi-
tive result
A weak false positive result
Although fiber shapes were
detected in the light micro-
scope, they did not have the
appearance of chrysotile
A case in which the debris
was so heavy in the as-
received preparation that no
chrysotile was detected
despite a relative high
chrysotile concentration
A very good separation of
gross amounts of chrysotile
This is the other sample of
two in the sample set that
contained a fairly heavy
concentration of interfering
particulate in the TPLS
fraction
-------
8. TPLS procedures will extract wide ranges of chrysotile concentrations from
real water samples. Some extractions resulted in mats of fibers too thick
to see through with either the TEM or LM, whereas, others contained only
a few or no fibers.
It was concluded from this study of the application of TPLS to actual water samples, that
TPLS is an effective approach to more sensitive detection of chrysotile by TEM or LM in water
samples than in transmission electron microscopy applied to unashed filtered deposits of as-
received samples. This is true even under the conditions of incomplete TPLS chrysotile
recovery. Also, light microscopy detected extracted chrysotile from samples in which both
LM and TEM examination of as-received filtered deposits revealed none.
Light microscopic counts were made on the carbon-coated TPLS deposits on the 0.2-Mm-
pore-size Nuclepore filters. Although possible interferences existed in a few samples in the form
of a platy mineral and filamentous organic structures that might be identified as chrysotile by
an inexperienced observer, they did not obscure the presence of significant quantities of
chrysotile or contribute to erroneously high counts. The results are presented in Table 8.
Four samples contained gross amounts of chrysotile making it impossible to perform
counting. An example of the concentration of fibers is shown in Figure 19. This is a photo-
micrograph taken at the edge of the filtered deposit extracted by TPLS. TPLS can handle these
large quantities without difficulty.
TPLS-Spot Test Analysis of Real Water Samples
The periodate-tetrabase catalytic spot test procedure, as previously described, was applied
to TPLS fractions filtered onto Millipore filters. The deposit was confined to a 4-mm-diameter
area by using the filter reservoir with a 3-mm-diameter hole. All deposits were dried before
reacting the chrysotile with manganous ion.
Figure 20 shows a comparison between spot test results on standard chrysotile concentra-
tions and on chrysotile concentrations extracts from 50-ml sample aliquots by TPLS of spot
test analyses of four real water samples.
In cases where no chrysotile was indicated by the spot test, some extraneous particulate
was observed. The color prints shown are not the same hue, and it is difficult to obtain prints
which are comparable. The positive transparencies from which the prints were made, however,
have the same hue and show differences in depth of color better. In view of this, it is recom-
mended that color transparencies of standards be viewed on a light box when comparisons of
analytical results from real water samples are made to standards in the filter deposit indicating
that the spot test procedure was selective for chrysotile in these cases.
Table 9 presents results of the estimated chrysotile concentrations in nine water samples.
45
-------
TABLE 8. LM ANALYTICAL RESULTS ON REAL WATER SAMPLES
Sample
C9305 1215U4WGO
N2 03021 805 D3WGO
Nl 0606 1335D2WGO
M30307 1500D2 WGO
SI 0326 0700 D4 WIO
M3 037 0730 Dl WIO
PI 01 30 0555 D2 WIO
P402120920D5 WGO
R203050615D1 WIO
C90305 1340D5 WGO
C90305 1350U5 WGO
Gl 0505 0450 D4 WIO
N80403 1730D6WGO
C3 05141 530 D3 WGO
C3 05141 530 U3 WGO
Chatfield Sample 3
(American Source)
Chatfield Sample 1
(Canadian Source)
Light Microscope Counts
0
0
439
Too many fibers to count
Too many fibers to count
Too many fibers to count
134
78
85
Too many to count
705
0
0
382
14
7,926*
594,500*
Nanograms
per Filter
0
0
32
>500
>500
>500
10
6
6.8
>500
50
0
0
21
1.2
400
28,000
Nanograms
per Liter
0
0
640
> 10,000
>1 0,000
>1 0,000
200
120
136
>1 0,000
1,000
0
0
420
24
5,300
373,000
*For Sample 3, three fields within #10 ring in the concentric circle eyepiece reticule were counted; for
Sample #1, one field delineated by ring #2 was counted.
46
-------
Figure 19. Light micrograph showing a heavy mat of TPLS-extracted
pure chrysotile from sample M3 0307 1500 D2 WGO.
-------
Series of standard chrysotile concentrations
Spot test results on real water samples
Figure 20. Comparison of spot test results on real water samples to spot
test results on a series of chrysotile standards.
TABLE 9. SPOT TEST RESULTS ON REAL
WATER SAMPLES
Sample
PI 0130 0555 D2
M3 0307 1500 D2
R2 0305 0615 Dl
C9 0305 1215 U4
M3 0307 0730 Dl
C9 0305 1350 U5
C9 0305 1340 D5
C9 0305 1215 U4
Estimated Concentration
Per Filter
48
-------
SECTION 5
INTERLABORATORY ANALYSIS OF FOUR WATER SAMPLES
Four water samples supplied by Dr. Eric Chatfield of the Ontario Research Foundation
(ORF) were analyzed at both Battelle and ORF by the TPLS-light microscopic analytical
procedure.
Instructions describing the procedure were supplied ORF by Battelle. The first attempt at
ORF to apply the technique as described revealed that the instructions required more clarifica-
tion. In particular, the isooctane remained cloudy after the separation procedure, which led the
ORF group to believe that they were not following the correct procedure. Cloudy isooctane is
normal, however. Another problem concerned the precise conditions for the optical microscope
count. As initially planned, Dr. E. J. Chatfield and Ms. M. J. Dillon visited Battelle's Columbus
Laboratories to obtain firsthand experience with the technique before attempting analysis of
the interlaboratory samples.
The samples supplied by ORF were from Thetford Mines, Quebec; Toronto, Ontario;
Asbestos, Quebec; and Lloyd Lake, Ontario. Both heavily loaded and lightly loaded samples,
which also contained some extraneous debris, were provided. By electron microscopy, ORF
estimated these samples to range from about 100 Mg/liter to 10 ng/liter of chrysotile asbestos.
Battelle did not analyze these samples by electron microscopy.
At ORF, some deviation from conditions in the instructions was necessary because of
equipment and supply limitations. In particular, Nuclepore filters were of 0.1-/urn pore size,
and backing Millipore filters were of 0.45-jum pore size. All shaking was by hand, and double-
distilled water was used instead of deionized water.
ORF detected some artifacts on the surface of some blank Nuclepore filters that might be
mistaken for chrysotile fibers. They state that in many cases the artifact can be recognized for
what it is: a glancing perforation. In others there is some difficulty in rejecting it as a genuine
fiber. No comparable artifacts were observed at Battelle.
The ORF researchers concluded that the two-phase liquid separation technique is a "most
effective chrysotile extraction method". They report that the technique allowed processing of
much larger volumes of dirty water samples than was possible by direct TEM methods. Also,
as Battelle reports, extraction into the organic phase is not totally specific to chrysotile, which
sometimes gives rise to difficulties in interpretation of the optical fiber counting. Dr. Chatfield
and Ms. Dillon conclude, however, that an experienced microscopist could in many cases dis-
tinguish the chrysotile from the other debris.
49
-------
Specific suggestions were made by ORF to assist in clarification of TPLS-light microscopic
analytical instructions. These have been heeded and the instructions modified accordingly.
A general suggestion was made concerning the requirements for optical microscopy. ORF
was unable to duplicate the exact light microscope magnification conditions employed at
Battelle. The distance between adjacent concentric circles of the eyepiece reticule was 10.0 ,um
with their microscope rather than 12.5 jum as with the Battelle microscope. Consequently,
Dr. Chatfield suggested that the standard curve relate total fiber length to chrysotile mass in-
stead of numbers of 12.5-pm lengths to chrysotile mass. Total fiber length is independent of
the features of a particular microscope, so long as an adequate magnification is in use. The
eyepiece reticule in any microscope then need only be calibrated and measurements expressed
as total fiber length.
Table 10 presents the comparison of the ORF and Battelle results of analyses on four
water samples. Agreement is approximately within a factor of 4 in all cases except for Sample
4, in which the Battelle result is ^0.1 that of ORF. In this case, the Battelle result agrees with
the electron microscopy result obtained by ORF. According to results from both laboratories,
all of these samples would be destined for more accurate analysis by electron microscopy, with
the possible exception of Sample 4.
The results of this interlaboratory comparison emphasizes the need for more extensive
round robin testing of the methodology to lead to further improvement of the reliability of
results.
50
-------
TABLE 10. COMPARISON OF ANALYTICAL RESULTS OBTAINED AT BATTELLE
AND ORF ON FOUR WATER SAMPLES
Sample Details
Sample
Number
1
3
4
5
Source
Lloyd Lake
Ontario
Thetford Mines
Quebec
Toronto
Ontario
Asbestos
Quebec
Water
Type
Raw
Treated
Raw
Treated
Description
Sand, day
Municipal water
General debris
Lake Ontario
Municipal water
Fine sand
Estimated Mass
by TEM
(jug/liter)
50
1.0
1.3 x 10~2
8.0 x 10~2
Optical
Number of
10.0 Aim
Segments
233
830
92
234
Counts
Area
Examined
(mm2)
0.0785
3.20
3.20
3.20
Equivalent
Number of
12.5 /on
Segments
186
664
74
187
Number of
12.5 /An Segments
Equivalent to
Diameter Scan
9478 ORF
4113 Battelle
830 ORF
5480 Battelle
92 ORF
9 Battelle
234 ORF
668 Battelle
Mass of
Chrysotile
(jug/filter)
0.60
0.28
0.065
0.35
0.0085
0.0013
0.020
0.050
Aliquot
Volume
(ml)
50
75
75
75
75
75
50
75
Calculated
Mass of
Chrysotile
(jug/liter)
16
3.7
1.16
4.7
0.15
0.0177
0.53
0.66
-------
REFERENCES
1. Speil, S. and Leinweber, J. P., "Asbestos Minerals in Modern Technology", Environmental
Research, March, 1969.
2. Rosenhaltz, V. L., Smith, F. T., "The Dielectric Constants of Mineral Powders", American
Mineralogist, 21, 115-120, 1936.
3. Mumpton, F. A., "Characterization of Chrysotile Asbestos and Other Members of the Ser-
pentine Group of Minerals", Siemens Review, 75-84, 1974.
4. von Reinders, W., Kolloid Zeitschrift, Vol. 13, p 235, 1913.
5. Chowdhury, S. and Kitchener, J. A., "The Zeta Potentials of Natural and Synthetic
Chrysotiles", Int. J. of Min. Proc., 2, 277-285, 1975.
6. Lai, R.W.M. and Fuerstenau, D. W., "Liquid-Liquid Extraction of Ultrafine Particles",
Trans. AIME, Vol. 241, 549-566, 1968.
7. Sherman, P., "Emulsion Science", Academic Press, p 6, 1968.
8. Bennett, H., Bishop, J. L., Wulfinghoff, M. F., "Practical Emulsions", Chemical Publishing
Company, p 33, 1968.
9. Sutheim, G. M., "Introduction to Emulsions", Chemical Publishing Co., p 161, 1946.
52
-------
APPENDIX A
INSTRUCTIONS FOR CARRYING OUT TWO-PHASE LIQUID
SEPARATION COUPLED WITH LIGHT MICROSCOPIC
DETECTION OF CHRYSOTILE IN WATER SAMPLES
The rapid analytical method developed at Battelle-Columbus for determining chrysotile
asbestos in water involves first the separation of chrysotile from other waterborne particulate
followed by quantitative analysis of the separated fraction by light microscopy. Separation is
accomplished by a procedure in which the selective transfer of chrysotile from water to iso-
octane is promoted by a reaction with an anionic surfactant. Subsequently, the separated
chrysotile is deposited on a Nuclepore filter and its concentration is determined by a light
microscopic counting technique.
The separation procedure is called two-phase liquid separation (TPLS). The selective trans-
fer of chrysotile occurs when the isooctane and water are shaken together in a separatory funnel
after the chrysotile surface is rendered hydrophobic through a surface reaction with an anionic
surfactant MO-70, dioctyl sodium sulfosuccinate. The presence of the surfactant causes an
emulsion to form, which is broken by "salting out" with sodium chloride. After the phases
have separated, the isooctane is filtered to deposit the chrysotile on a Nuclepore filter. Then,
the filtered deposit is coated with vacuum-evaporated carbon. The carbon coating serves to
make the preparation highly reflective and also increases the apparent fiber diameters to bring
more fibers into the detectable range of light microscopy.
Light microscopic analysis is carried out at 500X magnification using incident bright-field
(vertical) illumination to count fiber segments equal in length to the distance (12.5 Mm) be-
tween concentric circles in an eyepiece reticle. The counts expressed as total fiber length are
then translated into micrograms of chrysotile per filter deposit by reference to a standard curve
that relates total fiber length to mass of chrysotile per filter. Finally, the chrysotile mass Gzg)
per liter of water sample is calculated from the aliquot taken, assuming 75 percent recovery.
The following detailed procedure presents step-by-step the instructions for carrying out
TPLS coupled with light microscopic analysis.
Step 1: Shake the bottle of water sample with a mechanical shaker (Figure A-l) for approxi-
mately 5 minutes at «180 cycles per minute to create a uniform particulate
suspension.
Step 2: Take an aliquot either directly from the water sample, or, alternatively, filter an ali-
quot onto a Metricel DM450 filter.
Alternative 2A: Take a 75-ml aliquot directly from the uniform water sample. Adjust
the pH to 3.5 ±0.1 with 0.1 N HC1, add 25 ml of 0.01M MO-70, heat to 140°F
53
-------
while stirring, and allow to cool to room temperature. If acid additions overshoot the
correct pH value, correct by adding 0.1 N NaOH. Heat at least to 140°F; heating as
high as the boiling point has no adverse effect except to extend the time required for
cooling. Cooling may be accomplished by refrigeration or in an ice bath. Also, other
than 75-ml aliquots may be taken; then the amount of surfactant added must be ad-
justed to maintain the 2.5 x 10~3 M concentration, and the amount of isooctane must
be 1/3 of the total liquid volume.
Alternative 2B: Filter a relatively large aliquot (1 liter or more) of water sample to
deposit the particulate on 47-mm Metricel DM450 (0.45Mm pore size) filter(s)
(change filters if necessary as they become loaded to the degree that the water flow
stops). Before the filter deposits have dried, ultrasonically remove the deposited par-
ticulate from the filter(s) in a beaker containing 75 ml of water to which 25 ml of
0.01 M MO-70 was added to make a total volume of 100 ml. Adjust the pH to
3.5 ±0.1 with 0.1 N HC1, heat to 140°F, and allow to cool to room temperature.
Step 3: Combine the water phase from Step 2A or 2B with 50 ml of isooctane in a 250-ml
separatory funnel equipped with a Teflon stopcock.
Step 4: Attach the separatory funnel to a mechanical shaker and shake at ^150 ±25 cycles
per minute for 5 minutes (Figure A-2).
Step 5: Add 10 ml of 10% NaCl to break the emulsion (Figure A-3). Shake lightly to dis-
tribute the NaCl in the emulsion. Occasionally the isooctane layer will form under
the aqueous layer. Further gentle shaking will eliminate this condition.
Step 6: Remove the separatory funnel stopper during separation of the two liquid layers.
Drain off the water phase through the stopcock (Figure A-4). Allow time for the
water droplets to settle and drain them off before pouring the isooctane from the top
of the separatory funnel into the 25-mm Millipore filtering apparatus (Figure A-5).
It is extremely important to take care that no water is filtered along with the isooc-
tane. The isooctane is filtered through a 0.2-ptm-pore-size, 25-mm Nuclepore filter,
shiny side up, backed with an 8.0-jLim-pore-size, 25-mm Millipore filter to obtain a
uniform chrysotile distribution.
Step 7: Deposit the separated chrysotile onto the Nuclepore filter by aspirating the isooctane
through the filter.
Step 8: Combine the same water phase again with 50 ml of isooctane in the separatory funnel
and repeat the extraction procedure without adding additional sodium chloride, and
deposit the extracted chrysotile on the same Nuclepore filter. Repeat the extraction
procedure three more times to make a total of five extractions from the same water
aliquot deposited on the same Nuclepore filter.
Step 9: Remove the Nuclepore filter, allow it to air dry and tape it by the edges, deposit
side up, so that it lies flat on a 3 x 1-inch labeled microscope slide (Figure A-6).
Double sided adhesive tape may be used to hold the filter flat for microscopic
examination.
Step 10: Evaporate carbon onto the surface of filter bearing the deposit. (Evaporate 10.0
millimeters by 1.0 mm graphite rod at a distance of 20 cm at normal incidence.)
Step 11: Use vertical (incident bright-field) illumination to view the carbon-coated filter at
magnification with a light microscope equipped with an eyepiece reticle having
54
-------
10 concentric circles (Figure A-7). In the eyepiece reticle, the radius of the center
circle is equal to the distance between adjacent concentric circles.
Step 12: Position the filter beneath the objective lens so that a traverse may be made across
the full diameter of the particulate deposit.
Step 13: Determine the smallest detectable fiber diameter. One count by the light microscopic
method is registered for a fiber with the minimum detectable diameter and a segment
length equal to the distance between concentric circles. At 500X, this distance is
12.5 Mm in the Leitz Orthoplan microscope used at Battelle. For other microscope
setups the field delineated by the concentric circles must be calibrated using a stage
micrometer.
Step 14: Perform fiber segment counts on fibers that fall within the outside concentric circle
as the major diameter of the filtered deposit is traversed. As counts are made,
roughly estimate the number of minimum fiber diameters encountered in fiber
bundles and adjust the fiber segment count accordingly.
Step 15: Translate fiber segment counts into total fiber length by multiplying the number of
counts times the calibrated distance between adjacent concentric circles. Refer the
total fiber length to the standard curve (Figure A-8), which relates total fiber length
to chrysotile mass (jug) per filter.
Step 16: Calculate the chrysotile concentration (jug) per liter from the mass per filter obtained
from the standard curve as follows:
pig/filter x 100 x 1000 ml = /ig/liter
75 ml of water aliquot
This calculation is based on 75% recovery. When the outside circle of eyepiece reticle
delineates a field 250 ,um in diameter, and the major diameter of the deposit is
16 mm, one 250-Mm-diameter field area is 1/82 of the area of the path 250 /zm wide
and 16 mm long across the filter. Because the standard curve was derived using this
path area (a»4.0 mm^), areas measured with other microscope conditions must be de-
termined and compensations made for differences by dividing the total path area by
the area over which measurements were made and multiplying this ratio times the
number of fiber counts times the calibrated distance between concentric circles.
When the radius of the smallest circle in a set of 10 circles is equal to the spacing between
adjacent circles, the following area percentages may be used in calculations to normalize fiber
length data obtained by counting less area than that within the outermost circle.
Circle No. 10 100%
Circle No. 9 81%
Circle No. 8 64%
Circle No. 7 49%
Circle No. 6 36%
Circle No. 5 25%
Circle No. 4 16%
Circle No. 3 9%
Circle No. 2 4%
Circle No. 1 1%
55
-------
Making Up Solutions
It is imperative that all solutions be filtered before use. It is suggested that solutions be
filtered through 0.2-/xm-pore-size, 47-mm-diameter Millipore filters.
MO-70 (Dioctyl Sodium Sulfosuccinate)
MO-70 is obtainable from Mona Industries, Inc., Paterson, New Jersey 07524, in 70 per-
cent aqueous solution. To make up the 0.01 M stock solution, weigh out 6.34 ±0.02 g. Heat
the solution to 140-150°F to obtain a uniform solution while stirring. Dilute to 1 liter with
deionized water, and filter. (Shelf life — 1 month).
0.1 N HC1 and 0.1 N NaOH
Purchase vial of Dilut-lt (G. T. Baker Chemical Co.) for making up 0.1 N HC1 or 0.1 N
NaOH. Dilute the contents of the plastic ampoule to 1 liter with deionized water, and filter
through 0.2-pm-pore-size, 47-mm Millepore filter.
10% NaCl
Add 90 ml of deionized water to lOg of reagent grade NaCl and filter.
56
-------
Figure A-l. The mechanical shaker.
Figure A-2. Separatory funnels attached to mechanical
shaker with rubber bands.
57
-------
Figure A-3. Before (right) and after (left) breaking the
emulsion by "salting out" with NaCl.
58
-------
Figure A-4. Draining off the water phase.
:
Figure A-5. Filtering the isooctane.
59
-------
Figure A-6. A TPLS specimen prepared for light microscopic analysis. The
filtered TPLS fraction was taped to a microscope slide and
coated with vapor-deposited carbon in a vacuum evaporator.
60
-------
Figure A-7. Concentric circles in eyepiece reticle superimposed on image
of TPLS fraction from ORF water sample.
61
-------
125,000,000 p
12,500,000 —
I I I II I I I II I I 11 I I I Ill I III
0.0001 0.001 0.01 O.I 1.0 10
Pure Chrysotile, fj.g
Figure A-8. Standard curve which relates light microscope measurement of
total fiber length to mass (jug) of chrysotile per filter.
62
-------
APPENDIX B
INSTRUCTIONS FOR CARRYING OUT TWO PHASE LIQUID
SEPARATION (TPLS) COUPLED WITH SPOT TEST
DETECTION OF CHRYSOT1LE IN WATER SAMPLES
The rapid analytical method alternative to two-phase liquid separation - light microscopic
detection procedure is the TPLS - spot test procedure. Although spot test detection is less
sensitive than light microscopic detection (0.1 Mg versus 0.001 /xg), spot testing is more rapid,
and its sensitivity is adequate for some monitoring situations. Spot testing also has the advan-
tage of being routine, rapid, requiring no equipment as expensive as a light microscope, and
very little training. After separation of the chrysotile is accomplished by TPLS, spot test de-
tection can be run in 10 minutes.
The only equipment needed are a 25-mm Millipore filtering apparatus, a special reservoir
with a 3-mm hole for the filter apparatus, a heat lamp, a wash bottle, spot test chemicals, and
a spraying device such as is used in thin layer chromatography. The materials required are
25-mm-diameter, 0.45-jum-pore-size Millipore filters, spot-test chemicals (potassium periodate
and tetrabase), and white blotters.
A technician who has no formal scientific training can be taught to run the spot test de-
tection procedure in less than 1 hour. Semiquantitative results can readily be evaluated with a
little practice to determine whether there are changes in chrysotile concentration within the
range of detection of the method.
The chrysotile extraction procedure is the same as that recommended prior to light micro-
scopic detection; however, a Millipore filter is used in the spot-test technique rather than a
Nuclepore filter as is used for the light microscopic detection.
The selective transfer of chrysotile in the TPLS procedure occurs when the isooctane and
water are shaken together in a separatory funnel after the chryostile surface is rendered hydro-
phobic through a surface reaction with the anionic surfactant MO-70, dioctyl sodium sulfo-
succinate. The presence of the surfactant causes an emulsion to form, which is broken by
"salting out" with sodium chloride. After the phases have separated, the isooctane is filtered to
deposit the chrysotile on a Millipore filter. The deposit is confined to a 4-mm-diameter filter
area to enhance spot test detection.
Spot test detection involves first the reaction of manganous ion with chrysotile extracted
from a water sample followed by the detection of the reacted manganese by a spot test reaction
which is catalyzed by manganese. The manganese catalyzes the reaction between potassium
periodate and tetrabase [4, 4' Methylenebis (N, N-dimethylaniline)] to yield a bright blue
63
-------
reaction product. The depth of color developed is indicative of the quantity of manganese
present in the chrysotile structure.
The following detailed procedure presents step-by-step instructions for carrying out TPLS
coupled with chrysotile detection by the catalytic periodate-tetrabase spot test.
Step 1: Shake the bottle of water sample with a mechanical shaker (Figure A-l) for approxi-
mately 5 minutes at ^ISO cycles per minute to create a uniform particulate
suspension.
Step 2: Take an aliquot either directly from the water sample, or, alternatively, filter an ali-
quot onto a Metricel DM450 filter.
Alternative 2A: Take a 75-ml aliquot directly from the uniform water sample. Ad-
just the pH to 3.5 ±0.1 with 0.1 N HC1, add 25 ml of 0.01M MO-70, heat to 140°F
while stirring and allow to cool to room temperature. If acid additions overshoot the
correct pH value, adjust by adding 0.1 N NaOH. Heat at least to 140°F; heating as
high as the boiling point has no adverse effect except to extend the time required for
cooling. Cooling may be accomplished by refrigeration or an ice bath. Also, other
than 75-ml aliquots may be taken; then the amount of surfactant added must be ad-
justed to maintain the 2.5 x 10~3 M concentration, and the amount of isooctane must
be 1/3 of the total liquid volume.
Alternative 2B: Filter a relatively large aliquot (1 liter or more) of water sample to
deposit the particulate on 47-mm Metricel DM450 (0.45-jum-pore-size) filter(s)
(change filters if necessary as they become loaded to the degree that the water flow
stops). Before the filter deposits have dried, ultrasonically remove the deposited par-
ticulate from the filter(s) in a beaker containing 75 ml of water to which 25 ml of
0.01 M MO-70 was added to make a total volume of 100 ml. Adjust the pH to
3.5 ±0.1 with 0.1 n HC1 or 0.1 n NaOH, heat to 140°F, and allow to cool to room
temperature.
Step 3: Combine the water phase from Step 2A or 2B with 50 ml of isooctane in a 250-ml
separatory funnel equipped with a Teflon stopcock.
Step 4: Attach the separatory funnel to a mechanical shaker and shake at «* 150 ±25 cycles
per minute for 5 minutes (Figure A-2).
Step 5: Add 10 ml of 10% NaCl to break the emulsion (Figure A-3). Shake lightly to dis-
tribute the NaCl in the emulsion. Occasionally the isooctane layer will form under
the aqueous layer. Further gentle shaking will eliminate this condition.
Step 6: Remove the separatory funnel stopper during separation of the two liquid layers.
Drain off the water phase through the stopcock (Figure A-4). Allow time for the
water droplets to settle and drain them off before pouring the isooctane from the top
of the separatory funnel into the 25-mm Millipore filtering apparatus (Figure A-5).
It is extremely important to take care that no water is filtered along with the
isooctane.
A special filter reservoir is used to concentrate the chrysotile into a small area
(Figure B-l). [The special reservoir, used as a substitute for the standard reservoir, is
made with a 3-mm hole to concentrate the asbestos in a 4-mm-diameter filter area
(Figure B-l). A glass blower can readily make this special part.]
64
-------
Step 7: Aspirate the isooctane through a 25-mm, 0.45-/«n, Millipore filter to deposit the ex-
tracted chrysotile.
Step 8: Combine the same water phase again with 50 ml of isooctane in the separately funnel
and repeat the extraction procedure without adding additional sodium chloride, and
deposit the extracted chrysotile on the same Nuclepore filter. Repeat the extraction
procedure three more times to make a total of five extractions from the same water
aliquot deposited on the same Millipore filter.
Step 9: Remove the filter, air dry and store in a labeled plastic petri dish.
Step 10: Place the filter, deposit side up on the frit of a 25-mm Millipore filtering apparatus
and place a blank Millipore filter on top of the filter bearing the chrysotile deposit.
Clamp the standard Millipore reservoir (16-mm-hole) in position.
Step 11: Pour water into the reservoir and adjust the aspiration rate to 15 ml per minute.
Step 12: Aspirate 15 ml of 25% manganous sulfate (25 g MnSO4 • H^O/lOO ml solution)
through the filter.
Step 13: Wash five times aspirating each washing completely through the filter before adding
the next washing.
Step 14: Increase the aspiration rate to hold the filters to the frit and remove the filter
reservoir.
Step 15: Separate the filters and discard the top filter.
Step 16: Place the bottom filter deposit side up on a white blotter and dry under the heat
lamp at 145°F for 1 minute.
Step 17: Attach the filter deposit side up by its edges with cellophane tape to the white
blotter.
Step 18: Use a spraying device (Figure B-2) to spray the filter two times lightly with the spot
test working solution which is made up by mixing equal (5 ml) quantities of the fol-
lowing Solutions A and B:
Solution A: Saturated solution of potassium periodate in 2N acetic acid (11.5 ml
glacial acetic diluted to 100 ml with water).
Solution B: 1.0 g of tetrabase dissolved in 100 ml of 95 percent ethyl alcohol.
Step 19: Place the filter under the heat lamp at 145°F until the color of the spot does not
change, or if no spot becomes visible, until the filter background turns a light blue
color.
Step 20: Compare the color of the spot to colors developed in similar preparations made from
known chrysotile concentrations. The latter may be recorded on 35-mm color trans-
parencies that can be viewed to make comparisons to unknowns. Color prints were
not found satisfactory for this purpose.
A color Xerox copy of a spot test result obtained on an 0.8 /ig chrysotile standard is
shown in Figure B-3.
Note: Filter all solutions through 0.45-Mm-pore-size Millipore filters before use.
65
-------
Figure B-l. Special reservoir with 3-mm-diameter hole.
Figure B-2. Spraying device with vial containing spot test working solution.
66
-------
•:! I'-•*- •
^,-«*-.• ,.
Figure B-3. A typical spot test result obtained on an 0.8-/ig chrysotile standard.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. DEPORT \o.
EPA-600/4-78-066
3. RECIPIENT'S ACCESSION NO.
TITLE A\'D SUBTITLE
Development of a Rapid Analytical Method for Deter-
mining Asbestos in Water
5. REPORT DATE-
December ]978 issuing date
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
C.W. Melton, S.J. Anderson, C.F.
R.E. Heffelfinger
8. PERFORMING ORGANIZATION REPORT NO
Dye, W.E. Chase, and
9. PERFORMING ORGANIZATION NAME AND ADDRESS
BATTELLE Columbus Laboratories
505 King Avenue
Columbus, OH 43201
10. PROGRAM ELEMENT NO.
1BD713
11. CONTRACT/GRANT NO.
68-03-2199
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Research Laboratory - Athens, GA
Office of Research and Development
U.S. Environmental Protection Agency
Athens, GA 30605
13. TYPE OF REPORT AND PERIOD COVERED
Final 6/75 - 6/78
14. SPONSORING AGENCY CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The development of a rapid analytical method for determining chrysotile asbestos
in water that requires substantially less time per analysis than electron microscopy
methods is described. Based on the proposition that separation of chrysotile from
other waterborne particulate would greatly simplify the task of detection, the researc
effort was directed toward establishing separation and concentration techniques. This
investigation led to the development of a separation procedure whereby chrysotile is
extracted from a water sample into an immiscible organic liquid phase. The procedure
is called two-phase liquid separation (TPLS).
TPLS has been combined with a light microscopic intercept counting technique and
with acolorimetric spot test detection technique to result in two complete rapid anal-
ytical methods. The TPLS-light microscopic (LM) method requires more expensive equip-
ment than the TPLS-spot test method; however, the limit of detection of TPLS-LM
method is 1.0 ng at the 99 percent confidence level, whereas the limit of detection of
the TPLS-spot test method is approximately 100 ng. The TPLS-LM method, therefore, is
recommended as a first choice, and the TPLS-spot test method is recommended for use
under conditions that require no greater detection sensitivity than 100 ng per sample.
Experiments have shown that 75 percent of the chrysotile is extracted by TPLS.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Asbestos
Chemical analysis
Electron microscopy
Water pollution
07B
13B
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
80
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
68
-.: U.S. GOVERNMENT PRINTING OfFICEr 1979 -657-060 /1558
------- |