4 e c’mccrone associates, inc

CONSULTING ULTRAMICROANALYSIS. MICROSCOPY. SMALL PARTICLE PROBLEMS. SOLID-STATE CHEMISTRY

3 December 1973

Hearing Clerk Food and Drug Administration Room 6-86 5600 Fisheries Lane Rockvtile , Maryland 2 0 8 52

Gentlemen:

We are writing to comment on the Commissioner’s proposal covering “Asbestos Particles in Food and Drugs” as published in the Federal Register, Vol. 38 No. 188, pages 27076-27081, 28 September 1973. We are interested in helping the FDA choose a practical analytical procedure for the detection and identification of asbestos.

Our laboratory has analyzed many hundreds of asbestos and asbestos-containing samples by most of the various analytical procedures: polarized light microscopy, x-ray diffraction, scanning electron microscopy with energy-dispersive x-ray analysis, transmission electron microscopy with selected area electron diffraction, electron microprobe analysis and electron microscope microprobe analysis. We would like to begin by giving you our evalu- ation of each of these methods.

The methods to consider are: 1. Morphology by light and electron microscopy

2. Optical properties by light microscopy

3. Diffraction by x-ray or electron diffraction

4. Elemental analysis by electron or ion microprobe

The various instruments available are capable of a variety of combinations of these methods (See Table 1 attached to this letter). The advantages and disadvantages of each area are discussed individually.

The polarizing light microscope (LM) requires great skill and background obtained only with long practice. A microscopist trained in optical crystallography and familiar with the serpentines and amphiboles still requires hours of time per sample to find and identify each of the possible asbestos types. The problem is complicated by the presence of interfering fibers (3:l aspect ratio!), e.g., paper, talc on edge, diatoms, quartz, mineral wool, plant hairs and fibers etc. Finally, all of the asbestos types show considerable variation in all of their properties and these variations must bc familiar to the microscopist.

Dispersion staining (DS) greatly speeds up the detection and identification of asbestos fibers. It is the only practical way to apply optical crystallography to samples containing small per- centages of asbestos. The technique can be learned much more rapidly than optical crystal- lography yet it is securely based on optical crystallographic properties.

X-ray diffraction (XRD), also based on crystal properties, is somewhat less sensitive than dispersion staining but requires no more background and skill. The sensitivity can be in- creased from 0.5-l. 0% to 0.1-O. 5% by step-scanning. One advantage over LM and DS is smaller size limit - about 50 nm compared with about 5,000 nm (5 pm).

Scanning electron microscopy with an energy dispersive x-ray analyzer is an overrated tool. Many substances besides asbestos are fibrous and it is nearly impossible to obtain a good chemical analysis of submicrometer particles in a particulate matrix with EDXRA in the SEM; there is too much interference. Fitting the SEM with a wavelength dispersive x-ray analyzer greatly improves it usefulness.

The transmission electron microscope (TEM) with selected area electron diffraction is an ideal tool when the asbestos fiber size is d 0.2 pm. The shape of fibers down to 0.005 pm is immediately apparent on scanning a sample and the diffraction pattern is immediately discernible.

The electron microprobe analyzer (EMA) with its wavelength dispersive x-ray analyzer is better than the SEM fitted with EDXRA but is surpassed by an SEM fitted with wavelength dispersive x-ray analyzer. The better morphological image of the SEM tips the balance.

EMMA, the combination transmission electron microscope-electron microprobe analyzer is, by far, the most useful of the analytical tools. for asbestos. It may not be needed if the fiber size is large enough (about 0.5 pm in smallest dimension), but smaller particles can be characterized by morphology, diffraction and chemical analysis in this one instrument. It is ideal for water samples containing suspended asbestos fibers.

The ion microprobe, although a highly sensitive trace analytical tool, is not usually used for asbestos fiber analysis. It does, however, detect fine fibers and identifies them by qualita- tive chemical analysis (quantitative when compared with standards).

Summarizing then - if the size of the asbestos fibers is very small (as it nearly always is for water samples) LM is useless and a diffraction method must be used. If the size is large enough, say 1 x 3 pm, and the percentage high enough, say -(O. Ol%, then LM can do the job. If high sensitivity is not needed, say about 0.2% or greater, then XRD is the best method. Very fine particles require the TEM with selected area diffraction.

It should not be necessary to emphasize that no method is better than the analyst who makes the analysis. Few laboratories have all of the above instruments and highly skilled analysts for each. Instead, each tends to use the tools available and the methods with which their personnel are more or less familiar. In a situation as important as this, the choice should not be based on availability of tools ‘and skills. Fortunately, the tools for detection and identification of asbestos are neither the most expensive nor the most sophisticated.

. -. . - L

Page Three

FDA ilrl, . Cur procedure (in brief), based on the analysis of many hundreds of samples, is:

1.

2.

3.

Screen the sample by DS using Cargille high dispersion liquid 1.550.

Jf positive: a) identify the asbestos type by DS or XRD, b) determ ine amount by counting using criteria suggested in the FDA method of 28 September 1973, or by XRD. (We m ight add that a 3:l length to width ratio is not exactly a fiber. The Particle Atlas specifies 1O:l; quartz flakes are often =>3:1).

If negative: a) conolude that less than 0.01% of chrysotile or amphiboles are present, b) use TEM or E M M A with EDXRA and SAED to be sure submicroscopic asbestos particles are not present.

This is a complete procedure which can usually be shortened considerably in practice. For example, a m ineral sample or a sample containing a m ineral fraction will generally not have been subjected to a procedure that elim inates larger fibers, e.g., ultrasoneration or sedimentation. Hence TEM is not necessary and DS can be used.

Before describing in more detail precisely how we proceed, let me comment on the proposed FDA procedure.

1. To do optical crystallography on talc the particles must be well dispersed. One mg under an 18 m m cover slip is far too crowded if one hopes to find individual fine asbestos fibers; one-tenth that concentration is much better.

2. It is true that all refractive indices of chrysotile are less than 1.574, but so are the indices of numerous other elongated subst‘ances including ubiquitous paper and glass fibers, In other words, 3:l aspect ratio particles having all indices cl. 574 are very probably not asbestos. To be certain they are chrysotile requires:

a) remounting another aliquot in, say, 1.560 to see if n 11 is about that value and if the other two n’s are about 1.550. Worse still, one has to f.ind those now nearly invisible (without DS) fibers among m illions of other particles; or

b) scan the sample at room temperature in a hot stage so that a suspected fiber in 1.574 could be shown to disappear with sodium light on heating at about 25 + [ (1.574-I. 560)/ 0.00045] or 56” C for n 11 and about 25 +[(I. 574-1.550)/O. 00045] or 78” C for nl . One would at least have the same particle in view throughout the test but it would still take hours, or even days, to check 100 particles by this procedure.

3. Talc, often on edge, has one index less than 1.574 and it takes time to check the other index to be sure they are not chrysotile.

4. It is likewise true that all amphiboles have all indices higher than 1.590 but the same difficulties hold true here as are given above for chrysotile. The amphiboles are even worse, however, because there are more of them , each one varies considerably in optical properties and there are many possible interfering substances.

r 3. Counting the number of fibers is not a good measure of % asbestos. We find the size of individual fibers varies 106-fold, at least.

.I ’ *Page Four 1 . FDA a 6. It requires a highly skilled m icroscopist to identify the amphiboles; especially in . small amounts, in confusing m ixtures and as small fibers. Of all the m icroscopists in the world fewer than 1% would be so qualified. Nearly all of these qualified m icroscopists are m ineralogists and a few chemical m icroscopists. Very, very few industrial laboratories, especially in food and drugs, have such personnel nor are they available on the job market in the number that would be necessary if this procedure is adopted as written.

Now, what would we suggest as modifications that m ight improve the proposed method? We recognize that the proposed method is basically good in that it is based on crystallographic data, the classical approach to the problem . We would keep this base but shift to diffraction rather than crystal optics. The diffraction method is easier to teach, easier to learn, just as dependable, works for all fibers (large or small) and is more rapid. It is better than crystal optics in assessing the type of amphibole especially in complex m ineral series like tremolite - actinolite - ferroactinolite or the cummingtonite - grunerite series. It is, however, somewhat less sensitive than light m icroscopy. W ith step scanning, a time- consuming but potentially automatic procedure, 0. Z-O. 5% of a given asbestos m ight be detected; by ordinary slow continuous scanning 0.5-l. 0%.

Criteria for the certain identification of each asbestos type can be found in the literature. We use the following criteria for chrysotile and tremolite.

Chrysotile - An x-ray powder pattern showing both the 002 (7.3 f 0.1 A; I/IO = 1.00) and 004 (3.65f 0.05 A; I/IO = 90 f 20) reflections and showing no line in the 14.6 A chlorite region is necessary evidence for the presence of chrysotile. A line at 14.6 A in addition to, an:1 of the proper intensity relative to, the 7.3 A and 3.65 A lines identifies chlorite rather than chrysotile.

Tremolite - Necessary evidence for the presence of tremolite is an x-ray powder diffractionpatternwithIines at8.4fO.lA, 3.27& O.O5A, 3.03?~0.05Asnd2.92-fO.O5A. (Note: zinc and magnesium stearate may interfere at 8.4 A and calcite at 3.03 A. )

Similar criteria have been set for the other amphiboles. If greater sensitivity is required we recommend step-scanning. This must be done through the regions of the three strongest lines for each of the asbestos types.

If still greater sensitivity is desired or if, for any reason, light m icroscopy is the preferred method, we then suggest dispersion staining. The overriding reason for DS rather than the classical methods of optical crystallography is the following. One cannot by the latter methods mount a sample in the diagnostic refractive index mounting liquid for, say, chrysotile because chrysotile thereby becomes neariy invisible and, especially in trace amounts, small needles would not be detected. For this reason your proposed m .ethod rcquircs mounting in 1.574 so that chlysotile fibers can be seen and their index relative to 1.574 determ ined. However, many other substances also have indices lower than 1.574 and, although bircfringence, extinction, sign of elongation and contrast may elim inate some, this takes time to check. DS, on the other hand, allows the analyst to mount the sample in 1.550, the liquid most likely to make chrysotile disappear when classical methods are used. Now, however, with DS chrysotile fibers show two brilliant and distinctive colors, one for n

II and one for nl .

‘. . -* -Page Five . . FDA

A* . This uniquely identifies chrysotile fibers almost at a glance; glass fibers, paper fibers, talc on edge, quartz, splinters etc all show, at least, different colors and an experienced m icroscopist can quickly identify them .

Similarly for the amphiboles, we mount the sample in 1.605 rather than 1.590 and for a corresponding reason. In 1.605, tremolite is easily identified by either yellow and blue colors on crystals with oblique extinction or yellow and golden .magenta on crystals with parallel extinction (n 11 and n have distinctive and differen i!

, respectively). Furthermore, actinolite and anthophyllite colors in this same liquid. One is left only with amosite

(positive elongation) and crocidolite (blue fibers with negative elongation). All five of these amphiboles can thus be detected and identified in this one liquid.

In conclusion, if forced to analyze a sample by the proposed method, we would require several (2-S) hours of a skilled crystallographer’s time; this would average $250 per sample. The time can be cut to 0.5-l hour and the cost to about $50 when dispersion staining is used. It is also practical to use less highly trained m icroscopists for the dispersion staining procedure. Although we’re not sure, reading the proposed method, whether we are excluded from using dispersion staining, it is obvious the writer of the procedure did not visualize doing so.

We have just completed, for example, an intensive 2-day course in these methods for EPA personnel and have scheduled a similar course for food and drug industry personnel 25-26 February 1974. It is feasible to train a m icroscopist with little previous knowledge of crystallography how to detect and identify asbestos by dispersion staining in two days. Not so by the- classical method; here, we would need 10 days and even then less than half of the students could dependably analyze any and all samples for serpentine and amphiboles.

We have also enclosed copies of two papers from our laboratory on asbestos. One is, unfortunately, still in press and without the cited Figures. We hope, in any case, that some of this material based on our own experience may be helpful to you. If we can help in any further way, please be sure of our desire to cooperate.

Sincerely,

Walter C. McCrone Scientific Adviser

W C M :bb

TABLE 1

Instrumental Methods - Capabilities

Instruments Method Morphology Optics Diffraction Chemical

Analysis *

LM: light microscope 31 pm 21 firn NO NO DS: dispersion staining 25 firn 21 pm NO NO XRD: x-ray diffraction NO NO YES YES

SEM: scanning electron microscope 30.005 pm NO NO YES

TEM: transmission electron microscope

20.001 pm NO YES NO**

EMA: electron microprobe 35 pm NO NO YES

EMMA: electron microscope- 30.005 /.tm NO YES YES microprobe anal:,->;er

IMA: ion microprobe 25 pm NO NO YES

*diffraction methods indirectly analyze chemically by identifying the crystalline phase present; chemical analysis by electron optical instruments may utilize energy or wavelength dispersive x-ray analyzers

**may be fitted with energy or wavelength dispersive x-ray analyzers

Walter c. mccrone associates, inc.

.

- . . .

1 I

t%~crosc @ pe 0 Incorporating “Crystal Front”. Founded 1937 January-April 1970 Volume 18 No. 1 AnnualSubscription: 0. $15. Secondclasspostagepaidat New York. Published by: Microscope Publications Ltd., 2 McCrone Mews, Belsize Lane, London, N.W.3., England.

Identification of Asbestos Fibers by Microscopical

Dispersion Staining* Y. JULIAN and W. C. McCRONE

McCrone Associates, Inc., Chicago, Illinois, U.S.A.

Abstract The Cherkasov focal screening dispersion staining procedure has been successfully applied to the identification of asbestos. The various types of asbestos can be differentiated by noting the refractive index of the Cargille liquid giving matching wavelengths in the region near 550 nm. It is not necessary to use polarized light although the much more definitive data obtained with polarized light may eventually permit identification of the mine from which each asbestos came.

Although too few samples of amosite and crocidolitehavebeenstudied, it appears that these two types can be differentiated by dispersion staining. In Cargille liquid n will show colors in the

8°C blue

1.680 the central stop without polars magenta region (h,= 550- 650 nm) for

amosite and in the golden yellow region (h,=400- 500 nm) for croci- dolite. Furthermore, with polarized light crocidolite will show lower birefringence than amosite; in terms of h, difference in a given liquid amosite will show 160- 190 nm and crocidolite 120- 140 nm. Finally, the higher value of h, is observed for the vibration direction parallel to the length for crocidolite (three samples) but perpendicular to the length for amosite (two samples).

* Presented at INTER/MICRO-69, London, England.

1

Through publications1-3 and current research, the pneumoconiotic (lung hardening) and frequently cancer-producing hazards of asbestos are finally fully acknowledged. Hence, there is need of a method of determining qualitatively and quantitatively environmental pollution by respirable asbestos dust (l-7 pm).

Asbestos is a fibrous form of silicate rock. There are six types: chrysotile, amosite, crocidolite, aciinolite, tremolite and anthophyllite; only the first three are of commercial importance. All are amphiboles (minerals with chains of silica tetrahedra as their basic structure) except chrysotile which is a serpentine (mineral made up of layers of silica tetrahedra).

The property of asbestos that best lends itself to identification is refractive index. The fibers are so fine that electron microscopy would be necessary if morphology were to be used. There are no dependable differences in absorption color (except possibly for crocidolite) nor specific gravity (other than chrysotile) (Table I). If we wish to use chemical composition for identification of single asbestos fibers, we require highly sophisticated instruments and techniques-the electron microscope or an electron (or ion) microprobe. X-ray diffraction requires considerable sample and is not sensitive to small percentages in any sample (i.e., < 5-100/0). Very small asbestos fibers can, however, be identified simply and quickly using dispersion staining.

Experimental The McCrone dispersion staining ob,jective, based on the focal screening method of Cherkasov4 was used m this study of asbestos (Figure 2).

I

s80 640 700

Figure I. Dispersion staming data for a typical sample of crocidolrte.

2 MICROSCOPE (1970) 18

TABLE I

ASBESTOS

SERPENTINES

Composition

Spec. Grav.

Cryst. Syst.

Extinction

Sign (elong.)

Chrysotlle

3Mg0.2S102. 2Hz0

2.36 2.5

Monochnic

y/l LX=O‘

+

-~ Actinohte

2Ca0.4MgO. FeO.SSi02.HzO

X03-3.5

Monoclinic

yAL=10-15-

+

-~

Tremohte Anthophyllite

2Ca0.5MgO. 7Mg0.8S102. 8S102. HZ0 HrO

2.9-3.2 2.85-3.4-l-

Monoclinic Orthorhomblc

yAL=lO-21” y/lL=O” yAL=14-21”

*L= long directlon of fibers

AMPHIBOLES

Amosite Crocldohte

5.5FeO. 1 .SMgO.

8Si02.H20

2.6-3.0

NazO.FeqO,. 3Fe0&02

Hz0

MonoclinIc

3.0-3.45

Monoclinic

aAL=3-15

w

.

.

CENTRAL STOP

PREPARATION -- -- t

CONDENSER

SUBSTAGE IRIS

F+wre 2. The arrangement for foccrl xreenmg: the annular stop permrts passage of the matching wavelength and the central stop white lrght minus the matching wavelength.

Figures 4-i’. Central stop colors for chry and crocidolite (7) each with the approprratc 1.610 (Flgure 51, 1.670 (Ftgwe 61, 1.100 (I east-west w each figure.

4

sotile (Figure 4),, anthophyllite (5). amosite (6) * Cargille refrurtrve index lrqurd, 1.560 (Ft:ure4), :igure 7). The vrbration direction of the polar is

MICROSCOPE (1970) 18

Dlsperslon Stalnmg Colors

rrs observed Annular

stop

dark-blue blue-vmlet

blue blue-green

I green yellow-green

yellow orange

orange-red browmsh-red

lumination are often used for these methods are empirical.

nixture of colors observed is tern used. With the Cherkasov d and hence essentially pure

was determined as a function each of 26 asbestos samples.

rble II was found to be most better resolution and contrast; bers. The procedure involved : ly in those Cargille refractive -0lors to it.

oth polarized and unpolarized

tbration direction of the polar other asbestos fibers). p and y (crocidolite) or a and f3 ndom orientations under these b G n < y positions and all

3 a positions. The two extreme attons are designated a (highest :stos fibers other than crocidolite t IL,) for crocidolite.

often observed with asbestos Figure 1 shows the data for one s way. bestos fibers show nominally a re wavelengths observed during due to polarization by reflectton

MICROSCOPE (1970) 18

1 CHRYSOTILE

400 500 ho

600 700

FIgwe 3 The range of matchrng wavelengths shown by anthophyllrte and chrysotrle wrth unpolarized light The CarglUe Irqurd was chosen JO that most samples of each asbestos show

FIgares 8-11 Central stop colors for chr~sotrle (Figure 8) anthophyllzte (9) amosrte (10) and crocrdohte (11) m the same Cargrlle lrqurds as FIgares 4-7 respectrvelv lrght *as used for this series

Unpolanred

Y. JULIAN & W. C McCRONE

from the microscope mirror, there is a slight change in color when the stage is rotated. All our data were taken with the fibers normal to the mirror reflection vibration direction.

Figures 4-7 show the dispersion staining colors of representative samples of chrysotile, anthophyllite, amosite and crocidolite res- pectively. Each is shown in the various orientations (relative to the polarizer vibration direction) necessarv for determination of a, j3 and y. Careful study of these figures may indicate the simplified procedure used to orient the fibers at precise 90 degree angles to each other.

To determine if the simpler application of dispersion staining without a polarizer is adequate for identifying asbestos, the fields of view of Figures 4-7 were also taken with no polar (Figures 8-11). The h, curves obtained without polar were plotted in a single graph (Figure 12). The bunching of the curves of each different type of asbestos indicates that for identificaiion purposes a liquid can be chosen that will impart a characteristic color to one species of asbestos only. A polarizing microscope is not, therefore, required. However, we can anticipate that with additional samples of crocidolite and amosite there might be some overlap. For this reason, the distinction between the signs of elongation (Figures 6 and 7) and the crocidolite blue absorption color are mentioned as special aids for differentiation. In this case a polarizer is useful.

li i j i i iiii#

CROCIDOLITES

AMOSITES

ANTHOPHYLLITES

CHRYSOTILES

h o (NM, i, c Figure 12. Matchrng wavelengths shown by deferent samples of four types of asbestos when mounted rn the Cargrlle refracttve index fiquds shown and observed wrth unpolartzed light.

7

The characteristic color will vary somewhat for different samples of the same type of asbestos. So, the problem arose of determir&g the ranges of colors for each asbestos in the chosen liquid (the liquid in which most samples of any asbestos type give a color of wavelength 580-620 nanometers). To determine this range of colors for any type of asbestos, the matching wavelengths observed in the chosen liquid are plotted against the number of different samples showing that A,. The area under the resulting curve (colored in Figure 3 as per Table II) shows all the colors shown by our samples of anthophyllite and chrysotile. Too few samples of amosite or crocidolite have been studied to justify their inclusion in Figure 3.

Conclusion The limited results obtained indicate that asbestos can be identified by dispersion staining. More work on many more samples will establish the full range of h, values for each asbestos. Relating h, data to asbestos source may then also pinpoint the mine source of unknown asbestos samples.

Acknowledgement The authors are grateful to Dr. Bertram G. Woodland, Curator of Igneous and Metamorphic Petrology of the Field Museum of Natural History in Chicago for the samples of asbestos from various sources.

REFERENCES 1 H.M. Factory Inspectorate, (1968). Problems arising from the use of asbestos,

Memorandum of the Senior Medical Inspector’s Advisory Panel, Department of Employment and Productivity, London, Her Majesty’s Stationery Office.

2 Smith, K. W., (1955). Pulmonary disabilrty in asbestos workers. Arch. Ind. Health, 12 198-203.

3 Brodeur, P., (1968). The Magic Mineral, New Yorker Magazine, 12 October. 4 Brown, K. M., and McCrone, W. C., (1963). Dispersion Staining, Parts I and II,

Microscope, 13 311 and 14 39.

8 MICROSCOPE (1970) 18

APPENDIX (Conrmued)

Lmcoln Co.. Nevada Y El101

; a

I .609 1.602 I.596 I .5*89

I.611 1.616 1.623 1.630 1605 1.609 1.615 1.621 1.599 1.602 1.609 1.615 I.592 I .595 1.600 1.606

ICI -.

Califorma El4464

Pine Mt., Georgia

Encampment, Wyoming

Bedford, Virguua El2790

Orange River, South Africa

Westerburg, South Africa

South Africa

Lydenburg Dist., Transvaal

Penge, Transvaal

1.610 1.605 1.603 I.596

1.619 1.614 1.612 1.601

I.607 1.602 1.597 1.589

1.619 I.613 1.609 1.598

1.614 1.609 I .606 1.599

1.615 1.620 1.630 1.641 1.610 1.615 1.624 1.635 1.607 1.611 1.620 1.628 1.600 1.604 1.611 1.618

I.624 1.630 1.642 1.656 1.619 1.624 1.635 1.646 1.616 1.622 1.632 1.642 1.605 1.610 1.618 1.627

1.613 1.619 I.632 1.647 1.608 1.614 1.625 1.638 I .602 1.607 1.617 1.628 I.593 1.597 1.606 1.614

1.620 1.626 1.630 1.638 1.616 1.619 1.625 1.630 1.612 1.615 I .621 1.626 1.600 1.603 1.607 1.61 I

1.619 1.622 1.634 1.644 1.613 1.618 1.628 1.638 1.610 1.615 1.624 1.633 1.603 1.607 1.614 1.622

Crocidolites

1.688 1.694 1.684 1.690 1.680 1.686 1.674 1.678

1.686 1.692 1.682 1.687 1.679 1.683 1.674 1.678

1.691 1.696 1.685 1.690 1.680 1.685 1.674 1.679

Amosites

1.676 1.682 1.665 1.670 1.658 1.663 I .652 1.657

1.672 1.677 1.665 1.669 1.660 1.664 1.653 1.657

L 1973

1.700 1.712 1.726 1.696 1.707 1.721 1.692 1.702 1.713 1.682 1.690 1.698

1.697 1.708 1.718 1.692 1.701 1.710 1.688 1.697 1.706 1.683 1.691 1.699

1.701 1.712 1.723 1.695 1.705 1.714 1.690 1.700 1.710 1.683 1.692 1.700

1.687 1.699 1.712 1.676 1.687 1.698 1.668 1.678 1.689 1.662 1.672 1.682

1.682 1.692 1.701 1.674 1.682 1.691 1.668 1.676 1.684 1.661 1.669 1.676

MICROSCOPE (1970) 18

Y. JULIAN & ‘.+(. C. McCRONE

APPENDlX

Dispersion Staining Data, n ss”C of Cargtlle refractive index liquid

Chrysotiles

Sample

Thetford Mines, Quebec

Orientation Matchmg wavelength, he, nm 450 486(F) 520 590(D) 1.539 1.542 1.546 1.553 y?;;;’

1.535 1.539 1.543 1.550 1.556 1.533 1.539 1.543 1.547 1.556 1.528 1.531 1.534 1.541 1.546

Eden, Vermont E19048**

Tbetford Mines, Quebec E20436

Thetford Mmes, Quebec El2185

Kmg’s Mine, Quebec

Barquisimento, Venezuela

1.549 1.553 1.558 1.566 1.574 1.543 1.547 1.551 1.559 1.566 1.543 1.547 1.551 1.559 1.566 1.534 1.539 1.543 1.550 1.557

1.545 1.549 1.552 1.540 1.546 1.547 1.538 1.541 1.544 1.536 1.539 1.542

1.543 1.547 1.550 1.556 1.563 1.540 1.543 1.546 1.552 1.558 1.536 1.540 1.543 1.549 1.555 1.531 1.534 1.537 1.543 1.548

1.543 1.547 1.551 1.539 1.543 1.547 1.535 1.539 1.543 1.529 1.532 1.535

1.536 1.541 1.546 1 530 1.537 1.541 1.530 1.534 1.539 1.547 1.524 I 528 1.532 1.540

1.559 I .553 1.550 1.548

1.559 1.566 1.554 1.561 1.550 1.557 1.542 1.547

1.556 1.550

1.565 1.558 1.554 1.552

1.566 1.559 1.555 1.548

* no polars were used. ** sample identification number of the Field Museum of Natural Htstory,

Chicago, Ilhnois, U.S.A.

Anthophylhtes

Minas Gerats, Brazd I

1.604 1.613 1.620 1.633 1.645 El6696 I.6432 3.608 1.614 1.628 1.638

B I .599 1.606 1.612 1.625 1.636 a I.582 1.588 1.594 1.606 I.616

Maryland 1.617 I .623 I.630 1.642 1.658 E3713 z I.613 1.618 1.624 1.636 1.649

P 1.609 1.614 1.619 I .630 1.642 a 1.599 I.604 1.609 1.620 I.630

New Mexico 7 1 620 1.625 1.629 1.638 1.648 E3709 ; 1.616 1.621 I .625 I.634 1.642

I.609 1.612 1.616 1.620 1.628 a 1.601 1.604 1.607 1.613 I.618

Macon Co., N. Carolma y I.614 1.618 1.623 I .632 1.641 E3700 ; 1.610 I614 I.619 I .627 1.635

I.606 1610 1.614 1.622 1.629 a 1.594 1 597 1.604 1.607 1.613

l Detection and Identification of Asbcsto.s

by Microscopical Dispersion Staining

Walter C . McCronc, McCronc Associates, Inc. Chicago, IL GOGlG

There is a great need for a dependable, sensitive and rapid method for the

detection and identification of asbestos, Microscopical dispersion staining satisfies

all of these requirements. It is depcndablc because it is based on the measurement

of three refractive indices as well as the dispersion of those indices. Refractive

indices are among the most valuable identifyin g characteristics for small pnrticles.

It is sensitive because the refractive indices are “read” as bright dispersion staining

colors against a black background, These colors can be observed on particles well

below on{. 1 nicromctcr in diameter. It is rapid because any particle in the micro-

scopical preparation showing the optical propertics of asbestos signals its presence

by a unique color combination with polarized light, One p‘article of asbestos in a field

of view containing many thousands of other particles will be immcdintely apparent on

scanning the eye across the field of view.

Besides optical crystallographic methods like dispersion staining only DTA

and diffraction have this ability to tag a pnrticulnr crystalline phase in a mixture.

X-ray diffraction, however, is several orders of magnitude less sensitive than dis-

persion staining and rc .;uires much more time. There are, however, con~i>ounds

whose dispersion staining colors arc, at Icast at first glance, confused with chrysotile

colors. Hcl’c:, fO1~iutl:ltcl.)r, particle mor~~holo;;y is able to djffcrentiate between {,hesc

interfering suhslances. ~u:~sl~ is one esa111p~c?, lizxrtlilc :~lother . The ln(ler, llo\vev(:r ,

is 8 tabul;lr talclikc mineral and qu2rl.z sflol\3 eonchoiclal fracture into usually thin

flakes. 130th 31-c c:lsily diffcret~ti:LIccl I’rom fibrous asbestos by morphology.

. . e . Dispersion staining is, therefore, a straightforward technique easily applied -

by any microscopist with some knowlodge of optical crystallography. We have found

it cxtrcmcly useful for the rapid and routine cxaminntion of any p‘articulate samples

for any of the v;crious kinds of asbcstosl. The nccess,ary background information for

applying t.hc method is given in Figure 1 which plots the matching wavclcngth, hO, as

a function of refractive index of the Cargillc refractive index liquids used in these

determinations. 3Jost of the asbestos minerals have distinctive indices without overlap.

I( Qqpk

Figure 1 &ows, however, “V

that amosite and crocidolite A&z overlap partially. There

should be no confusion in this situation, however, since gamma for amosite overlaps

only with alpha for crocidolite. The data in Figure 1 are plotted as the averages of a

considerab? number of values for individual mine samples previously published’. It

is interesting to look a little bit more closely at this variation in dispersion st:lYning 4 y-g VA

data& mine to mine and Table I lists the matching wnvelc~~g~l~, hO, for gamma parsl- (

lel to the fiber lcngthjand alp2ln~erpenc~i~ul~ to the fiber 1cngth)for a group of more 1 I

than thirty asbestos samples from different p&arts of the world. There is some vcmiation

from sample to sample, indicating composition variations. However, all of the data t* re

in the same characteristic chrysotile region and show no overlap with xny of the fibrous

amphiboles.

Since dispersion staining is a relatively new technique requirjlig 8 certain

amount of skill, not. only in rending the matching wavelengths but also in adjusting the

microscope for best clisporsion staining colors, it socms well to summarize a few of

the common tlifficultie:;.

1. The rcfr:lctive indicc:s given in dispersion staining tables are not dispersion

data for that compound. To illust.rate this \vc c:In cite the data for the w inks of CJU:LI+, ,

-2-

I

486 nm 589 nm G5G nm

true refractive indicts, (3 1.550 1.544 1.542 dispersion staining data 1.538 1.544 1.547

The value 1.538 for w at 48G nm is nD for the Cargille liquid that matches quartz o at

486 nm. The acI~~al refractive index of that liquid at 486 nm is 1.550, the same as

quartz w. This operation, which simplifjcs the analytical procedure, is used for all

dispersion staining data. One can, of course, calculate the true refractive indices of

any substance from the dispersion staining data. The necessary data to do this can be

found in the table of dispersion of refractive index data for the Cargillc liq16~is.

2 e True refractive index data and dispersion staining data are id;, + ical at

589 nm, hence refractive i.ndex data at 589 run for any subs!nnce becomes dispersjt :

staining data for that substance. Chryso’-,n,ryl, for example, does not appear in the

dispersion staining tables but \..?nchel12 gives nD = 1.746 (a), 1.748 (p) and 7.5G (3’). %c

From these data one would n~ount a suspwted chrysobcryl in Cargille liquic!l.750 and

expect to see nnnulzr stop colors ranging from orange (~3 to greenish-yellow (^)‘) or

greenish-bl<le (0) to mngcntn-blue (y) with t,he central stop. This greatly ex%ends the

usefulness of dispersion staining.

3. Wc are often a&cd if the dispersion staining objective can be supplied with

a higher ma:;!lificnfion. This gives us the opportunity to point out that hi; ‘ser mxgnifj-

cation is not- dcsir:lble. One is trJ5ng to “resolve” color of the p:wticZes not the par-

ticks thcmscl~cs. Central stop dispersion staining is a darkficlcl p~~occdurc, hence a

lim it of tIctccIion is n function of light intcllsity and contrast.

- 3 -

4. There are a number of points of tcchniquc which greatly improve the

sensitivity of the disl, sion staining procedure, Thcsc include:

a.

b.

C.

d.

e.

f.

h.

The pxrticles must be well separated in the mounting liquid since nonstaincd particles close to or overlapping stained asbestos can mask their presence.

The dispersion stainin:, colors ?or chrysotile and the f&?&*5 amphiboles are more lzilliant if one uses the high dispersion Cnrgille set of refractive index li(,iddS.

In spite of the above injunction concerning high magnification, it is somctimcs useful to use a 20-25X ocular with the 10X dispersion staining objective.

The optics for the dip ersion staining objective, the axis of stage rotation, the substage qortures and lenses must be well aligned on the same optical ‘axis. It is a good idea to take special pains to :,lign the optical system and to maintain that microscope for dispersion staining examination only,

The probl<~m of glare from other p‘xticles in the field of view is solved to a great extent by j.aving a centered and nearly closed field clinpl~~~~~m in the optical system. This concentrate:: attention on particles in the center of the field and climinnt.cs well over 90% of the glare which makes it difffjcult to see very fine asbestos fibers.

The slicles :znd cover slips used for dispersion staining preps should be unusunlly clean since any optic::1 cliscontiuui!.jcs on my surface of the prep cause J glnrc md inte~~fcrc:~wjth visibility 01 the tlispm‘sion staining colors,

-4-

The color plates show the gcncr;ll nature of the dispersion staining colors

and the specific :lppea?:nncc of the various kinds of asbestos in their specific liquids.

l?igurc 2 shows the arrangement of annular am1 central stops in the objective back

focal plant. Thcsc may be a ccntcred 2-3 mm opening in any opaque film or a 3-4

mm dot of India ink on an 18 mm cover slip, respectively; although a dispersion

staining objective is avaiI:J,l, commercially . Note that the substage iris is closed to

allow only an axial beam of light to strike the object. The color series obtained with

each stop are shown in Figure 3. Chrysotile is shown in Figures 4-5 mounted in two

different standard Cargille refractive index media. Polarized light is used with nn

E-W vi& :ltion direction for Figure 4 and N-S for Figure 5. Observed A0 v~lucs arc

given in Table II,

Table II. Matching Wavclcngths for Cl:rysot.:’ ,

Refractive index, Parallel to

“D’ of Crugilk Iiquid fiber length

1.550 500 1.5GO 610

0osswise of fiber

570 640

l?ig-urc:.; 6-8 show anthophyllite, all with E-TV polar. Figur(.s G ‘and 7 show the

fibers mounlcd in 1 .GlO :tnd 1 . G20, respcctivcly . Figure S shows two diffcl-i?nt antho-

phyllitcs (Maryland and North CnrolintL) mounted in 1. G27. This figure, like thz ot,hcrs,

are double cxposurcs with the stgc rotated 90” between exposures, The two different

- 5 -

. ’ .

I . .

CONCLIJS!ON -

The combination of particle morphology and optics uniyucly idcntifics an>’ of

the fibrous asbestos compounds. The most rapid method for obtaining this informati~ ,

is throug!1 the use of dispersion staining. I+# ;xrly cnrric:d out, the dispersion stain-

ing method is capable to sensitivity in the ppm range.

REFERI’KCJ’S --

1. Julian, Y ahd W. C. McCrone, Microscope 18, l-10 (1970)

2. Winchell, A.N. and H. Winchell, Microscopical Charact!?rs of Artificial Inorganic Solid Substances, Academic Press, 19G4, p. ‘78

-G-

Table I

Matching Wavclcngth (A,) in II, D. 1,550 Liquid

Quchec; Lake Asbestos 510 610 Quchcc; King, Asbestos Corp. 510 61.0 Quebec; Asbestos Corp. 500 610 Quebec; Bell Mines 510 600 Qucbcc; ~Iohnsons 500 600 Quebec; Carcys, Bradford 480 590 Quebec; Flintkote 500 610 Quebec; Normandie 570 610 Ontario; Reeves 480 590 Ontario; Mwnro 560 610 Vermont; Hyde Park, GAF . 510 620 Vermont; Jcffery 500 580 New Foundland; Advocate 510 610 New Foundland; 590 620 Yukon; Clinton Creek 500 580 British Columbia; Cassiar 500 580 California; Pacific Asbestos Corp. 480 610 California; Coalings 590 630 Arizona; 600 620 Venezuela; 610 680 Rhodesian; 520 (460)* 620 (550) Rhodesian; Shabinn 480 580 Rhodesian; II:~vclocl~ C and G 490 590 Rhodesian; IIaveloclt IIVL 490 590 Rhodesian; IIavelocI~ VRA 500 630 Cyprus; 600 GGO Greece; Zandini 580 G20 Yugoslavia; 520 590 Italian; Balengera 500 (4CO) GO0 (510)" Russian; 500 GO0 Australia; Woodsrcef 610 680

rr lg 1 k

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Figure Captions

Ashcstos dispersion staining curves.

Schcmalic :wrangFment for annular and central stop dispersion staining. .

Color series for annular and ccntrnl stop dispersion staining.

Chrysotile in Cargille liquid nD = 1.550.

Ckrysotile in Cargille liquid nD = 1.560.

Anthofihyllite in Cargille liquid nD = 1.610.

Antbophyllitc in Car@& liquid nD = 1.620.

Two differs t anthophyllites in Cargille liquid nD := 1. G27.

Amositc in Cargille liquid nD = 1.670.

Amosite in Cargille liquid nl) = 1.680.

Amosite in Cargille liquid nD = 1 .G90.

Crocidolitc in Cargillc liquid nD = 1.700 .

One magenta chrysotile fiber in talc mounted in Cargille liquid nD = 1.555.

i

--- ----

“T I.“&”

L l-l- --

-- l

I I

I I