British Journal of Industrial Medicine 1981 ;38 :389-393

A technique to prepare asbestos air samples forlight and electron microscopyT W S PANG AND ANN E ROBINSON

From the Ontario Ministry ofLabour Occupational Health Branch, Toronto, Ontario, Canada M7A JT7

ABSTRACT The direct transfer technique used in preparing waterborne asbestos for analysis bytransmission electron microscopy is also suitable for preparing airborne asbestos collected on

polycarbonate filters for light microscopy, scanning, and transmission electron miscroscopy. Thesame area and even the same fibre may be examined by the three microscopic methods and individualfibres identified by optical and electron-optical techniques and by x-ray microanalysis.

Light microscopy (LM)l scanning electron micro-scopy (SEM),2 and transmission electron micro-scopy (TEM)3 4 are currently used to determine theconcentrations of asbestos in the work place and inthe environment. The LM method is still used tomonitor asbestos in the work place despite itslimitations5 6 because of the low cost per analysisand the vast amount of asbestos exposure data whichhave been collected.

Recently, Chatfield5 correlated the fibre counts asobtained by LM and TEM. There may have beensome overestimation by TEM, however, becauseultrasound was used in preparing the samples. Thedirect transfer technique4 described here does nothave this drawback and has been found to besuitable for preparing asbestos samples collected onpolycarbonate filters for LM, SEM, TEM, and the

particulate losses during sample preparation arenegligible.5 In fact, the same area or even the samefibre may be examined by the three microscopicmethods, and individual fibres may be identified byoptical and electron-optical techniques such aspolarised light microscopy dispersion staining,electron diffraction, and x-ray microanalysis.6Thus it becomes feasible to check the precision andaccuracy of the LM method and provide correlationsof the fibre counts as obtained by the three micro-scopic methods.

Sample preparation

The direct transfer method entails filtering a knownvolume of air through a polycarbonate filter. A thincoating of carbon, roughly 50 nm thick, is deposited

Filter section

'Petri dish

Fig 1 Jaffe-wick washer assembly.

Stack off i[ter papers

Received 1 September 1980Accepted 13 March 1981

389

on June 20, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.38.4.389 on 1 N

ovember 1981. D

ownloaded from

Panig and Robinison

Fig. 2Figs 2 and 3 LM micrographs of asbestos fibres in two grid openings of a carbon-coated TEM grid.

by vacuumevaporation on to the particulatematerialon a filter. A small section of the coated filter isplaced with the carbon side facing down on a TEMgrid in a Jaffe-wick washer containing chloroform(fig 1). The filter is dissolved in four to six hours, andthe particulate matter is left partially encapsulated on

the carbon film.Since the thin carbon film is isotropic and trans-

parent to light and electrons, it is possible to examinethe same particulate matter both by LM and TEM.The carbon film is also a good conducting mediumfor heat and electrons, therefore particulate matter

Fig. 4 Fig. 5Figs 4 and 5 TEM micrograph of asbestos fibres in part of the grid openings shown in figs 2 and 3.

390

on June 20, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.38.4.389 on 1 N

ovember 1981. D

ownloaded from

A techniique to prepare asbestos air samples for light and electron microscopy

Fig. 6

i::S ~~~~~~10'UMFig. 7Figs 6 and 7 SEM micrographs of asbestos fibres ingrid openings shown in figs 2 and 3.

may be examined also by SEM. Figures 2-7 showasbestos fibres in the same grid openings as observedusing the three microscopic techniques.

Optical microscopy

In the light microscopy method of the National Insti-tute for Occupational Safety and Health the asbestosfibres are fixed on to the chemically cleared cellulosefilter, and it is not possible to identify the asbestosfibres by optical techniques. Although a low tem-perature ashing step to remove the filter has beensuggested by other researchers,7 fibres may be lostduring ashing. In the direct transfer method thegrid is suitable for fibre enumeration by phasecontrast illumination, and additional optical tech-

niques may also be used to identify the asbestosfibres.

In the optical examination the grid is first im-mersed in an appropriate refractive liquid on a glassslide. A glass cover slip is placed carefully on top toavoid trapping air bubbles. The grid is then examinedat a magnification of about 50 x to check foruniformity of particulate deposition, breakage ofcarbon film, and the presence of undissolved filtermaterial. If the grid is found to be acceptable thegrid openings with intact carbon film are chosen in apredesignated or random manner for analysis. Tenor more grid openings from each of the five or more

grids prepared from the same filter are analysed at a

magnification of 450 x to provide an estimate of theconcentration of the asbestos fibres.The centre of the grid has a pattern that is used to

define the position of each grid opening so that thesame particle may be relocated for further examina-tion.

Figure 8 shows the carbon film of a grid openingobserved by phase contrast illumination. In additionto providing support for the particulate matter, thefilm replicates the pores of the surface of the poly-carbonate filter. These pores, which are roughly 0 45,Lm in diameter, are clearly resolved in this micro-graph, taken originally at a magnification of about150 x . The resolution and the quality of the imageprovided by the carbon support is comparable withthat obtained from the chemically cleared cellulosemembrane filter. The asbestos fibres, however, are

even more easily distinguished from the carbon filmbackground if a polycarbonate filter with 0 2 ,tmpore size is used, as in figs 2 and 3.

inPPUIEEEE' ias

Fig 8 LM micrograph showing pores in carbon filmas result of replicating surface ofpolycarbonate filterwith pore diameter of 0 45 lim. (Original micrographtaken at 150 x .)

391

on June 20, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.38.4.389 on 1 N

ovember 1981. D

ownloaded from

392

Identification of asbestos fibres

The particulate matter is partially encapsulated bythe carbon film so that the grid may be immersedseveral times in different refractive liquids to deter-mine the refractive indices of the asbestos fibreswithout loss of fibres. After each immersion, therefractive index liquid may be removed by placing thegrid back in the Jaffe-wick washer containingchloroform.

Other optical techniques, such as polarised lightmicroscopy and dispersion staining, may also beused to identify individual asbestos fibres. Forexample, crocidolite shows low relief or disappearswhen immersed in a refractive liquid of 1-69 andshows distinct dispersion colours when it is immersedin appropriate refractive liquids.7 Also, it is the onlyasbestos that is length fast-that is, the lowestrefractive index lies parallel to the length of thefibre. Therefore, under polarised light and in the pathof a gypsum compensator, a crocidolite fibre showsa blue colour when it is perpendicular to the slowvibration direction of the compensator and a yellowcolour when parallel to the slow vibration directionof the compensator.While the identification of asbestos fibres by their

dispersion colours or using a gypsum compensatorbecomes much more difficult for fibres less than1 ,um in diameter, some of their optical propertiesmay still be used to distinguish one from another.For instance, chrysotile and crocidolite fibres thatare known to be in a sample may be differentiated byusing a liquid having a refractive index of 1-62.

Fig 9 LM micrograph ofasbestos fibres in gridopening shown in fig 2. Change in contrast of asbestosfibres results from using a refractive liquid with an

index (n = 1-62) higher than that of chrysotile asbestos(n = 1J56).

Pang and Robinson

Chrysotile fibres show whitish (no colour) in phasecontrast illumination (fig 9) and crocidolite fibresshow dark grey to greenish colour. Thus the twotypes of asbestos fibres may be distinguished andtheir concentrations determined separately.

Electron microscopy

Figures 2-7 show the same area of the same gridopening as examined by light, scanning, and electronmicroscopy and indicate that the TEM gives thebest resolution. The identity of fibres previouslycharacterised by optical techniques may now beconfirmed by morphology, electron diffractionpatterns, and x-ray microanalysis. In addition smallasbestos fibres below the resolution of the lightmicroscope are shown.

In choosing the grid opening instead of randomfield of view as a unit of area for examination thedirect transfer method permits examination of thesame area or even of the same fibre by both light andelectron microscopy. Thus it becomes practicable todetermine both the precision and accuracy of thefibre identification and to compare enumeration bythe two methods. Evaluation of the variations orbias in fibre counting results obtained betweenanalysts or laboratories now becomes meaningful.This preparative technique provides also for a directcomparison of fibre enumeration on the same sampleby light, scanning, and transmission electron micro-scopy which have different limits of resolution.

Gentry and Spurny8 determined that the asbestoscollection efficiency of the polycarbonate filter mayvary from 0-3 to 0-7 depending on the size distri-bution of the airborne asbestos. They also showedthat the amosite fibres are able to penetrate thepolycarbonate filter with pore diameter less than20% of the fibre length. Therefore, the directtransfer technique described above may under-estimate the asbestos fibres below the I to 2 ,tm inlength and 0 2 ,um in diameter when a 0-2 ,tm poresize polycarbonate filter is used in sampling.

Conclusion

(1) The direct transfer method permits preparationof samples of airborne asbestos suitable for analysisby both light and electron microscopy. The reso-lution of small particles on the carbon film by lightmicroscopy is satisfactory, and additional opticaltechniques such as polarised light microscopy anddispersion staining may be used to identify theasbestos fibres.

(2) By analysing the same grid openings or thesame asbestos fibres by light microscopy as well aselectron microscopy, it is possible to check the

on June 20, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.38.4.389 on 1 N

ovember 1981. D

ownloaded from

A technique to prepare asbestos air samples for light and electron microscopy

accuracy and precision of the asbestos resultsobtained by the light microscopic method.

(3) Correlations between fibre enumeration resultsobtained by light and electron microscopy may beestablished so that it may be justified to use thelight microscopic method as a screening procedureto assess concentrations of airborne asbestos inworking environments.

We thank Mr Eric Lin of the department ofzoology of the University of Toronto for taking theSEM micrographs and Ms Wendy Dicker for herhelp in preparing the manuscript.

References

Liedel NA, Bayer SG, Zumalde RP. United States PublicHealth Service/National Institute for Occupational Safetyand Health membrane filter method for evaluatingairborne asbestos fibers. Cincinnati, Ohio: US Depart-ment of Health Education and Welfare, NIOSH, 1973.

2 Chatfield EJ, Dillon MJ. Some aspects of specimen prepara-tion and limitations of precision in particulate analysisby SEM and TEM. In: Johari 0, ed. Scanning electronmicroscopy. Chicago, Illinois: SEM Inc, 1978:487-96.

3Anderson CH, Long J M. Preliminary interim procedure forfibrous asbestos. Athens, Georgia: US EnvironmentalProtection Agency, 1976.

4Ontario Ministry of the Environment. An interim methodfor the determination of asbestos fibre concentrations inwater by transmission electron microscopy. Toronto,Canada: Ontario Ministry of the Environment, 1977.

Chatfield EJ. Measurement of asbestos fibres in the work-place and in the general environment. In: Ledoux RL, ed.Short course in mineralogical techniques of asbestosdetermination. Quebec: Mineralogical Association ofCanada, 1979:111-63.

6 Champness PE, Cliff G, Lorimer GW. The identification ofasbestos. J Microsc 1976;108:231-49.

7Ganotes JT, Tan HT. Asbestos identification by dispersionstaining microscopy. Am Ind Hyg Assoc J1980;41 :70-3.

8 Gentry JW, Spurny KR. Measurements of collectionefficiency of nuclepore filters for asbestos fibres. Journalof Colloid and Interface Science 1978 ;65:174-80.

393

on June 20, 2020 by guest. Protected by copyright.

http://oem.bm

j.com/

Br J Ind M

ed: first published as 10.1136/oem.38.4.389 on 1 N

ovember 1981. D

ownloaded from