Document gD8mg4JRX6Ee00YvBjvkrQ6jq

FILE NAME: RT Vanderbilt (RTV) DATE: 1989 DOC#: RTV082 DOCUMENT DESCRIPTION: Journal Article - The Regulatory and Mineralogical Definitions of Asbestos and Their Impact on Amphibole Dust Analysis Am. Ind. H>$. Assoc. J. 5t![l I j i u ) . 2211989) 4008 The Regulatory- and Mineralogical Definitions of Asbestos and Their Impact on Amphibole Dust Analysis JOHN W KKI.SF. and C- $HfU.f>f>N 1HOMPSON K.T. Vanderbilt Company. Inc , 30 Winlifld Street. Norwalk. CT 06S5i AHhoueha familiar occupaiiomil health lypic.ihciermfliftfifoieeriirilly Is not well understood, Sfemficant differences between mineealof. leal and regulatory definitions Sustain the confusion. Definitional ambiguity is addressed and itscffcrl upon the characterization of New York State Iremolilic (ale are investigated. Analysis oi asbestiform and nonasbesliform airborne dust populations clearly demonstrates the nonspecificityof the regulatorydefinition and fhc3;l aspect ratio "fiber" counting scheme. Shifting to a hightr aspect n tio would reduce false positives radically without a loss in sensitivity for true asbestos, Any change in aspect ratio, however, must be accompanied by a mineralogieslly correct definition of asbestos if proper mineral characterization is to be assured. Introduction its long, thin, strong, flexible fibers. Accordingly, to a min Few environmental health hazards have been as widely pub licized or viewed with a's much dread as asbestos. Despite eral scientist the term asbestos always includes some refer ence to the fibrous crystal growth pattern often described as Ibis attention, considerable confusion exists as to what the thO sbesiiform habit,- Mineraloglcally, asbestos is a mat generic term asbestos actually means. American regulatory ter of how a mineral grows, not simply a matter of one definitions arc incomplete and. in some instances, at odds mineral versus another or an arbitrary dimensional concept. with the mineralogical view of this substance. The purpose of this paper is to review this definitional problem and Several minerals, including those designated in United Stales' regulations, do grow in nature in an asbestiform demonstrate its effect on one controversial dusl envi ronmem. . habit. These would include the most commonly exploited forms of asbestos: ehrysottle, crocidoliie, and amosite. The Definitions Hogulatory The National Institute for Occupational Safety and Health fNlOSH) has established the definitions and analysis methods for asbestos used by almost a[( regulatory bodies in the United States. Under this scheme, asbestos is defined as any fiber of chrysotile.crocidolite, amosile, anthophyllitc, tremolite or actinolite. A fiber is defined as a particle with a length to width ratio (aspect ratio) of at least 3 :1and a length of 5 jt m or more as determined by the phase-contrast optical microscope (PCM) at a magnification of 450X to 500X.tl) While NlO SH acknowledges that this dimensional criteria and fiber counting method is not specific to asbestos* regula lory def/ni tions offer no further deserip tion of what is or is not asbestos. regulated asbestiform minerals, however, also occur in nature in a nonasbestiform habil. In alj cases, ihenonasbestiform habil is by far the mote common. Table 1 lists the asbestiform and nonasbestiform habits of the six regulated minerals and their separate Chemical Abstract Service numbers.The list conforms to the nomcndaturesei forth by the United States Department of the Interior.* It should be noted that the chemical composition is the same for each mineral in either growth habit. In all cases ex cept chrysolite, the internal crystal structure is identical as well. Also,Ihefirsuhreeminerals havebccnassigncd separate names to distinguish the different growth patterns, while the last three--anihophyilue. tremolite, and actinolite--have not. For these three the nonasbestiform analogs are com mon rock-forming minerals found throughout the earth's crust and, therefore, routinely encountered in many indus- Mineralgica! In the Glossary o f Geology, asbestos is defined simply as A commercial term applied to a group or highly fibrous silicate minerals that readily separate into long, thin, strong fibers of sufficient ticxibil* ity to be woven,are heal resistant and chemically inert, and possess a high electrical insulation and therefore are suitable for uses wr*re incomb usti ble, nonconductive or chemically resistant mate rial Is required.1'" While them cal and electrical n tncss arc propet s shared by almost all silicates, asbestos is unique because of iries. Figure 1 graphically depicts the basic difference in the two mineral growth patterns while Figure 2 contrasts the two macroscopically and microscopically. While nonasbestiform particles clearly differ from asbesiiform particles, e*->y would be counted as asbestos under the current ."-gutato/y.):! dimensional criterion lor a fiber when an ore is crushed, mi'led or otherwise reduced. Thus, while all asK -,'os is fibrou:. not all fibers are asbestos. It is also important to note the t isbcsiiform fibers cannot be created Iroronun&i'.estiforn. n.. if rials by crushing, milling, or grind ing */*- i.opieally. a |-aniele with an aspect ratio of 3:1 would not be considered a liber. Because the lermfther is ti ter; reted indifferent nays, its use m this papei will tic reside cd Cw>y5i*'W? /.<*>cai InCuih.gtHfi |>Afft'' Am tot) Jifia two: j Novfjcferr IOCS MJ WES000374 TABLE I . Asbeslitorm and Nonasbeslitorm Varieties of S elected Silicate Minerals and Their Chemical A bstract Service N um bers (CAS) AsbcsMorm Variety (CAS K) Chemical C om position Nonasbestilorm Variety (CAS tf) Serpentine Group Chrysolite (l?OOl-29-5) MgatSLOsHOH). ntigorue, liiaiflile (12135-06-3) Amphibole Group: CrocidOlilc (12001-28-ff Gruente asbestos (amosite) (1217?-73S')A Anthophyllite asbestos (77S3G-C7-5') Tremollle asbestos (77506-60-6'i Actinoiite asbestos (77S36-6S-4-) Na,rejrei(SI,0){OH.F)j (Mg.Fe)riSiaOKOH.F), (M9.FeWSi0](OH.F)> CayMgsfSijOjiKOH.F,) Ca,|Mg.Fe)5|Si0.)(OH.F);, ribackile (177B7-87-0) eurflmingionile-gruneriie (1 6 7 - 6 1 -4) amhophyittte (17060-76-3) trem olile (14567-73-8) ` actinotite (13768-OO-B) "'The pres once ol an asterisk following a CAS Registry Number indicates that me registra tion is (or ft substance which CAS floes nol treat in ils regular CA Index processing as a unique chemical entity. Typically, this occurs when the material is one ol variable compo sition: a biological orgams m. a botanical entity, an oil orexlract ol plant or animal origin, or a malnal that Includes some description ol physical specificity, such os morphology. in the interest of clarity to specific definitions only. To reflect '' the miner logical characteristics of asbestos in a definition, a group of mineral scientists agreed to the following.101 A. Asbcstos-~A collective mineralogica) term that de scribes certain silicates belonging to the serpentine and amphibole mineral groups, which'have Crystal lized in theasbestiform habif causing them lobe easily separated into long, litio, flexible, strong fibers when crushed or processed. Included in the definition arc chrysolite: crocidoliie.asbesliformgfuneritc(amosite); anthophyiliteasbestos; tremolile asbestos: and actinolite asbestos. B. Asbestos Fibers--Asbcstiform mineral fiber popula tions generally have the following characteristics when viewed by light microscopy. 1. Many particles wiihaspeci ratios ranging from 20:1 to 100:1 or higher (> 5 pm length) 2. Very thin fibrils genetally less than 0.5 pm in width, and 3. In addition to the mandatory fibrillar crystal growth, two or more ol the following attributes: (a) Parallel fibers occurring in bundles: tb) Fibers displaying splayed ends: . (c) Matted masses ot individual fibers: and (d) Fibers showing curvature*11 Mnnv if those who contributed to this definition and support n t listed criteria have published extensively on ibr problems associated with the MOSH definitions and the membrane filter m e t h o d T h i s definition has been incorporated in a proposed American Society for Testing and Materials (ASTM) medtod submitted to committee D-22.05 (January 14. 1988). The criteria have long been endorsed by the U.S. Department of the lmerior.rtUJ', While a I) mineral scientists may not agree with every entry in this definition, ii does present a more mineralogically accurate description of asbestos and asbestos fibers than docs the regulBiorydefinition. This is especially truewhenit is applied to a dust population rather than on a particle by particle basis. The definition, therefore, will be used in the remainder of this paper as the" mineralogical"definition or asbestos. It might be noted that the width criterion (0.5 pm) represents a dimension below which all individual "fibrils'* and clumps or masses o fibrils would be encountered in processed asbestos. Unprocessed clumps or masses may exceed this width, bur such particles would not be represen tative of common airborne asbestos libers. The Study Environment One of the most controversial workpla'ce exposures asso ciated with this definitional issue involves the mining and mi.Jing of New York State tremoliric talc. Accordingly, a study was undertaken io contrast dust data obtained it) this environment /i'h both the regulatory and mineralogical dis it sed above. N w Vqi Ic b>aic ircmolnic talc is an industrial grade talc u,,ed exti-sively i,i the ceramics, nie.and paint industries, since 1924 the R.T. Vanderbilt Company. Inc., las owned 1 operated the only \c w Vork State iremolitic talc mine. * >jvfjn&t) )989 We S000375 Talc mined from this operation vanes somewhat in mineral content but an assay of the ore generally reflects40%-60% tremolitc. l% -\0% anthophyllite. 20%-40% ialc,20%-30% serpentine (amigoriie-lizardilc). and 0%-2% quartz.08' The R .T. VahderbiJt Company states that allofthc tremoliteand anihophyllile in its talc products appear only in lire nonasbestiform habit.lISi' In 1980. however, NIOSH pub lished a technical report entitled Occupational Exposure to Talc Containing Asbestor111' specifically addressing this mineral dust exposure. In the report. NIOSH applied its regulatory asbestos definition to bulb and airborne dust samples collected at this mine and reported over 70% asbes tos for airborne fibers satisfying the 3:1 or greater aspect ratio and greater than 5-pro length limit (NIOSH PCM m ethod). Particles were identified as tremolile and anthophyllite by standard X-ray diffraction technique. M ethod of Study Sam ples for particulate analysis were collected on open faced. 37-mm diameter Milliporc type AA filters (0.8-pm pore size. Milliporc Corp.. Bedford, Mass.). Precalibrated M ine S afety Appliances'Modcl G pumps were used tod raw a ir through these filters at a rate of 1.7 L/min. Although fiber sampling technique has changed since, (his technique was used in order to compare results with data previously collected. Filters were changed throughout a full work shift ASBESTIFORM In th e a s b e s lif o rm habit, m ineral c ry stals g ro w in a s in g le d im e n sio n , in a straight line until they form long, th r e a d -lik e fib ers with asp e c t ratios ol 20:1 lo 1000:1 a n d h ig h e r . W h en p re s su re Is ap p lied , th e fib ers d o n o t s h a tte r b u t sim ply bend m uch like a wire. Fibrils of a sm a lle r d iam eter are produced as bundles of fibers are p u lle d a p a rt. T his bundling effect is referred to as polyfilam entous. NONASBESTIFORM In th e n o n a s b e s tifo rm variety, cry stal g ro w th is ra n d o m , form ing m ullldim enslonal prism atic patterns. W h e n p re s s u re is applied, the crystal traclu res easily, frag m en tin g into prism atic parm es. Som e of the par ticles or cleavage fragm ents are acicuiar or needle s h a p e d a s a result ol the tendency ol am phibole m inerals to cleave alo n g two dln,e" sions but not along the third. S ta ir-s te p clea v ag e along the e d g e s of som e particles is c o m m o n , and Oblique extinction is exhibited under the m ic ro sc o p e . C leavage fragm ents never show curvolure. Figure 1--Asbesliform and n o n asb esh lo rm graphics m tnij #t*f *5'C J (50) HBvnnbct as needed to pievem overloading. In all, 22 air samples were Obiained representing nine work activities in the R.T. Vanderbilt Co., GoUverncur, New York, mine and mill. Work activilies sampled included milling (H ardinge and WhceJeV mills), drying, packing, bag stacking, crushing, mine drilling, scraping, and tramming. Analyses were performed by The R.J. Lee Group, Inc., of Monroeville. Pennsylvania (Project No. 86-12318). Analytical techniques employed included phase contrast microscopy (PCM).polarized light microscopy(PLM),scanningtleciron microscopy (SF.M), computer-controlled scanning electron microscopy(CCSEM),and transmission electron microscopy (TEM). In accordance with NIOSH method 7400, all sam ples received PCM particle counts at400X magnification in Walton-Bcckett graticule measuring at least S-pm long with a 3; I or grcatcraspcct ratio. Beyond these specified parame ters, exact particle widths and lengths were not measured. For each sample, 100 fields or 100 particles, whichever came first, were counted (with a minimum of20 fields). In all, 2295 particles were counted and sized by PCM. A separate wedge was cut from each filter fo r PLM analy sis. Particles were lapped, then gently scraped from the wedge to a glass slide. Any remaining particles were cap tured by rolling a needle moistened with 1,592 refractive index (R l) liquid over the surface of the filter wedge (Rl selected for low-iron talc). Additional 1.592 Rl liquid was added to thesJideand used to wash particles from the needle onto the slide. It should be noted that this transfer technique could bias the PCM analysis if very fine particles were lost In the transfer. Additional analysis of particles not removed from the filter (another filter section) suggests such bias is unlikely for tremolile (see SEM partible width discussion below). PLM counts were made in a 1,592 RJ oil to differen tiate talc from allamphiboles on all 22 air samples. Follow ing this basic cut, tremolitc was differentiated from anthophyllite by angle of extinction (tremolitc has an inclined extinction of 14 to 17, whilcanthophyllitc exhibits parallel extinction). Since all asbestos exhibits parallel extinction, mineral habit (asbesliform or nonasbestiform) then was decided on the basts of criteria noted in the mineralogical definition. Depending on particle concentration for each of thc22 samples, 100 to 200 points were counted and charac terized at 100X magnification, yielding a minimum of 2200 panicles subjected to PLM analysis. If positive particle iden tification could not be made at IOOX total magnification, higher magnifications (up lo400X) were applied on a parti cle by particle basis. As in the PCM analysts, only particles with an aspect ratio of 3:1or greater and a length of 5 pm or more were so characterized. Although exact length and width measurements were not obtained, particles were sized by basic aspect ratio catcgories(/.e., thosc3:l orgrealer. 10:1 or greater, ere-). One additional step was taken in the PLM analysis in which panicles presumed to be anthophyllitc(> I.S92 RI) w -re tested for "transitional" phases (meaning talc eriwincd w .th or evolving from anihophyllitc an d /o r bio p; libles). Vhi-; was accom plished by finding pan icles which most d c .-'lv ap p ro : 'm ated ihexam esi; ' tnd rm rphological characteristics of these suspect purndrS on unrulier pornon WES000376 EXAMPLES Amphiboles with Separate Names: RAW ORE Amphiboles with the Same Name: EXAMPLES Amphiboles with Separate Names; MICROSCOPIC 265X Magnification, 2.75jum/Divi$ion Amphiboles with the Same Name: Irem olfe ASBESTIFORM trem olile NONASBESTIFORM nt irrt f<\) h,, I 150t M: bSil WES000377 of the filler and testing them at 1.608 Rl (the low gamma index lor anthophylliie). Because of problems inherent in this technique, testing the same pani'clc with different Rl liquids was not possible. Particles with an index of refraction between 1.592 and 1.608 were classed a s `'transitional.'*In all, 6 samples underwent this additional analysis. To lest further the differences and similarities between asbestifortn dust populations and the tremolitic tale dust environment, electron microscopy was employed on 5 sam ples most representative of common mine and mill exposures(e.,, product packaging).SUM with energy dispersive X-ray (EDX) first required the mounting of another 1/8 niter wedge from each sample on a carbon-coated stub. Fifty fields at 2000X magnification then were analyzed for count, size, and identity of all particles in every field with an aspect ratio greater than 3:1and a length greater than 5 ^m- For the five fillers, a total of 183 panicles were characterized in this way. Particles below and above a width of 0.25 pm Were noted as well. This width was selected primarilybecause it is used in references against which the findings of this study shall be com pared.1 These references generally refer to this width as the approximate lower resolution limit of the light microscope.'/'',, While other references report lower width scnslvly,lM,6, it generally is agreed this lower lirrtil varies with the quality of the microscope, use of dispersion staining and background contrast, magnification, and the microscopisL involved. CCSEM with EDX was used on the sam e carbon-coated filter wedges to scan a total of 2500 particles (500 per sample) at magnifications of 35X, I00X, a n d 50OX. Particles were sized by the preselected parame ters, and the chemical composition of all particles was noted. Particle distribution was expressed in volume percent and all trcmolite particles w'ir' counted. TEM with selected area electron diffraction (SAFO) also was employed on new carbon-coated filter wedges front the same five filters. Chem ical composition by EDX analysis and SAED patterns of individual fibers which measured 10 pm or greater on four grid squares per wedge were obtained a/lcr the filter matrix wasdissolved from the carbon film. While considerable data were lhus generated from this multiple analytical approach, only data summaries which directly address the definitional comparison are included in this paper. It should be noted lltai the EDX chemistries obtained through the CCSEM analysis and the SAED patterns obtained through TEM analysis were not adequate to distin guish talcand anlhophyllite. While an in-depth discussion of this problem is beyond the scope of this paper, in summary it should be said that talc may present the same X-ray spec trum as anthophylliie because talc displays a similar2:1 Si/ Mg ratio and overlapping range. KcgardingSAED patterns, talc in the fibrous form often reflects he ame 5.3 A spacing as anlhophyllite. Talc/anthophylliie in an intermediate or transitional phase poses further identification problems when electron diffraction analysis is restricted to one point per panicle. This is more fully described in other papers.<2,'2i) Study Results and Definitional Comparison Table II contrasts bulk trcmolite asbestos particles described in the literature118*to iremolile panicles reflected on five New Samples TABLE 11 Balio C om parison of Bulk Trem olile A sbestos* to N.Y. S tale Tremolile In Five Air S am p les'1by Optical and E lectron M icroscopy Ratio ol Tremolile Pal tides 3:1 aspect ratio (a.r.) or Greater to Total Tremolite O 5 yrn length) SEMC 10:1 a.r. or Greater to Total Tremolite > S pm L) SEM 20:1 a.r. or Greater to-Toial Tremolile |> 5 f is L ) SEM Tremolilc asbestos" i in t.6 1 in 2.6 . 1 In 4.6 TromolHe asbestos" tf total tremolite panic'es per sample (all sizes): 200 1 Ip 1,8 (approx. 55%) 1 in 2.3 (approx. 41%) 1 In 2.S (approx. 31%) 10:1 a.r. or Greater 103:1 a.r. or Greater Oplu SEM 1 in l.fr 1 in 1.6 1 >n 1.6 i in 1,2 1 in 1.3 (66%) Tremolite in 5 N.Y. air samples" a total iremolile particles {all sizes): 949 I in 6.2 (16%) CCSEM 1 1n 949 or greater (0.1%) CCSEM 0 in 949 (0%) CCSEM Opt. 1 in 101 or greater CCSEM li n 152 or greater 1 in 146 f gfcoier !0 6% ) 'D ata (rom U.S. D cpl. ol Interior. B ureau ol M ines R epori ol Investigation 5367, p a g e 13. T able 2 (1979)." -'' " P re s e n t study: CCSEM an aly sis o( 5 aif sa m p les ai 35X. MPT. a n d 6Q0X m i g n lllcalio n s. {2500 iota) parlicle count (all sizes)). O ptical IPC-M a n d PLM) and),-vi ol the sa m e 5 sam ples up to 400X ma9n>l>cationS 5 3 1 total p a rtic le s with a 3:1 a r or greater > 5 pm length). ' P an icles counted using SFM with m agntlicaiion up to en '- V ' " P a n ic le s c o u n te d using n p ucaH igni m icroscopy at 250X m m nlllcation (2C0 irem olile particles counted per n ite/ ' O O lained lio m C ahlorm a n o .'Ih er d e se n p b o n ol lit. r* tu .e(. W iic, ." 'led. "O b ta in e d lio m m useum sam ple (torn flaiasihan, Ind a 'Viley milled Am tod v; - lit 1 ODI w>vci|>r> i??,5 II) WES000378 TABLE III Average ol 22 Mine and Mill Air S am ples (2295 Particles) by C om position, Aspect Ratio 3:1 or G reater (> 5 fim length), a n d Mineral Habit by Light Microscopy**l Aspect Ratio: 3:1-10:1 Aof Total >10:1-20:1 > 20;i Particles per CC (TWA) 3:1-10:1 > 10:1-20:1 >20:1 Total Particles pet CC (9-hr TWA) % Asbestiform by Mincralogical Del. Trcmoiite 35.a 33 0 .45 .009 0 0.459 0 Transitional11 0.0 .76 0 0.00 .015 0 0.0)5 0 Tate SB.Z 4 60 0 .67 053 0 0 728 0 All particles 93.0 7.00 0 1.1? 0.002 0 1.210 0 *Mincral type and % by aspect ratio.were obtained by PLM analysis at 100X to 400X magnification. Total particles per cc were obtained by PCM at 400X magnification, l'%Ta Ic/anlhophyllile transl llonal particles were extrapolated from SdI 22 air samples based on a relractive index beiween 1.592 and 1.609 lor me gamma index. No pure amhophyllite particles were noled in me lieid analysed. York state tremolilic talc air samples by both optica! and electron microscopy. In this comparison, the ratio of tremo- lite particles which satisfy ihe regulatory definition of a fiber (3:1 o r greater aspect ratio, > 5 pm length) and those that exceed a 10:1 and 20;I aspecl ratio (> 5 /m length) are addressed. ' O f the 2500 total particles scanned by CCSEM on 5 air samples, 38% or949 were iremolite. Of these iremolite parti cles. 16% or 152 satisfied the regulatory size criteria.for a fiber. This represents a ratio of 1 lrcroolitc regulatory fiber in every 6.2 trcmolite panicles. In contrast, tremotiicasbestos reflected an average of I regulatory size fibey in every 1.7^ panicles (55%). Most striking, however, is (he difference reflected at 10:1 and 20:1 aspect ratios. For the New York state trcmolite, only I iremolite particle in 949 (total counted) exceeded a 10:1 aspect ratio (0.1%). Fortremoiite asbestos this ratio wasapproximately I inevery2.5 particles or 40%. At a 20:1 aspecl ratio or greater* no New York iremolite particles were counted, while I in every 3 (approx imately) were found for iremolite asbestos. Significant vari ation in these ratios was not noted under Optical microscopy Tor the same samples at the magnifications applied. While a bulk to airborne particle comparison is not ideal, the dimensional differences likely would be even greater if two airborne particle distributions were compared, since wider width, lower aspect ratio particles are more common in bulk particle distributions. Published particle distribu tions for airbbrne asbestos dust populations support this contention and support the basic dimensional similarity of iremolite asbestos lo other asbestiform minerals (sec the extended discussion on airborne particle aspect ratio distri butions below). Accordingly, on a iremolite to iremolite basis, an entirely different particle-size distribution would be expected in the New York stale tremolitic talc samples :f 'his iremolite were asbestifotm, Table III reflects the average of all 22 air samples by percent mineral composition, aspect ratio H I or greater), :id crystal growth habit (asbestiform or nonaibcsiiforui;. Results in this table rcllcct the combined appiua.ion of th PCM and PLM methods outlined above In the liclds analyzed by PLM. no parities exceeded a 20:1 a>[n'ci ratio or shout'd splayed ends, curvotuu-. m parallel fibers occurring in bundles. Using the mineralgica! definition, therefore, no asbestos was found: ho wever, 0.459 particles/ec would be noled if the regula lory definition were used (talc and transitional particles excluded). A total of 1.21 particles/cc would be reported if talc and transitional particles were counted. Proper characterization of talc, anchophylliie and transitional particles is extremely difficult in this ore body except by PLM, While PLM air sample data reflect no asbestiform libers, both talc and transiliooal par ticles can appearina fibrous, asbestiform and/or nonasbes. tiform habit in this ore body.07' If misclassified as aothophyllire, these asbestiform fibers would be characterized as asbestos under both the regulatory and mineralgica! defini tions. TEM SAED analysis with multiple electron diffrac tion patternsfeach indexed) confirmed the presence orboth nOnasbestifcrm and asbestiform transitional and fibrous talc particles in a random scan or fields not included in the PLM analysis. No effort to quantify these fibers was made. Because uf the rarity of these fibers and their marginal significance io the definitional distinctions being addressed here, further deiail in this area is beyond the scopeand intent of this paper. Tabic IV reflects a comparison of fiber counts obtained in (his study uiih data previously obtained in the same mine and mill (same or similar work activities). These data con firm'a marked difference in what is reported as asbestos, depending upon the definition used. Note that the average of all regulatory fibers counted by PCM (Column 2) shows far less variancebctwccn invesligatorsihan ihe percent of pani cles considered asbestifotm (Column 5). Mincralogical dis tinctions made reflect consideration of the characteristics described in the mincralogical definition. Although none of the particles in the study dust population exceeded a 20:1 aspect ratio by light microscopy, this factor alone did r n' dictate habit characterization Ior the 22 samples analyzed. Although the lack ol 20:1 aspect uiio particles in a dust population certainly suggests a nonasbestifor.r. dust envi ronment, aspect ratiosaloncare not pivotal toa mineralgi ca! sound definition of asbestos To test definitional specificity further, a comparison nf basic diniemicm.il charactewin'? common to asbestiform dust populations. nonasbesiilormtclcnvageiragmenOamphi- stt ;r -. Ili W e s o o o W '50 Source and Year R. Lee(1908) MSHA (1984-85)' Insurance (19a4)c NIOSH (1975)k Dunn (1902^ TABLE IV Historical Air Sam ples'' by Delinillonal Approach Average ot All Pertides/CC Mill and Mine Range PsrUciM/CC MIH and Mine"111 Definitional Approach %'PartcIes/ Classed as Asbestos t.21 0.14-3.56'" mineralogies! 0.00 2.39 O.M-iMff*" mineralogies! 0.40 1.0 1,36-2. IS"1 not glassed - 4.5 t.5-3.4r" " regulatory 72.00 0.65 0.03-1.38'"1 mincralogical but classifica tion completed on bulk sam ples only Particles/ Considered Asbestos 0.000 0.009u 3312 aAII panicles 3:1 or greater In aspect ratio. > 5 am In length and resolvable under the light microscope. "(n) * number ot air samples. ''Mine Safety and Health Administration Survey Reports dated: 7/17/85.7/30/85,5/22/65.6/12/84,1/9/84. UMSHAperlarms analysis lor liber type only on inters with elevated total liber goon Is. Ol the 38 tillers, 22were so analyzed. 01 these. 2 tillers were repottedas containing 2% asbestilorm libers. All other inters were lound or assumed to con lain 0%. ` 'Hartford Insurance Company Report dated November 1984 to R.T. Vanderbilt Company. Inc. ^NIOSH Technical Report, Occupational and Exposure to Taic Containing Asbestos. Table 7 (1980)1,11 l!Dunn Geoscience Core, report to R.T. Yenderbill Company (1905), bole d u st populations, and the study dust population -was population particle widths obtained by SCM. With regard to undertaken. Figure 3 compares airborne asbestiform and the tremolile found in the talc air samples (the only amphi- nonasbestiforrn particles which Tall above and below a width 'bolc noted), all tremolile particles (88 out of 183 total parti of 0.25 ftm r described in the literature,tM| with study dust >" cles) were wider than 0.25 pm. Particle widths noted in * % 100% % 1O0% TOO too ' BO 80 60 ACLMEPAHVIABGOELE FRAGMENTS' 40 GO N.Y. STATE TREMOLITIC talc 40 TREMOLlTE width 20 0% <025 width >0.25 20 0% <0.25 w id th >0.25 'From: J.G . Snyder. ft L. Virle. and J.M . Sep/PI: "Evaluallon ol the Phase Contrast Microscop" Method lor the detection ot Fibrous and Other Elongation Mineral Parlicuh tes by Comparison with a STEM Teen/uoi'c *Am. tnd. Hyg Assoc J. 4S(5)*.71-477 p9B?) Table IV. AVer ,ge 0117! samples. From: Auersce ol S air sam ples rmolyzed by SEM irepresems SOparticle* oul u> .0? total panicles), F ig u re 3 - A verage airborne p an icle W idti. c u m p a fia -n hy elec Iron m icroscopy (s'<i particles I Or greater asp ect r$no. 5 p m o < m ore length) Am Inti tifi ' i!CC J fiOl Noribc> 19*9 sis WES000380 TABUE V A spect Ratio C om parison Airborne Asbestos Panicles* (Mining and Sagging) > OjtS pm Width, > 5 pm Length Airborne Cleavage Fragments1' (Approx. 4500 Total Particles) > 0.25 pbt Width, > 5 pm Lenglh % 0/ Particles Seen a t V* Ol Parliclei Seen at: Aspect Ratio: 3:1 10:1 15:1 > 2 0 :1 Aspect Ratio: 3;t 10:1 15:1 > zon Crcidoifle 100 100 91.5 64.5 Amoslte 1O0 100 89.> 58.0 Chrysolite too 100 86.0 37.0 Average: lOO 100 89 53 cunrvmingtonUe 10O 2a 10 6 cumminglonile 100 32 7 3 acnnolite 100 . 15 A 3 grunerile/Bcliriolllc 100 6 0 0 tremotltic talc1' 100 7 N D U 0 Average: 100 17 0 2.4 'T a k e n from G.W. G ibbs an d C. Y. H w ung, Dimensions of Airborne Asbestos Fibers, jAfiC Scientific Pub. #30 Lyon, Fiance, pp. 79-65."" *Taken from A.G, Wylie, H.LYirle.ano E.Russok. "Characterizing and Discriminating Airborne Fibers: 1m- plicpiions lorthcNIOSH Method," American Industrial Hygiene Association Jootfiel.Voi. At. pp, 197-201.';' r Dala taken from the R.J.Lee Group Oust Analysis Project prepared for the R.7. Vanderbilt Co., me , 1968. Reflects PCM/PLM analysis ol 22 filters; % represents 2295 total particles. ND = not determined. asbesiiform dust populations by STEM differ markedly, with an average of 35% (ranging from 9% to 81%) reported to fail below a 0.25-vm width."*' The similarity between amphibole cleavage fragment particle width and tremolile widths noted in the study dust population, therefore, sug gests a nonasbestiform habit. It also might be noted that,, since all tremolile particles exceeded a 0.25-pin width, they should all be rcsolvablcatlhelowermagniiica cions used for both PCM and PLM analysis. Further, it is unlikely that particles o f this width would be lost in the transfer of parti cles from the litter to the glass slide in preparation for the PLM analysis. . In terms of aspect ratio. majordifferenccS between nonas- bestiform amphibole cleavage fragments and asbestiform particles also exist. Table V makes such n comparison for airborne particles which meet or exceed a 3:t aspect ratio and a greater than 5-pm length. Variances shown in this table typically are found in the literature.18'1'74' Figure 4 graphically depicts these data and further c!a rifles the differ ence. In terms of the study dost population, particle aspect ratio distribution is included inTable V under the cleavage fragment column where it best fits. Interestingly, total par ticulate aspect ratios noted in this study (based on 2295 particles) would represent the lowend of the cleavage frag ment line in Figured. Unfortunately,an airborne dust size characterisation for asbestiform tremolile could not be found for inclusion in this comparison. Although asbesti form iremolile is rare and is not exploited for commercial use. localized occurrences do exist :n the United States (i.e.. California. Montana). At least one industrial hygiene study exists of a mining operation containing asbestiform treino- Lle. but detailed airborne size characterization i; not avail- able.'1'" An aspect ratio d istrib u te.. Iv.vcver, wtu obtained on bulk asbestilorm tremolile rom ik|Mmine.""' Fcr parti cles longer than 5 Icli above I0;l. <0% above 15:1, and 52Cf above 20'-1. These ru-h/s correlate most closely to Sio theaverage airborne asbestos ratios teflected inTable Vand Figure 4 of 100%. 89%. and 53%. respectively. In summary, when the study dust population is contrasted with the minera logical definition--as well as the dimen sional characteristics of asbestiform and nonasbestiform particles reflected in the literature--the nonasbestiform nature of New York Slate tremolitic talc is quite apparent. The authors believe this reaffirms the nonspccificity of the NIOSH PCM method and the regulatory definitions it underpins when applied to mineral dust environments con taining common nonasbestiform cleavage fragments. Corrective Measures Given the differences between asbesiiiorm and nonasbestiform particulates, the least dramatic change necessary to improve specificity would involve an upward adjustment in the aspect ratio. As seen in Figure 4, airborne asbestiform particles exceed a 10:1 aspect ratio with very few less than 15;I. Cleavage fragments, in contrast, rartly exceed a 10:1 aspect ratio with fewer still exceeding 15:1. Any aspect ratio adjustment, however, should be applied as a screening tool only because there is some aspect ratio overlap between asbestiform and nonasbestiform particles. It, therefore, is considered essential that a miners logically correct definition of asbestos and criteria specific to asbestos should be reflected in regulations. Discussion Although it is not the intent of this caper toaddress health issues, the subject cannot be ignored In any discussion regarding the definition of asbestos, li can be argued, for example, that rcgulniorv definitions arc designed tonddiesx human health c neerns and not the realities of physical science. This argimem suffers, however, when it ts under- ity( -r" . I tW : lii'i-r'Wft I9S9 WES000381 NOTE: The majority ot cleavage fragments do not tall in this range . (most reflect lengths of< 5 nm). The 100%. iherefora, represents the starting point for 3:1 aspect ratio particle counting and noI the total w of airborne cleavage fragments Figure 4 --Airborne a sb e s lo s v e rso s cleavage fragm ent a sp e c t ratio com parison (parlicules with an a sp ect ratio of 3:1 o r g re a te r. > 5 pm length. > 0.25 p m width). From Table V. stood that health effects attributable to asbestos are not reasonably demonstrated for nonasbestiform exposures.01'381 Moreover, It can be argued that any environmen lal exposure ' ought io be studied and regulated for what it is. To do otherwise presents needless bias. It also has been argued that any change in the regulatory definition of asbestos would confuse the extensive data base developed for commercially used asbestos. Nonasbestiform amphiboles, however, cannot and are nol used for applica* lions typically reserved for asbestos (e.g...insulation, struc tural binding, fire proofing, brake linings,'etc.). Accord ingly, this asbestos data base would not be affected signifi cantly if a mtnera logically correct definition of asbestos were adopted. The definitional ambiguity discussed here relaLcs to dust populations which do contain nonasbestiform min eral cleavage fragments. Such environments commonly involve hard rock and aggregate mining operations and industries who use their mineral products (e.g.. ceramics, construction, paint, e'c.). Whatever asbestos data exist for these enviro nments may be misleading and, therefore, ought to be corrected. Conclusion M ajor differences in crystal growth patterns, lengths, and widths exist bclwccn usbestiform particles and common, hard rock-forming mineral d e.r >gc fragments. Current regulatory asbestos definitions and fiber gia.ttification methods do not address these disv.nut ms adeip a cly. Thus, nonasbestiform dust pOpt ' :t rns r-.-. and haw been mis taken ns .asbestiform. Confusion is likely to persist until ;i regulatory definition and analytical approach specific to asbestos is adopted. \ References 1. National Institute (or Occupational Safety and Health: USPHS/NIOSH Membrane Fitter Method lo r Evaluating Airborne Asbestos Fibers by N.A. Letdet. S.G. Bayer, R.O. Zumwalde, and K.A. Busch (Technical Report No. 79-127). Cincinnati. Ohio: National Institute for Occupational Safety an d Health, 1979, 2. National institute lor Occupational Safely and Health: N/OSHMar>ua/ofAna/yt/ea/Meinods.3ded.(DHHS/NIOSH Pub. No. 84-100). Washington. D.C.: Government Printing Oltice. 1984. Method #7400. 3. Gory, M,, R. McAfee, and C. Wolf: Glossary o f Geology. edited by R. Bales and J. Jackson. Pauls Church. Va.: American Geological Institute, 1972. p, 41. 4. United States Department ol the Interior: Se/ecterfSfrfce/e Minerals and Their Asbestilorm Varieties by W.J. Campbell. R.L. Blake, LL. Brown. H.E. Caiber. and J.J. SjObeft (Bureau of Mines Information Circular, l.C, 8751). W ashington, D.C.: Departm ent of the Interior, Bureau ol Mines. 1977. 5. `'Post-Hearing Comments in the Matter of Proposed Revi sions to the Asbeslos Standard.' Code o f Federal Regulations Title 29. 1910.1001, 1904. . 8. Wylie, A.G.: "The Relationship Between the Growth Habit ol Asbestos and the Dimensions ol Asbeslos Fibers." Paper presented ai the Society of Mining Engineer.' A n.ual Meet ing, Phoenix, Arizona January 1908. 7. Wylie, A.G., R.t.. Vlrta, and E. Russek: Characterizing and Discriminating Airborne Amphiboie Cleavage Fragmenis and Amosile Fibers--implications to; me NiOSH Method. Am Intf. Hyg Assoc. J. 46(41. 197-201 (1985). 0. Wylie, A.G.: Membrane Filter Method lor Estimating Asbeslos T>be< Exposure. Dehntttonsloi Asbestos and Other Health- Related Sti,;aies (aSTM STP 63a). edited by Benjamin -*n ml N*s I Movf*, >h l?5c art WES000382 Levadie. Philadelphia. Pa.; American Society lor Testing an d M aterials. 1984. pp. 105-11?. 9. Wylie, A.G-: Fiber Length and Aspect Ratio of Some Selected A sbestos Samples. Heblih Hazards o Asbestos Exposure. Ann. N.Y. Acad. Sci. 330:605-510 (1979). 10. Wylie, A.G.: Optical Properties ol the Fibrous Amphiboles. H eelth H azardsot A sbestos Exposure.Ann.N.Y. Acad. Sci. 330:611-620 (1979). 11. United S tales Department ol the Interior: Chemical and Physrcel Characterization ofAmoske, Chrysolite. Crocidolitc a n d N o n lib ro u s Tremolilo lo r Oral Ingestion Studies by the National Institute o f Environmental Health SciencesbyYt.J. Cam pbell. C.W. Huggins, and A.G. Wylie (U.S. Qureau ot M ines R eport ol Investigations #8452). Washington, D.C.: U nited States Department ol the Interior. Bureau ot Mines. I960. 12. United S lates Department ol the Interior: Relationship of M ineral Habit to Size Characteristics lo r Tremolile Cleavage Fragments and Fibers by W.J. Campbell, E.B. Steel. R.L. Vina, and M.H. Eisner (Bureau ol Mines Report o l Investiga tions #8307). Washington, D.C.: United Slates Department of the Interior. Bureau ol Mines, 1979. 13. U nited S tales Departm ent ol the Interior: Amphiboles in Soapstone Ridge, GA by R.L, Stake (U.S. Bureau ol Minos R eport ol Investigations #6627). W ashington. D.C.: United S tates Department of the Interior. Bureau o( Mines, 1962. 14. Zollai, T j H istoryof Asbestos-Related Mlneralogical Ter minology. Proceedings o f the Workshop on Asbestos: Definitions and Measurement Methods (National Bureau ol S tan d ard s Special Publicaron 506). Washington. D.C.:Gov- ernm ent Printing Obice, 1970. pp. 1-16, . 15. Z ussm an, J.: The Crysiat Structures of Amphibolo and Ser p en tin e Minerals. Proceedings o l the Workshop on Asbes- !os: Definitions and Measurement Methods (National Bureau o l Standards Special Publication 506). Washlnglon, D.C.:'' G overnm ent Printing Ollice. 1970. pp. 35-48. . ' 16. Thom pson, C.S- Discussion ol the Mineralogy ol industrial Talcs. In Proceedings of theSymposium on Tile, Washington, DC.. May $, 1973. W ashington, D.C.: United States Depart m ent ol the Interior. Bureau ol Mines. 1974, pp. 22-24. 17. R oss, M.: The Asbestos Minerals: Delinillons. Description, Modes ol Formation. Physical and Chemical Properties and H ealth Risk to the Mining Community. Proceedings ot the Workshop on Asbestos: Pelinltions and Measurement Meth ods (National Bureau ol Standards Special Publication 506). Washington, D.C.: Government Printing Office, 1978. pp. 49-6. 16. E ngel, A.E-J- The Precambrian Geology and Talc Deposit o f the Bafmat-Edwards Oistrlct, Northwest Adirondack M ountains, New York. (U.S. Geological Survey Open Pile R eport). W ashington, D.C.: Government Printing Ollice, 1962. p. 357. . 19. T h o m p so n , C.S.: C onsequences ol Using Improper Defini tions lor Regulated Minerals. In Definitions lor Asbestos end O ther Health-Related Silicates (STP-634). Philadelphia, Pa.: American Society lorTeslIng and Materials, 1984. p. 182. 20. Harvey, A.M.: Tremolile in Talc--A Clarification, in Indus tria i Minerals. W orcester Park. Surrey, England: Metal Bul letin Limited. 1979. pp. 23-59. 21. National Ins titule ol Occupational Safety and Health: Occu pational Exposure to Tate Containing Asbestos by D.P. Brown. J.M. D em ent. R.D. um w alde. j . , \ Gamble, w . Fellner. M.J. D etnco. and J.K. Wf.ot net (DHEW/NIOSH Publication No. $0-115). W ashmglon. D.C.: Government Printing Ollice. . i960. 100 p 22. Snyder, J.G., R.L. Vida, an d J.M. S e g re t: Evaluation ol the P h ase Contrast Microscopy Method lor the Detection of Fibrous 8nd O ther Elongated Mineral Particulates by Com parison with an STEM Technique. Am, tnd. Hyg. Assoc. J. 49:471-477 (1987). 23. G ibbs, G.W. and C.Y. Hwang: Dimensions of Airborne A sbestos Fibers. In Biologic Effects o l MineralFibers. Vol. I. edited by J.C . Wagner (IARC Scientific Publication (30). Lyon, France: World H ealth Organization. 1980. pp. 79-86. 24. ASTM: Standard Test M ethod D 4240-63 tor Airborne As bestos Concentration in Workplace Atmosphere. In Annual Book o f ASTM Standards. Vol. 11.03. Philadelphia. Pa.: ASTM. 1987. p. 420 25. Rpoker,S.J.,N ,P,V aughan,and JM.Le GuemOnlheVisibil ity Ol Fibers by Phase C ontrast Microscopy. Am. tnd. Hyg. A ssoc. J. 43:505-515 (1982). 26. P ang. T.W.S., W.L. Dicker, and M.A. Nazar: An Evaluation ol the Precision and Accuracy ol the Direct Transfer Method for the Analysis of A sbestos Fibers with Comparison to the NIOSH Method- Am. Ind. Hyg. Assoc. J. 46:329-335 (1984). 27. United States Department olthelftierior: The Phase Relation sh ip o t Talc and Am phiboles in a Fibro'vS Talc Sample by R.L. Vina (B ureau o( Mines Ftoport of investigations #8923), Washington, D.C.; United States Department of the interior, B ureau ot Mines, 1985. pp. 1- 11. 28. Veblen, D.R. and C.W. Burham:Triple-Chain Stopyribolcs: Newly Discovered Intermediate Products of the Retrograde Anthophyliite-Talc Transformation. Science 198:359-365 (1977). 29. McDonald, J.C ., A.D. McDonald, B. Armstrong, and P. Sbastien: Cohort Study ot Mortality ol Vermicutrte Miners Exposed to Tremolilo. Br. J. Ind. Med. 43:436-444 (1906). 30. Atkinson. G.R.. D. Rose, K. Thomas, D. Jones, end E.J. Chatlield: Collection. Analysis and Characterization o l Vermicuiiie Samples fo r Fiber Content and Asbestos Con tamination. W ashington. D.C.; United S lates Environmental Projection Agency. 1982. 31. M arsh, J-P-and B.T.Mossman: Mechanisms ol Inductionol O rithine Decarboxylase Activity in Tracheal Epithelial Cells by Asbeslilorm Minerals. Cancer Res. 48:709-714 (1988). 32. Sm ith, W.E., D.D. Hubert, H.J. Sobel, and E. Marquet: Bio logic Tests of Tremolite in Hamsters, in Oust and Disease edited by R.J. Lemen, J.M. Dement. Park Forest South. III.: Palhotox Publishers. 1979. pp. 335-339. 33. S tanton, M.F., M. Layard, A, Tegeris, E. Milles, M. May. E. M organ, and A. Smith: Relation ol Particle Dimension to Carcinogenicity in Amphlbole Asbestosis and O ther Fibrous Minerals. J. Natl. Cancer InsI. 67:965-975 (1961). 34. Lamm, S.H., M.S. Levine, J.A. Starr, and S.t-T irey: Analysis . . of Excess Lung C ancer Risk In Short-Term Employees. Am. ' ' J. Epidemiol, f 27(6): 1202-1209 (1988). 35. Tabershaw , |,R. and W.T. Slilte:The Mortality Experience ol U pstate New York Talc Workers. J. Occup. Med. 24:480-404 (19$2). 36. McDonald, J.C ., G.W. Gibbs, F.O.K, Liddell, and A.D, McDonald: Mortality Alier Long Exposure to Cumminglonlte- G runerite. Am. Rev. Resplr. Dis. i f 0:271-277 (1970J. 37. C ooper, W.C., Olio Wong, and ft. G raeb n er Mortality ol W orkers in Two Minnesota Taconile Mining and Milling Operations. J. Occop. Med, 30:506-511 (1986). 38. McConnell, E.E., H.A. Rutler, B.M. U lland,and J.A. Moore: C hronic Eltects ol D ie ta ry Expotuiv ,<> Amoslle Asbestos and Tremolite in F344 Rais. Environ. Health Petspect. 53:27-44 (1983). 6 May I9S8; Roiled iv ju>.. >9>}9 611 Irti H,i / '.n; ! isoI '^39 WES000383