Chemical Characterization andReactivity of Iron Chelator-treatedAmphibole AsbestosJulie Gold,1 Helena Amandusson,l Anatol Krozer,lBengt Kasemo,l Tore Ericsson,2 Giovanna Zanetti,3and Bice Fubini31Department of Applied Physics, Chalmers University of Technology,Goteborg, Sweden; 21nstitute of Earth Sciences, Uppsala University,Uppsala, Sweden; 3Department of Inorganic Chemistry, ChemicalPhysics, and Chemistry of Materials, University of Torino,Torino, Italy

Iron in amphibole asbestos is implicated in the pathogenicity of inhaled fibers. Evidence includesthe observation that iron chelators can suppress fiber-induced tissue damage. This is believed tooccur via the diminished production of fiber-associated reactive oxygen species. The purpose ofthis study was to explore possible mechanisms for the reduction of fiber toxicity by iron chelatortreatments. We studied changes in the amount and the oxidation states of bulk and surface ironin crocidolite and amosite asbestos that were treated with iron-chelating desferrioxamine,ferrozine, sodium ascorbate, and phosphate buffer solutions. The results have been comparedwith the ability of the fibers to produce free radicals and decompose hydrogen peroxide in a cell-free system in vitro. We found that chelators can affect the amount of iron at the surface of theasbestos fibers and its valence, and that they can modify the chemical reactivity of thesesurfaces. However, we found no obvious or direct correlations between fiber reactivity and theamount of iron removed, the amount of iron at the fiber surface, or the oxidation state of surfaceiron. Our results suggest that surface Fe3+ ions may play a role in fiber-related carboxylate radicalformation, and that desferrioxamine and phosphate groups detected at treated fiber surfaces mayplay a role in diminishing and enhancing, respectively, fiber redox activity. It is proposed that ironmobility in the silicate structure may play a larger role in the chemical reactivity of asbestos thanpreviously assumed. Environ Health Perspect 1 05(Suppl 5):1 021-1030 (1997)

Key words: asbestos fibers, iron chelators, iron mobilization, free radical release, hydrogenperoxide, electron paramagnetic resonance spectroscopy, surface analysis, X-ray photoelectronspectroscopy, Mossbauer spectroscopy

IntroductionIron in inhaled inorganic particulates may the latter might also originate as ironcontribute to, or be the direct cause of, redeposited from solution or endogenousthe pathogenicity of inhaled dusts and sources (1-8). Many in vivo and in vitrofibers. Both iron mobilized from the partic- investigations point to the crucial roleulate and iron at the surface are implicated; played by free radicals and active oxygen

This paper is based on a presentation at The Sixth International Meeting on the Toxicology of Natural and Man-Made Fibrous and Non-Fibrous Particles held 15-18 September 1996 in Lake Placid, New York. Manuscriptreceived at EHP26 March 1997; accepted 29 May 1997.

The authors thank H. Riedl, Chalmers University of Technology, Sweden, for his assistance with the atomicabsorption spectroscopy analysis. This research was carried out within the Swedish Biomaterials Consortium,which is financed by the Swedish National Board for Industrial and Technical Development and the SwedishNatural Science Council. Support was also received from Partek Insulation AB, Sweden and Finland.

Address correspondence to Dr. J. Gold, Department of Applied Physics, Fysikgrand 3, Chalmers Universityof Technology, S-412 96 Goteborg, Sweden. Telephone: 46 31 772 3369. Fax: 46 31 772 3134. E-mail:f7xjg@fy.chalmers.se

Abbreviations used: AAS, atomic absorption spectroscopy; at%, atomic percent; CO2-, carboxylate radical;CS, centroid shifts; DMPO, 5,5'-dimethyl-1-pyrroline-N-oxide; EPR, electron paramagnetic resonance; HRTEM,high resolution transmission electron microscopy; ICP-AES, inductively coupled plasma-atomic emission spec-troscopy; MS, Mossbauer spectroscopy; 'OH, hydroxyl radical; QS, quadrupole splittings; UICC, UnionInternationale Contre le Cancer; XPS, X-ray photoelectron spectroscopy.

species in causing asbestos-mediated tissuedamage (9-14). The highly reactivehydroxyl radical (OH) has been invokedin many of these processes. Asbestos fiberscan catalyze the production of free radicals,including hydroxyl radicals, in cell-free sys-tems (15-18), in cell culture (14,19), andin vivo (12).

Asbestos-related free radical production,as well as biological damage, can be sup-pressed or inhibited by iron chelators, suchas desferrioxamine and ferrozine, bothwhen present in the test system and whenused to pretreat fibers (17,20-24). On theother hand, other chelators such as ascor-bate enhance free radical production (25).Chelators may remove iron from asbestos,alter iron valence states via oxidation orreduction, or remain bound at surface ironsites, thus preventing the participation ofiron in redox reactions (15,26) or alteringthe overall redox potential of the asbestosfibers. In addition, chelators may stabilizeiron that is released into the surroundingmedium and thereby alter the dissolutionkinetics. The mechanisms involved in thesuppression of free radical production byiron chelator treatments have not beenclearly elucidated. These mechanisms arekey to understanding the specific role ofiron and iron-catalyzed reactions inasbestos toxicity.

Much attention has been focused on therole of asbestos-associated iron in the pro-duction of hydroxyl radicals from H202 viacatalytic Fenton and Haber-Weiss typereactions (15,17,22,27,28). Hydroxyl radi-cals are produced by the reduction of H202by Fe2+ in the Fenton reaction, and by thereaction of H202 with superoxide anion(02-) during the reduction of Fe3+ in theiron-catalyzed Haber-Weiss reaction.Another free radical of interest is the car-boxylate radical (CO2-1), which is formedfrom the abstraction of hydrogen from for-mate ions (HCO2y') on contact with thesolid (17). Recent reviews by Hardy andAust (29) and Fubini and Mollo (6) pre-sent a comprehensive overview of what iscurrently believed to be the role of iron infree radical production. A specific concerninvolves the influence and relative impor-tance of Fe3+ and/or Fe2+ ions available forredox-cyclic reactions at fiber surfaces.

Within the amphibole asbestos group,crocidolite [Na2(Mg+2, Fe+2)3Fe 32Si8O22(OH)2] and amosite [(Mg+2, Fe'2)7Si8O22(OH)2] contain equivalent amountsof iron [27 and 28 wt%, respectively, as

Environmental Health Perspectives * Vol 105, Supplement 5 * September 1997 1021

GOLD ET AL.

determined for Union Internationale Contrele Cancer (UICC) asbestos (30)], but theiron is found in different valence states andcrystallographic sites in these layered silicatestructures. Iron in crocidolite is present inboth 2+ (= 55 atomic percent [at%]) and 3+(= 45 at%) valence states, while only the 2+state of iron is found within the amosite crys-tal structure. It is important to note that themajority of iron in the outermost atomic lay-ers of both asbestos types is oxidized to Fe3+(8,31), as expected. Cation lattice sites Ml,M2, and M3 in these amphibole structureshave octahedral coordination; M4 is a cubicsite. Site population characterization indicatesthat the iron in amosite (as Fe2+) is found inall cation sites. The Fe2+ ions in crocidoliteare located in Ml and M3 sites, while Fe3+ions are found in the M2 site (32).

Several methods have been employedto purposely alter surface Fe2+ and Fe3+contents of fibers to study their influenceon the redox reactions mentioned above.By far the most common techniqueinvolves the use of different iron chelatorsthat specifically remove Fe2+ (e.g., fer-rozine) or Fe3+ (e.g., desferrioxamine) ionsor bind to these sites at the fiber surface.Asbestos surfaces have also been loadedwith iron of a specific valence state byincubation in concentrated ferric or ferroussalt solutions (4,7). In both cases, the abil-ity of silicates to catalyze DNA damage andgenerate oxygen radicals increased. Heattreatment in a reducing gas atmosphere isan effective method of producing Fe-reduced crocidolite with relatively long-term stability in air (10). Such a reductiontreatment activated crocidolite fibers toenhance free radical production and lipidperoxidation. However, in the same study,Fe3+-loaded crocidolite was shown to havea detoxifying effect. In general, it is clearthat not all asbestos-associated iron isinvolved; it appears that the valence andlocation of iron in the lattice, as well as itsavailability for mobilization, control theactivity of the fibers. Identification of theiron-binding sites responsible for fiberreactivity is yet to be determined.

The status of iron in asbestos isconventionally characterized indirectly byan analysis of how much iron can beremoved from the fiber by valence-specificiron chelators. Similarly, iron concentra-tion and valence state at chelator-treatedsurfaces are often described by how muchiron of each valence state was removedfrom the fiber, rather than what actuallyremained at the surface. This description isquestionable because surface reactions can

rapidly change the valence state of an ion.Knowledge of what remains at the surfaceis particularly important, as interactionsbetween fibers and the biological systemoccur or originate at this interface. For thisand other reasons, it is important to char-acterize the iron that is actually present atthe surface of native and iron-manipulatedamphibole asbestos. By characterization ofsurface and bulk modifications induced byiron chelator treatments, it should be possi-ble to better define the action of specificiron chelators and the role of iron in freeradical production.

In this paper we report the use of X-rayphotoelectron spectroscopy (XPS) andMossbauer spectroscopy (MS) to character-ize the nature of iron at the surface and inthe bulk, respectively, of crocidolite andamosite asbestos when treated with desfer-rioxamine, ferrozine, sodium ascorbate,and phosphate buffer iron-chelating solu-tions. The effects of iron chelator treat-ments were then compared with the abilityof the fibers to produce catalytic decompo-sition of H202 and generate free radicals,as determined in a cell-free system by theelectron paramagnetic resonance (EPR)spin trapping method.

MethodsFiber Preparationand Iron Mobilizaton

Samples of UICC crocidolite and amosite(33) in the nontreated, as received, condi-tion were analyzed after incubation in oneof the following iron-chelating solutions:desferrioxamine, sodium ascorbate, fer-rozine, or phosphate buffer. Ascorbate isboth a reducing agent and chelator: whenFe3+ is reduced to Fe2+, the change in ionsize facilitates its extraction from the sili-cate matrix. All solutions were prepared at10-mM concentrations in water. Fiberswere incubated at a concentration of 10mg fibers/ml chelator solution for 3 days at37'C and were continuously shaken andkept in the dark. No attempt was made tocontrol pH. We believed that the pHshould not change significantly during theincubation period, except for ferrozine,which most likely became acidic with incu-bation time. The use of buffer solutionscan confound results, as buffers affect ironmobilization from crocidolite (34).

The fibers were centrifuged and thesupernatants were filtered and analyzed fortotal iron content by atomic absorptionspectroscopy (AAS) and inductively coupledplasma-atomic emission spectroscopy

(ICP-AES) as a comparative technique.Fibers were rinsed in distilled water and vac-uum dried for 48 hr at room temperature. Aportion of each fiber type was packaged andstored under argon atmosphere until ana-lyzed by XPS or MS (3 weeks-3 months).Another portion of fibers was retained forfree radical and H202 decomposition test-ing and was stored in alr for approximately4 months prior to testing.

Characterization ofBulkand Surface Iron and AdsorbatesMossbauer Spectroscopy. Characterizationof the bulk iron in chelator-treated andnontreated asbestos fibers was performedby MS. The fibers were mixed with boronnitride and loosely packed into a metallicring. In the transmission measurements,the angle between the normal of theabsorber plane and the y-direction was heldat 54.7°, thus no texture effects were seenin the spectra (35). Accordingly, it waspossible to fit the spectra using symmetricaldoublets for Fe2+ and Fe3+ at the differentcrystallographic positions. The relativeamount of ferrous and ferric iron was cal-culated from the areas of the doublets. Thismethod assumes no sample thicknesseffects and equal f-factors at the differentcationic binding sites.

X-ray Photoelectron Spectroscopy.XPS is a surface-sensitive analyticalmethod in which analysis depth is limitedto the top 5 to 10 atomic layers, or < 10nm. The detection limit of XPS is 0.1 to1.0 at%, depending on the element.Changes in the surface chemical composi-tion of amosite and crocidolite fibers dueto chelator treatments have been evalu-ated with a PHI 5500C (Perkin-ElmerCorporation/Physical Electronics, EdenPrairie, MN) using a monochromatic AlKa X-ray source with system spectral reso-lution approximately 0.5 eV [(31); Gold etal., in preparation]. Charge compensationwas achieved by the use of low-energy elec-trons generated in an electron flood gun. Aminimum of four analysis spots on one ormore samples was used to calculate meanvalues of surface compositions.

To compare iron concentrations betweensamples, relative iron concentrations wereused and were calculated by dividing theatomic concentration of iron (determinedusing the Fe 2p peak) by the atomic concen-tration of silicon (Si 2p). To determine theoxidation state of iron, semiquantitativeevaluation of the high resolution Fe 2P3/2photoelectron peak (obtained at 0.025eV/step, 2.95 eV pass energy) was performed

Environmental Health Perspectives - Vol 105, Supplement 5 - September 19971022

CHEMICAL CHARACTERIZATION OF CHELATOR-TREATED ASBESTOS

by curve-fitting routines. This is necessaryas there is insufficient resolution to distin-guish the close-lying, multiplet-split peaksfrom Fe2+ and Fe3+ valence states. Curve-fitting procedures were developed to calcu-late the percentage of total iron in the Fe2+or Fe3+ oxidation states [(31); Gold et al., inpreparation]. Uncertainties in the curve fit-ting are inherendy large, but we are able tofollow trends that arise due to differentchelator treatments.

Characterization of Fiber RactivityFree Radical Generation. EPR spec-troscopy (Varian E109 EPR spectrometer[Varian, Palo Alto, CA] working in the Xband [9-9.5 Ghz] with a double resonantcavity) was used to detect the formation offree radical species in aqueous suspensionsof asbestos fibers via the spin-trappingmethod (16,17). The spin-trapping agent5, 5'-dimethyl- 1 -pyrroline-N-oxide(DMPO) forms stable radical adducts andwas used to quantitate carboxylate andhydroxyl free radical release in solution.Two target molecules were used: sodiumformate (F) and hydrogen peroxide 30%(P). They form the DMPO-COO'- andDMPO-'OH adducts, respectively. Eachis characterized by a typical EPR spectrum.The intensity of the EPR signal is a mea-sure of the number of radicals trapped byDMPO, and thus the number of radicalsformed in the solution. These two tests arereferred to as F-C (for carboxylate radicalformation from the abstraction of a hydro-gen atom from the formate ion) andP-OH (for hydroxyl radical formationfrom the decomposition of hydrogenperoxide via a Fenton mechanism orhomolytic rupture), respectively.

In the F-C test, 45 mg of fibers wasplaced in 2 ml Chelex-treated 0.5 M phos-phate buffer (Sigma Chemical, St. Louis,MO) (pH 7.4) containing 1 M sodium for-mate and 0.05 M DMPO. The suspensionwas incubated at 370C and shaken in a darkreactor. Aliquots of the suspension werewithdrawn at 55 min, then filtered and ana-lyzed at room temperature approximately 5min after the sample was withdrawn. TheEPR spectrum was recorded at microwavepower level 10 mW, scan range 100 G, andmodulation amplitude 1 G. The spectrumshows six lines with intensity ratios1:1:1:1:1:1, splitting constants aN = 15.2 G,aH= 18.5 G, and g-value = 2002.

In the P-OH test, the protocol is similarto that of the F-C test, but because of thelower DMPO adduct stability, incubationtimes are shorter. In this case, 45 mg of

fibers was placed into 2 ml solutionconsisting of 0.5 ml 0.5 M phosphate buffer(pH 7.4), 0.5 ml 30% H202 (diluted 1:50in phosphate buffer), and 1 ml 0.05 MDMPO. The EPR spectrum was recordedafter 30 min incubation. The spectrumshows four lines with intensity ratios1:2:2: 1, splitting constants aN = aH = 149.9G, and -value = 2054.

Because of an intrinsic variability fromone DMPO preparation to another,blanks were performed in each set ofexperiments. Blanks were run with solu-tion only (no asbestos fibers) and showedno radical formation. The intensity of thespectra was reported in arbitrary units cal-culated by dividing the height of the signal(in millimeters x 100) by the receiver gain.

Hydrogen Peroxide Decomposition.The percentage of H202 decomposed after300 min incubation (370C, shaken in darkreactor) of 50 mg treated or nontreatedfibers with an H202 solution (1 ml 0.1 MH202 + 30 ml distilled H20) was deter-mined by measuring the residual amountof H202 in 100 pl filtered suspension(17). Hydrogen peroxide was measuredusing an enzymatic assay whereby peroxi-dase catalyzes the oxidation of leuco crystal

violet by H202 via the method describedby Mottola et al. (36).

ResultsIron Mobilizaiion by Chelators

The total amount of iron mobilized fromasbestos fibers into the chelator solutionswas calculated on the basis of fiber weightor fiber surface and is shown in Figure IAand B, respectively. Based on the theoreti-cal composition and molecular weight ofcrocidolite and amosite, it was estimatedthat the amount of iron removed from thefibers due to chelator treatments was onthe order of 1 to 5 at% of the total ironpresent in the fibers. The least amount ofiron mobilized was found for the phos-phate buffer solution. All other chelatorsmobilized between 2 to 4 times moreiron, depending on whether iron removalis compared on the basis of weight versussurface area. Results obtained with the AASand ICP-AES methods are comparableexcept in the case of ascorbate-treated cro-cidolite. The reason for this discrepancy isnot clear.

Except for ferrozine treatments,slightly more iron could be mobilized

AX 250.0

cm 200E

E 150

Er 100

CD 50

-ii

CD 0

Desferrioxamine Ascorbate Ferrozine Phosphate Desferrioxamine Ascorbatebuffer

Crocidolite Amc

Ba) 25._

,I.

g -T500E10-

U,

Deeferriaxamine Ascorbate Ferrozin Phosphatebuffer

Crocidolite

- AAS {error ± 5%)_ICP-AES

Ferrozine Phosphatebuffer

osite

ill.~~~~~~~~~.Desferrioxamine Ascorbate Ferrozine Phosphate

bufferAmosite

Figure 1. Total iron mobilized by desferrioxamine, ascorbate, ferrozine, and phosphate buffer from crocidolite andamosite, as determined by MS and ICP-AES. (A) Iron release normalized by fiber weight. (B) Iron release normal-ized by fiber surface area (only AAS data presented). One set of measurements was made using each method. Theerror in the AAS measurement is estimated to be ± 5%.

Environmental Health Perspectives - Vol 105, Supplement 5 * September 1997 1 023

GOLD ET AL.

Table 1. Evaluation of Mossbauer spectra. Curve fitting parameters and resulting concentrations of iron ions indifferent cationic lattice sites in nontreated and chelator-treated crocidolite and amosite asbestos.

Crocidolite AmositeDl D2 D3 Dl D2

Doublet 1 2 3 1 2Centroid shifta(mm/sec) 1.13 1.11 0.39 1.15 1.08Quadrupole splittingb(mm/sec) 2.88 2.39 0.43 2.79 1.56Lattice site Ml M3 M2 Ml, M2, M3 M4Ion Fe2+ Fe2+ Fe3+ Fe2+ Fe2+

Relative concentrationc,%Nontreated 33.0 25.1 41.9 61.6 38.4Desferrioxamine 30.6 29.7 39.6 63.4 36.6Ascorbate 31.4 28.1 40.5 61.6 38.4Ferrozine 31.9 26.8 41.3 60.9 39.1Phosphate buffer 29.1 30.3 40.7 63.4 36.6

D, doublet. "Relative to metallic iron at room temperature. bPeak separation in the doublet. cRatio of the area ofthe doublet to the total absorption area x 100%.

from crocidolite (surface area= 8.3 m2/g)than amosite (surface area= 5.7 m2/g)fibers for the same chelator treatment(Figure 1A), and this might be due to thegreater surface area of crocidolite. If ironrelease is normalized by fiber surface area,as shown for the AAS data in Figure 1B,the amount of iron released from bothfiber types is equivalent, again except forferrozine. Ferrozine caused a significantly

A

0 X

E960

Velocity, mm/sec

greater amount of iron mobilization fromamosite than crocidolite.

Chemical CharacterizationofAsbestosBulk Iron. The results of the MS analysisreveal the distribution of bulk iron in thedifferent cation lattice sites of the asbestosfibers (Table 1). An example ofMS spectraobtained from crocidolite fibers is shownin Figure 2A. The room-temperature MSspectra of crocidolite were fitted by threedoublets with centroid shifts ([CS] relativeto metallic iron) and quadrupole splittings([QS] peak separation in the doublet), asindicated in Table 1. The variation in theseparameters was approximately±0.01mm/sec between samples. The three dou-blets were assigned to Fe2+ at M1 sites,Fe2+ at M3 sites, and Fe3+ at M2 sites. Theparameters are in agreement with Pollak etal. (37). In addition to the two doublets,there was also diffuse, broad absorption(not fitted, estimated to be a few percent oftotal signal) in the velocity range of 1.0 to1.9 mm/sec (Figure 2A). Signal intensity in

this range was not affected by chelator treat-ments. This signal may emanate from ironclose to or at the fiber surfaces, or from ironwithin the fibers in poorly defined locations.

The MS spectra of amosite (Figure 2B),also measured at room temperature, werefitted by two Fe2+ doublets with CS and QSas indicated in Table 1, again with a varia-tion of approximately±0.01 mm/secbetween the samples. These two doubletswere assigned to iron in the MI, M2, andM3, and M4 crystallographic sites, respec-tively, as in earlier assignments (32). Therewas also a weak, nonfitted absorption in theinterval 0.7 to 1.5 mm/sec that was notaffected by the different chelator treatments.

Based on the results from Table 1,amosite was found to contain only Fe2+,whereas crocidolite consisted of approxi-mately 59 at% Fe2+ and 41 at% Fe3+.Chelator treatments did not significantlyaffect bulk ferrous to ferric ratios.

Surface Compositionm The primary ele-mental constituents detected by XPS at thesurface (i.e., within approximately the top10 atomic layers) of both treated and non-treated crocidolite and amosite fibersinclude 0 (= 40-50 at%), C (= 12-30 at%),Si (=7-21 at%), and Fe ("4-14 at%). Mgcontent of amosite surfaces (= 1-3 at%) wasgreater than crocidolite surfaces (typically< 1 at%), as can be expected from theoreti-cal bulk compositions. Similarly, no Na wasfound on amosite surfaces, whereas 1 to 2at% was detected on crocidolite. Low levels(typically < 1 at%) of Ca, N, P, and K werealso detected.

The amount of iron detected at thesurface of the fibers as a function of chela-tor treatments is shown in Figure 3. Theiron content has been normalized by the sil-icon content in order to preclude artifactsthat might arise from compositional varia-tions in other elements when samples are

I 1

Co

.,_cn

LL

-4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0

Velocity, mm/sec

Nontreated Desferri-oxamine

Ascorbate Ferrozine

*ICrocidolite

Desferri Ascorbate Ferrozine Phosphateoxamine buffer

Amosite

Figure 2. Mossbauer spectra of nontreated (A) crocido-lite and (B) amosite. Details of the doublet structuresdetermined by curve fitting can be found in Table 1.

Figure 3. Relative surface iron content of nontreated and chelator-treated crocidolite and amosite, as determinedby the ratio of the atomic concentration of iron (Fe 2p) to silicon (Si 2p) detected by XPS. Average values and stan-dard deviations are shown.

Environmental Health Perspectives * Vol 105, Supplement 5 * September 1997

-1-0-en

E

as(0

1 024

CHEMICAL CHARACTERIZATION OF CHELATOR-TREATED ASBESTOS

compared. The Fe/Si ratio averaged approx-imately 0.8 for nontreated crocidolite andamosite surfaces.

Changes in surface iron content due tochelator treatments were observed; however,they differed for the two fiber types. Forcrocidolite, significantly lower iron concen-trations were observed with desferrioxam-ine- (Fe/Si = 0.53) and ascorbate-treated(Fe/Si = 0.64) crocidolite surfaces comparedto nontreated and phosphate buffer-treatedcrocidolite surfaces. However, there were nostatistically significant differences in surfaceiron content between nontreated, ferrozine,and phosphate buffer-treated surfaces. Foramosite surfaces, desferrioxamine, ferrozine,and phosphate buffer-treated fibers hadlower iron content (Fe/Si 0.67) than non-treated fibers. However, there was no statis-tically significant difference in iron contentof ascorbate-treated compared to nontreatedamosite surfaces.

The oxidation states of iron present atfiber surfaces were determined by deconvo-luting and curve fitting the Fe 2P3/2 photo-electron peak. An example of the curve fit isshown in Figure 4A for a spectrum obtainedfrom nontreated crocidolite. (Details on thespecific curve fitting procedure used can befound in Gold et al., in preparation.) OnlyFe2+ and Fe3+ oxidation states were detected.Distinct differences in the Fe 2P3/2 peakshapes of nontreated crocidolite and amositewere observed (spectra not shown). Thesedifferences were expressed in the curve-fittingresults shown in Figure 4B.

The most significant observation wasthat, despite the differing compositions ofbulk iron between the two asbestos types,the majority of iron at all surfaces was in theFe3+ state. There was on average approxi-mately 5 at% more Fe2+ on nontreatedamosite than crocidolite. This differencecould be expected considering differences inFe2+ content of their bulk compositions. Itis surprising that not all surface iron is oxi-dized to the Fe3+ state, considering thatthese UICC samples were prepared almost30 years ago and have been stored in air eversince (33). An average increase of 6 to 10at% Fe2+ content was measured on chelator-treated amosite surfaces. There was a slightincrease in Fe2+ on some ascorbate-treatedcrocidolite samples, but due to the largespread in the data, this increase was not sta-tistically significant. The other chelatortreatments did not alter the Fe2+ content ofcrocidolite surfaces.

Evidence of Adsorbed Chelators.Certain elements present in low quantities(< 1 at%) on the original asbestos were

observed in larger quantities at the surface ofchelator-treated fibers. The most significantfindings include elevated levels of nitrogenon desferrioxamine-treated amosite and cro-cidolite, as can be seen in Figure 5. Althoughthe levels are low and fall around 2 at%, they

are significantly greater than the levels ofnitrogen detected on all other surfaces,which fall around our estimated value for thedetection limit of nitrogen on these surfaces(= 0.5 at%). Nitrogen generally adsorbs onsurfaces from exposure to air, but in this

A10

9

L..

z

6

5-

716 714 712

Binding energy, ev710 708

B50 -

CA

30 -

° 20-0.-1+ 10U-

I T

Nontreated Desfer- Ascorbate Ferrozineoxamine

Crocidolite

Phosphate Nontreated Desferri- Ascorbate Forrozine Phosphatebuffer oxamine buffer

Amosite

Figure 4. (A) Example of the curve fitting procedure used in this study on the Fe 2p3/2 XPS peak obtained from non-treated crocidolite. Fe3+ states were represented by peaks 1 to 3; Fe2+ states were fit by peaks 4 and 5. (B)Percentage of total surface iron in the Fe2+ valence state by XPS on nontreated and chelator-treated crocidoliteand amosite. Iron was detected as either Fe2+ or Fe3+. Average values and standard deviations are shown.

2.5

2

w 1.5

o 1

Z 0.5

O

Crocidolite

T

T

Estimateddetection

limit T

Amosite

Figure 5. Evidence for desferrioxamine bound to crocidolite and amosite surfaces by nitrogen XPS spectra.Nitrogen concentration on desferrioxamine-treated fibers is elevated by approximately 1.5 at% compared to allother surfaces. Average values and standard deviations are plotted. The estimated detection limit for nitrogen onthese samples (= 0.5 at%) is indicated.

Environmental Health Perspectives * Vol 105, Supplement 5 * September 1997

munuestu ueeem- Ascrt rwerurzine rnuspnaeie Nunt.reeCIu ueiern- Aisucruaei rerruzineoxamine buffer oxamine

l

1025

GOLD ET AL.

study it probably also originates fromadsorbed organic molecules such as thechelating molecules.

Additional evidence for desferrioxam-ine binding was found in the intensityand the binding energy position of thecarbon Is photoelectron peak. Specificdifferences were noted in the high-bind-ing energy peak that typically appears as ashoulder in the C Is spectra. The high-binding energy C Is peak on crocidoliteand amosite surfaces exposed to desfer-rioxamine was shifted by approximately3.5 eV with respect to the main saturatedhydrocarbon peak (= 285.6 eV), whereason all other surfaces the shift was on aver-age approximately 4.4 eV (crocidolite) orapproximately 4.1 eV (amosite). In addi-tion, the intensity of this high-bindingenergy C Is peak tended to be higher ondesferrioxamine-treated surfaces than onall other surfaces (data not shown).

Similar indications of phosphate boundto phosphate buffer-treated surfaces wereobserved by XPS analysis. Average concen-trations of 0.8 to 1.2 at% phosphorus (asphosphate) and potassium were detected onphosphate buffer-treated surfaces, but wereat or below the estimated detection limitfor these elements (= 0.5 at%) on all othersurfaces. Although these signals are small,with a wide spread in the data, we believethat this is evidence that phosphate remainsadsorbed to phosphate buffer-treated fibers.

RedoxActiity ofAsbestosResults from EPR spin trapping measure-ments, which detected the release of car-boxylate (F-C test) and hydroxyl (P-OHtest) free radicals are shown in Figure 6Aand B and Figure 6C and D, respectively.It should be noted that the EPR signalshave overall low intensities that are at ornear detection levels. In fact, production of'OH on nontreated amosite surfaces wasbelow the detection limit. Low signalintensities are unavoidable, as both fibersare quite unreactive unless they aregrounded. Nontreated crocidolite hasslightly greater activity in both tests andgenerates more free radicals than amosite,but this could be an effect of the greatersurface area for crocidolite.

Variations in free radical release potentialfollowing chelator treatments were observedon crocidolite and amosite for most chela-tors. All chelator treatments eliminated orreduced fiber reactivity in the F-C test(Figure 6A, B). This effect was most dra-matic with amosite fibers (Figure 6A), as theDMPO-CO2-f signal was completely elimi-

AFresh sa e=65Fresh sample

BI = 102

Fresh sample

1=0 I 1=92DesferrioxamineDesferrioxamine

1=0 1=70FerrozineFerrozine

1=0Ascorbic acid

1=0Ascorbic acid

I=ND 1=48Phosphate bufferPhosphate buffer

C DI=ND

Fresh sample1=45

Fresh sample

1=0DesferrioxamineDesferrioxamine

I=ND

1=951=0FerrozineFerrozine

1=0Ascorbic acidAscorbic acid

1=48

1=45 1=133

Phosphate bufferPhosphate buffer

Figure 6. Free radical release from amphibole asbestos showing the effect of chelator treatments on the potentialfor free radical release. EPR spectra of the DMPO-C02- adduct collected during the F-C test are shown for (A)amosite and (B) crocidolite. EPR spectra of the DMPO-OH adduct collected during the P-OH test are shown for(C) amosite and (D) crocidolite. The intensity of the EPR peak is proportional to the number of free radicalstrapped. Abbreviations: I, intensity (arbitrary units); ND, nondetectable in the experimental conditions used.

nated with all chelator treatments. However,only ascorbate was able to eliminate CO2-generation by crocidolite (Figure 6B).

Interesting results were obtained in theP-OH test (Figure 6C, D). Although OHwas not generated by nontreated amositefibers, incubation in phosphate buffer stimu-lated fiber activity to a level wherebyDMPO-OH' could be detected (Figure 6C).Similarly, enhancement of the phosphatebuffer was observed with crocidolite; the EPR

signal increased 3-fold over nontreated fibers(Figure 6D). Surprisingly, ferrozine alsocaused an increase in 'OH activity forcrocidolite fibers. Desferrioxamine eliminatedall 'OH release from crocidolite as expected;ascorbate caused no change in fiber activity.

Indication of the catalytic activity ofasbestos fibers in the decomposition reac-tion of H202 was assessed by following theconsumption of H202 in aqueous fibersuspensions. The results for nontreated and

Environmental Health Perspectives - Vol 105, Supplement 5 * September 19971026

CHEMICAL CHARACTERIZATION OF CHELATOR-TREATED ASBESTOS

chelator-treated fibers are shown in Figure7A for crocidolite and Figure 7B for amosite.The H202 catalytic activity of both fibertypes is decreased after chelator treatments,with the exception of ascorbate-treatedamosite. Ferrozine treatments have the great-est reduction of fiber activity; desferrioxam-ine and phosphate buffer reduce the activityofboth fibers by approximately one-half.

DiscussionIron Mobilizatonby Chelator Treatments

Chelator treatments are responsible for themobilization of iron from crocidolite andamosite: in the absence of chelators, noiron is removed from these fibers. Asshown in Figure 1A, all chelators used inthis study except ferrozine mobilized moreiron from crocidolite than amosite fiberswhen compared on the basis of fiber

A100-

a 75-o

X 50-Co 5

25-

Original Phosphate Ascorbate Ferrozine Desferm-buffer oxamine

Iron chelator pretreatments

B

a,

0.1

dICDa,~0=-

Original Phosphate Ascorbate Farrozine Desferri-buffer oxamina

Iron chelator pretreatments

Figure 7. Catalytic decomposition of hydrogen perox-ide by (A) crocidolite and (B) amosite and the effect ofiron chelator pretreatments. The percentage of H202decomposed after incubation for 300 min with non-treated or chelator-treated fibers is shown.

weight. However, when normalized by sur-face area, the two fiber types released simi-lar amounts of iron, except in the case offerrozine treatments. In this case, signifi-cantly more iron was mobilized fromamosite than crocidolite. This can beexpected, as ferrozine is Fe2+-specific andamosite contains more surface and bulkFe2+ ions than crocidolite. However, thecorollary was not true: Desferrioxamine isFe3+-specific but did not remove more ironfrom crocidolite than amosite when com-pared on the basis of surface area. Oneexplanation for this contradiction is thatthe chelator solutions were not buffered.Ferrozine becomes acidic under such condi-tions, whereas the desferroxamine solutionis not as sensitive to the absence of buffer.The pH of chelator solutions can influencethe solubility of iron and hence the amountof iron mobilized. It has been shown byLund and Aust (34) that greater quantitiesof iron were mobilized by ferrozine at lowerpH than neutral or basic pH. Because fer-rozine only mobilizes Fe2+ ions, we observedmore iron mobilized from amosite than cro-cidolite due to the larger reserve of Fe2+ inamosite. This would also be the case if Fe2+ions were easier to remove from the silicatestructure than Fe3+ ions.

The two analytical methods used tomeasure the total amount of iron in super-natant solutions give qualitatively similarresults for different treatment solutions. Thelargest discrepancy occurs for ascorbate-treated crocidolite. Previous results alsoshow larger amounts of iron released fromascorbate-treated crocidolite compared toother chelators (25). The general trend iniron release for different chelator solutionsis in agreement with the literature afterconsideration of differences in incubationconditions (e.g., fiber and chelator concen-trations, chelator solvent, pH, temperature,and time) (8,17,25,34).

Our estimation of the amount of ironremoved during chelator treatments is onthe order of 1 to 5 at% of the total iron pre-sent in the fibers. When the amount of ironremoved per unit surface area, the numberof unit cells per unit surface area, and thenumber of iron ions within each unit cellare considered, the iron removed from theseasbestos fibers could have originated withinone to a few unit cell layers below the sur-face. This estimation depends on how manyiron ions were removed from each unit cell(assumed 50-100%). The depth of the sur-face modified region observed on desferriox-amine-treated crocidolite and amosite fibersby Mollo et al. (25), measured by high

resolution transmission electron microscopy(HRTEM), was on the order of 2 to 10 nm.Hence, this surface modified region couldrepresent the depletion of iron, and possiblyother M-site cations, from one to a few unitcell layers below the surface.

Iron Assodated with Asbestos Fibersafter Chelator TreatmentsChelator treatments did not affect therelative concentrations or lattice positions ofFe2+ and Fe3+ ions in the bulk of crocidoliteand amosite fibers. Our average measure-ment of 59% of the total iron in crocidolitein the Fe2+ valence state is in agreement withexpectations based on the theoretical compo-sition (= 55%). However, Pollak et al. (37)measured an Fe2+ concentration of only46% on UICC crocidolite by MS. The rea-son for this disagreement is most likely dueto their use of an extra doublet in the curve-fitting procedure around ±1 mm/sec in theMS spectra, which was assigned to Fe3+ inthe MI lattice site. We could not fit an addi-tional doublet in this region, perhaps due toa better signal-to-noise ratio in our data. Webelieve that iron contributing to the MSspectrum in this region probably resides atthe surface of the fibers or in poorly definedlattice sites, or possibly originates from ahighly mobile iron species, and cannot beclearly identified.

Because of the surface sensitivity of theXPS method, we were able to selectivelyprobe the composition and binding states ofelements only on the surface of the fibers.This corresponds to the modified surfaceregion that was observed by Mollo et al. (25)after chelator treatments. It is not surprisingthat the iron content within the surfaceregion is similar for nontreated crocidoliteand amosite fibers, as their total iron contentsare also similar to each other. However, theFe/Si ratio of 0.8 measured in this and otherrecent studies (Gold et al., unpublished data)is significantly higher than the values ofapproximately 0.6 that we measured on thesame fibers in previous years (31) and thatwere reported by Werner et al. (8). Differentsample preparation and XPS analysis meth-ods can explain this discrepancy (Gold et al.,in preparation). Despite the higher Fe/Siratio, we are still able to follow relativechanges in Fe/Si ratio due to chelator treat-ments. Because the iron content is measuredwith respect to the silicon content, it isimportant to keep in mind that simultaneousand/or faster dissolution of the supporting sil-icate structure will clearly affect measure-ments of surface iron content by XPS, aspostulated by Werner et al. (8).

Environmental Health Perspectives * Vol 105, Supplement 5 * September 1997 1027

GOLD ETAL

We measured reductions of up to 20 to30% in relative surface iron content (Fe/Siratio) on certain chelator-treated crocidoliteand amosite fibers that were stored underinert atmosphere and for short time periods(3 weeks-3 months) after treatments. Insome cases, the change in actual surface ironcontent may be larger or smaller due to pos-sible differences between Fe and Si dissolu-tion rates. Since relative surface iron contenteither decreased or remained unchanged,this is an indication that there is a net loss ofiron from the fiber after most chelator treat-ments. This gives additional support to thedaim that the modified region observed ondesferrioxamine-treated asbestos surfaces byHRTEM (25) is a surface-depleted zonerather than a deposited surface film.

For ferrozine- and phosphate buffer-treated crocidolite, there was no reduction insurface iron content, despite the fact that ironwas released into the chelating solutions. Oneexplanation for this is that, in these two cases,iron ion diffusion from the bulk to the sur-face has occurred. Diffusion probablyoccurred in all chelator-treated crocidolitefibers and in amosite, but only in these twocases (which had the least iron release of thefour chelators) was the initial surface concen-tration attained by the time ofXPS analysis.Another explanation is that there was somereprecipitation of iron from solution to thesurface, but not so much as to cause a netincrease in surface iron content.We would like to note here that in our

previous studies, ho changes in surface ironcontent were observed when chelator-treated fibers were allowed to age for longtime periods (up to 1 year) in air prior toXPS analysis (31). It is known that oxygen-induced segregation takes place in metals.This suggests, again, the possibility of irondiffusion through the subsurface lattice atroom temperature.

The majority of iron at both crocidoliteand amosite surfaces is in the Fe3+ valencestate, despite the fact that bulk concentra-tions of Fe2+ and Fe3+ determined by MSdiffer significantly for the two asbestostypes. This agrees with the findings ofWerner et al. (8) and is probably the result ofoxidation of surface iron due to exposureto air both before and after chelator treat-ments. However, because our fibers werestored in an inert atmosphere prior to XPSanalysis after chelator treatments, we werestill able to detect differences in Fe2+ andFe3+ concentrations at the fiber surfaces. Inour previous studies of chelator-treatedfibers stored in air (31), as in the study byWerner et al. (8), no significant changes in

surface iron oxidation states due to chelatortreatments were observed.

Curve fitting of different valence statesin the iron photoelectron peak is difficultand questionable due to overlapping multi-plet split states for the Fe2+ and Fe3+ ions(31,38,39). However, it can be successfullyused to follow changes in peak shapes dueto chelator treatments, especially if com-parisons with reference spectra are avail-able. We observed significant increases ofup to 10% in surface Fe2+ concentrationson chelator-treated amosite and ascorbate-treated crocidolite. We observed significantdifferences that followed the same trendwith more than one curve-fitting routine(data not shown). However, changes in theoxidation state of surface iron due to chela-tor treatments did not correlate with theamount and type of iron assumed to havebeen removed from the surface. For exam-ple, ferrozine, which specifically chelatesFe2+, mobilized the most amount of ironfrom amosite, yet the concentration of sur-face Fe2+ increased by almost 10% aftertreatment. A possible explanation for this isthat Fe2+ reprecipitates onto the surfacefrom the solution (although no evidence ofthe ferrozine molecule was detected) orthat Fe2+ diffuses up to the surface fromthe subsurface region.A comparison of Figures 1, 3, and 4

reveals no correlation between iron mobiliza-tion into chelator solutions and the amountand valence of iron remaining at the surfaceafter chelator treatments. This agrees withresults from Werner et al. (8), who con-cluded from similar types of experimentsthat it was not possible to correlate XPSresults with iron mobilization data. The pos-sibilities for rapid iron diffusion/migration tothe surface, oxidation of surface iron due toexposure to ambient atmosphere, and simul-taneous erosion of the silicate structure,complicate the analysis and do not allowsuch correlations to be made.

Detection ofAdsorbed ChelatorsEvidence for the adsorption of small quan-tities of desferrioxamine to crocidolite andamosite surfaces includes the following:* A higher N content (1-2 at%) on

desferrioxamine-treated surfaces.Nitrogen is a component of the desfer-rioxamine molecule and is part of thefunctional unit (hydroxamic acid) thatbinds to Fe3+ during chelation (40).

* The high-binding energy C Is photo-electron peak has a peak shift of approxi-mately 3.5 eV from the saturatedhydrocarbon peak, which is smaller than

the peak shift observed on all other sur-faces (=4.2 eV). Organic carbon boundwithin a N-C- 0 type of environ-ment would experience a binding energyshift between 3 to 4 eV, while carbonfunctional units such as O-C-O,O-C 0 would shift typicallybetween 4 and 4.5 eV (41). The latter isgenerally observed on oxide surfacesexposed to air. The observed change inpeak shift on desferrioxamine-treatedsurfaces can be indicative of the presenceof carbon bound in theN-CO typeof environment. This functional groupalso forms the active Fe3+ chelating cen-ter of the desferrioxamine molecule (40).The high-binding energy C is peakalso has slightly greater intensity ondesferrioxamine-treated surfaces com-pared to all other samples, which indi-cates the presence of a greater quantityof organic molecules on this surfacethan on the others.Despite the fact that these XPS peaks

have low intensity (1-2 at%), we feel thatthese three observations are significant andindicate that desferrioxamine molecules haveadsorbed and remained adherent to fibersurfaces. Desferrioxamine bound at surfaceiron sites might prevent the participation ofiron in redox reactions or alter the overallredox potential of the surface. If this is thecase, desferrioxamine treatments wouldreduce the ability of asbestos fibers to gener-ate free radicals and induce tissue damage.

XPS analysis has also shown low levelsof phosphate adsorbed at phosphate buffer-treated fiber surfaces, either as reprecipi-tated iron phosphates or as surface adsorbedphosphate groups. We propose that thisphosphate is associated with the elevatedfiber activity observed in the production of'OH from H202 in the P-OH test.

Catalytic Reactivity ofAsbestosThe measurements of free radical produc-tion by asbestos fibers using EPR and theDMPO spin trapping method resulted inlow signal intensities, as both crocidoliteand amosite produce few radicals unlessthey are ground (17). In general, signalintensities were higher from crocidolite thanamosite, and it could be argued that this isdue to the larger surface area of crocidolite.

In the F-C test, it has been proposedthat an oxidized surface site is reduced dur-ing the reaction [see Fubini et al. (17)]. Inthis study, all chelators decreased or elimi-nated (CO2- ) production in the F-C test,as was also reported by Fubini et al. (17);this effect was most pronounced on all

Environmental Health Perspectives * Vol 105, Supplement 5 * September 19971028

CHEMICAL CHARACTERIZATION OF CHELATOR-TREATED ASBESTOS

treated amosite samples and on ascorbate-treated crocidolite. This could be related tothe diminished quantity of oxidized ironsites at these surfaces (increased Fe2+ con-centration), as measured by XPS analysis,which indicates that Fe3+ is used in thehydrogen extraction reaction with carboxy-late. The usual effect of desferrioxamine,i.e., elimination of CO2- release from cro-cidolite (17), was not observed. This mightbe due to the fact that, although the rela-tive surface iron content was reduced, therelative amount of Fe3+ ions at the surfacewas not altered by chelator treatment(Figure 4B). Hence, there appears to besome correlation between the relative con-centration of Fe3+ ions in the surface regionand fiber activity in the F-C test.

In the P-OH test, it has been proposedthat a reduced surface site is oxidized anddrives the Fenton reduction of H202 tohydroxyl ion and hydroxyl radical (17).Phosphate buffer treatments enhanced theproduction of OH in the P-OH test forboth asbestos types. This behavior is possi-bly influenced by the ability of adsorbedphosphates or precipitated iron phosphatesto enhance reactivity. This seems likely,considering that phosphate strongly cat-alyzes the autoxidation of Fe2+ in solution(42). It is relevant to note that Fe2+ mobi-lization from crocidolite by ferrozine is pre-vented if phosphate buffer is used as thesolvent but is not prevented if NaCl solu-tion is used instead (34). Phosphate buffermight enhance hydroxyl radical productionby enhancing the oxidation of Fe2+ to Fe3+at asbestos surfaces. Note that phosphatebuffer was present in both F-C and P-OHtest solutions.

It is difficult to draw general conclusionsabout the effect of chelators on surface iron,the effects that changes in surface iron haveon the catalytic reactivity of asbestos fibers,and the active sites responsible for fiber reac-tivity. For example, ferrozine treatmentswere most effective in diminishing H202decomposition by crocidolite and amositefibers; however, ferrozine-treated crocidolitefibers had twice as much OH release thannontreated fibers. It is clear that the catalyticdecomposition of H202 does not coincidewith OH (or CO2-f) release, nor is it relatedto the surface iron content or its oxidationstates. The activities found must be ascribedto iron present at the surface when thesemeasurements were performed, assumingthat iron is the active constituent of themineral. We can conclude that fiber activitymust be dependent on iron that has attainedparticular locations at the crystal surface, or

surface iron that has bound a chelator mole-cule (thus exhibiting a different redoxpotential), or on any iron that might havediffused from the near-surface region to thesurface in the elapsed incubation time.

SummaryUnder our experimental conditions, weestimate that 1 to 5 at% of the total ironcontained in asbestos fibers was released intosolutions of desferrioxamine, ascorbate, fer-rozine, and phosphate buffer; iron mobiliza-tion was least in phosphate buffer solutions.There was no apparent correlation betweenthe amount of iron mobilized by chelatortreatments and the amount of iron remain-ing at fiber surfaces, which either decreasedor remained unchanged compared to non-treated fibers. Although there was no signifi-cant change in the relative concentration ofbulk Fe2+ and Fe3+ due to chelator treat-ments, the relative concentration of surfaceFe2+ species increased on ascorbate-treatedcrocidolite and all chelator-treated amositesamples. Hence, iron mobilization data can-not be used to predict XPS results, and viceversa. However, XPS results have indicatedthat low levels of desferrioxamine and phos-phate remained bound to both crocidoliteand amosite surfaces.

Most, but not all, chelator treatmentsreduced the reactivity and catalytic activityof crocidolite and amosite asbestos.However, phosphate buffer treatmentsenhanced hydroxyl radical production byboth crocidolite and amosite, as did fer-rozine treatment of crocidolite. The usualeffect of desferrioxamine, the elimination offree radical release from crocidolite, was notobserved in the case of CO2- generation.However, the interpretation of the free rad-ical release data may be complicated by lowsignal intensities.

Our results indicate that Fe3+ is impor-tant for the hydrogen extraction reactionfrom carboxylate functional groups. Noother correlations were found between ironmobilized by chelator treatments, irondetected by XPS at the treated surface, andthe effect of iron chelators on chemicalreactivity of crocidolite and amosite asbestos.We propose that iron mobility in the sili-cate structure may play a larger role in thechemical reactivity of asbestos fibers thanpreviously assumed.

ConclusionsX-ray photoelectron spectroscopy hasconfirmed that iron chelators can modifythe amount and valence states of iron atthe surface of crocidolite and amosite

asbestos in addition to mobilizing iron fromthe fibers. However, we found no direct cor-relation between the amount of ironremoved from fibers and changes in surfaceiron content (Fe/Si ratio). Observationscould be complicated by: a) iron ion diffu-sion from the bulk to the surface, b) pre-cipitation of iron complexes from thesolution, and c) simultaneous dissolutionof the supporting silicate structure. Thevalence state of iron found at chelator-treated fiber surfaces does not reflect theamount of Fe2+ or Fe3+ mobilized intosolution, as assumed by the specificity ofeach chelator for iron in a particular oxi-dation state. This can be expected becauseof the reactivity of surface iron with oxy-gen when exposed to air, although it doesnot explain all of our observations. Wecannot exclude the possibility of ironmigration from subsurface layers and theprecipitation of iron or iron complexesfrom the solution.

Few correlations were found betweenfree radical release and H202 decomposi-tion activities of iron chelator-treatedfibers, and between these quantities andthe amount of iron removed, the amountof iron remaining at the surface, and itsoxidation state. Interpretations becomecomplicated by adsorbed chelator species:Desferrioxamine and phosphate detectedby XPS at fiber surfaces may play a role indiminishing and enhancing, respectively,iron activity in hydroxyl radical produc-tion. It is most likely that the catalyticactivity of asbestos fibers is governed byseveral factors, such as iron mobility throughthe structure and the arrangement of ironions at the surface, rather than the totalquantity of iron at the surface and itsoxidation state as analyzed by XPS.

REFERENCES

1. Lund LG, Aust AE. Mobilization ofiron from crocidolite asbestos by cer-tain chelators results in enhanced croci-dolite-dependent oxygen consumption.Arch Biochem Biophys 287(1):91-96(1991).

2. Lund LG, Aust AE. Iron mobilizationfrom crocidolite asbestos greatlyenhances crocidolite-dependent forma-tion of DNA single-strand breaks in4X174 RFI DNA. Carcinogenesis13(4):637-642 (1992).

3. Gulumian M, Bhoolia DJ, TheodorouP, Rollin HB, Pollak H, van Wyk JA.Parameters which determine the activityof the transition metal iron in crocidoliteasbestos: ESR, Mossbauer spectroscopic

Environmental Health Perspectives * Vol 105, Supplement 5 - September 1997 1029

GOLD ET AL.

and iron mobilization studies. S Afr Tydskr Wet 89:405-409(1993).

4. Ghio AJ, Kennedy TP, Stonehuerner JG, Crumbliss AL,Hoidal JR. DNA strand breaks following in vitro exposure toasbestos increase with surface-complexed (Fe3+). Arch BiochemBiophys 31(1):13-18 (1994).

5. Lund LG, Williams MG, Dodson RF, Aust AE. Iron associatedwith asbestos bodies is responsible for the formation of singlestrand breaks in (p X174 RFI DNA. Occup Environ Med51:200-204 (1994).

6. Fubini B, Mollo L. Role of iron in the reactivity of mineralfibers. Toxicol Lett 951:82-83 (1995).

7. Hardy JA, Aust AE. The effect of iron bindin on the ability ofcrocidolite asbestos to catalyze DNA singTe strand breaks.Carcinogenesis 16(2):319-325 (1995).

8. Werner AJ, Hochella MF, Guthrie GD, Hardy JA, Aust AE,Rimstidt JD. Asbestiform riebeckite (crocidolite) dissolution inthe presence of Fe chelators: implications for mineral-induceddisease. Am Mineralogist 80:1093-1103 (1995).

9. Kamp DW, Graceffa P, Pryor WA, Weitzman SA. The role offree radicals in asbestos-induced diseases. Free Radic BiolMedicine 12:293-315 (1992).

10. Gulumian M, Bhoolia DJ, du Toit RSJ, Rendall REG, PollakH, van Wyk JA, Rhempula M. Activation ofUICC crocidolite:the effect of conversion of some ferric ions to ferrous ions.Environ Res 60:193-206 (1993).

11. Moyer VA, Cistulli CA, Vaslet CA, Kane AB. Oxygen radicalsand asbestos carcinogenesis. Environ Health Perspect102(10):131-136 (1994).

12. Schapira RM, Ghio AJ, Effros RM, Morrisey J, Dawson CA,Hacker AD. Hydroxyl radicals are formed in the rat lung afterasbestos inst illation in vivo. Am J Respir Cell Mol Biol10:573-579 (1994).

13. Gilmour PS, Beswick PH, Brown DM, Donaldson K. Detectionof surface free radical activity of respirable industrial fibres usingsupercoiled OX174 RFI plasmid DNA. Carcinogenesis16(12):2973-2979 (1995).

14. Kamp DW, Israbian VA, Preussen SE, Zhang CX, WeitzmanSA. Asbestos causes DNA strand breaks in cultured pulmonaryepithelial cells: role of iron-catalyzed free radicals. Am J Physiol268:L471-L480 (1995).

15. Weitzman SA, Graceffa P. Asbestos catalyzes hydroxyl andsuperoxide radical generation from hydrogen peroxide. ArchBiochem Biophys 228(1):373-376 (1984).

16. Zalma R, Bonneau L, Guignard J, Pezerat H, Jaurand M.Formation of oxy radicals by oxygen reduction arising from thesurface activity of asbestos. Can J Chem 65:2338-2341 (1987).

17. Fubini B, Mollo L, Giomello E. Free radical generation at thesolid/liquid in iron containing minerals. Free Radic Res23:593-614 (1995).

18. Fubini B. Use of physico-chemical and cell-free assays to evalu-ate the potential carcinogenicity of fibres. In: Mechanisms ofFibre Carcinogenesis (Kane AB, Saracci R, Boffetta P, WilbournJD, eds). IARC Sci Publ No 140. Lyon:International Agencyfor Research on Cancer, 1996 35-54.

19. Vallyathan V, Mega JF, Shi X, Dalal NS. Enhanced generationof free radicals from phagocytes induced by mineral dusts. AmJ Respir Cell Mol Biol 6:404-413 (1992).

20. Weitzman SA, Weitberg AB. Asbestos-catalyzed lipid peroxida-tion and its inhibition by desferroxamine. Biochem J225:259-262 (1985).

21. Turver CJ, Brown RC. The role of catalytic iron in asbestosinduced lipid peroxidation and DNA-strand breakage inC3H1OT1/2 cells. BrJ Cancer 56:133-136 (1987).

22. Kennedy TP, Dodson R, Rao NV, Ky H, Hopkins C, BaserM, Tolley E, Hoidal JR. Dusts causing pneumoconiosis gener-ate OH and produce hemolysis by acting as Fenton catalysts.

Arch Biochem Biophys 269(1):359-364 (1989).23. Goodglick LA, Kane AB. Cytotoxicity of long and short croci-

dolite asbestos fibers in vitro and in vivo. Cancer Res50:5253-5163 (1990).

24. Faux SP, Howden PJ, Levy LS. Iron-dependent formation of 8-hydroxydeoxyguanosine in isolated DNA and mutagenicity inSalmonella typhimurium TA102 induced by crocidolite.Carcinogenesis 15(8): 1749-1751 (1994).

25. Mollo L, Merlo E, Giamello E, Volante M, Bolis V, Fubini B.Effect of chelators on the surface properties of asbestos. In:Cellular and Molecular Effects of Mineral and Synthetic Dustsand Fibers, Vol H85 (Davis JMG, Jaurand M-C, eds).Berlin/Heidelberg:Springer-Verlag, 1994;425-432.

26. Weitzman SA, Chester JF, Graceffa P. Binding of desferoxam-ine to asbestos fibers in vitro and in vivo. Carcinogeriesis9(9):1643-1645 (1988).

27. Graf E, Mahoney JR, Bryant RG, Eaton LW. Iron-catalyzedhydroxyl radical formation. J Biol Chem 259(6):3620-3624(1984).

28. Pezerat H. Role of transition metal compounds in the capacityof dusts to generate electrophilic species. In: Cellular andMolecular Effects of Mineral and Synthetic Dusts and Fibers,Vol H85 (Davis JMG, Jaurand M-C, eds). Berlin/Heidelberg:Springer-Verlag, 1994;359-368.

29. Hardy JA, Aust AE. Iron in asbestos chemistry and carcinogenic-ity. Chem Rev 95(1):97-118 (1995).

30. Timbrell V. Characteristics of the International Union AgainstCancer standard reference sample of asbestos. In:Pneumoconiosis: Proceedings of the International Conference,April 1969, Johannesburg, South Africa. (Shapiro HA, ed). CapeTown, South Africa:Oxford University Press, 1970; 28-36.

31. Amandusson H. XPS Analysis of Asbestos Fibers [M.S. Thesis].Chalmers University of Technology, Gothenburg, Sweden,1995.

32. Hawthorne FC. Crystal chemistry of the amphiboles. In:Amphiboles and Other Hydrous Pyriboles-Mineralogy,Reviews in Mineralogy, Vol 9A (Veblen DR, ed).Washington:Mineralogical Society ofAmerica, 198 1;1-102.

33. Timbrell V, Rendall REG. Preparation of the UICC standard ref-erence samples of asbestos. Powder Tech 5:279-287(1971/1972).

34. Lund LG, Aust AE. Iron mobilization from asbestos by chela-tors and ascorbic acid. Arch Biochem Biophys 278(1):60-64(1990).

35. Ericsson T, Waippling R. Texture effects in 3/2-1/2 Mossbauerspectra. J Phys Coll 37:719-723 (1996).

36. Mottola HA, Simpson BE, Gorin G. Absorptiometric determi-nation of hydrogen peroxide in submicrogram amounts withleuco crystal violet and peroxidase as catalyst. Anal Chem42:410-412 (1970).

37. Pollak H, Waard Hd, Gulumian M. Mossbauer spectroscopicstudies on three different types of crocidolite fibres. S Afr TydskWet 89:401-404 (1993).

38. Brundle CR, Chuang TJ, Wendelt K. Core and valence levelphotoemission of iron oxide surfaces and the oxidation of iron.Surface Sci 68:459-468 (1977).

39. McIntyre NS, Zetaruk DG. X-ray photoelectron spectroscopicstudies of iron oxides. Anal Chem 49(11):1521-1529 (1977).

40. Keberle H. The biochemistry of desferrioxamine and its relationto iron metabolism. Ann NYAcad Sci 119:758-768 (1964).

41. Beamson G, Briggs D. High Resolution XPS of OrganicPolymers: The Scienta ESCA300 Database. Chichester,England:John Wiley & Sons, 1992.

42. Lambeth D, Ericson G, Yorek M, Ray P. Implications for invitro studies of the autoxidation of ferrous ion and the iron-cat-alyzed autoxidation of dithiothreitol. Biochim Biophys Acta719:501-508 (1982).

1030 Environmental Health Perspectives - Vol 105, Supplement 5 * September 1997