Document pMQgdpj9YwOww6wEE84wk6Ja

EDITOR Seath : that e data 2 any loyees in the of talc in 9.6a of ge Perals gamon gamon chouler Mines ; ers and 17 A ets and : Ratc 3. National National Smith - Natrics e and jersey School School Jersey . . LETTER TO THE EDITOR S = .. . | MINERALOGICAL CHARACTERIZATION OF COSMETIC TALC PRODUCTS e ; . , i | The entitled purpose of this letter Consumer Talcums is to comment on the and Powders Mineral paper by Rohl et al and Chemical Char- acterization which appeared in the Journal of Toxicology and Environ- mental Health 255-284 1976 The paper conveys an erroneous impression that it is reasonably common to detect the presence of asbestos in counter talcum powders The authors imply in fact that half of the products studied are of questionable quality when they state that 10 of the 20 products examined contained detectable amounts of tremolite and , anthophyllite - + principally asbesti The article is misleading and its publication demands clarification in light of the following points 1 a comprehensive understanding of mineralogy makes it apparent that the reported results are impossible to achieve 2 the analytical methods described do not and can not lead to the conclusions reported and 3 the conclusions drawn are without sound scientific basis and are therefore misleading and invalid These points are discussed below The portions of the paper by Rohl et al that describe the geologic generally occurrences and properties of talc and associated minerals are correct with one major exception The paper very clearly makes the point that the predominant occurrences of tremolite anthophyllite and serpentine are as the asbestiform varictics and readers other than . -professional geologists or mineralogists will almost surely reach the ' - erroneous conclusion that occurrences of the asbestiform varieties are the rule rather than the exception It is a fact that the asbestiform varietics of these minerals are relatively rare and that such occurrences probably account for less than % of the known world occurrences of each mineral Further tremolite and anthophyllite are erroneously stated to have fibrous cleavage whereas in fact their cleavage is prismatic Tremolite asbes. tos is essentially a mineralogical curiosity The analytical methods described by the authors for identification and quantification of tremolite anthophyllite and serpentine are invalid The 1223 Journal of Toxicology and Environmental 1223-211222236-1226 1977 Copyright '1977 by Hemisphere Publishing Corporation 1 pe 0052292 JNJ 000343099 : 1224 LETTER TO THE EDITOREDITOR ray diffraction XRD methods described cannot be used to identiis the Saited amphibole species present to distinguish serpentine from chlorite or the quantify serpentine or quartz because of interferences from chlorite and talc respectively Sh ; a! - Unequivocal identification and quantification of the amphibole species present on the basis of the tremolite 8.38 A 110 peak and the anthophyllite 8.26 A 210 peak is impossible Examination of the powder diffraction file published by the joint Committee cn Powder Diffraction Standards JCPDS for amphiboles 52 amphibole cards in the JCPDS file reveals that all 52 have their 110 or 210 diffraction peaks within 0.2 A of each other For Cuka radiation a change in spacing of 0.2 ^ at 10.5 20 the region step scanned is -11 20 is equivalent to a 20 angular . change of only 0.25 Slight compositional and structural variations in the amphiboles readily cause small shifts in the position of the 110 and 210 peaks Similarly very small changes in the position and nature of the sample surface will cause small peak shifts As a result of general peak overlap among all amphiboles unpredictable peak shifts and mismeasurement of peak position it is not possible to identify any specific amphibole on the basis of the 110 or 210 peak even if the sample is 100 pure amphibole of one variety It is clearly impossible to identify and quantify trace to minor amphiboles by the methods described in this paper Bulk chemical analyses for major minor and trace elements has questionable value in the identification of an amphibole present in tracteo minor amount The authors used the serpentine 004 peak at 3.66 A 24.32 20 for Cuka for identification of serpentine Chlorite has a major diffraction peak 004 in this region It is well known that chlorite and serpentine are difficult to differentiate by all known methods of characterization but that their 004 XRD peaks are separate enough to generally allow unambiguous determination of the presence of both phases when they are present in quantities sufficient to give definable peaks However the 004 peak of chlorite varies between 3.53 and 3.60 A -25.23 20 for Cuka depending on the particular compositional and structural variety present To attempt quantification of trace serpentine in talc by using the 004 peak is not possible since talcs almost invariably contain chlorite far in excess of serpentine and chlorite will therefore interfere The authors acknowledge that Stanley and Norwood were confronted with the chlorite- serpentine interference problem but they still claim to have been able to use XRD for identification and quantification of trace serpentine in the ; presence of chlorite samples 12 and 15 In addition if any of the clay minerals kaolinite halloysite or dickite are present they will also interfere with the XRD identification and quantification of serpentine The XRD peaks of serpentine chlorite kaolinite dickite and halloysite are typically broad and often poorly resolved further complicating identification and almost certainly eee eed 0052293 JNJ 000343100 DITOR \ the or to 2 and cecies I the wder Action ; filc 0.2 10.5 gular ins in ) and ire of eneral and any f the ble to the tra ibole 0 for ction le are 1 but allow y are 004 6 for ariety g the te far thors oritele to > the ckite and orite porly ainly LETTETRO THE EDITOR , 1225 eliminating any possibility of quantifying trace amounts of serpentine by "+ XRD if any of these interfering phases are present An additional complication is presented by halloysite if it is present " Halloysite has a tubular structure similar to that of chrysotile and _ therefore cannot be distinguished from chrysotile by electron microscopy The authors used the quartz 211 peak at 1.54 A 60.08 20 for Cuka for quantification of quartz in spite of the fact that talc has one " large peak and several small peaks in this 29 region that can interfere The of - problem involves quantification trace to miner amounts of quartz by using a weak quartz peak 211 relative intensity 15 that is interfered with by a strong talc peak 060 relative intensity 55 A recent ray diffractogram of very pure talc run in this laboratory had the angle shoulder of this talc peak extending to 59.7 20 for example completely overlapping the 211 quartz peak position Quantification of trace to minor amounts of quartz on the basis of its 211 peak is clearly not possible in the presence of major amounts of talc The interference and identification problems discussed above become apparent when examining errors associated with the linear equations of the form y = + bx given by the authors to correlate peak area with percent mineral content table 3 From table 3 comparing the known mineral content with the calculated mineral content using peak areas and equation of fit the error for each data point can be calculated The stated detection limits of 2.0 0.7 1.4 and 0.1 for anthophyllite chrysotile quartz and tremolite respectively have very little meaning when the errors for each data point are examined The detection limits of 2 and 1.4 for anthophyllite and quartz are less than the lowest standard observed prepared % and are therefore strictly theoretical rather than Using the authors calibration equation for the quartz the calculated valuc for the % standard is an overestimation of the true level by 56 At the stated detection levels for chrysotile and tremolite 0.7 and 0.1 respec- tively the calculated values from the calibration equations are overestimations of the true levels by at least 100 for chrysotile and 500 for tramolite Linear calibrations for these types of samples and concentration ranges are not ordinarily expected and the use of linear equations to fit the data is questionable especially in view of the previous discussion of interferences and invalid procedures For example the data for chrysotile are distinctly nonlinear figure2 but the authors still use a linear equation to fit the data I also wish to point out that the quotation from the paper by Stanley 2nd Norwood is a paraphrase rather than a direct quotation Also there is an error in the quoted fiber size of 0.5 X 0.2 ...mrequiring transmission electron microscopy size stated by Stanley and Norwood is 0.5 ^ 0.02 m . ems The entire paper reflects a strong bias to assume that all serpentine or and amphibole minerals are asbestiform The fact that the authors added Sme te aa 0052294 JNJ 000343101 * 1226 LETTER TO THE EDITOR EDITOR - asbestos minerals to talc in the preparation of their standards is by no means an assurance that the samples later examined also contained asbestos Use of such standards has the effect of prejudging the samples . before they are ever examined and in this case has led to impossible . . results and invalid conclusions In the above discussion I have described some of the errors and invalid procedures of mineralogical identification and quantification as used by the authors It is therefore apparent that the stated major purpose of their paper namely to determine the mineralogical and chemical composition with particular emphasis on the quantitative determination of _ tremolite anthophyllite serpentine minerals and quartz has not been achieved Accordingly the results reported are without analytical basis and the conclusions drawn are invalid and misleading Jerome B. Krause Colorado School of Mines Research Institute . Golden Colorado GER wey a. : a Te eT wae ~ Sle cae oe aie ta an ee ol 0052295 JNJ 000343102