Document x1rmxvVOZq26KyQ8eXgx3JnN1

COLGATE- PAL MOL I VE COMPANY Research and Development Department 909 River Road, Piscataw ay, N. J. 08854 CHRISTOPHER H. COSTELLO A ssociate Director of Research September 6, I973 Dr. Robert Schaffner Deputy Director - Office of Technology Eureau of Foods - FDA Building 8 Food <pnd Drug Administration 56OO Fishers Lane Rockville, Maryland 2052 Dear Dr. Schaffner: Ever since the question of asbestos in talc came up approximately two years ago, cur Analytical people at Colgate have been study ing the problem in an effort to determine meaningful methods of detection of asbestos. Our people have had several conferences with Dr. Lewin and other experts in this area prior to launching this program. In addi tion our company has acquired a step-scanning capability for our X-ray diffraction apparatus. Yesterday J. received the final re port on the development of a method for the determination of cry- sotile asbestos in talc which.is estimated to.have a lower limit of detection of 0.2$. ( I note from the PUfK SHEET of A.ugust 27 that FDA has acquired the services of Dr. John W. Stuart to undertake a program to try to identify asbestos in talc through both differential thermal analy sis and X-ray diffraction. It occurs to us that the work of our Dr. Simko and his associates would be most helpful to Dr. Stuart in his work and we are therefore sending several copies of our research report on this subject in the hope that you will fine it useful. We intend to publish this work eventually,but in the interim we would be happy to have Dr. Simko discuss his work with Dr. Stuart, or whomever you might designate. Dr. Simko will be willing to go to Washington for this purpose. We would welcome Dr. Stuart here also. Until a more sensitive method is determined, it is our intention to use this method as a basis for establishing standards for Colgate talc. WnUMSfc c Dr. Robert Schaffner 2 September 6 , 1973 Please let me near from you in regard to this matter. I am sending a copy of the research report to Dr. Alfred Weissler as well be cause of his involvement in this problem in the past. I trust you will both find the report of interest. Very truly yours, CHC:il Enc. cc: Dr. Alfred Weissler C. H. Costello C c Colgate-Palmolive Company Research and Development Department Piscatavay, New Jersey August 15, 1973 9 Research Report 2675 (Analytical Method) DETERMINATION OP CERYSQTILE ASBESTOS IN TALC BY X-RAY DIFFRACTION Author: Pasquale Brscese Analytical Section PC 1*7 RPA 62130 Referenced Lab Notebook: 5075 Previous Research Report: None Ill ABSTRACT An x-ray diffraction technique for the determination of chrysotile asbestos in talc is described. Quantitative analysis is made by measuring the area under a selected diffraction peak using the step-scan procedure. The (002) chrysotile peak d = 7.3 A0 is ideal for this measurement; however, the talc (U.C. Regal Talc used in this study) contains Kaolinite (00T) line d = 7*18 A .as a contaminant which interferes vith this chrysotile line. Therefore, the (00U) line, d = 2.6k A0 , of chrysotile was used. To compensate for its veaker intensity, a longer exposure for each step was used. This was statistically calculated0to be 80 seconds. The method is estimated to have a lower limit of detection of 0.2$ chrysotile in talc. iv TABLE OF CONTENTS ABSTRACT...... ............... ........ ............... iii LIST OF ILLUSTRATIONS.................. v LIST OF TABLES......... DIGEST OF REPORT.......... v 1 EXPERIMENTAL DETAILS ............... 2 1. 2. * 3 U. 5. 6. 7. Instrumental Conditions....... 2 Estimation of Expected Intensities.... ... ,, 2 Procedure................ L Calibration Curve............... U Analytical Error Due to Counting Statistics 5 Variation of Standard Deviation "with Counting Time............ 7 Lower Limit of Detection (L.L.D.).......... 9 RESULTS......... 10 1. Reproducibility and Precision of the Instrument...... 2, Recovery Study............. .10 .10 DISCUSSION............................. CONCLUSION AND RECOMMENDATIONS......... ............ 12 INDEX TERMS..... ..................................... 13 REFERENCES.................................. 11 lU t V LIST OF ILLUSTRATIONS Figure 1 Working Curve: Intensity Net Counts in O.U Angle vs. Concentration, Weight % Chrysotile Figure 2 Variation of Standard Deviation With Counting Time f *^ 9 LIST OF TABLES Table 1 Standard Component s1 Weights in Grams Table 2 Results: Intensity-Net Counts in 0.i Angle vs. Concentration Weight % Chrysotile Table 3 Analytical Error Due to Counting Statistics Table h Reproducibility and Precision of the Instrument Table 5 Recovery Study v ;; t L .r \ ^: c C c r V f DIGEST OF REPORT Introduction --- - The characteristics of x-ray diffraction patterns provide a specific method for the identification of crystalline compounds. This technique can he applied either qualitatively or quantitatively to the examination of solid materials. The allegation that commercial samples of talc contain asbestos minerals contaminants which could he hazardous to health when inhaled led to the development of this method. Major interest is focused on chrysotile asbestos, a fibrous variety of serpentine, which shews major diffraction peaks at d = 7*3 A and d = 3 .6U A while the amphibole group, another variety of asbestos, shows a maximum at d = Q.b A. (1,2) This report presents details for the quantitative analysis by x-ray diffraction of chrysotile asbestos in talc at the (00*O line corresponding to the major diffraction peak at d = 3.6^ A . ' -$*> - 2- * EXPERIMENTAL DETAILS 1* ' Instrumental Conditions Apparatus: Philips Electronics Instruments x-ray generator, type No. 12215/0, vas used in combination with a high angle vertical goniometer, scintillation counter and step-scanner, -- GuK^ radiation, Nickel filtered, generated at 35 kv and 18*Ma, Slit system: divergence scatter receiving *I o O.OOo inch 1 Signals from the detector vere fed through a pulse height descriminator and the integrated counts per step printed out on a teletypewriter. The step-scanning arrangement permitted the measurement of small integrated areas of 0.01 per step. A total of HO steps covering an angle of O.U 2 theta from 2^.20 up to 2k.60 vas used for peak area, and an angle of 0.2 2 theta from 23-50 up to 23-70 and 26.15 up to 26,35 vas used tor background. Ip. every sample prior fto counting for quantitative analysis, a qualitative scan is necessary over the angle from 23 up to 27 2 theta to ensure that the area considered for background is free from interferences, I , Counting time per step vas 80 seconds and total counting time 80 x 0 = 3,200 seer .ids for peak area and 3,200 seconds for background. (6 ) Reagents*: !f North arolina Regal Talc from Charles Mathiev, Inc. Chrysotile Asbestos obtained from Whittaker, Olaik and Daniels, Inc. * Both materials vere supplied through the help of the Skin Products Section. -( 2 . Estimation of jgxpected Intensities Talc, a natural hydrous magnesium silicate vith the formula Mg^Si^O (0H)p crystallographically is monoclinic and triclinic 'in form! Geologically, the mineral occurs in rock masses and co-exists with! a large number of other hydrated magnesium mineral species--- -- - C - 3- Chrysotile, a naturally*fibrous mineral, is a hydrated silicate of magnesium -which approximates in composition the formula Hi Mg Si 0 , The structure is monoclinic, four molecules of H^Mg Sip O ^in^the ^ unit cell. Each silicon atom is surrounded by four oxygen atoms arranged at the point of a tetrahedron. Together these groups form chains extending parallel to the vertical axis. The binding forces between these chains are weak, thus accounting for the fibrous structure. It includes most of the silky amianthus of serpentinerocks and much of what is popularly called asbestos, (h) Qualitative examination by x-ray diffraction of N.C. Regal Talc showed mainly talc, with the following minor constituents: chlorite, kaolinite, phlogopite, dolomit and alpha quartz. No detectable quantities of asbestiform minerals, either the amphibole or serpentine group were seen, (2 ) Of the above constituents, none absorbed Copper K, radiation strongly in relation to talc, this is known from'study of the appropriate chemical elements constituent to each compound. So, absolute counts versus concentration of chrysotile should give a straight line. The intensity of the (OOU) line of pure chrysotile d = 3.6^ A gave 5^+0 counts/second net by step-scanning. (in terms of counts per second per percent nis = 5.^+0 c/s/Je.) Calculation of mass absorption coefficients for Cu K radiation gave 32.07 for talc and 29.88 for ichrysotile. The following equation gives the c/s/% expected in a mixture of talc and chrysotile. - /* 32.07 S. 02 .rid 5.03 Practically no corr "ction is needed due to absorption in the mixture of talc and chrysotile. (3,5) The value obtained for the mass absorption coefficients are based on the formulas for talc, Mg Si, 0 (OH) , and chrysotile Mg SipCVCOH)^. (Card No. 19-770 for talc an* 9-^** for chrysotile edited by th Joint Committee on Powder Diffraction Standards.)* L f # Joint Caminititee on JCPDS Powder Diffraction Standards ^ X601 Park Lane, Swarthmore, Pennsylvania 19081 c 3. Procedure - k .* No special sample preparation was necessary for the talc. North Carolina Regal talc is already finely-ground to pass a 325 mesh sieve, microns opening, as received. The chrysotile asbestos obtained from Whittaker, Clark & Daniels, Inc., #1*1 0 , was partially ground and needed further grinding to diminish the particle size to an average of 50 microns. This was accomplished by 2 hours grinding in a Waring Blender.* The ground fibers were checked by microscope for size. * The blending assembly can be obtained from Waring Products Service * Center, New Hartford, Connecticut 06057s Model No. 5018. An exactly-known weight of chrysotile was added to the talc. The mixture, tightly closed in a weighing bottle, was placed in another jar and insulated with tape in the jar to avoid breakage. This jar was placed in a mixer /mill* and each sample mixed for 1 hour with the help of two plastic balls. * Spex Model 1810, Spex Industries, Metuchen, New Jersey. Each sample, after be .ig properly mixed, was loaded in a flat aluminum x-ray sample.holder. Care was taken to ensure proper loading of the powderi To minimize orientation effects and to standardize the procedure for each sample, t h e >following three steps were found necessary: 1. Load the sample holder and press the powder with a spatula. 2. Remove the excess with a Gillette razor blade. 3 . Make the surface flat and smooth by pressing with a g~ourd glass plate. Finally, place the sample in the x-ray instrument and take counts over the peak and background areas. r 1*. Calibration Curve ------------- !---- ' I Chrysotile can be detected with a high degree of assurance by a qualitative x-ray scan when present at approximately 5# and higher. Consequently, a calibration curve covering the range from 1 to 5% only was necessary. Samples used for calibration are listed in Table _1_._ . -- IT - c Table 1 Standard Components*Weights in Grams Talc Chrysotile Tot al % Chrysotile 2.1(75 2.1*50 2.1*25 2.1+00 2.375 0.025 0.050 0.075 0.100 0.125 2.500 2.500 2.500 2.500 2.500 - 1.0 2.0 3.0 1*.0 5.0 Results obtained by x-ray analysis are listed in Table 2 and plotted in Figure 1. Table 2 Results: Intensity Ret Counts in 0 A Angle vs. Concentration in Weight % Chrysotile Concentration Weight % Chrysotile Intensity Net Counts Q.k Angle 1.0 2.0 3.0 1+.0 5.0 1 26.1*00 39.100 61.870 86.1+30 102,550 Ret Counts = Total Counts (Peak Area - Background Area) !u Using the method of least squares, the following equation was found for the relationship'between intensity and concentration. y = intensity (net counts) x = concentration in vt. % chrysotile t... Equatior of the line: y = 3381 + 19-9^3 x. 5. Analytical Error Due to Counting Statistics (5) A series of measurements for 1, 2, 3, 1+ and 5%'chrysotile in talc gave the following counts. Table 3 columns 2 and 3. e The standard deviation (?) in percent of a measurement: involving a given number of counts at peak area Kp and a given number of counts at background area Kb is G ($) = 100 iM. Kp + - Kb Kb In Table 3, 2CT is shown as counting error and in terms of % chry sotile, columns 4 and 5 Table 3 Analytical Error Due to Counting Statistics Concentration Kp -- Counting Error '26 Kb Error 2( % Chrysotile 1% 353,080 326,1+1+0 6.2% 0.062% 2% 360 ,Tl+0 321,61+0 k.2% O.OQH! 3% 386,390 32l+ ,520 2.1% 0 .0 7 6 % k% 1+29,190 31+2,760 2.0% 0.080$ 5% +33,390 330,81+0 1 .1 % 0.085$ Variation of Standard Deviation With Counting Time (5,6) A series of measurements of a sample containing 1% chrysotile in talc gave a total count 353,080 for the peak area and 326,1+1+0 for the background area. (1+0 steps, 80 seconds per step). Taking the same number of steps, i.e., 1+0, and different counting time, i.e., 5, 10, 20, 1+0. 60, 80, 100 and 200 seconds per step, the following equations are applicable. Total Counting Time x 353^080 = Counts for peak area = Kp Tot a l C_q ^ _t_i n^___T]_me_ x 326,1+1+0 = Counts for background = Kb,. C.(JU v where: \fKp + Kb G{%) = 100 Kp - Kb From the data obtained, a curve is constructed in Figure' 2 showing the standard deviation as a function of counting time. Ko great advantage occurs by increasing the counting time above 80 seconds. rii## 7 . Lover Limit of Detection (L.L.D.) (5,6) The lover limit of detection is normally defined as that concen tration vhich gives a count rate equivalent to a background reading plus tvice the standard deviation of the background, This assumption will be correct in 95% of cases studied {9 5 % confidence limit). where: Fb = 326,1+1+0 net counts of background /Fb~ = 571.35 counts for 1(6') 2^Fb = 111+2.70 counts for 2(6) Equation of the line y = 3381 + 19,967 x x Let y = 0 * x = -0.17$ Let y = 111+3 counts = 2(Tabove background x = -0.11# L.L.D. = = x at y = lll+3 - x at y = 0 = 0.06# As a figure of merit, it is better to multiply the L.L.D. by 3, where 3 is a factor vhich corrects,for instrumental drift, Therefore, L.L.D. = + 0.18# chrysotile. V/r- . r RESULTS Reproducibility and Precision of the Instrument The reproducibility of the x-ray instrument vas checked by daily analyses on a standard for 6 consecutive days. The standard used contained 5$ chrysotile. Table b Reproducibility and Precision of ftfe Instrument Standard: 5% Chrysotile in Talc Pay Intensity: Net Counts O.U -Angle % Chrysot; 1 102,550 2 102,230 3 103,979 k 100,039 5 9 7 ,2 6 k 6 101,55k 5.13 ? 5 .1 2 % 5 .2 1 % 5.01 % U .87^ 5.08$ Av. = 5 .07% SD = + 0.12 RSD = 2.1% Recovery Study A sample of M6670 *50 finished product Cashmere Bouquet Talc was checked for chrysotile with the present method and gave zero % chrysoti]e. Chrysotile was added to it for a recovery study. The results are listed in Table 5* Table 5 Total Weight in Grams of Talc and Chrysotile Recovery Study < r % Chrysotile Added % Chrysotile ' Found % Recovery 2.5 0.2 0.33 165 2.5 0.5 0.6l 122 2.5 1.0 0.93 93 2.5 2.0 1.73 86 2.5 5.0 b<9 98 The very high recovery at the 0.2$ level and the jmore acceptable results at the 0.5% level, limit the technique to the order of 0 - 11 - * DISCUSSION The method presented can he applied satisfactorily to detect chrysotile asbestos in talc. It was developed especially for the evaluation of chrysotile asbestos. In principle, however, it can be applied to any other type of asbestos subject to the limitation of producing suitable calibrations. The lower lii :.t of detection was calculated as 0.2%. Further improvement of the detection limit for this method will require the use of a more powerful x-ray tube, generator andv improved revolution provided by the use of a crystal monocromator, It was an unfortunate case that talc invariably contains , among other contaminants, kaolinite, making it difficult to use the more intense line (002) of chrysotile due to the proximity of kaolinite, 7*18 A , and chrysotile, J.3 A lines. An early attempt to use this line after thermally destroying kaolinite at 500C proved that along with kaolinite some chrysotile was also destroyed. (3) Carrying the method furthe. .to use the 002 line can be very rewarding for the sensitivity that can be attained, calculated to be k6% better. However, this is not without sacrificing a lot of time since the parameters involved are many. | !? By looking at the less intense chrysotile line (00^), free of interference, and by selecting the proper parameters, favorable results were obtained. I Tctal time required for ^he analysis of a single sample is estimated at 3-1/2 hours. C - 12 CONCLUSION AND RECOMMENDATIONS X-ray diffraction offers.a specific, accurate and precise method for the determination of chrysotile asbestos in talc at 0.5% and above. C c C 13 - INDEX TERMS i. X-ray Diffraction 2 Chrysotile Asbestos 3. Talc - 1^ REFERENCES 1. J. V. Crable, Am. Ind. Hygiene Assoc. Journal, p. 293, May-June (1966). 2. Analytical Data on Reference Clay material. "American Petroleum Institute Project ^9 Clay Material Standard" , Columbia University New York, July 1950. 3. Klug and Alexander, "X-ray Diffraction Procedure" John Wiley & Sons, Inc., pp. ^08-^09, Appendix V, p, 677* Danas, "Textbook of Mineralogy", Vth Edition by W. E. Ford, John Wiley & Sons, Inc., 19^2. 5. R. H. Jenkins and B. De Vries, Worked Examples in X--ray Spectrometry, Philips Technical Library Springer-Verlap, New York, Inc., 1970, 6. P. Briscese Notes: Practical Applications "X-ray Powder Diffractometry". State University of New York at Albany. C