\Hw...‘ Q. an i 4*. u o .- .L an ~ g n. «3.. . «L .. «V4. «W 3 a .' Pom-B 53‘. fif‘ o.— .l..« R“ b. q o .I 3:. .3” J 1:”. fr” “a. 3\:&‘3 K: »- ‘ K: 1}“ E '3. 3 .‘l ‘RACT 2% av . 4:1 o 0348‘ S m? Ca Am J . nu .. r... mg b. 8”.» “I.” ‘& KM? .5 g; z _ __23::_:__E3::5:213:23:g THES‘S ' u I, \ a _.--.—_-—-.--q..__‘_ .. - _ - I L ' This is to certify that the thesis entitled "X-ray'Diffraction of Common Silica Minerals and.Possible Applications to $011 Genesis" presented by Sidney Solomon Pollack has been accepted towards fulfillment of the requirements for Master of Mdegree inMence Major professor Date MAW— 0-169 TF -. _L».._.~.—h_.‘.‘a—._“- A--- o C I Ln-.. & us. .$ X—RAY DIN-RA STIOI‘I OF COEIMON SILICA L'ZINEPALS AIID POSSIBLE APILICATIONS TO SOIL GET-SIS By Si dney Solomon Pollack AN ABSTMCT Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Soil Science 1953 Approved g. >27 . 7:311: \‘\ var. .\ ' ' s L: Sidney Solomon Pollack 1 ABSTRACT This study is an attempt to determine if quantitative determina— tions of quartz using x-ray diffraction could be used as a basis for calculation of net changes that have taken place in the formation of soils of cool temperate regions. To determine the uniformity of x-ray diffraction of quartz and silica minerals sixteen samples of quartz and other silica finerals (flint, chert, opal, chalcedony, jasper) were studied qualitatively and quantitatively using a Geiger counter Xaray spectrometer. Qualitative examination indicated that all the specimens contained quartz. Only the Jasper contained any identifiable crystalline impurity. Preliminary quantitative results showed that the varieties of quartz varied in the intensity of their xhray diffraction patterns by as much as 30 percent. The reason for this seems to be in part the na- ture of the quartz, and in part to inaccuracies of the method used. The reproducibility of the quantitative determinations in these preliminary studies was poor. Some changes in the method that decrease the standard error of the estimates are presented. Even though quartz does vary in the intensity of its x-ray diffrac- tion pattern, it may still be possible to use it as an indicator mineral. minerals This would be true if the quartz/throughout a given profile were similar. 3le230 X-RAY DIFFEACTICK OF COTTON SILICA KIKZRALS A33 POSSIBLE AEELICATIOYS TO SOIL GEIESIS By Sidney Solomon Pollack THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of ms was or SCIENCE Department of Soil Science 1953 ACKITOT'LED G". '2‘5IT The author wishes to express his gratitude to Professor E. P. Whiteside with Whose inspiration and aid this investigation was under- taken. He is also indebted to Kr. Don Van Farrowe for his advice and permission to use the Geiger counter Xpray Spectrometer at the Michi— gan Industrial Health Laboratory of the Michigan Department of Health, Lansing, Michigan. TABLE OF CONTENTS CHAPTER Page I m PROBLEM A-ND IVMHOD OF ATTACK o o o o o o o o o o O 1 H Statement of the problem.. . . . . . . . . . . . . . Importance of the study . . . . . . . . . . . . . . Minerals studied . . . . . . . . . . . . . . . . . . Method of attaCk.. . . . . . . . . . . . . . . . . . II REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . Diffraction of Xkrays by crystals . . . . . . . . . Diffraction of Xkrays by crystalline powders . . . . Use of X-ray film technique . . . . . . . . . . . . Use of Geiger counter Xeray spectrometer . . . . . . III EXPERIMENTAL WORK . . . . . . . . . . . . . . . . . . Method'used . . . . . . . . . . . . . . . . . . . . Preparation of the standard curve for quartz. . N mmmm KKK-KW NH Operation of the spectrometer . . . . . . . . . H H Preparation of silica samples . . . . . . . . . |-’ H Preliminary Results . . . . . . . . . . . . . . . . Preliminary qualitative studies . . . . . . . . 11 Preliminary quantitative studies . . . . . . . 13 Discussion of preliminary results . . . . . . . . . 1} Modifications of the original method . . . . . . . . 15 Use of modified McCreery Slide . . . . . . . . 15 Grinding of quartz to reduce variability . . . l6 TABLE OF OONTWITS (Cont.) CHAPTER Page Use of the modified method . . . . . . . . . . . 18 Preparation of new standards and samples . 18 Results using the modified method . . . . . 18 Discussion of variation of standard curves . . . 23 Iv smamar.....................2n Qualitative results . . . . . . . . . . . . . . 24 Quantitative results .. . . . . . . . . . . . . 2M Needs for further study . . . . . . . . . . . . 25 BIBLIOGRAPHY...................26 HPMDIX O O O O O O O O 0 O O O O O O O O O O O O 28 LIST OF FIGURES FI SUP. - td li Relative diffraction intensities of the 3.35 A° spacing of quartz and the 3.16‘A0 spacing of fluorite for eight of the minerals listed in Table I . . . . . . . . . . . . . . . . . . . . . 13 Relative diffraction intensities of the 3.35 A? spacing of quartz and the 3.16 A9 spacing of fluorite for eight of the minerals listed in Table I . . . . . . . . . . . . . . . . . . . . . 2 Diffraction intensity patterns of Jasper and rock crystal quartz for interplanar spacings between ‘40 (2 theta) and 64° (2 theta). Arrows indicate probable hematite lines. (The background differ- ence between the two curves is due to different scale factors used on the recording unit). . . . . 3 Three standard curves prepared using the Xpray diffraction method for quartz determination on the same instrument. . . . . . . . . . . . . . . . h Standard curves prepared using the modified method. Curve (1) was prepared using the ratios of peak:heights and curve (2) was prepared using the ratio of weights of areas under the peaks. . . LIST OF TABLES TABLE Page I Percentages of quartz found in different specimens of quartz and related silica minerals . . . . . . . 1“ II Repeated estimations of quartz in the sample Of 8mm q‘mrtz O O 0 O I O O O O I O O O O C O O O 15 III Effect of grinding on the diffraction intensity of (Ward's) quartz particles less than H4 microns in diameter . . . . . . . . . . . . . . . . . . . . 17 IV Percentages of quartz found in different specimens of quartz and related silica minerals using the mdified math-0d o o o o e o e o e o e o o O o o o O 20 V Quartz estimations made from a standard curve prepared using Iq .= Weight of area under the quarts peak Icf Weight of area under the fluorite peak 21 C? iER I TFZE PEOPLE! AITD ICETI-IOD OF ATTACK One of the aims of pedologists is to measure quantitatively the differences between horizons of a soil profile. Marshall and Haseman (13) have used indicator minerals as a basis for calculation of changes that ( have taken place in soil formation and development. An indicator min- eral should be highly resistant to weathering and found in measurable amounts in the soil. Since quartz is a highly resistant mineral and is the dominant mineral in the soils of the Forth Central United States it might serve as a useful indicator. Statement g£_the Problem The purpose of this study was to examine the uniformity of the Xpray diffraction characteristics of diiferent varieties of quartz. A constancy of Xpray diffraction intensities has been assumed in the quantitative estimations of minerals by Xsray methods. If this con- stancy exists in the numerous varieties of quartz known, this method for quantitative determination of quartz in soils might be very useful in studies of soil genesis. Importance g£_the Study After measuring the amount of quartz in sand fractions of the un- weathered material, and the various differentiated soil horizons, comp parisons would indicate the relative uniformity of the parent material 2 and the amount of weathering in the non—quartz fraction. The size dis— tribution of quartz could be used as a check on the uniformity of the parent material. The ratio of quartz size fraction in a soil horizon to similiar quartz in the parent rock could then be used to calculate loss or gain of weight of different horizons and their constitutents. minerals Studied Quartz is a member of the silica family, the most common of min— erals, and is composed of many different varieties. Using Dana's Textbook q§_flineralogy(7) the sixteen specimens* studied in this paper would be classified as follows: I. Quartz A. Phenocrystalline Rock crystal Ward's rock crystal Amethyst Smoky quartz Rose quartz Xilky quartz Geode quartz** NO‘\JIJZ’\_~JNH o B. Cryptocrystalline 8. Chalcedony 9. Flint 10. Basanite ll. Jasper 12. Chert l3. (Chert) 7‘Tovaculite"”'I * Quartz specimens, except for Ward's rock crystal and rock crystal were supplied through courtesy of Dr. S. G.I&agquist of the Geology Department of Michigan State College. ** Classification made by Bake, g3, El, (6). Ce Others 1H. Itacolumite II. Opal 15. Wbod l6. Tripolite I:_!ethod 9;; Attacjg The samples were prepared for study using the procedure for quartz estimation in use at the Michigan Industrial Health Laboratory at Lansing, Michigan. Quantitative measurements of the ratios of the Xsray diffraction intensity of the 2:2§_a9 spacing in quartz relative to the diffraction from the igléfing spacing in fluorite, used as an internal.slang§rd, were made on all the specimens. The percentages of quartz were estimated from the relative intensities of the quartz and fluorite diffraction on a standard working curve showing these rela— tionships for a pure quartz and calcite mixture. A.Philip's Geiger- counter Xeray Spectrometer was used to measure diffraction intensities. The diffraction patterns of the minerals were also recorded with the same instrument. These patterns were studied for evidence of any crystalline impurities and any other peculiarities of the minerals. CHAPTER II REVIEW OF LITERATURE Diffraction g£_§}rays by_Crystals The fact that Xprays could be diffracted by crystals has been known since Laue's famous experiment in 1912 (12), just seven years after the discovery of Xgrays by Roentgen. However it was not until 1919 that Hull pointed out that powdered crystalline substances can diffract X-rays, each substance giving rise to a unique diffraction pattern. Diffractiqn_g£_§7ravs by_Crystalline Powders Hull (9) also suggested that this property could be used as a method of quantitative analysis. Clark in his book (4) gives a good discussion of the theory of X—ray diffraction and its application to powders. Clark and Reynolds (5) in 1936, using an internal standard made use of Hull's suggestion. Use f Efrem Film Techniques They used an X_ray film technique to record their intensity measurements. Whiteside (16) in 19H7, making use of the internal standard in soil studies for the first time, measured quantitatively quartz, albite, dolomite, and calcite in the silt fraction (2 - 50 microns) of loses in Illinois. Use q; Geiger-counter Erray Spectrometer Ehillippe and White (15) in 1950, used a similar method in a min- eralogical study of the fine silt (2 — 20 microns) and coarse silt (2O - 50 microns) in 12 Indiana soils. However, instead of using the Xpray film technique of Whiteside they used a Geiger counter Xpray spectrometer. V The Geiger counter spectrometer is reported to be less time cons suming and to provide a more accurate means of measuring intensity than the photographic method. It also has the advantage of more pre- cise angle measurements and gre ter resolution.(10). Alexander, Klug and Kummers (1) using a manual scanning and count- ing technique found that reproducibility with their spectrometer was poor unless the particles of material examined were less than 15 mic- rons in equivalent diameter. They also found that intensity varied with the size fraction, the smaller particles diffracting with greater intensity. “Ballard, Oshry and Schenk (2) have described methods of \- ,-.--""""" .. ._ "‘"m-‘n-a. m -‘ "‘ grinding particles to the desired size. Determinations made by dif- ferent workers (3, ll,‘15) and different Geiger counter X—ray spectro- meters have been found,under suitable conditions, to be accurate to about i 5 percent of the quantity of the mineral present. CHAETER III EC}; "- ’I‘E‘IE‘J. VIC? Yethod Used The standard working curve used in the preliminary studies was the sane one used at the Iichigan State Industrial Healt. Laboratory, and prepared by them in 1952 (See Figure 3, curve labeled 1952). The procedure described below is the one used in its preparation and the preparation of unknown samples. This method will be called, for fur ture reference, the original method. Preparation<§_the standard curve for gpartz. The quartz used in determining the points on the working curve had been secured from the Rational Bureau of Standards. Standard mineral mixtures pr pared by mixing varying amounts of quartz with a diluent, calcium carbonate, to :- /" make a total of 2 grams. To each of these mixtures ranging from 10 a" " . . ,,- 4? " ‘3 _.fi I " I f'.lo percent to 100 percent quartz, 1 gram of calcium fluoride, used as an .a g internal standard, was added. These samples were then ground, using a Fisher hard steel, motor powered, mortar and pestle to pass through a 325 mesh sieve having openings of Efi’microns. They were then placed in a glass vial and mixed 10 to 15 minutes with a spatula. The standards were then ready for mounting to be X—rayed. The samples were mounted in aluminum slides or cells about #0 mm. square, with a rectangular hole 10 by 20 millimeters, with a depth of about 1 millimeter. This provided a maximum.focal area of 200 square millimeters. A piece of glass was fastened to one side of the cell by 7 means of scotch tape. The mixture was then poured into the cell and carefully leveled off with a glass slide or spatula, so that the sur- face of the sample was level with the top of the slide. A smooth sur— face having no pits or lint protruding is desired. However, it is also necessary that the particles are not compressed in such a way as to lose their random orientation and that no sorting of the sample occur on the surface of the mount. A slide prepared in the above manner was then placed in the slot in the spectrometer and the X-ray diffraction intensities from the 335 A9 spacing of quartz and the 3.16 A9 spacing of the internal stan- dard, fluorite, were recorded as the goniometer rotated from 25° to 30° 2 theta. The types of curves recorded are shown in Figures 1A and 13. Operation gf.the spectrometer. The Phillip's Geiger counter X- ray spectrometer type “2202 used,included a basic Xpray production unit, a wide range, high and low angle Geiger counter Goniometer, and elec- tronic circuit panel and an automatic recorder. The Xéray unit was adjusted to 157milliamperes at 35 kilovolts. It was allowed to warm up for 20 minutes, after which the amperage did not fluctuate. This indicated a steady output of X—rays and anal- yses were then begun. The Xéray tube contained a tungsten filament and a copper target. With electron bombardment of the target X—rays of various wave lengths were produced. A nickel filter .0007 inhhes thick filtered out all but the copper K‘kradiation. J CHERT FIGURE IA l J J J a d J ITACO- TRIPO- MILKY CHAL- NOVAC- AME- OPAL LUMITE LITE QUARTZ CEDONY ULITE THYST ' QUARTZ Relative diffraction intensities of the 3.35 A0 spacing of quartz and the 3-16 1° spacing of fluorite for eight of the minerals listed in Table I. The larger peak in each pair is quartz and the smaller peak is fluorite. SMOKE Geo: WARD'S L. ~ { k FLINT ROCK JASPER sASAN- QUARTZ QUARTZ ROCK CRYSTAL ITE CRYSTAL QUARTZ FIGURE | 9 Relative diffraction intensities of the 3.35 A0 spacing of quartz and the 3.16 AP spacing of fluorite for eight of the minerals listed in Table I. The larger peak in each pair is quartz and the smaller peak is fluorite. H ROSE QUARTZ 10 A slit near the source of the Xprays and the divergence slit, with an angular aperture of 1°, enabled the X-raying of approximately 2/3 of the maximum focal area at 26.50 (2 theta) the angle at which the diffraction of the 3.35 A0 spacing of quartz is measured. With a larger angle a smaller portion of the maximum focal area is exposed to the radiation so that less of the sample is X—rayed when the dif- fraction intensity of the 3.16,A9 spacing of fluorite is measured. Bragg's Law is used to evaluate the diffraction pattern. The receiving slit which defines the width of the diffracted beam recorded by the Geiger counter had a width of 0.003 inch. A scatter slit having a u° aperture, placed in front of the Geiger tube insures that it receives only rays diffracted from the sample. The electronic circuit panel causes the rays to be recorded as peaks on a graph. To obtain the measurements necessary for one ratio, the goniometer was set at a scanning speed of 1° (2 theta) per minute, and the sample was rotated from g§3_33“39? (2 theta). From 12 to 18 ratios of the height of the diffraction intensity peak of the quartz 3.35 A9 spacing to the fluorite 3.16 A? spacing were used in calculating each point on the standard working curves. This was accomplished by making N or 6 measurements at each of three slightly different positions on the slide. Half of the readings were made with 2 theta increasing, the other half while it was decreasing. The heights of the peaks were measured from a base line judged to be an average of the background. It was noted that the intensity peaks of diffraction occur at slightly different angles depending on whether the goniometer was ascending or descending. This is due to some slight play in the actuating mechanism. 11 Preparation 9: silica samples. Each mineral specimen was examined and a portion thought to be representative was chipped off. The sam- ples were then ground to pass through a 130 mesh sieve, having an open- ing of 0.1 millimeters, using a manual mortar and pestle of hard steel. The procedure used in the preparation of the standards was followed from this point, with a few exceptions that are noted below. At first instead of the original method of mixing the samples with the internal standard they were shaken through a 325 mesh sieve 3 times, and then mixed in a mechanical mixer for 2 or 3 minutes. This was later changed to mixing the sample on a lint free surface 10 to 15 minutes. The second modification consisted of shaking small portions of the mixed samples through a 200 mesh sieve to break up aggregated chunks just before mounting. Instead of the entire sample only the part that went through the sieve was used in making the slide. This led to er- roneous results which are discussed below. Diffraction patterns of the silica samples diner than “H microns) were obtained to detect the presence of crystalline impurities. Preliminarngesults Preliminarl qualitative studies. The diffraction patterns of the unmixed samples indicated that all contained quartz but only the jasper contained any observed crystalline impurity. The diffraction patterns of a pure quartz sample and the jasper sample are shown in Figure 2. Using Hanawalt's (8) tables the lines present in the jasper sample but not in the quartz sample were judged to be due to hematite. 12 ROCK CRYSTAL QUARTZ PEAK 0F CHART—9 JASPER PEAK 0F CHART-'9 J A! FIGURE 2 Diffraction intensity patterns of Jasper and rock crystal quarts for interplanar spacings between no (2 theta) and 6M0 (2 theta). Arrows indicate probable hematite lines. (The background differ- ence between the two curves is due to different scale factors used on the recording unit). 13 Preliminary quantitative studies: Table I lists the amounts of quartz estimated in mineral specimens studied using the above described method. Figures 1A and 13 show examples of the recorded intensities of each of the samples studied. The broad peaks of flint and chalce- dony as compared to the others, are indicative of very fine crystalline material. In addition to the silica specimens examined, the author ran determinations on a number of the standard quartz-calcite mixtures to check the accuracy of his technique. Discussion 9: Preliminary Eesults A statistical analysis of the results listed in Table I (See ap- pendix for method used in calculation) revealed that differences of 9.5 percent or more of the quartz content of a sample would be signi- ficant at the 5 percent level. It was mentioned above (See: Preparation of silica samples) that making a slide of a small portion of the powdered sample sieved through a 200 mesh sieve led to erroneous results. It was believed that this sieving caused sorting of the mineral and the internal standard. In some cases the quartz was overestimated, while in others it was under- estimated. This was discovered when a mechanical shaker was obtained — to break up the small aggregates of the sample - and the whole sample was sieved prior to the slide preparation. Ending of the faulty sieving did not resolve all problems. Al- though there was no longer a sorting effect, something was causing a high variability between separate determinations of the same sample. Seven estimations of the quartz content of smoky quartz are listed in TABLE I IN PERCIHTAGES OF QEARTZ FOUHD IH DIFmefiHT SPECIMEHS OF QUARTZ AND RELATED SILICA HIHERALS Sample No. of Average-oi Individual Quartz Ratios Quartz % Determinations (when Used more than one was made) 1. Flint an 142.5 i 1.3 1111.0 f 1.5; 111.0 1': 0.1 2. Jasper 12 994.0 t 1.0 3. Chalcedony 12 45.0‘i 0.9 u. Quartz, var. - 53.0 i 2.5; 52.0 i 2.3; Amethyst 36 52.0 1' 1.5 55.0 i 1.9 6. Nowacunte 12 60.0 i" 1.0 7. Tripolite 12 63.5 1‘ 1.5 s. Chert 2L1 63.5 t 2.0 60.5 i 2.6; 66.5 1’ 0.01 9. Quartz, Geode 24 71.0 1' 2.1 72.0 i v.2; 70.0 i 1.0 10. Quartz, Var. Basanite 12 72.0 i 1.5 11. Quartz, Rose 12 82.5 :t 2.1 12. Quartz, (‘?ard's)* 12 89.5 1' 2.5 13. Quartz, Milky 12 92.0 t 2.5 111. Itacolmnite 24 92.5 :1: 1.8 92.5-: 1.6; 92. 5+ _ 3. s 1.5. Quartz, Smoky as 88.5 i h.1 72.0 i 2.6; 110. 0 + 3. 3 96.0 i 6.7. 16. Quartz ‘21?" 98.01" 1.7 97.0 111.0;9 9.0 i 2.0 17. 667: Quartz + toys I I Calcium Carbonate** 2H 59.5 i 1.7 53.0 i 0.1; 66.0 i 0.1 16'. 70<5Q artz + 30.6 Calcium Carbonate" 21+ 67.0 i 1.5 63.0 i 1.3; 70.5 i 0.1 19. 80% Quartz +'20$ Calcium Carbonate'" 12 80.5 1' 1.6 * Used as a pure quartz standard (S. S. S. A. P. 12:u1r-u19, 19u7{: ** Standards used by the State Health Laborathry, Lansing, Hichigan. **‘ Quartz estimates using the original X;ray diffraction method, see page 11. Table II. At the 5 percent level three of the seven samples , instead of one in twenty, are significantly different from the other four. TABLE II REPEATED ESTIXATICHS O? QUARTZ IV THE SAHELE OF SUCK? QUARTZ V“ V w— —: -— '“ Dre—termination Estimated Quartz Percent Standard Error 1. 72 2.6 2. 76 1.3 3. 110 3.3 1+. 96 6.7 5- 97 2.7 6- 95 3.9 7. 88 5.0 - It was thus realized that modifications of the method would be necessary to achieve accurate and reproducible results. hodifications g£_the 0riginalDMethod Q§g_g£_§'modified_McCreerg_slide. The first change in procedure consisted of preparing the slide in a different manner. McCreery's (1%) method of not manipulating the face of the slide to be X;rayed was adopted. As in the original method a clean piece of glass was taped tightly to the aluminum cell. The sample was not sieved as McCreery suggested into the hole but simply poured. It was then compressed to form.a smooth surface against the glass fastened to the slide. The same precautions as before were taken to insure a smooth slide. 16 The excess powder was removed with a razor blade. .Another piece of glass was taped over the surface Just leveled. This left a slide with powdered sample sandwiched between two pieces of glass. The first piece of glass and tape placed on the cell was then removed. The surface then exposed was examined for imperfections and if none were found, it was X;rayed. Using this type of slide the variability was reduced only slightly Grinding g£_gg§3£g_tg_reduce variabiligy, The other attempt to reduce the variability involved the use of finer textured samples. . Since separation of the quartz into size fractions below NM microns was thought to be too laborious, it was decided to grind these samples again and use the whole sample. The first step in this attempt was an experiment to determine how further grinding would effect the diffraction intensity of quartz smaller than MM microns. Four samples of Ward's rock crystal weighing 1.2 grams were weighed out and ground for different lengths of time, 30, 60, 120 and 2&0 minutes, using the power driven grinder. Methanol was added to prevent the samples from sticking to the mortar. After grinding the remaining methanol was allowed to evaporate and the aggre- gates formed on drying were destroyed by sieving through a 325 mesh sieve. To one gram of each sample was added 0.5 gram of calcium.f1uoride that had passed through a 325 mesh sieve. These were then mixed on a lint free surface for 10 to 15 minutes using a spatula. Slides were prepared by the modified McCreery method and the mixtures were X-rayed. 17 Sixteen ratios were used in calculating the average ratio, of the intensities of the quartz and fluorite diffraction spacings. These were obtained by making four mounts of each sample and making two mea- surements at each of two different positions on each slide. Half the measurements at each position were made with 2-theta increasing, and half while it was decreasing. Table III shows that grinding reduced the diffraction intensity of ward's rock crystal quartz. It can also be seen that grinding re- duced the variability of the determinations, as is shown by the lower standard error of the ground samples. TABLE III EFFECT OF GRINDING ON THE DIFFRACTION INTENSITY 0F (wann's) QUARTZ PARTICLES LESS THAN nu MICRONS IN DIAMETER Length of Grinding Iq/Icf Standard Error 0 5.29 .251 30 minutes ”.67 0093 30 minutes n.73 .101 60 minutes n.35* ‘ .107 120 minutes n.51‘ .097 2h0 minutes 3.82 .085 2M0 minutes 3.78 .057 To checkzthe validity of the standard error calculation two de- terminations were made on samples ground for two different lengths of time. The differences between the two measurements of each sample 18 were not significant indicating that the standard error calculation apparently could be used as a measure of the variability between separ- ate determinations of the same sample. Use 9§_the Modified Method Preparation g£_new standards and samples. Since it was found that grinding reduced the variability of separate determinations made on the same sample, it was decided to prepare a new standard working curve. Quartz similiar to that used in the preparation of the original standard working curve was not available. In its place was used the rock crystal quartz, sample 16 in Table I. With calcite as a diluent. and calcium fluoride as the internal standard, mineral mixtures of 30 percent, 60 percent, and 90 percent, quartz were prepared. The modifications described in the preceding section were used in the preparation and X>raying of the standards. The standard curve thus prepared is shown in Figure 3, 1953. Results using the modified method. New portions of some of the silica samples were prepared and XLray d using the modified method. The results of the estimates of their quartz content based on the new standard working curve, (See Figure 3, curve labeled 1953) are listed in Table IV. These results, like the preliminary results in Table I, show that various samples diffract Xerays with different intensities. However, there is a decrease in the range of estimations. Although fewer amn- ples were run using the revised method they seem to fall into two groups. . 5.530 ezmommd n mane .._ 8. cm or 2. em on oc on on o. o m _ _ . _ _ _ m a r: 1.N m m I 1m nmm_ r. .L¢ :um_ T 4332. .3 2% C meander 3. 3A 1 n A ouaaosam auaoaeo mo hauesopcw ma Hem “Annmaok mo meson NV wqaosam ed mm.m mom. “page no heinous“ a“ UH .oafiuoooaa causes .0 Magus nomads on» .3 woken no.3 as: 3.30 mmma 08a .3383 came one means @303 anchorage as women no.3. ease serge mmma one Hmma one icon—933w. case on» no nouagaaaopoc 39qu no.“ cameos ”“0303.“ch assign on... wages venomous. mega assuage e939 20 TABLE IV PERCENTAGES 0F QUARTZ FOUND IN DIFFERENT SPECIMENS OF QUARTZ AND RELATED SILICA MINERALS USING THE MODIFIED METHOD Sample % Quartz Standard Error 30% Standard 30% 1.1% 30% Standard 32.5 0.9 60% Standard 57.5 009 90% Standard 92.0 2.0 90% Standard 90.0 2.5 Ward's 93.5 2.0 Ward's 99.5 2.1 Amethyst 91.5 2.2 Ghalcedony 58.0 l.0 Deode 93-5 2.3 Chart 63.5 2.0 The phenocrystalline samples fall into a group having'a high percent- age of quartz. Rock crystal, Ward's rock crystal, geode and amethyst have a range from 91.5 percent to 100 percent of quarts. These differ- ences may be due to unequal grinding and may not be real differences. The cryptocrystalline samples have low quartz contents. Chert shows 63.5 percent and chalcedony 58 percent of quartz. A statistical analysis (See appendix) indicates that differences of N.9 percent are significant at the 5 percent level. It was mentioned in the Preliminary Quantitative Results section, that some specimenspflint and chalcedom'. had broader bases of the 21 intensity peaks than the other samples. The data collected previously was therefore expressed in another manner, in an attempt to see how it would effect the quartz estimations. This was done by cutting the peaks from.the graphs using a razor and determining the ratio of the weight of all the quartz peaks to the weight of the fluorite peaks. A standard curve (see Figure h) was prepared using the above method and estimations were made on some of the samples listed in Sible V. TABLE v QUARTZ ESTIMATIONS MADE FROM A STANDARD CURVE PREPARED BY USING LL: WEIGHT OF AREA UNDER THE QUARTZ PEAK Icf WEIGHT 0F AREA mmER THE FLUORITE PEAK ____ Sample % Quartz Estimated 30% Standard 28 30% Standard 31.5 60% Standard to 90% Standard 90 90% Standard ‘ 89.5 Ward's 100 Amethyst ' 97.5 Chalcedony 72 Using this type of standard curve the percent quartz increased some- what in both the chalcedony and amethyst while the Hand's rock crystal remained the same. Nkmdbo hzwomwm v wane; 00. Ca on on 00 on 0' On ON 0. -o _ a _ _ _ _ _ _ _ x 2.3? cow... .8 O x _¢¢o.+ en». . a 3 I 9| mw_b 0 II I. m Inc 0 Judson on» .393 means we magmas: mo ago." on» mdamd cohaoam out Amy 05.50 .33 magmas: Moog mo moans." on... magma nonsmoum F. as: AC o>.3.o .uonuca damages on» madam 69530.5 motes 6.3333 In 23 Discussion g§_Variation g£_Standard Curves Including the curve Just mentioned four standard curves have been prepared at the Health Laboratory, using the same spectrometer as a basis for the determinations. Three of them (Figure 3) were prepared using the measurement of peak heights to determine the ratio of the intensity of quartz diffraction to that of an internal standard. The same samples were used by different operators in preparing the 1951 and 1952 curves (Figure 3). The author used the modifications already mentioned in treating samples be used for preparing the 1953 curve. The reason for the wide variance of the curves is probably the tech— nique of the person making the sample mounts for the determinations. A large part of the error of the method is probably due to this part of the procedure. CHAPTER IV SUMMARY Qualitative Results Qualitative diffraction patterns of the silica varieties studied indicated quartz to be present in all of them. No crystalline impur. ities were identified in any of the samples except that hematite was found to be present in the Jasper. Quantitative Results The modified method used in estimating quartz indicated.that dif- ferent silica specimens varied in intensity of Karay diffraction typ- ical of quartz. This was taken to mean differences in actual quartz content. The phenocrystalline varieties contained from 91 percent to 100 percent of quartz. The cryptocrystalline varieties contained con- siderably less than 100 percent quartz. Chert had 63.5 percent and chalcedony 58 percent. A.modification in preparing the standard curve to take into account the broad peak of chalcedony raised its quartz estimation 1“ percent; to 72 percent of quartz. Three standard curves prepared by three different people, using the same equipment are shown. Differences are attributed largely to variation in technique of people using this method. 25 Needs for Further Study Further study is needed to determine if the phenocrystalline varieties actually differ in quartz content or whether the differences are due to sample preparation. This might account for the low estimates of the cryptocrystalline varieties, but it is improbable. The low diffraction intensities of the latter varieties may be due to the fineness of the crystalline material, the presence of amorphous material, or impurities, or less perfectly crystalline material. Thin sections of chalcedony, smoky quartz, amethyst and milky quartz indicated no amorphous material present. Although.there may be some impurities in most of the sam- ples it is doubted that any but the jasper contain amounts greater than 5 percent. It would therefore, seem that the differences of diffraction are due to the nature of the quartz in the specimens or the way the diffraction is measured. If quartz is to be measured quantitatively using the.xaray diffraction method, the cause of differb ences in quartz specimens must be found. Even though varieties of quartz do vary in the intensity of X; ray diffraction, it may stillln possible to use it as an indicator mineral in the study of soil genesis. This would be true if the quartz throughout a given profile were a similar varidty. It might be suf- ficient to establish only that the quartz throughout the profile was entirely of the phenocrystalline group. 1. 9. 10. 11. 12. 13. BIBLIOGRAPHY Alexander, D., Klug, H. P. and Hummers, E. "Statistical Factors affecting the intensity of X—rays." Journal of Applied Physics, 19: 7A2.753. 19nd. Ballard, J. H., Oshry, H. I. and Schenk, H. H. "Sampling, Mixing and grinding technique in the preparation of samples for quanti- tative analysis by X—ray diffraction and spectrographic methods,“ J. Optical Soc. Amer.. 33: 667. 19u3. Carl, H. F. "Qunatitative mineral analysis with a recording X-ray diffraction spectrometer,“ Amuican Minera10gist, 32: 508-517. 19u7. - Clark, G. L. ’Applied g-razs, New York: McCraw-Hill Book Company, Inc., 19110, 647 pp. Clark, G. L., and Reynolds, D. H.,"Quantitative analysis of mine dusts, an X—ray diffraction method,” Ind. and Eng. Chem., Anal. Ed., 8: 36—uo. 1936. Dake, H. C., Fleener, F. L., and Wilson, 3. H. MFamilz Minerals, New York: McGraw—Hill Book Company, Inc., 1938, 304pp. Ford, W. E., revised by, Dana's Textbook of Mineralogy, l4th edition, New York: John Wiley and Sons, Inc., 1951, 851 pp. Hanawalt, J. D., Rinn, H. W. and French, L. H.,"Chemical Analysis by xaray Diffraction," Ind. and Eng. Chem., Anal. Ed., 10: M57- 512, 1938. Hull, A. w. "A new method of chemdcal analysis," Journal of the American Chemical Society, ’41: 1168-1175. 1919. Klug, H. P., Alexander, L. and Hummers, E. ”X-ray diffraction of crystalline dusts," J. of Ind. Hyg. Tox., 30: 166—171. 19148. Klug, H. P., Alexander, I... and Hummers, E. "Quantitative analysis with an X-ray spectrometer." Anal. Chem., 20: 607-609. 19118. Kraus, E. H., Hunt, W. P., and Ramsdell, L. S. Mineralog. New York: McGraw-Hill Book Company, Inc., 1936. 638 pp. Marshall, C. E. and Hasemn, J. P., “Quantitative evaluation of soil formation and development by heavy mineral studies: Grundy silt loam." Soil Sci. Soc. of Amer. Proc., 7: 1448-953. 19342. 27 19. McCreery, G. L., "Improved mount for powdered specimens used on Geiger counter Xéray spectrometer." J. Amer. Ceram. Soc., 32: 141-146. 1949. 15. Phillippe, M. M. and White, J. L. "Quantitative estimation of minerals in the fine sand and silt fractions of soils with the Geiger counter thay spectrometer." Soil Sci. Soc. Amer. Proc., 15: 138-lu2. 1950. 16. Whiteside, E. P. "Preliminary Karay studies of lease deposits." Soil Sci. Soc. Amer. Proc., 12: #15-419. 19u7. APPENDIX Calculation of percent quartz required for significant difference between means at the 0.05 level. 29 X = % Quartz = Standard Error x r x2 r3 xr 30.0 1.h 900 1.96 u2.0 32.5 0.9 1056.25 0.81 29.25 57.5 0.9 3306.25 0.81 51.75 58.0 1.0 336u.00 1.00 58.00 90.0 2.5 8100.00 6.25 225.00 91.5 2.2 8372.25 n.8u 201.30 93.5 2.3 87u2.25 5.29 215.05 92.0 2.0 8u6u.00 u.00 18u.0 98.5 2.0 9702.25 n.00 197.00 99.5 2.1 9900.25 u.u1 208.95 806-5 19-3 65939-75 37-37 1539.30 11 = 11 f = 7.3318 ? = 1.751: X? - LZXMZY) 1539.30 - 180654193) b: n = 11 x2 - gap? 65939.75 - L__51_806. 72 n 11 129.26 = b= 686835"' 0.0182 = a + bX a = f - bi = 1.7511- (.0182) 73.32 = 1.754 - 1-33“ Y = .1420 + .0182 (X) .If the Y axes is translated so that when X = 0 Y = 0, the regres- sion equation becomes: Y = 0.011} X 30 Let h = given mean Let g== mean significantly different at 0.05 level. Then t=h+g-h r(h+8-E r01 + g - 1T) \/(0.011+)?(h + g)?+ 0029211?" 0.0111. V‘Zh? + 211g + g2 t = L \/0.011+ 21:2 + 2hg + g2 g (t) (0.019) V312 + Zhs 4* 82 1: 2.1115 for 11+ degrees of freedom at 0.05 level. g = (2.1M5)(0.01I+) ‘/2h2 4- Zhg + g2 g2 = (n.601)(0.000196) 2n2 + 211g + g2 (0.000902) 2n2 + 212.2: + 82 0.0018014 h2 + 0.00180h hg + 0.000902 g2 g = u .. 0.000902) - 0.0018Mh2 - 0.0018011 hg o = 0.999098g2 - 0.001801ch - 0.00180th 0.00180u :t h 70.0018011113 + h(o.001801+) (0.999098) 09 ll 2(0-999095) = 0.0018011 i h V0.00180h + 0.0072095 1.998196 0.0018011 :1: h \/0.009113 1.998196 (00041801! 1' 0.095101: 1.998 31 = 0'0 2 = " = .' g 1—21-.998 11 0.0486 b “.936 or =0.0 6 = .48 h= . g Til—.998 h 00 6 117% From before: h = given mean g = mean significantly different at 0.05 level. )4.95% is used since it is the larger percentage. '9 U V“ I ' A .“ ~ I t i . I I. I atom USE 0:111, ‘ 'E' unu- ROOM uxs. ’ : o. .8 '54 ’ ; c P- . "m . ,_ c if" '33.", 13;: 1”\' K" II 2‘ pig,“ A fist/+‘s _:’". o: . ' "8 urMJEII-wnvluPllhfiian r? i . Firth)?- inr uni-[LIE‘bnla ,2... u I MHHHGANSTATELHNVERSHYIJB n * .|( NH” H I‘M! IJIHIF i‘llmlll l I 3 1293 03175 5071