Determination of Mineral Constituents of Rocks by Infrared

May 1, 2002 - Click to increase image size Free first page. View: PDF | PDF w/ Links. Citing Articles; Related Content. Citation data is made availabl...
1 downloads 11 Views 633KB Size
1169

V O L U M E 25, NO. 8, A U G U S T 1 9 5 3 high decade resistance. The neutral filters are then removed and the sample is placed in the chamber (in the goniometer a t 90” to the incident beam in the case of flat samples) and the intensity is determined. The transmittance of the sample is then given by the ratio of the first reading to that of the second and the ratio is multiplied by the combined transmittance of the neutral filters, the transmittance of the neutral filters having previously been determined by an earlier measurement with the instrument. By these means, samples with a transmittance as low as 1O-IO may be determined. Reflectance properties of surfaces are determined by placing the flat sample in the goniometer a t fixed angles with respect to the incident beam and measuring the relative intensities of reflected light a t various angular settings of the detector. The smaller incident nosepiece and the receiver nosepiece should be used in these measurements. The results obtained for a white bond paper are shown in Figure 8. Such measurements should be useful in comparing the optical properties of papers manufactured by various processes or papers to which coatings or filters have been added. The instrument also serves as a glossmeter for comparison of coated glasses, painted surfaces, ground surfaces, sheen of textiles, ctc. ACKNOWLEDGMENT

The author wishes to express his thanks to the American Instrument Co., Silver Spring. Nd., for its cooperation in the design and development of the inqtrument. He is especially in-

debted to Justin J. Shapiro of t’hat company for his many helpful suggestions. LITERATURE CITED

(1) Alfrey, T., Bradford, E. B., Tanderhoff, J . W., and Oster, G., J . Opt. SOC.A m e r . , in press. (2) American Instrument Co., Silver Spring, AIcl., BUZZ. 2202A. (3) Brice, B. A., Halwer, M., and Speiser, R., J . Oyt. Soc. A m e r . , 40, 768 (1950). (4) Corning Glass Works, Corning, X. Y., “Glass Color Filters.” ( 5 ) Doty, P., and Bteiner, R. F., J . Chem. P h y s . , 18, 1211 (1950). (6) Garlick, G. F. J., “Luminescent Materials,” Chap. 8, Oxford,

Oxford University Press, 1949. (7) Hadow, H. J., Sheffer, H., and Hyde, J. C., Can. J . Research, 27B, 791 (1949). ( 8 ) Hermans, J. J., and Levinson, S.,J . Opt. Suc. Attier., 41, 460 (1951). (9) Johnson, I., and LalIer, I-.li., J . Am. Chem. Soc., 69, 1184 (1947). (10) Xommaerts, W.F. H. AI,, J . Colloid Sci., 7 , 71 (1952). (11) Oster, G., Chem. Recs., 43, 319 (1948). (12) Oster, G., J . Gen. Physiol., 33, 445 (1950). 9, 525 (1952). (13) Oster, G., J . Polymer Sei., (14) Oster, G., Trans. Faraday Soc., 47, 660 (1951). (15) Speiser, R., and Brice, B. A . , J . O p t . SOC.A m e r . , 36, 364 (1946). (16) Valcouleurs, G. de, Conipt. rend., 229, 35 (1949). (17) Weber, G., Biochem. J . , 51, 146 (1951). (18) Zimni, B. H., J . Chem. Phys., 16, 1099 (1948). (19) Zimm, B. H., J . Polymer Sci., 10, 351(1953). RECEIVED for review February 19, 1953.

Accepted June 1, 1953.

Determination of Mineral Constituents of Rocks by Infrared Spectroscopy JOEY M. HUNT AND DANIEL S. TURNER Research Laboratory, The Carter Oil C o . , Tulsa. Okln. It was desired to develop an infrared spectroscopic method for the qualitative and quantitative analysis of the mineral constituents of sedimentary rocks for use in petroleum exploration research. The method developed consists of grinding rocks to a fine powder and examining the powder as a film on a conventional sodium chloride window. The mineral constituents of the rocks are identified by comparing their spectra with the spectra of pure minerals. Quantitative analyses within 10% of the amount present can be made for minerals which have sharp, well-defined absorption bands such as quartz, kaolinite, orthoclase, calcite, and dolomite. Errors in the analysis are caused by nonuniformity of the sample film and scattering of the infrared radiation. The technique has been used in analyzing oil well cuttings, cores, drilling mud, and surface samples that are obtained in connection with various geological problems related to petroleum exploration.

I

K A previous paper ( 3 ) a powder film technique was described by which it is possible to obtain well-defined infrared spectra of minerals and other inorganic compounds. The present paper describes the application of this technique with some modifications to the determination of the mineral constituents of rocks, both qualitatively and in some cases quantitatively. The importance of grinding mineral samples to a particle size smaller than the wave length of the infrared radiation has been emphasized ( 5 ) . The presence of large particles tends to scatter

the radiation, so that only a small percentage of the incident radiation is transmitted and the absorption bands are very poorly resolved. Several techniques for obtaining small particles of solids have been reported in the literature. Mulling a sample in Sujol or Fluorolube (a misture of completely fluorinated hydrocarbons available from the Hooker Electrochemical Co.) is the most common method and was used by Miller and Wilkins in their excellent and extensive study of the spectra of inorganic chemicals (6). Suspension of a solid with aluminum stearate in cyrbon disulfide has been used for the quantitative analysis of organic chemicals (I). Vnfortunately, neither of these techniques was applicable to the quantitative determination of minerals in rocks, because of the greater hardness of the rocks and their higher densities. Grinding and sedimenting a rock to ohtain a desired particle size have been used for qualitative studies (3-5). They are not satisfactory for quantitative analysis, owing to the separation of the mineral constituents of a rock in a sedimenting column. I n a typical rock the clay minerals will be concentrated in the fine particle fractions, whereas quartz 1% ill be concentrated in the large particle fractions. An unequal distribution of rock constituents can result from differences in particle size, shape, and density. The method ultimately adopted in this work involves grinding all of a rock sample to a particle size of less than 5 microns in a cyclonic type of jet pulverizer known as the Micronizer (obtained from the Sturtevant Mill Co., Park and Clayton Sts., Boston, 22, Mass ). Rock samples ground in this unit are fine enough to run directly as a powder film mount. Quantitative determinations of certain minerals in the rocks are made by comparing absorbances (optical densities) with those obtained from known con-

ANALYTICAL CHEMISTRY

1170 centrations by weight of pure minerals. The application of this method to unknown rock samples gave quantitative results comparable to those obtained with the x-ray Geiger counter spectrometer. PROCEDURE

Y

Z

2

GULF COAST LA

p

I

I

100 110 WAVE LENGTH IN MICRONS

0

IiO

IlO

140

IS0

160

Q IO0 From 10 to 20 grams of F the rock to be analyzed c are ground to pass a 250Z 80 mesh screen. This powder Y is then fed into a Micronm izer a t the rate of 5 to 10 P 60 w grams per minute under a jet pressure of 90 to 100 pounds per square inch. In it the particles are broken up by violent collision with each other in a cyclone chamber. The 20 30 40 50 60 70 80 90 100 110 120 130 140 150 1 160 micronized powder is then 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 examined with a microWAVE LENGTH IN MICRONS scope to determine if over Figure 1. Infrared Spectra of Sedimentary Deposits 90% of it consists of particles smaller than 5 microns. The Dresence of many large particles may necessitate passing the sample incorporating the spectra of some of the more common silicates in through the Micronizer again. This may happen with very this region. This pattern, shown in Figure 2, contains the spectra hard materials such as quartz sand. of the silicate minerals kaolinite, illite, montmorillonite, attapulA few milligrams of the micronized powder are placed on a gite, muscovite, and albite. The intensity and shape of the instandard sodium chloride window and a few drops of isopropyl alcohol are added to form a paste. The paste is smoothed out frared absorption bands for these minerals are seen to differ conwith a polished glass microscope slide and after the alcohol siderably. By placing this template directly over the spectrum evaporates, a thin powder film is left adhering to the window of a sample being analyzed, the qualitative identification of the surface. The edge of this film is cleaned off to a premarked silicate minerals is simplified. It is important, of course, to adposition on the window and an infrared spectrum of it is obtained in the usual manner. All the spectra in this paper were recorded just the thickness of the powder film to obtain a spectrum with on a Baird Model B double-beam infrared recording spectrothe silicate absorption bands a t about the same intensity as in the photometer. template. Referring to the template and the spectra in Figure 1, it appears For quantitative analyses, the powder film is examined microthat the principal clay mineral in the recent sediment is illite. scopically, before recording, to evaluate the uniformity of the film The steplike infrared pattern in the wavelength region of 11 and the extent of coverage-that is, the percentage of uncovered microns is characteristic of most illites and has not been found in areas. If the film is very uneven, or if there is more than 5y0of the spectra of other common rock-forming minerals. The specbare space, the mounting should be repeated. After recording, the weight of the powder film is determined, care being taken not to lose any of the powder between the recording and weighing operations. When only qualitative analyses are being made, these steps can be eliminated.

-

d ,ii,

,

v ,

QUALITATIVE ANALYSIS

The principal constituents of most sedimentary rocks are quartz, the carbonates (dolomite, calcite), and the clay minerals. Spectra of these and other pure minerals have been presented (3-5). I n Figure 1 the infrared spectra of a typical recent and ancient sedimentary deposit are shown. The first sample is a mud deposit from the Gulf Coast. The second sample is a reef limestone, Silurian in age, from Indiana. The presence of quartz in both of these samples is evident from the characteristic double band a t wave lengths of about 12.5 to 12.8 microns. The absorption bands a t wave lengths of 7.0 and 11.4 microns are due to a carbonate mineral. This is identified as calcite in both samples from the absorption band a t a wave length of about 14.0 microns. Dolomite is present in much smaller amounts in both samples, its characteristic band being a t about 13.7 microns. The clay minerals, kaolinite, illite, and montmorillonite, are more difficult to identify, because the distinctive features of their spectra all overlap in the wavelength region from 9 to 12 microns. [This characteristic absorption region of silicates has been described thoroughly in a recent paper by Launer (.5).1 I n order to simplify recognition of these minerals, a template pattern was drawn

80

90 100 11.0 WAVE LENGTH IN MICRONS

Figure 2.

12.0

Mineral Identification Template

V O L U M E 25, NO. 8, A U G U S T 1 9 5 3

1171 era1 minerals in the wave-length range from 12 to 16 microns are shown in Figure 3 for comparison. The percentage of a mineral in a rock that can be detected qualitatively varies considerably, depending on the intensity of the diagnostic absorption bands of the mineral. Quartz, for example, can be detected in concentrations as low as 1% in typical sedimentary rocks from its strong bands in the wave-length region of 12.5 to 12.8 microns. Less than 2OY0 montmorillonite is difficult to detect in the same rocks, because its spectrum lacks any strong diagnostic absorption bands. Lower percentages of montmorillonite have been detected, but only in the absence of other common clay minerals such as illite and kaolinite. Opal is not readily detected in the presence of quartz, since the diagnostic bands of both are in the same region, 12.5 to 12.8 microns. Most of the carbonate minerals can be detected in as low as 2y0 concentration, owing to their strong diagnostic bands in the wavelength region from 13 to 14 microns. Certain polymorphs can be detected in the presence of each other. Aragonite is readily distinguishable from calcite, with which it is chemically identical. I n Figure 4 the diagnostic bands of these minerals in the wavelength region of 14 microns are shown. Sulfates such as gypsum and barite can be recognized in low concentrations from their strong bands in the wave-length regions of 10 and 15 microns.

I

QUARTZ

c

GYPSUM

1

I

12.0

I

I 14.0

13.0 WAVE

I

15.0

LENGTH IN MICRONS

Figure 3. Infrared Absorption Spectra of 3Iinerals in Wave Length from 12 to 16 Microns Curves displaced vertically

CS2 +ALKAT€RGE C

truni of the limestone is not sufficiently diagnostic in the wavelength range from 8 to 11 microns to identify the clay mineral in the sample. By dissolving the limestone in hydrochloric acid and concentrating the clay mineral it was possible to identify it as montmorillonite. Templates similar to this can be made up for any group of minerals whose absorption bands are concentrated in a specific region of the spectrum, and are very useful for qualitative interpretations. Minerals have fewer and more intense infrared absorption bands than most organic compounds, which makes less interference in a mixture containing several substances. However, many minerals have no narrow and well-defined absorption bands in their spectra suitable for absorbance measurements. Another feature in the spectra of minerals is their strong absorption a t longer wave lengths. Gypsum, barite, orthoclase, and other silicates all have strong absorption bands a t wave lengths greater than 15 microns. The infrared absorption bands of sev-

YI

T

:n.: 20

I

PURE ARAGONITE

MONTMORILLONITE

Suspended i n carbon disulfide and Alkaterge-C Curves displaced vertically

I

14.0

I

14.0

WAVE LENGTH IN MICRONS Figure 4.

ILLITE

WAVE LENGTH IN MICRONS Figure 5. Infrared Absorption of --OH Group in Clay Minerals

PURE CALCITE I

KAOLIN I TE

Infrared Spectra of Polymorphs

I

One disadvantage of the dry powder mount is that there is too much light scattering in the short wave-length ranges from 2 to 4 microns. The authors and others ( 5 ) have attempted to obtain well-defined absorption bands in this range by reducing the particle sizes below 1 and 2 microns. However, this was effective in only a few instances. An alternative method, similar to the method used successfully by Dolinsky ( 1 ) with organic chemicals, is to suspend the powder in carbon disulfide. The mineral is first ground to the desired particle size (less than 5 microns) and then is shaken vigorously with a solution of carbon disulfide containing 1% or less of a detergent such as Alkaterge-C (substituted oxazoline available from Commercial Solvents Corp.). The suspension is introduced into a standard 0.5-mm. infrared liquid cell to obtain

1172

ANALYTICAL CHEMISTRY

the spectrum. Unfortunately, a detergent has not yet been found which stabilizes the mineral suspensions for more than about 10 minutes, so only a part of the spectrum is examined from one mount. The method appears to show considerable promise for determining the bonded and unbonded hydroxyl content of minerals] as shown in Figure 5. Here the spectra of the clay minerals kaolinite, illite, and montmorillonite are shown in the wave-length region from 2 to 4 microns. The presence of unbonded -OH shows up a t 2.75 microns and bonded -OH in t h e region around 3.0 microns. The bands are well defined a n d ultimately it may be possible to make quantitative measurements from them. The dotted line shows that Alkaterge-C does not interferewiththis analysis; in fact, when thedouble-beam spectrometer is used, the carbon disulfide detergent solution can be put in the reference beam to eliminate its spectrum. QUANTITATIVE ANALYSIS

In addition to the qualitative identification of minerals in rocks, it has been possible to determine quantitatively minerals having sharp, well-defined absorption bands. Included in this group are quartz, kaolinite, orthoclase, calcite, dolomite] and other carbonates. The method used consists of selecting a diagnostic absorption band of a mineral which is in a wavelength position where there is a minimum of interference by other minerals. Infrared spectra of several different weights of the mineral are obtained and working curves are made plotting absorbance against the weight of powder in milligrams on a unit area of the salt block. A series of such curves is shown in Figure 6. In each case the base-line method (2) of computing absorbances is used, the base lines being drawn between regions of no specific absorption. Generally the positions of the base lines were determined from the examination of a large number of spectra of the mineral involved. Data points for the curves in Figure 6 were obtained from the spectra of the pure minerale, 2nd also controlled mixtures of the pure minerals.

by bare spaces in the film (7). I n order to eliminate bare spaces, a thicker film is used for quantitative analyses than for qualitative work. The thicker films are made simply by adding more powder to the sodium chloride window. Qualitative data can be obtained from film thicknesses varying from 0.2 mg. per sq. cm. on up. Quantitative data are obtained from films in the range from about 0.6 to 1.4 mg. per sq. cm. On the salt blocks used with the Baird double-beam instrument this is equivalent to a sample weight of from about 7 to 15 mg. When films are made up in the higher weight ranges and examined with the petrographic microscope, it is found that in most cases more than 90% of the window area is covered. Even with complete coverage] however, the film is never completely uniform. This is an error inherent in the method, but with sufficient practice in making these films, and by checking actual thickness variations with a microscope, the films can be kept fairly uniform. .4n additional error arises from scattering of the infrared radiation by whatever large particles are present. This can be minimized by proper grinding of the sample. In Table I some data are presented relating to the analysis of synthetic mixtures of pure minerals. Montmorillonite cannot be determined quantitatively. It was added to these mixtures merely to evaluate its effect on the analyses of the other minerals. In one case it was possible to make up a synthetic mixture containing calcite, dolomite] quartz, and montmorillonite to about the same composition as that of a natural rock. When the spectra of the synthetic sample and the natural rock were superimposed they were found to be virtually identical. In Table I1 some infrared analyses of sedimentary rocks are compared with x-ray analyses made with the Geiger-counter spectrometer. Both methods are accurate within about 10% of the amount present. In most cases there is good agreement between the infrared and x-ray. APPLICATIONS

In actual application infrared spectra have been used in solving various geological problems. The polarizing microscope has long been the standard method of determining the mineral composition of rocks. It is virtually useless with fine-grained rocks such as clays and shales because of their small particle size. Infrared analysis has been particularly useful in the quantitative deter-

“I

Table I.

a

MILLIGRAMS OF SAMPLE

Infrared Analysis of Synthetic JZixtures

Components Dolomite Calcite 41ont r u o illonitea ~ Quartz Dolomite Kaolinite Quartz Dolomite Calcite Montmorillonitea By difference

Compoqition, Weight 77j Known Deterniined 36 34 46 45 20 19 20 23 40 37 40 10 6 J 14 14 72 75 B 8

Table 11. Infrared and X-Ray Analysis of Sedimentary Koclcs

Figure 6. Relation of Absorbance and Weight of Powder Sample Xo.

On some of the curves there is a considerable scatter of points, such as for absorbance measurements a t the 12.8-micron band of quartz. This is because in the powder film method there are certain inherent errors which limit the accuracy of a quantitative analysis to about 5 to 10% of the amount present. The principal error in making quantitative analyses of powder films by infrared arises from the nonuniformity of the powder film. Fluctuations in film thickness cause some error, but .the greatest error is caused

1

2

3 4

Components Quartz Kaolinite Quartz Calcite Quartz Dolomite Calcite Quartz Dolomite Calcite

Compo%ition,Weight Infrared X-ray 33 31 45 41 9 11 71 76 7 3 7 5 84 86 8 7 19 21 40 41

1173

Figure 7 . Infrared Spectra of Yatural Sediments

20

30

40

60

5.0

70

80 90 100 IlO WAVE LENGTH IN MICRONS

120

80 9.0 10.0 I10 WAVE LENGTH IN MICRONS

I20

I3.0

i40

I5O

160

IO0

80 -

OOLOGAH 20

3.0

40

FM.

5.0

60

7.0

130

140

150

160

.o WAVE LENGTH IN MICRON:

Z

a

100

I-

c

DRILLING MUD 0 20

3'0

410

510

610

7b

810 910 Id0 WAVE LENGTH I N MICRONS

IlO

Ii.0

;l,o

16.0

100

TRAVIS PEAK S S . MISSISSIPPI

WAVE LENGTH IN MICRON:

100 80

-

60 -

,o WAVE LENGTH

IN MICRONS

inination of the calcite, dolomite, and quartz content of suchrocks,andithasbeenapplied to several more limited problems, such as following the penetration of drilling mud constituents into a formation. I n Figure 7 the spectra of R few nat>ural sediments are shown. At the top is the spectrum of a doloniitic shale from the Green River formation in I-tah. It is possible to identify the absorption peaks of dolomite, calcite, and quartz from this spectrum. The absorption from 9.5 t o 10 microns is due to a clay mineral, possibly niontmorillonite. The second spectrum in Figure 7 is of some nodules found in the Oologah formation of Oklahoma. The doublet a t about 7.0 microns characteristic Of the phosphate radical. The general nature of this curve identifies this material as the phosphate mineral apatite. The third spectrum is of a recent sediment,, mud from the Laguna 3Iadre on the Texas Gulf Coast. Absorption peaks due to calcite, quartz, and illite :IWapparent in this spectrum. .A side-wall core of the TravisPeaksmdstone in Mipsissippi was examined for (.vitlence of penetration 1)y the tlrilling mud. The spectrum of the drilling mud and core are shown in the fourth and fifth curves, respectively. The drilling mud contained about 15% calcite n-hich was ]licked up in thecourseof drilling through higher formations. There was no calcit'e in t'he sand to begin n-ith, hut penetration of the mud introduced a small amount, as m n be seen from the small carhonate peak a t about. 7 microns. Some kaolinite also Tvould have been introduced ; however, there was more in the sand to begin n-ith than there was in the drilling mud, ne can be seen from a comparison of t,he peaks a t about 11 microns. The last curve in Figure 7 is the spectrum of R

ANALYTICAL CHEMISTRY

1174 mineral whose composition could not be easily determined in the It was identified as barite from the teristic peaks of this mineral a t wave lengths of 10.2 and 15.7 microns. The few examples cited show that the infrared spectrophotometer can be a useful instrument in both the qualitative and quantitative determination of the mineral constituents of rocks. LITERATURE CITED

(1) Dolinsky, M.,J . Assoc. Ofic. Agr. Chemists, 34, 748 (1951).

(2) Heid, J. J., Bell, M. F., and White, J. v., I.vD. ENG.CHEM., ANAL.ED.,19,293 (1947). (3) Hunt, J. M., Wisherd, &I. P., and Bonham, L. C., ANAL.CHEM.. 22,1478 (1950). (4) Keller. W. D.., SDotts. . J. H.. and Binns. __ D. L.. Am. J. Sci.. 250. 463 (1952). ( 5 ) Launer, J. L., Am. Mineralogist, 37,764 (1952). (6) Miller, F.A., and Wilkins, C. H., ANAL.CHEM.,24, 1253 (1952). (7) Wilchinsky, Z., private communication.

RECEIVED for review March 31, 1953. Accepted May 16, 1953. Presented at the Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 7, 1952.

Semiquant itative Spectrographic Method for Analysis of Minerals, Rocks, and Ores C. L. WARING AND C. S. ANIVELL U . S. Geological Survey, Washington, D . C . The quantity and complex nature of materials received for analysis in the spectrographic laboratories of the U. S. Geological Survey have emphasized the need for a spectrographic method to determine a maximum number of elements in a limited time with a reasonable degree of accuracy. The semiquantitative method described determines 68 elements in one arcing of a 10-mg. sample. The method has been used to complete 245,000 determinations during a 3-year period. Each determination is reported as a concentration range or bracket (0.001 to 0.01, 0.01 to O . l % , etc.). A chemical check of 500 such determinations showed 92% in agreement; the remaining 8570 agreed to within one bracket. The method requires a minimum of sample handling, thus reducing the chances of contamination, detects low concentrations of elements, and is rapid. Analyses have been completed on n wide variety of materials.

T

HE spectrographic laboratories of the United States Geo-

logical Survey receive for analysis each year a large number of samples of radioactive minerals, rocks, and ores in connection with the investigation of radioactive raw materials, a program which the Survey is undertaking on behalf of the Atomic Energy Commission. It is desirable to know the trace element content of this material, and for many purposes it is necessary to gain aome knowledge of the amounts of major constituents present without taking the time necessary to make chemical analyses. There is need for a spectrographic method to determine a maximum number of elements in a limited time with a reasonable degree of accuracy. In spectrographic parlance such a method is termed “semiquantitative” and the results are usually reported in orders of magnitude of weight percentages of the elements (not the oxides). A survey of the literature (2-4) reveals that several similar methods are being applied in other laboratories on materials of a different nature with different standardization procedures. As a starting point, pom-dered samples were used in order to eliminate costly dissolution techniques. As it was not intended to provide complete quantitative data, the internal-standard, buffer, and carrier distillation methods were not considered, nor could any definite advantage be anticipated in using the cathode-layer method with its critical optical alinement. ilhrens ( 1 ) compared the cathode-layer and anode-excitation methods and found them approximately equally sensitive.

An investigation of the various excitation sources indicates that the direct-current arc gives the best sensitivity or, in other words, produces a higher degree of sample excitation so that lines emitted by elements in low concentrations may be recorded. The direct-current arc supplied by the Multisource (Applied Research Laboratories) produces a high degree of sensitivity, with the added advantage of simple operation. Because of these advantages the Multisource was selected to excite the samples, which were mixed with graphite and placed in the crater of a graphite electrode a t the positive side of the arc. The purpose of the graphite addition was to prevent the formation of mobile beads of molten salts, oxides, or metals, to assist in the volatilization of elements of high boiling points or of elements in extremely nonvolatile compounds, and to steady the arc with a minimum of spraying or mechanical loss of sample. The method has been applied to the following materials: Minerals

Allanite Alunite Anthophyllite Apatite Auerlite Bastnaesite Bayleyite Betafite Billietite Bostonite Brannerite Britholite Carnotite Celestite Chalcopyrite Columbite-tantalite Corvusite Cyrtolite Davidite Diderichite Eschynite Euxenite Feldspars Fergusonite Florencite

Fritxcheite Galena Garnet Guadarramite Hewettite Huebnerite Hummerite Ianthinite Idocrase Johannite Magnetite Martite 3Ielilite Microlite Monazite Montroseite Xiccolite Novacekite Parisite Perovskite Pickeringite Pitchblende Pyrite Renardite Samarskite

Bauxite ores Boron ores Clays Coal ash Diabase Granite Leach products

Rocks Lead ores Lignite ash Limestone Pegmatites Phosphate rocks Placer concentrates Rhyolite

Sabugalite Scheelite Schoepite Schroeckingerite Siderite Smaltite Sphalerite Sphene Stibnite Studtite Thorite Torberni te Tornebohmite Tourmaline Tyuyamunite Uraninite Uranocircite Uranophane Uranothorite Vanoxite Volborthite Wernerite Zippeite Zircon

Rock salt Sandstones Shale Sulfur ores Syenite Volcanic rocks