Thermoluminescence as a Tool for Scientific Research

F. Daniels, Ch. Boyd, and D. Saunders*)

Thermoluminescence is the emission of light that occurs when a body is heated to a temperature below the incandescence temperature. It is observed in crystals (for example, alkali halides) that have first been irradiated with X-rays or radioactive rays and then subjected to subsequent rapid heating. Radiation of high energy knocks out electrons, some of which become localized in defects of the crystal lattice. If, as the temperature is raised, the increase in kinetic energy proves sufficient, these electrons are released and light is emitted.

Thermoluminescence is extremely sensitive to changes in the structure of crystals. Changes in the concentration of minute impurities, differences in the method of crystallization, and various methods of physical treatment all have a strong influence on the intensity of thermoluminescence. Similar sensitivity to structure is also characteristic of other properties of a body connected with its solid state. Such, for example, are catalytic activity, the magnitude of the dielectric constant and magnetic susceptibility, absorption of light, fluorescence, and phosphorescence. Proceeding from the close connection between thermoluminescence and the above-mentioned properties, the idea arose of the possibility of studying these properties with the aid of thermoluminescence.

The use of thermoluminescence for the analysis of minerals and for quality control of feldspars in the ceramic industry was described by Deribere^1,2. Garlick et al.^3 used thermoluminescence to investigate the mechanism of phosphorescence. Other interesting applications of this phenomenon were mentioned in a number of preliminary communications^4–6. In the present article, achievements in the application of thermoluminescence to the solution of various scientific problems are briefly reviewed.

*) Science 117, 343 (1953).

APPARATUS

The method of measuring thermoluminescence with the aid of so-called glow curves was developed by Urbach\(^7\) and improved by Randall\(^8\) and Boyd\(^9\). At the present time the measurement procedure has been so simplified that a large number of the most diverse thermoluminescent substances can readily be investigated.

The light intensity, measured by a photomultiplier, and the temperature, determined by a thermocouple, are recorded simultaneously on a self-recording potentiometer. With uniform heating of the specimen at a rate of \(1^\circ\mathrm{C}\) per second, a graph of the type shown in Fig. 1 is obtained. The temperature course is represented by a straight line; the intensity—

Fig. 1.

of the light, as can be seen, passes through a series of maxima and minima. The crystalline substance is prepared in the form of a powder or of a square plate with a side of \(1\ \mathrm{cm}\) and a thickness of \(1.5\ \mathrm{mm}\), obtained by cleaving crystals or by cutting the mineral with a diamond saw. Because of the scattering of light within the specimen, the radiation intensity of the powder is usually small; however, on the other hand, owing to the averaging of inhomogeneities of the specimen, the reproducibility is considerably better. The diameter of the powder particles is usually brought by grinding to \(0.1\text{--}0.2\ \mathrm{mm}\); approximately \(20\ \mathrm{mg}\) of such powder is transferred to a thin glass plate and moistened with a drop of water containing a negligible amount of detergent. After drying, the powder adheres sufficiently well to the glass. The uniform rise in temperature shown on the graph is achieved by regulating the heating of a silver plate by means of an autotransformer with smooth adjustment; as the plate is heated, the rate of heat removal increases, and it is necessary gradually to increase the current.

The measuring apparatus is enclosed in a light-tight box. After the temperature corresponding to dark-red heat is reached, which makes observation of thermoluminescence difficult, the heated plate is removed and, with the aid of dry ice, is rapidly cooled in preparation for the next measurement.

Crystals were obtained chiefly in two special γ-irradiators, consisting of two concentric aluminum tubes, the space between which is filled with cobalt powder \(^{10}\). Activation (formation of Co\(^{60}\)) was carried out in a nuclear reactor. By this method it was possible to create a cavity 2.2 cm in diameter and 8 cm long, uniformly irradiated by γ-rays with energies of 1.1 and 1.3 MeV and intensities of 6000 or 1400 roentgens per hour, which corresponds to 4 or 1.3 curies.

MECHANISM OF THERMOLUMINESCENCE

Thermoluminescence is observed only in substances possessing an ordered structure (for example, crystals) or a semi-ordered structure (for example, glass). In addition, electrically this substance must be an insulator or a semiconductor. Finally, upon irradiation with ionizing radiation, the absorption of the latter must cause the formation of a cloud of electrons in the lattice.

Thermoluminescence in certain crystals can be induced by using various forms of high-energy radiation—hard or soft X-rays, γ-rays, α- and β-particles. In most crystal lattices, in order to detach an electron and impart to it energy enabling it to migrate through the crystal, an energy on the order of 10 eV is sufficient; however, for thermoluminescence to arise it is necessary that local levels be present which are capable of accepting electrons.

There are several types of possible local levels: a) defects and vacant sites in the crystal lattice, arising during formation of the crystal or produced as a result of mechanical pressure or thermal treatment; b) defects caused by thermal motion, the number of which increases with temperature; c) distortions caused by impurity ions, the size of which differs from the size of the ions composing the crystal lattice; d) ionic dislocations or “holes” arising under the action of radioactive radiation.

One of the most frequently encountered types of local level is a vacant site formed as a result of the removal of a negative ion; an example is the absence of a chlorine ion in the sodium chloride lattice. Filling such an empty site with an expelled electron leads to the formation of the so-called \(F\)-center, which acts as a whole in the absorption of light. It is believed that

coloration of crystals subjected to irradiation by X-rays, or the coloration in fluorite and other minerals containing radioactive impurities, owes its origin to the formation of similar \(F\)-centers and other states of localized electrons \(^{11,12}\). It has long been known that, upon heating, the coloration disappears \(^{13}\); it is less well known that such disappearance of coloration is usually accompanied by thermoluminescence.

The number of localized electrons in the crystal lattice may be large and, nevertheless, thermoluminescence may not be observed upon heating, since for this there must exist a definite mechanism for converting the excess energy of the electrons into visible light. It is commonly assumed that this mechanism is based on certain emission centers in the lattice, associated with impurities, as occurs in many artificial phosphors and natural fluorites \(^{14—16}\); it is also possible that these centers are caused by displacement of lattice ions or by other disturbances of the latter.

The influence of impurities on the character of thermoluminescence was discussed by Alt and Steinmetz \(^{17}\). In our investigations it was shown that individual peaks of the glow curves of fluorite correspond to light of different colors. This indicates that the local levels are closely connected with definite quenching centers, which, as was shown, are due to rare-earth impurities \(^{16}\). It may be thought that in the case under consideration the electrons are localized in defects formed by foreign ions.

The formation by \(\alpha\)-particles of ionic dislocations in various materials was studied by Seitz and Sleater \(^{18}\). The colored rings observed around radioactive inclusions in minerals are apparently due to electrons localized in dislocations created by \(\alpha\)-particles. In some feldspars such rings give very intense thermoluminescence, and from this it is concluded that the local levels responsible for this luminescence were formed as a result of bombardment by \(\alpha\)-rays. Changes in the density of crystals as a result of irradiation by X-rays were observed by Estermann \(^{20}\). These changes were attributed to the formation of vacant sites in the lattice under the action of the radiation. In lithium fluoride, with each subsequent irradiation by \(\gamma\)-rays from \(\mathrm{Co}^{60}\), an increase in thermoluminescence is observed. It may be thought that the increase in sensitivity is the result of the formation of additional local electron levels by the \(\gamma\)-rays.

Each peak on the glow curve corresponds to a definite local level of electrons. Peaks corresponding to low temperatures can be eliminated if heating is stopped at a sufficiently low temperature. The intensity of the thermoluminescence corresponding to a given peak is determined by the difference between the spe-

with the rate at which electrons fall onto local levels under the action of radiation and the rate of their thermal removal at the irradiation temperature. Upon a sudden increase and subsequent stabilization of the temperature of the irradiated crystal, the intensity of luminescence decays according to a quite definite law, from which one can calculate the activation energy of the given type of localized electron.

The character of thermoluminescence under continuous irradiation has been studied in considerable detail, and the data obtained will be reported separately. As a rule, saturation is observed at a certain intensity of the incident radiation, depending on the type of crystal and the nature of the radiation. Under prolonged irradiation, the height of the low-temperature peaks decreases and, at the same time, new high-temperature peaks are formed. Alkali halides and limestones reach saturation at approximately 100,000 roentgens when irradiated with \(\gamma\)-rays from \(\mathrm{Co}^{60}\). Continuous irradiation may lead either to a constant thermoluminescence intensity or to a decrease in its intensity.

TYPES OF THERMOLUMINESCENT MATERIALS

All alkali halides thermoluminesce. In general, they are characterized by two large peaks on the curve of thermoluminescence intensity as a function of temperature, whose positions depend on the sizes of the ions; in atoms of smaller size, thermoluminescence begins at a higher temperature, and the coloring \(F\)-centers have an absorption maximum located at shorter wavelengths1. In lithium fluoride, both ions of which are small, the “coloration” caused by irradiation is observed in the ultraviolet region, and the thermoluminescence maxima are observed approximately at \(220^\circ\mathrm{C}\) and \(320^\circ\mathrm{C}\). In addition to alkali halides, the following substances thermoluminesce after irradiation with \(\gamma\)-rays: calcite, dolomite, fluorite, aluminum oxide, magnesium oxide, gypsum, quartz, glass, certain catalysts, feldspars, certain fired clays, and ceramics.

A large number of inorganic crystals are being tested for thermoluminescence, and the method of measuring it is being improved toward increasing the sensitivity of the apparatus. As a rule, after irradiation with \(\gamma\)-rays, solid transparent or translucent crystals with a simple structure most often thermoluminesce.

Naturally, in a number of cases the light emitted during thermoluminescence is not detected by the apparatus used, since it may lie in the ultraviolet or infrared regions of the spectrum. It is impossible to investigate organic substances in air because of the interfering effect of light arising as a result of combustion. There are, however, grounds for thinking that if certain organic substances are irradiated at the temperature of liquid air—

and then heat them to room temperature or higher, thermoluminescence will arise in them. If alkali halides are irradiated at the temperature of liquid air, new peaks appear on the glow curve at low temperatures1. When irradiation is carried out at room temperature, this radiation is recorded as fluorescent, since it is emitted during irradiation.

The presence of impurities strongly affects the form of the glow curves (thermoluminescence); until a theory has been created that explains the difference in the activation energies of the various local levels, it will be difficult to determine whether a given peak is characteristic of a given crystal lattice or of a given impurity, even if present in a negligible amount.

Preliminary work has been completed on studying the influence on thermoluminescence of impurities and of mixtures of salts that have crystallized together from a common melt[^22]. If, for example, 1 mole percent of silver chloride is added to molten sodium chloride, the intensity of the thermoluminescence of the crystallized salt proves to be 100 times greater than in the case of sodium chloride alone. Traces of cupric chloride and manganous chloride, on the other hand, quench thermoluminescence. If potassium bromide is added to sodium chloride, both salts are fused, and the melt is recrystallized, it turns out that the curve showing the dependence of thermoluminescence intensity on mole percent has a maximum, and the glow intensity at the maximum is considerably higher than when sodium chloride or potassium bromide is taken separately. Apparently the lattice distortion associated with the presence of ions of different sizes creates holes into which electrons knocked out by γ-rays fall.

The thermoluminescence spectrum over the entire visible region depends on the type of crystal and is rather sensitive to contamination.

APPLICATIONS OF THERMOLUMINESCENCE

Dosimetry

In many crystals the intensity of thermoluminescence is almost proportional to the amount of absorbed γ-radiation, and on the basis of this fact considerable work has been carried out on the practical application of this property for γ-ray dosimetry[^23]. For these purposes the best crystal proved to be lithium fluoride. With this crystal it was possible to measure, with great convenience and high accuracy, doses from 10 to 1000 roentgens or more by means of a photomultiplier and a microammeter1. With a nonportable apparatus it was possible to measure doses of less than 1 roentgen. Lithium fluoride is stable, insoluble, and is not affected by moisture. For pro-

for carrying out measurements an artificial crystal weighing about \(1/2\) gram is sufficient. Initially crystals with an area of about \(1 \text{ cm}^2\) were used, but recently small beads of powdered lithium fluoride enclosed in a thin metallic case have been used. Experiments are being carried out with the aim of obtaining greater uniformity, but even now it may be noted that measurements of various types of rays have proved quite satisfactory. Dosimetric crystals, or pills, are small and can easily be used under such experimental conditions as are inaccessible to ordinary radioactivity meters. They have been successfully used in one hospital to measure the intensity of internal irradiation of cancer patients who had received a dose of radioactive isotopes. The crystals were swallowed by the patient and collected again after a day or two; the dose in roentgens was determined by comparing the intensity of thermoluminescence with its intensity in a crystal that had received a definite dose of X-rays.

Thermoluminescent analysis

Every substance is characterized by its glow curve, which has certain peaks at a specified temperature, and in this sense glow curves are similar to spectrograms with their lines at quite definite wavelengths. The glow curves of each type of crystal depend on impurities, on the thermal treatment of the crystal, and on its previous history. This means that measurements of thermoluminescence cannot be used for the analysis of chemical compounds. However, they may find applications in the identification and control of the composition of minerals. Thus, for example, the glow curves of different types of aluminum oxide depend on impurities and previous thermal treatment, but all samples of one type have identical glow curves. If it is necessary to deal with an oxide of some one type, then the latter can easily be determined from its glow curve. Another example. The thermoluminescence curve of ordinary window glass depends strongly on the presence of impurities, chiefly iron oxide; if we are interested in glass of a definite quality, then the thermoluminescence glow curve could be used as a convenient method for checking the quality of each batch of glass. Among the materials that can be classified, identified, and controlled by the indicated method, one may name limestones, ceramics, and optical crystals.

A set of approximately 30 different sorts of clay was subjected to a comprehensive investigation and, in particular, to chemical analysis and to measurement of viscosity and absorption of infrared rays. Some, but not all, of these clays had characteristic thermoluminescence glow curves.

luminescence, undoubtedly deserving of further study. It may be expected that the inorganic part of some soils has characteristic glow curves that can be used for the identification and classification of soils.

Often the light emitted on heating a material that has been subjected to preliminary irradiation has a characteristic color. Limestones emit yellow light, dolomites—orange. It is possible that the existence of this orange light will find application in a field method for prospecting magnesium carbonates, although it is not excluded that this color is characteristic not of magnesium carbonate itself, but of an impurity present in it.

Catalysis

Boyd and Hirshfelder²¹ observed that thermoluminescence curves may prove useful for studying the effectiveness of surface catalysts. The same internal stresses and lattice defects that promote the bringing together of adsorbed molecules to distances optimal for chemical reactions may also lead to the formation of local levels onto which electrons knocked out by γ-rays fall, and from which electrons are removed as a result of heating. It is possible that the intensity and character of thermoluminescence may find application for measuring catalytic effectiveness. In any case, the thermoluminescence glow curve can be used to monitor the process of preparing a given catalyst. For this purpose, a satisfactory standard catalyst must have a glow curve of characteristic form.

Fig. 2.

To test the above, an alumina catalyst was studied; it turned out that the thermoluminescence intensity of a satisfactory catalyst made from this material is considerably greater than that of a poor catalyst made from the same substance. Several types of commercial catalysts were investigated. The results of measurements for three catalysts are shown in Fig. 2. The area under peak 2 correlates with catalytic effectiveness, as is shown graphically in Fig. 3. Many catalysts,

in particular, catalysts of the colloidal type do not thermoluminesce, while in others it is not possible to establish a clear correlation. Nevertheless, the number of catalysts tested is so large that

Fig. 3.

one may assert that thermoluminescence is a new means of evaluating catalytic properties and may prove useful in the search for new catalysts.

Radiation Damage

Most high-energy radiations produce damage in crystals. The study of this phenomenon is important in the design of nuclear reactors and for the interpretation of changes in the structure of crystals caused by radioactive rays over millions of years.

The fraction of the energy of γ-radiation that can be stored in a crystal in the form of localized electrons is very small. With the aid of a photomultiplier calibrated against a standard lamp and thermocouple, it was found that less than 1/10,000 of the energy of γ-radiation absorbed by lithium fluoride is emitted in the form of thermoluminescent light24.

Thermoluminescence is one of the methods for studying the process of energy accumulation under bombardment by high-energy rays and can also be used to investigate those damages that occur in this process. Atomic displacements, if they occur, can lead to the accumulation of larger amounts of energy than is observed in thermoluminescence, but instruments for measuring heat are far less sensitive than instruments for measuring light. Thus, up to the present time

thermal measurements did not permit establishing the accumulation of heat in crystals irradiated by $\gamma$-rays. For the final clarification of this question, investigations have been begun with more sensitive apparatus and with longer exposures to radioactive radiation. In some crystals containing considerable amounts of uranium, bombardment by $\alpha$-rays led to strong displacements in the crystal lattice. These displacements are so great that the X-ray diffraction pattern characteristic of the given crystal completely disappears. When the temperature of such minerals, which have undergone considerable destruction, is raised, the stored energy is released in the form of heat and the X-ray diffraction pattern characteristic of the normal lattice is restored. In one case the amount of energy released reached 25 calories per gram^24.

Thermoluminescence of Rocks

Many rocks and minerals exhibit thermoluminescence even without preliminary exposure to X-rays or $\gamma$-rays in the laboratory. Most limestones, ground to millimeter size, emit a bright white or orange light for several seconds when heated on a pan brought to the temperature of a dull red heat. Upon reheating after cooling, no light emission is observed. However, a short exposure to X-ray or $\gamma$-radiation restores the property of thermoluminescence, and by suitably choosing the dose one can obtain thermoluminescence with the former intensity. Natural thermoluminescence is often mentioned in the geological literature^13, ^25, ^26, but it was generally regarded as a kind of curiosity. Some investigators attributed this light to the combustion of organic material. Meanwhile, it has repeatedly been shown that the cause of thermoluminescence is traces of uranium, thorium, or other radioactive elements contained in the rock as impurities^6, ^27. These radioactive impurities may be present at concentrations of less than $10^{-6}$, but $\alpha$-, $\beta$-, and $\gamma$-rays are emitted over millions of years, and the thermoluminescence effect is partly cumulative. It turns out that the natural thermoluminescence of rocks is often a more sensitive means of detecting traces of radioactivity than a Geiger counter or a scintillation counter. The connection between radioactivity and thermoluminescence in certain minerals was previously noted by Elsworth^25, Koler^28, and Alt and Steinmetz^17.

An unexpectedly large number of minerals possess natural thermoluminescence. Of more than 3000 specimens investigated, chiefly limestones and granite, visible thermoluminescence was detected in approximately 75 percent of the cases. There is no doubt that, with the use of more sensitive apparatus,

one will be able to observe thermoluminescence in a greater number of cases. Almost all limestones and acidic volcanic rocks possess natural thermoluminescence. Calcium and magnesium carbonates emit light whose color changes from light yellow to orange; potassium and sodium feldspars emit light from white to blue-violet.

It is precisely these minerals that are the principal source of the thermoluminescence observed in the rocks investigated, such as limestone fossils or inclusions in shales and the cementing material between grains of sandstone. Some fluorites containing uranium thermoluminesce especially strongly. Thermoluminescence is also observed in quartz, nepheline, topaz, halite, and spodumene, if they contain impurities of uranium or thorium.

In one children’s amateur collection of 36 minerals, thermoluminescence was observed in more than one third of the specimens. In another collection of 65 rock-forming minerals, natural thermoluminescence was observed in 34 specimens. The light intensity varied from a level sufficient for reading a newspaper to an intensity barely detectable.

Fig. 4.

In all minerals with natural thermoluminescence, the brightness of the latter could be considerably increased by irradiation with γ- or X-rays, and many of the minerals that did not glow could be made thermoluminescent by brief irradiation with the indicated rays. Of the 65 minerals mentioned above, 47 proved to be thermoluminescent after irradiation with γ-rays.

In limestone, the increase in thermoluminescence caused by irradiation with γ-rays at a dose of 140,000 roentgens is characterized by the glow curves shown in Fig. 4. Peak 2, corresponding to a temperature of approximately 300°C, changed almost not at all, but peak 1 at 215° became seven times more intense. In the irradiated specimen, at lower temperatures two rather intense peaks are found which are completely absent in natural (unirradiated) specimens, since, because of the comparatively high temperature of the Earth, electrons are not retained at the local levels. When irradiating in the laboratory, however, the time between irradiation

and measurement is so small that transitions from local levels do not play a noticeable role. The kinetic energy corresponding to the peak at \(300^\circ\)C is so large that almost all localized electrons remain at local levels over geological time, and there is no doubt that still more intense irradiation by natural radioactive rays would lead to an increase of peak 2 up to saturation.

The thermoluminescence of rocks is heterogeneous in character, i.e., on heating some components emit light whereas others do not. If thin samples of limestone or granite are annealed and placed on a hot plate, then by focusing a camera on the surface of the sample it is possible to obtain sharp photographs in the light emitted by the heated (below the incandescence temperature) samples.

Fig. 5.

An example of such a photograph is given in Fig. 5. The luminescence of the sample is due to the transition of electrons from local levels of the calcium carbonate lattice contained in the limestone fossil sample; the electrons reached the local levels under the action of radiation from negligible impurities of uranium. The brighter photograph on the left was obtained with the same sample, but after irradiation with \(\gamma\)-rays at a dose of 140,000 roentgens. In the thermoluminescent photograph of the surface of a granite sample, bright spots situated in the immediate vicinity of inclusions of high radioactivity are clearly visible.

At high altitudes cosmic rays may be partially responsible for the activation of thermoluminescence, but in rocks underground or in rocks containing uranium or thorium in concentrations above \(10^{-6}\), the role of cosmic rays is quite negligible in comparison with radioactivity.

The glow curves of rocks are highly variable: the height of the peaks and the corresponding temperature change from rock to rock. The higher the concentration of uranium or thorium and the greater the age of the rock, the more intense the thermoluminescence, since in this case the rock has been subjected to irradiation for a longer time. No less important factors are the presence or absence of chemical impurities and physical defects. The chemical environment during crystallization and the physical action of temperature and pressure on rocks over the course of their geological history exert a great influence on the nature and intensity of thermoluminescence.

Stratigraphy

Since glow curves are sensitive to the chemical and physical conditions under which a rock originated, they can thereby be used to determine whether two samples of sedimentary rocks were formed at the same time and under similar conditions, despite the geographical separation of the places where they were formed. The belonging of two rock samples to one and the same bed is usually determined by the character of the corresponding fossil remains, and such determinations are of practical value in oil exploration. If, during drilling, a definite sequence of beds is observed that leads to the discovery of an oil field, then the repetition of this same sequence in another place may mean that the same basic geological characteristics are also present there. Thermoluminescence glow curves may serve as an additional means of correlating sedimentary formations. In some rocks, for example in volcanic ash or in Precambrian limestone, there are no fossil remains and, consequently, they cannot serve as a means for identifying beds.

The reliability of the thermoluminescent method was tested on ten samples taken from one and the same bed of a large limestone quarry. The different beds could be visually differentiated over a distance of almost half a mile. The thermoluminescent curves of all ten samples taken along this bed were similar to one another in form, but differed strongly from those obtained for samples taken from other beds lying several feet above or below the given one.^29

Extensive investigations have been carried out to determine the possibility of using thermoluminescence as a means of correlating limestones. For this purpose, glow curves obtained after excitation by γ-rays prove more convenient than natural thermoluminescence curves, since a larger number of peaks suitable for identification is obtained in the low-temperature region, whereas under natural conditions, under the influence of the earth’s temperature, these peaks have disappeared. As a rule, in the glow curve

for most limestones there are four peaks: at 120–140°, 150–190°, 210–250°, and 290–310°C. Special studies on correlating limestones with their thermoluminescence glow curves have been carried out by Saunders^27,29 and Bergström^30. A special investigation of subsurface stratigraphy by means of glow curves was carried out by Parks^31.

Determination of Age

If it were possible to introduce corrections for chemical impurities and physical defects and for self-absorption of light in the rock, it would become possible to determine the age of a rock from the intensity of thermoluminescence and from the uranium content, or from thorium, measured by the α-activity of the sample. Each α-particle creates a certain number of localized electrons responsible for thermoluminescence. The amount of light should be proportional to the product of the number of α-particles emitted per year by the number of years elapsed since formation of the crystal. This method of age determination is considerably less convenient than the radioactive-carbon method, but the latter is applicable only to carbon-containing materials whose age does not exceed 25,000 years. The thermoluminescence method is still less reliable than methods based on determining the ratios of the contents of lead and uranium, helium and uranium, or potassium and argon, but these radioactive methods also have their limits of applicability. Any new method for determining the age of rocks and minerals deserves attention, and at present the thermoluminescence method, which gives an age counted from the time of the last crystallization of the mineral, is being intensively studied. For limestones it was developed by Zeller^32 and for fluorites by Saunders^33. Alpha activity was determined over a large surface of the rock with the aid of a scintillation counter. If the uranium content is of the order of 10^-6, then it is precisely the determination of α-activity that limits the accuracy of the method. In one method, the area under the peak in the natural glow curve corresponding to a high temperature is used as the standard of comparison; in another method, the number of localized electrons which are assumed to have been produced by α-rays is brought to saturation with the aid of γ-rays. The absorption of light in the sample itself is determined experimentally, which makes it possible to introduce corrections for self-absorption of light in semitransparent samples. Limestones of known geological age serve as standards.

The main uncertainty of the method described is due to chemical impurities and defects of the crystal lattice, which affect the intensity of thermoluminescence. It may be hoped that more satisfactory results will be obtained with the aid of a polonium source with an activity of 1 curie. Bombardment of the surface of a thin sample by α-rays from this source during

of one second is equivalent to bombarding limestone with uranium rays for a million years. By this method it is possible to remove, to a significant extent, external contamination. If radiation damage could be neglected, then the time of bombardment by a polonium source required to obtain a thermoluminescence intensity equal to the natural thermoluminescence of the rock would be a direct measure of the number of $\alpha$-particles emitted since the time of formation of the rock. Knowing the number of $\alpha$-particles emitted in the rock per unit time, it is easy to calculate the time required to create the observed thermoluminescence.

At the present time, the possibility is being investigated of determining the firing date of limestones and ancient clay vessels. Other applications of thermoluminescence in laboratory technique and in applied problems are also being studied.

CITED LITERATURE

1. M. Deribere, Argile 188, 5 (1938).
2. M. Deribere, Rev. sci. 76, 383 (1938); Chem. Zentr. Blatt 11, 397 (1944).
3. G. F. Garlick, Luminescent Materials, London, Oxford Univ. Press (1949).
4. F. Daniels, C. A. Boyd and D. F. Saunders, Science 109, 440 (1949).
5. F. Daniels and D. F. Saunders, Science 111, 461 (1950).
6. F. Daniels and D. F. Saunders, Final Rept., U. S. Atomic Energy Comm. Research Contract No. AT (11—1)—27 (1951).
7. F. Urbach, Sitzber. Akad. Wiss. Wein. Math. naturw. Kl., Abt. IIa, 139, 363 (1930).
8. J. T. Randall and M. H. F. Wilkins, Proc. Roy. Soc. (London) A184, 347, 366, 390 (1945).
9. C. A. Boyd, J. Chem. Phys. 17, 1221 (1949).
10. D. F. Saunders, F. F. Morehead, Jr. and F. Daniels, in press.
11. S. C. Lind and D. C. Bardwell, J. Franklin Inst. 196, 357 (1924); Am. Minerologist 8, 171 (1923).
12. S. Meyer and K. Przibram, Sitzber. Akad. Wiss. Wein. Math. naturw. Kl., Abt IIa, 123, 643; Neue Jahrb. Mineral. 1, 277 (1916); Chem. Zentr. Blatt 2, 633 (1916).
13. C. Doelter, Lab. Mineral., Univ. Vienna, Radium 7, 58 (1911).
14. H. Steinmetz, Fortschr. Mineral., Krist., Petrogr. 21, 107 (1937); Neues Jahrb. Mineral. Geol. 1, 378 (1937).
15. H. Steinmetz and A. Gisser, Naturwiss. 24, 172 (1936).
16. E. Iwase, Sci. Papers Inst. Phys. Chem. Research (Tokyo) 22, 233 (1933); 23, 22 (1933); 23, 153, 212 (1934).
17. M. Alt and H. Steinmetz, Zeits angew. Mineral. 2, 153 (1940).
18. F. Seitz, Discussions Faraday Soc. 5, 271 (1949).
19. J. C. Slater, J. Applied Phys. 22, 237 (1951).
20. W. Estermann, J. Lievo and O. Stern, Phys. Rev. 75, 627 (1949).
21. L. F. Heckelsberg, Doctoral dissertation, Univ. Wisconsin (1951).
22. C. A. Boyd, Doctoral dissertation, Univ. Wisconsin (1948).

1. F. Daniels, Report of Symposium IV, Chemistry and Physics of Radiation Dosimetry; Technical Command, Army Chemical Center, Maryland, Part I, 148 (1950).
2. F. F. Morehead and F. Daniels, J. Phys. Chem. 56, 546 (1952).
3. H. V. Ellsworth, Can. Dept. Mines, Geol. Survey, Econ. Geol. Ser. No. 11, 55 (1932).
4. M. A. Northup and O. L. Lee, J. Opt. Soc. Amer. 30, 206 (1940).
5. D. F. Saunders, doctoral dissertation, Univ. Wisconsin (1950).
6. A. Köhler and H. Leitmeier, Zeits. Krist. 87, 146 (1934).
7. D. F. Saunders, Bull. Am. Assoc. Petroleum Geol. (in press).
8. R. F. Bergstrom (being prepared for publication).
9. J. M. Parks, Jr. Bull. Am. Assoc. Petroleum Geol. (in press).
10. E. J. Zeller (in press).
11. D. F. Saunders (in press).
12. J. B. Ockerman and F. Daniels (being prepared for publication).

1. 8* UFN, vol. LI, no. 2. ↩↩↩