Thermoluminescence and Age Determination
for Archaeological Samples

Thermoluminescence (TL) can be found in many different archaeological samples or fractions separated by chemical means. If so the age or time since the last heating to over 300° C can be estimated when the background rate for the location(s) is measured.

Instruments for glow curves etc. are available from dosimetry (TLD). From my work at the Harshaw Chemical Co. supplying materials to Farrington Daniels, John R Cameron and others some things about non-radiation-induced peaks in glow curves (NRI-TL) and the handling of these materials can improve the age determination. Even so with the problems of location and background, the accuracy for age can never equal carbon-14. This is a better clock, but TL can apply to more materials. For the age of ancient earthworks both methods are an improvement over inches of black earth (BEA).

Some archaeological samples with a potential for TL that have been reported in fieldwork on mounds in Ohio include: (1) Potsherd, (2) Burnt rock, especially igneous, (3) Burnt clay, (4) Wood ash, (5) Heat tempered bone and antler, (6) Parts of pottery kilns, mixture of ash, clay and rock, (7) Wood charcoal, and (8) Meteorites.

Materials not likely to have TL are glass, slag, sedimentary rock, hydrothermal crystals or deposits, hydraulic cement, kaolin and anything that changes in heating to about 400°C.

While shortwave UV (<2500A) and beta rays can excite TL the background for virtually all locations will be cosmic rays or hard gamma rays from potassium -40. Penetration for activation is not a problem. Light passing through glass will not generate TL..

Since none of the samples listed are transparent, powder will give the best values. It is important that this be of a limited particle size range, 80 to 170 mesh, .177 to .088mm. Thin slices can be used, but both are subject to error from NRI-TL.

Surface distortion, work hardening, for most steps in the procedure can be measured by glow curves as low temperature peaks (NRI-TL). The processing must to be adapted to each class of samples. Once established the same steps must be followed for crushing, etching, screening and heat treatment before the glow curve is run.

For crushing or scraping, temperature is most important, 40-50°C is generally preferred. A liquid like water or alcohol helps and the tools must be hard and sharp. In crushing to pass 80 mesh discard the coarse material after two times to avoid the build up of NRI-TL.

Use a so called ‘diamond crusher’. This consists of a ½ or 3/8 inch drift punch. Working in a shallow flat bottomed hole in a stainless steel cylinder or block. Strike with one good blow of a hammer and separate by wet screening through 80 mesh.

Next etch this through 80 mesh powder, scrapings or drillings with an appropriate solvent that does not leave a coating or residue. Glow curves can indicate the best choice for each type of sample. Some etchants to be considered include: 5% HF, 5% HBF4, 10% H2SiF6, (15% HNO3 + 10% HF), (70% HNO3 + 48% HBr), (5% Na2CO3 at 80° C followed by acetic acid) and water. After etching remove the fines that pass 170 mesh for crushed samples. For scrapings form artifacts or drillings remove all fines that pass 325 mesh to minimize damage.

By wet screening, use the part on 170 mesh or with scrapings on 325 mesh.

Dry and preheat the powder to remove the last of the NRI-TL from the glow curves.

For processing a number of samples in the preheated step, fold the powder in a 2" square of aluminum foil. Samples can be identified by writing on the foil with a #2 pencil on a glass surface. Then after heating this can be recovered as if by magic by smoothing out the foil on glass with a finger. There marks can be recovered for temperature as high as 600°C.

After the glow curve measurements the response or sensitivity can be calibrated to get a response vs. radiation dose curve for the sample. This calibration should be carried as high as the TL response initially found, because of the sensitivity may not be linear.

The preheating step for LiF and CaF2 is usually 140°C for one hour. These samples should not to be much different. To be a valid measure of age use the high temperature peaks rather than total light output.

For scraping, a diamond point will make the least NRI-TL and the motion should not be back and forth. An orbital sander or a drill have an acceptable type of action.

Fresh samples of charcoal or ash should be exhaustively leached with water to be comparable with samples for ancient mounds.

In testing charcoal, remove the carbon by long treatment with 70% HNO3 + 48% HBr about 100°C. This is called wet ashing.

Instrumentation and TLD has been applied to potsherd, but the aforementioned preparation of samples should minimize the error from NRI-TL. The comparison between grinding at 20-25°C and crushing at 40-50°C is like night and day.

The need for development of TL age is urgent considering the failure of C-14 for the Pyramids and the Shroud of Turin.

From members of the David H. Koch Pyramids Radiocarbon dating Project (1999), a weighted average in 1984 of ages by C-14 was 374 years older than historical. 163 new dates in 1995 just added more random ages.

As a fluoride chemist since 1929 my first thoughts were for M+3FCO3 in wood and coal. This does not decompose in boiling acid in testing, The carbonate removal step in C-14 measurements use the mineral Basnasite or Bastnasite. However upon reading an old text for Botany 101 the explanation for the "old wood" problem is obvious but no solution.

In sunlight green plants make carbohydrates (sugar, starch, cellulose etc.) and Oxygen from CO2 and H2O.

For CO2 and pure H2O, the age of wood would be zero. For CO2 in water as (HCO3)-1 the age would be over 5860 years.

Less than 7% of the wood would need to come from hard ground water to test 400 years older than history. For TL age the clock starts at the time of burning for all "ages" of wood.

Some work may be needed for time to leach K-40 after burning or heating vs. rainfall per year at the location but both ashes and charcoal can be tested for TL age.

Of the many tests on the Shroud of Turin at its last showing only C-14 showed it to be a 13th century fake (or visa versa)

Assuming the Shroud is real, C-14 measurements prove that it had been exposed to a high intensity source of cosmic rays and the lack of microscopic pigment confirms that the colored parts could be due to color centers not duplicated by man today.

copyrightSeptember 1999 Dr. Carl F. Swinehart

copyright2000 CKFS Web Page Design