UW Luminescence Dating
Laboratory
The University of
Washington Luminescence Dating Laboratory offers a dating service for ceramics,
lithics, and sediments. The laboratory specializes in archaeological
applications and is particularly interested in research projects in which
luminescence can solve archaeological problems not assessable by other dating
techniques. The current director, James
Feathers, is a professionally trained archaeologist but has been involved
in the luminescence field since 1986. The laboratory has produced more than 300
dates, almost all from archaeological contexts, since Dr. Feathers became
director in 1993.
Luminescence dating is a
rapidly expanding field. Recent advances in technique and instrumentation have
improved both the accuracy and precision of the method, with the result that
luminescence dating is becoming an important player in quaternary science. The
advantage luminescence dating has over other techniques is the ability to date
directly events of archaeological and geological interest: the last heating of
ceramics and lithics and the last exposure of light for sediments. This often
eliminates the need for associational arguments and the uncalibrated loss of
accuracy involved therein. The technique is not as precise as radiocarbon, but
errors of about 10 percent or better are commonly obtained.
Costs are negotiable, but
generally we charge $400 per ceramic sample and $500 per lithic sample. An
initial sediment sample costs $800, but additional samples from the same
locality will be cheaper depending on the difficulty in dating. University
indirect charges, currently 14.7 percent, are added to each project. Turn
around times for ceramics and lithics are about 6 months, depending in part on
sample queue, but quicker times can be negotiated in special circumstances.
Sediment dating usually takes longer. In no case can we guarantee a successful
date, although our past performance suggests that we achieve dates that agree
with independent dating evidence (which itself is not always accurate) more
than 85 percent of the time.
For researchers applying
for grant money to fund luminescence dating, the laboratory will be happy to
assist in writing the grant proposal at no charge.
Clients interested in using
our dating service are advised to contact the laboratory before collecting
samples for a feasibility assessment. For sediment dating, it is also advisable
to have a luminescence specialist visit the site. Collection instructions are
included as a link to this site. Also linked are a laboratory bibliography and
current projects being undertaken by the laboratory. We also provide web site
links to other luminescence laboratories throughout the world.
Instrumentation: The laboratory is equipped with
three glow ovens: an automated Daybreak 1100 reader capable of measuring TL,
OSL, and IRSL; an earlier Daybreak version reader for TL only; and a Littlemore
reader capable of measuring TL, OSL and IRSL. We have two alpha counters, one
high resolution gamma spectrometer and a flame photometer for measuring
radioactivity. Other resources include two calibrated Sr-90 beta sources, one
Am-241 alpha source, and an Applied Physics solar simulator.
James Feathers
University of Washington
Box 353100
Seattle, WA 98195-3100 USA
Ph: 206-685-1659 or 206-782-9421
FAX: 206-543-3285
Email: jimf@u.washington.edu
What is Luminescence
Dating?
Luminescence dating is
based on the accumulation of stored energy that occurs in some crystalline
materials as a function of natural radioactivity. This energy may be released
in the form of a luminescence signal by exposure to sufficient heat or light,
which in effect act as zeroing mechanisms. The intensity of a luminescence
signal is proportional to the amount of stored energy, which in turn is
proportional to the amount of absorbed radiation. If the radiation dose rate is
a constant function of time, the luminescence intensity can then be related to
the time since the last zeroing event (exposure to heat or light). Since the
major sources of natural radioactivity are radionuclides with exceptionally
long half-lives, a constant dose rate can be assumed, except where geological
processes have altered the radioactive concentrations. This latter situation
can often be detected and sometimes corrected.
The following expression is
used to derive ages in luminescence dating:
Age (t) = equivalent dose (Gy)/dose rate (Gy/t).
Gy stands for gray, the
unit of absorbed radiation and t is unit time. The equivalent dose is the
amount of radiation dose equivalent to what was necessary to produce the
sample's natural luminescence signal. It is measured by calibrating
luminescence sensitivity against artificial irradiation applied in the
laboratory. It requires isolating signals that have been thermally stable over
time and which increase with applied radiation. The luminescence signal can be
stimulated in the laboratory by heating (thermoluminescence, TL) or by exposure
to intense light, either from the visible spectrum (optically stimulated
luminescence, OSL) or from the infrared (infrared stimulated luminescence,
IRSL). The dose rate is the sum of alpha, beta, gamma and cosmic dose rates in
the sample and its immediate environment.
For introductions to the
method, the reader can consult two books by Martin Aitken, Thermoluminescence
Dating (1985,
Academic Press), which provides an overall discussion of luminescence and is
still the best reference for pottery dating, and An Introduction to Optical
Dating (1998,
Oxford University), which is a more current review of sediment dating. More
specialized reviews on a number of topics can be found in a special volume of Radiation
Measurements (1997,
Vol. 27 #5/6), edited by Ann Wintle. A review of archaeological applications
can be found in an article by Richard Roberts in that same issue and also in Journal
of Archaeological Method and Theory (1997, Vol. 4) by the director of this laboratory.
Sample Collection
Instructions
Luminescence dating is
applied to nonmetallic materials. Two classes of materials can be dated: heated
materials (primarily ceramics and burned lithics) and buried sediments. The
former is dated to the last heating to about 500°C or so. The latter is dated
to the last exposure to light.
Heated
Materials
Selection Criteria:
Any coherent ceramic
material that has been heated to at least 500°C is potentially suitable.
Incidental ceramics (e.g., house daub, hearth fragments) may be variable in
this regard even though often ideal subjects. Consequently, select pieces most
likely on contextual and physical grounds to have been heated to the highest
temperatures.
Lithics are more
problematic. There are at least two concerns. One is the possibility for
translucent pieces to be optically bleached. When samples are being collected
specifically for luminescence dating, they should immediately be placed in
opaque containers upon excavation to minimize this problem. For surface finds
or lithics exposed to light for long periods after excavation, restriction to
opaque specimens is recommended. The other concern is selecting samples that
have been heated to at least 450°C. Burned chert or flint can be recognized by
discoloring or the appearance of luster, crazing or pot-lid fractures, although
none of these can indicate how high the sample has been heated. Unfortunately,
a sufficient heating can only be ascertained by the luminescence measurements
themselves. While the physical characteristics can be used to select samples, a
good precaution is to submit several samples, only some of which may prove to
be suitable for dating. It should be noted that flaking properties optimized by
heat treatment require temperatures around 370°C, too low for luminescence.
Better prospects are lithics that have been accidentally fired in hearths or
heat treatment failures (fired too high). Suitable lithics for dating do not
have to be flints or cherts. Fire cracked rocks from fire pits have been
successively dated.
Finally, luminescence can
be a destructive technique. The laboratory will do what it can to save valuable
portions of the sample, but in general do not expect to have anything returned.
In special cases, dating can be performed on only very small portions of
samples.
Size
The outer 2 mm layer of the
sample must be removed in the laboratory to eliminate effects of light exposure
and influence from environmental beta radiation. After removal, enough material
must be left over for analysis. While improved techniques have reduced the
amount of sample required for luminescence measurements, larger pieces are
still advantageous since more can be done with them. Ceramics at minimum should be about 5 mm thick and 3 cm
in diameter; better precision is obtained with larger pieces. Flint or chert
artifacts need to be at least 10 mm thick and 5 g in weight.
Context
Samples should be selected
from contexts that maximize two parameters: 1) compositional homogeneity within
a sphere of 30cm radius around the sample, and 2) contextual stability over the
post-depositional period. Both requirements stem from facility in measuring the
sample dose rate. Gamma radiation has a range in soils of about 30 cm, so
determining the gamma dose rate requires knowledge of the radioactive
composition of the environment within 30 cm. This is more difficult the more
complicated the environment is. Numerous clasts or stratigraphic boundaries can
create a geometric complexity that can frustrate an accurate dose rate measure.
Also composition that has changed through time (for example, by leaching or
other movements that redistribute the radioactive impurities) will complicate
the dose rate assessment. Less than optimal conditions, however, should not
deter dating. Often it is possible to account for the uncertainties, although
the precision may be less and the cost more. Some complicated contexts,
moreover, are not necessarily complicated in terms of radioactivity. If all the
different strata and clasts have a similar radioactivity, the dose rate is
quite easy to determine. It is always advisable to contact the laboratory
before collecting samples to discuss environmental complications and the
probability for successful dating.
Treatment
Avoid excessive heating of
the sample (do not heat above 100°C) and exposure to ionizing radiation
(ultraviolet, infrared, x-rays, gamma-rays) above ambient levels. Any such
treatment can reduce the luminescence signal and produce too young ages.
Airport and postal inspection X-rays are usually not of significant dosage, but
avoid them if possible.
Do not perform any
unnecessary treatment of the samples. Particularly avoid contact with materials
of unknown composition that may be absorbed by the sample (e.g., glues,
preservatives, detergents). Supply a list of treatments that have been applied.
Contaminants can affect both the luminescence signal and the radioactive assay.
Washing with distilled water is okay.
Sediment Sample
At least 100 g of sediment
representing that in which the sample was buried, or on which it was resting in
the case of surface items, should be supplied in a plastic bag. Care should be
taken to avoid any sorting as different size fractions will typically differ in
composition as well. If there is a scatter of small stones (pebble size) or
small artifacts, these should be included and larger sample submitted. Exposure
of the soil to sunlight, ultraviolet, x-rays, etc., does not matter.
The sediment sample is
required for assessment of the environmental gamma dose rate to which the
sample has been exposed (up to 30 cm). The sediment should come from within 30
cm of the sample but will be representative of the full environment of the
sample only to the extend the latter is homogeneous. In complicated
environments different samples for each area of suspected different composition
(clasts, different strata, features, etc.) will need to be supplied, assuming
the geometry is not too complicated. In exceeding complex environments, in
situ measurements
of the environmental dose rate is preferred. A common procedure is to bury a
small copper capsule (about 1 cm in diameter by 3 cm long) containing a
sensitive dosimeter (calcium sulfate is commonly used) for a full year in the
same location from which the sample was extracted. The laboratory supplies the
capsules. On the day they are to be buried, the capsules must be heated to
400°C, or until they begin to glow red (avoid going too high at risk of melting
the solder sealing the ends). This zeroes the dosimeter. Heating can be done
with a propane torch, camp stove or other portable heat source. The capsule can
be inserted in the sediment by driving a 30cm long plastic conduit (of slightly
larger diameter than the capsule) into the profile at the sample location. A
string or wire is tied around the capsule, which is pushed to the back of the
conduit with the string extended out the front for easy retrieval in a year's
time. A full year is recommended to account for annual changes in moisture
content (if this varies greatly from average during the time of burial, such
information should be provided). On the day of retrieval an accompanying
"travel monitor" capsule is heated to 400°C. The monitor accompanies
the capsule retrieved from the ground to record the gamma dose received en
route to the laboratory. An alternative to dosimeters is to use a portable
gamma spectrometer. The laboratory does not have one at present, but access to
one is possible under special circumstances. A problem with in situ radiation measurement is that the
sample environment is destroyed to some degree in the acquisition of the
sample, so that the dosimeter never measures the radiation from the exact same
environment as the sample, only something close to it, the significance of
which varies directly with the complexity of the environment. Nevertheless,
on-site measurements are beneficial for corroboration of laboratory
derivations.
In selected cases ceramics or
lithics can be dated without a sediment sample. This works only where the
samples all come from the same environment so that the external radiation is
effectively constant. If all the samples can be assumed to be of the same age,
a ratio scale date for them can be obtained by a technique called isochron
dating. If the samples are of different ages, only interval scale dates can be
obtained. In either case several samples need to be dated, so costs are higher.
The laboratory should be consulted for feasibility of dating without sediment
samples.
Water Content
Because the radiation
absorption coefficient of water differs from than of sediments or pottery, a
correction must be made for the water content of the sample and its
surroundings. Some approximation must be made of the average water content, how
this varies between contexts and with respect to surface conditions, how it
might vary seasonally and whether any long term variation is known. The
location of the water table with respect to the sample is also desirable. If
current water conditions are representative, a sample can be collected in an
air tight container and sent to the laboratory for measurement. Other
information that might be useful in making an approximation are seasonal and
long-term climatic variation and grain size distribution of the sediments
(clay, silt, sand, etc.). While usually not precisely known (except in
permanently saturated or permanently dry conditions), any information related
to water content is useful in improving the precision of the dates. Water
content of lithics is much less important than ceramics because of much lower
absorption capabilities.
Other Information
- Drawings and, if
possible, photographs of the contexts from which the sample was taken. The
former should roughly show sample locations and approximate distances to any
boundaries, such as change in soil type, and whether stones or other clastics
are within 30 cm. This information is useful to judge the appropriateness of
the sediment sample for the environmental radiation dose rate and other context
related matters.
- Latitude, longitude and
altitude of the site, and depth of burial of the sample. This data is necessary
to compute the cosmic ray contribution to the dose rate.
- Anticipated ages of the
samples, if known. This helps select procedures and appropriate artificial
irradiation exposures. While knowing the expected age will not influence our
results, it does prompt us to look for explanations when our date does not
correspond with expectations.
Buried
Sediments
Selection Criteria
Since the dating event is
last exposure to light, the most important consideration in selecting sediment
samples for dating is the likelihood that the sample was exposed to sufficient
light prior to deposition and burial. Sufficient light may be as little as a
few minutes of direct sunlight, although complete bleaching usually takes
several hours. Aeolian deposits have a high likelihood of long exposures
because of airborne time. The bleaching of fluvial deposits depends on the
sediment load and the turbity of the flow, deposits from meandering streams
generally being well bleached while glacial outflow tending to be less so.
Colluvial deposits can be problematic, depending on the nature of the
deposition and the grain size of the sediments. Buried soils have been
successfully dated because of cycling of A-horizon sediments to the surface by
pedoturbation. Bleaching is also affected by shade conditions and cloud cover.
Any kind of crystalline
material is potentially datable, but most research has been conducted on quartz
and feldspars. The latter tends to have a longer dating range than quartz but
can suffer from anomalous fading. Use of other material will require more basic
research and thus entail a higher cost. Analysis can be done on coarse grains
(generally 90-125µm for quartz, 125-250µm for feldspar) or fine grains (usually
4-11µm), but other sizes have been used as well.
Where sediments down a
profile are being selected, it can be useful to have one sample from the
surface that represents a "zero-age" sample that can be used to test
for extent of bleaching.
Sample Size
Any single sample should be
a minimum of 250g, but larger samples are desirable. If the same sample is to
be used for radioactivity measurements, at least 350g is needed. More optimal
size is 500 to 1000g. In special circumstances very small samples may be
sufficient. The vertical extent of the sample will affect resolution depending
on the rate of deposition. For example, a sample drawn from a 10cm profile section
that accumulated in 1000 years will not be able to resolve at a scale less than
1000 years. The number of samples depends on the specific problem to be
addressed and financial resources, but more than one sample from any given
context will certainly increase precision.
Collection Procedure
Exposure of the sample to
light must be avoided at all costs to avoid an underestimation of the age. This
means the sample must be collected under conditions that minimize light and
must be stored in opaque containers. There are several solutions to the
collection problem. One way is to collect the samples at night, preferably with
no moon, aided by low intensity red/orange filtered flashlights. Red to orange
light causes insignificant reduction to the luminescence signal. Exposure to
within 1 cm of the sample can be done in daylight; then under night light the
overlying cm is scraped off and the sample excavated. More convenient are
collections in daylight. A common daylight solution is to drive an opaque pipe
(plastic or stainless steel) into the sediment and cap both ends. The ends,
which will have been exposed, can be removed in the laboratory. It is important
to make sure the sediment is secure, not loose. Otherwise mixing of the exposed
ends with the unexposed center will occur during transport. If the sample does
not fill the entire pipe, some kind of stopper is needed to keep the sediment
from moving around. In capping the ends or stopping, do not use material that
will oxidize or decompose when in contact with the sample (often damp) and thus
contaminate it. Aluminum foil, for example, will oxidize when in contact with
wet soil for long periods of time. In sediments that are too hard to drive a
pipe in, a block of sediment can be removed as long as it keeps its integrity.
The outer surfaces are then removed in the laboratory. Samples can also be
collected by coring. Various split-core samplers with light tight sample
containers are available. In general, the procedure is to take one core that
will be exposed to light in order to define the stratigraphy and the depth of
the desired sample. Then core down to just above where the sample will be
collected and attach the split core sampler to retrieve the sample.
Radioactivity
Where the sample comes from
a very homogeneous deposit, the sample itself can be used to measure
radioactivity. Where the environment is more complex, considerations similar to
those for heated materials must be taken into account. In a sediment profile
where movement through time of radionuclides, either up or down, is suspected,
several 70-100g samples down the profile will be useful to model the
distribution of radioactivity. No special precautions are needed for these
"radioactivity" samples, except to avoid altering their composition.
Other information
For context considerations,
treatments, moisture contents, data for cosmic ray dose, drawings, and age
estimations, the same procedures as for heated materials should be followed.
Moisture contents are somewhat more critical for sediments since they have a
larger influence on the dose rate than for pottery or lithics.
Laboratory Bibliography
Readhead, M. L., Dunnell,
R. C., and Feathers, J. K., 1988, Recent addition of potassium: a potential
source of error in calculating TL ages. Ancient TL 6(1):1-4.
Feathers, J. K., 1993,
Radioactive disequilibrium in the TL dating of southeast Missouri pottery, Radiation
Protection Dosimetry 47:655-658.
Feathers, J. K., 1993,
Thermoluminescence dating of pottery from southeastern Missouri pottery and the
problem of radioactive disequilibrium. Archeomaterials 7:3-20.
Kaylor, R., Feathers, J.
K., Gottfried, M., Hornyak, W. F., and Franklin, A. D., 1993, Important
date/strange material.Ancient TL 10:40-44.
Vandiver, P. V., Feathers,
J. K., Kaylor, R., Gottfried, M., Yener, K. A., Hornyak, W. F., and Franklin,
A. D., 1993, Thermoluminescence dating of a crucible fragment from an early tin
processing site in Turkey. Archaeometry 35:295-298.
Feathers, J. K., 1995,
Optically stimulated luminescence of heated materials. In Ceramic Cultural
Heritage,
proceedings of the International Symposium of the 8th CIMTEC-World
Ceramic Congress and Forum on New Materials, edited by P. Vincenzini, pp.
368-374. Techna, Faenza
Dunnell, R. C., and
Feathers, J. K., 1995, Thermoluminescence dating of surficial archaeological
material. In Dating in Surface Context, edited by C. Beck, pp. 115-137. University of New
Mexico Press, Alburquerque.
Kaylor, R., Feathers, J.
K., Hornyak, W. F., Franklin, A. D.,1995, Optically stimulated luminescence in
Kalahari quartz: bleaching of the 325°C peak as the source of the luminescence.
Journal of Luminescence 65:1-6.
Roosevelt, A. C., Lima da
Costa, M., Lopes Machado, C., Michab, M., Mercier, N., Valladas, H., Feathers,
J. K., Barnett, W., Imazio da Silveira, M., Henderson, A., Sliva, J., Chernoff,
B., Reese, D. S., Holman, J. A., Toth, N., and Schick, K., 1996, Paleoindian
cave dwellers in the Amazon: the peopling of the Americas. Science 254:1621-1624.
Feathers, J. K., 1996,
Luminescence dating and modern human origins. Evolutionary Anthropology 5:25-36.
Feathers, J. K., 1997, The
application of luminescence dating in American archaeology. Journal of
Archaeological Method and Theory 4:1-66.
Feathers, J. K., 1997,
Luminescence dating of early mounds in northeast Louisiana. Quaternary
Science Reviews (Quaternary Geochronology) 16:333-340.
Feathers, J. K., 1997,
Luminescence dating of sediment samples from White Paintings Rockshelter,
Botswana. Quaternary Science Reviews (Quaternary Geochronology) 16:321-331.
Saunders, J. W., Mandel, R.
D., Saucier, R. T., Allen, E. T., Hallmark, C. T., Johnson, J. K., Jackson, E.
H., Allen, C. M., Stringer, G. L., Feathers, J. K., Frink, D. S., Williams, S.,
Gremillion, K. J., Vidrine, M. F., and Jones, R., 1997, A mound complex in
Louisiana at 5400-5000 years before present. Science 277:1796-1799.
Feathers, J. K., and Rhode,
D., 1998, Luminescence dating of protohistoric pottery from the Great Basin. Geoarchaeology 13:287-308.
Michab, M., Feathers, J.
K., Joron, J.-L., Mercier, N., Selos, M., Valladas, H., Valladas, G., Reyss,
J.-L., and Roosevelt, A. C., 1998, Luminescence dates for the paleoindian site
of Pedra Pintada, Brazil. Quaternary Science Reviews 17:1041-1046.
Feathers, J. K., in press,
Why luminescence dating deserves wider application in American archaeology. In It's
About Time, edited
by Stephen Nash, University of Utah Press.
Feathers, J. K., and Bush,
D. A., in press, Luminescence dating of Middle Stone Age Deposits at Die
Kelders, Journal of Human Evolution.
Feathers, J. K., and
Migliorini, E., Luminescence dating at Katanda – A reassessment. Quaternary
Geochronology,
submitted.
Feathers, J. K., Dating the
Howiesons Poort assemblages at Diepkloof, western Cape, South Africa, by
luminescence methods. South African Journal of Science, submitted.
Feathers, J. K.,
Thermoluminescence of Southwestern United States ceramics: a date list. Ancient
TL, in preparation.
Feathers, J. K.,
Thermoluminescence of Eastern North American ceramics: a date list. Ancient
TL, in preparation.
Dykeman, D. D., Towner, R.
H., and Feathers, J. K., Tree-ring and thermoluminescence dating techniques:
methods for confidently dating early Navajo sites, American Antiquity, in preparation.
Feathers, J. K., Brooks,
A., and Yellen, J. Reassessing dating evidence at Katanda. Science, in preparation.
|