of Hopewell Archeology in the Ohio River Valley
Volume 7, Number 1, December 2006
3. Development of a Protocol to Detect and
Classify Colorants in Archaeological Textiles and its
Application to Selected Prehistoric Textiles from Seip
Mound in Ohio. PhD Dissertation, The Ohio State University,
Columbus Ohio, 2005
By Christel M. Baldia
The goal of this dissertation research
was the development of a protocol to study colorants
applied to archaeological perishable materials such
as textiles even if these colors are no longer visible
to the unaided eye. The protocol is composed of a sequence
of non-destructive or minimally destructive methods
designed to yield a classification of the colorants
that were used prehistorically as inorganic or organic
and pigment or dye. This protocol was then applied to
selected textiles from Hopewellian Seip Mound Group
in southern Ohio to test its effectiveness on actual
The protocol consists of a succession of analytical
methods that have been adapted to be used with very
small sample sizes. If these are sequenced properly,
the efficacy of the protocol is further optimized, thereby
maximizing the acquisition of critical data while minimizing
the need for large amounts of sampling material, and
thus preserving the integrity of the artifacts.
The methods used were forensic photography using different
lighting conditions (simulated daylight, infrared and
ultraviolet), optical and scanning electron microscopy
with energy dispersive X-ray analysis (EDS), and inductively
coupled plasma mass spectrometry (ICP-MS) for elemental
analysis. All methods were first tested on replicated
materials thereby establishing suitable parameters for
their application to archaeological textiles. During
the course of working with the replicas, limitations
of the analytical methods were discovered and addressed
for their use on archaeological materials, i.e. a limited
quantity of material with an unknown chemical composition.
These materials have potentially undergone degradation
processes and could have been exposed to a variety of
contaminants, which all must be considered during the
analysis. For example, the digestion of the sampled
material for the ICP was refined and a more appropriate
instrument was selected based on the results of working
with the replicas.
To reach the goal of using a minimal amount of sampling
material, it is essential that the series of steps within
the protocol are performed in the suggested sequence.
One step builds on the previous one with several key
tasks that must be performed before continuing with
First, a comprehensive and systematic visual examination
of the textile fragments (obverse and reverse side)
must be conducted. Much can be learned if this is done
meticulously. For instance, many details that had not
been expected were discovered when the textiles were
turned to the reverse side. The lighting conditions
must be controlled for this process to guarantee reproducibility.
Otherwise, the results will differ as the lighting temperature
Then suitable textiles that represent types within
an assemblage based on the results of the visual examination
are selected. For instance, the Hopewell textiles were
grouped by commonalities in color and physical condition
such as charring. Magnification should be used if necessary
so no details are overlooked while also controlling
Next, non-destructive forensic photography is used
as a precursor to all the other steps. Before any other
analytical method can be employed, the photography of
the textiles in different lighting conditions must be
performed because it reveals different chemical signatures
due to colorant/substrate interaction even if these
are no longer visible. This optical behavior is used
to discriminate areas of diverse chemistry that can
be correlated to colorant application or contamination.
Thereby, the photography facilitates selective sampling
of these areas, while areas of like chemistry do not
need to be sampled. Thus, purposive sampling enables
focused stratified sampling, increasing the opportunities
for critical data acquisition while decreasing the need
for the copious sampling of the material.
At this point, particulate matter should be collected.
This dust-like particulate matter could consist of small
textile fiber fragments and contaminants, which gives
the first indication to the researcher about the textiles’
state of degradation. The more particulate matter there
is, the more likely the textiles are severely fragile
due to degradation or mineralization. Furthermore, the
particulate can give detailed information about the
textile as a whole, and it can be used for optical microscopy
and possibly other analyses that pertain to the continuous
textile such as infrared spectroscopy.
After that, a detailed macroscopic examination, which
also should be done in controlled lighting conditions,
must be performed. Information about the physical state
of the fibers can be gained. For instance, some of the
Seip textiles showed many fractured and fragments of
fibers within a yarn structure that still appeared to
be intact, therefore making it very fragile. Furthermore,
the colorant penetration and levelness of color can
be determined, and adhering particulate can be observed.
This should be followed by the sub-sectioning the
samples to divide the materials for further analysis.
Subsequently, optical microscopy (OM) of the sub-samples
can be performed to reveal fiber morphology and optical
behavior. Additionally, the particulate that was collected
earlier should be studied. This process should not be
hurried since it takes some time to get accustomed to
the samples and to recognize what is important in these
samples. Images of these micrographs should be collected,
and if a digital camera is used, the colors that are
seen on the screen should be calibrated and matched
to the colors seen in the microscope.
Next, scanning electron microscopy (SEM) on the sub-samples
should be performed. One of the strengths of SEM is
the ability to capture detailed surface morphology that
may not otherwise be detected. Furthermore, the great
magnification that can be achieved with SEM shows details
such as degraded scales from hair fibers or even the
medulla cells that otherwise cannot be detected by optical
While collecting images with the SEM, energy dispersive
x-ray analysis (EDS) of the fibers and all their components,
i.e. fibers and particulate adhering to them should
also be performed. The EDS only gives the relative ratio
of elemental composition of fibers and adhering materials,
which cannot replace quantitative analysis. However,
EDS is a good qualitative method to detect elemental
composition. EDS constitutes a key step that allows
the evaluation of carbon compared to the zero baseline,
hence indicating the presence or absence of organic
compounds such as dyes. If organic components are present,
organic analysis methods should follow as the next step,
while the inorganic path of analysis should be taken
if inorganic constituents are present. Furthermore,
the relative ratios of elements detected by EDS in different
areas of one fiber can be compared to each other, to
other fibers or to the elemental content of the particulate
adhering to the fibers. Thereby, EDS can give information
about ratio of organic and heavy elements, presence
of mineral based colorants, the degree and variability
of fiber mineralization, and possible contaminants.
For the inorganic path of analysis, such methods as
ICP-MS/OES or LA-ICP-MS can be used. For these analyses,
the potential problems that may occur during the digestion
process that prepares spectrometry samples to be analyzed
with various potentially suitable instruments were explored.
When dealing with archaeological textile materials,
it must be assumed that the samples will not digest
well and that sample size is very small; and therefore,
appropriate adjustments must be made. Knowing the relative
ratio of elements present in the samples from results
of the EDS will ease this process greatly, because appropriate
replicas can be created, and the most likely successful
digestion agent can be chosen to perform the spectrometry.
For the organic path of analysis, such methods as gas
or liquid chromatography followed by mass spectrometry,
micro-infrared (IR) and Raman spectroscopy must be explored.
It must be assumed that problems similar to those found
when preparing the samples for ICP-MS will also be found
when other methods such as chromatography are used.
Therefore, a successful trial run of every analytical
method with replicated materials must be conducted before
using artifacts. Thereby, subsequent analyses of the
artifact material are most likely to be successful without
having to be repeated; thereby the amount of sample
material that is needed will be kept at the absolute
Based on an initial visual examination, eleven Seip
textiles were selected and divided into three main color
groups: (1) yellow/brown, (2) turquoise/white, and (3)
charred. These are representative of textiles from the
actual assemblage. An extensive, painstaking visual
examination under controlled light and description of
the selected textiles’ obverse and reverse sides
was conducted. Then both sides of the selected textiles
were photographed in UV, warm and cool visible, and
IR lighting. Based on the findings of the forensic photography,
purposive sampling of the artifacts was conducted. Although
the sample sizes were small, they were representative
of the studied textile assemblage.
The elemental composition of the materials from the
three colors in this group did not show any differences
between the colors. All colors contained a large amount
of copper, some iron and small amounts of soil minerals,
but they also contained large amounts of carbon, and
some sulfur indicating organic materials in the fibers.
It was concluded that the organic constituents of the
fibers had been partially replaced by copper in a mineralization
process. While these textiles were not reported to have
been in contact with copper, they must have been saturated
by copper corrosion products carried by ground water,
i.e. they were near copper albeit not directly adjacent
to it. The encrustations that were observed in the optical
microscopy and the severe brittleness of the fibers
support this statement.
The textiles belonging to the turquoise/white group
were made of milkweed fibers that were painted with
different pigments. These colorants had not penetrated
into the fibers, but adhered to the fiber surface. Different
lighting conditions during the photography showed various
dissimilar aspects of the patterns, indicating differences
in chemical signatures, and thereby different colorants
that had been applied. The elemental analysis indicated
large amounts of copper, and small amounts of other
elements. It was concluded that the white color was
likely kaolin and that some of the other colors had
been mixed with the kaolin or some other types of clay.
These textiles were relatively stable when comparing
them to the state of degradation of those from the two
The charred textiles were extremely fragile. Patterns
no longer visible in fluorescent white light were visible
using photography, and the simulated daylight showed
them best. Ovate motifs in blue, ochre color and different
shades of grey were found. When magnified, it could
not be determined if the colored fibers were penetrated
by dyes, because they are too charred to transmit light.
However, different inorganic particulates adhered to
the outside of these colored fibers, and some of these
deposits were iridescent. Large amounts of iron were
found in these colored fibers, but also some copper.
The orange/red substance without any fiber material
showed the same spectra as did the fiber but with lesser
carbon peak, thereby verifying that the fibers still
contain some amount of organic material. Fibers without
any colorant on them had less iron and higher carbon
and calcium peaks.
Two textiles were identified as composite upon examination.
Both consisted of a combination of several layers of
materials: fabric, leather and matting. Due to the complicated
nature of these specimens, they were only described
but could not be addressed otherwise.
Considerations for Further Research
All research seems to create as many questions as
it provides answers, and with that it provides room
for more work. This research is no exception. Based
on the findings of this study, these are some suggestions
for further work.
- For the digestion process to prepare samples for
spectrometry, ultrasound needs to be applied to the
nitric acid/sample mixture to achieve better digestion..
- Different potential digestion solutions or a combination
of these such as hydrochloric acid (HCL) and hydrofluoric
acid (HF) should be explored. Since these can cause
problems such as the matrix effect, they must be tested
with replicated materials..
- The phytochemistry of many plants that were used
by Native Americans has not been analyzed yet. Colorant
constituents must be identified in such genera that
are known to yield dyes such as the native Indigofera
- Standards of North American dye plants and their
colorants must be created for potentially applicable
methods such as Infrared and Raman spectra.
- The methods to detect organic dye constituents such
as micro-Raman, micro-IR, GC-MS need to be explored,
tested with replicated materials and then applied
to actual artifacts..
- The compositional data reported herein should be
explored as to which inorganic pigments could have
rendered the color to the textiles..
- Quantitative elemental analysis should be conducted
to link the colorants from the textiles to potential
color producing minerals.
- Composite images from the pictures that were taken
should be created, thereby creating a likeness of
what the textile might have looked like in the past,
but also to potentially differentiate and sequence
tasks in the production process.
- The research done by Song and Thompson on the structures
of the Seip textiles should be correlated with the
chemical analysis and microscopy from this research.
- Trace element analysis of copper artifacts should
be done and compared to the copper content of the
- With the discovery of a bast fiber that had not
been identified before, new aspects of Seip material
culture came to light. This bast needs to be identified.
- The two textiles that were identified as composite
herein need to be studied in a separate project.
- The insect piece that was found should be identified,
and further research should consider when insect infestation
of the textile occurred.
The dissertation will be available in full length to
the public through Ohio Link in 8/02:
In the mean time, please refer to:
Baldia, Christel M. and Kathryn A. Jakes
2006 Photographic Methods to Detect Colorants
in Archaeological Textile. Journal of Archaeological
Sciences; in press.
2006 Toward the Classification of Colorants in
Archaeological Textiles in Eastern North America.
In: Archaeological Chemistry: Analytical Techniques
and Archaeological Interpretation, American Chemical
Society Monograph Series; in press.
Baldia, Christel M, Kathryn A. Jakes and Maximilian
2006 Polychrome Hopewell Textiles: Dye Technology
at Seip Mound in Southern Ohio In Proceedings
Textile History Forum, Meeting on October 6-7th at
the Museum in Winterthur DE.
Print this Page
word .doc (46K)
Social Implications of the Colorant Application
Technology to Textiles from the Hopewellian Seip Mound
Site. American Antiquity.
Small Things in Big places: Textiles and Colors
from the Seip Burial Mound Group in Ohio. In
Acts of the XVth UISPP Congress, University of Lisbon,
Portugal, 4-10 September 2006. BAR International Series,