Content uploaded by Giacomo Chiari
Author content
All content in this area was uploaded by Giacomo Chiari on Jan 18, 2014
Content may be subject to copyright.
ABSTRACT. — The field of pigment analysis is
explored through a series of examples taken from
the authors’ work experience. Some analyses are
easy to make, other may require the combination of
various techniques, still others require the use of
sophisticated equipment such a synchrotron or
neutron diffraction analyses. Sometimes the results
obtained do not agree with the accepted theories
about the use of specific pigments (e.g. Egyptian
Blue), or a new, unexpected substance is found to be
used as a pigment (e.g. magnesium oxalate or lead
sulfate). Investigations into illuminated manuscripts
revealed the presence of a variety of verdigris, only
matched by our laboratory synthesis of this
variant. The discovery that the binders used were of
size, resulted in the presence of copper-proteinate
paints from reaction of the media with verdigris. The
red pigment in a manuscript by Von Ems was found
to be made from rhubarb, which is the first time this
colorant has been detected from illuminated
manuscripts. Examination of an Egyptian ushabti of
the New Kingdom revealed a layer of tridymite
white pigment overlying a calcite ground. This
unusual discovery may signify that the Egyptian
pigment palette is much more extensive than
previously thought.
RIASSUNTO. — Lo studio dei pigmenti è affrontato
attraverso una serie di esempi presi da esperienze di
ricerca. Alcune analisi sono molto semplici da
eseguire, altre richiedono la combinazione di più
tecniche analitiche, altre ancora prevedono l’uso di
strumenti sofisticati come il sincrotrone o analisi di
diffrazione con neutroni.
A volte i risultati ottenuti non sono in accordo con
le teorie già attestate sull’utilizzo di pigmenti specifici
(es. il blu egizio), o si scopre l’utilizzo inaspettato di
alcune sostanze per ottenere dei pigmenti (es. ossalato
di magnesio o solfato di piombo). Gli studi di
manoscritti miniati hanno rivelato la presenza di una
varietà di “verdigris” che è stata preparata, in questa
variante, solo dal nostro laboratorio di sintesi. La
scoperta che i leganti erano nel giusto rapporto, è
confermata dalla presenza di pitture con proteinati di
rame, prodotti dalla reazione del supporto con il
verdigris. In un manoscritto di Von Ems, si descrive
la produzione del pigmento rosso a partire dalla pianta
del rabarbaro, che è stato individuato per la prima
volta in un manoscritto miniato.
Esaminando un ushabti egizio del Nuovo Regno è
emerso un livello di pigmento bianco a tridimite
posto su di un fondo di calcite. Questa scoperta
inconsueta, può significare che la tavolozza dei
colori Egizi è molto più ampia di quanto fin ora
conosciuto. (tradotto dall’editor).
KEY WORDS: Pigment analysis, Maya Blue,
Egyptian Greens, Verdigris
Per. Mineral. (2004), 73, 227-237 http://go.to/permin
SPECIAL ISSUE 3: A showcase of the Italian research in applied petrology
Pigment analysis: potentialities and problems
GIACOMO CHIARI* and DAVID SCOTT
Getty Conservation Institute, Los Angeles, California and UCLA/Getty
Archaeological and Ethnographic Conservation Program
* Corresponding author, E-mail: gchiari@getty.edu
An International Journal of
MINERALOGY, CRYSTALLOGRAPHY, GEOCHEMISTRY,
ORE DEPOSITS, PETROLOGY, VOLCANOLOGY
and applied topics on Environment, Archaeometry and Cultural Heritage
INTRODUCTION
Practically all art employs pigments of one
kind or another. This fascinating area of study
frequently makes use of the latest scientific
tests to establish their identity, yet the world of
pigments is a comparatively well-known one,
which has been the subject of several classic
studies, such as the compilation by Gettens and
Stout, originally published in 1942. As long
ago as 1910, Laurie was able to write a useful
account of Egyptian pigments, yet such is the
increase in our knowledge that many
discoveries since that time have resulted in the
need for more recent review, such as that for
Egyptian pigments by Lee and Quirke in 2001.
Despite the recent date of that review, we have
discovered the use in ancient Egypt of still
more pigments, some of which will be
discussed in this short paper. There are many
historic accounts of how pigments were made,
such as the compendium of principally
Venetian manuscripts by Merrifield (1849) and
«La fabbrica dei Colori» (1986), which
contains information about use and
manufacture.
For wall-paintings, we have a recent and
useful review by Howard (2003), which
concentrates on the Medieval period, and
reveals the complexity of some of the colour
schemes employed by Gothic and Mediaeval
artists. Over the last fifteen years a series of
volumes on pigment has been produced by the
National Gallery of Art, Washington, which
promise to remain standard works of reference
for many years to come (Feller 1986; Roy
1993; Fitzhugh 1997), and which document
preparation, identification, geographic location
and period of use for many of the important
pigments of ancient art.
The examination and characterization of
pigments is an interesting task for the
experienced conservation scientist, since some
of them are relatively easy to identify, once
their characteristics and the history of their use
is known and understood. The task, therefore of
analyzing these pigments should not be too
difficult, since until recent times:
(a) Their number is limited to between 50-
100 common types.
(b) The composition of most of them is well
known and published.
(c) They can be categorized in a range of
simple colours, so that the number of possible
alternatives is further restricted (e.g.: a few
blues, a few yellows or blacks etc.
Furthermore, one can make use of other
information such as the time and geographic
provenance of the work of art. In fact, it may
have been retouched with pigments only
available at a later date, in which case further
information can be revealed.
All this is true when the pigments are pure and
the colour can be easily studied and identified.
In this case, simple observation, by use of the
naked eye, or with the help of optical
microscopy can help to solve the problem of
identity. Microscopy can be carried out directly
to observe the surface of the work of art, or
more efficiently by means of a cross section,
which allows for the understanding of the
succession of layers of pigment which may be
present. In case of doubt, a relatively simple X-
ray fluorescence analysis (XRF) can help to
solve the problem. For example, a red-orange
pigment, for which XRF shows only mercury,
allows for an almost certain identification:
cinnabar. But it is not always that easy: in fact,
X-rays penetrate beyond the surface, so that
information from many layers may be mixed
together, with the result that all elements heavier
than sodium present in a multi-layered pigment
or paint layer, would be detected. Therefore,
when the analysis point selected is part of a
white, pink, gray, yellow, orange, red or dark-
purple colour, and XRF only reveals the
presence of lead, the choice still remains among:
lead white PbCO3.Pb(OH)2, used both as white
pigment and as a substrate; massicot (yellow)
PbO; litharge (orange) PbO; minium (red)
Pb3O4[PbO2.2PbO] or plattnerite (an alteration
of lead white, which may be coloured dark
purple or almost black) PbO2. In fact, oxygen,
carbon and hydrogen are not visible by XRF.
In such a complex situation one may have to
refer to some other analytical techniques, such
G. CHIARI and D. SCOTT
228
as X-ray diffraction or Raman spectroscopy,
which can provide the extra information
needed for an unambiguous identification of
the compound rather than the elemental
composition only, which is obtained using x-
ray fluorescence analysis.
Another problem faced by the conservation
scientist, is that the pigments which are present
may be present in such low concentrations or
amounts that they are not easily detected by the
analytical techniques employed. When using
micro-invasive techniques one tends to remove
samples of the order of the milligram, using a
microscalpel or other small tool. This amount
is usually sufficient to carry out an analysis,
given the tremendous progress in analytical
instrumentation provided by modern
technology.
In the case of paintings, one must consider
the fact that the painted layer, which is the
ultimate goal of the analysis, is generally very
thin (~ 50 µm) and may represent only a small
fraction of the sample removed. It is therefore
not unusual for the pigment of interest to fall
below the sensitivity threshold of the method
employed. In this case, the experienced
investigator may be able to work with small
clues, determined from the context of the
investigation.
For example, on the facade of the Casa
Velasquez, in Santiago de Cuba, very small
remnants of a turquoise blue were found. If this
pigment proved to be Maya Blue, more
evidence of the trade of the pigment from
Mesoamerica to Cuba could be ascertained. X-
ray diffraction showed quartz, calcite and
gypsum only (See Fig. 1a). These minerals
obviously did not justify the blue colour. After
a simple acidic attack the pattern in Figure 1b
was produced, in which palygorskite (the clay
composing, together with indigo, Maya Blue,
discussed later) is unambiguously present. The
presence of this particular clay mineral in the
context of this blue pigment is an important
clue for the investigator that the pigment is, in
fact, Maya blue.
If the pigments under investigation are
known and properly described from previous
studies the task of recognizing them is easier.
Although very rarely, it is still possible to find
«new pigments» or alterations such as bronze
corrosion products, never studied before. Since
the majority of the techniques work by
comparison of the actual sample with standard
references contained in databases, an unknown
product may be very difficult to identify and to
characterize.
As an example, let us consider the verdigris
group of pigments, which are based on copper
acetate. These can be simply made by exposing
Pigment analysis: potentialities and problems 229
Fig. 1b – XRD of the same sample after acidic attack.
Palygorskite was present.
Fig. 1a – XRD pattern of a blue paint residue as sampled.
There is no evidence of Maya Blue.
copper sheets to stale wine or sour vinegar. The
verdigris pigment group, consisting of various
copper acetates, have been important as pale
blue or turquoise pigments from antiquity, but
especially in the late mediaeval and renaissance
period, when numerous alchemical recipes
were recorded for their manufacture.
Most of these recipes produce variations on
the copper acetates, usually basic salts, such as
Cu(CH3COO)2[Cu(OH2)]2, which is a light
blue colour, often used as pigments for a
variety of purposes (Scott 2002). These basic
copper acetates could be recrystallized from
vinegar solutions to form the neutral copper (II)
acetate dihydrate, Cu(CH3COO)2.2H2O, which
is green in colour, and was preferable in oil
painting, as being less susceptible to
undesirable reactions or colour changes than
the basic salts.
The identification of some later Medieval
pigments based on the copper acetates can be a
complex task, because the alchemical recipes
for their preparation becomes increasingly
subtle. For example, a fifteenth-century
manuscript, MS 1243 in the Biblioteca
Riccardiana, Padua, (Merrifield 1849), gives
this recipe for a durable azure:
«Mix well one part of sal ammoniac and
three parts of verdigris with oil of tartar until it
is soft and paste-like….then place it in a
glassed vessel under hot dung for a day;
afterwards you will find that the green has
turned to best blue…..»
«Oil of tartar» can be identified as potassium
carbonate, and our synthesis in the laboratory
to replicate this preparation, produced a
mixture of blue and colourless phases, which
gave a complex set of X-ray diffraction data
partially matching potassium copper acetate
2K(CH3COO).Cu(CH3COO)2and ammonium
copper acetate acetic acid C14H50CuN4O20.
Thus, this recipe does indeed produce a durable
blue colour, but identifying this particular blue
concoction in an illuminated manuscript would
be far for easy, especially since the pigment
mixture may react with the media. The
complete identification of some of these blue
and green pigments in works of art continues to
be an interesting analytical challenge for the
years ahead.
For other pigments of unknown composition
it took years to unravel the mystery. An
interesting example is that of Maya Blue. This
beautiful pigment was made by the Maya by
mixing the clay palygorskite with indigo and
warming the mixture to around 100°C, which
produced a blue pigment of exceptional
stability and permanence. To try to discover
some of the reasons for the remarkable
properties of this pigment and its structure, not
only the usual techniques of analysis have been
employed, but also very advanced scientific
tools from different disciplines, such as various
types of spectroscopy, thermal analyses,
electron microscopy, synchrotron radiation and
neutron radiation for highly advanced
diffraction studies. (See Chiari et al., 2003;
Giustetto & Chiari, 2003 and references
therein). For Maya Blue, even a trial
experiment for dating the pigment using 14C
AMS was performed, unfortunately with
dubious results.
WHY DO WE ANALYZE PIGMENTS AT ALL?
To understand and reconstruct the painting
technique (this has value on itself, just as in
other branches of archaeometry).
To reconstruct the history of the use of a
given pigment: this may be of great help in
indirectly dating a painting or even more
frequently, a retouch or repainting. A typical
example is that of the «braghe» or censure
panels used to cover some nudity in
Michelangelo’s Last Judgment. (Chiari, 1996).
To aid in the conservation of the work of art
or to suggest possible problems or alterations
of existing pigments.
Sometimes one can discover unexpected
results: for example, Egyptian blue is one of the
most widely utilized pigments of Egyptian and
Western civilizations up to the fall of Roman
Empire. It was widely used throughout the
Ancient Middle East, starting from about 3600
BC, when it was first synthesized. It was
G. CHIARI and D. SCOTT
230
obtained by heating at 800-900°C a mixture of
calcium carbonate, silica and copper compounds
(possibly bronze scrapings). It is possible that
the Egyptian Blue «secret» traveled along the
Silk Road and reached China, where the
manufacture was readapted to the local mineral
sources. Han Blue, in fact, contains barium
instead of calcium, and has more or less the
same colour, while its homologous member
containing strontium is Han Purple. It is
universally accepted that the manufacture of
Egyptian Blue, in spite of the recipe left by
Pliny, was lost with the advent of the Middle
Ages. Only in the second half of the 20th century
was the recipe recovered thanks to modern
analysis (Schippa and Torraca, 1957), which
proved the formula to be CaCuSi4O10, identical
to the rare mineral cuprorivaite (Pabst, 1959),
although several 19th century scientists were
beginning to unravel the truth, such as Sir
Humphry Davy who wrote about the
constitution of Egyptian blue in 1815, and was
also able to synthesize the pigment. Pliny
mentioned, as ingredients, sand and Cu
compounds, without specifying if the sand was
lime or silica based. Very likely, the sand that
Pliny saw adding to the mix contained both
silicon and calcium. The Renaissance trials to
duplicate the process failed because the sand
added was either silica- or lime-based. Therefore
the unequivocal finding of Egyptian blue on a
painting of the 12th century (Nicolaus Johannes
Last Judgment in the Vatican Museums,
unpublished result) may be puzzling. The
pigment is not occasional or attributable to a few
small retouches. On the contrary it constitutes
the whole dark blue sky in the background. Do
we have to, on the basis of one finding, change
completely the theory regarding the history of
Egyptian blue? Perhaps we should, if all other
possible explanations of a well documented fact
can be excluded. In this case there is a
possibility that the pigment derives from an
ancient small object made of sintered Egyptian
blue, (of which there are many examples) and
therefore the direct synthesis of the pigment is
not the only possible interpretation.
Nevertheless, there is a suggestion that in the
12th century somebody knew how to synthesized
the prestigious pigment. For this reason it is
important to analyze works of art and cumulate
our knowledge, because only when several
findings of this type will corroborate each other,
an important chapter of art history may have to
be re-written.
At least in one specific case a pigment
(haematite) when used on mural paintings can
give directly the date of a painting. This
technique is recent and not completely
developed (since it requires a non trivial
calibration procedure). (Chiari & Lanza, 1997,
1999). It is based on the fact that the small
grains of haematite are magnetized, and
therefore can preferentially orient themselves
toward the magnetic pole while suspended in
the medium used to execute the mural
paintings. When the paint dries, the grains are
trapped in that position (Pictorial Remanent
Magnetization, PiRM) and, provided that no
change in orientation of the support took place
between the moment of painting and the
measure, one can retrieve the position of the
magnetic pole by statistically interpreting the
results of a number of samples (at least 7-8).
The sensitivity of the method depends upon
several factors, beside the measurement errors,
such as the shape of the curve of the magnetic
pole movements at the time to be dated. The
Secular Variations of the magnetic pole
position are not well known due to the lack of
precise data, but at least from 1820 on (when
precise measurements started to be carried out)
they are large enough to guarantee a sufficient
precision (10-50 years). The winding of the
curve creates intersections and therefore
multiple dating for the same magnetic values.
For this reason, some ideas of the date may be
required to discriminate amongst various
possible dates. The application of the method is
limited since it is rather destructive (multiple
samples of about 1 cm2need to be retrieved for
the measurement). In many situations though,
given the extremely common use of haematite
for mural paintings, both in fresco and tempera,
it is possible to find borders or solid red
backgrounds that can be reasonably sacrificed
Pigment analysis: potentialities and problems 231
in order to obtain a dating that cannot be
otherwise achieved.
HOW DO WE ANALYZE PIGMENTS?
The number of techniques now available is
large. The specialist can therefore pick and
chose according to the type of object to be
analyzed, its value, the importance of the
information retrievable, the instruments which
are available, the time and the budget for the
work. In the last few years the technological
advance in the instrumentation was such that
the amount of sample needed for an analysis is
often under a milligram, thus allowing one to
classify most techniques as micro-invasive.
Furthermore, the number of techniques which
can be called non-invasive, since they do not
require a sample to be taken at all, increases
every day. Often the information retrievable
using non-invasive techniques is not complete,
but the use of more than one complementary
technique normally produces the requested
results. The most important thing for the
analyst is to be result-oriented and not
instrument bound (although one tends to be
naturally attached to the technique that one
uses more frequently and, knowing it better, to
attribute to it a larger capability than it has).
Often the simple techniques of optical
examination and micro chemical tests can
provide excellent results with pigment
identification problems. This is because of the
limited number of them and the acquired
experience of the optical microscopist, which
renders polarized light microscopy such a
useful tool for examination and identification.
A trained eye can recognize without fault
several pigments. This may lead to the
temptation of just «trusting your eyes». That
can be dangerous because many times, together
with the main pigment that one may be able to
recognize at sight, there are smaller quantities
of other pigments, which can be interesting. In
other instances our eye may completely fail. It
is the case of a curious finding, still not
understood at all: in the small Church of San
Fiorenzo, in Bastia Mondoví (Piedmont, Italy)
there is a 1474 fresco cycle: on the cupola a
Christ figure is holding a white scroll. The eye
immediately suggests «Bianco di San
Giovanni», which is a form of good-quality
calcite, but since it could have been lead white
as well, a small sample was taken to be
analyzed by XRD. Very surprisingly it turned
out to be glushinskite, a magnesium oxalate.
The most simple explanation was that the
mortar contained magnesium and the oxalate
was the result of a reaction similar to the one
producing weddelite or whewellite (the two
well studied calcium oxalates). Analysis of the
mortar showed that it was made of pure calcite
and glushinskite was limited exclusively to the
white manuscript. Other white parts in the
same painting only contained calcite (pointing
to Bianco di San Giovanni as white pigment).
These types of findings are intriguing, puzzling
and many times are not published at all (it is
hard to write even a short paper on a single
analysis, quite simple per se, leading to a
sporadic finding). On the other hand publishing
sporadic result may be the only way to group
them with those of other researcher and
determine if they are indeed a simple curiosity
after all, or the result of the application of a
previously unknown technique, although rare.
In the case of San Fiorenzo, the only
explanation one can imagine for the result is
that a painter in 15th century found some pure
magnesium oxalate (or some alchemist
prepared some for him) and used it as a
pigment. But until at least some more examples
can be found, the whole matter remains a
simple curiosity. Another example of this type
is the find of anglesite (lead sulfate) on a
lunetta attributed to Antelami and located on
the entrance of the Sant’Andrea church in
Vercelli, Piedmont, Italy (Fig. 2a). The
preparation layer for all faces and hands of the
protagonists of the scene is made of lead white.
The background, of a slightly warmer tonality,
more similar to the stone used for the carving is
instead anglesite, i.e. lead sulfate. On the top of
the lunetta there is a frieze portraying flowers
and an angel who takes to heaven San Andrea
G. CHIARI and D. SCOTT
232
soul. All faces and hands are made with
anglesite. This substance is not known to be a
pigment, to our knowledge. Was it brought
back from England by the bishop who built the
church? Was it commonly used but we do not
know it because we trust our eyes too much?
Another interesting example where simple
assumptions as to the nature of a white pigment
were proved to be erroneous concerns the case
of an Egyptian New Kingdom wooden painted
Ushabti in the collections of the Department of
Religion, University of Southern California. A
micro scalpel was used to obtain a minute
sample of pigment which was mounted for
polarized light microscopy. This sample proved
to have all of the optical characteristics of
calcite, and since this is well-known to have
been extensively employed by the Egyptians as
a white ground and paint, this seemed perfectly
reasonable and in accord with visual
estimation. We decided to check on the
identification of calcite by x-ray diffraction of
another micro sample from the Ushabti, but
this produced a surprise: the white pigment
identified proved to be tridymite, which is a
metastable form of quartz, usually associated
with the high-temperature decomposition of
alpha-quartz. As a result of this, we decided
that this apparently simple white pigment
required further study: The pigment micro
sample used for the diffraction work was
transferred to a stub and examined in low
vacuum mode in the ESEM. The first scanning
electron photomicrograph showed the presence
of rhombohedral calcite crystals, which was
confirmed by EDAX analyses.
Pigment analysis: potentialities and problems 233
Fig. 2 – Sant’Andrea in Vercelli. Lunetta attributed to
Antelami. The white pigment on the faces of the main
scene is lead white, as expected. In the top frieze and in the
background anglesite (lead sulfate) was used instead.
Fig. 3 – Egyptian Ushatbi. The white paint with an
underlayer of calcite and a finishing layer of tridimite can
be seen near the waist.
However, when the particle on the stub was
rotated into a different orientation, the scanning
electron image revealed an assemblage of tiny
hexagonal prismatic crystals, obviously not
calcite. Further study showed that the
Egyptians had used calcite for the ground of
the Ushabti, and had employed tridymite as the
final white paint, since the white of the
tridymite is even purer and more intense than
calcite. The two whites can be clearly
differentiated by crystal morphology.
Tridymite actually forms hexagonal crystals as
a pseudomorph after beta-tridymite, and is
itself monoclinic, but the preservation of the
pseudohexagonal variety is quite common and
an ESEM photomicrograph of the tridymite
from the Ushabti is shown in Figure 3. Where
does this tridymite come from? The mineral
forms from quartz on heating to temperatures
between 870-1470°C and can be found as a
natural mineral or could have been prepared
from heated and crushed quartz. Tridymite is a
metastable polymorph of quartz, but the fact
that it has survived here on the Egyptian
Ushabti shows that it can remain stable for
thousands of years without change. Until
further examples are found and studied, we will
not know how common the use of this white
pigment was in ancient Egypt. (See Fig. 4).
Another difficult case is provided by a
German illuminated manuscript of the 14th
century AD in the collections of the J. Paul
Getty Museum (Scott et al., 2001). This
manuscript, the Barlaam und Josaphat, by
Rudolf Von Ems, with illustrations by the
Diebolt Lauber Atelier dates to 1469 AD. The
manuscript is on paper, which has become
embrittled, with the result that many small
pieces have broken away. One of these
detached fragments with red, light green and
dark green pigment was available for taking a
micro sample for further study. The red
pigment exhibits a pronounced craquelure,
including severe cupping and cracking in some
areas and a fine white precipitate, having the
appearance of a salt, can be seen on the red
under the binocular microscope. We attempted
to identify the green pigment using in situ
examination on a Siemens D5005 x-ray
diffractometer equipped with Gobels mirrors.
This was not successful in giving a good
diffractogram, given the extremely small
amount of pigment. We next tried FTIR with
similar problems: no identification could be
made from the spectrum obtained. Application
of 0.2 microlitres of water also failed to
produce any extractable components. Because
of these difficulties we prepared a micro
sample of the green pigment for polarized light
G. CHIARI and D. SCOTT
234
Fig. 4 – View under the ESEM of the tridymite crystals
forming the outer white pigment layer overlying calcite in
the Ushabti shown in Figure 3.
Fig. 5 – Verdigris curved crystal array of acicular
aggregates (cross polars x120) from a laboratory synthesis
of the verdigris used on the German Von Ems manuscript.
microscopy, which showed particles of RI less
than 1.662 which were not malachite, since the
refractive index of malachite is greater nor of
chrysocolla, whose colour and conchoidal
fracture are characteristic. A micro sample of
the green pigment was then examined using
Debye-Scherrer x-ray powder diffraction, when
a reasonable set of d-spacings could be
measured, but once again there was a problem:
the set of d-spacings did not match any known
mineral defined in the ICDD files or other
pigment types published in the conservation
literature. The problem was only solved since
for the three years prior to this study we had
been working on issues of identification
and synthesis of verdigris pigments, both
from literature sources and historical recipes
that were replicated in the laboratory.
The XRD data showed a close match to
verdigris salt B4, of nominal composition
Cu(CH3COO)2.Cu(OH)2.5H2O which is a
water-soluble verdigris variety, made by the
action of sour vinegar on copper strips at high
humidity. The binder that had been used for
this pigment was found to be egg by GC-MS,
and since the proteins in the egg media
interacts with the verdigris pigment to produce
a copper proteinate green, this could help to
explain some of the difficulties of general
identification, as well as the embrittlement of
the paper support due to the leaching of the
soluble acetate and cupric ions, well-known to
be very damaging to cellulose fibers. A darker
green copper-containing pigment layer which
had been applied over this underlying green
was also examined and its binder determined to
be glue by GC-MS analysis. This material,
which is non-crystalline appears to be made
from a verdigris pigment dissolved in glue to
make a thicker, more opaque dark green paint
to apply over the verdigris in egg tempera
medium used for larger green areas of the
design. The red pigmented area also proved to
be very difficult to identify. ESEM studies
showed that the pigment consists of a very
uniform layer with sharp-edged cleavage.
ESEM-EDS examination revealed mostly
calcium, aluminum, sulfur, oxygen and carbon
as the major elements present. The results of
further studies suggested that an organic red
pigment had been laked onto an aluminum-
based mordant, such as alum, (potassium
aluminum sulphate), which was well-known
from early periods. Initially we used non-
destructive florescence spectrometry
employing remote sensing with a fiber optic
probe to identify the red colourant. This was
not successful and our attempts to match a
micro sample employed for UV/Vis
spectroscopy with known standards also failed.
We eventually succeeded using thin layer
chromatography. The sample was evaluated
together with thirty-four standards, and a match
was found with rhubarb. To further evaluate
the data from the UV/Vis spectrometry, a new
spectrum of rhubarb root in concentrated
sulphuric acid was run and found to be a close
match to our manuscript sample. Rhubarb root
(colouring agent chrysophanic acid), is a red in
either acidic or alkaline conditions. Since
rhubarb was used as a fabric dye it is possible
that this pigment began as a clothlet dye that
was then extracted and laked with alum for use
on the manuscript. The palette of this German
manuscript is quite restricted and uses a limited
range of colours and inexpensive materials;
which may explain the use of simple verdigris
and rhubarb as the principal palette for this
impressive manuscript. Instead of expensive
cinnabar and ground malachite, the restricted
palette makes good use of what were probably
locally made materials. The description of this
analysis may give an idea of how difficult the
identification of a pigment could be. (See Fig. 5).
Further difficulties with green pigments were
found with a late Greco/early Roman Egyptian
cartonnage fragment from the Archaeology
Research Collection of the Department of
Religion at the University of Southern
California. A light green was analyzed by XRF
in situ and showed strong peaks for iron, with
minor peaks for potassium and magnesium,
suggesting that this pigment is a green earth.
XRD also in situ, using a Siemens D5005 and
Gobels mirrors showed that this green was a
mixture of celadonite and glauconite, typical
Pigment analysis: potentialities and problems 235
for green earth pigments , but very rarely
mentioned in Egyptian contexts where the
usual greens encountered are malachite or
Egyptian green frit. The eye of Horus on this
cartonnage is painted in a dark olive-coloured
green which is an unusual shade, and has in
places decayed to a brown discoloured
surfaced. XRF analysis showed the presence of
copper, while polarized light microscopy
revealed isotropic particles, clearly non-
crystalline which could either be a copper
resinate or proteinate. GC-MS analysis
revealed the presence of a glue media and the
presence of copper in the green pigment
suggests that it was either made with a
verdigris pigment dissolved in glue to make a
copper proteinate, or the verdigris has reacted
with the glue media over time to form the
copper proteinate green. Since the reaction
between verdigris and glue is fairly rapid, this
would soon have been noticed and may have
been deliberately employed by the ancient
Egyptians.
CONCLUSIONS
As the above examples show (and the list can
be much longer) the task of identifying
pigments on art objects can be very interesting
and challenging. It is a common mistake to
think that the number of pigments used by man
is limited and therefore their identification is
straightforward. In fact, often what remains on
the decorated surface is not the original
pigment but an alteration, which depends upon
the substances which came in contact with the
work of art accidentally, from the exterior
world, or that were used during the
manufacture of the object for various tasks.
When this happens in general any reference
standard from the pigments class is lost and the
search extends to a much larger set of
chemicals. Unfortunately, all the restrictions in
sampling dictated by the necessity of non
intrusive analysis remain. This is perhaps why,
beside the beauty of dealing with wonderful
masterpieces, pigment analysis still remains a
very fulfilling and challenging endeavor.
REFERENCES
CHIARI G. (1999) — Analisi dei panneggi censori.
In: Michelangelo – La Cappella Sistina –
Rapporto sul Restauro del Giudizio Universale –
Due Volumi. Istituto Geografico De Agostini.
341-351.
CHIARI G., GIUSTETTO R. and RICCHIARDI G. (2003)
— Crystal structure refinements of palygorskite
and Maya Blue from molecular modelling and
powder synchrotron diffraction. Eur. J. Mineral.,
(in press).
CHIARI G. and LANZA R. (1997) — Pictorial
remanent magnetization as an indicator of secular
variation of the Earth’s magnetic field. Phys.
Earth Planet. Int., 101, 79-83.
CHIARI G. and LANZA R. (1999) — Remanent
magnetization of mural paintings from the
Bibliotheca Apostolica Vatican, Rome). J. Appl.
Geophys., 41, 137-143.
DAVY H. (1815) — Some experiments and
observations on the colours used in painting by
the ancients. Philosophical Transactions of the
Royal Society of London 105 97-124 W. Nichol:
London.
FELLER R.L. (1986) — Artists’ Pigments: a
handbook of their history and characteristics Vol.
I. National Gallery of Art: Washington & Oxford
University Press.
FITZHUGH E.W. (1997) — Artists’ Pigments: a
handbook of their history and characteristics Vol.
III. National Gallery of Art: Washington &
Oxford University Press.
GETTENS R.J. and STOUT G.L. (1942) — Painting
materials, a short encyclopedia. [Dover Reprint:
1966]. Dover Publications: New York.
GIUSTETTO R. and CHIARI G. (2003) — Crystal
structure refinement of palygorskite from
neutron powder diffraction. Eur. J. Mineral., (in
press).
HOWARD H. (2003) — Pigments of English Medieval
Wall Painting. Archetype Publications: London.
LAURIE A.P. (1910) — Materials of the Painter’s
Craft in Europe and Egypt London: T. N. Foulis.
LEE L. and QUIRKE S. (2001) — Painting materials.
In: (Ed. P. T. Nicholson and I. Shaw) Ancient
Egyptian Materials and Technology 104-120.
Cambridge University Press.
PABST A. (1959) — Crystal structure of
cuprorivaite. Acta Crystallographica, 12, 733-
739.
PALACHE C., BERMAN H. and FRONDEL C. (Eds)
(1951) — Dana’s System of Mineralogy 7th
Edition. John Wiley & Sons: New York.
RINARDI S., QUARTULLO G., MILANESCHI A.,
PIETROPAOLI R., OCCORSIO S., COSTANTINI SSCALA
G. CHIARI and D. SCOTT
236
G., MINUNNO G. and VIRNO C. (1986) — La
Fabbrica dei colori. Il Bagatto Ed., Roma.
ROY A. (1993) — Artists’ Pigments: a handbook of
their history and characteristics Vol. II. National
Gallery of Art: Washington & Oxford University
Press.
SCHIPPA G. and TORRACA G. (1957) — The pigment
Egyptian blue. Rassegna chimica (Rome),
9, 3-9,
SCOTT D.A., KHANDEKAR N., TURNER N., SCHILLING M.
and KHANJIAN H. (2001) — Technical examination
of a 15th century German illuminated manuscript on
paper: a case study in the identification of materials.
Studies in Conservation 45, 101-112.
SCOTT D.A., DENNIS M., KHANDEKAR N., KEENEY J.,
CARSON D. and DODD L.S. (2002) — Technical
examination of an Egyptian cartonnage of the
Graeco-Roman Period. Studies in Conservation
46, 122-132.
SCOTT D.A. (2002) — Copper and Bronze in Art:
Corrosion, Colorants, Conservation. Getty
Publications: Los Angeles.
Pigment analysis: potentialities and problems 237
