The Grolier Codex: A Non Destructive Study of a Possible Maya Document using Imaging

and Ion Beam Techniques

Jose Luis Ruvalcaba
1
, Sandra Zetina
2
, Helena Calvo del Castillo
1
, Elsa Arroyo
2
, Eumelia
Hernndez
2
, Marie Van der Meeren
3
, and Laura Sotelo
4

1
Instituto de Fisica, Universidad Nacional Autnoma de Mxico, Apdo. postal 20-364, Mexico
DF, 01000, Mexico
2
Instituto de Investigaciones Estticas, Universidad Nacional Autnoma de Mxico, Mexico DF,
Mexico
3
Coordinacion Nacional de Conservacion del Patrimonio Cultural, Instituto Nacional de
Antropologa e Historia, Mexico DF, Mexico
4
Centro de Estudios Mayas, Instituto de Investigaciones Filolgicas, Universidad Nacional
Autnoma de Mxico, Mexico DF, Mexico

ABSTRACT

The Grolier Codex has been a controversial document ever since its late discovery in
1965. Because of its rare iconographical content and its unknown origin, specialists are not keen
to assure its authenticity that would set it amongst the other three known Maya codes in the
world (Dresden, Paris Codex and Madrid Codex).
The document that has been kept in the Museo Nacional de Antropologa in Mexico City,
after its exposure in 1971 at the Grolier Club of New York, has been analyzed by a set of non-
destructive techniques in order to characterize its materials including paper fibers, preparation
layer and color compositions. The methodology included UV imaging, IR reflectography and
optic microscopy examinations as well as Particle Induced X-ray Emission (PIXE) and
Rutherford Backscattering Spectrometry (RBS) using an external beam setup for elemental
analysis. All the measurements were carried out at 3MV Pelletron Accelerator of the Instituto de
Fsica, UNAM. The aim of this work is to verify if the materials in the Grolier Codex match
those found in other pre-Hispanic documents.
From the elemental composition we concluded that the preparation layer shows the
presence of gypsum (CaSO
4
), color red is due to red hematite (Fe
2
O
3
) and black is a carbon-
based ink. These results agree with previous analyses carried out by Scanning Electron
Microscopy (SEM-EDX) on few samples. However, the presence of Maya Blue in the blue
pigment cannot be assured. The examination using UV and IR lights shows homogeneity in the
inks and red color but dark areas that contain higher amounts of K in the preparation layer. This
paper discusses the results obtained for the UV-IR examinations and the elemental analysis. A
comparison with other studies on pre-Hispanic and early colonial codex is presented.

INTRODUCTION

There are sixteen codices from pre-Hispanic Mexico, only three of which come from the
Maya area: the Madrid Codex, the Dresden Codex and the Paris Codex. If the controversial
Grolier Codex is authentic, it would be the fourth Maya pre-Hispanic document known to this
date.
The Grolier Codex discovery was strange; it is the only pre-Hispanic codex found in the
Mater. Res. Soc. Symp. Proc. Vol. 1047 2008 Materials Research Society 1047-Y06-07
20
th
century, with the exception of a few archaeological fragments. The Mexican collector Jos
Senz bought the manuscript in 1964. It was supposed to have been found in a dry cave, in
Chiapas [1]. In 1971 Michael Coe presented the document at the Grolier Club in New York. Two
years after its presentation, a facsimile was published with an iconographic study identifying the
manuscript as a Post-classic Maya Venus calendar with Toltec features [2].
The painted section of the Grolier Codex (Figure 1), also known as Saenz Codex, consists
in a 125 cm long strip of bark paper screen folded in 11 pages: both sides are prepared with a
white layer, but only one is painted. Each page has a maximum of 19 cm height and 12.5 cm
length, but the dimensions of the support vary substantially because of the losses. Three paper
fragments, unpainted and without preparation, are associated with the manuscript, one of them
has the remains of a red line.

Figure 1. The Grolier Codex

Over the white, thick and uniform preparation layer figures of glyphs, gods, priests and
warriors with black and red lines are depicted. A few areas are filled with plain colors: brown,
red and black, only page 11 has a pale blue-green color. A lot of preparatory drawings in brown,
black or red washes that consistently differ in design consistently from the final outline can be
seen.
There are many reasons to question the validity of this codex, apart from its recent
discovery. Some scholars, [2] attributed the Aztec resemblance to a Maya-Toltec style, and noted
that some renowned Maya codices were not completely painted, even though they were
prepared, because it seems that the priests over-painted them. Others think that the combination
of Central Mexican and Maya iconography in the context of a Venus calendar that repeats some
of the Dresden Codex images in an inconsistent reading is probably the result of a falsification
[3, 4].
Though radio-carbon dating of a free-standing sheet of paper placed it at AD (1230 70),
when Maya culture was receiving strong Toltec influences, detractors insist that despite the fact
that the paper is antique, the painting might be the work of an experienced forger that has had
access to the other three codices, particularly to the Dresden Codex. Although some scientific
analyses (SEM-EDX, FTIR) have been practiced on some samples [5], a comprehensive study of
the whole document has never been done. In this work, a general examination of this codex has
been carried out using non destructive techniques such as ultraviolet (UV) imaging, infrared (IR)
reflectography and optic microscopy examinations as well as Particle Induced X-ray Emission
(PIXE) and Rutherford Backscattering Spectrometry (RBS) using an external beam setup for
elemental analysis.

EXPERIMENT

The methodology proposed consists of a global analysis of the object without taking
samples. The initial stage involved imaging techniques for a general examination of the codex,
then characteristic UV and IR images can be registered and related to each part of the document.
A lead sulfide Hamamatsu Vidicon tube camera was employed to perform IR reflectography
with IR LED lighting (940 nm). The entire document was additionally registered with UV
imaging of long wavelength (365nm). In a second stage, a detailed technical examination was
made with a stereomicroscope. Finally for elemental analysis, about 60 spots in representative
pages were analyzed with Particle Induced X-ray Emission (PIXE) and Rutherford
Backscattering Spectrometry (RBS) using an external beam set-up. PIXE and RBS in-air
measurements were carried out at the 3MV Pelletron Particle Accelerator of the Instituto de
Fsica (UNAM) using a 3 MeV proton beam of 1 nA and 1 mm in diameter. The pages were
placed on a PVC stand rotated 60 from the horizontal plane (Figure 2) and protected with PVC
sheets on top that left only the analyzed areas uncovered.


Figure 2. Analysis of the Grolier Codex by our external beam set-up.

Pages analyzed were 1, 2, 3, 5, 6, 8, 9 and 11. Measurements were taken in different
points including fiber, preparation layer and pigments. Two detectors were used in the PIXE set-
up; a Si-Pin for the detection of light elements and a LEGe for the trace elements detection. For
RBS, a particle detector (Ortec R series) was used, placed at 45 to the beam direction. Reference
materials of NIST SRM 2704, SRM 2711 and pure SiO
2
were used for calibration of PIXE and
RBS detectors.

RESULTS AND DISCUSSION

Elemental composition by PIXE and RBS

The preparation layer is made of CaSO
4
(14.9% S, 49% Ca, 32.4% O) with some amount
of strontium (2%). On the other hand, the PIXE spectra of red pigments showed a great amount
of iron (average 23.9 %), together with the elements present in the preparation layer (Ca, S, O,
Sr). Also, some magnesium, aluminum, silicon, titanium and manganese could be found.
Mexican pre-Hispanic red pigments were made from either an organic colorant (carminic acid)
or from inorganic pigments. No mercury or lead has been found; the presence of iron and
elements that are commonly found in soils, indicate the use of natural red ochre, also called
hematite red (Fe
2
O
3
plus soil material such as clay) [Table I]. The black lines showed high
amounts of carbon in the RBS spectra.
The traces of preparatory drawing, a thin red, brown or black wash does not contain an
important iron presence. From Al/Fe ratios, the iron concentration in them resembles more that
of the preparation layer than the one in red pigments [10].

Table I. Elemental concentrations determined by PIXE (%) for the pigments present in the
Grolier Codex. Uncertainties are 10%.

Elements in preparation layer Elements attributed to the pigments and inks
Color/Inks
O Ca S Sr Mg Al Si K Cl Ti Mn Fe Cu Zr
P1 33.3 49.3 13.3 1.14 0.32 0.25 0.59 1.02 - - - 0.67 0.082 -
P2 30.3 51.0 14.7 1.97 - 0.18 0.59 0.65 - - - 0.53 0.084 -
P6 33.5 48.2 15.1 1.33 - 0.27 0.48 0.56 - - - 0.35 0.190 -
Black
P11 31.7 49.5 15.0 1.36 0.12 0.20 0.50 0.98 - - - 0.54 0.064 -
P1 36.3 37.8 9.3 1.10 - 0.65 0.87 1.09 - 0.084 0.085 12.7 0.067 -
P2 30.6 26.9 3.62 1.19 0.17 0.54 1.15 0.67 - - 0.129 35.0 0.048 -
P6 49.3 26.8 9.7 0.86 0.16 0.55 0.97 1.28 0.005 0.067 0.030 10.1 0.120 -
P8 55.5 12.0 2.11 0.65 - 0.43 0.88 0.21 - 0.089 0.089 28.0 0.049 -
Red
P11 27.1 22.8 3.66 0.76 0.05 1.07 2.23 0.69 0.034 0.187 0.061 33.8 0.041 -
P2 36.6 31.9 26.0 2.39 - 0.49 0.98 0.79 - - - 0.76 0.120 -
P6 40.8 27.9 5.24 0.76 - 1.12 1.71 0.40 - 0.146 0.08 21.7 0.095 -
Previous
drawing
P8 27.9 51.7 14.6 2.40 0.07 0.19 0.64 1.42 0.018 - - 0.86 0.243 -
Blue P11 29.0 54.6 8.15 1.38 0.87 0.52 2.70 0.98 0.049 - 0.014 1.52 0.26 0.012

Blue pigments in the Maya culture are usually Maya Blue pigments [6, 7, 8]. Maya
blue consists of colorant (indigo) fixed on palygorskite or other similar clay. Indigo is an organic
colorant extracted from the Indigofera suffruticosa plant and palygorskite is an ino-phyllosilicate
that belongs to the sepiolite family. The Maya Blue appeared around the 8th century and was
used up to 1580 [9].
While Indigo being an organic colorant cannot be identified with PIXE, elements
present in palygorskite are possible to detect with this technique. The PIXE spectra of the blue
shade show in fact a composition that would match that of palygorskite: Mg, Al, Si, K, Cl, Mn,
Fe, Cu, Zn. However, as PIXE does not provide information of the compound but only of the
elements present in it, and there are other materials also found in Mexican artifacts that should be
considered as well, it is not possible to assert the presence of Maya Blue. A comparison with the
analysis of two Maya blue mural painting fragments from Calakmul and Tulum sites did not
show a good correspondence with the expected elemental profile of palygorskite and sepiolite
clays. Though the presence of these clays cannot be certain we are able to conclude that no
modern synthetic pigments have been found in the blue paint.

Infrared reflectography

The red color has a gray (middle absorbance) in IR reflectography, a common behavior of
iron oxide earth pigments, and the black has a strong absorbance, usually seen in carbon black
(Figure 3). It was observed with this technique that all the preparatory drawing lines are not IR
absorbent so they are not seen (even though they are drawn in three different colors: red, black
and brown) which means that the materials used either were organic, or were applied in a very
low concentration and do not contain carbon black. In the IR reflectography imaging the brown
degradation stains in the edges totally disappear.

UV imaging

The white preparation layer presents a slightly lilac tone under UV lighting, commonly
seen in gypsum, which agrees the PIXE-RBS identification. All the painting lines have a strong
purple response to UV (Figure 3).


Figure 3. Grolier Codex, page 7. UV lighting, visible light and IR reflectography images.

The brown stains on the edges had an unexpected behavior under UV: A strong
fluorescence that turns from orange to dark violet. In a detailed examination, these stains do not
permeate the surface, they have a halo effect as if two or three subsequent drops of dye or ink
were carefully applied on top. Besides the borders are very well defined and in the spots where
the preparation layer has been lost, the area beneath is unaffected. The degradation stains appear
like coat of ink (Figure 4). PIXE analysis indicated that in these regions the amount of K
increases and it can be related to the stains.



Figure 4. Halo effect, detail, UV lighting.

Microscopy technical examination of degradation

The fibers in the paper support are jointed in vertically oriented bundles, while the
common direction observed in other Maya codices (Dresden and Madrid) is horizontal; the paper
is also thinner. The pages without painting have a crossed pattern of bundles and are composed
of only one layer of filaments, while in the painted sections, at least two superimposed layers can
be detected, so it is possible that they are not the same kind of paper. There have not yet been
any fiber tests to verify the species used in the manufacture of the paper.
The preservation of the manuscript is heterogeneous, all the edges are lost, eroded and
stained, in contrast, the central parts are very well preserved, and the colors are unexpectedly
bright. No single page is complete; all have lost between 10 to 50% of the paper. Some edges are
too sharp, as if they were deliberately cut. In the microscope the fibers of the paper and the thick
gypsum preparation layer have a clear and straight incision or a sudden disruption (Figure 5). In
some cases the red lines used in the original design are painted over losses of gypsum
preparation and painted the paper fibers.


Figure 5. Sharp cut edges and red painting over losses.

The yellowish brown stains at first sight show degradation due to moisture appearance:
pages 9, 10, 11 are entirely tainted. As it has been noted, under UV the stains look like an
intentional alteration, showing a pattern of applied inks. It should also be considered that if the
stains were indeed the product of moisture, the liquid should have affected more of the gypsum
coat and the pictorial layers, because of the water reactivity of gypsum.
On the darker areas of the codex some regions contain clumps of dust and diverse
materials adhered to the surface. Some sections present cracks and loses in the preparation layer,
not always related to a natural cause of degradation (moisture, folds, etc).

CONCLUSIONS

Results obtained through external PIXE and RBS analyses indicate that no modern
inorganic materials are present. The preparation layer consists of gypsum (CaSO
4
). The presence
of carbon-based black has been established by the use of RBS. Red pigments are made of red
ochre (Fe
2
O
3
) and blue shades show some clay composition.
Although a gypsum composition preparation layer has been found in Colombino Codex,
a Mixtec pre-Hispanic document [10], it is known that in the Madrid and Dresden codices the
preparation composition is calcite (CaCO
3
). This is the only material analysis that has been
performed on those documents, so it is quite difficult to compare the results obtained with the
results in this study. On the other hand, red ochre has not been observed in most of the codex
already examined in our researches, (e.g. Colombino, de la Cruz Badiano, Azoyu [12]) a red
organic colorant was used instead. More pre-Hispanic codices must be studied in order to
establish the patterns of use of materials in codex manufacture and writing.
From what can be achieved with PIXE and RBS techniques, the Grolier Codex contains
materials used in pre-Hispanic times, although the composition of the blue pigment could not be
definitely established. Further analysis needs to be done for the identification of this blue
pigment and also on organic materials.
The most unexpected features in the Grolier Codex are the degradation patterns. The
stains and cut edges of the losses seem like an induced degradation, the UV lighting examination
and the microscopy observation led to question the nature of the deterioration process. The
irruption of the design painting lines over the degradation would not be easily explained if the
document was indeed produced in the 13
th
century and then eroded and degraded after the
moment of its production. The identification of the organic compounds in the codex seems
crucial to establish the origin of the stains and in consequence the possibility to find out if they
are a natural process or a forgery.
Although we are a bit closer to the determination of its authenticity, other factors must be
considered, such as the iconographic content and the historical context. Materials analysis is just
one of the methods that bring to light new questions.

ACKNOWLEDGMENTS

Authors would like to thank technicians K. Lpez and F. Jaimes for their support at the Pelletron
particle accelerator during PIXE-RBS measurements. Financial support was by projects MEC
MAT2002-180, UNAM-DGAPA-PAPIIT IN403302, CYTED Proy.VIII.12, and CONACyT
Mexico grant U49839-R.

REFERENCES

1. J.Alcina Franch, 1992. Cdices Mexicanos, MAPFRE, Madrid, 219-220.
2. M.D. Coe, 1973. The Maya Scribe and His World, The Grolier Club, Nueva York.
3. E. Thomson, 1972. A Commentary on the Dresden Codex: A Maya Hierogplyphic Book,
American Philosophical Society, Philadelphia.
4. C. F. Baudez, 2002. Arqueologa Mexicana, Vol. X, num. 55, 70-79.
5. V. Rodrguez-Lugo, D. Mendoza-Anaya, L. E. Sotelo, Microstructural Study of the Grolier
Codex by Means of LV-SEM, Acta Microscpica, October, (2001), 252-253,
6. M. Snchez del Ro, P. Martinetto, A.Somgyi, C.Reyes-Valerio, E. Dooryhe, N. Peltier, L
Alianelli, B. Moignard, L. Pichon, T. Calligaro, J.C. Dran, Spectrochimica Acta Part B,
(2004) 1619-1625.
7. M. Snchez del Ro, A. Sodo, S.G. Eeckhout, T. Neisius, P. Martinetto, E. Dooryhe, C.
Reyes-Valerio; Nuclear Instruments and Methods B, 238 (2005) 50-54.
8. M. Snchez del Ro, P. Martinetto, C. Reyes-Valerio, E. Dooryhe, M. Surez; Archaeometry
(2006) 115-130.
9. M. Matteini, A. Moles, 2003. La Chimica nel Restauro, I materiali dellarte pittorica, Nardini
Editore, Firenze.
10. R. C. Gonzlez Tirado, Masters Thesis, Monfort University, 1998.
11.C. Lpez Binnqist, PhD Thesis, University of Twente, Netherlands. Twente University
press, Enschede. 2003.
12. J.L Ruvalcaba and C. Gonzlez Tirado 2005. Anlisis in situ de documentos histricos
mediante un sistema porttil de XRF in La Ciencia de Materiales y su Impacto en la
Arqueologa. Vol II, Academia Mexicana de Ciencia de Materiales A.C. D. Mendoza, J.
Arenas y V. Rodrguez coord., Ed. Lagares, Mxico. p. 55-79.