Spectrochimica Acta Part B 65 (2010) 852–858

Contents lists available at ScienceDirect

Spectrochimica Acta Part B

j ourna l homepage: www.e lsev ie r.com/ locate /sab

Analytical note

Inorganic pigment study of the San Pedro Gonzalez Telmo Sibyls using total reflectionX-ray fluorescence☆

Cristina Vázquez a,b,⁎, Graciela Custo b, Néstor Barrio c, José Burucúa c, Susana Boeykens a, Fernando Marte c

a Universidad de Buenos Aires, Facultad de Ingeniería, Paseo Colón 850. C1063ACU, Buenos Aires, Argentinab Comisión Nacional de Energía Atómica, Gerencia Química, Av. Gral Paz 1499, B1650KNA, San Martín, Argentinac CEIRCAB-TAREA, Universidad Nacional de San Martín (UNSAM), Escuela de Humanidades, Campus Miguelete, 25 de Mayo y Francia, B1650KNA, San Martín, Argentina

☆ This paper was presented at the 13th ConferenFluorescence Analysis and Related Methods (TXRF 200915–19 June 2009.⁎ Corresponding author. Comisión Nacional de Energ

Av. Gral Paz 1499, B1650KNA, San Martín, Argentina.E-mail addresses: vazquez@cnea.gov.ar (C. Vázquez)

nbarrio@unsam.edu.ar (N. Barrio), gburucua@unsam.edfmarte@unsam.edu.ar (F. Marte).

0584-8547/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.sab.2010.06.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received 25 October 2009Accepted 15 June 2010Available online 25 June 2010

Keywords:Total reflection X-ray fluorescenceInorganic pigmentSibylsArchaeometry

This article describes the study carried out on a series of oil paintings on canvas from the eighteenth centurythat were restored at Centro de Producción e Investigación en Restauración y Conservación Artística yBibliográfica - Tarea (CEIRCAB-Tarea), Buenos Aires, Argentina: the San Pedro González Telmo Sibyls.Experimental study was undertaken to identify inorganic pigments and the technique used in theirconfection; and, in this way, try to add information about their local origin. Therefore special emphasis wasput to infer technologies used in the manufacturing of these paintings. Elemental analysis was performed bytotal reflection X-ray fluorescence spectrometry (TXRF) and complemented by optical and polarized lightmicroscopy. Microsampling was carefully done over areas of the paintings which were damaged and where asmall additional loss will not be noticed. This investigation has shown that a variety of pigments were used,namely earth pigments (red and yellow ochres), white lead, vermilion, etc., and they were used either pureor in mixtures. This characterization helped conservators in their decisions regarding a better understandingof the deterioration processes. In addition, this research about the material composition allowed the arthistorians and restorers the possibility to obtain information about where, when or by whom The San PedroGonzález Telmo Sibyls may have been painted.

ce on Total reflection X-ray), held in Gothenburg, Sweden,

ía Atómica, Gerencia Química,

, custo@cnea.gov.ar (G. Custo),u.ar (J. Burucúa),

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

During year 2005, a series of paintings belonging to San PedroTelmo church, was studied and restored at CEIRCAB-Tarea; thesepaintings are among the most important colonial painting fromArgentina. Beside the quality itself of the series, its importance residesin the fact that this is one of the few complete series of Sibyls. It iscomposed by twelve paintings: Helespóntica, Cumea, Cumana, Pérsica,Líbica, Tiburtina, Frigia, Délfica, Rodia, Eritrea, Sanbethea and Samia.These are based on engravings by Crispín de Passe [1], where the Sibylsappear prophesizing about different episode of Christ's life. Ten out ofthese paintings are from XVIII century; and the other two, Délfica andTiburtina, are copies made during 1864 when they were requestedprobably to replace the two deteriorated originals. This request alsoincluded the restoration of the others paintings of the series [2].

Regarding the possible origin of this wonderful group, there aretwo positions. Some researchers believe in a workshop from Cuzcowhen some others attribute this series to a Spanish origin [3–5]. It isimportant to highlight that the paintings were extensively studiedfrom the historical and aesthetical points of view but not from thematerial one. So, in this context a characterization study of thepaintings was crucial in trying to elucidate the probable source.Although some late investigations may have finally attributed them aSpanish origin [6], fact that is supported from the organic pigmentsstudies from this series (unpublished data).

The original palette is not vast in comparison with the amount ofcolors used in the areas where impaintings and old interventions arelocated—especially all over the garlands and garments.

It must be said that the San Pedro Telmo's series of Sibyls has agreat potential value for technological and historical investigation. Infact, they were kept together for more or less two hundred years.

The size of the twelve paintings is about 117×92 cm. A very finehandmade flax canvas was laid over cedar stretchers using coarsenails. Both the stretchers and the supports appeared to us verydifferent from those usually found in the colonial paintings. After therestoration treatment the twelve frames (initially covered by blackpaint), proved to be accomplished with the jaspeado technique,a delicate imitation of marble furrowed by fine red and black veins.This kind of decoration falls in the widespread seventeenth century

853C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

Spanish tradition related with polychrome sculpture and altarpiecespractice [7].

The paint technique is very simple: over an absorbent ground,composed essentially by a natural clay, we find no more than two orthree color layers, where traditional pigments such as lead white, earthcolors, bone black, calcium carbonate, vermilion and perhaps indigo,were mixed at different levels and proportions. The radiographicexamination shows a subtle under-modeling treatment witch wascovered by delicate brushstrokes and some impasto layers in the lightzones. The use of organic colors—particularly red lakes—is quite evidentin the flesh areas and also in the grutescos at the bottom sections.Comparing the protected margins with the front surface, a severediscoloration caused by light exposure could be easily discovered [8].

From the twelve paintings samples were taken in order to have afull representation of the palette used by the artists; samples werealso taken from the areas were re-paints were located in order tounderstand the history of the series.

Although analytical techniques have evolved to non-invasive ones,in the specific case of the “Sibyls” samples were required for otherreasons (not related to the main subject of the article). The paintingscame to the workshop for restoration and some little samples weretaken from parts where the conservation treatments were applied.Samples were taken by restores regardless of the research we carriedout. The main reason was to take them for studying the ground layer(because it was considered the possible reason of the deterioration

Fig. 1. Selected sample location of Persica sibyl with the corresponding TXRF spectrum showiand in all subsequent figures are directly reproduced from the Spanish software.

process), something you cannot do with portable XRF equipment. So,since we had samples available and, not least important, we have notaccess to a portable XRF equipment, we decided to use the instrumentwe reported in our article [9].

2. Experimental

2.1. Analytical techniques

Elemental analysis was performed by total reflection X-rayfluorescence spectrometry (TXRF) and complemented by scanningelectron microscopy-energy dispersive spectroscopy (SEM-EDS)optical and polarized light microscopy. TXRF is well suited to theanalysis of artist's pigments where it is mandatory to keep the samplesize as low as possible. It is evident that TXRF can only be used forinorganic pigments. The majority of the old pigments is inorganic andwas made mostly from natural minerals. Their identification by TXRFis based on the fact that most of them can be characterized by theirobvious color and almost one to three constituents regarded as keyelements. The proposed sampling requires only a few small grains ofsample which are deposited on the reflector surface and illuminatedby the totally reflected X-ray beam. This μg-amount of material is avaluable advantage doing practically a non-destructive determination[10,11].

ng Pb L lines attributed to basic lead carbonate. Note that the spectra shown in this figure

854 C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

Microscopy observation was applied in order to provide strati-graphic data with the aim of getting information about the paintingtechniques.

The complete methodology carried out in this study permitted anadequate and fast way for the identification of the San Pedro GonzálezTelmo Sybil's inorganic pigments.

2.2. Sampling

In this research the ex situ procedure was employed selecting theremoving modality for this purpose using a carbon-tungsten (widia)tip [12]. Widia is a recommended material for these samplers becauseof its hardness and the little probability of finding it in the analyzedsamples.

For all paintings, sampling sites were selected after discussionwithart historians, conservators, and a meticulous observation with amagnifier aimed to detect the most appropriated areas of thepaintings. Once the site was selected, a minuscule sample of thepaint layer was scraped off carefully and immediately kept in a cleancontainer to avoid external contamination. For TXRF analysis, thesample was directly adhered onto the reflector. Figs. 1–4 presentsome examples of the sample locations in the paintings studied by

Fig. 2. Selected sample location of Tiburtina sibyl with the corresponding

TXRF and optical and polarized microscopy. Sampling location wasrecorded together with all the information from the paintings,including technical reports on the conservation state.

2.3. Instrumentation

The optical microscopy observation and the production ofcorresponding photographs were obtained by a Leica DM EPPolarization microscope after including the samples in an appropriateresin and polish them with a successive series of sand papers.

Multielemental analysis was carried out using an X-ray fluores-cence system with total reflection geometry. The spectrometerconsists of a Seifert X-ray generator and a fine focus X-ray tube witha Mo anode. The detection and data acquisition system consists of a80 mm2 Si(Li) detector with 166 eV FWHM for 5.9 keV, a 0.008 mmthick Be window, an Ortec 672 fast spectroscopy amplifier and a ADCNucleus PCA2. The total reflection module designed at the Atomin-stitut der Östereichischen Universitäten, fitted with a cut-off-filteredradiation from a fine focus diffraction molybdenum anode X-ray tubewas employed. Excitation conditions were 50 kV and 30 mA in allcases. The acquisition time for each spectrum was 500 s. The minutesample mass cannot be measured exactly so that the concept of

TXRF spectrum showing Zn K lines attributed to Zn white pigment.

Fig. 3. Selected sample location of Eritrea sibyl with the corresponding TXRF spectrum showing Fe and Mn signals ascribed to an umber.

855C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

internal standardization cannot be applied. Nevertheless, selectedconcentration ratio (previous normalised to sum=100) can be usedfor the stoichiometric composition comparison.

3. Results and discussion

3.1. Ground layer

Ground layer was first observed under optical microscopy; itsappearance is heterogeneous and grey in color. The thickness relationwith the color layer is around three to one showing a highly skill onthe performance of the technique in order to achieve the compositionof figurative elements of the painting. The ground layer shows white,grey, and some black particulates, these have a vastly distribution onsize. It also possible to observe cavities distributed all over the layer.All these characteristics of the ground layer may account for some ofthe deterioration consequences observed on the paintings. It shouldbe noted that results from these observations were compared withdata from CEIRCAB-Tarea archives, were almost four hundred colonialpaintings with their corresponded laboratory analysis, are recordedand no equivalent was found. Ground layer on colonial paintings arecomposed mostly by iron oxides. SEM-EDAX analyses of this layershow Si, Ca, and Al as main components with K, Cu, and Fe assecondary constituents; these elements and their relative proportionscould be explained by the use of aluminum silicate as principal

component of the ground layer. Claywas a common ground layer usedin Spain during the XVIII century [13].

On the other hand the ground layer from the XIX centurypaintings, Délfica and Tiburtina, are clearly different with browncolor and white and red particulates. The ground layer was executedby hand over a machine made cotton canvas.

3.2. Pigments

The pictorial palette used on these paintings is relatively limited incontraposition to the vast one used in the restoration of the XIXcentury. These re-paints are mostly placed all over the garlands andgarments compromising large areas. It is important to note thatsampling zones were selected after an exhausted examination withdifferent illumination systems which allowed us to detect interven-tions from original areas.

White pigment used on the confection of this series was basic leadcarbonate, white lead; this compound was used as pigment itself,substrate for other organic dyes, mixedwith other pigment in order toobtain different tonalities, and also in the composition of the fleshes.As example, Fig. 1 shows the TXRF spectrum obtained for Pérsica sibylevidencing Pb L series. It should be noticed that samples from theDélfica and Tiburtina sibyls showed Zn as pigment (Fig. 2). This is animportant fact since Zn white pigment was first commercialized as anartist pigment by Winsor and Newton circa 1850, just a few years

Fig. 4. Selected sample location of Cumana sibyl with the corresponding TXRF spectrum showing Pb as a major component.

856 C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

before the restoration took place in San Telmo, Argentina. AlthoughWinsor and Newton had introduced this pigment, known as ChineseWhite, for water color in 1834 it is until 1850 when they solved thedrying problem caused by the use of linseed oil as medium. Thisinformation is showing the existence of a fluid web of commercial-ization between the Americas and Europe [14].

TXRF spectra from areas where ochre are located showed thepresence of Fe and Mn ascribed to an umber that is a dark earthypigment containing manganese oxides in addition to iron oxides.Fig. 3 shows as an example of this assumption in the case of Eritreasibyl.

In the case of blue pigments we shall differentiate those originalareas from the repaintings. The results from samples coming from thefirst areas showed Pb as a major component, this fact makes us tothink in the use of indigo, this is an organic dye obtained from plantsand it was precipitated over amatrix of leadwhite, andwas confirmedby microchemistry tests. The TXRF spectrum from Cumana sibyl inFig. 4 shows this hypothesis. On the other hand, blue samples fromimpaintings and those from Délfica and Tiburtina sibyls have a highestproportion of Fe evidencing the probable use of Prussian blue as isshowed in Fig. 5 for the Délfica sibyl.

Red pigments analyzed have shown S and Hg as major compo-nents, and these elements are typical constituents of Vermilion (HgS)known in its natural state as cinnabar as is shown in Fig. 6 for Frigiasibyl. However, in some red areas where the painting was hiddenunderneath the frame or in the borders of stretchers, we could

observe a significant intense red coloration just next to an ochre area.This ochre zone, product of the deterioration; that compromises thegarment and frieze of Samia Sybil, was before red. We have sampledboth the red and the deteriorated zone and the results showed,basically, sulfur, calcium, and lead; the compounds we could presumefrom these elements are basic lead carbonate (2PbCO3·Pb(OH)2) andcalcium sulfate (CaSO4·2H2O gypsum hydrate) as in shown in Fig. 7.

This makes us think in the use of an organic lacquer precipitateover the inorganic matrix. The use of an organic lacquer is supportedby the fact of the deterioration process observed on these areas, astrong discoloration; if the red pigment was based on lead, red leadPb3O4, its deterioration would not be the observed.

4. Conclusions

The analytical development of the last decades, in this specific casethe use of TXRF, has provided conservators, restores, and arthistorians with tools that allow them a deeper and more unambig-uous understanding of the cultural heritage. The knowledge built inthis way also gives us the chance to better comprehend thedeterioration processes associated with artworks and, in this way,permit the opportunity to prevent further damage. TXRF has provedto be very suitable and reliable method for the identification ofinorganic pigments used for paintings on canvas. Although TXRF willnot replace other methods for pigments analysis certainly comple-ment them. Inorganic pigment information is essential for restoration

Fig. 6. Selected sample location of Frigia sibyl with the corresponding TXRF spectrum showing typical constituents of Vermilion (HgS) known in its natural state as cinnabar.

Fig. 5. Selected sample location of Delfica sibyl with the corresponding TXRF spectrum showing a highest proportion of Fe evidencing the probable use of Prussian blue.

857C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

Fig. 7. Selected sample location of Samia sibyl with the corresponding TXRF spectrum showing signals ascribed to basic lead carbonate (2PbCO3·Pb(OH)2) and calcium sulfate(CaSO4·2H2O gypsum hydrate) and a detail of original coloration on the area covered by the frame.

858 C. Vázquez et al. / Spectrochimica Acta Part B 65 (2010) 852–858

and conservation procedures as well as to allow a completeknowledge on the use of pigments in different times, areas andschools.

References

[1] N. Barrio, J. Burucúa, F. Marte, A. Rodríguez Romero, Las 12 Sibilas de la Parroquiade San Pedro G. Telmo, Universidad Nacional de Gral. San Martín, San Martín,2005 (In Spanish).

[2] Academia Nacional de Bellas Artes, Patrimonio Artístico Nacional. Inventario deBienes Muebles. Ciudad de Buenos Aires. II Primera parte, ANBA (con el apoyo deThe Getty Foundation), Buenos Aires, 2006 (In Spanish).

[3] H.G. Bauzá, La tradición sibilina y las Sibilas de San Telmo, Fondo Nacional de lasArtes, Buenos Aires, 1999.

[4] H.H. Schenone, “Pintura”, en Historia General del Arte en la Argentina, AcademiaNacional de Bellas Artes, Buenos Aires, 1983 tomo II (In Spanish).

[5] M.J. Sanguinetti, San Telmo. Su pasado histórico, Ediciones República de SanTelmo, Buenos Aires, 1965 (In Spanish).

[6] J.M. Morales Figueroa, José Miguel, Las Sibilas en el arte de la Edad Moderna,Europa Mediterránea y Nueva España, Málaga, Universidad de Málaga, Consejeríade Innovación, Ciencia y Empresa, 2007 (In Spanish).

[7] I.C. Bristol, The Imitation of Natural Materials in Architectural Interiors, PaintedWood: History and Conservation. Symposium Organized by the Wooden ArtifactsGroup of the American Institute for Conservation, The Getty ConservationInstitute, Williamsburg, Virginia, 1994.

[8] D. Sanders, J. Kirby, Light induced color changes in red and yellow lake pigment,National Gallery Technical Bulletin 15 (1994) 79–97.

[9] K. Huhnerfu, A. von Bohlen, D. Kurth, Characterization of pigments and colorsused in ancient Egyptian boat models, Spectrochim. Acta Part B 61 (2006)1224–1228.

[10] A. von Bohlen, Total reflection X ray spectrometry: a versatile tool for ultra-microanalysis of objects of cultural heritage, e-Preservationscience, Morana Edit 1(2004) 23–34.

[11] K. Hühnerfuβ, Characterization of pigments and colors used in ancient Egyptianboat models, Spectrochim. Acta Part B 61 (2006) 1224–1228.

[12] C. Vázquez, A. Albornoz, A. Hajduk, D. Elkin, G. Custo, A. Obrustky, Total reflectionX ray fluorescence and archaeometry: application in the Argentinean culturalheritage, Spectrochim. Acta Part B 63 (2008) 1415–1419.

[13] A. Palomino. Modo de Imprimar o Aparejar los Lienzos y Otras Superficies ParaPintar. El Museo Pictórico y Escala Óptica (1715). Madrid, Aguilar, 1988, 3 tomos;tomo 2, 125–134 (In Spanish).

[14] R.J. Gettens, G.L. Stout, Painting Materials. A short Encyclopedia, New York DoverPublications, 1966.