archaeometry and cultural heritage: the contribution of mineralogy

Cover photoFrench faience test pieces from the Le Bois d'Épense (LesIslettes, NE France) factory, early 19th century. Width of theclay balls 2-3 cm. Photo M. Maggetti (pers. coll.).

01

Seminarios de la Sociedad Española deMineralogía

Volumen 09

Editores:José Miguel Herrero

Marius Vendrell

Archaeometry andCultural Heritage:the Contribution of

MineralogySeminario celebrado en

Bilbao, 27 de Junio de 2012

02

Seminarios de la Sociedad de Españolade Mineralogía

© Sociedad Española de MineralogíaMuseo Geominero del Instituto Geológico yMinero de Españac/ Calle de Ríos Rosas, 2328003 Madrid

http://www.ehu.es/[email protected]

EditoresJosé Miguel HerreroMarius Vendrell

Diseño y maquetaciónSoma Dixital, S.L.

ImpresiónLugami Artes Gráficas, S.L.

Depósito legalCA-602-2004

ISSN1698-5478

Impreso en España - Printed in Spain

2012

Socie

dad

Esp

añ

ola de Mineralogía

03

ForewordThe Spanish Mineralogical Society (SEM) has published eight previous thematic volumesconcerning dif ferent topics of interest for mineralogists. In this new volume we presentthe key notes of six invited lecturers to the International Seminar "Archaeometry andCultural Heritage: the Contribution of Mineralogy" held in Bilbao on June 27, 2012,during the SEM-SEA 2012 congress, a joint meeting of the Spanish Clay Society and SEM.These talks deal with the contributions in scientific disciplines like mineralogy, petrologyand geochemistry to the characterization, dating and provenancing of ar tworks and man-made and natural ar tifacts that constitute par t of our tangible Cultural Heritage. Newadvances in both conceptual and methodological studies are also reviewed. The aim ofthe seminar is to open up to new ideas to researchers in this field and par ticularly, toshow how Mineralogy may provide critical clues in the study of Archaeology, History,Architecture and Fine Ar ts and Restoration. The interest for these studies in our countryis growing, as we can see in the SEM-SEA meeting proceedings, where 24 papers havebeen presented about this topic.

The 2012 seminar is being addressed by scientists of international prestige covering awide range of leading topics: Dr. Domínguez-Bella, who talks about prehistoric lithic toolsand prestige objects, Dr. Prudêncio, who focuses on archaeological ceramics or Dr.Maggetti who make a study of glazed ceramic French faiences. The prehistoric smeltingtechnologies is the topic of the lecture of Dr. Ar tioli; Dr. Hradil introduces us to the non-invasive / non-destructive analysis of pigments in ar tworks; and Dr. Rodríguez-Navarroexplains the use of gypsum and lime in historical buildings.

We par ticularly wish to thank the authors who have accepted our invitation to par ticipatewith generosity. We are also very thankful to the institutions (Education Depar tment ofthe Basque Government, The University of the Basque Country UPV/EHU, Science andTechnology Faculty), companies and entities (Ente Vasco de la Energía, Fisher Scientific,Bilbao Turismo) and to the UPV/EHU Mineralogy and Petrology Depar tment staff. Thesehave provided financial and / or logistics suppor t to the organization of the Seminar andfor the preparation of this volume and its distribution to the SEM-SEA 2012 attendants,and to SEM and SEA members.

José Miguel HerreroMarius Vendrell

Bilbao, Junio 2012

04

Index / Índice

Archaeomineralogy of prehistoric artifacts and gemstones.Arqueomineralogía de artefactos y gemas prehistóricos.Dr. Salvador Domínguez-Bella

Trace element geochemistry and mineralogy for solving problems inprovenance and production technologies of Pre-historic ceramics.La geoquímica de elementos traza y la mineralogía en la resolución dela proveniencia y tecnología de producción de la cerámica prehistórica.Dra. María Isabel Prudêncio

Technology and Provenancing of French faience.La tecnología e investigación de la procedencia de la fayenza francesa.Dr. Marino Maggetti

Archaeometallurgy: the contribution of mineralogy.La arqueometalurgia: la contribución de la mineralogía.Dr. Gilberto Artioli

Microanalysis of pigments in art works.Microanálisis de los pigmentos en obras de arte.Dr. David Hradil & Janka Hradilová

Binders in historical buildings: traditional lime in conservation.Ligantes en edificios históricos: la cal tradicional en la restauración.Dr. Carlos Rodríguez-Navarro

05

29

41

65

79

91

05

Archaeomineralogy of prehistoricartifacts and gemstones/ Salvador Domínguez-Bella

Abstract

The study of the compositional nature, geological and geographical origin of the tools and jewe-llery used by man in prehistoric times has been since the nineteenth century the scientific goalof some researchers in the fields of mineralogy and petrology. This interest in heritage researchstudies is experiencing a significant growth in recent decades, with a great development of inter-disciplinary collaborations, both from the field of archeology as for the conservation, restorationand management of artistic and cultural heritage. The application of physico-chemical techniques,common in studies of mineralogy, petrology and analysis of materials to the resolution of the fas-cinating questions posed from the archaeometry and the fact that we dispose of a growing num-ber of analytical techniques with higher experimental performance make that this line of researchhas grown in the interest of researchers. Here several studies worldwide about some of the mostwidely used mineral substances throughout history and in different geographical areas and thetools and prestige objects manufacture, used by human societies are summarized. Finally, we pre-sent several examples of archaeometric studies carried out on minerals and fossil resins, usedduring the Prehistory of the Iberian Peninsula, western France and North Africa in the elaborationof tools, jewellery and objects of prestige.

Resumen

El estudio de la naturaleza composicional y del origen geológico y geográfico de las herramien-tas y joyas usadas por el hombre desde la Prehistoria, ha constituido desde antiguo, el obje-tivo científico de algunos investigadores de las áreas de mineralogía y petrología. Este inte-rés por los estudios patrimoniales está experimentado un gran crecimiento en las últimasdécadas, con un mayor desarrollo de las colaboraciones interdisciplinares, tanto desde elcampo de la arqueología como de la conservación, restauración y gestión del patrimonio artís-tico y cultural. La aplicación de técnicas físico químicas, habituales en los estudios de mine-ralogía, petrología y análisis físico-químico de materiales, a la resolución de las apasionantesincógnitas planteadas desde la arqueometría y el hecho de disponer de cada vez mayor núme-ro de técnicas analíticas y con mayores prestaciones experimentales, hacen que esta líneade trabajo haya crecido en el interés de los investigadores. Se resumen varios de los estu-dios a nivel mundial sobre algunas de las sustancias minerales más utilizadas a lo largo dela historia y en diferentes áreas geográficas, en la elaboración de herramientas y objetos deprestigio, usados por las sociedades humanas. Finalmente se muestran varios ejemplos deestudios arqueométricos realizados sobre sustancias minerales y resinas fósiles, utilizadasdurante la Prehistoria de la península Ibérica, el oeste de Francia y el Norte de África, en laelaboración de herramientas, joyas y objetos de prestigio.

Key-words: mineralogy, archaeometry, cultural heritage, Iberian Peninsula, raw materials, prehistoricjewels, conservation.

1. Introduction

The archaeomineralogy is itself a mineralogical sub-discipline with a history of not very longtradition, at least in the Iberian Peninsula. This specialization of mineralogy has been develo-ped in parallel to archaeometry studies applied to materials in archaeological and artistic heri-tage. The first reference to this term appears in Mitchell, 1985 and contrary to the disciplines

Dpto. de Ciencias de la Tierra. Universidad de Cádiz. Campus Río San Pedro. 11510 Puerto Real (Cádiz), Spain

06

of Geoarchaeology, has not yet had recogni-tion as such, although in recent years therehas been a large increase in the interest ofmineralogists in the study of archaeologicalmaterials (Turbanti Memmi et al. 2011).

A current definition of archaeomineralogy appe-ars in Rapp (2003 and 2009), as "the study ofminerals and rocks used by ancient societiesacross space and time, as tools, ornaments,building materials and raw materials for metals,ceramics and other processed products".

In recent years archaeomineralogy studieshave been increasing. This may be due toseveral factors:

• Increase in interdisciplinary studies inarcheology, with the participation of manyspecialists from different scientific disci-plines (Price & Burton 2011), includingamong them the mineralogy and petrology.

• Higher number of analytical techniquesavailable for studies of mineralogical andgeochemical characterization of the sam-ples for study and greater analytical preci-sion and capacity of them.

• Increased interest from the field ofarchaeology and restoration of historic andartistic heritage in the potential of thesetechniques on obtaining more and betterarchaeological and historical information.

2. Application of mineralogical studies indifferent archaeological problematic

Mineralogical disciplines provide valuableinformation to design solutions to many ofthe problems, from an archaeological point ofview, for the study of historical and artisticheritage and the restoration of works of artand monuments.

However there are problems inherent in deve-loping partnerships, like this:

• From the mineralogy and geology in gene-ral, it is possible to have a perception ofmaterials, types and natural processesthat originated them within a broader regio-nal geological context, so that the back-

ground in geology and mineralogy shouldbe considered a fundamental and neces-sary part, both for research archaeometricto those relating to heritage conservation.So we need a good understanding of whathave been the physical-chemical systemsand processes involved in the genesis ofeach mineral or rock.

• Besides it is particularly important that thereis reciprocity in the transmission of dataamong mineralogists and scientists in gene-ral, working on heritage and humanisticcounterpart specialists, this is archaeolo-gists, museum and collections curators andcultural heritage managers (Artioli & Angelini2011). This interdisciplinary collaboration,consistent and objective, will provide a com-plete picture of the issues to be addressedin each case and will avoid producing an"innocent archaeometry" (Ramos et al.1998) in which the analytical data are pre-sented only as annexes to the archaeologi-cal work and very often with no connectionbetween the conclusions of both.

Recent contributions as book edition of Rap(2009) on archaeomineralogy or monographicworks as the one published in the EuropeanSociety of Mineralogy (Turbanti Memmi et al.2011) show the growing interest in this line.Many minerals and rocks as prehistoric arti-facts used as gems in antiquity have beenstudied archaeometrically in recent decades.

The variety of compositions, colors and geolo-gical origins that these present is very signi-ficant and there are many archaeological pro-blems that are related to these lithologiesand raised by archaeological research, insome cases for a long time (Damour 1864).This is equally applicable to other mineralsused as pigments since prehistory to the pre-sent (Domínguez-Bella 2010a; Domingo et al.2012).

Ver y dif ferent mineralogical techniqueshave been implemented for its resolution inthe last century, due to the evolution oftechnological and analytical capacities forits identification and characterization, inconstant evolution, especially in recentdecades (Ar tioli 2010).

Archaeomineralogy of prehistoric artifacts and gemstones Salvador Domínguez-Bella

07

3. Analytical techniques at thearchaeometry of prehistoric artifactsand jewellery

Different work strategies can define, such ascombined techniques (ie, the analysis techni-ques applied to the same object at differenttimes) and / or simultaneous (ie, severalanalytical techniques are applied to the sameobject simultaneously), since in many cases,a single technique is not sufficient by itself toresolve complex problems.

The pre-treatment of the samples and theamount of the same required for analysisvaries depending on the analytical techni-que to be employed as well as its nature. Ingeneral for non-invasive techniques, sam-ples can be analyzed without any prepara-tion (Fig. 1 & 2), while in other more aggres-sive as the polarized light microscopy therequired amount of sample is generally gre-ater. In most cases, samples can be prepa-red quite easily, as occurs in many geologi-cal laboratories available in research cen-ters or companies.

Today a great effor t is being focused in thedevelopment and optimization of portable

instruments in order to per form in-situanalysis, in museums, institutions and thefield, including XRF spectrometers, opticaland Raman sensors, X-ray diffractometersand other instruments. Scientific and pro-tection of cultural world heritage institu-tions have been working hard in the lastyears to have laboratories capable to makefast and cheap analysis. Despite this, it isespecially important to per form a study andprevious programming of the technique ortechniques that are most suitable for thestudy of a particular type of material and tohave a knowledge of the protocols for mea-surement, the instrumental configuration,procedures calibration, and ultimately, opti-mization and suitability of each techniqueor set of techniques with respect to thenature of the material to investigate.

We currently have a wide range of analyticaltechniques with more or less invasive inrelation to the amount of sample affected,although many of them, traditionally des-tructive sample preparation, can be appliedon a non-destructive and analytical resultsquite reliable. This would be the case, forexample, the X-ray diffraction (XRD) and X-Ray Fluorescence diffusion wave (WDXRF),applied to archaeological materials of smallsize and with a flat sur face (Figs. 1 & 2).

Among the usual techniques in archaeometrywe can mention: Optical microscopy (MSC),Polarized light microscopy (PLM), X-RayDiffraction (XRD) (powder or direct method),X-Ray Fluorescence diffusion wave (WDXRF),

Seminario SEM 09 Depósito legal: CA-602-2004 / ISSN: 1698-5478

Fig. 1. Direct XRD analysis of archaeological objects. Sillimanite axefrom a Neolithic site of Cadiz province, SW Spain. SCCYT, CádizUniversity

Fig. 2. Sample-holder for direct and non-destructive WDXRF analysis ofa sillimanite adze from the Neolithic of Brittany (France). SCCYT, CádizUniversity

08

Fourier Transform Infrared Spectroscopy(FTIR), Scanning Electron Microscopy withEnergy Dispersive X-ray Analyzer (SEM-EDX),Raman microscopy (RM), Optical EmissionSpectroscopy (OES), Induced CoupledPlasma Mass Spectrometry with laserAblation (ICP-MS-LA), Inductively-coupledPlasma Optical Emission Spectroscopy(Laser Ablation) (ICP-OES-LA). Also other tech-niques reported as: Laser microprobe massanalyzer or Laser induced mass analyzer(LAMMA / LIMA), Atomic absorption spectro-metry (AAS), Flame atomic emission spec-troscopy (FAES), Atomic-emission spectros-copy (AES), Instrumental neutron activationanalysis (INAA), Electron diffraction (ED), ionprobe (secondary ion mass spectrometry)(SIMS), radiometry, emission spectroscopy,spectrophotometry, etc.

Many of these or other qualitative, semiquanti-tative or quantitative techniques, classified asnondestructive analysis techniques are:Optical microscopy with visible light (OM),Cathodoluminescence microscopy (CL),Ultraviolet microscopy (UV), Infrared micros-copy (IR), IR absorption spectrometry (Fouriertransform IR microspectrometry)(FTIR/FTIRM), Laser Raman microprobe (orLaser Raman spectroscopy) (LRM/LRS),Portable X-ray Fluorescence (PRXF), micro X-Ray Diffraction (µXRD) Electron microprobe(EPMA), Transmission electron microscopy(TEM), Synchrotron X-ray fluorescence (SXRF),X-ray absorption fine structure (XAFS), PromptGamma Activation Analyses (PGAA), Elasticrecoil detection analysis ERDA, Extended X-rayabsorption fine structure (EXAFS), Nuclearmagnetic resonance/Proton magnetic reso-nance (NMR/PMR), etc.

An important role is played by isotopic(Carbon-Oxygen-Sulfur) determination techni-ques, especially in provenance of, for exam-ple, metamorphic rocks as marbles or inminerals as cinnabar used as pigment(Minami et al. 2005).

The experimental parameters and the featu-res and availability of each of these techni-ques are very different, and its use alone orin combination, depending on the type ofsubstance and the specimen to analyze.

4. Prehistoric Artifacts

Many examples of minerals and rocks usedsince the beginnings of prehistory for the ela-boration of different kinds of artifacts andtools can be mentioned.

We emphasize the use of silicate mineralsand rocks, and metamorphic rocks in relationto mineralogical and petrological studies ofminerals and rocks used since thePalaeolithic (Domínguez-Bella et al. 2010),such as quartzite, volcanic and plutonic rocksand sedimentary rocks as silicified sandsto-nes, the group of flint and radiolarite (Navazoet al. 2008), as well siliceous minerals asmicrocrystalline quartz varieties (agate, car-nelian, onyx, etc.), single crystals of rockcrystal (Domínguez-Bella and Morata 1995),volcanic glass as obsidian (Tykot 2002) ororganic substances such as fossil resins(Beck et al. 1964 & 1965), etc.

There is a extensive bibliography on the cha-racterization and identification of sourceareas for these materials based on the use ofmany analytical techniques, including someclassical techniques predominant in the geo-logical sciences as OM, XRD, XRF, and morespecific as RM, FTIR, ICP-MS-LA, PIXE, PIGE,INAA, µXRD, PXRF, CL, Carbon-Oxygen-Sulfur-Strontium isotopes, etc. These specific tech-niques vary depending on the substanceunder consideration and the technologicaldevelopment of new analytical techniques inrecent decades (Price & Burton 2011).

Through these studies, we have nowadaysgreat information on mobility of raw materialsin prehistory, such as obsidian in the centralbasin and the western Mediterranean (Tykot2002), in Central Europe (Rosania et al.2008) or in Mesoamerica and South America(Jiménez-Reyes et al. 2001; Rivero-Torres etal. 2008; Tenorio et al. 1997; Seelenfreundet al. 2005); of siliceous materials as flintand radiolarite in Europe by using more orless sophisticated techniques as AAS, XRD,ICP-MS and ICP-AES (Navazo et al. 2008), orwith simple microfacies studies with MO(Della Cassa 2005).

The bulk of the analytical work on polished


09

rocks, has taken place on the most commonlithologies in the Neolithic and Chalcolithic,which have a regional or continental distribu-tion from production centers, such as theaxes of prestige elaborated in jadeite and HPgreen rocks of alpine origin (Petrequin et al.2012; Cassen et al. 2012; D'Amico et a.l2003), or the dolerites of Plusulien, France(Le Roux 1999), the amphibolites in the Westof Iberian Peninsula (Lillios 1997), hornfels inNE Spain (Risch & Martinez 2008; Clop2004), sillimanites-fibrolites of the IberianPeninsula and France (Goer de Herve et al.2002; Domínguez-Bella et al. 2004; Aguayode Hoyos et al. 2006; Pailler 2009) or flintfrom Casa Montero, Madrid (Bustillo et al.2009) or Benzú, Ceuta (Ramos et al. 2008).

Rocks formed under conditions of high pres-sure (HP), deserve a special attention suchas green rocks as jadeite, eclogite, etc.,which have been used in different periodsand geographical areas since prehistorictimes (Ruvalcaba et al. 2008; Cassen et al.2012), often polished and with a high valueas prestige goods.

Mineralogical, petrological and geochemicalstudy techniques have been used for the classi-fication and determination of their source areas(D'Amico et al. 2003; Sheridan et al. 2010).

4.1. Archaeomineralogy of prehistoric artifacts,examples in the Iberian Peninsula, West ofFrance and North Africa

From the end of the XIXth century to thebeginning of XX, geologists worked with theprehistorians in the examination of archaeo-logical materials that composed these lithicelements to attempt the characterization ofthe constituent rocks and their geological andgeographical origin. The works of Quiroga(1885) and of San Miguel de la Cámara(1918) were the pioneer studies in minera-logy and petrography of archaeologicalobjects in Spain.

But these petrographical studies were not con-tinued for many reasons since this informationcould not be correlated with that one given bythe geological outcrops, mainly because thebasic geological works had not been done,

which would have given an adjudication of theanalyzed rock type with a definite geologicaloutcrop. At that time, works of geological carto-graphy were beginning to start, with the crea-tion of the Commission of the Geological Mapof Spain. On the other hand, in those days, agreat number of prehistorians still consideredmore important the object itself than itsarchaeological significance.

To these obstacles we also have to add the factthat the analysis methods were expensive anddestructive. It is not till the 80 and 90 decadesof the XXth century that more or less systema-tic petrographic studies will return, in this typeof materials in Spain. (Domínguez-Bella yMorata 1995; Domínguez-Bella et al. 2004).

In recent years archaeomineralogical studieshave been ongoing on archaeological materialsabiotic prehistory, especially the so-calledstone industry, these materials can separatetwo groups, industry and polishes carved.

The study of the lithic industry has a strongmineralogical and petrographic and geoche-mical component, so that participation ofthese disciplines is very important to deter-mine the mineral nature of the object andtheir petrographic, paleontological and geo-chemical features. These factors are of greatinterest both from the determination of thesource area of these raw-materials and thefeatures and physical properties of the rock.

The determination of the source areas of mine-ral raw materials is one of the main issues ofconcern archaeomineralogical studies, thesestudies allow to obtain great archaeological infor-mation, both on the strategies for obtaining oflithic resources, by prehistoric societies, theirexploitation techniques if they exist (under-ground mining for example) (Camprubí et al.2003; Bustillo et al. 2009), mobility of thesegroups in the territory, the use of the lithic mate-rial and determination of transport over short,medium or long distances, in case there areorganized networks for such distribution, asoccurs with some precious or exotic materials,which can travel long distances (up to thousandsof kilometers), from its geological source areasto where they are deposited (Domínguez-Bella etal. 2002; Cassen et al. 2011; 2012; Querré et


10

al. 2012; Rubalcaba et al. 2008).

The study and determination of mineralogicaland textural properties of many rocks and mine-rals is also of archaeometric interest, as theseproperties can be direct determining factors fora particular use by prehistoric societies of thematerial. These first steps in the "materialsscience" are interesting examples in many of thestudies currently being developed, such as lithicassemblages in the Palaeolithic and Neolithic,indicating high levels of technical knowledge,obtained certainly through the experimentation.By means statistical analysis applied to minera-logical-petrological classifications of tools andtheir relationship with their technological type, isverified in many cases a deliberate selection ofcertain raw materials for a particular use, suchas flint and radiolarite in the Palaeolithic environ-ment of the Strait of Gibraltar as Embarcadero ofPalmones river, Algeciras or Benzú rock-shelter,Ceuta (Domínguez-Bella et al. 2004).

In the studies that we are doing since 1994on these materials, in southern Spain andnorthern Africa it has been shown or inferredallochthonous or autochthonous provenance(Domínguez-Bella and Morata 1995;Domínguez-Bella, Perez and Morata 2000;Ramos and Giles 1996; Domínguez-Bella etal. 2000, 2004 and 2006).

Within the group of analyzed knapped stonematerials, many minerals and rocks, espe-cially siliceous, constitute in percentage themain group in prehistoric lithic industry.These include flint, radiolarites and jasper;we can join with other siliceous sedimentaryand metamorphic rocks as silicified sandsto-nes and quartzites (Hernández et al. 2012).

4.2. Metamorphic rocks in the Atlantic Band ofCádiz, SW Spain

4.2.1. Sillimanite/fibrolite, amphibolites, marbles

The sillimanite-fibrolite (Al2SiO5) is a meta-morphic mineral of high temperature andvariable pressure, which appears in high-grade metamorphic rocks. This is not anexceedingly rare mineral in metamorphic envi-ronments, although it is more limited when itcomes to centimeter or decimeter-sized nodu-

les. This relatively wide distribution in meta-morphic areas makes it difficult to define thesource area of the samples, usually veryhomogeneous in macroscopic appearance.

The size of the nodules of sillimanite seemsto be a determining factor when a geologicaloutcrop is susceptible of constitute a sourcearea for the manufacture of polished stonetools of sufficient size.

Polished tools of these lithologies, are widelyrepresented in the recent prehistory of theAtlantic Band of Cádiz and other peninsularareas and of the Northwest of France, theproblem of their possible origin is still open tonew theories and analysis, even in progress.In the case of SW Spain and Portugal, wethink that sillimanites can be completelyallochthonous materials to this region, giventhe scarcity and small size of sillimanitenodules that appear in the Betic Cordilleras.However some small outcrops have appearedwith this mineral in the province of Malaga(Aguayo de Hoyos et al. 2006) that accordingto these authors could be the source area forthe productions in the area.

We are currently working on the possibledetermination of the source areas of thismineral, relating the geological materialanalysis of the most important sites in theIberian Peninsula, North Africa and France,as are those of the province of Segovia,Madrid and Avila (Sierra de Guadarrama),points of the Serranía de Ronda, SierraMorena, west peninsular zone (Zamora,Salamanca), Brittany and Massif Central(France), Tetouan area (Morocco), amongothers.

Within these lithologies, generally scarce in thesouthern Iberian geological environments, themost important in terms of their proportion inthe archaeological record in the Atlantic regionof Cadiz are amphibolites, some metamorphicrocks such as marbles and quartzites and afew volcanic as tuffs (Domínguez-Bella et al.2004). Some of these exotic lithologies wouldbe possible source area, outside the scope ofthe Betic Cordilleras. In the Iberian Peninsulathere are several places you could find amphi-bolites similar to those described in the


11

archaeological record of the Atlantic Band ofCádiz. The possible source areas closer to thearea would be in volcano-sedimentary sequen-ces, the southwest sector of the peninsula, par-ticularly in the Ossa-Morena zone (provinces ofHuelva, Seville, Badajoz and Alentejo area,south Portugal) and the amphibolites, the sour-ce could also be in the zone north of Huelva,Seville and south of the province of Badajoz(Domínguez-Bella & Morata 1995; Domínguez-Bella et al. 2004) and various points inPortugal (Lillios 1997). A future line of workshould address the petrological, geochemicaland mineralogical different outcrops of theserocks in the western peninsula and the BeticCordillera in order to be able to establish adatabase that allows discrimination of differentexposures and allow the determination of -sour-ce areas for these archaeological materialswidely distributed in recent prehistory, of whichthere is already evidence that were exploited bymeans open pit quarry, somewhere in the wes-tern peninsula.

Regarding the articles made of marble, theyhighlight the bracelets in one piece, made byturning from fragments of this rock type andrelatively common in the Late Prehistory ofthe area. We have analyzed some significantrecords in areas bordering the Atlantic Bandof Cádiz, including: El Jadramil (Arcos de laFrontera) (Domínguez-Bella 2003), Ardales(Málaga) (Domínguez-Bella et al. 2001 and2004) and Villamartin, Cadiz. Its origin hasnot yet been determined analytically.

4.2.2. Serpentinites and peridotites

In this type of ultrabasic rocks we have stu-died some objects recovered in the archaeo-logical record of the area around the Strait ofGibraltar. They are usually colored beadsmade in dark green serpentine, as in the siteof Cantarranas-La Viña (Domínguez-Bella,Perez & Morata 2000) with an allochthonousorigin, due to the absence of these rocks inthe region. Other site is the cave of Benzú(Ceuta), placed in north Africa (Chamorro etal. 2003 & Domínguez-Bella et al. 2006),where the origin of these materials is local,since there is an outcrop of ultrabasic rocksin the city of Ceuta, close to the cave, whichmay be the source area of these materials

currently under study geochemical and mine-ralogical XRD-XRF and MO.

5. Prehistoric Jewellery

The wide variety of minerals and organic mate-rials used as luxury or prestige jewellery (Bard1999), their geological origin or source areas,exploitation and trade routes of transportationand exchange, make these studies from theanalytic sciences of great interest for archaeo-logists and museum and collections curators.It is one of the most exciting in the archaeo-mineralogy (Guillong & Günther 2001;Kosmowska-Ceranowicz 1990 & 2003;Domínguez-Bella 2004; Ruvalcaba et al.2008; Querré et al. 2012; Calligaro et al.1998 & 1999) and one of the archaeologicalmaterials that has attracted more interestfrom the archaeometric point of view, the gem-mological minerals and substances.

There are interesting examples in the humanhistorical record since the Palaeolithic, but aspecial abundance of these objects appear inthe recent prehistory of Europe and Africa(Neolithic-Chalcolithic) or pre-Columbian timesin America. Gemmological materials havebeen widely distributed over large commercialnetworks throughout history, as in the case ofrubies, sapphires and emeralds, etc., highlyappreciated in Roman times and later(Calligaro et al. 1998; 1999; Aurisicchio et al.2005; Giuliani et al. 2000).

Notable examples in the minerals used injewellery may also be some green mineralsfrom prehistory which have been a constantacross cultures and geographies.

They highlight examples such as jade inMesoamerica or Asia (Casadio et al. 2007),turquoise in Mesoamerica and South America(Domínguez-Bella & Sampietro 2005; Hull etal. 2008), in Asia in the west Europe (VázquezVarela 1983; Domínguez-Bella 2004; Querréet al. 2012).

From among the green mineral, as well asother colors, deserves special attention jade,a material highly valued since prehistorictimes in different cultures, chronologies andgeographical areas. During the last decade


12

different methods of analysis of this mineralsubstance using a wide range of analyticalmethods have developed to determine theirgeological sources, early jade workingmethods, the detection of heating processesin jade, burial and surface alterations.

Of particular interest is the use of non-invasi-ve Raman microscopy (RM) in the study ofMesoamerican jadeite pebbles fromGuatemala (Gendron et al. 2002), This appli-cation of RM to jadeite could become a routi-ne approach in archaeometry for identifica-tion and provenance studies, especially asinexpensive portable Raman microprobes aredeveloped with improved spectral resolution(up to 8 cm_1).

Other techniques as X-ray fluorescence spec-troscopy (XRF) and external beam particle-induced X-ray emission (PIXE) have beenapplied to the study of jades. Chinese jadesdating from the Neolithic period (5000 to 1700BCE) to the Han dynasty (206 BCE to 220 CE),composed of nephrite, has been analyzed inorder to determine the minor elemental com-positions of these objects, and their geologicsource in China (Casadio et al. 2007).

Many analytical techniques already cited arebeing used in the study of prehistoric jewe-llery, especially those of greater use in the

field of mineralogy as OM, XRD, WDFRX,PXRF, RM, CL, PIXE, PIGE, etc. Also of inte-rest are some analytical techniques for thestudy of fluid inclusions, especially importantin some of the gems, in relation to their gene-tic and therefore its source and origin area. Asynthesis of specific analytical techniques forthese studies has been described byAnderson & Mayanovic (2003). The sameoccurs with some applications of isotope stu-dies in determining the source areas in gems(Giuliani et al. 2000).

5.1. Archaeomineralogy of gemstones in thePrehistory of Iberian Peninsula, France andNorth of Africa

While the Iberian Peninsula and in general,south-western Europe, are not rich in gemmo-logical materials, there are some exceptionsthat have been of great importance in theexploitation, processing, transportation anddistribution of minerals and rocks used in theproduction of precious or prestige objects,not always gems, along prehistory in this geo-graphical area.

Minerals of the silicate group are also com-mon in prehistoric jewellery, so we can findexamples similar to the Near East steatite(Allen et al. 1975), such as beads of talc, cli-nochlore and micas in many different prehis-


Fig. 3. Multifactorial analysis distribution in geological and archaeological Neolithic variscites from ICP-MS-LA geochemical data (Domínguez-Bellaet al., 2002).

13

toric sites in the Peninsula, where they appe-ar as pendants or beads necklace. There aremany examples in the Iberian Peninsula inrecent prehistory sites as Katillotxu dolmenin Biscay (Quintana 2009) or the site ofLeceia, Portugal (Cardoso 2002). Other sili-cate minerals such as clinochlore have beenanalyzed in recent prehistoric sites, as in thetumulus of the Higueras Valley, Toledo(Domínguez-Bella 2010).

Depending on the type of material, its compo-sition and origin have many different analyti-cal techniques employed in the work ofarchaeometric characterization carried out inrecent 50 years. Thus, for the majority of sili-cate compounds, oxides, sulfides, phospha-tes, carbonates, etc., has been used traditio-nal techniques such as XRD, XRF, OM, FTIRand in recent years, especially in samplesbelonging to the funds of museums andcollections, new non-destructive techniquessuch as PIXE, PIGE, XR microdiffraction, RM,EDX, ESEM, LIBS, PXRF, etc., are being used.

Techniques such as mass spectroscopy,inductively coupled plasma, laser ablation(LA-ICP-MS) (Domínguez-Bella et al. 2003)that can provide a detailed geochemistry ofthe samples and that together with statisticalstudies of factorial analysis of data have allo-wed us to identify possible source areas oforigin of products such as archaeologicalvariscites (Domínguez-Bella et al. 2002) (Fig.3). The same occurs with techniques such asPIXE, PIGE that we are applying to thesephosphate minerals within the projectCALLAIS, CHARISMA program, in develop-ment since 2010 and using the facilities ofAGLAE in the Louvre, Paris (Fig. 4). We arecurrently working on multivariate statisticaltreatment of data obtained over a wide sam-pling of variscitas and turquoise of the IberianPeninsula and France, as well as archaeologi-cal samples from all known geological sitesin Spain, Portugal and France, with the colla-boration of Museums as Huesca, Braga,Bilbao, British Museum, etc. Some of theresults of these and previous analyses havebeen published (Querré et al. 2012).

The X-ray fluorescence (WDXRF) and portable(PXRF) also allows a quite precise analytical,

especially for the major elements in the sam-ple and with a non-destructive character(Domínguez-Bella & Bóveda 2011).

5.1.1. Variscite and turquoise

Another green mineral of interest is the variscite(Al2O3PO4•nH2O), which geological rarity andinterest by the Neolithic man is a clear exampleof precious material in prehistory. In Europethere are not many geological areas containingthis phosphate, usually associated with its iso-morphic variety, the strengite (FePO4•2H2O) andsometimes other phosphates such as turquoise(CuAl6(PO4)4(OH)8•4H2O) (Moro et al 1992a, b,1995b). Underground mining techniques forselective procurement of this mineral are wellknown in Gavá, Barcelona (Fernández-Turiel et al1990; Camprubí et al 2003) and distribution ofthis material in the N and NE of the IberianPeninsula (Munoz-Amilibia 1971; Guerra et al.1995; Fernández Vega & Pérez Cañamares1988; Edo et al. 1998) and the SE of France(Villalba et al. 1998).

Mineralogical and geochemical analysis byICP-MS-LA, XRF, etc., of these preciousobjects in phosphate minerals were used todetermine their source and distribution areas


Fig. 4. Variscite necklace beads from the Neolithic site of SaintMichael, Carnac, France. Non-destructive analysis by PIXE-PIGE.AGLAE, Louvre Museum, Paris. CALLAIS project.

(Guerra et al. 1995; Domínguez-Bella et al2002 & 2003).

The variscite appears among others in nec-klace beads recovered from the burials of thedolmens of Alberite I (V and IV millenniumsBC) (Domínguez-Bella & Morata 1995) andTomillo (Domínguez-Bella et al. 2002), (IV-IIImillennia BC) in the province of Cadiz. It is agreen mineral which has had a great impor-tance in Neolithic-Chalcolithic societies insouthwest Europe.

In this area and many others of the peninsu-la and in some places of France, it is relati-vely frequent the appearance of objects ela-borated in this mineral, usually necklacebeads, which in the dolmen of Alberite repre-sented 7% of over recovered 1000 necklacebeads.

The XRD study of some of these beads reve-aled a monomineral nature, correspondingwith the type variscite Palazuelos. IR spectraof these samples showed absorption bandscharacteristic of the molecular groups (OH)-and (PO4)3-, coinciding also with the variscite.

These beads have a similar mineralogy, andmicroscopic examination shows that it isgenerally monomineral samples, massive,fine-grained, pale green, colorless, non ple-ochroic and low relief, interference colors ofthe second order. The qualitative chemicalanalysis performed by means EDX as expec-ted, showed the presence of P and Al.

The green beads of these chronologies pre-sent different mineral compositions (varisci-te, turquoise, talc, muscovite), with differentgeological origins and many times geographi-cal (Damour 1864; Cardoso 2000; FernandezVega & Cañamares Perez 1988; Huet B.Gonçalves 1980 & 1982; Muñoz-Amilibia1971; Rojo et al. 1995; Querré et al. 2012).

It seems that the green color of theseminerals was the main feature wanted bythese communities for the elaboration ofthese objects, whatever their composition,hardness, etc. The green color should be asign of prestige or some ritual significance,highly desired by the dominant members inthe communities. This idea is shown by thefact that it is only in burials of the domi-

14


Fig. 5. Selection of objects made from jade, fibrolite and variscite from the cist at Saint-Michel (photos: S. Cassen and C. Le Pennec, collectionof the Soc. Polymatique du Morbihan, musée de Vannes).(in: Cassen et al., 2011) and in: www.jungsteinSITE.de

15

nant groups in which the large necklacebeads and pendants of these mineralsappear, as in the megaliths of FrenchBrittany (Cassen et al. 2011 & 2012),Galicia (Dominguez-Bella & Bóveda 2011),Por tugal, Extremadura and Andalusia, orthe pit burials Culture in Catalonia(Domínguez-Bella 2004).

The variscite and turquoise beads are associa-ted in these funerary environments (Fig. 5)with other minerals that also had to have a sig-nificance of prestige or ritual, such as cinna-bar, amber, rock crystals and large flint bladesand axes, chisels and adzes polished rocks,idols and palettes for pigments (Ramos andGiles 1996; Cassen et al. 2011).

The emergence and expansion of variscitenecklaces become important in the Neolithicsouth-western Europe from the sixth millen-nium BC and their use lasts until the RomanEmpire, where it is mined in western Spain,replacing and / or imitating emeralds.

The variscite outcrops in south-westernEurope with special geological features arefound almost exclusively in the IberianPeninsula. Vein deposits of this mineral,associated with Silurian slates with black sili-ceous levels and quartzites, appear in thePalaeozoic of North Portugal (Ervedosa), theprovinces of Zamora (Palazuelo de lasCuevas, Bercianos, El Bostal, etc.) andHuelva (Encinasola), Galicia (Punta Montalvo)


Fig. 6 Large distribution networks over long distances for Iberian variscites have been identified from these sources in the French Brittany (Querréet al., 2012)

16


Fig. 7. Selection of turquoise beads from the Tafí Culture, Tucumán province, Argentina and direct XRD diagrams of the representative litholo-gies of the beads (turquoise, green mica, opal). (Domínguez-Bella & Sampietro, 2005).

17

and the Catalan Coastal Cordillera (Gava,Moncada) in Barcelona. For the studies con-ducted so far, it seems that the only depositsof variscite and turquoise exploited inPrehistoric Europe were in the IberianPeninsula. Mines of prehistoric turquoise andvariscite at Gavá (Barcelona) are well known,although there are clear indications alreadyconfirmed of prehistoric exploitation inEncinasola (Huelva) and Palazuelos de lasCuevas (Zamora), where exploitation is mani-fest in Roman times (Campano Rodriguez &Sanz 1985; Domínguez-Bella 2004).

Variscite mining of Catalonia (Gavá mines)seems that supplied the northern and north-eastern Spain and southern and westernFrance (Guerra et al. 1995; Edo et al. 1998).The deposits of western Spain seem thatvariscites distributed to the southwest andwestern half of the Iberian Peninsula(Domínguez-Bella 2004), although there isevidence of a large distribution networks overlong distances, variscites have been identi-fied from these sources in the French Brittany(Querré et al. 2012) (Fig. 6).

Other variscite deposits of the IberianPeninsula that has certainly been exploitedsince prehistoric times are Encinasola (Huelva)and the area of Palazuelos-The Bostal-Bragança (Moro et al. 1992 a-b; Moro et al.1995 a-b; Domínguez-Bella 2004) located inthe SW and NW of the Iberian Peninsula, whichhave been analyzed in the last 18 years(Dominguez-Bella & Morata 1995; Merielles etal. 1989; Domínguez-Bella 2004 ; Domínguez-Bella et al. 2004; Odriozola et al. 2010; Querréet al. 2012). We have made in recent years, dif-ferent geochemical analysis along with statisti-cal studies by factor analysis of analytical dataobtained by ICP-MS-LA, XRF and SEM-EDX, witha geochemical model of 13 variables, carriedout both on geological samples and archaeolo-gical samples (Dominguez Bella et al. 2002).Recently, further analysis by PIXE-PIGE, XRF,PXRF, XRD are being carried out on these geo-logical sites of the Iberian Peninsula, includingminor ones like Punta Montalvo, Coruña andseveral archaeological sites in Spain, Portugaland France, in an International Project,CALLAIS, in the CHARISMA program of the EU7th Framework. We also propose new isotopic

studies for these variscite materials as a com-plementary line of future research in the deter-mination of source areas.

In the case of turquoise, very different analy-tical techniques have been used for mineralo-gical characterization and source areas iden-tification. Techniques as Arc emission spec-trometry analysis, Electron microprobe,Instrumental neutron activation analysis,Spectrometry, X-ray diffraction and X-ray fluo-rescence are employed throughout the world.As an example, we can cite the study on a setof cylindrical beads and zoomorphic of TafíCulture (300 B.C. - 800 A.D.), Tucuman,Northern Argentina, possibly from the tur-quoise deposits in northern Chile (Fig. 7)(Domínguez-Bella & Sampietro 2005). Thesame method as the variscites has been


Fig. 8. Smoky rock crystal, monocristaline quartz from the Dolmen deAlberite I, Villamartín, Cádiz, SW Spain. (Domínguez-Bella & Morata,1995).

18

employed in the Iberian turquoise, since ithas the same paragenesis in the studieddeposits.

5.1.2. Single cr ystal quar tz

The presence of quar tz cr ystals is relati-vely common in Prehistoric funerary envi-ronments, in the southwest peninsular, inthe province of Cadiz, as in many otherpar ts of the Iberian Peninsula and Europe.Spectacular examples are the large singlecrystal of smoky quar tz that appeared inthe dolmen of Alberite (Fig. 8), a rockcrystal in the dolmen of El Juncal, Cadiz ora quar tz cr ystal of Triassic age, as foundin the silos of La Esparragosa archaeologi-cal site (Chiclana).The first is a pegmatitic quar tz, accompa-nied by small traces of feldspar. The pos-sible source area of that can be located,according to Dominguez-Bella and Morata(1995) in pegmatitic rocks environments,perhaps in the Sistema Central of Spain,with which pegmatite quar tz cr ystals havestrong similarities in their morphology, andlocated several hundreds of kilometersaway where it appeared.

In the case of the bipiramidal quar tzcr ystal appeared in La Esparragosaarchaeological site (Chiclana), it is acr ystal type “Jacinto de Compostela”, greyin color and a size of about 3 cm. Thesecrystals are common in gypsum and claydeposits of the Keuper facies, of Triassicage in the Betic Cordilleras. Outcrops ofthese materials extending along a bandfrom southwest to nor theast across theprovince of Cadiz, according to the predo-minant direction of the Betic Cordillera,several of them exist in this area of theAtlantic band of Cádiz, in Iro River Basin inwhere the site of La Esparragosa is pla-ced. Thus, the origin of this cr ystal wouldprobably be local, being of an exceptionalsize, it was possibly collected.

5.1.3 Amber in the Prehistor y of theIberian Peninsula and Europe

In addition to many of the minerals andgems known and used since ancient

times, we could include in this group ofsubstances other compounds of organicorigin as fossil resins, which also havebeen used in jeweller y or as objects ofprestige, from prehistor y to present.

From the Upper Palaeolithic, pieces ofamber are recorded at sites attributed tothe Magdalenian in locations of CentralEurope, also in enclaves of theHamburgian culture. The Neolithic recordsare also very prominent and many beadsare well known since ancient in theMegalithism of the Iberian Peninsula,especially dolmens in Por tugal. They havegenerally been classified as jet, due to thetotal absence of analysis.

The ambers of Baltic origin are composedmainly by succinite (Beck et al. 1965;Stout et al. 2000) and are used in Europesince at least the Iron Age, with a wides-pread use during the Roman Empire, atboth the European and Mediterraneanareas, where amber routes were well esta-blished, crossing Europe from nor th tosouth. This characterization has been wor-ked since the 70's (Savkevich & Shaks1964; Beck & Vilaça 1995; Kosmowska-Ceranowicz 1990, 1999 & 2003) on geolo-gical and archaeological ambers, espe-cially in European sites.

For organic compounds such as amber,the fundamental techniques that havebeen used since the beginning of theanalytical archaeometric have been FTIR(Beck et al. 1964) and some other par-tially destructive as elemental analysis(Domínguez-Bella et al. 2001).Considering the great experience on thetopic of the schools in Nor thern andEastern Europe, it follows that one of thebest methods of identification and classifi-cation of resinites is infrared spectroscopy(FTIR) (Savkevich & Shaks 1964; Beck etal. 1965).

Angelini & Bellintani (2005) published astudy about five dif ferent localities andtypes of European geological ambers, andalso included Italian geological ambersfrom seven dif ferent deposits. They emplo-


19

yed in the analysis Fourier transform infra-red spectroscopy (FTIR) and dif fuse-reflec-tance infrared Fourier transform (DRIFT).

Other techniques of analysis of geologicaland archaeological ambers are the GC-MSand thermal pyrolysis analysis, X-ray dif-fraction and scanning electron microscopyand mass liquid chromatography withspectrometric detection. In some cases,the use of polarizing transmitted lightmicroscope in the study of natural poly-mers in thin sections is another impor tanttechnique for the identification of themineral inclusions in the sample, the tex-tural relationships between them, theInvestigations of the fluid inclusions andthe fossil animals and plants fragments. Inorder to obtain a fingerprint related to theorigin amber, good results are obtained bydirect mass spectrometric techniques, asthe atmospheric pressure photoionization(APPI) (Tonidandel et al. 2008). Theresults using the X-ray dif fraction vir tuallydoes not allow in general, to obtain infor-mation of interest (Domínguez-Bella et al.2001). However there are cases where itprovides information about the type ofamber inclusions or minerals, such asoccurs with the romanite (Teodor et al.2009).

These new techniques for this cataloguing ofdeposits and varieties of ambers and the cre-ation of analytical databases of the same,will undoubtedly improve the determination ofthe origin of archaeological ambers. This isespecially important to identify the sourceareas of different ambers of local origin,which are different chemically and geneticallyin relation to the Baltic succinite (Teodor etal. 2009; Kosmowska-Ceranowicz 1999).

These studies have been generally associatedwith the search of the geological source sitesin each region. In the Iberian Peninsula, someambers have been characterized in differentarchaeological sites; in southern peninsular(Dominguez-Bella & Morata 1995), Portugal(Vilaça et al. 2002), north of Spain (Alvarez etal. 2005; Peñalver et al. 2007), central Spain(Domínguez-Bella 2010) and Galicia(Domínguez-Bella and Bóveda 2011) (Fig. 9).


Fig. 9. Neolithic necklace of amber and variscite from the ChousaNova dolmen, Galicia, NW Spain (reconstruction). (see Domínguez-Bella & Bóveda, 2011).

Fig. 10. FTIR diagram from three archaeological amber beads of thenecklace from Chousa Nova dolmen, Galicia, NW Spain, and compa-rison with a succinite Baltic amber sample. (Domínguez-Bella &Bóveda, 2011, modified).

20

In the case of amber there are severalrecords in the area of the province of Cadiz,SW Spain, with necklace beads made ofamber and in a Neolithic chronological con-text, such as the Dolmen Alberite I, whoseorigin seems to be far off, since the samplescorresponds to a simetite, species unknownin geological outcrops of the IberianPeninsula, at least until the present(Domínguez-Bella et al. 2001). We arecurrently working on Galicia archaeologicalambers (Domínguez-Bella & Bóveda 2011)(Fig. 10) and on two new Neolithic sites inthis region of southern Spain, having identi-fied new ambers beads or pendants withoutBaltic provenance.

6. Conclusions

The interest from the fields of archeology andheritage management and study of the mine-ralogy and archaeometry is growing in recentdecades in a very strong worldwide, with anincreasing number of scientific publicationsand outreach related to the archaeometricapplication of techniques, where the minera-logy plays an important role in the study andconservation of heritage.

We agree with the idea of mineralogists colle-agues in the need for a real interdisciplinarywork in the archaeomineralogical andarchaeometric determinations which we havebeen doing in our research field in recentdecades.

There are a great number of mineralogicalanalysis techniques currently available, thechoice of which one or ones are best suited toeach case studies posed problems and it is animportant matter to determine a priori. It isusually necessary more than an analytical tech-nique for a complete study of the samples andthe resolution of the issues raised.

Undoubtedly, the physical availability of equip-ment and the economic cost of the use of cer-tain analytical techniques will largely conditiontheir use in solving archaeometric problems,the number of samples to be analyzed, theexistence of scientific equipment in thesurroundings and the available budget. Accessto large facilities has been provided in recent

years, at least at European level, with programssuch as CHARISMA, the 7th FrameworkProgram for transnational access to largeequipment.

The development of analytical databases fordifferent substances such as amber, ancientmetallurgical products, marble, flint, ceramic,etc. undoubtedly facilitate future research inthe field of archaeometry, but require largeinvestments or transnational projects whichmust overcome many administrative difficul-ties, economic and technical to be of homolo-gated use. The different analytical techniquesand technological developments, and experi-mental protocols make it difficult to homogeni-ze the results contained in these bases, atleast for specific techniques. Databases ofimages of OM, CL, in research areas such asmarbles and rocks or ceramics in ancient timesare for example very interesting.

A massive growth has been experienced in theIberian Peninsula in recent years regarding thenumber of researchers involved in archaeo-metry work, although there is still a long wayoff, given the historical gap in relation to otherneighbouring countries. The increasing interna-tional collaborations are softening rapidly thesedifferences.

From our own experience we can note that theinformation provided by mineralogical techni-ques applied to archaeometry is providingvaluable data for historical reconstruction in dif-ferent periods of the Prehistory Peninsular,European and North African. This is applicableto the aspects related to daily use objects instone, ceramic, bone, metal, etc., as to othersrelated to rituals, symbolism and power, whereminerals like pigments and minerals, rocks andsubstances used in jewelry and votive or pres-tige objects would be taken into consideration.

We have not only been able to characterize geo-chemically many compounds mineralogical andminerals, rocks, ceramics and pigments, butwe have also obtained interesting informationabout the source areas of these materials andhence their mobility at short, medium or longdistance, during prehistory. Notable examplescan be used as a pigment cinnabar in manymegalithic tombs of the Neolithic and the


21

Chalcolithic, in which the determinations mine-ralogical, geochemical and especially the S iso-topes seem promising lines of work. Amber,with a presence since the Paleolithic in cavesof the North Spain, has been used as a jewelin the Neolithic-Chalcolithic of Galicia, Castileand Andalusia, with a possibly peninsular ori-gin, while from the Iron Age, is becoming a prio-rity Baltic provenance.

Tools or prestige objects manufactured in polis-hed rocks have also provided valuable archaeo-logical information, after determining their localor distant origin, possibly transported over longand organized exchange networks duringrecent prehistory. So it is with some knappingstone products, especially in siliceous rocks asflint and radiolarite, where objects such aslarge blades of flint, are transported hundredsof kilometers along the peninsula and Europe.Another of the minerals used in the manufactu-re of polished stones, used as tools or objectsof prestige are the sillimanite axes and adzes,with great peninsular diffusion which extendsto the West European and even the BritishIsles, with relatively few areas-source and inwhich analytics we are still immersed.

Materials used as jewels of prestige as varisci-te and turquoise, with geological source areasquite well known and almost exclusive to theIberian Peninsula, show large geographicalmobility, exceeding of a thousand kilometers asoccurs with jadeite and other HP alpine rocks intheir distribution network throughout WesternEurope. In this case, mineralogy, geochemistryof trace elements and multivariate statisticaltreatment of analytical data, are allowing theirpotentially-source geological areas.

These are undoubtedly exciting topics and withgreat future expectations in the scientific deve-lopment of archaeometry and mineralogy,which certainly are of great importance inunderstanding the complex human history andfor some mineralogists, continue to constitutea “fatal attraction”.

7. Acknowledgement

I thank many archaeologists, geologists, che-mists, etc. colleagues and collaboratorswhich over the past two decades have helped

and taught me to increase knowledge andpassion for this branch of mineralogy. My sin-cere thanks and apologies for mistakes to allof them. Financial help from different rese-arch projects contribute (for example,PB96/1520 and HAR2008-0669-C03-02,Ministry of Education and Science of Spainand Callais Project, CHARISMA PROGRAM.7th Framework Program and Research GroupHUM-440 of Junta de Andalucía). Their contri-butions are greatly appreciated. Finally Iwould like to thank the organization of thisseminar, especially J.M. Herrero for their invi-tation and assistance in the editorial workand A. Durante for her unconditional help.

8. References

Aguayo de Hoyos, P., Puga Rodríguez, E., LozanoRodríguez, J.A., García González, J.D. & CarriónMéndez, F. (2006) Caracterización de fuentes dematerias primas para la elaboración de herra-mientas de silimanita, de los yacimientos de laDepresión de Ronda, durante la PrehistoriaReciente. In: G. Martínez Fernández, A. MorgadoRodriguez y J.A. Afonso Marrero (coords).Sociedades prehistóricas, recursos abióticos yterritorio. Ed.: Fundación Ibn Al-Jatib. Servicio dePublicaciones Universidad de Granada. 249-277.

Álvarez Fernández, E.; Peñalver, E. & Delclós, X.(2005) La presencia de ámbar en los yacimien-tos prehistóricos (del Paleolítico Superior a laEdad del Bronce) de la Cornisa Cantábrica y susfuentes de aprovisionamiento. Zephyrus 58,159-182.

Allen, R.O., Luckenbach, A.H. & Holland, C.G.(1975) The application of instrumental neutronactivation analysis to a study of prehistoric stea-tite ar tifacts and source materials.Archaeometry, 17, 69-84.

Anderson, A.J. & Mayanovic, R.A. (2003)Electron, nuclear and X-ray probe microanalysisof fluid inclusions. In: I. Samson, A. Andersonand D. Marshal, (eds.) The Analysis andInterpretation of Fluid Inclusions. MineralogicalAssociation of Canada short course. 32. 323-351.

Angelini, I. & Bellintani, P. (2005) Archaeologicalambers from Northern Italy: an FTIR–DRIFT study


22

of provenance by comparison with the geologicalamber database. Archaeometry, 47, 441-454.

Artioli, G. (2010) Scientific methods and theCultural Heritage. An introduction to the applica-tion of materials science to archaeometry andconservation science. Oxford University Press,Oxford, 536 p.

Artioli, G. & Angelini, I. (2011) Mineralogy andarchaeometry: fatal attraction. Eur. J. Mineral.23, 849-855

Aurisicchio, C., Corami, A., Ehrman, S., Graziani,G. & Stella Nunziante Cesaro, S.N. (2006) Theemerald and gold necklace from Oplontis,Vesuvian Area, Naples, Italy. Journal ofArchaeological Science, 33, 725-734

Bard, K. (1999) Jewelry. Encyclopedia of theArchaeology of Ancient Egypt. Routledge. 462-465. 1280 p.

Beck, C.W. (1986) Spectroscopic investigationsof amber. Appl. Spectrosc. Rev. 22, 57-110.

Beck, C,W., Wilbur, E. & Meret (1964) Infraredspectra and the origin of amber. Nature, 201,256-257.

Beck, C.W., Wilbur, E.; Meret, S.; Kossove, D. &Kermani, K., (1965) The infra-red spectra ofamber and the identification of Baltic amber.Archaeometry, 8. 96-109.

Blasco A., Edo M., Fernández Turiel J.L., GimenoD., Plana F. & Villalba J. (1992) Aplicación de téc-nicas geológicas al estudio de materialesarqueológicos: el ejemplo de las cuentas devariscita catalanas y el complejo minero de CanTintorer (Gavá, Barcelona). Cuaternario yGeomorfología 6, 71-80.

Beck, C.W. & Vilaça, R. (1995) The Provenienceof Portuguese Archaeological Ar tefacts, a casestudy from Moreirinha (Beira Baixa), Trabalhos deAntropologia e Etnologia, 35, 209-219.

Bustillo, M. A., Castañeda, N., Capote, M.,Consuegra, S., Criado, C., Díaz-Del-Río, P.,Orozco, T., Pérez-Jiménez, J.L. & Terradas, X.(2009) Is the macroscopic classification of flintuseful? A petroarchaeological analysis and cha-

racterization of flint raw materials from theIberian Neolithic mine of Casa Montero.Archaeometry, 51, 2. 175-196.

Calligaro, T., Mossmann, A., Poirot, J.P. & QuerréG. (1998) Provenance study of rubies from aParthian statuette by PIXE analysis. NuclearInstruments and Methods in Physics ResearchSection B: Beam Interactions with Materials andAtoms, 136–138, 846-850.

Calligaro, T., Poirot, J.-P. & Querré, G. (1999)Trace element fingerprinting of jewellery rubiesby external beam PIXE. Nuclear Instruments andMethods in Physics Research Section B: BeamInteractions with Materials and Atoms, 150, Iss.1–4, 628-634.

Camprubí, A., Melgarejo, J.C., Proenza, J.A.,Costa, F., Bosch, J., Estrada, A., Borell, F.,Yushkin, N.K. & Andreichev, V.L. (2003) Miningand geological knowledge during the Neolithic: ageological study on the variscite mines at Gavà,Catalonia. Episodes, 26, 4. 295-301.

Cardoso, J.L. (2000) The for tified site of Leceia(Oeiras) in the context of the Chalcolithic inPor tuguese Estremadura. Oxford Journal ofArchaeology, 19, 37-55.

Casadio, F., Douglas, J.G. & Faber, K.T. (2007)Noninvasive methods for the investigation ofancient Chinese jades: an integrated analyticalapproach. Anal Bioanal Chem., 387, 791–801.

Cassen, S., Boujot, C., Domínguez-Bella, S.,Guiavarc’h, M., Le Pennec, C., Prieto Martinez,M.P., Querré, G., Santrot, M.H. & Vigier, E.(2012) Dépôts bretons, tumulus carnacéens etcirculations à longue distance. In: Pétrequin,Cassen, Errera, Klassen, Sheridan and A.M.Pétrequin (dir.). JADE. Grandes haches alpinesdu Néolithique européen, Ve au IVe millénairesav. J.-C. P. Cahiers de la MSHE Ledoux, n° 17.Série: Dynamiques territoriales, n° 6. T. I. Ch.16, 918-995.

Cassen, S., Pétrequin, P., Boujot, C., Domínguez-Bella, S., Guiavarc’h, M. & Querré, G. (2011)Measuring distinction in the megalithic architec-ture of the Carnac region: from sign to material.In: Furholt, M.; Lu_th, F. and Mu_ller, J. (eds.)Megaliths and Identities. Dr. Rudolf Habelt


23

GmbH, Bonn, 225-248.

Chamorro, S., Domínguez-Bella, S., & Pereila, F.(2003) Geología del yacimiento de Benzú.Análisis arqueométrico de la industria lítica y lasmaterias primas minerales. In: J. Ramos, D.Bernal, V. Castañeda (Eds.), El Abrigo y la Cuevade Benzú en la Prehistoria de Ceuta, Consejeríade Educación y Cultura de Ceuta, UNED Ceutaand Universidad de Cádiz, 169–205.

Clop, X. (2004) La gestión de los recursos mine-rales durante la prehistoria reciente en el nores-te de la Península Ibérica. Cypsela 15, 171-185.

D’Amico, C., Starnini, E., Gasparotto, G. &Ghedini, M. (2003) Eclogites, jades and other HP-metaophiolites employed for prehistoric polishedstone implements in Italy and Europe. Specialissue 3: A showcase of the Italian research inapplied petrology. Per. Mineral., 73, 17-42.

Damour M. (1864) Sur le Callais nouveau phos-phate d´alumine hydraté recueilli dans un tombe-au celtique du Morbihan. Comptes Rendus del´Academie des Sciences Paris., LIX, 936-940.

Domingo, i., García-Borja, P. & Roldán, C. (2012)Identification, processing and use of red pig-ments (hematite and cinnabar) in the Valencianearly Neolithic (Spain). Archaeometr y. Doi:10.1111/j.1475-4754.2011.00650.x

Domínguez-Bella, S. (2004) Variscite, a prestigemineral in the Neolithic-Aeneolithic Europe. Rawmaterial sources and possible distribution rou-tes. Slovak Geological Magazine, 1-2. 151-158.

Domínguez-Bella, S. (2010a) Aplicaciones de lastécnicas experimentales y la mineralogía a laarqueometría. Los pigmentos de cinabrio del dol-men de Alberite I, Villamar tín (Cádiz). In:Domínguez-Bella, S. Ramos Muñoz, J., GutiérrezLópez, J.M. and Pérez Rodríguez, M. (Eds.)Minerales y rocas en las Sociedades de laPrehistoria. 235-244.

Domínguez-Bella, S. (2010b) Objetos ornamenta-les en el Calcolítico del centro de la PenínsulaIbérica. Estudio analítico de las cuentas de collarde los enterramientos prehistóricos del Valle delas Higueras (Toledo). In: Domínguez-Bella, S.Ramos Muñoz, J., Gutiérrez López, J.M. and

Pérez Rodríguez, M. (Eds.) Minerales y rocas enlas Sociedades de la Prehistoria. 275-284.

Domínguez-Bella, S., Álvarez Rodríguez, M.A. &Ramos Muñoz, J. (2001) Estudio analítico de lascuentas de collar de ámbar del dolmen deAlberite (Villamartín, Cádiz). Naturaleza química ymineralógica e implicaciones sobre su origen. In:B. Gómez Tubío, B., Respaldiza, M. A. y Pardo, M.L. (eds.): III Congreso Nacional de Arqueometría(Sevilla 1999). 621-630.

Domínguez-Bella, S. & Bóveda, M.J. (2011)Variscita y ámbar en el Neolítico gallego. Análisisarqueométrico del collar del túmulo 1 de ChousaNova, Silleda (Pontevedra, España). Trabajos dePrehistoria, 68, 2, 369-380.

Domínguez-Bella, S, (coord.), Calado, D.,Cardoso, J.L, Clop, X. & Tarriño, A. (2004) Rawmaterials in the Neolithic-Aeneolithic of theIberian Peninsula. Slovak Geol. Mag., 10, 1-2.17-42.

Domínguez-Bella, S., Chamorro, S. Ramos, J. &Bernal, D. (2006) Materias primas minerales ygeología en el entorno del Abrigo y la Cueva deBenzú (Ceuta). In: G. Martínez Fernández, A.Morgado Rodriguez y J.A. Afonso Marrero(coords). Sociedades prehistóricas, recursosabióticos y territorio. Ed.: Fundación Ibn Al-Jatib.Ser vicio de Publicaciones Universidad deGranada. 119–133.

Domínguez-Bella, S. & Morata Céspedes, D.(1995) Aplicación de las técnicas mineralógicasy petrológicas a la arqueometría. Estudio demateriales del dolmen de Alberite (Villamartín,Cádiz). Zephyrus XLVIII. 129-142.

Domínguez Bella S., Morata D., De la Rosa J. &Ramos Muñoz J. (2002) Neolithic trade routes inSW Iberian Peninsula ?. Variscite green beadsfrom some Neolithic sites in the Cadiz province(SW Spain): Raw materials and provenanceareas. Proceedings of 32 InternationalSymposium on Archaeometry. México D.F. 2000.Electronic Book. Universidad Nacional Autónomade México.

Domínguez Bella, S., Nieto, J. M., Nocete, F.,Calvo Ramos Muñoz, J. & Sáez R. (2003) ICP-MSen Arqueometría. Variscitas arqueológicas en el


24

oeste peninsular, Origen de las materias primasgeológicas. Boletín de la Sociedad Española deMineralogía 26-A. 113-114.

Domínguez-Bella, S., Pérez, M. & Morata, D.(2000) Mineralogical and petrological characteri-zation of polished lithic material from La Viña-Cantarranas Neolithic/Aeneolithic site (Puerto deSanta María, Cádiz, Spain). Krystallinikum, 26.57-65. Moravian Museum.

Domínguez-Bella, S.; Ramos Muñoz, J;Castañeda, V.; García, M.E.; Sánchez, M.;Jurado, G. & Moncayo, F. (2004) Lithic productsanalysis, raw materials and technology in the pre-historic settlement of the river Palmones(Algeciras, Cádiz, Spain). Archaeometry Section2. Acts XIVe Congres de L´UISPP. Lieja, Belgium2001. British Archaeological Repor ts. (BAR)Internacional Series 1270, Oxford. 47-55.

Domínguez-Bella, S. Ramos Muñoz, J., GutiérrezLópez, J.M. & Pérez Rodríguez, M. (Eds.) (2010)Minerales y rocas en las Sociedades de laPrehistoria. Ed. Grupo HUM-440, Universidad deCádiz. 396 p. ISBN: 978-84-9389-45-04.

Domínguez-Bella, S., Ramos, J., Maate, A.,Pérez, M., Vijande, E. & Martínez, J. (2012)Matières premières des industries polies néoli-thiques du Détroit de Gibraltar. In : Marchand, G.et Querré, G. (dir.) Roches et sociétés de laPréhistoire, Presses Universitaires de Rennes,195-204. ISBN 978-2-7535-1781-3.

Domínguez-Bella, S. & Sampietro Vattuone, M. M.(2005) Collar Beads from the Tafí Culture (FirstMillennium AD), Tucumán, Argentina: RawMaterials Characterization and Provenance. In:Proceedings of the 33rd International Symposiumon Archaeometry, 2002. Vrije Universiteit, Ámster-dam , Geoarchaeological and BioarchaeologicalStudies, 3. 75-78.

Edo, M., Plana, F., Fernández Turiel, J.L.,Gimeno, D., Blasco, A. and Villalba, M.J. (1998)La caracterización de la variscita del complejominero de Can Tintorel, una experiencia aplicadaal conocimiento del sistema de bienes de presti-gio durante el Neolítico. In: T. Orozco, J.Bernabeu, Xavier Terradas Batlle, (coord.) Losrecursos abióticos en la prehistoria: caracteriza-ción, aprovisionamiento e intercambio. ISBN 84-

370-3450-7, 83-110.

Fernández-Turiel, J.L., Gimeno, D., Plana, F.,Blasco, A., Edo, M. and Villalba, M.J. (1990)Estudio de las mineralizaciones fosfáticas delcomplejo minero de Can Tintorer (Gavà,Barcelona) y comparación con las cuentas proce-dentes de ajuares arqueológicos. Bol. Soc. Esp.Mineral., 13-1, 86-87.

Fernández Vega A. & Pérez Cañamares E. (1988)Los objetos de adorno en “piedras verdes” de laPenínsula Ibérica. Espacio, Tiempo y Forma.Serie I. Prehistoria, I. 239-252.

Forestier F.H., Lasnier B. & L’Helgouach J.(1973) A propos de la “callaïs”. Decouverte d’ungisement de variscite à Pannencé (Loire-Atlantique). Analyse de quelques “perles vertes”néolithiques. Bul. Soc. Prehis. Franc. 70, 173-180.

Gendron, F., Smith, D.C. & Gendron-Badou, A.(2002) Discovery of Jadeite-Jade in Guatemalaconfirmed by Non-Destructive Raman Microscopy.Journal of Archaeological Science, 29, 8, 837-851.

Ghough, L.J. and Mills, J.S. (1972) The composi-tion of Succinite (Baltic amber). Nature, 239,527-528.

Goër de Herve, A. de, Servelle, C. & Surmely, F.(2002) Les haches polies du site de Chastel-sur-Murat (commune de Chastel-sur-Murat, Cantal,France) Comptes Rendus Palevol, Vol. 1, Issue 2,123-128.

Giuliani, G., Chaussidon, M., Schubnel, H-J., Piat,D.H., Rollion-Bard, C., France-Lanord, C., Giard,D., de Narvaez, D. & Rondeau, B. (2000) Oxygenisotopes and emerald trade routes since anti-quity. Science, 28, Vol. 287, no. 5453. 631-633.

Guerra M., Delibes G., Edo M. & Fernandez TurielJ.L. (1995) Adornos de calaita en los ajuares dol-ménicos de la provincia de Burgos. Apuntessobre su composición y procedencia. I Congrésdel Neolitic de la Península Ibérica. Gavá-Bellaterra. Rubricatum, 1. 239-250.

Guillong, M. & Günther, D. (2001) Quasi ‘non-destructive’ laser ablation-inductively coupled


25

plasma-mass spectrometry fingerprinting of sap-phires. Spectrochimica Acta Par t B: AtomicSpectroscopy, 56, 7, 1219-1231.

Hernández, V., Jorge-Villar, S., Capel Ferrón, C.,Medianero, J., Ramos, J., Weniger, G.-C.,Domínguez-Bella, S., Linstaedter, J., Cantalejo,P., Espejo, M. & Durán Valsero, J.J. (2012)Raman spectroscopy analysis of Palaeolithicindustry from Guadalteba terrace river, Campillos(Guadalteba county, Southern of IberianPeninsula). Journal of Raman Spectroscopy, inpress.

Huet B. Gonçalves A. (1980) Elementos de ador-no de cor verde provenientes de estaçoesarqueológicas portuguesas. Importância do seuestudo mineralógico. Trabalhos do InstitutoAntropologia Mendes Correa 40, 5-22.

Huet B. Gonçalves A. & Reis M. (1982) Estudomineralogico de elementos de adorno de corverde provenientes de estaçoes arqueológicasportuguesas. Trabalhos do Instituto AntropologiaMendes Correa 43, 1-18.

Hull, S., Fayek, M., Mathien, F.J., Phillip Shelley,P. & Roler Durand, K. (2008) A new approach todetermining the geological provenance of turquoi-se ar tifacts using hydrogen and copper stableisotopes. Journal of Archaeological Science, 35,Issue 5, 1355-1369.

Jehlicka, J., Jorge Villar, S. E. & Edwards, H. G.M. (2004) Fourier transform Raman spectra ofCzech and Moravian fossil resins from freshwatersediments. J. Raman Spectrosc. 35. 761–767.

Jiménez-Reyes, M., Tenorio, D., Esparza-Lopez, J.R., Cruz-Jiménez, R. L., Mandujano, C. & Elizalde,S. (2001) Neutron Activation Analysis ofObsidians from Quarries of the CentralQuaternary Trans-Mexican Volcanic Axis. Journalof Radioanalytical and Nuclear Chemistry 250,3, 465-471.

Kosmowska-Ceranowicz, B., (1990) The scientificImportance of Museum Collections of Amber andother Fossil Resins. Warszawa Prace MuzeumZiemi, 41. 141-148.

Kosmowska-Ceranowicz, B. (1999) Succinite andsome other fossil resins in Poland and Europe,

Estudios del Museo de Ciencias Naturales deAlava. 14, 73–117.

Kosmowska-Ceranowicz, B. (2003) Amber fromliquid resin to jeweller y. Ed. J. Popiolek,Bucarest.

Le Roux C.T. (1999) L‘outillage de pierre polieen métadolérite du type A: les ateliers dePlussulien (Côtes-d‘Armor): production et diffu-sion au Néolithique dans la France de l‘ouest etau-delà. (Rennes 1999).

Lillios, K.T. (1997) Amphibolite tools of thePortuguese Copper Age (3000–2000 B.C.): Ageoarchaeological approach to Prehistoric econo-mics and symbolism. Geoarchaeology, 12, 2,137-163.

Ludden J., Feng R., Gauthier G., Stix J., Shi L.,Francis D., Machado N. & Wu G. (1995)Applications of LAM-ICP-MS analysis to minerals.Canadian Mineralogist 33, 419-434.

Marchand, G. & Querré, G. (eds.)(2012) Rocheset sociétés de la Préhistoire. Entre massifs cris-tallins et bassins sédimentaires. PressesUniversitaires de Rennes,512 p. ISBN: 978-2-7535-1781-3.

Meirelles C., Ferreira N. & Reis M.L. (1987)Variscite occurrence in Silurian formations fromnorthern Portugal. Comun. Serv. Geol. Portugal73, 21-27.

Minami, T., Imai, A., Bunno, M., Kawakami, K. &Imazu, S. (2005) Using Sulfur Isotopes toDetermine the Sources of Vermillion in AncientBurial Mounds in Japan. Geoarchaeology, 20, 1,79–84.

Miranda, J. (Ed.) (2008) PIXE 2007. X-RaySpectrom. 37. 2. Special Issue, 93-197. DOI:10.1002/xrs.

Mitchell, R. (1985) Archaeomineralogy: an anno-tated bibliography of North and Central America(1840-1981). Southern Illinois University,Edwardsville.

Moro M.C., Cembranos M.L. & Fernández A.(1995)a. Estudio mineralógico de las variscitas yturquesas silúricas de Punta Cor veiro


26

(Pontevedra, España). Geogaceta 18, 176-179.

Moro M.C., Gil M.L., Cembranos M.L., Pérez delVillar L. & Fernández A. (1995b). Las mineraliza-ciones estratiformes de variscita(Aluminofosfatitas) silúricas de los Sinformes deAlcañices (Zamora) y Terena (Huelva) (España).Bol. Geol. Min. 106-3, 233-249.

Moro M.C., Gil M., Cembranos M.L., Pérez delVillar L., Montero J.M., Fernández A & HernándezA. (1992)a. Aluminofosfatitas silúricas de laPenínsula Ibérica: análisis preliminar. Actas IIICongreso Geológico de España. Salamanca.España. T. 3, 212-217.

Moro M.C., Gil M., Montero J.M., CembranosM.L., Pérez del Villar L., Fernández A. &Hernández Sánchez A. (1992)b. Característicasde las mineralizaciones de variscita asociadas alos materiales silúricos del Sinforme de Terena,Encinasola (Provincia de Huelva). Comparacióncon las de la provincia de Zamora. Bol. Soc.Española de Mineralogía, 15, 79-89.

Muñoz-Amilibia A. M. (1971) La “calaita” en elPaís Vasco. Munibe 23, 347-354.

Navazo, M.; Colina, A.; Domínguez-Bella, S. &Benito-Calvo, A. (2008) Raw stone materialsupply for Upper Pleistocene settlements inSierra de Atapuerca (Burgos, Spain). Flint charac-terization using petrographic and geochemicaltechniques. Journal of Archaeological Science,35. 1961-1973.

Odriozola, C.P., Linares-Catela, J.A. & Hurtado-Pérez, V. (2010) Variscite source and sourceanalysis: testing assumptions at Pico Centeno(Encinasola, Spain). Journal of ArchaeologicalScience, 37, 3146-3157.

Pailler, Y. (2009) Neolithic Fibrolite Working inthe West of France. In: B. O'Connor, G. Cooneyand J. Chapman ed., Materialitas: working stone,carving identity (9-10 march 2007, Dublin),Oxbow Books and The Prehistoric Society, Oxford,113-126.

Pailler, Y. (2009) Neolithic Fibrolite Working inthe West of France. In: B. O'Connor, G. Cooneyand J. Chapman ed., Materialitas: working stone,carving identity (9-10 march 2007, Dublin),

Oxbow Books and The Prehistoric Society, Oxford,113-126.

Peñalver, E., Álvarez-Fernández, E., Arias, P.,Delclòs, X. & Ontañón, R. (2007) Local amber ina Palaeolithic context in Cantabrian Spain: thecase of La Garma A. Journal of ArchaeologicalScience 34, 843-849.

Pétrequin, P., Cassen, S., Errera, M., Klassen,Sheridan, A. & Pétrequin A.M. (dir.)(2012) JADE.Grandes haches alpines du Néolithique europé-en, Ve au IVe millénaires av. J.C. P. Cahiers de laMSHE Ledoux, n° 17. Série: Dynamiques territo-riales, n° 6. Tomes I & II. 1524 p.

Price, D.T. & Burton, J.H. (2011) An Introductionto Archaeological Chemistr y, Methods ofAnalysis. Springer 73-126, DOI: 10.1007/978-1-4419-6376-5_4.

Querré, G., Domínguez-Bella, S. & Cassen, S.(2012) La variscite ibérique: exploitation, diffu-sion au cours du néolithique. In: Marchand, G. etQuerré, G. (dir.) Roches et sociétés de laPréhistoire, Presses universitaires de Rennes,307-316. ISBN 978-2-7535-1781-3.

Quintana, J.C. (coord..)(2009) El ConjuntoMonumental de Katillotxu (Mundaka): una miradaal Megalitismo cantábrico. Illunzar, 7, 07-09.

Quiroga y Rodríguez, F. (1875) El microscopio enlitología. Separata de Anales de la Soc. Esp.a deHistoria Natural. t. IV. M. S. de Uhagon. 409-420.

Ramos Muñoz, J., Bernal, D., Domínguez-Bella,S., Calado, D., Ruiz, B., Gil, M.J., Clemente, I.,Durán, J.J., Vijande, E. & Chamorro, S. (2008)The Benzú rockshelter: a Middle Palaeolithic siteon the North African coast. Quaternary ScienceReviews. 27, Is. 23–24, 2210–2218.

Ramos Muñoz, J., Domínguez-Bella, S. & MorataCéspedes, D. (1998) Alternativas no adaptativaspara la integración de técnicas mineralógicas ypetrológicas dentro de una arqueología comoproyecto social. Revista Atlántica-Mediterráneade Prehistoria y Arqueología Social, vol. 1.Servicio de Publicaciones de la Universidad deCádiz. 223-239.

Ramos Muñoz, J., Domínguez-Bella, S., Morata


27

Céspedes, D., Pérez Rodríguez, M., Montañés,M, Castañeda, V., Herrero, N. & García Pantoja,M.E. (1998) Aplicación de las técnicas geoar-queológicas en el estudio del proceso históricoentre el V y III milenios a.n.e. en la comarca dela Janda (Cádiz). Trabajos de Prehistoria, 55 (2),163-176.

Ramos Muñoz F. & Giles F. (eds.) (1996) El dol-men de Alberite (Villamartín). Aportaciones a lasformas económicas y sociales de las comunida-des neolíticas en el NW de Cádiz. ServicioPublicaciones, Universidad de Cádiz. 366 p.

Rapp, G. (2009) Archaeomineralogy. 2nd ed.,Natural Science in Archaeology. Springer-Verlag.Berlin. 348 p. DOI 10.1007/978-3-540-78594-1_1.

Rivero-Torres, S., Calligaro, T., D. Tenorio, D. &Jiménez-Reyes M. (2008) Characterization ofarchaeological obsidians from Lagar tero,Chiapas Mexico by PIXE. Journal ofArchaeological Science 35, 3168–3171.

Risch, R. & Martínez Fernández, F. (2008)Dimensiones naturales y sociales de la produc-ción de hachas de piedra en el noreste de lapenínsula ibérica. Trabajos de Prehistoria, 65, 1,47-71.

Rojo M. A., Delibes de Castro G., Edo M. &Fernández Turiel J. L. (1995) Adornos de calaítaen los ajuares dolménicos de la provincia deBurgos: apuntes sobre su composición y proce-dencia”. Rubricatum, 1, 239-250.

Rosania, C.N., Boulanger, M.T., Biró, K.T.,Ryzhov, S., Trnka, G. & Glascock, M.D. (2008)Revisiting Carpathian obsidian. Antiquity 82(318). Online: www.antiquity.ac.uk.

Rubalcaba-Sil, J.L.; Manzanilla, L.; Melgar, E. &Lozano Santa Cruz, R. (2008) PIXE and ionolumi-nescence for Mesoamerican jadeite characteriza-tion. X-Ray Spectrom. 37. 2. 96-99.

Salvado Canëlhas M.G. (1973) Estudo radiográfi-co de "calaítes" portuguesas. Rev. GuimaraesLXXXIII, 125-144.

San Miguel de la Cámara, M. (1918) Estudiopetrográfico de siete hachas neolíticas pulimen-

tadas, de la colección del Ilmo. Sr D. LuisMariano Vidal. Publicaciones de la Sección deCiencias Naturales, Facultad de Ciencias de laUniversidad de Barcelona, 35-41.

Savkevich, S. & Shaks, I. (1964) Infrared absorp-tion spectra of Baltic amber (Succinite), Zh.Prikladnoi Chimii, 37. 1120-1122.

Seelenfreund, A., Miranda, J., Dinator, M.I. &Morales, J.R. (2005) Caracterización de obsidia-nas del nor te y centro sur de Chile medianteanálisis de fluorescencia de rayos x. Chungara,Revista de Antropología Chilena, 37, Nº 2, 245-253.

Sheridan A., Field D., Pailler Y., Pétrequin P.,Errera M. & Cassen S. (2010) The Breamorejadeitite axehead and other Neolithic axeheadsof alpine rocks from central southern England.Wiltshire Archaeological and Natural Histor yMagazine, 103, 16-34.

Smith, G.D. & Clark, R.J.H. (2004) Ramanmicroscopy in archaeological science. Journal ofArchaeological Science 31, 1137–1160.

Stout, E.C., Beck, C.W. & Kosmowska-Ceranowicz, B., (1995) Gedanite and Gedano-Succinite, In: Amber, Resinite and fossil Resins.ACS Symposium Series 617, 130-148.

Szakmány, G., Kasztovszky, Z., Szilágyi, V.,Starnini, E., Friedel, O. & Biró, K. (2011)Discrimination of prehistoric polished stone toolsfrom Hungary with non-destructive chemicalPrompt Gamma Activation Analyses (PGAA). Eur.J. Mineral. 23. 883-893.

Teodor, E., Litescu, S.C., Neacsu, A., Truica, G. &Albu, C. (2009) Analytical methods to differentia-te Romanian amber and Baltic amber forarchaeological applications. Central EuropeanJournal of Chemistry, 7 (3), 560-568.

Tenorio, D., Jiménez-Reyes, M. & Lagarde, G.(1997) Mexican obsidian samples analysed byPIXE and AAS. International Journal of PIXE. 7, 1-2, 17-24.

Tonidandel, L., Ragazzi, E., Roghi, G. & Traldi, P.(2008) Mass spectrometry in the characteriza-tion of ambers. I. Studies of amber samples of


28

different origin and ages by laser desorption ioni-zation, atmospheric pressure chemical ionizationand atmospheric pressure photoionization massspectrometry. Rapid Commun. Mass Spectrom.22, 630–638.

Trevisani, E., Papazzoni, C.A., Ragazzi, E. &Roghi, G. (2005) Early Eocene amber from the“Pesciara di Bolca” (Lessini Mountains, NorthernItaly). Palaeogeography, Palaeoclimatology,Palaeoecology, 223. 260-274.

Turbanti Memmi, I., Ionescu, C. & Schüssler, U.(eds.) (2011) Mineralogical Sciences andArchaeology. Eur. J. Mineral, 23, 6, 847-1008.ISSN: 0935-1221.

Tykot, R.H. (2002) Chemical Fingerprinting andSource Tracing of Obsidian: The CentralMediterranean Trade in Black Gold. Acc. Chem.Res., 35, 618-627.

Vázquez-Varela J. M. (1975) Cuentas de “calaíta”en la Península Ibérica. Datos para la revisióndel problema. Gallaecia 1, 25-30.

Vázquez-Varela J. M. (1975) Estudio mineralógicode cuentas verdes procedentes de la necrópolismegalítica de Monte da Mora, O Saviñao (Lugo).Boletín do Museo Provincial de Lugo, 175-178.

Vilaça, R.; Beck, C. W. & Shout, E. C. (2002)Provenience analysis of prehistoric amber ar ti-facts in Por tugal. Madrider Mitteilungen.Deutsches Archaölogisches Institut Abteilung 43,61-74.

Villalba, M. J., Edo, M. & Blasco, A. (1998)Explotación, manufactura, distribución y usocomo bien de prestigio de la calaita en elNeolítico. El ejemplo del complejo de CanTintorer. In: G. Delibes de Castro (ed.) Mineralesy Metales en la Prehistoria Reciente, 41-70.Universidad de Valladolid.


29

Trace element geochemistryand mineralogy for solvingproblems in provenance andproduction technologies ofPre-historic ceramics/ Maria Isabel Prudêncio

Abstract

Compositional analysis of archaeological ceramics and raw materials, together with archaeologicaldata is a powerful approach to solve questions dealing with provenance, production technology, andregional and interregional interaction patterns. The chemical and mineralogical characterization of theregional potential raw materials is very important when there is a lack of kilns structures, which is thecase of prehistory. The vessels are made mostly by clays, the composition of the ceramics reflectingthe original raw materials. Conclusions may be inferred concerning the type of raw materials and theestablishment of the resource procurement zones. A special care has to be paid when interpreting thecompositional results of the archaeological ceramics, since the composition of the broken fragmentsthat archaeologists recover in excavations is the combination of several parameters, starting with theinitial composition of the raw materials, its manipulation to make the vessel, the use and the burialenvironment. Compositional studies of Chalcolithic ceramics from archaeological sites in Portugal pla-yed a very important role, particularly the trace elements distribution, to determine provenance. Ingeneral, the results point to local productions, and thus to the circulation of the idea rather than theproducts. These results help to conceptualize the diffusion models and to recognize the importanceof certain typologies and their significance in an archaeological point of view.

Resumen

El análisis composicional de las cerámicas arqueológicas y sus materias primas, junto a los datosarqueológicos, son una potente herramienta para resolver cuestiones relativas a la proveniencia, tec-nología de producción y modelos de interacción regional e interregional. La caracterización químicay mineralógica, de las materias primas potenciales de la región, es muy importante cuando se care-ce de estructuras de hornos, como es el caso de la prehistoria. Los recipientes se han realizadoprincipalmente con arcillas, en los que la composición de la cerámica refleja las mater-ias primasoriginales. Se pueden extraer conclusiones, teniendo en cuenta el tipo de materias primas, estable-ciendo las zonas de aprovisionamiento de recursos. Hay que considerar como un caso especial lainterpretación de los resultados de composición de las cerámicas arqueológicas, ya que los frag-mentos rotos que recuperan los arqueólogos en sus excavaciones suponen una combinación devarios parámetros, desde la composición inicial de las materias primas, su manipulación para reali-zar la vasija, el uso y finalmente el ambiente de enterramiento. Los estudios composicionales, y enespecial los de distribución de elementos traza, de la cerámica del Calcolítico de yacimientos arqueo-lógicos portugueses, han jugado un papel muy importante para determinar la proveniencia. En gene-ral, los resultados apuntan a producciones locales y por tanto, a la circulación de la idea, más quede los productos. Estos hallazgos ayudan a conceptualizar los modelos de difusión y a reconocer laimportancia de ciertas tipologías y su significado en el punto de vista arqueológico.

Instituto Tecnológico e Nuclear, Instituto Superior Técnico, Universidade Técnica de Lisboa, Estrada Nacional 10,Sacavém, Portugal.

Key-words: ceramics, clays, trace elements, mineralogy, production technology, provenance, pre-history

30

1. Introduction

Ancient ceramics are well known for theirenormous potential of information in ar-chaeological studies, being used to datesites, trace trade patterns, understand socialand economic relationships, and so forth. Inarchaeology the term ceramics usually refersto cooking and serving utensils and artobjects manufactured of clay. For potters theterm clay means the basic ingredient in cera-mic manufacture, composed of plastic parti-cles (clay) and natural nonplastic grains.Compositional analyses of pottery togetherwith classical archaeological approacheshave been largely used in solving a broadvariety of questions. In fact, the application ofthe techniques of chemistry, physics, geo-logy, and materials science provide a basisfor understanding many questions aboutmanufacturing techniques, history of techno-logy, production organization, functional rela-tionships between specific resource manu-facturing combinations, and patterns of local,regional, or extraregional distributions ofpotery. Thus, better understand a culture,time period or human interaction. These arethe main re-search issues of ceramic techno-logical (and ecological) analysis. An overviewof these types of studies is well shown in thelitera-ture (Rice 1987; Chappell 1991; Hector1992; Velde & Druc 1999).

The compositional analysis of ceramic mate-rials in archaeological studies is carried outaiming particularly to understand how theceramic might have been used and to deter-mine the location and techniques involved inits manufacture. Furthermore, these objecti-ves can be achieved on broken fragments aswell as intact vessels. Numerous issues des-erve significant attention at the start of amineralogical and chemical characterizationstudy:

• the objectives of the analysis must be defi-ned so they can be translated into chemi-cal and mineralogical terms;

• the requirements of sample selec-tion foreach technique must be understood;

• the understanding of the limitations of the

conclusions one can get from the results;and

• the translation of mineralogical and geo-chemical data to archaeological propo-sals.

The use of resources is very important todeciphering the provenance of a ceramicobject, i.e., the geographic area where theobject was produced, especially when noarchaeological evidence of pottery work-shops are found which is the case of themajority of prehistory. Compositional charac-terization, as applied to pottery, is orientedtoward quantitative and qualitative descrip-tion of its mineral and chemical components.Establishing local and regional patternsneeds a regional clay deposits inventory,followed by studies of minera-logical and che-mical characterization.

Mineralogical analysis of ancient pottery isparticularly useful in the understanding of themanufacture process, and appropriate techni-ques may provide information about temperaddition, oxidation-reduction conditions andtemperatures of firing. Results may also beobtained about post-depositional processes.

Petrographic analysis is an important tool inthe determination of pottery provenance bythe identification of fingerprint minerals,when comparing different clay deposits.

Chemical analysis of sherds, especiallyconcerning trace elements, is a power fulapproach in ceramics characterization andprovenance studies, since geochemicalinterpretations may contribute largely tosolve archaeological problems.

Among the materials and processes involvedin making a pot, the most important is clayand its manipulation to make the vessel.

Therefore it is important to know how the dif-ferent elements make clays, what the lefto-vers are in the claymaking process, and themineralogical and geochemical fingerprintswhich can give one a clue to the geographicorigin of the clay and temper materials foundin the finished product. Hence a discussion

Trace element geochemistry and mineralogy...Prudêncio, M.I.

31

of pottery must include clay and its origin,composition and properties.

In ancient times, sources of clays tended tobe those easily available. Soils or surfacesediments are likely candidates (Fig. 1).Some soils or sediments can be adapted toform a ceramic without further treatment. Ifthe proportions of clay, silt and sand are notadequate, tempering by mixtures of materialsof the same source or from different sourcescan be done. All these procedures duringmanufacture, together with eventual modifi-cations of the ceramic composition duringuse and burial, can lead to compositionalalterations that can difficult the establis-hment of provenance.

Studies of the mineral transformations of claysinduced by firing and their comparison with themineralogical composition of sherds may helpto determine the type of clay used and the tem-peratures range they could have been exposedto during their production. In this comparison itis necessary to take into account that ceramicsare not merely fired clays. Their compositiondepends not only on the chemical and mineralo-gical composition of the clays, but also on thegrainsize distribution, possible addition of diffe-rent types of nonplastic grains as temper, mixtu-re of clays, maximum heating temperature, hea-ting ratio, duration of firing, redox atmosphereand post-depositional alteration (Maggetti,1982; Rice, 1987; Moropoulou et al., 1995;Velde & Druc, 1999; Trindade et al, 2010).

Compositional data are independent ofother common categorizations of potteryused by archaeologists, such as styles,type classes, or shape categories, yet theycan be used for comparing such groupingsor creating new ones. In combination withceramic ecology, ceramic technology can,for example, characterize properties of bothresources and pottery and permit compari-sons between them. The resources shouldbe compared with care to the possibilitiesof a given production site.

Today clay and other ingredients used forceramic production can come from sites thou-sands of kilometres from the production site.However, in the past, the potter was moretied to the local resources for his production,since clays did not travel more than severalkilometres; tens at most if water transportdid not intervene, sands and so forth evenless. In this way the potter would have adap-ted local resources to answer specific needs,functional and aesthetic.

The methodological approaches to studyarchaeological ceramics vary depending onthe nature of the raw materials used and themanufacturing techniques. Research beganby a careful examination of the broken pot-tery fragments that archaeologists recover intheir excavations. Ceramic characterization isdone with three main objectives – classifica-tion, production technology and provenance.The more commonly used techniques to cha-


Fig. 1. Surficial clay materials derived by weathering of schists (A) and sediments (B).

32

racterize archaeological vessels focus on themineralogical and chemical constituents.There are several methods that can beemployed. However it should be noted thatspecial considerations in applying thesemethods to archaeological pottery must betaken into account.

The very first step (and one of the mostimportant) in an archaeological ceramicsstudy concerning a compositional study, isthe selection of sherds from a large collec-tion. In general there is an enormous range ofrecognized variation in visible characteristics,and it is a difficult task to break down acollection into a much smaller number ofsherds that represent the variations in theentire collection. The ideal starting point isthe identification of a problem and the cate-gories of archaeological pottery from whichthe sample will be drawn are generally thosetraditionally used in archaeology - types,decorative classes, and wares, as well as

chronological and regional variations (Fig. 2).Compositional analyses of the archaeologicalceramics aim to know texture features, andthe mineralogical and chemical composition.Several techniques can be used such as opti-cal microscopy, X-ray dif fraction (XRD),Mössbauer spectroscopy, instrumental neu-tron activation analysis (INAA), X-ray fluores-cence, among others. Chemical analysis ofceramics and of raw materials, determining

the larger number of chemical elements aspossible, is particularly useful in provenancestudies. In this way, large data sets are gene-rated, which need to be processed by compu-ter with statistical programs. One importantthing to not forget is that statistics analysisis only a tool to help us with very largeamount of data matrixes (variables and sam-ples) and remember that the results obtainedare an approximation of the reality. The attri-bution of one pot to one particular site orcomposition group is based on statistical pro-babilities. The results obtained by statisticalanalysis must be checked taking into accountgeochemical considerations.

In some cases, an important step prior to anystatistical analysis is normalization of the che-mical elements contents. To compensate forgrainsize and mineralogy effects on trace ele-ment concentrations, thus diminishing errone-ous interpretation of ceramic provenance, acommon approach used in determining regio-nal geochemical baselines is to normalize geo-chemical data using one element as grainsizeand mineralogical proxy, that is to express theratio of the concentration of a given element tothat of the normalizing factor.

Normalization using a conservative elementhas been commonly used in environmentalstudies, especially in verifying whether thevariation of elements in sediments is indeedthe result of anthropogenic and/or naturalactivities. Hence normalization in archaeolo-gical ceramics studies is defined as a proce-dure to compensate for the influence of natu-ral (geological and ceramic burial time) andanthropogenic (technology of production) pro-cesses on the measured variability of theconcentration of elements, emphasizing theimportance of taking into account geochemi-cal behaviour of the element chosen for nor-malization, and not purely based in statisticalconsiderations (Prudêncio et al., 2006; Dias& Prudêncio, 2008).

Normalization of elemental concentration toan immobile, conservative element providesdeeper insights about other elements distri-bution than the consideration of the absoluteconcentration itself alone. Therefore, thebest option appears to be the use of a con-


Fig. 2. Pre-historic sherds with different pastes and decorations (pho-tographs by António Valera).

33

servative element, and among these, Scappears to be the more appropriate to norma-lize chemical data then other elements; eventheir variability is higher, since the raw mate-rials are related with a specific geologicalcontext which has to be considered. Sc isstructurally combined in clay minerals andmicas, being a good tracer of phyllosilicates,particularly in sediments containing Al-silica-tes in all size fractions.

Several multivariate statistical methods canbe employed by using statistic programs (e.g.STATISTICA, Statsoft, 2011). The resultsobtained must be integrated and interpretedtaking into account the significance of corre-lations between the chemical elements(variables) found in ceramics and their mea-ning in terms of raw materials composition,and possible alterations due to the produc-tion technologies (sieving, sedimentation,add of temper, mixing clays, etc.), eventualmodifications during the use of pots, andpostdepositional processes.

A number of different analytical techniqueshave been applied with varying degrees ofsuccess to characterize archaeological mate-rials, but all of them need to have multiele-ment capability and sufficient sensitivity todetect traces of elements in the variousmatrices. Among them the analytical methodwith one of the longest and most successfulhistories of application for provenance rese-arch has been instrumental neutron activa-tion analysis (INAA). The real success story ofthis technique, however, comes from theinvestigation of ceramics dating from throug-hout archaeological times. INAA is a sensiti-ve technique useful for quantitative multiele-ment analysis of major, minor, and trace ele-ments, which concentration can have diffe-rent meanings in minerals or in rock descrip-tions, as well as in ceramics (Prudêncio,2009; Glascock & Neff, 2003).

Among trace elements, rare earth elementsdistributions can be particularly useful in dis-tinguishing clayey materials resources. Thesubtle variations in the properties of REEmake them sensitive to mineral/melt equili-bria, as well as to weathering conditions afterthe breakdown of primary minerals and the

formation of new mineral phases, sedimen-tary sorting, and diagenesis. These elementshave very similar chemical and physical pro-perties, which is the result of the nature oftheir electronic configurations. The dominantoxidation state is the +3 state, and there is asmall but steady decrease in ionic radius withincreasing atomic number (for a given coordi-nation number). The REE, therefore, tend tooccur in nature as a group. They are lithophi-le, in that they concentrate predominantly inthe silicate rather than the metal or sulphidephases when they coexist. The differencesexisting among the REE lead to differences intheir relative behaviour in response to thechemical environment, making this group par-ticularly useful in geochemistry since theycan be a pointer of the genesis processes ofthe rocks and minerals and subsequent alte-rations (Prudêncio et al. 1993, 1995;Gouveia et al. 1993; Burt, 1989).

The application of the INAA method in sup-port of provenance research has been largelyused over the past few decades in theInstituto Tecnológico e Nuclear (IST/ITN). Thefirst analyses of archaeological ceramicswere conducted in the late 1970’s by usingthe Portuguese research reactor (RPI), andproceeded generating a large database forceramics and raw materials from different


Fig. 3. Map of Portugal, with the location of the archaeological sites -Fraga da Pena, Monte do Tosco and Porto Torrão.

34

chronologies and archaeological sites (Diaset al., 2000; 2001; 2002; 2003a, 2003b;2005; 2007; 2010; Prudêncio et al., 1988;2003; 2006; 2009). In the late 1990’s, aresearch group especially devoted toarchaeometry (measurements techniques inarchaeology), particularly compositional anddating studies (TL-OSL) of ceramics and rawmaterials, was formed in ITN – “CulturalHeritage and Sciences”, nowadays named“Applied Geochemistry & Luminescence onCultural Heritage” (GeoLuC). INAA has alsobeen used for precise and accurate determi-nation of the contents of natural radioactiveelements such as potassium, rubidium, tho-rium and uranium aiming to evaluate thedose rate, which is fundamental for TL-OSLdating of cultural materials and archaeologi-cal contexts, as well as geological contexts(Prudêncio et al, 2007; Burbidge et al.,2009; 2010).

The making of pots in pre-history

Potter’s clay is a material of “high” clay con-tent, with a high or certain plasticity whenwetted and worked. In addition to clay mine-rals, nonplastic grains may occur in differentproportions. Also non-plastic grains may beadded by the potter to the clay material toimprove the working to make the pot that isreducing the plasticity of the waterloving clayparticles. The material added is in generaltermed as temper. Inclusion is also a termused by archaeologists to refer any nonplas-tic material in the paste (mineral grains, rockfragments, grog or crushed shells).

The manufacturing process include severalsteps (forming, drying and firing), which lead tophysical and chemical reactions influencingceramic production: plasticity and the role oftemper, shrinkage during firing, and nonplasticexpansion during firing. The “visible” reactionsare mainly change in colour, and retraction andexpansion processes related to the quality ofthe paste, the type and quantity of the rawmaterials, the particle size, the amount ofwater present, the preparation of the paste andthe forming techniques used.

The function or quality of a pot is influenced bythe raw materials used. A coarse paste, fairly

porous, with poorly oriented particles, and low-firing, is more defected resistant. It is a multi-purpose paste often found in prehistoric wares.However in some cases a careful selection ofthe clays or a preparation of the materials toobtain a particular paste for the production of aspecific kind of ware was also done in pre-his-tory (Velde & Druc, 1999).

2. Case studies. Geochemical and mine-ralogical characterization of Chalcolithicceramics from Portuguese sites

Three relevant case studies of ceramics fromChalcolithic to early Bronze Age archaeologi-cal sites from Southern and Central Portugal(more recent up to north) are presented (Fig.3). They are specially focused on a chemicaland mineralogical characterization of bothpottery and potential raw materials from theregion, aiming the establishment of prove-nance and production technology, thus contri-buting to the discussion of the circulation/diffusion of this kind of pottery (Cabral etal., 1988; Dias et al., 2000; 2002; 2003a;2005; Dias et al., in press; Valera, 2006).

Chalcolithic ceramics from the three sites inclu-de the typical typologies of this chronology, par-ticularly the combed incisions and the BellBeakers. These typical pots were decoratedwith patterns stamped on the surface in para-llel bands, sometimes filled with white paste,but there are often traces of lime and paintdecoration as well. Cord impressions are com-mon, some are all over cord or all over orna-mented, others are incised or stamped withvarious geometric motives, often arranged inzones (Valera, 2006).

Classical archaeological studies regardingthe prehistoric ceramics together with labora-tory research has been providing significantresults concerning questions related with pro-duction technology, raw materials exploitationstrategies, provenance and mechanisms ofcirculation.

2.1. Porto Torrão, Ferreira do Alentejo

Bell Beakers from the Chalcolithic settlementof Ferreira do Alentejo (Beja, Portugal), PortoTorrão archaeological site, as well as pre-


35

Beaker ceramics, modern ceramics, andpotential raw materials from site surroundswere studied in order to establish whetherthe Bell Beakers found at the site were impor-ted or produced locally (Fig. 4).

The geological context of the site comprisesgabbrodiorites, quartz, porphyries, Silurian andDevonian schists, greywackes, Palaeogene-Miocene and Pliocene sediments. Nowadayslocal potters prefer to use raw materials deri-ved by weathering from gabbros and diorites,the so called “Barros de Beja”.

The materials studied included:

• sherds from two different archaeologicalcontexts - the pre-Beaker and the Beakerlevels;

• clay samples from available geological back-ground in the area, including soils derived byweathering from gabbros and diorites, andfrom schists and greywackes, as well asPalaeogene-Miocene and Pliocene sedi-ments; and

• samples of modern ceramics made by localpotters.

Among the chemical elements studied REEpatterns, particularly the europium anomaly,showed that weathered gabbros and dioriteswere used as raw materials to produce BellBeakers and also pre-Beaker ceramics. Thegeographic area of origin of this type of geolo-gical materials is shown in Fig. 5. The sametype of surficial soils is still used to produceceramics by the local potters.

Thus the results obtained for ceramics fromPorto Torrão Chalcolithic site and for availableclay materials, give an important contributionto the Bell Beakers circulation problematic, asthey clearly indicate that they were not introdu-ced in the area as a result of some trade or“prestige good” distribution network, but wereproduced locally.

2.2. Monte do Tosco, Alentejo

Monte do Tosco is an enclosure Chalcolithicsite, reoccupied during the early Bronze Age,where a late Bell Beaker context has beenexcavated (Fig. 6). The sherds studied inclu-de Chalcolithic ceramics, Bronze Age cera-mics from the Bell Beakers context, commonware, and crucibles. Raw materials sampleswere collected from regional/local clays (wea-thered schists, greywackes, diorites and gab-bros and Tertiary sediments).


Fig. 4. Sherds from Porto Torrão archaeological site, Ferreira doAlentejo (photographs by António Valera).

Fig. 5. Geological map of southern Portugal with the location of thePorto Torrão archaeological site and the area of the gabbro-dioritecomplex (red dashs).

36

Ceramics from Monte do Tosco were dividedin three chemical groups (Fig. 7):

• Group one (1) embraces 80% of Chalcolithicceramics;

• the second group (2) comprises all theBronze Bell Beakers (60%) and otherBronze and Chalcolithic ceramics;

• the third group (3) is mainly composed ofBronze Age ceramics (80%)

Chalcolithic ceramics can be differentiatedfrom Bronze Age ceramics by a chemical com-position more correlated with the basic rocks ofthe region, namely clays derived by weatheringof quartzodiorites, diorites and gabbros. BellBeakers ceramics have a more homogeneouscomposition than other Bronze Age ceramics.Bell Beakers and other Bronze Age ceramicspresent a similar chemical composition, with achemical signature also found in local clays -weathered schists and sedimentary clays.

Hence, at Monte do Tosco site, a local pro-

duction for all ceramic typologies can be deli-neated, including Bell Beakers. Still a fewoutliers are defined comprising the analysedcrucibles and three Bell Beakers (Fig. 7),pointing to the use of different raw materials,probably a different provenance.

2.3. Fraga da Pena, Beira Alta

The Fraga da Pena archaeological site is afortified settlement located in a huge granitictor, impressively conspicuous in the surroun-ding country, occupied at the transitionChalcolithic - early Bronze Age (last quarter ofthe third millennium BC) (Fig. 8).

The material culture indicates an occupationwhere symbolic activities could overcome thedomestic ones (Valera, 2006).

Among the several typological groups iden-tified, four were selected for an archaeome-tric study: Bell Beakers (Fig. 9) decoratedvessels with combed incisions; vessels withmorphological and decoration patterns ofChalcolithic tradition; and Bronze Age newmorphologies.

Clays available in the settlements


Fig. 6. Monte do Tosco archeological site (Alentejo, Portugal (photo-graphs by António Valera).

Fig. 7. Plot of means for each cluster (K-means method) using chemi-cal results of Monte do Tosco ceramics as variables.

Fig. 8. Fraga da Pena archeological site, central Portugal (photographby António Valera).

Fig. 9. Bell Beakers from Fraga da Pena archaeological site, centralPortugal (photographs by António Valera).

37

surrounding area are residual materials ofgranites, veins of quar tz, aplitedolerite,and schists. Dolerites provide the mostclayey samples. The aplite pegmatiteveins also provide argillaceous materials,ver y rich in alkali feldspars. Plagioclasereaches the higher levels in granites. Ironoxides and pyroxenes reach the higherlevels in dolerites. Micas occur in allcases. A good dif ferentiating indicatorwithin these samples is the phyllosilica-tes, plagioclase, iron oxides and pyroxe-nes propor tion.

Clays derived by weathering of granite anddolerite can be distinguished by their Na andK contents related to plagioclase and alkalifeldspar, and other elements with geochemi-cal affinity to ferromagnesian minerals, suchas Cr, Co, Sc and Fe. Although it is possibleto clear differentiate these geological mate-rials, significant variations of the REE con-tents may occur within the same type of claysdepending on the weathering degree of theparent rock.

Macroscopic observation and petrographicanalysis showed significant amounts ofnonplastic grains, a high irregularity of thegrain size distribution, lumps of clay joinedto form the paste, giving a very irregulartexture, and a very nebulous and irregular

aspect of the orientation of the clays. Thusthe claytemper mixture materials used tomake the vessels were cer tainly notground and sieved.

Compositional studies of all ceramic typo-logies from Fraga da Pena revealed thatBell Beakers have in general a similarcomposition. The main geochemical featu-res enhance three main groups, which arerelated with the regional geological mate-rials:

• Group 1 comprises mostly Bell Beakers;

• Group 2 is mainly composed of Chalcolithicand Bronze Age typologies, combed incsions ceramics, and a few Bell Beakers;

• Group 3 includes Bell Beakers and Chalcolithicand Bronze Age typologies (Fig. 10).

Phyllosilicates, mainly mica, illite, chloriteand smectite occur in Fraga da Pena cera-mics. XRD allowed the identification of sig-nificant amounts of smectite and interstra-tifieds in some sherds. The results obtai-ned point to low temperature firing proces-ses, not higher than 500 ºC – 600ºC (Diaset al., 2005).

The chemical patterns of the Bell Beakers are


Fig. 10. Plot of means for each cluster (K-means method) using chemical results of Fraga da Pena ceramics as variables.

38

similar to the dolerites samples, with signifi-cant amounts of elements associated to ferro-magnesian minerals. The dissimilarities foundmay be due to the differences among diffe-rent dolerite veins and/or different stages ofweathering as revealed by the REE fractiona-tion. The chemical composition of Chalcolithicand some Bronze Age typologies points to theuse of more acid rocks, like local/regional gra-nites, with higher contents of elements rela-ted with the presence of feldspars (Na, K, Rb,Cs), higher amounts of REE, specially LREE,Zr, Hf, Ta and Th, and a depletion in elementsof the first transition series.

Textural features obtained by petrography,together with mineralogy and chemistry, indi-cate that weathered dolerites were used forthe production of fine pottery. Coarser pastesappear to have been done using weatheredgranites. In both cases non-plastic grains ofgranite origin are found. For the production ofthe Bell Beakers, specially the “nail printed”ones, a careful production technology occu-rred, with the use of well selected materialsof dolerite origin.

In general non plastic grains occur in greatamounts added by the potter in the processof assembling the paste to produce cera-mics. The high irregularity of grain size distri-bution, the irregular texture, and the verynebulous and irregular aspect of the orienta-tion of the clays indicate that raw materialswere not ground or sieved and well mixed.

Bell Beakers, especially the “nail printed”ones, present a thinner paste with wellselected nonplastic grains, indicating acareful mixing and working of the clayresource and temper grains, thus a diffe-rent process of making the pot. Thereforemost of the Bell Beakers from Fraga daPena, especially the “nail printed” ones,were produced using a careful local produc-tion technology.

3. Contribution of compositional data forsolving archaeological problems relatedwith pottery circulation

The methodological approach by usingcompositional data and archaeological evi-

dence to solve ceramics mechanisms ofcirculation, including the Bell Beakersissue, has been very useful. In the threecase studies (Por to Torrão, Fraga da Penaand Monte do Tosco archaeological sites),local production was ascer tained for themajority of the cases. Never theless someexogenous ceramics were found.

The mineralogical assemblage of pre-histo-ric ceramics in general does not presenthigh temperature mineral phases, pointingall of them to low firing temperatures -below 500ºC.

In general, the results obtained point tolocal productions including Bell Beakers.However in some cases a careful selectionof the clays and/or a dif ferent preparationof the raw materials to obtain a more finepaste for the production of special wareslike “nail printed” were also done in theFraga da Pena archaeological site.

In a few cases the circulation of the“Product” was suggested. In the threearchaeological sites most of the outliersfound are Bell Beakers. The transpor t ofthese fragile ceramic objects over greatdistances, enhance the impor tance ofthese typologies over a cer tain period.This approach helps to conceptualize thedif fusion models of ceramics.

Thus, the results indicate that severalceramic types including Bell Beakers weremade with local/regional resources andonly occasionally impor ted.

Acnowledgements

The help of Jose Miguel Herrero by impro-ving and editing the manuscript is grate-fully acknowledged.

4. References

Burbidge, C.I., Dias, M.I., Prudêncio, M.I., Rebêlo,L.P., Cardoso, G. & Brito, P. (2009) Internal a acti-vity: localisation, compositional associations andeffects on OSL signals in quartz approaching bsaturation. Radiation Measurements, 44,494–500.


39

Burbidge, C.I., Rodrigues, A.L. Dias, M.I.,Prudêncio, M.I. & Cardoso. G. (2010) Optimisationof preparation and measurement protocols forluminescence dating of small samples from a suiteof porcelains and faiences. MediterraneanArchaeology and Archaeometry, 10, No. 4, 53-60.

Burt, D.M. (1989) Compositional and phase rela-tions among rare earth elemenks. In Geochemistryand mineralogy of rare earth elements (B.R. Lipinand G.A. McKay, eds). Reviews in Mineralogy, 21,Mineralogical Society of America, 259--307.

Cabral, J.M.P., Prudêncio, M.I., Gouveia, M.A. &Morais Arnaud, J.E. (1988) Chemical and mineralo-gical characterization of Pre-Beaker and Beaker pot-tery from Ferreira do Alentejo (Beja, Portugal).Proceedings of the 26th International ArchaeometrySymposium. Editor: Farquhar, R.M.; Hancock,R.G.V.; Pavlish, L.A. , Publisher/Distributor:University of Toronto. Archaeometry Laboratory,172-178.

Chappell, J. (1991) The potter’s complete book of clayand glazes. Watson-Guphill Publications, NY, 416 pp.

Dias, M.I. & Prudêncio, M.I. (in press) Geochemicaland mineralogical characterization of Bell Beakersfrom Portuguese sites. A contribution to the esta-blishment of provenance and circulation. -In: SALA-NOVA, L. (ed): Mécanismes de circulation desvases campaniformes, Paris.

Dias, M. I., Prudêncio, M. I. & Gouveia, M.A.(2001) Arqueometria de cerâmicas islâmicas dasregiões de Lisboa, Santarém e Alcácer do Sal(Portugal): caracterização química e mineralógica,in GARB, Sítios Islâmicos do Sul Peninsular (eds.IPPAR/MC e Junta da Extremadura — Consejaríade Cultura), 257–81.

Dias, M.I. & Prudêncio, M.I. (2007) "Neutron acti-vation analysis of archaeological materials: anoverview of the ITN NAA Laboratory, Portugal.Archaeometry 49, 2, 381-391.

Dias, M.I. & Prudêncio, M.I. (2008) On the impor-tance of using Sc to normalize geochemical dataprevious to multivariate analyses applied toarchaeometric pottery studies. MicrochemicalJournal, 88, 136-141.

Dias, M.I., Prudêncio, M.I. & Gouveia, M.A.

(2003a) Geochemical study of clay materials inFornos de Algodres region (Central Portugal) in anarchaeometric view. In: E. Dominguez, G. Mas & F.Cravero (eds), A Clay Odissey, Elsevier, 65-70.

Dias, M.I., Prudêncio, M.I. & Rocha F. (2003b)Amphorae Production at Occidental Lusitania:Identification of Raw Material and Production.-In:Pérez-Rodríguez, J.L. (ed.): Applied Study ofCultural Heritage and Clays, CSIC, Madrid,187–200.

Dias, M.I., Prudêncio, M.I. (2008), On the impor-tance of using scandium to normalize geochemicaldata preceding multivariate analyses applied toarchaeometric pottery studies, MicrochemicalJournal 88, 136-141.

Dias, M.I., Prudêncio, M.I., Gonçalves, M.A.,Sequeira Braga, M.A. & Gouveia, M.A. (2000)Geochemical and mineralogical diversity of claymaterials in Fornos de Algodres region (CentralPortugal) and its implications on provenance stu-dies of ancient ceramics.- Proceedings of the 1st

Latin American Clay Conference, Funchal 2000, VolII, 237-244

Dias, M.I., Prudêncio, M.I., Gouveia, M.A.,Trindade, M.J., Marques, R., Franco, D., Raposo,J., Fabião, C.S. & Guerra, A. (2010) Chemical tra-cers of Lusitanian amphorae kilns from the Tagusestuary (Por tugal). Journal of ArchaeologicalScience 37, 784–798.

Dias, M.I., Prudêncio, M.I., Valera, A.C. SequeiraBraga, M.A. & Gouveia, M.A. (2002): Provenanceand Technology of Pre-Historic Pottery From Fornosde Algodres (Portugal): The Fraga da Pena archao-logical site.- British Archaeological Reports, Inter-national Series 1011, 253-264.

Dias, M.I., Valera, A.C. & Prudêncio, M.I. (2005)Pottery production technology through out the 3rd

millennium B.C. on a local settlement network inFornos de Algodres, central Por tugal. In:Prudêncio, M.I., Dias, M.I. & Waerenborgh, J.C.(eds), Trabalhos de Arqueologia, Série MonográficaTrabalhos de Arqueologia, 42, IPA, 41-50.

Glascock, M.D. & Neff, H. (2003) Measurement.Neutron activation analysis and provenance rese-arch in archaeology. Science and Technology 14,1516-1526.


40

Gouveia, M.A., Prudêncio, M.I., Figueiredo, M.O.,Pereira, L.C.J., Waerenborgh, J.C., Morgado, I.,Pena, T. & Lopes, A. (1993) Behaviour of REE andother trace and major elements during weatheringof granitic rocks, Évora, Por tugal. ChemicalGeology, 107, 293-298.

Hector, N. (ed) (1992) Chemical characterization ofceramic pastes in archaeology. Monographs inworld archaeology 7. Prehistory Press, Madison,Wisconsin.

Kin, F. D., Prudêncio, M.I., Gouveia, M.A. &Magnusson, E. (1999) Determination of rare earthelements in geological samples: a comparativestudy of instrumental neutron activation analysisand inductively coupled plasma mass spectro-metry. Geostandards Newsletters, 23, 47-58.Maggetti, M. (1982) Phase analysis and its signifi-cance for technology and origin. Pp. 121-133 in:Archaeological Ceramics (J.S. Olin, editor).Smithsonian Institution Press, Boston.

Marques, R. (2007) Geoquimica e mineralogia deargilas do Cretácico de Taveiro e Aveiro, Portugal.MSc thesis, University of Aveiro, Portugal, 109 pp.

Moropoulou, A., Bakolas, A. & Bisbikou, K. (1995)Thermal analysis as a method of characterizingancient ceramic technologies. Thermochimica Acta2570, 743-753.

Prudêncio, M.I., Dias, M.I. Gouveia, M.A.,Marques, R., Franco, D., Trindade, M.J. (2009)Geochemical signatures of Roman amphorae pro-duced in the Sado River estuary, Lusitania(Western Por tugal). Journal of ArcheologicalScience, 36, 873-883.

Prudêncio, M.I., Dias, M.I., Raposo, J., Gouveia,M.A., Fabião, C., Guerra, A., Bugalhão, J., Duarte,A.L. & Sabrosa, A. (2003) Chemical characterisa-tion of amphorae from the Tagus and Sado estua-ries production centres (Portugal). In: S. di Pierro,S., Serneels, V. & Maggetti, M. (eds): Ceramic inthe Society, 245-253.

Prudêncio, M.I., Gouveia, M.A. & Sequeira Braga,M.A. (1995) REE distribution in actual and ancientsurface environments of basaltic rocks (centralPortugal). Clay Minerals, 30, 239-248.

Prudêncio, M.I., Marques, R., Rebelo, L., Cook,

G.T., Cardoso, G.O., Naysmith, P., Freeman,S.P.H.T., Franco, D., Brito, P., Dias, M.I. (2007)Radiocarbon and blue optically stimulated lumines-cence chronologies of the Oitavos consolidateddune (Western Portugal). Radiocarbon, 49, nº2,1145-1151.

Prudêncio, M.I., Oliveira, F., Dias, M.I., SequeiraBraga, M.A., Delgado, M. & Martins, M. (2006)Raw materials identification used for the manufac-ture of Roman “Bracarense” ceramics from NWIberian Peninsula. Clays and Clay Minerals, 54, 5,638-649.

Prudêncio, M.I., Waerenborgh, J.C. & Cabral, J.M.P.(1988) Chemical and mineralogical characteriza-tion of a cretaceous clay from the Lousã Basin(central Portugal). Clay Minerals, 23, 4, 411-422.

Rice, P.M. (1987) Pottery Analysis, A source book.-The University of Chicago Press, Chicago, 559 p.StatSoft, Inc. (2011)

StatSoft, Inc. (2011) STATISTICA (data analysissoftware system), version 6. www.statsoft.com

Trindade, M.J., Dias, M.I., Coroado, J. & Rocha, F.(2010) Firing tests on clayrich raw materials fromthe Algarve Basin (Southern Portugal): Study of themineral transformations with temperature. Claysand Clay Minerals, 58, 188-204.

Valera, A.C. (2006) Calcolítico e transição para aIdade do Bronze na Bacia do Alto Mondego: estru-turação e dinâmica de uma rede local de povoa-mento. PhD Thesis. Universidade do Porto.

Velde, B. & Druc, I.C. (1999) ArchaeologicalCeramic. Materials. Origin and utilization. Springer-Verlag, Berlin, Heidelberg, 299 pp.


41

Technology and Provenancing ofFrench faience/ Marino Maggetti

Abstract

An overview is given of the recipe and the “chaîne opératoire” of French faiences, as revealed inancient textbooks and archives of the 18th and 19th century. The preparation of the calx, the glaze,the glazing, the decoration with in-glaze and on-glaze colours, the kiln and the three firings (first,second, third) are described and the chemical, mineralogical and technological aspects assessed.Archaeometric analyses of French faiences are scarce. Scientific analyses of such crockery, mainlyfrom the workshops of Le Bois d’Épense/Les Islettes and Granges-le-Bourg, constrain the natureof the clay, the firing temperatures, the glaze composition and the colour pigments. Successful attri-butions of faience objects are shown. However, such provenancing need much more robust andwellknown chemical reference groups.

Resumen

Se presenta una vision general de la receta y la “cadena operativa” de la Fayenza francesa, talcomo se concebía en antiguos manuales y archivos de los siglos XVIII y XIX. Se describen los pro-cesos de preparación de cenizas, vidriado, la aplicación del vidriado, la decoración con colores enel vidriado o post-vidriado así como los aspectos químicos, mineralógicos y tecnológicos involucra-dos en el proceso. Hay poca información arqueométrica de fayenza francesa. Los análisis científi-cos de vajillas vidriadas, principalmente de los talleres de Le Bois d’Épense/Les Islettes y Granges-le-Bourg, ponen de manifiesto la naturaleza de la arcilla, temperaturas de cocción y composiciónde los vidriados y pigmentos. Se muestran los resultados de asignar piezas de fayenza a los talle-res de procedencia. Sin embargo, los estudios de proveniencia necesitan consolidarse aún más ydisponer de grupos de referencia químicos conocidos.

Key-words: Tin glazed earthenware, French faience, calcareous clay, glaze, biscuit, firing temperatures,archaeometry.

1. Introduction

Ancient and modern ceramic products can be studied from different perspectives. To the petrologist,they are artificial rocks subjected to relatively high temperatures and recrystallized to become ther-mometamorphic products, analogue to those naturally formed through metamorphism (Maggetti2001). In contrast to natural rock forming processes, pressure is insignificant in the genesis of suchobjects, because the kiln can be considered as a technical system. In such open systems, neitherpore solutions, present before firing, nor gaseous reaction products, which may have been producedduring the high temperature process, have an influence on the transformation, because they canleave the system at any time. Similar to natural rocks, ceramic objects consist of an assemblage ofcrystalline and amorphous phases, which can be analysed with the same petrographic, mineralogi-cal and chemical methods such as micaschists, which are formed from clay during metamorphism.

Faience is, in the French nomenclature, an earthenware body coated with a white glaze opacifiedby tin oxide (SnO2) crystals (Fig. 1). These cassiterites are present as small particles up to 10 µmacross or as bigger clusters. The finely distributed tiny crystals refract and reflect light almostwithout absorption losses (Vendrell et al. 2000). Hence the ceramic object takes on a white andopaque appearance even though it is coloured in the bulk. Tin-glazed earthenware pottery resem-bles therefore, from the outside, porcelain.

Department of Geosciences, University of Fribourg, Fribourg, Switzerland.

42

This paper deals with the technique and theorigin of 17th to 19th century French faiencetableware. Faience tiles and older faiences ofthis country are not addressed.

2. Discovery and spreading of tin-glazetechnique

2.1. Mesopotamia and China

The origin of the tin glaze technique can betraced to what today is Iraq (Caiger-Smith1973, Soustiel 1985). An important produc-tion centre was Baghdad, the palace cityerected by the second Abbasid Caliph AbuJa’far al-Mansur in 762 AD in the locationwhere the Silk Road crosses the Tigris River(Fig. 2).

It is still a matter of debate if the tin-glazetechnique was invented as reaction of Islamicpotters from Mesopotamia and Persia to theChinese competition, trying to emulate thegleaming white surface and the hard, com-pact body of Chinese porcelain (Watson1987) or if this technology appeared in Basraas a genuine local invention already during700-750 AD, several decades prior to theknown import date of Chinese ware (Mason &Tite 1997, Mason 2004).

At the end of the 9th century, in Baghdad,Basra, Samarra and other places, the tin-gla-zed pottery technique was well established.This novel glaze technique resulted in twoadvantages. On the one hand, the off-whitecolour of the ceramic body was masked by theopaque glaze and on the other hand, the whiteglaze surface provided an ideal canvas for pain-ted decorations. Consequently, the Islamic

Technology and Provenancing of French faienceMaggetti, M.

Fig. 1. Backscattered electron images of end 18th century faiencesfrom the workshop Le Bois d’Épense/Les Islettes. (a) Sample BEI 6ashowing on top of the tin oxide opacified glaze a stroke of blue in-glazepainting; (b) Sample BEI 54 with a yellow in-glaze painting on top of theglaze. Photo M. Maggetti.

Fig. 2. Major centres of Islamic potter ymanufacture in the Middle East.

43

craftspeople discovered the inglaze blue,brown, green and yellow colours, which werelater adopted by the Chinese potters.

A further innovation of the Islamic crafts peo-ple was the lustre decoration, possibly inspi-red by the religious ban on using gold for pro-fane purposes. Hence only gold-like lustrevessels conformed to the religious command-ments and statutory laws. In terms of techno-logy, lustre ware was a transfer from glass toceramic technology, invented in Basra andFustat during the 8th and 9th centuries AD(Caiger Smith 1985).

At the end of the 12th century AD the firstceramic objects appeared in Iran, showingoverglaze decoration, a technique called‘painting with seven colours’ (blue, green,brown, violet, red, black and white) (Weiss1970, Soustiel 1985) and known as mina’ior ladjvardina wares. Analyses by Mason &Tite (1997) and Mason (2004) throw muchlight on the development of Mesopotamiantin glazes. The oldest (c. 700-750 AD) tin-opacified glazes from Basra were an impro-vement of the traditional local opaque gla-zes that owed their opacity to gas bubblesand fine crystallites of quar tz, feldspar,wollastonite and diopside. Cassiterites areconcentrated in a slip-like layer at the inter-face ceramic body/glaze. PbO is very low(around 1 wt.%, Fig. 3). Slightly younger(750-800 AD) glazes contain like earlierglazes gas bubbles, silicates and cassiteri-tes as opacifying agents. The latter, howe-ver, are spread throughout the entire glazeand SnO2 contents are slightly elevated (3-4 wt.%, Fig. 3). The glassy coatings arealkali-lime glazes with 1-2 wt.% PbO. Tinoxide concentrations of later glazes (800-975 AD) reach 4-8 wt.%, sufficiently high toeschew additional opacifying agents. Owingto the elevated PbO content (3-11 wt.%)these glazes can be classified as alkali-lead-lime glazes. Glazes from Baghdaddated c. 800-900 AD show substantiallyincreased PbO (31-41 wt.%) and SnO2 (6-8wt.%) contents. Tin-opacified lead-alkaliglazes with 25-35 wt.% PbO from EgyptianFustat of 975-1025 AD show even higherSnO2 contents (9-16 wt.%) compared to theBasra or Baghdad glazes.

The Mesopotamian Islamic tin-glazed wares,painted or not, with or without lustre, were allproduced from calcareous (16-22 wt.% CaO)clays very similar in composition (50-55 wt.%SiO2, 10-12 wt.% Al2O3, 5-7 wt.% MgO, 5-6wt.% FeO and 1-2 wt.% Na2O; Mason, 2004).Firing temperatures of blue-painted waresfrom Basra were estimated at 850-1050ºC(Tite 1988). Since Baghdad paste contains inaddition some glass fragments it is conceiva-ble that this ceramic production centre useda technology dif ferent from Basra.Mesopotamian lustres were studied byPradell et al. (2008 a, b) and Colomban &Truon (2004).

2.2. Spread of tin-glazed pottery

Very early the Arabian westward expansionreached via North Africa Spain that was con-quered between 711 and 719. In the wake ofthe new rulers Islamic potters settled inSpain and introduced this technology.However, it is unknown when the firstSpanish tin-glazed ceramic objects were pro-duced. From the 12th to the 15th centuries tinglazing flourished in pottery centres such asTalavera-Puente, Paterna-Manises and


Fig. 3. SnO2 and PbO concentrations in Mesopotamian tin glazes(Mason & Tite 1997, Mason 2004).

44

Sevilla. In the 13th century the Hispano-Moresque lustre decoration technique wasdeveloped to high perfection, in particular inmanufactories of Malaga and Manises.These potters combined the cobalt-blue pain-ted decoration with the metallic lustre coatingand thus revived a technique that had beenlost in 1224 by the destruction of Rajj inPersia by the advancing Mongols.

Many lustre-decorated ceramic plates andcups display the coat-of-arms of French andItalian noble families thus showing how highlyesteemed this tableware was. These objectsdid not serve their foreign clients as dinnerwa-re for daily use anymore but were consideredprestigious objects of art fit to grace theirostentatious sideboards. From Spain this typeof ceramic ware was introduced to Francewhere in mid-13th century in Marseille tin gla-zed pottery was made, and eventually to Italy.From there the knowledge of production of tin-glazed ceramic crossed the Alps and extendedinto the northern parts of Europe. In its wakea continuous stream of new manufactures wasestablished in France, Germany, TheNetherlands, England, Switzerland etc.Eventually the faience technology reached theNew World (Olin et al. 1978, Maggetti et al.1984, Olin and Blackman 1989, Jamieson andHancock 2004). Tin glazed ware is calledHispano-Moresque ware in Spain, maiolica inItaly, faience in France and delftware inNorthern Europe.

3. Archaeometric studies of Frenchfaiences

Such studies are very scarce. Some show nochemical analyses at all (Démiansd’Archimbaud & Picon 1972, Picon &Démians d’Archimbaud 1978, Vallauri et al.1978, Picon 1993, Rosen 1997a,b), otheronly mean values (Vallauri & Leenhardt 1997,Picon 2000, Rosen 2001, 2007) and few thecomplete list of analyses (Carette et Deroeux1985, Dufournier & Deroeux 1986, 1987,Dufournier 1989, Schmitt 1990, Pellet 1993,Rosen 2000b, Bernier 2003, Meunier &Bouquillon 2004, Maggetti et al. 2009a, b,Rosen 2009, Rosen et al. 2009). Glaze andpigment analyses are discussed in these andother papers (Bouquillon 2000, Oger et al.

2002, Dufournier et al. 2004, Marco deLucas et al. 2006, Rosen et al. 2007,Maggetti et al. 2009a, b).

4. Technology of French faience

4.1. First French faience productions

The first French tin-glazed earthenware wasproduced in Marseille at the beginnings ofthe 13th century (Marchesi et al. 1997,Rosen 2000a). Local workshops createdthere glaze-less objects or objects with atransparent lead-glaze, both in a Middle-Agetradition, as well as tin-glazed pieces in anew, Islamic technique. This technology spre-ad rapidly throughout France. But during thefirst 300 years (end of the 13th until the early16th century), only tin-glazed tiles were produ-ced for a wealthy clientele (Rosen 2000a).Italian potters moved ca. 1550 to France andopened before the end of the 16th century tinglaze pottery workshops in Lyon, Montpellier,Nevers and Cosne-sur-Loire.

These establishments were the starting pointfor the extraordinary expansion and successof French faience in the 17th and 18th centu-ries (Figs. 4, 5). Around the mid-16th centuryin Faenza the biscuits were coated with a par-ticularly thick pure white and matt glaze.These bianchi di Faenza were so popular andcommanding throughout Europe that to thisday in France the name of the city becamethe synonym for French tin-glazed pottery,


Fig. 4. Map of 17th century French faience manufactures. Redrawnfrom Faÿ-Hallé & Lahaussois (2003) and completed.

45

faïence. In the time span 1550-1600, Lyon’stin glaze potters were named ‘white earthpotters‘ or ‘vase makers in the style ofVenice‘, and those from Nevers ‘white croc-kery makers in the style of Venice‘ (Rosen2000a). The term faience appeared for thefirst time in 1604 in Nevers (France) wherethe ceramist Jean-Baptiste Conrade was labe-lled ‘sculpteur en terre de fayence’ (sculptorof faience earth) (Rosen 2000a).

4.2. Technical treatises of the 18th and19th centuries

The manufacturing techniques of this specialkind of pottery are well known thanks to con-temporary publications (Diderot 1756, Boscd’Antic 1780, Anonymous 1783, Boussemart1786, Boyer 1827, Bastenaire-Daudenart1828, Harlé 1831, Brongniart, 1844), morerecent papers (Deck 1887, Munier 1957,Montagnon 1987, Rosen 1995, 2009,Peiffer 2000, Bastian 2002-2003, Maggetti2007a) and the mid 18th century Caussymanuscript (de la Hubaudière & SoudéeLacombe 2007).

4.3. ‘Chaîne opératoire‘ of French faience

4.3.1. Nature and treatment of the clays

18th and 19th centuries French faience is

manufactured from clays in several produc-tion steps (Fig. 6). The composition of idealfaience clays is 58 wt.% SiO2, 30 wt% Al2O3,5 wt.% Fe2O3 and 7 wt.% CaCO3 (Bastenaire-Daudenart 1828). In order to obtain the cal-careous body typical of tin-glazed potteries(Caiger-Smith 1973, Tite et al. 2008, Tite2009), either clays which already contain theideal amount of carbonates (calcite and dolo-mite), or artifical mixtures of several kinds ofclay are used. The French faience makers ofthe 18th and 19th centuries preferred mixtu-res by far, ranging from two as in the case ofLe Bois d’Épense (Liénard 1877), Nevers(Bosc d’Antic 1780) and Rouen (de laHubaudière & Soudée-Lacombre 2007) tothree (Aprey, Froidos, Lavoye, Paris,Rarecourt, Salvange, Thionville, Tours andWaly, Bosc d’Antic 1780, Brongniart 1844,Liénard 1877) or even four clays (Paris,Bastenaire-Daudenart 1828). French faien-ces with non or low calcareous paste are rareand show technical drawbacks (Munier 1957,Caiger-Smith 1973, Picon et al. 1995,Thornton 1997).

Archaeometric analyses confirmed the blendingof two clays, a low and a high calcareous one,for Meillonnas (Rosen 2000b) and Nevers(Rosen 2009), and the use of one clay inGranges-le-Bourg (Maggetti et al. 2009c).Chemical analyses proved the calcareous natu-re of faiences from Arthé (25-32 wt.% CaO,Pellet 1993, Rosen 2001), Dijon (14-26 wt.%,Rosen 2001), Le Bois d’Épense (17 –24 wt.%,Maggetti et al. 2009a), Montpellier (17-28 wt.%,Rosen et al. 2009), Moustiers (15-32 wt.%,Rosen et al. 2009), Nevers (19-30 wt.%, Rosenet al. 2009, Rosen 2009), La Rochelle (11-21wt.%, Rosen et al. 2009), Villers-le-Pot (21-28wt.%, Rosen 2001). CaO mean values of faien-ces from Ancy-le-Franc (20 wt.%, Rosen 2001),Chevannes (17 wt.%, Rosen 2001), Meillonnas(17 wt.%, Rosen 2000b) and Vausse (23 wt.%,Rosen 2001) fit well in this scheme. Faiencesfrom Granges-le-Bourg are magnesium-rich(MgO 5-10 wt.%) and were made from dolomiticmarls (Maggetti et al. 2009c, d). Five faiencesfrom Varages (South of France) have more than11 wt.% MgO (Schmitt 1990).

There are five reasons why calcareousclays were widely used (Tite 2009):


Fig. 5. Map of 18th century French faience manufactures. Redrawnfrom Rosen (2001)

46

(1) They are very common;

(2) High CaO-bodies shrink considerably duringcooling, putting the glaze under compres-sion;

(3) CaO acts like a bleach during firing, givingpale buff colours;

(4) The thermal expansion coefficient of calca-reous bodies matches those of lead-alkaliglazes and

(5) Microstructures remain essentially unchan-ged over the 850-1050ºC firing range.

Calcareous bodies show significantly lessfiring shrinkage and have a greater rigidityand compressive strength as non calcareouswares.

The raw materials extracted from river sedi-ments or a quarry must undergo several ope-rations to remove the rougher elements andto make them homogeneous. Finely ground


Fig. 6. “Chaîne opératoire” of the four faience types (without lustred faience). (A) White faience. Below, photo of a plate from the faience manu-facture Du passage de la Cour-Robert in Fribourg, Switzerland, ~1790-1810. Archaeological Survey of the Canton Fribourg, Switzerland, inv. FPL-CRI No 1646. Photo G. Bourgarel. (Bourgarel, 2007); (B) Faience with in-glaze decoration. Below plate from the faience workshop of Le Bois d’É-pense/Les Islettes, France, ~ 1800, diam. 22,5 cm, inv. 943.1.32. Collection and photo of the Museum Bar-le-Duc (Rosen, 2007a); (C) Faiencewith on-glaze decoration. Below plate with a cock, after 1830 (?), from a Lorraine manufacture, most probably Lunéville or Saint-Clément. (pers.coll.). Photo M. Maggetti; (D) Faience with in- and on-glaze decoration. Below plate with stamp Paul Hannong from Strasbourg, ~ 1735-1748, dia-meter 24,8 cm. Collection and photo J. Bastian.

47

temper can also be added, or two or moreclays mixed. The clay was first ground, thenplaced in suspension in water in vats orbarrels and well stirred to separate the grains(washing). The suspension was filtered, pou-red into a large earthen pit and once againwell stirred. The suspension was then allo-wed to decant, the unwanted coarse particlessettled to the bottom to form a coarse inferiorlayer. Specialized workers removed the finersuperior layers and poured them into anothervat. This decanting process was repeatedseveral times.

The excavations made by Jean-JacquesThévenard in the faience manufacture of LesAuges in Langres, revealed the presence of awhole series of clay pools (Fig. 7). The pastewas then stocked in a cellar and allowed torot for several weeks, even months. Beforethe material taken from these rotting-cellarscould be used, it had to have the exactamount of water necessary, obtained eitherthrough drying to remove a surplus, or byadding liquid.

Homogenization was the following step, duringa process called marching (Fig. 8). The pastewas spread on a hard surface, in the shape ofa circle a few centimeters thick. A specializedworker, trained in this delicate task, tread onthe clay, walked it, going from the center out,and then the reverse. The effect of this opera-tion was not only to homogenize the clay, butmade it also possible for the worker to detectnumerous undesirable tiny particles andforeign bodies. Then the paste, shaped intoballs of different sizes was forcibly beaten withwooden bats or iron bars in order to extract

any air it could still contain.

4.3.2. Shaping, dr ying and first firing

Once prepared, the mass of plastic clay couldbe shaped according to two main techniques,throwing and moulding. The first makes itpossible to obtain symmetrical objects suchas platters, jugs, etc. (Fig. 9). The second isused to shape non-symmetrical objects suchas handles, spouts, decorations but also forsymmetrical shapes (Fig. 10).

In order that they may dry, the objectsthrown on the wheel or moulded were pla-ced on long wooden planks, placed on shel-ving in a specific, well-aired room, or evenin the shaping workshop (Fig. 8).Consequently, a lot of space was necessaryfor this important and delicate phase, as inthe case of the drying of the moulds, whoseaim it was to remove as much water as pos-sible from the objects to prevent their shat-tering during firing.


Fig. 7. Clay washing pools. Faience manufacture of Les Auges, middleof the 19th century (Langres, France). Excavations J. J. Thévenard.Photo M. Maggetti

Fig. 8. Schematic view of the inside of a manufacture with a worker"walking" the clay (b), mounds of kneaded clay (c), a worker shapingan object (d) and long shelves to dry the moulds and the objects eitherthrown or moulded. Diderot (1756, plate II).

Fig. 9. Wheel throwing. The worker in front is shaping and his colleaguebehind is turning, i.e. finishing the object (Brongniart, 1844, plate XLV).

48

During the drying process, water evaporationprogresses outward in and its consequenceis a loss in volume, called drying shrinkage.The pieces are thus subjected to tensionsand it was necessary to ensure that this eva-poration took place very slowly and uniformly,without any risk of cracking or distortion.

The usual procedure, in the production line offine ceramics, is to examine the objects todetect any cracking before placing them inthe kiln, because even a very fine crack canturn into breakage of the raw (green) objectduring firing (Casadio 1998). These dry andcracked pieces are not rejected, but recycledin the shaping phase, as the raw mass canbe crushed and rehumidified. This type of dis-card leaves no trace and is not quantifiable.After forming and drying, the green leather-hard ceramic bodies were fired in a kiln givinga porous, coloured biscuit (Fig. 11). Duringfiring water and other volatile compounds willevaporate.

Mineralogical changes and textural evolutionsof calcareous illitic clays during oxidativefiring were studied by many authors (Peters &Jenni 1973, Maniatis & Tite 1975, Tite &Maniatis 1975, Heimann 1982, Maggetti1982, Shoval 1988, Echallier & Mery 1992,Diminuoco et al. 1998, Cultrone et al. 2001,Trindade et al. 2009 and literature therein). X-ray diffraction analysis of experimentally firedsamples of a given clay is a popular methodto estimate ancient firing temperatures. Suchheat related phase changes of a dolomiticclay with a chemical composition matching

those of the faiences from the manufactureof Granges-le-Bourg are shown in Fig. 12.

The X-ray diffraction patterns of biscuits fromthe manufactures of Le Bois d’Épense/LesIslettes (Maggetti 2007b) and Granges-le-Bourg (Maggetti et al., 2009c) were comparedwith published results (and Fig. 12), giving esti-mated maximum firing temperatures of 950ºCfor the first (or biscuit) firing. This holds mostprobably for all other French manufactures, asa highly fired biscuit wouldn’t be enoughporous to adsorb sufficient tin glaze suspen-sion (Bastenaire-Daudenart 1828, p. 395).

4.3.3. Preparing the tin oxide bearingglaze

This lead glaze, opacified with tin oxide, beginsto melt at low temperatures around 900°C andshould show a thermal dilatation coefficientsimilar to that of the biscuitted ware to avoidcrazing or scaling. The raw material is alwaysquartz sand (SiO2), as pure as possible, with asignificantly low iron content, to avoid the deve-lopment of a yellowish color during the oxidati-ve second firing. In order to lower the meltingpoint of quartz from 1713°C to approximately900-950°C, it is necessary to add fluxes suchas lead oxide and alkalis (potassium, sodium).French faience makers in the 18th and 19th

centuries used sodium rich fluxes, mostly seasalt (NaCl) and less soda (Na2CO3), as indica-ted in contemporaneous glaze recipes, whichvary greatly from one manufacture to the next.

The tin glaze, in its final stage, is a whitish pow-der obtained through a series of complex ope-


Fig. 10. Different gypsum moulds from the Bois d'Épense/Les Islettesmanufacture with a few biscuits from these moulds (pers. coll.). Early19th century. Photo M. Maggetti.

Fig. 11. Biscuits from the faience manufacture Le Bois d’Épense/LesIslettes, France (pers. coll.). Photo M. Maggetti.

49

rations. This begins with the preparation of acalx or calcine which consists in merging andoxidizing tin with lead in proportions goingroughly from 1/3 tin to 2/3 lead up to 1/5 tinto 4/ 5 lead (in weight), which means 50 to 16kg of tin for 100 kg of lead. To cut down on tin,a costly metal, it was even possible to go downto 10 kg of tin for 100 kg of lead, by addingquartz to the frit (see below). The quantity of tinbears a direct influence on the quality of thevitreous glaze, which becomes whiter and shi-nier as the amount of tin increases. Bothmetals were heated together in a small oven ofa specific type. Tin melts at 232°C, lead at327°C. When both metals are melted, the hea-ting process is continued, the oven door is ope-ned allowing air to enter. The liquid becomesoxidized, producing yellowish ash, pushed tothe back of the kiln by a worker with an appro-priate tool. The oxidation process continues assoon as the surface of the metallic liquidcomes once more in contact with the oxygen inthe air and the process of ash raking is conti-nued until there is no more liquid.

X-ray diffraction analyses of a calcine from themanufacture of Arthé detected three phases:massicot (PbO, 64.55%), lead stannate(Pb2SnO4, 31.46%) and cerussite (PbCO3,3.52%). The chemical composition is 91.01wt.% PbO, 7,94 wt.% SnO2 and 0.58 wt.% CO2(Total 99.53 wt.%, Pellet 1993). Lead stannateis a well-known yellow pigment in artist painting(Maggetti et al. 2009b).

Alkaline fluxes are highly soluble in theaqueous suspension of the tin glaze. They the-refore have to be stabilized, like the lead oxi-des, harmful to the workers' health. This is thesecond stage consisting in preparing a vitreousmass called frit. The lead stannate ashes, thatis the calcine, are removed from the oven,mixed with very pure sand and flux. The propor-tions vary greatly according to the place wherethe production occurs and even inside thesame factory. As an example, a recipe ofBastenaire-Daudenart (1828, p. 330) shows100 parts calcine (Sn:Pb = 30:100 parts) +100 parts Nevers sand (SiO2) + 12 parts salt(NaCl) + 6 parts minium (Pb3O4) + 5 partsAlicante soda (Na2CO3), or recalculated byBrongniart (1844, vol. II, p. 25) 44 calcine + 44sand + 8 salt + 2 minium + 2 soda (Total100%). Bosc d’Antic (1780) gives three recipeswithout any supplementary flux, estimating thatthe calcine’s PbO acts as flux. Some recipesmention the addition of a colouring metallicoxide to lightly tint the glaze, such as cobalt forexample or copper.

The mixture calcine + sand + fluxes was finelyground and then placed on a sand layer in thelower part, that is on the hearth of the faiencekiln. The exception is the workshop ofClermont-en-Argonne where the mixture wasfired in ceramic containers at the hearth of thekiln (Liénard 1877, p. 169). It melted almostcompletely thanks to the high temperatures inthis area. These were estimated 60 or 70


Fig. 12. Mineral stability in experimental firing of a dolomitic marl (sample GLB 1) from the manufacture Granges-le-Bourg. Portlandite is a post-firing hydrate of lime CaO.

50

degrees of the Wedgwood pyrometer(Brongniart (1844, II, p. 26), corresponding toca. 1100ºC. After cooling, the frit was seen asa whitish vitrous product, where the grains ofthe sand bed stuck to its base can still be seen(Fig. 13). Scanning electron microscopy revea-led the inhomogeneous granulometry and dis-tribution of the cassiterite particles. X-ray fluo-rescence analyses of three representative sam-ples from Granges-le-Bourg show rather cons-tant SnO2 concentrations (9-12 wt.%), andwider variations in SiO2 (38-51 wt.%) and PbO(26-40 wt.%). A sample from the manufactureof Pré d’Auge has 10 wt.% SnO2 (Dufournier etal. 2004). Such concentrations match those ofthe faience tin glazes. The frit was then crus-hed, extracted from the place where it wasfired, kneaded by hand to separate it fromimpurities and very finely ground in mills toobtain powdered opaque and white tin glaze.

4.3.4. Glazing

The powdered raw glaze was then diluted inwater, in which clay or other substances couldbe added to maintain the frit particles in sus-pension and to ensure adhesion between theglaze powder and the body after drying

(Parmelee 1948). "Often, potters would add tothe glaze some organic product such as flour-based glue, gum or honey to make its suspen-sion in water easier, since the high density ofthese coatings drove them speedily to the bot-tom. In other cases, clay is added to the sus-pension; it will act in the forming of the glazeduring the firing. All these additions give, more-over, greater cohesion to the film deposited onthe surface of the ceramic. […] When a trans-parent or opaque glaze has to be tinted throug-hout, the coloring agent, carefully crushed, canbe added to the aqueous suspension, but itcan also be mixed to the lead and tin oxides inthe frit." (Picon et al. 1995, p. 49).

The tin glaze suspension can theoretically beadded to the leather-hard, unfired object, as isthe case for other glazes, but French treatisesonly mentions the application of the glaze ontobiscuitted ware. Powdered tin-lead glaze, pla-ced directly onto a dry and unfired clayey objet,tends to shrink when fired. To avoid this, theglaze must be applied in a thick, costly layer.The advantage of a biscuit is that it does notreact when the aqueous suspension is placedon its surface, and that the dissociation of thecarbonates already took place during the firstfiring. A well fired, calcareous biscuit will theo-retically no longer give off carbon dioxide duringthe second firing. On the other hand, an unfiredcalcareous paste will give off a lot of CO2,which appears after firing in the shape of gasbubbles in the vitrous tin glaze.

The bisquitted pieces were dipped either byhand or with pincers (Fig. 14). To glaze theinterior of an object, the acqueous glazeslurry was poured inside and the surplus pou-red out after a certain time. Another techni-que consisted in using brushes or cloths dip-ped in the suspension. Thanks to the highdegree of porosity of the biscuits, the waterwas rapidly absorbed and a layer of powderedglazed formed on the surface of the ware.

The objects were removed and let dry (Fig.15). The traces left by fingers or holes due tothe use of pincers then had to be smoothedover with a brush. The objects were ready forthe second firing, if the aim was to producewhite dishes (Fig. 6A). Uniformly white table-ware was in all French manufactures the gre-


Fig. 13. A piece of tin-glaze frit showing a smooth vitrous break (a) andthe rough base (b). To the latter, grains of quartz are stuck from thesand bed on which the powdered mixture was deposited before thefiring. Sample GLB 80. Width 3 cm. Faience manufacture of Granges-le-Bourg (France), excavations D. Morin. Photo M. Maggetti

51

atest part of the production.

SEM-EDS analyses of 23 tin-glazes from LeBois d’Épense revealed an overall content of46 wt.% SiO2, 9 wt.% SnO2, 31 wt.% PbO, 4wt.% K2O and 2 wt.% Na2O (Maggetti et al.2009a). Chlorine (1 wt.%) indicates the addi-tion of salt, a common practice in this manufac-ture. The relatively high alumina content indica-tes the addition of a clay phase to the aqueousglaze suspension, which was totally dissolvedin the tin-glaze during the firing. However, it can-not be ruled out that some alumina present inthe glaze results from diffusion processesoccurring from the body into the glaze (Tite etal. 1998, Molera et al. 2001).

The tin-opacified lead-alkali glaze shows veryfew bubbles and relictic quartz grains, as wellas newly crystallized phases, such as potash

feldspars and a SiO2-polymorph, probably cris-tobalite acording to its shape (Fig. 1).Cassiterite particles are not homogeneouslydispersed, but form small SnO2-clusters. Theglazes of Granges-le-Bourg have similar chemi-cal compositions with SnO2 around 9.1 wt.%(Maggetti et al. 2009d).

4.3.5. Decorating

Painters decorated the pieces freehand, basingthemselves on illustrations, often engravings,or by using the technique of stencilling (Bastian2002, p. 99). The use of stencils, generallymade of metal, became widespread only in the19th century (Rosen 1995, p. 41). The drawing,first made on paper and pierced through by aneedle, was rubbed with a stamp containing apowder which disappeared during the firing,such as wood charcoal, and made it possible,by transfer, to obtain a preliminary outlinewhich would guide the painter's hand. AsRosen (1995, p. 41-42) notes: "Only a few pie-ces received elaborate designs. Most of themonly bore elementary patterns executed rapidlyby workers who were not always qualified pain-ters. In that case, they could place the objecton the girelle of the wheel, and traced with anunmoving hand the outlines which they thenfilled in with different patterns."

The printed design, a process started inEngland towards the middle of the 18th century,was not used very much for tin-glaze faience,even though it appeared early on in France onporcelain (ca. 1759, Préaud & d’Albis 2003, p.346) and white earthenware or terre de pipe(Garric 2006, p. 16-17).

4.3.6. Ceramic colours

French ceramic colours of the 17th and 18th

century were, with few exceptions, colouredglass, finely crushed, in order to make it possi-ble to draw lines, even very fine ones. Theircolour is due to:

(1) The addition of colouring metallic oxidesto a colourless glass melt;

(2) The incorporation of natural or syntheticcrystals in a colourless or coloured glassmelt;


Fig. 14. Glazing and touching up. Two workers are dipping biscuit pla-tes in the liquid containing the glaze powder. In the background, twofemale workers are touching up the raw glazed objects. The one infront is removing the glaze powder from the bottom of a plate with abrush or a piece of felt, so that it will not adhere to the support duringfiring, the other is removing with a blade the excess thickness of theglaze or their drops. (Brongniart, 1844, Plate XLVI).

Fig. 15. Faience biscuits covered with raw white tin glaze, drying in thesun. Grottaglie, Italy. Photo M. Maggetti.

52

3) A mixing of milled, coloured crystals with acolourless or coloured glass powder.

Thus we can divide ceramic colours into twolarge categories, transparent colours andopaque colours. D'Albis (2003, p. 162) cha-racterizes them as follows: "If a colouringoxide is melted into a colourless substanceand has passed on its colour to it, the colouris then transparent. It is said that the colouris "in solution" (in the glass). If on the otherhand, the colouring oxide or the blend ofcolouring oxides is merely mixed in its powderstate to the colourless substance and if thelatter, then called flux, only agglomerates thecolouring grains between themselves and onthe porcelain during the firing, without as arule becoming coloured itself, then the colouris opaque. The colouring grains themselvesare responsible for this. It is then said thatthe colour is "in suspension" (in the glass)”.

One of the methods for preparing transparentcolours consisted in putting transparentglass powder in a crucible, adding a low per-centage of a colouring oxide, melting thewhole in a kiln, crushing the newly formedglass, reducing to a powder in the colour mill,washing it and filtering it. This process of frit-ting-crushing-washing-filtering was usuallyrepeated several times.

For opaque colours, one could either usenaturally refractory substances, with low solu-

bility in a glass such as the Armenian bole orsynthetic refractory products prepared on thespot or that could be purchased. They weremixed in a crucible with a transparent orcoloured crushed glass, then placed in a kilnand melted. After cooling, the procedure wasthe same as for transparent colours.

Brongniar t (1844, II, p. 505-688) devotes alarge part of his treatise to the techniqueand the preparation of ceramic pigmentsand colouring agents, which shows howimportant he felt this technological aspectwas. In the 18th and at the beginning of the19th centuries, each workshop normallyprepared its own colours and kept its secretvery closely. The archive sources and thetreatises published reveal countless reci-pes and numerous trials, whether success-ful or unsuccessful. These essays arebrought for th by archaeological excavationsas test pieces, that is to say small tumblersor ceramic objects (Fig. 16).

It is easy to understand the reticence therewas to reveal these recipes because thecolours, these transparent or opaque vitre-ous powders, the tin glaze and the biscuit, allthree submitted to the same second or thirdfiring temperatures, must have had thermaldilation coefficients which were comparableamong themselves. Moreover, if the objectswere used for culinary purposes, there wasthe need to find on-glaze colours with a che-


Fig. 16. Test pieces from the faience manufacture Le Bois d'Épense / Les Islettes manufacture (pers. coll.). Width of the clay balls 2-3 cm. PhotoM. Maggetti.

53

mical composition which would resist liquidssuch as vinegar, tea or coffee and with a highresistance to abrasion to avoid being scrat-ched by metallic objects such as knives orforks. All these requirements were not easyto bring together and one understands thateach factory needed a lot of time, energy andfinancial backing to put the finishing touchesto their recipes, using local and foreignresources to prepare the paste, the tin glazeand the colours, while at the same time kee-ping in mind their specific physicochemicalproperties.

4.3.7. In-glaze painting

In this decoration technique (Bastian 2002,p. 95 ff.), painters worked directly onto thedry tin glaze powder of the biscuitted pieces,which was very fragile (Fig. 17).

It took someone with a steady and well-trai-ned hand, because it was almost impossibleto repair an erratic brushstroke. When correc-tions were required the faulty part had to becarefully scraped off and the glaze includingthe colour pigments reapplied.

Considering the high temperature and thelong time of the second (glaze) firing, in-glazecolours become closely fixed to the tin glaze.But very few colours resist at such high tem-peratures, the palette contains but fewcolours: blue, green, brown to violet, yellow,black, red and white. The first three are trans-parent and their hue is due to the "diluted"presence of cobalt in the case of blue, copperfor green and manganese for brown to violet.

Yellow (Normally Pb2Sb2O7) is a very ancientopaque colouring agent used by theEgyptians as early as 1600 BC (Clark et al.1995). Its synthesis is a result of a dry mix-ture of powdered antimony and lead oxides,obtained by oxidizing antimony and lead sul-phides (Stibnite Sb2S3 and galena PbS), thewhole melted in a crucible at temperaturesaround 900°C (Shor tland 2002). Theseyellow crystals, finely crushed, are thenmixed to powdered colourless, transparentglass, or coloured glass, if the painter wantedother shades and tones for the desiredcolour. Black is often obtained through man-ganese and iron oxide crystals, either alone,or mixed. Red, a delicate colour to obtainbecause it is not very stable at high tempera-tures such as those of a second firing andturns into drab brown, is therefore little usedin faience. What were used instead were cal-cined ferruginous sands such as Armenianbole or, later on, Thiviers red (Rosen 1995,Rosen et al. 2007). White is a colour contai-ning more tin oxide as the tin glaze, used forexample in the technique called Bianco sopraBianco.

The powdered ceramic in-glaze colours weremixed with water and a binder, for examplestarch, in small bowls. During glaze firing thepigment particles were coated by a thin she-ath of molten glaze.

In-glaze colours of the manufacture Le Boisd’Épense were studied by Maggetti et al.(2009 a, b). Electron microscope examina-tion showed that the blue drawings are inclu-sion-free glazes with wavy boundaries to theunderlying tin-glaze (Fig. 1a). They are lead-alkali glazes with a small amount of dissolvedcobalt (0.2-0.8 wt.%). NaCl was used to makethe blue glass, as evidenced by the presenceof chlorine. Significant arsenic concentrationsclassify these blue pigments (smalts) intogroup 4 of Gratuze et al. (1997), derivingfrom cobalt ores of the Saxonian Schneeberg(Germany). Purple colours contain minoramounts of cassiterite and relictic Mn-Fe-crystals. Manganese and iron contents inthese transparent colours lie between 2.5and 4.6 wt.% respectively.

The black colour is an association of very tiny


Fig. 17. In-glaze painting on tiles at the manufacture SchreiberKeramik AG, Matzingen, Switzerland. Photo M. Maggetti.

54

and rounded particles. Elemental mappingsshow that the latter consist of distinct iron-, man-ganese- or iron + manganese-rich pigments.

The rare opaque red colour is a complex mix-ture of: (1) very rare cassiterites; (2) rare,large corroded potash feldspars; (3) large (upto 10 µm “crumbly” pigments (fired yellow orred ochre); and (4) small (approximately 1µm) rectangular or square particles (Pb felds-pars?) (Molera et al. 1993, Fortina et al.2005).

Early 19th century antimony-based opaqueceramic colours are yellow, tawny and green(Maggetti et al. 2009b). The first is generatedby lead antimonate crystals (Pb2Sb2O7),which are incorporated into an uncolouredglassmatrix (Fig. 1b). According to SEM-EDSmeasurements, these pigments contain iron.The tawny colour is the optical result of thecombined presence of similar yellow, iron-bearing lead antimonate particles in a Fe-rich, brownish glass matrix. The green opa-que colour is produced by the combination ofa blue cobalt glass and yellow Pb-Sn-Fe-anti-monate crystals. Relictic cores of zoned pig-ments lighten the recipes, according to whichthe pigments were produced. First, they weresynthesised by calcination, ground and thenmixed with a colourless, brown or blue glasspowder. The resulting powder mixture wasadded to a liquid agent and used as high tem-perature ceramic colour.

Analytical studies of red decorations fromfour production centers (Argonnes, Nevers,Rouen and Thiviers) showed a complex situa-tion (Marco de Lucas et al. 2006, Rosen etal. 2007). Thiviers used a particular, locallyavailable reddish sandstone called Grès deThiviers or Thiviers Red. This rock is compo-sed of !-quartz grains cemented by goethite(!-FeOOH). During firing, the latter transformsto red hematite (!-Fe2O3). Raman spectra ofdifferent Nevers red hues revealed that the18th century pigments were a mixture of leadantimonate Pb2Sb2O7 with iron oxide andthat Thiviers Red was used for the mid 19th

century faiences. The manufactories ofRouen and Le Bois d’Épense/Les Islettesdidn’t employ Thiviers Red, but Armenianbole for the former and lead animonate with

an iron-rich glaze matrix for the latter.

4.3.8. The faience kiln

The kiln is an essential structure in a faienceworkshop. At the time under study here,wood-burning kilns were used, intermittent,with a vertical draught and direct flame (hori-zontal semi-cylindrical kilns). The kilns inFrench faience workshops in the 18th and19th centuries were made of refractory bricks(Rosen 1995, p. 47). In the Middle Ages,kilns had only one laboratory in which green,leather-hard objects for the first firing and rawglazed biscuits for the second firing were pla-ced simultaneously. The latter were stuckedin saggars. These kilns can still be foundtoday (Caiger-Smith 1973, Amigues 2002).French kilns of the 18th and 19th centuriescould have two floors or laboratories, separa-ted by a perforated platform to let hot airthrough (Fig. 18).

The bigger was used for the second firing ofraw glazed pieces and was situated in thelower part of the kiln. The smaller, named"globe" or first firing room was for biscuitfiring. There, the green leather-hard objectwas thoroughly dried and became ceramised,that is, it was transformed into a hard object—the biscuit— and its shape was thus con-solidated. The unfired pieces were normallysimply piled up "in charge" in the top labora-tory where the temperature was not so high,


Fig. 18. Schematic aspect of a faience kiln with its fireplace, the soleand the laboratory for second firing. In the latter two different methodsof placing the wares are shown, that is with saggars or rafters and thebiscuit laboratory upon it. Diderot (1756, plate IX).

55

and fired at the same time as the raw glazedbiscuitted pieces placed in the lower labora-tory. When these objects were placed in thekiln it had to be in such a way as to guaran-tee they would not touch to avoid becomingstuck together with the molten glaze. It wasalso necessary that the pieces be kept awayfrom the flames and substances resultingfrom the combustion, such as ashes andsmokes, or from projections, to avoid thedeterioration of glaze and decoration.

To this purpose, potters used very specificrefractory materials when placing their wares inthe kiln, by using two techniques: rafters, alsocalled en chapelle (Fig. 18, to the left of thesecond firing laboratory) and saggars, alsocalled encastage (Fig. 18, to the right of thesecond firing laboratory). The first consists inplacing the objects to be fired on rafters or onslabs separated by pegs or reels (Fig. 19).These scaffoldings were wedged with bits ofclay of different shapes, named accots.

In the second technique, the pieces were pla-ced, according to their shape, in box-like contai-ners either round, or oval, or rectangular, calledsaggars (Fig. 20) where they rested on triangu-lar supports, the saggar-pins (they can alsohave other shapes, named cockspurs or tri-vets). The latter had to be as sharp-edged aspossible so as to leave only a very small tracewhen in contact with the fired object. The insi-de of the saggars was usually glazed to protect

as much as possible the objects to be firedfrom the projection of particles. The saggarswere piled up in the kiln, isolated from gasesand flames by round tiles pierced with a holeand having all cracks filled in with clay putty.

These technical or kiln ceramics which came indirect contact with the flames and the atmos-phere of the kiln, that is the saggars, the raf-ters used to place the objects in the kiln andthe reels had to be resistant at high temperatu-res, go through as many firing cycles as possi-ble, and bear the weight of the pile up.Refractory mixtures with a high mechanicalresistance had to be made. Each manufacturehad its own recipe. Non-calcareous clays werecommonly used, to which were added a temperin the form of coarse sand and crushed cera-mics (grog).

There is only one archaeometric study of kiln fur-nitures from a French faience workshop.Maggetti et al. (2009d) studied thirty-nine sam-ples (firing plates, saggars, spacers, props andaccots/wads) from the manufacture of Granges-le-Bourg. The kiln furniture is chemically inhomo-geneous. Firing plates and saggars belong to aCaO- + MgO-poor group, well suited to supporthigh firing temperatures as well as several firingcycles. They correspond to decarbonatised toplayers of local dolomitic Triassic marls, temperedwith local quartz sands. Contrasting, props, spa-cers and wads show markedly higher CaO andMgO, and were made using local dolomiticmarls, which were probably not as well proces-sed as the faience paste.

The potters obviously employed two major


Fig. 19. Slabs, firing plates or shelves for placing wares in the kiln andcolumnar props or reels from the faience manufacture Le Bois d'Épen-se/Les Islettes (pers. coll.). Width of the slabs ca. 20 cm. Photo M.Maggetti.

Fig. 20. Saggars and triangular, nail-like saggar-pins or pegs from thefaience manufacture Le Bois d'Épense/Les Islettes (pers. coll.).Height of the saggars ca. 16 cm. Photo M. Maggetti.

56

recipes, based on refractory and non-refrac-tory clays. Ceramic objects with high fluxeswill melt around 1100ºC and are thereforenot very well suited to resist the firing tempe-ratures of a faience kiln or to support manyfiring cycles. The use of such clays for thesecond group of kiln furniture is thereforepuzzling and not yet understood. Plate andsaggars are covered inside with a tin glaze.For obvious financial reasons, the pottersused significantly less tin oxide (< 6 wt.%SnO2), added much more crushed quartz andapplied the watery glaze suspension with amuch thinner stroke as for the faience. Theabsence of any significant reaction zone atthe body/glaze interface indicates that theglaze suspension was applied on alreadyfired (biscuit fired) plates and saggars.

4.3.9. Second (glost or glaze) firing

Each ceramic firing is divided up into two pha-ses, the firing phase proper which includes therise in temperature up to the maturing tempe-rature of 950-1050ºC, and a second coolingphase in which the kiln is no longer stoked(Picon 2002). In the first, reducing atmosphe-res alternate with oxidizing ones while in thecase of faience, the second phase is throug-hout oxidizing because air is allowed to pene-trate inside the kiln. A vivid description of sucha faience firing is given by Amigues (2002).According to Bastenaire-Daudenart (1828, p.390-391) a French kiln would continuously firefor 25-27 hours, with a first slow fire stage (5-6 h), a second more sustained fire stage (5h)and a long sustained fire stage at maximumtemperatures (15-16h).

The firing, the crowning step in the faience-making process, is not an easy matter, butfull of risks, because the distribution of theheat inside a kiln is often very uneven, asmuch as several hundred degrees differencedepending on the specific spot in the kiln(Flame "T up to 550ºC, Wolf 2002), yieldingmisfired pieces, either under or over fired, orbecause the piled up objects crumbled as aresult of a poor saggar recipe or throughexcessive heat. Therefore, one must not besurprised at the large quantity of waste foundduring workshop excavations. Biscuit wasterscould be used as testing material (drawing,

painting, glazing), while the faience of lesserquality were sold at lower prices or incorpora-ted into building materials (Amigues 2002, p.197). Even the very worst quality was not lostbecause it was used to fill in ditches or tolevel the ground. The end products of theglost firing were white faience tableware (Fig.6A) or in-glaze decorated pottery (Fig. 6B).

Experimental firings in kilns with one firingchamber, built and functioning according toancient techniques, have given reliable tem-perature-time curves (see Amigues 2002;Wolf 2002). On the other hand, no experimen-tal firing was carried out in a two-chamberkiln. It can be guessed that the temperature-time curve of the second firing is quite simi-lar to that of a one-laboratory kiln, but preci-se experimental data for the temperatures ofthe biscuit laboratory are lacking. "The firingis in two parts, first the objects are fired asbiscuit at a temperature ranging from cherryred to whitish red, and then with the glaze ata slightly higher temperature." (Brongniart1844, 1877 tome II, p. 20).

Inferred second firing temperatures in thekilns from Le Bois d’Épense/Les Islettes andGranges-le-Bourg were, according to X-ray dif-fraction analyses, between 950 and 1050ºC(Maggetti 2007b, Maggetti et al. 2009c).These results are in good agreement with themicrostructure of the well sintered bodieswith voids due to the dissociation of carbona-tes (Fig. 21) and tiny calcsilicate phases(diopside, gehlenite, plagioclase) crystallizedduring firing (Fig. 22).


Fig. 21. Backscattered electron images of the body of a faience fromGranges-le-Bourg. D = void, former dolomitic grain, F = K-feldspar, Q =quartz. Photo M. Maggetti.

57

It has often been observed that the firing trig-gers the building up of a reaction zone betwe-en body and glaze, called interface body-glazeor body-glaze layer. A detailed discussion ofthis phenomenon can be found in technicalhandbooks (Munier 1957, Hamer & Hamer2004). With progressing firing, glaze melt willsoak into the porous biscuit. The fluxes of theglaze (and the body) will selectively attackand dissolve refractory particles of the body,especially clay minerals. This is the reasonfor the Al-richness of the reaction zone. Oncooling, Pb-feldspar will therefore crystallizein the body-glaze layer (Molera et al. 1999).The thickness of this interface depends onseveral factors (Molera et al. 2001, BenAmara 2002):

(1) Mineralogical nature of the clay;(2) CaO-content of the body;(3) Na-content of the glaze; (4) Green, leather-hard body or biscuitted

body;(5) Firing conditions (T max., soaking time,

cooling time).

Modern industrial applications favour thedevelopment of such an interface as it pre-vent to a certain degree crazing and shive-ring. In the glaze, wollastonite and diopsidecrystals can be concentrated in a small layerclose to the interface body-glaze (Mason &Tite 1997, Molera et al. 2001, Bobin et al.2003, Fortina et al. 2005, Zucchiatti et al.2006). Body-lead glaze reactions (up to 300-400 µm from the body/glaze contact into theglaze) are documented by conspicuous ele-

mental diffusion profiles, for instance for alu-mina and lead oxides (Molera et al. 2001).These phenomena must be taken intoaccount when measuring the chemical com-position of a glaze.

4.3.10. On-glaze painting

This technique was introduced into theFrench faience community around 1740 byPaul Hannong of Strasbourg (Blondel 2001,p. 261). For the on-glaze painting, a binderwas added to the powdered colour such as,for example, natural gum or turpentine andturpentine oil (Bastian 1986, p.134, Bastian2003, p. 106) so as to obtain a more or lessviscous mixture. It was possible to use on theone hand powders with pure tints and on theother hand blends, for example blue andgreen, which gave different colours. Duringthe firing process, the organic substancesevaporated, the powder melted and boundwith the mollified tin glaze. The applying ofcolours on the glost fired tin glazed objects,in-glaze decorated (Fig. 6D) or not (Fig. 6C),needed a third firing intended to fix the pat-tern, in a special reverberatory furnace, alsoknown as muffle furnace. Temperatures inthis furnace used to be lower (estimated 400to 850°C, Rada 1989, p. 83, Endres et al.1991, p. 42, De Plinval de Guillebon 1995, p.20, Peiffer 2000, p. 97) than those of thesecond high temperature firing (inferred 950-1050°C) and allowed for a much larger colourpalette to be used.

In rare cases, for the same object, artistsused the mixed technique (Fig. 6D). To dothis, they applied in a first step, certain in-glaze colours on the unfired tin glaze, and ina second step, after the glost firing, the on-glaze colours which were then fired in themuffle furnace. If the thermal dilatation coef-ficients or the fixation temperatures of the on-glaze colours were too far apart, it was pos-sible to fire each colour individually, whichmeant placing the object several times in themuffle furnace. Obviously, these techniquesentailed a higher cost.

In order to obtain special effects, the paintercan use the luster technique which requires afourth firing under reducing conditions. This


Fig. 22. Closer view of a well fired faience body from Granges-le-Bourg.The tiny crystals around the quartzes and in the glassy matrix are calc-silicates (diopside, gehlenite, plagioclase). Photo M. Maggetti.

58

technique was not used in France in theperiod studied here.

5. Provenancing of French faiences: twocase studies

Attributions of ceramics to a pottery works-hop must be based on so-called chemicalreference groups. Each reference group com-prises a minimum of 20-30 chemical analy-ses of unequivocally local products (biscuits,wasters) from a specific manufacture. Up todate, there are only four reference groups ofFrench faience workshops which fulfil allrequisites, i. e. (1) enough chemical analy-ses; and (2) all analyses published. Theseare Granges-le-Bourg (Maggetti et al. 2009c),Montpellier (Rosen et al. 2009), Moustiers(Rosen et al. 2009) and Nevers (Rosen2009). Three new reference groups (Le Boisd’Épense, Lunéville and Saint-Clément) willbe published very soon.

5.1. 17th century faiences: Nevers or “LeCroisic”?

France had only few faience manufactures inthe 17th century (Fig. 4). A characteristic pro-duct painted in the compendiario style (Fig.23) is attributed by many specialists withoutany real proofs to the “Le Croisic” workshop(Vince 1982).


Fig. 23. Faience plate attributed to the “Le Croisic” manufacture(Vince 1982). Photo J. Rosen.

Fig. 24. Result of the cluster analysis of the 17th century Nevers refe-rence group (Rosen, 2009) and the “Le Croisic” faience piece of Fig.21. Redrawn and simplified from Rosen et al. (2009).

Fig. 25. Factor analysis of the 18th century Nevers reference group(blue rectangles), genuine La Rochelle faiences (from the excavationsof the La Rochelle manufacture, red dots) and faiences attributed toLa Rochelle, found in the waste dumps of the Nevers manufactures(green rectangles). Redrawn from Rosen et al. (2009).

59

The authors situated this hypotetical works-hop at the estuary of the river Loire in theAtlantic. Using the reference group of 17th

century faiences from Nevers (Rosen 2009)it could be shown that this sample of so-called typical “Le Croisic” faience from theMusée du château des Ducs de Bretagne inNantes (French Brittany) match the chemicalcomposition of the Nevers faiences (Fig. 24,Rosen et al. 2009). It was therefore made inNevers and shipped on the navigable Loire toNantes and the Atlantic. Faiences fromNevers eventually reached Canada and werefound in excavations in Quebec.

5.2. 18th century faiences: Nevers or LaRochelle?

A comparative factor analysis using aNevers 18th centur y reference groupagainst a reference group from the manu-facture of La Rochelle, a harbour on theWest coast of France, clearly distinguishesboth (Fig. 25). A whole set of faience pie-ces with so-called “La Rochelle” decora-tions (Morin 1990, p. 104), found in thewaste dumps of Nevers manufactures,definitely belong to the Nevers referencegroup (Rosen 2007b, Rosen et al. 2009).

6. References

Amigues, F. (2002) Technique de fabrication de lacéramique valencienne. Le calife, le prince et lepotier, Les faïences à reflets métalliques,Catalogue de l’exposition au Musée des BeauxArts de Lyon (2.3.-22.5.02), Musée des Beaux Artsde Lyon et Réunion des musées nationaux, Paris.180-213.

Anonymous (1783) Fayencerie (Art de la), dansEncyclopédie méthodique. Ar ts et MétiersMécaniques, Dédiés Et Présentés à Monsieur LeNoir, Conseiller D’Etat, Lieutenant Général DePolice, &, (ed. Panckoucke, C. J. & Plombteux, C.),tome deuxième. 506-528.

Bastenaire-Daudenart, F. (1828) L’art de fabriquerla faïence blanche recouverte d’un émail opaque.Paris.

Bastian, J. (1986) Les Hannong: étude des décorspeints sur les faïences et porcelaines à Strasbourg

et Haguenau (1721-1784). PhD dissertation,University of Strasbourg, France.

Bastian, J. (2002-2003) Strasbourg. Faïences etporcelaines 1721-1784. 2 tomes, EditionsM.A.J.B., Strasbourg.

Ben Amara, A. (2002) Interaction glaçure plombifère-terre cuite: recherches sur le mode d’application dumélange glaçurant. PhD dissertation, UniversitéMichel de Montaigne Bordeaux 3, France. 259-284.

Bernier, M. (2003) Caractérisation typologique,microscopique et chimique des faiences du XVIIIesiècle du site Saint-Ignace de Loyola en Guyanefrançaise. Cahiers d’archéologie du CELAT no. 14,série archéométrique no. 4, Université Laval,Canada.

Blondel, N. (2001) Céramique vocabulaire techni-que, Centre des monuments nationaux, Monum,Editions du patrimoine, Paris.

Bobin, O., Schvoerer, M., Ney, C., Rammah, M.,Daoulatli, A., Pannequin, B. & Gayraud, R. P.(2003) Where did the lustre tiles of the Sidi Oqbamosque (AD 836-863) in Kairouan come from?Archaeometry, 45, 4, 569-577.

Bosc d’Antic, M. (1780) Observations sur l’Art dela Faiencerie, Œuvres de M. Bosc d’Antic, Docteuren médecine... Contenant plusieurs Mémoires surl'art de la Verrerie, sur la Faïencerie, la Poterie,l'art des Forges, la Minéralogie, l'Electricité, & surla Médecine. Tome Premier, Paris, 258-283.

Bouquillon, A. (2000) Observations sur la techni-que et la composition des glaçures de quelquescarreaux de pavement Renaissance provenant deBrou, Longecourt-en-Plaine, Thouars, Langres,Polisy et Rouen, In: Rosen, J. et Crépin-Leblond,T. (dir.), Images du pouvoir – Pavements de faïen-ce en France du XIIIe au XVIIe siècle, Musée deBrou et Réunion des musées nationaux, Brouand Paris. 185-187.

Bourgarel, G. (2007) La production. Les Formes.In: Maggetti, M. (dir.), la faïence de Fribourg. Faton,Dijon, France, 14-31.

Boussemart, J., de (1786) Les procédés de fabri-cation de la faïence. Manuscript published in:Dansaert, G., 1876, Les Anciennes faïences de


60

Bruxelles, facsimile 1922, Van Oest et Cie.,Bruxelles. 130-150.

Boyer, M. (1827) Manuel du porcelainier, du faïen-cier et du potier de terre; suivi de l’art de fabriquerles terres anglaises et de pipe, ainsi que les poê-les, les pipes, les carreaux, les briques et les tui-les, 2 volumes, Roret, Paris. II. 83-128.

Brongniart, A. (1844) Traité des Arts Céramiquesou des Poteries considérées dans leur Histoire,leur Pratique et leur Théorie, 3 vol. et 1 atlas.Paris, 1844; 2è éd. revue, corrigée et augmentéede notes et d’additions par Alphonse Salvetat,1854; 3è édition avec notes et additions parAlphonse Salvetat, 1877; Facsimile 1977 of thethird edition 1877 by Dessain & Tolra, Paris.Caiger-Smith, A. (1973) Tin-Glaze Pottery in Europeand the Islamic World, Faber & Faber, London.

Caiger-Smith, A. (1985) Lustre Pottery Technique,Tradition and Innovation in Islam and the WesternWorld. London.

Carette, M. & Deroeux, D. (1985) Carreaux depavement médiévaux de Flandre et d’Artois (XIIIe-XIVe s.) , Mémoires de la Commission départemen-tale d’Histoire et d’Archéologie du Pas-de-Calais, t.XXII-1.

Casadio, R. (1998) Note techniche sulla fabbrica-zione di vasellame a Faenza nel XVI secolo.Guarnieri, C. (éd.), Forni e fornaciai a Faenza nelXVI secolo, Faenza, 127-139.

Clark, R. J. H., Cridland, L., Kariuki, B. M., Harris,K. D. M., & Withnall, R. (1995) Synthesis,Structural Characterization and RamanSpectroscopy of the Inorganic Pigments Lead TinYellow Types I and II and Lead Antimonate Yellow:Their Identification on Medieval Paintings andManuscripts. Journal Chemical Society DaltonTransactions, 2577-2582.

Colomban, P. & Truong, C. (2004) Non-destruc-tive Raman study of the glazing technique in lus-tre potteries and faience (9-14th centuries): sil-ver ions, nanoclusters, microstructure and pro-cessing. Journal of Raman Spectroscopy, 35,195-207.

Cultrone, G., Rodríguez-Navarro, C., Sebastian, E.,

Cazalla, O. & de la Torre, M. J. (2001). Carbonateand silicate phase reactions during ceramic firing.Eur. J. Miner. 13, 621-634.

D’Albis, A. (2003) Traité de la porcelaine deSèvres, Faton, Dijon, France.Deck, Th. (1887) La faïence, Quantin, Paris. Newedition 1903, Picard & Kaan, Paris, France.

De la Hubaudière, C., & Soudée Lacombe, Ch.(2007) “L’art de la fayence” des Caussy, faïenciersà Rouen et Quimper au XVIIIe siècle. Lilou,Quimper, France.

De Plinval de Guillebon, R. (1995) Faïence et por-celaine de Paris XVIIIe-XIXe siècles. Faton, Dijon,France.

Démians d’Archimbaud, G. & Picon, M. (1972) Lescéramiques médiévales en France méditerranéen-ne. Recherches archéologiques et de laboratoire.Colloques Internationaux CNRS, la céramiquemédiévale en Méditerranée occidentale, 15-42.

Diderot, D. (1756) Fayence. In: Encyclopédie OuDictionnaire Raisonné Des Sciences, Des Arts EtDes Métiers, Par Une Société De Gens De Lettres(eds. Diderot, D. & d’Alembert, J.), Briasson, Paris,1751-1780, vol. 6 (Et-Fn), 454-460.

Diminuoco, P., Messiga, B. & Riccardi, M. P. (1998)Firing processes of natural clays. Some microstruc-tures and related phase compositions.Thermochimica Acta, 321, 185-190.

Dufournier, D. (1989) Céramologie, technologiecéramique. Actes du colloque de Lille, Travaux duGroupe de Recherches et d’Etudes sur laCéramique dans le Nord Pas-de-Calais, N° horssérie de Nord-Ouest Archéologie.

Dufournier, D. & Deroeux, D. (1986) Les carreauxdécorés du « groupe de l’Artois occidental, caracté-risation et diffusion ». Terres cuites architecturalesau Moyen Age. Colloque des 7-9 juin 1985, Muséede Saint-Omer, Mémories de la Commission dépar-tementale d’Histoire et d’Archéologie du Pas-de-Calais, T. XXII », 186-206.

Dufournier, D. & Derouex, D. (1987) Carreaux defaïence en Artois au XIIIe siècle. Nouvelles appro-ches. Bulletin de la Commission départementaled’Histoire et d’Archéologie du Pas-de-Calais, XII, 2,


61

141-154 and pl. XVII-XXI.

Dufournier, D., Thiron, D. & Bouquillon, A. (2004)Les carreaux de pavement faïencés du Pré-d’Auge(Calvados) (milieu XVIIe – début XVIIe siècle), TECH-NE, 2004, 20, 96-101.

Echallier, J.-C. & Mery, S. (1992). L’évolution miné-ralogique et physico-chimiques des pâtes calcairesau cours de la cuisso: experimentation en labora-toire et application archéologique. Documents etTravaux IGAL, 16, 87-120.

Endres, W., Krause, H.-J., Ritscher, B. & Weber, C.(1991) Reichenbacher Steingut. Grafenau,Germany.

Faÿ-Hallé, A. & Lahaussois, C. (2003) La faïenceeuropéenne au XVIIe siècle, le triomphe de Delft.Editions de la Réunion des musées nationaux,Paris, France.

For tina, C., Santagostinao Barbone, A. &Turbanti, Memmi, I. (2005) Sienese « archaic »majolica: a technological study of ceramicbodies and coatings. Archaeometry, 47, 3, 535-555.

Garric, J. M. (2006) Terres blanches, grès et faïen-ces fines 1750-1850. Collections de l’abbaye deBelleperche, Conseil Général de Tarn-et-Garonne,Catalogue d’exposition de l’abbaye de Belleperche15 mai – 30 septembre 2006, Abbaye deBelleperche, France.

Gratuze, B., Soulier, I., Blet M., & Vallauri, L.(1997). De l’origine du cobalt: du verre à la cérami-que. Revue d’Archéométrie, 20, 77-94.

Hamer, F., & Hamer J. (2004) The Potter’sDictionary of Materials and Techniques, 5th edition,A & C Black, London and University of PennsylvaniaPress, Philadelphia.

Harlé, S. H. (1831) Mémoire sur la fabrication dela faïence à Nevers. Bibliothèque de l’Ecole desMines, Paris.

Heimann, R. (1982) Firing technologies and theirpossible assessment by modern analyticalmethods. In: Olin, J. S., Franklin, J. D. (eds.)Archaeological Ceramics. Smithsonian Institution,Washington D. C., 89 – 96.

Jamieson, R. W. & Hancock, R. G. V. (2005).Neutron Activation Analysis of Colonial CeramicsFrom Southern Highland Ecuador. Archaeometry,46, 4, 569-583.

Liénard, F. (1877) Les faïenceries de l’Argonne.Mémoires de la Société Philomatique de Verdun,VIII, 2, 111-224.

Maggetti, M. (1982) Phase analysis and its signifi-cance for technology and origin. In: Olin, J. S.,Franklin, A.D. (eds.), Archaeological Ceramics.Smithsonian Institution Press, Washington D. C.,121-133.

Maggetti, M. (2001). Chemical Analyses of AncientCeramics: What for? Chimia 55, 923-930.

Maggetti, M. (2007a) Technique de la faïence fran-çaise (Fin XVIIIè/Début XIXé siècle). In: Maggetti,M. (dir.), la faïence de Fribourg. Faton, Dijon,France, 14-31.

Maggetti, M. (2007b) Analyses scientifiques descéramiques de la manufacture du Bois d’Épensedite « des Islettes », In: Rosen, J. (dir.), la faïence-rie du bois d’Épenses dite « des Islettes », unemanufacture à l’étude. Edition Ville de Bar-le-Duc,Bar-le-Duc, France, 44-55.

Maggetti, M., Westley, H. & Olin, J. (1984)Provenance and Technical Studies of MexicanMajolica Using Elemental and Phase Analysis. - In:J.B. Lambert (ed), ACS Advances in ChemistrySeries, No. 205, Archaeological Chemistry III,American Chemical Society, 151-191.

Maggetti, M., Rosen, J., & Neururer, C. (2009a)Grand feu colours used in the faience manufactureLe Bois d’Épenses/Les Islettes (North-EasternFrance, 18/19th centuries). In: Moreau, J.-F.,Auger, R., Chabot, J. & Herzog, A. (dir), ProceedingsActes ISA 2006, 36th International Symposium onArchaeometry, 2-6 May 2006, Quebec City,Canada, 307-316.

Maggetti, M., Neururer, Ch., & Rosen, J. (2009b)Antimonate opaque glaze colours from the faiencemanufacture of Le Bois d’Épense (19th century,Northeastern France). Archaeometry, 51, 5, 2009,791-807.

Maggetti, M., Morin, D., & Serneels, V. (2009c)


62

High-Mg faiences from Granges-le-Bourg (HauteSaône, France). In: Biro, K. T., Szilagyi, V., Kreiter,A. (eds), Proceedings of the conférence EMAC’07,9th European Meeting on Ancient Ceramics, 24-27October 2007, Budapest, Vessels: Inside and out-side, 207-216.

Maggetti, M., Morin, D., Serneels, V., & Neururer,C. (2009d) Contrasting recipes for the kiln furnitu-res of the faience manufacture Granges-le-Bourg(Haute Saône, France). Studia Universitatis Babes-Bolyai, Geologia, 54, 2, 5-8.

Maniatis, Y. & Tite, M.S. (1975) A scanning elec-tron microscopic examination of the bloating offired clay. Trans. Brit. Ceram. Soc., 74, 229-232.

Marchesi, H., Thiriot, J. & Vallauri, L. (dir.) (1997)Marseille, les ateliers de potiers du XIIIe s. et lequartier Sainte-Barbe (Ve-XVIIe s.), Documentsd’archéologie française, no 65, Maison desSciences de l’Homme, Paris.

Marco de Lucas, M. C., Moncada, F. & Rosen, J.(2006) Micro-Raman study of red decorations inFrench faiences of the 18th and 19th centuries. J.Raman Spectroscopy, 37, 1154-1159.

Mason, R. B. (2004) Shine Like the Sun. Lustre-Painted and Associated Pottery from the MedievalMiddle East. Mazda Publishers and Royal OntarioMuseum, Toronto, Canada.

Mason, R. B. & Tite, M. S. (1997) The beginningsof tin-opacification of the pottery glazes.Archaeometry, 39, 1, 41-58.

Meunier, F. & Bouquillon, A. (2004) Les cérami-ques du château de Madrid au bois de Boulogne,Leur histoire. TECHNÈ, 20, 25-27.

Molera, J., Pradell, T., Martinez-Manent, S. &Vendrell-Saz, M., (1993) The growth of sanidinecrystals in the lead glazes of Hispano-Moresquepottery. Applied Clay Science, 7, 483-491.

Molera, J., Pradell, T., Salvadò, N. & Vendrell-Saz,M. (1999) Evidence of tin oxide recrystallization inopacified lead glazes. Journal of American CeramicSociety, 82, 2871-5.

Molera, J., Pradell T., Salvado, N. & Vendrell-Saz,M. (2001) Interactions between clay bodies and

lead glazes. J. Am. Ceram. Soc., 84, 5, 120-128.

Montagnon, G. (1987) Histoire des fayenciers deNevers et de leurs fabriques de 1585 à nos jours,Maison de la Culture, Nevers, France.

Morin F. (1990) Les faïenceries de Marans et de LaRochelle, Rupella, La Rochelle, France.

Munier, P. (1957) Technologie des faïences,Gauthier-Villars, Paris, France.

Oger, D., Bouquillon, A., Colinart, S, Elias, M.,Musculus, G., Pouthas, C. & Gratuze, B., (2002) Lafaïence de Tours (1840-1910). Avisseau et le souf-fle de la Renaissance. TECHNÈ, 16, 26-32.

Olin J. S., Harbottle, G. & Sayre E. V. (1978)Elemental Compositions of Spanish and Spanish-Colonial Majolica Ceramics in the Identification ofProvenience. In: Carter, G. F. (ed), ArchaeologicalChemistry II, Advances in Chemistry Series 171,Washington (D.C.): American Chemical Society,200-229

Olin, J., S., and Blackman, M. (1989)Compositional Classification of Mexican MajolicaCeramics of the Spanish Colonial Period. In: Allen,R. O. (ed), Archaeological Chemistry IV, Advancesin Chemistry Series 220, Washington (D.C.):American Chemical Society, 87-111.

Parmelee, C. W. (1948). Ceramic glazes. IndustrialPublication Inc, Chicago.

Peiffer, G.-J. (2000) L’art des céramiques. Une his-toire complète des techniques, Dessain et Tolra,Paris, France.

Pellet, C. (1993) La faïencerie d’Ar thé enBourgogne auxerroise. Le Bleu du Ciel, Vézelay,France.

Peters, T. & Jenni, J.P. (1973) MineralogischeUntersuchungen über das Brennverhalten vonZiegeltonen. Beiträge zur Geologie der Schweiz,Geotechnische Serie, Lieferung 50, Kümmerly undFrey, Bern, Switzerland.

Picon M (1993) Analyse des faïences en laboratoi-re: objectifs et difficultés. Actes du Colloque “Lescarreaux de faïence stannifère européens du XIXe

siècle”. Bulletin du GRECB, 15.


63

Picon, M. (2000) Observations sur la technique etla composition des pâtes de carreaux de pavementprincipalement bourguignonnes, XIVe-XVIIe siècles,In: Rosen, J. et Crépin-Leblond, T. (dir.), Images dupouvoir – Pavements de faïence en France du XIIIe

au XVIIe siècle, Musée de Brou et Réunion desmusées nationaux, Brou et Paris, 182-184.

Picon, M. (2002) Les modes de cuisson, les pâtes etles vernis de la Graufesenque: une mise au point. In: Genin, M. & Vernhet, A. (eds.), Céramiques de laGraufesenque et autres productions d’époque romai-ne/Nouvelles recherches/Hommages à BettinaHoffmann, Archéologie et Histoire romaine no 7,Montagnac, France, 139-164.

Picon, M. & Démians d'Archimbaud, G. (1978) Lesimportations de céramiques italiques en Provencemédiévale: état des questions. ColloquesInternationaux CNRS, la céramique médiévale enMéditerranée occidentale, 125-135.

Picon, M., Thiriot, J. & Vallauri, L. (1995)Techniques, évolutions et mutations. Le vert et lebrun, De Kairouan à Avignon, céramiques du Xè auXVè siècle, Marseille, France, 41-50.

Pradell, T., Molera, J., Smith, A. D. & Tite, M. S.(2008a) The invention of luster: Iraq 9th and 10th

centuries AD. J. Archaeol. Sci., 35, 1201-1215.

Pradell, T., Molera, J., Smith, A. D. & Tite, M. S.(2008b) Early Islamic luster from Egypt, Syria andIran (10th to 13th century AD). J. Archaeol. Sci.,35, 2649-2662.

Préaud, T. & d’Albis, A. (1991) La porcelaine deVincennes. Paris.

Rada, G. (1989) Les techniques de la céramique.Paris, France.

Rosen, J. (1995) La faïence en France du XIVè auXIXè siècle, Histoire et Technique, Editions erran-ce, Paris, France.

Rosen, J. (1997a) Analyses de faïences en laboratoi-re (Fluorescence X), méthodologie, premiers résul-tats, et limites. Salon international de la céramiquede collection et des arts du feu, Paris, 40-52.

Rosen, J. (1997b) X-ray fluorescence analyses offrench modern faiences (XVI-XIX AD): first results.

Key Engineering Materials Vols. 132-136, 1483-1486.

Rosen, J. (2000a) La faïence française du XIIIe auXVIIe siècle. Dossier de l’Art, no 70.

Rosen, J. (2000b) La manufacture de Meillonnas(Ain). Etude d’une fabrique de céramique régionale,1760-1870. Editions Monique mergoil, Montagnac,France.

Rosen, J. (2001) Faïenceries françaises du Grand-Est. Inventaire. Bourgogne-Champagne-Ardenne.XIVe – XIXe siècle. Comité des Travaux historiqueset scientifiques, Paris, France.

Rosen, J. (2007a) Catalogue et évolution des pro-ductions. In: Rosen, J. (dir.) La faïencerie du Boisd’Épense dite „des Islettes“, une manufacture àl’étude. Edition Ville de Bar-le-Duc, 102.

Rosen, J. (2007b) Ces faïences de Nevers qu’ondit « de la Rochelle ». Sèvres, Revue de la socié-té des amis du musée national de céramique,16, 31-45.

Rosen, J. (2009) La faïence de Nevers 1585-1900.Tome 1. Histoires et techniques. Faton, Dijon,France.

Rosen, J., del Carmen Marco de Lucas, M.,Moncada, F., Morin, A., & Ben Amara, A. (2007) Lerouge est mis: analyses des rouges dans la faïen-ce de “grand feu“ du XVIIIe et du XIXe siècle (avecle rouge de Thiviers). ArchéoSciences, 30, 95-107.

Rosen, J., Picon, M., & Maggetti, M. (2009)Revisiting the origin of some french faience throughchemical analysis. – In: Moreau, J.-F., Auger, R.,Chabot, J. & Herzog, A. (dir), Proceedings Actes ISA2006, 36th International Symposium onArchaeometry, 2-6 May 2006, Quebec City,Canada, 297-306.

Schmitt, A. (1990) Moustiers et Varages: quelquesproblèmes de composition chimique. Actes du Vecolloque national de céramologie, Paris 1 et 2décembre 1990. Paris, Editions Varai, France, 30-32.

Shortland, A. J. (2002) The use and origin of anti-monate colorants in early Egyptian glass.Archaeometry, 44, 4, 517-530.Shoval, S. (1988) Mineralogical changes upon hea-


64

ting calcitic and dolomitic marly rocks.Thermochimica Acta, 135, 243-252.

Soustiel, J. (1985) La céramique islamique.Fribourg: Office du Livre Fribourg, Suisse et Paris,Éditions Vilo.

Thornton, D. (1997) Maiolica Production inRenaissance Italy. In: Freestone I. & Gaimster D.(eds.), Pottery in the Making, World CeramicTraditions, London, 116-121.

Tite, M. S. (1988) Inter-Relationship BetweenChinese and Islamic Ceramics from 9th to 16thCentury A. D. In: Farquhar, R. M., Hancock, R. G. V.and Pavlish, L. A. (eds). Proceedings of the 26thInternational Archaeometry Symposium. UniversityPress, Toronto, Canada, 30-34.

Tite, M.S. (2009) The production technology ofItalian maiolica: a reassessment. J. Archaeol. Sci.,36, 2065-2080.

Tite, M. S. & Maniatis, Y. (1975) Scanning electronmicroscopy of fired calcareous clays. Trans. Brit.Ceram. Soc, 74, 1, 19-22.

Tite, M. S., Freestone, I., Mason, R., Molera, J.,Vendrell-Saz, M., & Wood, N. (1998) Lead glazes inAntiquity – Methods of production and reasons foruse. Archaeometry, 40, 2, 241-260.

Tite, M. S., Pradell, T., & Shortland, A. (2008)Discovery, production and use of tin-based opaci-fiers in glasses, enamels and glazes from the LateIron Age onwards: a reassessment. Archaeometry,50, 1, 67-84.

Trindade, M. J., Dias, M. I., Coroado, J. & Rocha, F.(2009) Mineralogical transformations of calcare-ous rich clays with firing: A comparative study bet-ween calcite and dolomite rich clays from Algarve,Portugal. Applied Clay Science, 42, 345-355.

Vallauri, L., Broecker, R. & Salvaire, M.C. (1978)Les productions de majoliques archaïques dans leBas-Rhône et le Roussillon. ColloquesInternationaux CNRS, la céramique médiévale enMéditerranée occidentale, 413-427.

Vallauri, L. & Leenhardt, M. (1997) Les produc-tions céramiques, In: Marchesi, H., Thiriot, J. &Vallauri, L. (dir.), Marseille, les ateliers de potiers

du XIIIe s. et le quartier Sainte-Barbe (Ve-XVIIe s.).Documents d’archéologie française, no 65,Maison des Sciences de l’Homme, Paris, France,166-174.

Vendrell, M., Molera, J. & Tite, M. S. (2000) OpticalProperties of Tin-opacified glazes. Archaeometry42, 2, 325-340.

Vince J. (1982) Faïences et poteries de Nantes etsa région. Musées du château des Ducs deBretagne, Nantes, France.

Watson, O. (1987) Islamic pots in Chinese style.Burlington Magazine, 129, 304-306.

Weiss, G. (1970) Ullstein Fayencebuch. Eine Kunst-und Technikgeschichte der Fayencen mitMarkenverzeichnis. Ullstein, Frankfurt/M, Berlin,Wien.

Wolf, S. (2002) Estimation of the production pra-meters of very large medieval bricks from St.Urban, Switzerland. Archaeometry, 44, 1, 37-65.

Zucchiatti, A., Bouquillon, A., Katona, I. &D’Alessandro, A. (2006) The “Della Robbia Blue”:a case study for the use of cobalt pigments in cera-mics during the Italian Renaissance.Archaeometry, 48, 1, 131-152.


65

Archaeometallurgy:the contribution of mineralogy/ Gilberto Artioli

Abstract

Modern mineralogy is a discipline that is intrinsically suited to face archaeometric problems, especiallyin the field of archaeometallurgy, which requires contributions from areas as diverse as geochemistry,petrology, materials science, metallurgy, archaeology, engineering, and many more. Arguably, it is showthat mineralogy may provide the necessary frame to put into use the information derived from diffe-rent sources, and combine it into a unifying interpretation. According to the mainstream of the metalproduction cycle, the most significant areas of investigation in archaeometallurgy are: (1) the charac-terization and identification of ore sources (the mining stage); (2) the reconstruction of the smeltingtechnologies for reduction of the metal (the smelting stage); (3) the interpretation of the metallurgicalmanufacturing processes in the production of the artefacts (the metallurgical stage); (4) the recons-truction of the use and diffusion of the metal objects (the physical lifetime of the object) and the incor-poration in the archaeological record and their preservation (the afterlife stage). Examples will be dis-cussed of the contribution of mineralogy to all steps of the archaeometallurgical cycle.

Resumen

La mineralogía moderna es una disciplina intrínsecamente adecuada para abordar los problemasarqueométricos, especialmente en el campo de la arqueometalurgia, donde se requieren aportes deáreas tan diversas como la geoquímica, petrología, ciencia de los materiales, arqueología, ingenieríay otras. Podríamos decir que la mineralogía puede proporcionar los fundamentos necesarios para usarla información obtenida de diferentes fuentes y combinarla para ofrecer una interpretación unificada.De acuerdo con la línea central del ciclo de producción de los metales, las áreas más destacadas enla investigación arqueometalúrgica son: (1) la caracterización e identificación de las materias primas(la fase de minería); (2) La reconstrucción de las tecnologías de fusión con reducción del metal (lafase de fundición); (3) la interpretación de los procesos de manufactura metalúrgica en la fabricaciónde artefactos (la fase metalúrgica); (4) la reconstrucción del uso y difusión de los objetos metálicos(el tiempo de vida físico del objeto) y su incorporación al registro arqueológico y su conservación (lafase postvida). Se discuten aquí distintos ejemplos de la contribución de la mineralogía a todas lasfases del ciclo arqueometalúrgico.

Key-words: Mineralogy, Metals, Archaeometallurgy, Archaeometry, Ore Minerals, Metallography, Texture,Provenancing, Geochemical Tracers.

1. Introduction

For a long time mineralogy as a discipline has been confined to the application of optical minera-logy and X-ray diffraction to the characterization of natural minerals, mainly within academic curri-cula in the Earth Sciences. In the last decades however modern mineralogy has contributed tomany technical developments in collaborations with physics and chemistry, and now it is a matu-re discipline that is naturally suited to serve as a bridge and interface between the traditionalEarth Science areas (petrology, geochemistry, structural geology, sedimentology, etc.) and themore materials science-oriented areas (physics, chemistry, engineering, metallurgy, etc.).

It is argued that mineralogy today may contribute deeply to cultural heritage studies in terms

Dipartimento di Geoscienze and Centro CIRCe, Università di Padova. Via Gradenigo 6, I-35131, Padova, Italye-mail: [email protected] URL: http://www.geoscienze.unipd.it/personal/artioli-gilberto

66

of thorough knowledge of materials and opti-mization of investigation techniques (Artioli2010a, Artioli and Angelini 2011). Being atruly interdisciplinary science, mineralogy the-refore may well form a strong foundation forscientists working with cultural heritage mate-rials, especially in the area of archeometa-llurgy.

As a matter of fact mineralogy naturally dealswith complex mineral systems most often com-posed of a large number of chemical elementsand several crystal-chemical phases, it hassolid roots on crystallography and the atomicstructure of matter, and it is technically andconceptually equipped to face the methodologi-cal and experimental challenges involved in thetechnical analysis of complex materials, and inthe interpretation of the processes acting uponthem (Fig. 1). Therefore mineralogy may well bethe discipline possessing an adequate balanceof the interdisciplinary knowledge required toface archeometallurgical problems (Artioli2010b: pp 305-348).

2. Archaeometallurgy

Pure and alloyed metals played a major rolein the technological evolution of mankind (Fig.2). It is indeed not surprising that post-Neolithic archeological periods are mostlydefined on the basis of the metal technology:

Copper Age (i.e. Chalcolithic, Eneolithic),Bronze Age, and Iron Age. To these we maywell add the Steel Age, dominating techno-logy and production after the industrial revo-lution. We are now in a period of extensiverecycling and use of advanced alloy composi-tes (Ashby 1987 & 2001).

Archaeometallurgy deals with all aspects ofmetal production, distribution and usage inthe history of mankind (Fig. 3, Rehren andPernicka 2008). To date the vast majority ofarchaeometallurgical investigations deal withthe early part of the use of metals (i.e. up tothe 2nd millennium BC), which include thetechnological use of a limited numbers ofmetallic elements (Cu, Pb, Au, Ag, Sn, Fe)and their alloys: mostly copper-based binaryand ternary alloys (bronzes and brass: Cu-Sb,Cu-As, Cu-Sn, Cu-Zn, Cu-Sn-Pb, Cu-Sn-Zn), sil-ver-based alloys (Ag-Cu, Ag-Au), and pewter(Sn-Pb) (Table 3.14 in Artioli 2010 summari-zes the approximate time and place of intro-duction of the different metals and alloys inthe past).

However, the archaeometallurgical aspectsconcerning the introduction of Fe and itsalloys (steel: Fe-C, Fe-C-P, etc.) and the sub-sequent use of all other known metals areequally fascinating. The recent paper byBourgarit and Plateau (2007) dealing with

Archaeometallurgy: the contribution of mineralogy Gilberto Artioli

Fig. 1. The schematic diagram of some of the disciplines and knowhow forming the backbone of modern mineralogy (modified from Artioli andAngelini 2011.

67

the early use of Al and the possibility to dis-tinguish the later electrolytic productionfrom the early chemical reduction is a niceexamples of the problems and techniquesinvolved in modern archaeometallurgical

analysis.

A shortlist of reference textbooks and mono-graphs is listed at the end of the paper as anintroductory aid to archaeometallurgy.


Fig. 2. Relative importance of the use of materials through the prehistory and history of mankind following Ashby (1987, 2005).

Fig. 3. The archaeometallurgical cycle (from Rehren and Pernicka 2008).

68

3. The contribution of mineralogy toarcheometallurgy

Based on the schematic diagram shown inFig. 3, we may distinguish several steps inthe production and life cycle of metallicobjects. The examples below are intended toshow the contribution of mineralogy to thecharacterization and to the archaeometallur-gical interpretation of the materials pertai-ning to some of the major stages of the cycle.

3.1. The mining stage: ore extraction andprovenancing

Our knowledge on mining and ore treatmentin historical times has the support of writtensources, oral accounts, toponyms and familynames, and even direct evidence of galleries,mine surveys and maps, treatment plants,written records of production and trades, andthe like. However, the more we venture intothe past the more the evidence is confusedand fragmentary.

When we deal with prehistory, there are nodirect sources of information, and the mate-rial evidence is scarce and very often conta-minated or altered by successive exploitationin late periods. Therefore the investigation ofprehistoric archaeometallurgy is a fascinatingand challenging task, aimed to reconstructthe links between ore deposits, mining activi-ties, ore treatment, and metallurgical sites.

One of the major problems is that the availablepieces of archaeological evidence (prehistoricmines, roasting and smelting sites, technicalceramics, smelting slags, furnaces, crucibles,raw metal ingots) are almost never found in thesame site: there are lots of pieces in the puzz-le, though there is hardly a complete image. Inthe most general case it is virtually impossibleto complete all aspects of the investigation foran individual site or area, simply because oftime, cost, and record biases.

Only a few extensive long-term projects comeclose to delineating all the major features ofthe metallurgical chain in a geographicalarea, from the mine survey to the productionand diffusion of metal. These are the casesof the prehistoric copper extraction in the

Arabah (Timna and Faynan) region in theNear East where, over thirty years, long pro-jects have investigated and detailed out allavailable evidence at dif ferent levels(Rothenberg and Merkel 1995, Rothenberg1999, Weisgerber 2003, Hauptmann 2007),reaching a sound and complete interpretationof the regional archaeometallurgy.

A recent long term project is the HiMAT one(www.uibk.ac.at/himat/), aimed to investiga-te the history of mining activities in the Tyroland adjacent areas, with impact on environ-ment and human societies (Tropper andVavtar 2009, Schibler et al. 2011).

In general, the main questions involved in theinvestigation of prehistoric metal extractionsites are:

(1) what metal was extracted and what wasthe nature of the mineral charge?

(2) can we trace the provenance of the smel-ted or treated ore minerals?

(3) when did the extraction activity startedand how long did it last? and

(4) what was the technology for metal extrac-tion from the ores?

The answers often require investigations atvery different scales: from the submicronscale of segregations and impurities inmetals and slags, to the large geographicscale of the regional distribution of ore depo-sits and metal objects.

As an example, many of these questions con-cerning the rise and development of coppermining and metallurgy in the Alpine areafrom the end of the Neolithic (approx. in the5th millennium BC) to the end of the BronzeAge (approx. the beginning of the 1st millen-nium BC) are still unanswered.

Our knowledge before the Iron Age is veryfragmentary and we are confronted for exam-ple with mines showing early exploitation butlittle or no evidence of reduction slags (i.e.Libiola and Monte Loreto, Ligurian Alps, Italy:Maggi and Pearce 1998; Saint Véran,Queyras, France: Bourgarit et al. 2008,2010), or with several smelting sites where asubstantial amount of slags have been


69

found, but they show no straightforward con-nection with the ore deposits (i.e. Luserna,Trentino: Nicolis et al. 2007; Renon, AltoAdige).

One further paradox is that there was abun-dant metal circulating since at least the 4th

millennium BC, as exemplified by the Icemancopper axe (approx. 3200 BC: Fleckinger2007) and a number of coeval artefacts, butthere is no clear evidence of large scalemining or smelting until half a millenniumlater. The only earlier recognized smeltingsites are small scale metallurgical activitiesconfined to the Brixlegg site, in Tyrol(Höppner et al. 2005) and Belovode, Serbia(Radivojevic et al. 2010). Several sites withChalcolithic slags are known in the EasternAlps (Artioli et al. 2007) although they mostlydate to the mid 3rd century BC or later. Theyapparently indicate a sizeable mass copperproduction only towards the latter part of theAlpine Copper Age.

In the last few years a long-term project focu-sed on prehistoric Alpine copper metallurgy(AAcP, Alpine Archaeocopper Project:http://www.geoscienze.unipd.it/aacp/welco-me.html) was launched in order to investiga-te in detail the relationship between ores,slags, and metal in the area. The core of theproject is the development of an extensivedatabase of Alpine copper mines containingmineralogical, petrological, chemical, isoto-pic, and genetic information on the ore depo-sits (Artioli et al. 2008a).

Solid minerogenetic interpretation of thedeposits (Nimis et al. 2012) and advancedstatistical analysis of the ore chemical andisotopic data is the basis to understand theapplicability of the measured geochemicaltracers (Pb and Cu isotopic ratios, nearly 60minor and trace elements, including REE ele-ments) to metallurgical processes, includingslagging and reduction smelting. The discrimi-nant analysis developed on the copper oredatabase was tested on the well-characteri-zed area of Agordo, Alpi Bellunesi, Veneto,whose Valle Imperina copper mine was theprincipal source of copper for the Republic ofVenice. The results of the Agordo case studyindicate that the developed discriminant

models can be applied to the smelting slagsand to the unalloyed raw copper extractedfrom the ores (Artioli et al. 2008b).

Application of the tracing models to themanufactured objects requires a careful che-mical and metallographic examination of themetal, in order to understand the nature ofthe alloy, the presence of remnants of the ori-ginal charge as inclusions and segregations,and possible biases due to alteration or con-tamination. It may indeed be dangerous topursue separately the investigation of theores, of the slags, and of the metal objects:strict synergy is needed to understand thecomplexity of the metallurgical cycle.

Recent applications of the models based onthe database of geochemical tracers havemet some success in

• firmly locating in the Monte Fondoli area(Pfunderer Bergwerk, Chiusa, Alto Adige)the source of the copper ores used for allknown Chalcolithic smelting slags in theBrixen area (Millan, Gudon,Circonvallazione Ovest: Angelini et al.2012), and

• identifying the source of the ores used forthe production of the Late Bronze Ageobjects found in the Monte Cavanerohoard (Chiusa di Pesio, Cuneo, WesternAlps: Artioli et al. 2009).

These examples highlight the patient work ofthe mineralogist in provenancing studies, thatis to critically filter the geological, geochemi-cal, mineralogical, and archaeological informa-tion into a consistent and unifying interpreta-tion. Key issues to be discussed are:

(1) the development, consistency, and data-retrieval protocols of geochemical andisotopic databases;

(2) the availability of chemical and isotopicdatabases for different ore deposits; and

(3) the mineralogical, geochemical and statis-tical significance of the chemical and iso-topic tracers used for provenance discri-mination.


70

3.2. The smelting stage: metal production

Our knowledge concerning the ancient extrac-tion processes of metals from the ores is moreand more fragmentary as we go back in time,in analogy to our knowledge about mining ope-rations. For recent times, starting from the fun-damental 16th century volumes of VannoccioBiringuccio (De la pirotechnia, published in1540) and Georg Agricola (De re metallica,published in 1556) (Fig. 4), we have ampledocumentation, though sometimes the detailsare difficult to interpret and reproduce.

However for earlier times there are no directaccounts of the metallurgical operations, sothat we have to rely on the archaeometric analy-sis and the archaeological reconstructions ofthe objects found in archaeometallurgical sites(see table 2.1 of Hauptmann 2007). The scar-ce evidence is based on: stratigraphy of thesites, furnaces, crucibles and other technicalceramics, ore fragments and smelting slags.Generally most of these evidences are sparseand removed from the mining sites.

A discussion of the logistic rationale for the

location of the metallurgical smelting siteswith respect to the ore mines is out of placehere, it is sufficient to say that only in a fewplaces there is a clear and recognized connec-tion between mining and metallurgical activi-ties: for example in the cited Faynan area(Hauptmann 2007) or at Cabrières, Francewhere a small metallurgical village has beenexcavated close to a known mining area(Ambert and Vaquer 2005).

In many instances furnaces and slag heapsare quite far from the ores, or even in areastotally deprived of ore minerals, especially forpre-Bronze Age activities (Hauptmann 2007,Artioli 2010). The reasons for the transporta-tion of the ores are subject to debate: in thecase of early Chalcolithic smelting the follo-wing of Neolithic trade patterns from sourcesto settlements have been postulated(Hauptmann 2007), whereas the Bronze Agesmelting sites are close to the ores, thoughtheir geographical distribution may follow therequirements for large amount of fuel or othertechnical considerations.

Concerning the smelting technology, if furna-


Fig. 4. Title pages of the two fundamental volumes on mining and metallurgy published in the 16th century: (left) De la pirotechnia of VannoccioBiringuccio and (right) De re metallica of Georg Agricola.

71

ces and technical ceramics (crucibles, tuye-res, bellows, etc.) are available, they may gre-atly help in the technical reconstruction of thereduction process.

Mineralogical analysis of the ceramics andthe furnace components may be important tothe identification of the ores being treated,the temperatures of firing, and the parame-ters of the smelting operations, though mostcommonly these data are obtained by thedetailed investigation of the smelting slags(Bachmann 1982). The mineralogical, textu-ral, and chemical analysis of the slags yieldinformation on the type of minerals in thecharge, the smelting temperatures, the visco-sity, the different steps of metal reduction,the efficiency of the process, the redox con-ditions, the cooling rates, etc. The informa-tion is commonly extracted by a combinationof X-ray diffraction, optical microscopy, scan-ning electron microscopy, electron probemicro-analysis, and X-ray fluorescence spec-troscopy (see Artioli 2010: box 3.k).

The great majority of the ancient smeltingslags were derived from the processing ofcopper, iron, and lead ores, mostly sulphides.Tin, antimony, and zinc slags are also known;they are scarce but they can be rather impor-tant to pinpoint the source of alloying ele-ments in bronzes and brass.

By far most of the slag studies in the literatu-re involve copper smelting and iron smel-ting/smithing. The major difference betweenthe two processes is that metallic copper(melting temperature 1080°C) can be produ-ced through a fully molten state (i.e. the gan-gue and the slag are more viscous than therunning matte and metal), whereas the mol-ten state of metallic iron (melting temperatu-re 1540°C) could not be reached beforemodern cast iron production, so that the pre-industrial technologies required squeezingthe fluid slag out of the solid metal at hightemperature by forging.

The chemical, mineralogical and textural cha-racteristics of the different types of smeltingslags can be found in the cited literature(Bachmann 1982, Hauptmann 2007, Artioli2010, and references therein).

Here, it is important to remind that a neatchange in copper smelting technology appe-ars to take place at about the Chalcolithic-Bronze Age transition, as evidenced by themarked differences in the Chalcolithic slags

(Bourgarit 2007) in terms of texture, hetero-geneity, incompleteness of the melting pro-cess, absence of fluxing, low density, and lowreducing conditions with respect to the laterBronze Age slags (Anguilano et al. 2002,Mette 2003). Fig. 5 shows typical microsco-pic textures of Chalcolithic copper smeltingslags in the Alpine region.One further note should be mentioned on theinvestigations of the smelting sites. Very fre-quently in the past during the archaeologicalexcavations of the sites the metallurgicalslags were not considered as important asthe metals, technical ceramics, or otherarchaeological finds, thus producing loss ofvaluable knowledge.

Modern excavations now pay due attention tothe distribution and quantification of theslags (even micro-slags, such as the ham-


Fig. 5. Optical micrographs of fayalitic olivine (top: crystal chains, bot-tom: euhedral crystals) crystallized in the glassy matrix of Chalcolithiccopper slags from Bressanone (Angelini et al. 2012).

72

merscales produced during iron smithing andforging, for example Angelini et al. 2011) inthe archaeological record, so that this infor-mation can be fully used to interpret andreconstruct the metallurgical activities.Interaction between the archaeometallurgistand the archaeologist during excavation isfundamental.3.3. The manufacturing stage: metallurgyand metallography

The physical properties of a metal object (i.e.hardness, ductility, etc.) are dependent on thesize, shape, and orientation of the crystallites(i.e. the micro-texture) in the material. In turn,the metal texture is determined by its thermo-mechanical history, that is the temperature-time path (i.e. cooling rates, annealing tempe-ratures) and the mechanical stress imposedduring manufacturing. On one hand, modernmetallurgical engineering is interested in desig-ning and optimizing the production processesin order to meet specific physical properties,

on the other hand archaeometallurgical investi-gations aim to measure the physical propertiesin order to reconstruct the ancient manufactu-ring processes of the objects. Both investiga-tions rely in the measurement and quantifica-tion of the metal texture, which may be experi-mentally performed by optical metallography,electron backscattering diffraction, or crystallo-graphic texture analysis (see details in Artioli2010, box 3.l).

Reflected-light optical metallography (Scott1991, Wang and Ottaway 2004), based onpolishing and chemical etching of a small por-tion of the metal surface, is by far the most dif-fused, fast and cheap technique, though it isgenerally invasive and only yields 2D informa-tion of the crystalline arrangement (Fig. 6). Itcan be used as a micro-invasive technique onlyby careful micro-sampling of pertinent areas ofthe unaltered metal (Angelini et al. 2009).

Electron backscattering diffraction (EBSD) is


Fig. 6. Reflected light optical metallographic images of copper crystallites that underwent different thermo-mechanical processes: (a) copper den-drites produced by casting, (b) thermally annealed and partially recrystallized cast copper still showing remnants of the original dendrites, (c) slipsystems produced by mechanical working of the metal, and (d) worked copper through several cycles of mechanical hardening and thermal anne-aling, showing recrystallized grains and twin boundaries.

73

based on the point-by-point analysis of theorientation of the crystallite through interpreta-tion of the Kikuchi lines (Schwartz et al. 2000);the 2D texture is obtained by mapping thedomains with equal crystal orientation. It maybe performed micro-invasively on submillime-ter–sized samples, though the analysis andinterpretation of polyphasic metals may be acomplicated issue.

The non-homogeneous spatial distribution ofcrystallites in a metal sample implies a non-homogeneous distribution of Bragg intensityin the Debye diffraction rings. The measure-ment of the variation in intensity along the dif-fraction rings allows recalculation of the so-called orientation distribution function (ODF)of the crystallites, that is the function statis-tically describing the crystallographic orienta-tion of the crystals in the samples with res-pect to the sample orientation (Kocks et al.1998, Popa 2008). Orientation distributions

are generally represented graphically bymeans of the pole figures in direct (i.e. thesample space) or inverse (i.e. the crystalspace) space (Monaco and Artioli 2011). Themeasurement of the crystallographic texturalfeatures of the sample (CTA: crystallogra-phic texture analysis) therefore is a powerfulalternative to perform metallographic analy-sis in a totally non-invasive mode (Artioli2007, Artioli 2010 pp 343-348).

It should also be reminded that for metalsand average absorbing materials, textureanalysis performed with X-rays gives only anappropriate description of the surface layerof the sample, whereas the same analysisperformed with deeply penetrating neutronsmay yield a more appropriate interpretation ofthe bulk. Fig. 7 shows an example of the polefigures resulting from the full CTA analysis ofa prehistoric copper axe: the data allow com-plete interpretation of the thermo-mechanical


Fig. 7. (a) Eneolithic copper axe from Castelrotto/Kasteruth, Bolzano, Italy (Courtesy of the Museo Archeologico dell’Alto Adige), with the diagramshowing the reference orientation directions (RD, TD, ND) used in the diffraction experiment; (b) direct pole figures projected from the calcula-ted ODF; and (c) inverse pole figures recalculated from the ODF along the main directions of the object (modified from Artioli et al. 2003). Thedata indicate that the axe was cast in a bivalve mould and subsequently cooled very slowly, allowing for an extreme iso-orientation of the crysta-llites.

74

history and the interpretation of the metallur-gical techniques employed to produce theobject.

Further information on the manufacturing tech-niques and assemblage of composite objectsmay be obtained by routine and advances ima-ging techniques (i.e. 2D radiography and 3Dtomographic reconstructions). Neutron imagingis especially useful for metals because of thepenetrating power of neutron beams (Lehmannet al. 2005; Casali 2006). Fig. 8 (fromLehmann 2008) shows the comparison of highenergy and neutron radiographies on a complexcomposite statue.

3.4. The life and death of metal objects

Surface analyses of archaeological metal findsand art objects are aimed to

• verify the presence of marks derived fromthe manufacturing process, polishing, andtool use (Fig. 9);

• analyse the composition of the surfacelayers to check for original patinas, protec-tive layers, or chemical treatments; and

• assess the alteration and corrosion stateof the metal for conservation purposes(Scott 2002, 2009).

A number of characterization techniques areused to investigate the external metal surfa-ces, through the corrosion layers, to the pris-tine metal: high magnification optical micros-copy, scanning electron microscopy, X-ray dif-fraction depth profiling, X-ray and neutronimaging, X-ray fluorescence spectroscopy,proton induced X-ray emission, X-ray photoe-mission spectroscopy, and many more.

Fig. 10 shows the use of hard X-ray tomogra-phic reconstruction techniques for the non-invasive visualization of the alteration layersin a strongly corroded Bronze Age copperingot (Artioli et al. 2011).

The interpretation of the surface layers of themetal in terms of the original compositionand the subsequent corrosion processesmay be very difficult, particularly in presenceof polymetallic alloys, original surface treat-ments and patinas, or complex soil-metal andatmosphere-metal interactions.

The archaeometrical problems involved in thecharacterization of metal surfaces are com-


Fig. 8. A sealed Buddha figure (left) (about 20 cm in height, photo) Sakyamuni, Bhumisparsa Mudra, West-Tibet, 14th–15th century. It was inves-tigated with X-rays (150 kV tube voltage, middle) and thermal neutrons (right). The X-ray image only shows the outer metallic brass cover, whe-reas neutrons can penetrate the metal and show the hidden organic cultual enclosures (wood, dried plants, rope) (from Lehmann 2008).

Fig. 9. Example of manufacturing tools marks on the surface of an IronAge cauldron. See the Iron Age Chiseldon cauldrons project at theBritish Museum for details. http://www.britishmuseum.org/rese-arch/projects/featured_project_chiseldon.aspx)

75

plex and various (Giumlia-Mair 2005), theyinclude the assessment of the corrosion pro-cesses, the identification of original patinas,the reconnaissance of re-deposited phases,the identification of fake alteration layers inforgeries (Craddock 2007), and so on.

In principle many of these problems are rela-ted. Sometimes the identification of externalcontamination or diffusion phenomena rela-ted to corrosion is straightforward. In othercases the identification of the actual proces-ses is much more difficult. For example,enrichment of Sn at the surface of bronze isa very common feature that can be producedby very different mechanisms: inverse segre-gation during casting, preferential dissolutionof copper with respect to tin during corrosion,or explicit superficial tinning of the object(Tylecote 1985).

Simple chemical analysis is sometimes not suf-ficient to resolve the issue and careful analysisof the micro-textural features (such as theinterstitial position and the morphology of there-precipitated crystalline phases, presence ofchemical coring, etc.) are needed.

4. References

Agricola G (1556) De re metallica. Engl. Transl.:Herbert Clark Hoover and Lou Henry Hoover(1950) Dover Publications Inc., New York.

Ambert P, Vacquer J (eds) (2005) La premièremétallurgie en France et dans les pays limitro-phes. Société Préhistorique Française,Carcassonne.

Angelini I, Molin G, Artioli G (2009) L’atelier meta-llurgico di Monte Cavanero. Indagini chimiche emetallografiche. In: Venturino Gambari M (ed) Ilripostiglio del Monte Cavanero di Chiusa di Pesio(Cuneo). LineLab.Edizioni, Alessandria, Pp. 107-165.

Angelini I, Busana MS, Francisci D, Bernardi L,Bacchin A, Molin G (2011) Excavation and prelimi-nary archaeometric investigations of iron smithingslags from the Roman workshops atMontebelluna (Treviso, Italy). Archaeometallurgyin Europe III. Bochum, June 29-July 1 2011.Metalla (Bochum) Sonderheft 4: 99.

Angelini I, Gallo F, Artioli G, Nimis P, Tecchiati U,Baumgarten B (2012) Mineralogical and isotopiccharacterization of the Late Chalcolithic slags fromBressanone/Brixen (Nor thern Italy). 39thInternational Symposium on Archaeometry,Leuven, Belgium, 28 May - 1 June 2012, Abst n.16, pp. 65.

Anguilano L, Angelini I, Artioli G, Moroni M,Baumgarten B, Oberrauch H (2002) Smeltingslags from Copper and Bronze Age archaeologicalsites in Trentino and Alto Adige. In: D’Amico C(ed) Atti II Congresso Nazionale di Archeometria.Bologna 29 January-1 February 2002, PàtronEditore, Bologna, pp 627-638, 2002.

Artioli G (2007) Crystallographic texture analysisof archaeological metals: interpretation of manu-facturing techniques. Appl. Physics A89:899–908.


Fig. 10. (a ) Virtual 2D section of the X-ray computed tomography scanof a corroded Bronze Age ingot (Artioli et al. 2011). The segmentedfalse-colour image shows the copper metal core (orange), the cupritelayer (purple) and the external layer of secondary minerals (green). (b,c, d) The 3D exploded view of separate layers. The external corrosionlayers very often preserve the original shape of the corroded object,which is not recognizable in the leftover metal.

76

Artioli G (2010a) Mineralogy and cultural heritage.Chimia 64:712-715. DOI: 10.2533/chi-mia.2010.712

Artioli G (2010b) Scientific methods and the culturalheritage. Oxford University Press, Oxford, pp 1-552.Artioli G, Dugnani M, Hansen T, Lutterotti L,Pedrotti A, Sperl G (2003) Crystallographic textu-re analysis of the Iceman and coeval copper axesby non-invasive neutron powder diffraction. In:Fleckinger A (ed) La mummia dell’età del rame. 2.Nuove ricerche sull’uomo venuto dal ghiaccio.Collana del Museo Archeologico dell’Alto Adige,Vol. 3, Folio Verlag, Bolzano, pp 9–22.

Artioli G, Angelini I, Burger E, Bourgarit D (2007)Petrographic and chemical investigation of theearliest copper smelting slags in Italy: towards areconstruction of the beginning of copper meta-llurgy. Proc. 2nd Intern. Conference“Archaeometallurgy in Europe 2007”, Aquileia,17-21 June 2007. Printed in the Selected PapersVolume, AIM, Milano, pp. 12-20, 2009.

Artioli G, Baumgarten B, Marelli M, Giussani B,Recchia S, Nimis P, Giunti I, Angelini I, OmenettoP (2008a) Chemical and isotopic tracers in Alpinecopper deposits: geochemical links betweenmines and metal. Geo.Alp 5:139-148.

Artioli G, Nimis P, Gruppo Arca, Recchia S, MarelliM, Giussani B (2008b) Gechemical links betweencopper mines and ancient metallurgy: the Agordocase study. Rend. Online Soc. Geol. It. 4:15-18.

Artioli G, Angelini I, Giunti I, Omenetto P, Villa I(2009) La provenienza del metallo degli oggetti diMonte Cavanero: considerazioni basate sugli iso-topi del Pb e sulla geochimica delle mineralizza-zioni cuprifere limitrofe. In: Venturino Gambari M(ed) Il ripostiglio del Monte Cavanero di Chiusa diPesio (Cuneo). LineLab.Edizioni, Alessandria, pp.167-178.

Ar tioli G, Angelini I (2011) Mineralogy andarchaeology: fatal attraction. Eur J Min 23:849-855. DOI: 10.1127/0935-1221/2011/0023-2119

Artioli G, Parisatto M, Angelini I (2011) Highenergy X-ray tomography of Bronze Age copperingots. Archaeometallurgy in Europe III. Bochum,June 29-July 1 2011. Metalla (Bochum)

Sonderheft 4:247.

Ashby MF (1987) Technology of the 1990s: advan-ced materials and predictive design. Phil. Trans.Royal Soc. London. Series A, Mathematical andPhysical Sciences 322:393–407.Ashby MF (2001) Drivers for material develop-ment in the 21st century. Prog. Mater. Sci.46:191–199.

Ashby MF (2005) Materials selection in mechani-cal design. 3rd Edition. Butterworth Heinemann,Oxford. pp. 603.

Bachmann H (1982) The identification of slagsfrom archaeological sites. Institute of ArchaeologyOccasional Publication 6, Institute of Archaeology,London.

Biringuccio (Vannoccio Biringuccio) (1540) De lapirotechnia Engl. Transl.: Cyril Stanley Smith andMartha Teach Gnudi (1959) Basic Books, NewYork.

Bourgarit D (2007) Chalcolithic copper smelting.In: La Niece S, Hook D, Craddock P (eds) Metalsand mines. Studies in archaeometallurgy. TheBritish Museum–Archetype Publications, London.pp. 3–14.

Bourgarit D, Plateau J (2007) When aluminiumwas equal to gold: Can a ‘chemical’ aluminium bedistinguished from an ‘electrolytic’ one?Historical Metallurgy 41:57–76.

Bourgarit D, Rostan P, Burger E, Carozza L, MilleB, Artioli G (2008) The beginning of copper massproduction in the western Alps: the Saint-Véranmining area reconsidered. Historical Metallurgy42:1-11.

Bourgarit D, Rostan P, Carozza L, Mille B, Artioli G(2010) Vingt ans de recherches à Saint-Véran,Hautes Alpes: état des connaissances de l‘activi-té de production de cuivre à l’âge du Bronzeancient. Trabajos de Prehist. 67:269-285. DOI:10.3989/tp.2010.10039

Casali F (2006) X-ray and neutron digital radio-graphy and computed tomography for cultural heri-tage. In: Bradley D, Creagh D (eds) PhysicalTechniques in the Study of Art, Archaeology andCultural Heritage, Volume 1, Chapter 2, Elsevier,


77

Amsterdam, pp 41-123 DOI: 10.1016/S1871-1731(06)80003-5

Craddock PT (2007) Scientific investigation ofcopies, fakes and forgeries. Butterwor th-Heinemann, Oxford. pp. 628.Fleckinger A (2070) Ötzi, the iceman. The fullfacts at a glance. Folio Verlag, Bolzano.

Giumlia-Mair A (2005) On sur face analysis andarchaeometallurgy. Nucl. Instrum. Meth. Phys.Res. B 239:35–43.

Hauptmann A (2007) The archaeometallurgy ofcopper. Evidence from Faynan, Jordan. Springer,Berlin – New York. pp. 388

Höppner B, Bartelheim M, Huijsmans M, KraussR, Martinek K-P, Pernicka E, Schwab R (2005)Prehistoric copper production in the Inn Valley(Austria), and the earliest copper in CentralEurope. Archaeometry 47:293–315.

Kocks UF, Tomé CN, Wenk HR (eds) (1998)Texture and anisotropy. Preferred orientation inpolycrystals and their effect on materials proper-ties. Cambridge University Press, Cambridge.

Lehmann EH, Vontobel P, Deschler-Erb E, SoaresM (2005) Non-invasive studies of objects fromcultural heritage. Nuclear Instruments andMethods in Physics Research Section A:Accelerators, Spectrometers, Detectors andAssociated Equipment, 542:68-75, DOI:10.1016/j.nima.2005.01.013

Lehmann EH (2008) Recent improvements in themethodology of neutron imaging. Pramana – J.Phys. 71:653-661.

Maggi R, Pearce M (1998) Les mines prehistori-ques de Libiola et Monte Loreto (nouvelles foui-lles). In: Frere-Sautot M.-Ch. (ed.)Paleometallurgie des cuivres. Monique Mergoil,Montagnac. pp. 89–94.

Mette B (2003) Beitrag zur spatbronzezeitlichenkupfermetallurgie im Trentino (Sudalpen) im ver-gleich mit anderen prehistorischen kupferschla-ken aus dem Alpenraum. Metalla 10:1–122.

Monaco HL, Ar tioli G (2011) Experimentalmethods in X-ray and neutron crystallography. In:

Giacovazzo C (ed.) Fundamentals ofCrystallography. IUCr Texts on Crystallography,3rd Edition, Vol. 15. Oxford Science Publications,Oxford, Ch. 5, pp 301-416

Nicolis F, Bellintani P, Artioli G, Cappellozza N(2007) Archaeological excavation at Pletz vonMozze, Luserna, Trentino, Italy. A case study forassessment of late Bronze Age metallurgical acti-vities. Proc. 2nd Intern. Conference“Archaeometallurgy in Europe 2007”, Aquileia,17-21 June 2007.

Nimis P, Omenetto P, Giunti I, Artioli G, Angelini I(2012) Lead isotope systematics in hydrothermalsulphide deposits from the central-easternSouthalpine (Northern Italy). Eur. J. Miner. 24:23-37. DOI: 10.1127/0935-1221/2012/0024-216

Popa NC (2008) Microstructural proper ties:Texture and macrostress effects. In: Dinnebier RE,Billinge SJL (eds) Powder diffraction: Theory andpractice. RCS Publishing, Cambridge, pp 335-372

Radivojevi_ M, Rehren T, Pernicka E, _ljivar D,Brauns M, Bori_ D (2010) On the origins ofextractive metallurgy: new evidence fromEurope. J. Arch. Sc. 37:2775-2787. DOI:10.1016/j.jas.2010.06.012.

Rehren T, Pernicka E (2008) Coins, ar tefactsand isotopes – Archaeometallurgy and archaeo-metr y. Archaeometr y 50:232–248. DOI:10.1111/j.1475-4754.2008.00389.x

Rothenberg B (1999) Archaeo-metallurgical rese-arches in the Southern Arabah 1959–1990, partI: Late Pottery Neolithic to Early Bronze Age IV.Palestine Explor. Quart. 131:68–89.

Rothenberg B, Merkel JF (1995) Late NeolithicCopper Smelting in the Arabah. Inst. Archaeo-Metall. Studies Newsletter 15:1–8.

Schibler J, Breitenlechner E, Deschler-Erb S,Goldenberg G, Hanke K, Hiebel G, Hüster-Plogmann H, Nicolussi K, Marti-Grädel E, PichlerS, Schmidl A, Schwarz S, Stopp B, Oeggl K (2011)Miners and mining in the Late Bronze Age: a mul-tidisciplinar y study from Austria. Antiquity85:1259-1278.

Schwartz AJ, Kumar M, Adams BL (2000) Electron


78

Backscatter Dif fraction in Materials Science.Springer Verlag, Heidelberg.

Tropper P, Vavtar F (2009) Mineralogical/petrolo-gical and geochemical survey of historic miningsites in the Tyrol and adjacent areas in the cour-se of the Sonder forschungsbereich HiMAT(History of Mining Activities in the Tyrol andAdjacent Areas). Geo.Alp 6:1–10.

Tylecote RF (1985) The apparent tinning of bron-ze axes and other artifacts. J. Hist. Metal. Soc.19:169–175.

Wang Q, Ottaway BS (2004) Casting experimentsand microstructure of archaeologically relevantbronzes. Archaeopress, BAR International Series,Vol. 1331.

Weisgerber G (2003) Spatial organization ofmining and smelting at Feinan, Jordan: Miningarchaeology beyond the history of technology. In:Craddock P, Lang J (eds) Mining and metal pro-duction through the ages. The British MuseumPress, London. pp. 76–89.

5. Selected introductory and referencevolumes on archaeometallurgy

Bachmann H (1982) The identification of slagsfrom archaeological sites. Institute of ArchaeologyOccasional Publication 6, Institute of Archaeology,London.

Craddock P (1995) Early metal mining and produc-tion. Edinburgh University Press, Edinburgh.

Craddock P, Lang J (eds) (2003) Mining and metalproduction through the ages. The British MuseumPress, London.

Hauptmann A (2007) The archaeometallurgy ofcopper. Evidence from Faynan, Jordan. Springer,Berlin – New York.

La Niece S, Hook D, Craddock P (eds) (2007)Metals and mines. Studies in archaeometallurgy.The British Museum – Archetype Publications,London.

Porter DA, Easterling KE (2009) Phase transfor-mations in metals and alloys. 3rd Edition. CRCPress, Boca Raton, FL. p. 520.

Scott DA (1991) Metallography and microstructu-re of ancient and historic metals. The GettyConservation Institute, Los Angeles.

Scott DA (2002) Copper and bronze in art.Corrosion, colorants, conservation. The GettyConservation Institute, Los Angeles.

Scott DA, Eggert G (2009) Iron and steel in art:corrosion, colorants, conservation. ArchetypeBooks. pp 196.

Scott DA (2010) Ancient metals: microstructureand metallurgy. Volume 1. Copper and copperalloys. Conservation Science Press. pp 351

Tylecote RF (1987) The early history of metallurgyin Europe. Longman, London. pp. 391.

Tylecote RF (1992) A history of metallurgy. 2nd

Edition. The Institute of Materials, London. P.205. (1st edition 1976, The Metals Society,London).

Wang Q, Ottaway BS (2004) Casting experimentsand microstructure of archaeologically relevantbronzes. Archaeopress, BAR International Series,Vol. 1331.


79

Microanalysis of pigments inpainted artworks/ David Hradil (1, 2) / Janka Hradilová (2)

Abstract

This paper summarizes the current knowledge concerning the materials analysis of paintedartworks. In the introduction materials analysis is interrelated with other disciplines within abroad context of the research of cultural heritage. Current methodological and instrumentaldevelopments in the area are described with a special attention given to non-invasive analyti-cal methods. The article also briefly deals with the relation of natural science, technology andhumanities in the investigation and conservation of the fine art. In more detail historical pain-ting techniques are discussed and paint layers are described as composed materials contai-ning both organic and inorganic compounds in characteristic layer sequences. The aim ofmaterials microanalysis of paints is to get both technological and materials signatures, whichboth could be specific for the period, region or workshop. Therefore the identification of orga-nic and inorganic materials in paints should always be accompanied by the description of thelayer stratigraphy and finding other detailed characteristics that relate to the origin of the pain-ting. Mineralogy plays an important role in the analysis of paint layers, because the majorityof historical pigments are minerals. When using the tools of mineralogical analysis, particu-larly the X-ray diffraction methods, one can get an idea of the process of formation and/orfabrication of pigments and its regional provenance. It also helps to identify the provenanceof the painting itself. A substantial part of this paper deals with the mineralogical analysis ofhistorical pigments. Carefully selected case studies demonstrate the usefulness of themethod to solve important tasks of the origin of artworks.

Resumen

En este artículo se hace un resumen del estado actual de conocimiento del análisis de losmateriales que constituyen las obras de arte pintadas. En la introducción se interrelaciona elAnálisis de Materiales con otras disciplinas dentro de un amplio contexto de la investigacióndel Patrimonio Cultural. Se describen los actuales avances metodológicos e instrumentalesen esta área, con especial atención a los métodos analíticos no invasivos. El artículo tambiéntrata brevemente la relación de las Ciencias Naturales, la Tecnología y las Humanidades enla investigación y conservación de las obras de arte. Se discuten con más detalle las técni-cas pictóricas históricas y se describen las capas pictóricas como materiales compuestos,que contienen tanto compuestos orgánicos como inorgánicos en secuencias de capas carac-terísticas. El interés del microanálisis de los materiales de las pinturas es encontrar huellas,tanto tecnológicas como materiales, que puedan ser específicas de un período, región o taller.Por esto, la identificación de los materiales orgánicos e inorgánicos en las pinturas debe venirsiempre acompañada por la descripción de la estratigrafía de las capas y el hallazgo de otrascaracterísticas de detalle relacionadas con el origen de las pinturas. La Mineralogía desem-peña un papel importante en el análisis de las capas pictóricas, ya que la mayoría de los pig-mentos históricos son minerales. Al recurrir a los instrumentos del análisis mineralógico, y enparticular a los métodos de Difracción de Rayos X, podemos hacernos una idea de los proce-sos de formación y/o fabricación de los pigmentos y su proveniencia regional, Esto tambiénayuda a identificar la procedencia de la misma obra pictórica. Una parte importante de esteartículo se extiende sobre el análisis mineralógico de los pigmentos históricos. Se han selec-

(1) Institute of Inorganic Chemistry of the AS CR, v.v.i., ALMA laboratory, 1001 Husinec-Rez, Czech Republic, [email protected] (2) Academy of Fine Arts in Prague, ALMA laboratory, U Akademie 4, 170 22, Prague 7, Czech Republic

> >

80

cionado a este propósito algunos casos prác-ticos que demuestran la utilidad de estemétodo para resolver cuestiones importan-tes sobre el origen de las obras de arte.

Key-words: Conservation science, materials micro-analysis, powder X-ray microdiffraction, paintingtechnique, mineral pigments, clay minerals.

1. Introduction

Materials analysis of tangible cultural herita-ge is a key part of conservation science.Conservation science is a branch of naturalscience that is concerned with the materialaspects of works of art, their deteriorationand conservation. Conservation science, con-trarily to conservation technology, falls to thefundamental research of materials, describesthe processes of their interactions and deve-lops new tools for their description. Technicaland technological solutions are then derivedfrom the current knowledge obtained.

Conservation science relies on measurementsof physical and chemical properties. Such mea-surements are central to analytical chemistry,that branch of chemistry concerned with deter-mining the qualitative and quantitative identityof a substance (Mazzeo et al. 2011). Recently,non-destructive analysis of cultural heritageobjects represents one of major trends in analy-tical chemistry (Chiari 2008; Martin et al.2010). Better understanding is the first neces-sary step for better preservation, conservation,restoration and protection and gives also decisi-

ve arguments for an art-historic evaluation ofworks of art (Fig. 1).

It should, however, be said that without theparticipation of other scientific branches, acomprehensive analysis is only hardly possi-ble – this is simply given by the diversity ofmaterials that form cultural heritage objects(stones, paper, plaster, glass, pigments etc.).Substantial limitations of any analysis ofworks of art are given by the fact that no sam-

Microanalysis of pigments in painted artworksHradil, D. & Hradilová, J.

Fig. 1. Interrelation of the fields of science (in grey), technology (in blue) and humanities (in yellow) in the cultural heritage investigation and pro-tection.

Fig. 2. Sampling. Conservators should make a compromise betweenrepresentativeness of the collected sample and minimizing a risk ofthe painting damage

81

pling is usually allowed or only minute sam-ples are possible to get. Furthermore theirinternal heterogeneity can be considerable(Fig. 2).

Because of these reasons in Europe in lastdecades, non-invasive (integrity of the objectis preserved, no sampling is necessary) andnon-destructive micro-analytical (sample istaken, but analyzed as is, not consumed ormodified) methods were widely tested in cul-tural heritage field – especially in the case ofpaintings. This growing trend can easily bedemonstrated on the increasing number ofpublications in top-impacted analytical jour-nals. Current developments can be dividedinto two directions

• developments in physical principles andinstrumentation – e.g. portable instru-ments (Grieten and Casadio 2010; Milianiet al. 2010) and

• developments in measuring and interpreta-tion strategies – including unconventionalapproaches (Cardell et al. 2009; Palmieriet al. 2011).

2. Paints as composed materials

In cultural heritage, composed materials appe-ar as an integral part of historic objects, arti-facts and works of art of any kind. The mostimportant fact is that their internal heteroge-neity brings a crucial information related eitherto the technology of manufacturing, natural pro-cess of formation, or the process of secondarymodification, deterioration etc. From the analy-tical point of view composed materials taken (ifever) in minute amounts from objects repre-sent unconventional samples that cannot behomogenized or modified without a risk ofloose of information described above.

A comprehensive methodology that includesan advanced description of internal structu-res and materials composition of different artobjects made of composed materials is notdeveloped yet. Close collaboration and com-posite action of several scientific branches isneeded to achieve inevitable synergic effectsrouting to develop such a methodology basedon using up-to-date analytical and imaging

instruments (Jonge and Vogt 2010; Yoneharaet al. 2010). The necessity of an interdiscipli-nary approach to solve this problem is evi-dent.

In a routine practice of the investigation of pain-tings, imaging and analytical methods are notyet fully interrelated. Conservators use conven-tional UV-Vis photographs, IR reflectography(IRR) or radiography (XR) and/or computer tomo-graphy (CT) to visually interpret technical signa-tures of the author visually – without any kno-wledge of materials used (Fig. 3). Analytical che-mists, on the other hand, frequently perform“blind” analyses of materials without any kno-wledge of how they are organized in the layersequence, what are their interrelations andwhat these “textural patterns” mean in terms ofthe painting technique. Materials and structu-ral/technological signatures should be, howe-ver, studied together in a substantially higherlevel of understanding than is common today.

Cultural heritage objects can be categorizedaccording to technological signatures. Usingthis clue, easel and wall paintings together


Fig. 3. Use of infrared reflectography (IRR) and X-ray radiography (XR)to visualize internal structure of the painting or, like in this case, touncover an original painting completely over-painted by another pain-ting (adopted from Hradil et al, 2006).

82

with polychrome sculptures, for example,belong to one single group – all these objectsare classified as fine art (not applied art) andcontain color (painting) layers applied in logi-cal sequences following similar technologicalrules (Fig. 4). This group of painted artworkswill only be considered in the following text.

Paint layers represent composed mixtures oforganic and inorganic compounds (pigmentsand their binders) of very different properties,which are further composed to characteristiclayer sequences; the description of this inter-nal structure of paints together with the identi-fication of organic binders (glue, egg yolk,whole egg, linseed oil etc.) is relevant for inter-preting of historical painting technique withperiod and regional specificity.

In the traditional European paintings, the tech-nological principles were held relatively strictly.The layer sequence starts with ground layersthat served as a coating material to preparethe surface of a wooden panel or canvas. Bytheir color, they were also closely related to thetechniques and materials in the subsequentpainting. Different types of (under)drawingsand/or priming (“imprimitura”) belong also tothe so-called preparatory layers.

Then the sequences of paint layers follow.Each the painting sequence is usually finishedby varnish based e.g. on egg white, natural oilsand resins and, later on, on synthetic acryla-tes. Original painting sequence can very fre-quently be over-painted by later re-paints.

In the European fine art, one of the firstfamous manuals of technical experiences,material preparations and painting itself, is theearly renaissance work by Cennino Cenniniissued first in 1437 (Cennino 1978). A verygood review of treatises and ‘recipe-books’from the 12th to the 18th centuries is given by


Fig. 4. Simplified description of the layer strati-graphy in cross-section of the painting frag-ment embedded in polyester resin (in the mid-dle) as visible in white (a) and UV (b) lights andin backscattered electrons (c)

Fig. 5. Degradation and color change of minium (Pb3O4). In the envi-ronment of dissociated carbonic acid minium disproportionates asfollows: Pb3O4 (orange minium) + 2H2CO3 # PbO2 (brown-black platt-nerite) + 2PbCO3 (white cerussite) + 2H2O; this process of darkeningis showed on the example of Pre-Romanesque murals in Slovakia (a),where some blackened parts of figures (b) were later retouched by car-bon black (c) – not further respecting the original coloring.

83

Merrifield (1967). Then, further works on pain-ting techniques with a general scope should bementioned, e.g. Eastlake (1960), Laurie(1967), Berger (1973).

To properly interpret the painting technique isthus not only needed to identify materials butalso to locate them in the layer sequence (Rosiet al. 2011). Heterogeneity of the sourcematerials from which the paint layers are madeis further important for understanding theirprovenance and/or properties and thus itshould not be lost or overlooked during theanalysis. To properly interpret these sourcematerials in the context of their use in the his-tory combining knowledge of analytical che-mistry and geosciences is needed.

On the other hand, the description of secon-dary-developed textures – e.g. corrosion andweathering rims, surface layers etc. is relevantfor better understanding of deterioration pro-cesses and virtual reconstruction of originalcolors (Fig. 5). A description of zones of degra-dation and explaining deterioration processes

is another popular field of recent investigationsof works of art (Monico et al. 2011).

3. Microanalysis of pigments

3.1. Mineral and synthetic pigments

Pigments in color layers of paintings areeither natural (mineral) applied simply aftermere mechanical pre-treatment and separa-tion not-affecting their chemical or phasecomposition (e.g. malachite, azurite, celado-nite, cinnabar, orpiment etc.), or manufactu-red - prepared by chemical reactions fromsubstantially different raw materials (e.g.smalt, lead white, verdigris).

In a relatively lower extent, also organic pig-ments are prepared from organic dyes by pre-cipitation on inorganic substrate (e.g. redorganic lakes). Gradually, a lot of mineral pig-ments have been substituted by their artifi-cial analogues (e.g. artificial ultramarine, arti-ficial orpiment, vermillion etc.), however, onthe other hand, some of them survived until


Fig. 6. Grain size and morphology is related to the origin of minerals. Aggregate of small uniform particles of synthetic titanium white(anatase – TiO2) used in modern paints since 1920 (a) vs. weathered grain of natural anatase in the same scale as a common admix-ture in clay-based pigments used since the pre-history (b); microfossils indicating a natural origin of the chalk (limestone) used inste-ad of ar tificially prepared calcium carbonate or carbonatizated lime (c); sharp and elongated crystals of neo-formed salts in porousspace of color layers as an example of secondary deterioration process (d).

84

the present days (e.g. clay-based pigments –“earths”). During the 19th century industrialproduction of new synthetic pigments (e.g.zinc white, synthetic alizarine etc.) has star-ted. Fundamental works on pigment origin,preparation, use and identification are givene.g. by Feller (1986), Roy (1993), Fitzhugh(1997) and Berrie (2007).

Natural and artificial pigments differ in contentof various admixtures, variable grain size andcrystallinity. These properties affect the opa-city and color fastness of the pigment (Fig. 6).

The knowledge about when and how the histo-rical pigments were being replaced by thesynthetic ones contributes to the dating ofpaint layers. Mineralogical analysis describingthe process of the formation of individualmineral pigments in nature helps to indicatethe regional provenance of painting materialsby comparing them with materials from refe-rence sources (Table 1).

Modifications of the pigments structures,different admixtures and contaminations ofthe product could also be related with diffe-rent fabrication processes and thus with dif-ferent workshops. To assign the authorshipof anonymous paintings the materialsanalysis should always be combined withthe description of ar tistic style within com-parative studies.

3.2. Methodology

Complete fingerprint of each painter can befolded from the material composition ofcolors (materials fingerprint), the method oftheir preparation, mixing and spreading inlayers (technological fingerprint) and artisticexpression (style fingerprint). The first two fin-gerprints provide extra information about theregion and the period of artwork origin.

Obtaining a complete fingerprint todayrequires costly transpor tation of the pain-ting for a long distances and extensivesampling - otherwise the evaluation ishighly subjective and it results in the gro-wing of the trade with forgeries. The costsof analyses, the transpor t of works of ar t,insurance costs, etc., which are usuallytoo high for regional museums, churchesand other owners, have their share on thegrowing interest in non-invasive analyticaltechniques. Despite a significant boom inthis field in last years, it must be saidthat without sampling the description ofauthor’s fingerprint will always be onlypar tial (Fig. 7).

Once samples are taken from the prestigiousworks of art, they become a part of a physi-cal archive and may not be lost or consumedduring their analysis. Therefore, the maineffort is directed to analyze a small amount


table 1. Comparison of mineralogical composition of five selected red clay-based grounds of paintings from the Czech collections andtheir attribution to three different source materials from historical mining deposits – Horenec in the Czech Republic (Bk), Kronach innorthern Bavaria (C) and yet unknown locality in northern Italy (Bs)

85

of a sample by the most methods in a non-destructive way to obtain maximum informa-tion without any loss of material. Sample pre-treatment has to be minimized and, ifapplied, it has to be acceptable for most or,ideally, all techniques which are necessaryfor complete examination of the sample. Thisimplies testing of particular methods, namelytheir detection limits, accuracy, and correct-ness of measurements because they can dif-fer significantly from those of the standardprocedures.

Routinely used laboratory methods for ins-tant investigation of fragments of paints andtheir cross-sections are light microscopy (LM)and scanning electron microscopy coupledwith energy dispersive spectrometry (SEM-EDS). They provide preliminary informationabout the morphology, stratigraphy, and ele-mental composition of paints. The semi-quan-titative elemental analysis can further beimproved by quantitative and trace elementsanalysis via particle induced X-ray emission(PIXE) and/or scanning electron microscopycoupled with wavelength dispersive spectro-metry SEM-WDS (Schreiner et al., 2007).

Focusing on the identification of inorganic crystalphases, X-ray diffraction-based techniques (XRD)are very effective tools for their indisputabledetermination (Svarcová et al. 2010), besidesmore frequently used spectroscopic methods

such as Raman and/or infrared spectroscopy(Vandenabeele et al. 2007, Bacci et al. 2001).The main disadvantage of Raman spectroscopyin the mineralogical analysis of paints is a highrisk of undesirable fluorescence promoted eitherby organic stuff (binders, sealing compoundsetc.) and/or other fine-grained materials suchas, for example, clays (Fig. 8).

X-ray diffraction, contrarily to Raman spectros-copy, enables indisputable identification ofcrystal phases, including double salts, hydra-tes, and unstable materials, using a huge num-ber of reference patterns available in an arrayof databases. Furthermore, XRD can be usedfor quantification of phases in mixtures.

Implementation of mono or poly-capillarysystems into the conventional diffractome-ters enables analysis of small volume sam-ples or even their selected parts without anydestructive pre-treatment (Svarcová et al.2010). Introduction of X-ray microdiffractioninto material characterization was made pos-sible by synchrotron radiation sources, whichare noted for high intensity X-ray beams ena-bling collection of diffraction data from smallareas (typically of the order of 0.001– 0.1mm) in a reasonable time (Dooryhée et al.2005). Laboratory equipments for routinepowder X-ray microdiffraction have becomemore frequent in last years and, as we alre-ady proved, the resulting data are generally


Fig. 7. X-ray fluorescencespectra (a) obtained bynon-invasive punctualmeasurement by hand-held equipment (b). Useof mixed Pb-Sb-Sn yellow(i.e. Pb2SbSnO6.5) inste-ad of pure Naples yellow(i.e. bindheimite -Pb2Sb2O7) is here indica-ted by a joint occurrenceof Sb and Sn in yellowcolor together with Zn –zinc white (ZnO); use ofthis uncommon pigmentis one of author’s finger-print of the 19th centurypainter Friedrich vonAmerling.

86

comparable with those obtained on synchro-tron-based microdiffraction (Svarcová et al.2011a). In our workplace, for example,PANalytical X’Pert PRO diffractometer isused, equipped with a conventional X-ray tube(CoK! radiation, 40 kV, 30 mA, point focus),a glass collimating mono-capillary with an exitdiameter of 0.1 mm, and a multichannel posi-tion-sensitive detector (X’Celerator) with ananti-scatter shield (Fig. 9).

3.3. Case studies

The authors have been dealing with mineralo-gical analysis of pigments systematically formany years and have published its results onregular basis.

The mineralogical analysis provides e.g. indi-cators which make it possible to discern pro-venance of various earthy pigments (contai-ning clay minerals) found in microsamples

(Hradil et al. 2003 and 2011).

Based on distinguishing of structural modifi-cation of bindheimite (Pb2Sb2O7) by Sn andidentification of its accompanying mineralsone can trace back procedures and raw mate-rials to produce various types of lead-basedyellows (Hradil et al. 2007).

Experimental mineralogical research has ledto the novel description of the crystal structu-re of basic copper acetate (verdigris) – a pig-ment used by painters since the Middle Ages(Svarcová et al. 2011b).

Practically no attention has been paid in lite-rature to description of clay structures in thetraditional painting pigments. Most articleson analyses of paintings are happy with thestatement that the painting contains anearthy pigment, iron ochre or red clay, withoutfur ther specifying their compositions.Meanwhile, there is a significant diversity ofsuch materials in terms of natural processesof their formation.

European Medieval paintings often contain inten-sely-colored materials, originally coming fromsurface layers of just uncovered polymetallic oredeposits, where clay minerals were practicallymissing and for which the technical term “earth”should not be even used. On the other hand,these raw materials often contained specificadmixtures of metals (e.g. Pb, As or Zn), whichmay help to identify their origin.


Fig. 8. Raman spectra ofnatural kaolinite fromGeorgia (U.S.A.), where onlythe peaks of admixed anata-se (A) are clearly distinguis-hable; without additionalinformation this spectracould be misinterpreted astitanium white (adopted fromKosarová, 2011).

Fig. 9. Powder X-ray diffraction equipment at the Institute of InorganicChemistry of the AS CR in Rez, Czech Republic.

87

Within detailed mineralogical research ofselected Gothic altarpieces in Nor thernSlovakia and Lesser Poland it was, for exam-ple, found that the original iron red containsgoethite, hematite, calcite, and Zn-dolomite.Zinc containing dolomite represents a finger-print indicating that the source material wasmined in the oxidation zone of the carbonate-hosted Zn–Pb ore deposit near Cracow,Poland, which has been active since the 12th

century. Thus one could conclude that worksof the so-called Master of Matejovce altarpie-ce workshop can be recognized not only bytheir conservative style known as the ‘Cracowschool’ of painting but also by very characte-ristic materials features indicating the autho-r’s preference to use pigments from regional

sources, differently to other workshops of thesame period and place (Hradil et al. 2008)(Fig. 10).

In the case of mineral pigments, which wereformed by weathering or other alteration ofsilicate rocks, the decisive factor is thecorrect identification of clay structures andtheir closely associated phases, e.g. iron oxi-des. The main reason why such informationis often missing is particularly the limitedavailability of material from the painting,which reduces the possibility to apply stan-dard procedures for clay investigation – sepa-ration and orientation of particles, exchangeof interlayer cations etc. (Moore andRaynolds 1997).


Fig. 10. X-ray intensity maps of selected elements in the original beige layer of the background, toned by red grains (Altarpiece ofMatejovce, right wing); X-ray intensities correspond to relative concentrations of individual elements within the measured area, but theconcentration maxima indicated by the highest intensity are different for each element (Ca maximum 100 wt%, Fe maximum 60 wt%,Mg maximum 45 wt%, Pb maximum 60 wt%, Zn maximum 5 wt%); grain +1 represents iron oxides with some admixtures of Zn-bearingminerals, and grain +2 is Zn-dolomite.

88

Completely unaddressed remains the issue ofreaction of expandable clay structures in the pain-ting layer with humidity and organic binders,which result in organo-clay complexes, both inpainting and preparatory layers. However, theirstability in the painting may be very significantfrom the viewpoint of preservation of the pain-tings. The only organo-clay historical pigment,which has been carefully investigated, is the so-called Maya blue – a very stable complex of paly-gorskite and natural indigo (Doménech et al.2009; Giustetto et al. 2011). This pigment, howe-ver, has not been found in European paintings.

Results of last investigations demonstratedhow a correct description of basal diffrac-tions of clay minerals in the low-angle regionleads to a direct determination of provenanceof anonymous 17th century paintings (Hradilet al. 2010), or, how the source locality of thewhite earth (kaolin) used in the preparatorylayer of Bohemian Gothic murals can be indi-cated by a calculation of the kaolinite crysta-llinity index (Svarcová et al. 2011a). Widerange of other examples could be given.

4. Conclusion

An estimated 90% of tangible cultural herita-ge is not studied by modern scientificmethods. Often (and often unnecessarily)only the famous works of art in famous galle-ries are subjected to detailed investigations.

Is this correct? As mentioned in this paper,

the way to reduce this disparity is the currentdevelopment of non-invasive analytical techni-ques that allow obtaining exact data in situ,even in distant places - regional galleries,rural churches, etc.

In the restoration/conservation practice the aim-less collection of analytical data still prevailsover the targeted analyses. Each expertise in thefield of cultural heritage requires its make-tomeasure methodology reflecting the diversity ofmaterials as well as variety of tasks. When appl-ying this approach, targeted mineralogical analy-sis of paint layers could directly lead to datingand to determining the provenance of anony-mous works of art.Often we also hear the opinion that artworks

must first to be preserved and then (when enoughmoney is left) discovered and explored. But is itnot rather the opposite? Material properties we donot keep forever, they decompose sooner or later,and we can only slow down this process. But theywere created for our inspiration and knowledge,which is not a subject of any further decay. First ofall, this knowledge itself is a substantial part ofour cultural heritage. And a respect to the kno-wledge is something what we often lack.

5. Acknowledgement

The authors are grateful to all their colleaguesfrom laboratory ALMA as well as to all friendlycollaborating restorers for fruitful cooperationin this research. Financial supports by theMinistry of Education of the Czech Republic


Fig. 11. Interstratifiedillite / smectite indica-ted by interplanar dis-tance of ca 1.31 nm atthe diffraction patternof reference Italian17th century paintingby G.B. Langetti andon one anonymouspainting; this clay fin-gerprint together withalmost the same mine-ralogical compositionof the ground helpedto locate the anony-mous painting to theregion of northernItaly.

89

(RVO 61388980), and Ministry of Culture ofthe Czech Republic (DF12P01OVV048) is kindlyacknowledged.

5. References

Bacci, M., Fabbri, M., Picollo, M. & Porcinai, S.(2001) Non-invasive fibre optic Fourier transform-infrared reflectance spectroscopy on painted layers- Identification of materials by means of principalcomponent analysis and Mahalanobis distance.Anal Chim Acta, 446, 15–21.

Berger, E. (1973) Quellen fur Maltechnik wahrendder Renaissance und deren Folgezeit (16.– 18.Jahrhundert) in Italien, Spanien, den Niederlanden,Deutschland, Franreich und England: Nebst dem deMayerne Ms. Reprinted from 1901. Martin SandigoHG, Walluf b. Wiesbaden.

Berrie, B. (Ed.)(2007) Artists’ Pigments, vol. 4.,Archetype publications, London.

Cardell, C., Guerra, I., Romero-Pastor, J., Cultrone,G. & Rodríguez-Navarro, A. (2009) Innovative analy-tical methodology combining micro-X-ray diffraction,scanning electron microscope-based mineralmaps, and diffuse reflectance infrared Fouriertransform spectroscopy to characterize archeologi-cal artifacts. Analytical Chemistry, 81, 604-611.

Cennino, C. (1978) Il Libro dell’ Arte. Thompson Jr.,D.V. (translator), The Craftsman’s Handbook.Reprinted from 1437. Dover Publications, NewYork.

Doménech, A., Doménech-Carbó, M.T., del Rio,M.S. & Pascual M.L.V.D. (2009) Comparative studyof different indigo-clay Maya Blue-like systemsusing the voltammetry of microparticles approach.Journal of Solid State Electrochemistry, 13/6, 869-878.

Dooryhée, E., Anne, M., Bardiès, I., Hodeau, J.L.,Martinetto, P., Rondot, S., Salomon, J., Vaughan,G.B.M. & Walter, P. (2005) Non-destructivesynchrotron X-ray diffraction mapping of a Romanpainting Appl Phys A, 81, 663–667.

Eastlake Sir, Ch.L. (1960) Methods and Materialsof the Great Schools and Masters. Reprinted from1847. Dover Publications, New York.

Feller, R.L. (Ed.) (1986) Artists’ Pigments, vol. 1.,Archetype publications, London.

Fitzhugh, E.W. (Ed.) (1997) Artists’ Pigments, vol.3., Archetype publications, London.

Giustetto, R., Levy, D., Wahyudi, O., Ricchiardi, G.& Vitillo, J.G. (2011) Crystal structure refinementof a sepiolite/indigo Maya Blue pigment usingmolecular modelling and synchrotron diffraction.European Journal of Mineralogy, 23/3, 449-466.

Grieten, E. & Casadio, F. (2010) X-ray fluorescenceportable systems for the rapid assessment of pho-tographic techniques in notable art collections: theAlfred Stieglitz Collection. X-Ray Spectrometry,39/3, 221-229.

Hradil, D., Grygar, T., Hradilová, J. & Bezdicka, P.(2003) Clay and iron oxide pigments in the historyof painting. Applied Clay Science, 22/5, 223-236.

Hradil, D., Fogas, I., Miliani, C. & Daffara, C.(2006) Study of late 18th – 19tth century paintingsof „Vienna school“ by non-invasive analyticalmethods. Technologia Artis 2006 – Proceedings ofthe 1st conference of ALMA in Prague, 20-28. /InCzech/

Hradil, D., Grygar, T., Hradilová, J., Bezdicka, P.,Grünwaldová, V., Fogas, I. & Miliani, C. (2007)Microanalytical identification of Pb-Sb-Sn yellow pig-ment in historical European paintings and its diffe-rentiation from lead tin and Naples yellows. Journalof Cultural Heritage, 8/4, 377-386.

Hradil, D., Hradilová, J., Bezdicka, P. & Svarcová, S.(2008) Provenance study of Gothic paintings fromNorth-East Slovakia by hand-held XRF, microscopyand X-ray microdiffraction. X-ray Spectrometry, 37,376-382.

Hradil, D., Hradilová, J. & Bezdicka, P. (2010) Newcriteria for classification of and differentiation bet-ween clay and iron oxide pigments of various ori-gins. Acta Artis Academica 2010 – Proceedings ofthe 3rd interdisciplinary conference of ALMA,Prague, 107-136.

Hradil, D., Písková, A., Hradilová, J., Bezdicka, P.,Lehrberger, G. & Gerzer, S. (2011): Mineralogy ofBohemian green earth and its microanaytical evi-dence in historical paintings. Archaeometry, 53/3,


90

563-586.

Chiari, G. (2008) Saving art in situ. Nature, 453/8,159.

Jong,e M.D. & Vogt, S. (2010) Hard X-ray fluores-cence tomography —an emerging tool for structu-ral visualization. Current Opinion in StructuralBiology, 20, 606-614.

Kosarová, V. (2011) Characterisation of clay andiron-based pigments in paint layers by diffraction andspectroscopic methods. Diploma thesis, Masarykuniversity, Faculty of science, Brno. /In Czech/

Martin, M.C., Schade, U., Lerch, P. & Dumas, P.(2010) Recent applications and current trends inanalytical chemistry using synchrotron-basedFourier-transform infrared microspectroscopy.Trends in Analytical Chemistry, 29, 453-463.

Laurie, A.P. (1967) The Painter’s Methods andMaterials. Dover Publ., New York.

Mazzeo, R., Roda, A. & Pratti, S. (2011) Analyticalchemistry for cultural heritage: a key discipline inconservation research. Anal Bioanal Chem, 399,2885–2887.

Merrifield, M.P. (1967). Original Treatises on theArts of Painting. Dover Publications, New York.

Miliani, C., Rosi, F., Brunetti, B.G. & Sgamellotti, A.(2010) In Situ Noninvasive Study of Artworks: TheMOLAB Multitechnique Approach. Accounts ofChemical Research, 43/6, 728-738.

Monico, L., Van der Snickt, G., Janssens, K., DeNolf, W., Miliani, C., Verbeeck, J., Tian, H., Tan, H.,Dik, J., Radepont, M. & Cotte, M. (2011)Degradation Process of Lead Chromate inPaintings by Vincent van Gogh Studied by Means ofSynchrotron X-ray Spectromicroscopy and RelatedMethods. 1. Artificially Aged Model Samples, 2.Original Paint Layer Samples. Analytical Chemistry,83, 1214-1231.

Moore, D.M. & Raynolds, R.C. (1997) X-RayDiffraction and the Identification and Analysis ofClay Minerals, 2nd ed. Oxford Univ. Press, NewYork.

Palmieri, M., Vagnini, M., Pitzurra, L., Rochci, P.,

Brunetti, B.G., Sgamellotti, A. & Cartechini, L.(2011) Development of an analytical protocol for afast, sensitive and specific protein recognition inpaintings by enzyme-linked immunosorbent assay(ELISA). Anal Bioanal Chem, 399, 3011–3023.

Rosi, F., Federaci, A., Brunetti, B.G., Sgamellotti,A., Klementi, S. & Miliani, C. (2011) Multivariatechemical mapping of pigments and binders in easelpainting cross-sections by micro IR reflection spec-troscopy. Anal Bioanal Chem, 399, 3133-3145.

Roy, A. (Ed.) (1993) Artists’ Pigments, vol. 2.,Archetype publications,, London.

Schreiner, M., Melcher, M. & Uhlir, K. (2007)Scanning electron microscopy and energy dispersi-ve analysis: applications in the field of cultural heri-tage. Anal Bioanal Chem, 387, 737–747.

Svarcová, S., Kocí, E., Bezdicka, P., Hradil, D. &Hradilová, J. (2010) Evaluation of laboratory pow-der X-ray micro-diffraction for applications in thefield of cultural heritage and forensic science. AnalBioanal Chem, 398, 1061 – 1076.

Svarcová, S., Bezdicka, P., Hradil, D., Hradilová, J.& Zizak, I. (2011a) Clay pigment structure charac-terisation as a guide for provenance determination– a comparison between laboratory powder micro-XRD and synchrotron radiation XRD. Anal BioanalChem, 399/1, 331-336.

Svarcová, S., Klementová, M., Bezdicka, P., Rasocha,W., Dusek, M. & Hradil, D. (2011b) Synthesis andcharacterization of single crystals of the layered cop-per hydroxide acetate Cu2(OH)3(CH3COO)•H2O.Crystal Research and Technology, 46/10, 1051-1057.

Vandenabeele, P., Edwards, H.G.M. & Moens, L.(2007) A decade of Raman spectroscopy in art andarcheology. Chem Rev, 107, 675–686.

Yonehara, T., Yamagochi, M. & Tsuji, K. (2010) X-ray fluorescence imaging with polycapillary X-rayoptics. Spectrochimica Acta Part B, 65, 441–444.


91

Binders in historical buildings:Traditional lime in conservation/ Carlos Rodríguez-Navarro

Abstract

Mankind has used many types of binders throughout history: from mud in vernacular earthen archi-tecture to current high-performance cement. However, gypsum and lime have been by far the mostcommon and important binders in historical buildings. Their use expands several millennia and isrepresented in the architectural feats of all kind of civilizations. Here, the history, technology anduses of gypsum- and lime-based binders are reviewed. We focus on the study of the different stepsof the gypsum and lime cycles as a conceptual framework to underline past applications and perfor-mance of these binders, and their current use in architectural conservation. Special emphasis is paidto the technology and uses of lime as a versatile and compatible conservation material.

Resumen

La humanidad ha usado muchos tipos de ligantes a lo largo de la historia: desde el barro en la arqui-tectura vernácula de tierra a los actuales cementos de altas prestaciones. Sin embargo, el yeso yla cal han sido, con diferencia, los ligantes más comunes en edificaciones históricas. Su uso seextiende durante varios milenios y está presente en los hitos arquitectónicos de todo tipo de civili-zaciones. Se revisan aquí la historia, tecnología y usos de los ligantes basados en el yeso y la cal.Nos centramos en el estudio de las diferentes etapas en los ciclos del yeso y la cal, como un entra-mado conceptual para destacar las aplicaciones y comportamientos de estos ligantes en el pasadoy su actual uso en la restauración arquitectónica. Se hace un especial énfasis en la tecnología y losusos de la cal como un material de restauración versátil y compatible.

Key-words: Binders, lime, portlandite, gypsum, calcite, mortars, plasters, historical buildings, carbona-tion, hydration, topotactic.

1. Introduction

A binder is defined as a material which acts as a glue or cement when mixed with an aggre-gate (e.g., sand) and water to form a fresh plaster, render, mortar or concrete. Following set-ting and hardening, such composite materials play a structural and/or decorative role in a buil-ding (Lea 1970). Since the advent of sedentism and building technology, which could be tra-ced back to the Pre-Pottery Neolithic A in South-West Asia (e.g., Jordan river valley), ca.11,700 to 10,500 years ago (Finlayson et al. 2011), mankind has used many different bin-ders. Mud (clay-rich earth) was among the first binders used (Houben and Guillaud 1994) asexemplified by the adobe walls of Jericho (Israel) dated to 8300 BC (Allen and Tallon 2011) orthe earthen structures (rammed earth) in Çatalhöyük (Anatolia, Turkey) dated to 6000-7500BC (Mellaart 1967).

However, the use of this type of binder does not involve pyrotechnology (i.e., no heat tre-atment was required). As a result, ear then structures display a low strength and areprone to weathering and damage. With the advent of pyrotechnology ca. 12000 yearsago (Kingery et al. 1988), a new class of binders emerged with superior proper ties interms of strength and durability, as well as versatility and applicability. Among them,gypsum and lime have played an outstanding role in building histor y from the Neolithicuntil nowadays.

Dpto. Mineralogía y Petrología, Universidad de Granada, Fuentenueva s/n, 18002 Granada

92

Here, the history, technology and uses ofgypsum and lime binders will be reviewed.Special emphasis will be paid to the case oflime, as this has been the binder most profu-sely and ubiquitously used by mankind.Furthermore, lime has recently emerged asthe binder of choice for architectural conser-vation. Therefore, its current use in heritageconservation will also be outlined.

2. Gypsum binders

2.1. Gypsum as a building and decorativematerial

Gypsum-based mortars and plasters havebeen used since ancient times, especiallyto cover masonry, as decorations, or as asupport for mural paintings (Elsen 2006).Due to their setting upon addition of water,gypsum-based binders could be consideredthe first “hydraulic binder” used by mankindsince ca. 9000 years before present(Gartner 2009).

Gypsum was the most common binder inAncient Egypt both for masonry and decorativepurposes since Pharaonic times (Lea 1970;Lucas and Harris 1962). Lucas and Harris(1962) report the use of gypsum in severaltombs and as mortar and plaster in the Gizapyramid. Other studies have confirmed its pre-sence in mortars from the pyramids of Medium(2600 BC), Cheops (2500 BC) and Unas(2250 BC) (Regourd et al. 1988).

Gypsum plasters were commonly used in theMiddle East and in countries around theMediterranean basin, especially during theMiddle Ages (e.g., Islamic Architecture) (Elsen2006). In the area around Paris, gypsum mor-tars were thoroughly used in gothic buildingssuch as the Cathedrals of Chartres andBourgues (Adams et al. 1992), which mayexplain why gypsum-based binders are knownas “Plaster of Paris”. Despite its low strengthand poor durability in humid environments,examples of gypsum mortar applications innorthern Europe are numerous. For instance,

Binders in historical buildings: Traditional lime in conservationCarlos Rodríguez-Navarro

Fig. 1. Gypsum-based “mocárabes” in the Corral del Carbón, a XIV c. Islamic building in Granada (Spain).

93

Kawiak (1991) describes the use of gypsummortars in a XII c. church in Wislica, Poland.

Up until nowadays gypsum has been appliedeither as mortar and plaster for surface finis-hing, or as cast and/or carved decorations(Arredondo 1991 Luxán et al. 1995). A signi-ficant example of this latter application is thecase of the “mocárabes” and decorated plas-ters (“yeserías”) from the Alhambra and otherIslamic buildings in Granada (Spain) (Fig. 1).

At the Alhambra (and by extension in mostMedieval Islamic architecture) two maingypsum-types were used:

• “black gypsum” (yeso negro), made ofhighly impure hemihydrate (50-60 wt %)containing significant amounts of ashesproduced during calcination, and

• “white gypsum” (yeso blanco) a purer mate-rial made of > 66 wt % hemihydrate(Rubio-Domene 2006).

Gypsum renderings were extensively appliedas a base for mural painting. In all thesecases the major component after setting was

gypsum, although in some ancient mortarsand plasters, anhydrite has been also found(Regourd et al. 1988) along with calcite(Kawiak 1991). The presence of anhydriteposes the question about the possible originof anhydrous calcium sulfate after dehydra-tion of gypsum at relatively low T (< 30 °C)and RH (<< 70 %) (Charola et al. 2008).Conversely, it could be argued that anhydritecrystals are unhydrated relicts of the originalbinder prepared following calcination at highT (> 350 °C) of gypsum mixed with calcite,what is known as “Keene’s cement” or“Estrich Gips” (Sayre 1976; Kawiak 1991).

A variation of plaster of Paris is the so-called“scagliola” which is a composite material pre-pared by mixing soft-burned gypsum (hemihy-drate or bassanite, see section 2.2), glueand pigments with water. Such a compositematerial has been used as an imitation ofmarble, thereby its name “stucco marble”. Tosome authors, however, scagliola is agypsum-based plaster made of purer andfiner hemihydrate particles, which typicallymakes the set material whiter than normalgypsum plasters (Rubio-Domene 2006).Rubio-Domene (2006) indicates that the so-


Fig. 2. The gypsum cycle. Scanning electron microscopy (SEM) images show examples of lens-shaped gypsum prior to dehydration, bas-sanite phases formed after dehydration (modified from Singh and Middendor f, 2007) and gypsum in a set plaster.

94

called “stucco” is made by mixing inert anhy-drite with glue.

2.2. Calcium sulfate phases

The common calcium sulfate minerals aregypsum (CaSO4•2H2O, monoclinic, I2/c),bassanite (CaSO4•0.5H2O, trigonal, P312),and anhydrite (CaSO4, orthorhombic, Amma).Here we report the space groups compiled inHawthorne et al. (2000); note however thathemihydrate and anhydrite form differentpolymorphs (see below).

Gypsum is the most abundant sulfate mineralon the Earth surface and has a very widespre-ad occurrence in evaporite deposits and othersediments, in fumaroles, and in ore deposits(Vaniman and Chipera 2006). Evaporation ofsea water initially leads to gypsum precipita-tion, followed by the crystallization of anhydri-te (CaSO4) in the residual brine due to a reduc-tion in water activity. Anhydrite and bassanitealso occur in situations where thermal activityhas naturally dehydrated precursor gypsum, orin acid alteration environments (see review byCharola et al. 2007).

2.3. The gypsum cycle

The use of calcium sulfate as a binder invol-ves a series of steps which form the so-called“gypsum cycle” or “calcium sulfate cycle”(Fig. 2).

The first step consists in the thermal dehy-dration of gypsum to form a hemihydrate (themineral bassanite). The second and final stepof this cycle involves the hydration and set-ting of bassanite to form gypsum which actsas a binder. Below we will study these twosteps separately.

2.3.1. The dehydration of gypsum

The use of gypsum as a binder involves itsthermal treatment to form the hemihydrate(“plaster of Paris”), according to the dehydra-tion reaction:

CaSO4•2H2O # CaSO4•0.5H2O + 1.5H2O (1)

This reaction starts at 42°C (315 K) (i.e.,

equilibrium transition T between bassaniteand gypsum) at standard atmospheric pres-sure (Charola et al. 2007). Because at this Tthe reaction is extremely slow, T > 373 K isnormally required in order to accelerate thedehydration kinetics. Industrial plaster ofParis is produced under humid conditions atca. 413 K (Charola et al. 2007). Dependingon the dehydration procedure, two hemihydra-te varieties are obtained, !- or $-CaSO4•0.5H2O. Heating gypsum in a humidatmosphere produces the first variety, whileheating in a dry atmosphere results in thesecond variety (Freyer and Voigt 2003; Singhand Middendorf 2007).

Under near-equilibrium conditions, heating to~373-473 K (dehydration T depends on thetype of precursor bassanite: Freyer and Voigt2003) produces %-CaSO4 (also known asCaSO4(I)), according to:

CaSO4•0.5H2O # CaSO4 + 0.5 H2O (2)

The anhydrous product is called soluble anhy-drite, a metastable phase which rapidly rehy-drates under normal atmospheric conditions.As in the previous step, some overstepping (Thigher than the equilibrium decomposition T)is required to speed up the reaction.

Further heating at 633 K leads to the forma-tion of anhydrite ($&CaSO4 or CaSO4(II))which does not react with water (Prasad et al.2005). Heating of anhydrite at 1453 Kresults in the formation of !&CaSO4(CaSO4(I)) (Charola et al. 2008).

Le Chatelier (1887) was the first to show thetwo steps dehydration of gypsum. Since hispioneering work, much has been publishedon the mechanisms and kinetics of gypsumthermal decomposition (see reviews by Freyerand Voigt 2003; Charola et al. 2008).

There is however no consensus on the actualmechanism of thermal decomposition ofgypsum. For instance, the two-step dehydra-tion process has been challenged by Prasadet al. (2005) that reported direct dehydrationof gypsum to form %&CaSO4, followed by itsrapid hydration to yield bassanite. There isalso a persisting controversy regarding the


95

existence of several hydrated (metastable)phases with formula CaSO4•nH2O, where theactual value of n is not well established forsome of these phases (see Adam 2003).

Crystallographic studies (using XRD and/orneutron diffraction) show that the thermaldecomposition of gypsum crystals most pro-bably follows a topotactic mechanism (Abriel etal. 1990). Topotaxy can be defined as a solid-state chemical change in which the reactantprovides a template for product phase genera-tion (Galwey 2000). This means that there is aclear structural relationship between precursorand product phases due to a good matchingbetween their crystallographic structures. Insitu TEM-SEAD analyses have confirmed thatthe dehydration reaction is in fact topotactic(Sipple et al. 2001).

2.3.2. The hydration of bassanite

The setting and hardening of plaster of Parishas been the subject of extensive researchover the last 100 years, although it is a pro-cess not fully understood (Singh andMiddendorf 2007). Hemihydrate hydrationfollows the overall (exothermic) reaction:

CaSO4•0.5H2O + 1.5H2O # CaSO4•2H20 (3)

Both the kinetics and properties of setgypsum are dependent on the type of bassa-nite precursor phase (!& or $&CaSO4•0.5H2O). The !&phase shows a shor-ter induction period and produces a set plas-ter with higher strength than the $&phase(Singh and Middendorf 2007).

Most researchers agree that the hydration ofbassanite is a solution mediated process (Singhand Middledorf 2007). In contact with water,bassanite crystals start to dissolve generating asaturated solution with respect to this phase.Bassanite is more soluble than gypsum: 0.65 wt% calcium sulfate hemihydrate dissolves in waterat room T, against 0.2 wt % in the case ofgypsum. The resulting solution rapidly reachessupersaturated with respect to gypsum leadingto its crystallization.

Precipitation of gypsum reduces the activity ofboth Ca2+ and SO4

2- in solution, which further

promotes the dissolution of the remaining bas-sanite and the precipitation of more gypsum.Typically, newly formed gypsum appears asµm-sized acicular crystals which form a highlyporous, yet strong, 3D-mesh structure (Singhand Middlendorf 2007). Both the morpho-logy/size of gypsum crystals and the kinetic ofthe hydration reaction can be modified by theaddition of different inorganic and organiccompounds (Solberg and Hansen 2001). Forinstance, addition of Ca(OH)2 accelerates set-ting (Ridge and King 1976) whereas additionof carboxylic acids (and their salts) have beenshown to delay hemihydrate setting, leading tothe formation of larger equant gypsum crystalsas opposed to the fibrous interlocking crystalsin additive-free set pastes (Singh andMiddlendorf 2007; Gartner 2009). As aresult, the flexural and compressive strengthof gypsum plaster is significantly reduced(Lanzón and García-Ruiz 2012).

In addition to precipitation in solution, Melikhovet al. (1991) indicate that a topochemical reac-tion is responsible for the formation of gypsumpseudomorphs after bassanite. This is consis-tent with the reported pseudomorphic transfor-mation of hemihydrate into gypsum in a watervapor atmosphere (Triollier and Guilhot 1976), aprocess which appears to be topotactic. Notehowever that a tightly interface-coupled dissolu-tion-precipitation reaction could also explain theformation of gypsum after a pseudomorphicreplacement of bassanite. For details on such areplacement process see the recent review byPutnis (2009). Overall, these observations sug-gest that the mechanical properties of the setmaterial may strongly depend on the textural pro-perties of the precursor bassanite (in addition tothe curing and setting conditions, as well as theabsence/presence of additives and impurities).

2.4. Deterioration and conservation ofgypsum mortars and plasters

Gypsum plasters and mortars are prone todecay (Cotrim et al. 2008). This is primarily dueto the low strength of gypsum, which has a hard-ness of 2 in the Mohs scale, and its relativelyhigh solubility (2.1 g/L at 20°C). Gypsum plas-ters and mortars are therefore highly suscepti-ble to chemical weathering (dissolution) andloss of structural stability, particularly in humid


96

environments and outdoors exposure. Althoughthey undergo very similar decay problems asthose of lime mortars, they are however lesssensitive to air pollutants (Sayre 1976).

Gypsum can undergo deliquescence and repre-cipitation under fluctuating relative humidity(RH) conditions. Its equilibrium relative humidity(RHeq) at 25°C is 99.9 %, so when the RH ishigher than its RHeq, gypsum crystals dissolve.When the RH drops, crystals reprecipitate as anon-coherent powder, and may even inducecrystallization damage (salt weathering) whenthey grow within the pores of adjacent buildingmaterials (mortar, bricks or stone) (Charola etal. 2007). At T > 42°C, transition to anhydritecan take place, thereby resulting in a significantweakening of the material. Such a transition canalso occur at lower T if RH < RHeq (Charola etal. 2007).

Little has been published on the conservation ofgypsum plasters and mortars. In general, lostmaterial tends to be replaced with new gypsum-based materials. Consolidation and protectionof this type of building and decorative materialhas been performed using the same approa-ches used for other materials such as stone orlime mortars (Sayre 1976; Ashthur 1990;Cotrim et al. 2008). Much research is neededto develop and apply specific conservation tre-atments for gypsum-based materials. VanDriessche et al. (2012) recently reported thedirect precipitation of nanosized bassanitewhich eventually transforms into gypsum atroom T. This technology offers a new, potentiallyeffective way to consolidate gypsum-based buil-ding materials via the application of a suspen-sion of nanosized bassanite. When transformedinto gypsum, it could act as a consolidant in asimilar manner as nanolimes applied on porousbuilding materials (see section 3.5).

3. Lime binders

3.1. Lime-based building materials: defini-tions, history and uses

The term lime s.l. refers both to calcium oxide(CaO) or quicklime, the product of the calcina-tion of calcium carbonate (calcite or aragonite:CaCO3), and to the compound obtained afterthe hydration of the oxide, i.e. calcium hydroxi-

de (Ca(OH)2), the mineral portlandite, alsoknown as slaked lime or hydrated lime(Boynton 1980; Gárate Rojas 1994). This termalso applies to the products of the hydration ofCa and Mg oxides formed after the calcinationof magnesium limestone and, in particular,dolomite (CaMg(CO3)2). Calcitic lime is com-monly known as fat lime, while dolomitic ormagnesian limes are commonly called magrelimes (Cowper 1927; Gárate Rojas 1994).

Archaeological evidence shows that lime wasused in the construction of some of the floorsand paving of the ruins excavated inÇatalhüyück (Mellaart 1967), dated between10000 and 5000 BC (Von Landsberg 1992;Kingery et al. 1998). These archaeological fin-dings, along with the remains of 4500 years oldlime kilns found in Khafaje, Mesopotamia, con-firm that lime was a common building materialin the Levant during the Neolithic (Davey 1961).

The Egyptians also used lime as a binder.Some coatings of lime in different pyramidshave been dated ca. 4000 BC (Boynton 1980).However this is challenged by Lucas and Harris(1962) and Ghorab et al. (1986) who indicatethat the Egyptians did not use lime in construc-tion (they used gypsum) until Roman times.Other ancient civilizations, like India, China andthe different cultures of pre-Columbian America(e.g., Mayans and Aztecs) systematically usedlime as a building material (Gárate Rojas1994).

The type of lime first used hardened when expo-sed to air, and was called air lime. The Greekand Roman civilizations discovered that calci-nation of marly limestones, i.e., with a concen-tration of aluminosilicates (clays) > 10 wt %,yielded a binding material that hardened under-water (hydraulic setting) and had improvedmechanical properties (Malinowski 1981).

Upon calcination of impure limestones, claysdehydroxylate at 400 to 600°C. The resultingsilica and alumina combine with CaO formedafter the decomposition of CaCO3 at 950 to1250º C, to produce calcium aluminates andsilicates (Callebaut et al. 2001). These limesare called natural hydraulic limes. Dicalciumsilicate (C2S) is the main phase that reacts withwater causing their hydraulic setting, unlike in


97

the case of cement where tricalcium silicate(C3S) is the main hydraulic phase (Callebout etal., 2001).

In addition to natural hydraulic lime, the so-called artificial hydraulic limes have been alsoused. They were discovered by the Phoeniciansand perfected by the Greeks and the Romans.Artificial hydraulic limes were obtained bymixing lime with a pozzolanic material, nametaken from the town of Puzzoli, located nearbythe Vesuvius in Italy, where a tuff with highhydraulic capacity was extracted. A pozzolanicmaterial contains highly reactive silica and alu-mina. When combined with Ca(OH)2 in the pre-sence of water generates new products, mainlyhydrated calcium silicates and aluminates, withsuperior binding or cementing properties(Mertens et al. 2009).

The Greeks used the so-called "Santoriniearth”, a tuff, as a puzolanic additive mixedwith lime for the manufacture of hydraulic limemortars as those found in Thera (Santorini,Greece) (Alejandre Sánchez 2002). TheGreeks also used a technique called "polis-hing" consisting in the application of coatingsmade of lime and crushed limestone mixedwith pozzolana (Malinowski 1981; AlejandreSánchez 2002). The Romans used lime inconstruction since the last two centuries of theRepublic (200-100 BC). In addition to air lime,they routinely used lime mixed with either natu-ral (pozzolana s.s.) or artificial (brick powder)pozzolanic materials, thus obtaining the well-known opus ceamentitious and the cocciopes-to described by Vitruvius (30 BC).

After the fall of the Roman Empire, natural andartificial hydraulic limes, including Romancement, together with traditional air lime werethe most common binders in constructionsince Byzantine time, though the Middle ages,Renaissance and Baroque, until the discoveryof Portland cement in the early 19th century byAspdin (Lea 1970).

However, since mid-18th century, traditionallimes began to be replaced by high performan-ce artificial hydraulic limes, made by resear-chers such as Smeaton (1791) or Vicat (1837).In 1824 Aspdin patented a process for obtai-ning a cement which hardened under water

(hydraulic setting) and acquired the strength ofPortland limestone, one of the most resistantstoned used in British architecture, hence thename "Portland cement". See details on suchpioneering works in the report by Haswell(1865). The homogeneity and uniform proper-ties, easy of application and high mechanicalstrength after a rapid setting explains theimmediate success and massive use ofPortland cement until present day.

However, in recent decades, lime has re-emerged as the optimal material for the con-servation of historic structures and buildings(Teutonico et al. 1994), replacing cement insuch interventions. There are several rea-sons for the revival of lime (Elert et al. 2002).Compared with cement, lime is more compa-tible from a mechanical, physical and chemi-cal point of view when applied in historicalstructures (brickworks and/or stone) (Elert etal. 2002; Lanas and Alvarez 2003; Hansen et


Fig. 3. Examples of the high incompatibility between Portland cementand historic structures: a) Portland cement applied on serpentine andmarble at the Chancilleria (Granada, Spain); b) salt efflorescence anddetachments associated to Portland cement applied on limestoneashlars (Granada Cathedral, Spain).

98

al. 2008). For instance, the absence of alka-lis and sulfates in lime prevents salt weathe-ring due to the formation of deleterious car-bonates and sulfates of Na/Ca or Mg, as canbe seen after the application of Portlandcement (i.e., Rodríguez-Navarro et al. 1998;Maravelaki-Kalitzaki et al. 2003; Hansen etal. 2008). Crystallization of such salts cau-ses extensive damage in both new construc-tion and historic structures (Fig. 3).

In successive sections recent advances inthe study of air limes are presented. We willfocus on the case of high calcium or calciticlimes as these have been the most profuselyused by mankind and tend to be preferablyapplied in architectural conservation (Hansenet al. 2008).

The most important aspects that determinelime properties, reactivity, applicability andsuitability as a binder in conservation will bestudied within the framework of the "limecycle". Finally we will discuss aspects rela-ted to recent studies on the characteristicsof historic lime mortars, and the design andtesting of conservation mortars as well asmethods for their actual consolidation andprotection.

3.2. The lime cycle

The lime cycle consists of three stages (Fig. 4):• calcination; • hydration or slaking; and • carbonation (or air setting).

Below we will present a detailed description ofeach stage of this cycle.

3.2.1. Calcination

The first stage of the lime cycle involves thecalcination of limestone at a sufficiently hightemperature as to cause the decomposition ofcalcite according to the reaction:

CaCO3 # CaO + CO2' (4)

This reaction is strongly endothermic, so theproduction of calcium oxide requires highenergy consumption and the use of suitablekilns. Traditional lime kilns involved the firingof wood or coal along with limestone blocks,resulting in lime mortars with a high contentof ash (Luxán and Dorrego 1996). Kilns usedsince the industrial revolution placed the fueland the limestone in separate compartments(Boynton 1980). Thus, the resulting limes do


Fig. 4. The lime cycle: Electron microscopy images show reactant and products of the different stages of the cycle. ACC: AmorphousCalcium Carbonate.

99

not present any ashes and have homogene-ous characteristics, something difficult toachieve through the use of traditional kilns.

The temperature of dissociation of CaCO3 is898°C at pCO2 = 1 atm. This T is reduced thelower the pCO2 is, and increases as pCO2 incre-ases (Boynton 1980). CO2 released by carbona-te decomposition (plus CO2 generated by thecombustion of wood or coal in traditional kilns)must be evacuated to avoid the re-carbonationof CaO, or to hinder carbonate decarbonation.

The presence of other gases, e.g. water vaporreleased during the combustion of wood in tra-ditional lime kilns, may speed up the decompo-

sition process and reduce calcination T (Berutoet al. 2003). The characteristics of the oxideproduct are strongly dependent on T and dura-tion of the calcination process (Boynton 1980).CaO crystals formed a low T, and/or after arelatively short period of calcination, are nano-sized, show a very porous structure and arehighly reactive (Fig. 5a).

Calcination in air at T above 900°C and/orduring long periods of time favors the orientedaggregation of CaO nanocrystals and the sub-sequent sintering and densification(Rodríguez-Navarro et al. 2009). These pheno-mena can lead to the formation of completelyinert or "dead burnt" CaO at T ca. 1400°C. In


Fig. 5. The thermal decomposition of calcte: a) SEM image of the porous structure of CaO crystals formed after calcite (see pseudo-morph in inset). Preferred orientations of CaO crystals are indicated with arrows. The left inset shows the (100) pole figure of CaOcrystals with a preferred orientation with respect to the parent calcite (the projection plane is parallel to the cleavage plane of calcite);b) TEM photomicrographs of oriented CaO nanocrystals formed upon calcite decomposition following e-beam irradiation in the TEMchamber. Insets show the corresponding SAED pattern of precursor (calcite) and product (CaO) phases. Four sets of product CaOcrystals with preferred crystallographic orientation are identified in the SAED as shown by the orientation of their reciprocal lattice vec-tors (1 to 4); c) projection of the calcite structure on the cleavage plane; d) loss of CO2 from former CO3

2- groups in the calcite (10.4)planes; and e) final transformation (via a topotactic mechanism) into oriented CaO crystals. Modified from Rodríguez-Navarro et al.(2009).

100

general, the most reactive limes are thosesubjected to a relatively "soft" calcination pro-cess ("soft burning") (Boynton 1980).

An adequate knowledge of the mechanism ofCaCO3 calcination is crucial to optimize thisprocess and to obtain oxides with desired pro-perties. Despite many studies, there is noconsensus on how such a reaction occurs(Rodríguez-Navarro et al. 2009). We recentlyinvestigated the thermal decomposition of cal-cite, both studying the kinetics of the reactionand its textural and microstructural aspects(Rodríguez-Navarro et al. 2009). X-ray diffrac-tion (XRD), two-dimensional XRD (2D-XRD),scanning (FESEM) and transmission electronmicroscopy (TEM-SAED) allowed us to deter-mine that the calcination of calcite is pseudo-morphic and topotactic (Figure 5b).

The release of CO2 from CO32- groups existing

along the (10.4) planes of calcite, results in arealignment of oxygen and Ca atoms to formCaO nanocrystals that preserve a closecrystallographic relationship with the parentcalcite crystal (Figure 5c-e). Pseudomorphsare highly porous (up to 54% porosity) due tothe molar volume difference between calciteand calcium oxide. CaO nanocrystals (initiallywith size ca. 5 nm and with surface area of upto 80 m2/g) show the following crystallogra-phic relationships with the precursor calcite:{10.4}calcite//{110}CaO and <441>calci-te//<110>CaO. These results enabled us topropose a new model for the thermal decom-position of calcite, which is schematicallyshown in Fig. 5.

The nanosized character of CaO crystals for-med topotactically pre-determines the kineticsof the following stage of the lime cycle,namely, the hydration step, as well as the pro-perties of the resulting hydroxides. On theother hand, the close textural and crystallo-graphic relationship between carbonate andoxide crystals means that the textural andmicrostructural characteristics of the carbona-te will largely determine the properties of theresulting oxides. This assertion is corrobora-ted by various studies showing that the surfa-ce area, porosity, pore size and crystal sizedistribution of the carbonate rock, in additionto the presence of impurities, determine the

properties, in particular the reactivity, of pro-duct calcium oxide (Elert et al. 2002).

Depending on the composition and texturalcharacteristics of carbonate rocks, it will betherefore necessary to establish the optimalparameters for calcination (temperature andduration) (Boynton 1980).

3.2.2. Hydration

In contact with water calcium oxide reacts for-ming calcium hydroxide (Ca(OH)2), the mineralportlandite, according to the reaction:

CaO + H2O # Ca(OH)2 (5)

Reaction (5) is highly exothermic. Thus, safetyprecautions have to be taken during traditionallime slaking, both due to possible burns and tothe corrosive effect of splashes (pH 12.4). If thehydration occurs with the stoichiometric amountof water, a dry powdered precipitate is formed.Dry hydrate is the standard industrial productnowadays. If excess water is added, the finalproduct is an aqueous dispersion of hydroxidecrystals. This way a lime paste (or lime putty) isformed.

There are numerous studies on the effect ofoxide/water ratio, oxide grain size, temperatureof water, presence of alcohol or additives in theslaking water, and agitation in the properties ofthe hydroxide formed after slaking (see reviewby Elert et al. 2002).

Various mechanisms for the transformation ofcalcium oxide into portlandite have been propo-sed (Ramachandran et al. 1964; Beruto et al.1981; Wolter et al. 2004; Sato et al. 2007). Itappears that a part of Ca(OH)2 directly precipi-tates from the supersaturated solution formedupon dissolution of CaO in the slaking water.However, there is evidence suggesting that thehydration of a considerable amount of CaOoccurs as a solid state reaction between theoxide and water vapor generated as a result ofthe increase in temperature experienced duringhydration (Wolter et al. 2004).

Yet, it is not clear how these two processesoccur and what their relative merit is in the ove-rall slaking process. Beruto et al. (1981) propo-


-

101

se that vapor phase hydration of CaO is contro-lled by the structure of the reactant and product,being a pseudomorphic reaction that generatesnanometric Ca(OH)2 crystals. It could be hypo-thesized that given the structural similarity bet-ween (111)CaO and (001)Ca(OH)2 planes, thereplacement of O= in the oxide by OH- groupscould easily result in the topotactic formation ofportlandite. This would imply that the micros-tructural characteristics of the oxide would be(at least partially) inherited by the hydroxide.

This mechanism could help to explain whydepending on the characteristics of the oxide(e.g., surface area, porosity and particle size)hydrated limes with different properties aregenerated, even when slaking is performedunder identical conditions (Elert et al. 2002).

Rodríguez-Navarro et al. (2005) have shownthat there are significant microtextural differen-ces between traditional slaked limes (lime put-ties) and commercial (powder) hydrated limes.The formers have a greater proportion of parti-cles with size < 1 µm. Apparently, the drying pro-cess undergone by dry hydrated limes results inboth reversible (non-oriented) and irreversible(oriented) aggregation of portlandite crystals.Oriented aggregation results in an increase ofthe effective size of Ca(OH)2 particles (aggrega-tes), which results in a significant reduction ofsurface area and reactivity, and a worsening ofrheological properties (viscosity, plasticity andworkability) (Ruiz-Agudo and Rodríguez-Navarro2010). This is consistent with studies carriedout at the beginning of the 20th century whichconcluded that slaked lime behaved as an irre-versible colloid once dried (Ray and Mathers1928).

In the case of lime putties, there is a range ofphenomena that occur after slaking if the pasteis kept underwater during long periods of time(months or years). This is the so-called“ageing”, which was used since Roman times toimprove the properties of slaked lime (Cowper1927; Ashurst 1990). Pliny indicates that therewas a law in ancient Rome which establishedthat lime putties should be "aged" for at leastthree years prior to their application (Rodríguez-Navarro et al. 1998).

This tradition, based on empirical observations,

has survived to our day. However, until relativelyrecently, there were doubts about whether ornot ageing produced any improvement in theproperties of lime paste, and if this was thecase, it was not known which was the causeand/or mechanism of such improvement.Rodríguez-Navarro et al. (1998) showed thatafter ageing of lime pastes for 2 to 10 months,there was a reduction in the size of portlanditecrystals and an increase in their surface area.Planar (hexagonal platelets) with overdeveloped{0001} forms appeared in parallel with thedisappearance of larger prismatic crystals.Preferential dissolution of the prism faces along(0001) planes resulted in an overall increase inthe amount of plate-like portlandite crystals. Inaddition, secondary precipitation of sub-micro-metric, planar portlandite crystals also occurredduring ageing. All these smaller planar crystalsled to an overall increase in surface area, reac-tivity, dynamic viscosity and plasticity(Rodríguez-Navarro et al. 2009; Atzeni et al.2004; Ruiz-Agudo and Rodríguez-Navarro2010).

However, Ruiz-Agudo and Rodríguez-Navarro(2010) indicate that depending on the characte-ristics of the oxide (hard vs. soft burnt lime), therheological evolution of slaked lime putties wasdifferent, although in all studied cases a gene-ral improvement was observed after longperiods (several months) of ageing.

3.2.3. Carbonation

Carbonation of calcium hydroxide in the presen-ce of atmospheric CO2 occurs according to theoverall reaction:

Ca(OH)2 + CO2 # CaCO3 + H2O (6)

The resulting CaCO3 crystals act as a binder orcement due to their interconnected microstruc-ture (Beruto et al. 2005) joining the rest of theelements in the mix (aggregate).

The setting of air lime mortars and plastersbegins with an initial period of drying and shrin-kage and is followed by the carbonation reac-tion. The reaction interface advances throughthe porous system from the surface towardsthe unreacted core. This process consists ofseveral steps:


102

• (i) diffusion of CO2 (gas) through the openpores,

• (ii) dissolution of Ca(OH)2 in the pore water, • (iii) absorption and dissolution of CO2 in the

alkaline pore water forming carbonic acidH2CO3;

• (iv) its immediate dissociation into bicarbona-te and carbonate ions,

• (v) reaction between Ca2+ and CO32– ions for-

ming CaCO3 through nucleation and growth.

All these processes are interrelated and alteringthe kinetics of one process influences theothers (Van Balen and Van Gemert 1994;Cultrone et al. 2005; Cizer et al. 2012).

The volume, geometry and size of the pores, aswell as the water content, play an importantrole in the progress of carbonation. This is pri-marily due to the fact that water largely controlsthe rate of CO2 diffusion, which along with CO2dissolution (forming carbonic acid) are ratedetermining steps for carbonation. As waterevaporates during drying, an effect which ispromoted by the exothermic nature of carbona-tion (Moorehead, 1986) and the less hydrophi-lic character of newly-formed CaCO3 if compa-red with Ca(OH)2 (Beruto and Botter 2000;Beruto et al. 2005), pores tend to stay open.This accelerates CO2 diffusion towards theinterior of the material which is carbonating.

However, the porosity is slightly reduced due tothe precipitation of CaCO3 crystals, which havea molar volume 11.7% higher than that ofCa(OH)2 (Moorehead 1986). Overall the kineticsof carbonation is very slow, an the reduction inpore size (Lawrence et al. 2007) could furthercontribute to a reduction in carbonation rateover time. This may help explaining why uncar-bonated portlandite has been found in lime mor-tars several centuries old (Marchese 1980),although this is a matter of controversy (Adamset al. 1998).

CaCO3 formed after carbonation may appearas three anhydrous polymorphs, calcite, ara-gonite and vaterite, depending on the condi-tions of the reaction, (i.e. pH, T, supersatura-tion) and the presence of impurities or additi-ves. Calcium carbonate may also precipitateas an amorphous phase (Amorphous CalciumCarbonate, ACC), which is hydrated and

metastable. ACC, which usually precipitatesas small spheres of less than 1µm in diame-ter, turns into calcite as the reaction progres-ses. As we will see later, this phase seems toplay a key role in the process of carbonationof Ca(OH)2 (Cizer et al. 2012).

The textural characteristics of CaCO3, as wellas the cementing structure formed after car-bonation, appear to depend on the propertiesand characteristics of precursor portlanditecrystals and on the conditions of carbonation(i.e., water content, T and RH, pCO2, pH, andionic activity). There are studies on theeffects of the size, surface area, and habit ofportlandite crystal on the kinetics of carbona-tion. In general, the smaller the portlanditecrystals are, the higher their reactivity duringcarbonation is (Van Balen 2005). In turn, thetextural features of the portlandite precursorand the kinetics of carbonation determine thephysical and mechanical properties of lime-based materials (Cizer et al. 2012).

Regarding the mechanisms of Ca(OH)2 carbo-nation, two possible routes exist: a) precipita-tion in solution and b) replacement of portlan-dite by calcium carbonate. The latter canoccur either via a solid state reaction (Gillott1967; Matsuda and Yamada 1973;Moorehead 1986) or through a tightly cou-pled process of dissolution/precipitation(Putnis 2009).

Carbonation in solution has been thoroughlystudied because it is important in the indus-trial precipitation of calcium carbonate usedin paper, paint, plastic and pharmaceuticalindustries (see overview by Cizer et al.2012).

This reaction takes place according to theOstwald rule of stages following the precipita-tion sequence:

ACC # vaterite # (aragonite) # calcite

(Ogino et al. 1987). Some studies show theinitial precipitation of ACC nanoparticles inlime mortars subjected to carbonation(Rodríguez-Navarro et al. 2002; Cizer et al.2012), which is followed by the formation ofscalenohedral and, finally, rhombohedral cal-


103

cite crystals (Cultrone et al. 2005; Cizer et al.2012) (Fig. 4).

Occasionally, the formation of aragonite orvaterite has been observed (Zanco 1994). Itshould be noted that the use of a calciticaggregate may promote the direct epitaxial(homoepitaxial) precipitation of calcite, i.e.,not following the Ostwald’s rule. Such aregrowth should generate a coherent interfa-ce between the aggregate and the binder.This could help to explain why conservationlime mortars prepared with calcitic aggrega-tes tend to show a higher strength than thoseprepared using dolomitic or siliceous aggre-gates (Lanas and Alvarez 2003).

With respect to the second mode of carbona-tion, that is, the transformation in solid stateor via coupled dissolution/precipitation, thereis no consensus on what is the actual mecha-nism of reaction. Gillott (1967) and Matsudaand Yamada (1973) indicate that in the pre-sence of low amounts of water, portlanditecrystals transform into calcite through a solidstate reaction controlled by the structure,i.e., topotactic transformation. In his classicstudy, Moorehead (1986) indicates that inaddition to the precipitation of calcium carbo-nate in solution, which he considered asecondary process, carbonation occurredfollowing a solid state replacement.Nonetheless, he noted that water was neces-sary for the reaction to progress. This castssome doubts about the solid state nature ofthe carbonation process. Beruto and Botter(2000) and Beruto et al. (2005) show thatcarbonation was significantly accelerated forRH values > 75%, when more than 4 H2Omonolayers were adsorbed onto portlanditecrystals. Ca(OH)2 dissolution into the adsor-bed water enabled the carbonation to pro-gress at a sufficiently high speed. Thisimplies a pseudomorphic replacement viacoupled dissolution/precipitation, whichshould be much faster than a solid statereaction (Putnis, 2009).

Beruto et al. (2005) called attention to the factthat the direct precipitation of calcite from a satu-rated Ca(OH)2 solution generates dispersed par-ticles, without any degree of cohesion or mecha-nical resistance, while through a mechanism of

pseudomorphic replacement, as describedabove, a very cohesive and mechanically resis-tant paste of calcium carbonate is obtained.

Although the mechanism of carbonation via thispseudomorphic replacement is not yet fullyunderstood, it has very important implications.Amongst them are the following: the size, mor-phology and level/type of aggregation of calciumhydroxide crystals determine how it will be thetexture and microstructure of calcite crystals for-med after carbonation. This may help explainingwhy depending on the type of calcitic slaked lime,mortars with different properties and durabilityare obtained after carbonation.

For instance, mortars prepared with agedlime putties show faster carbonation and ahigher mechanical strength than mortars pre-pared with fresh lime putties or with dryhydrate (Cazalla et al., 2000). Rodríguez-Navarro et al. (2002) show that mortars pre-pared using aged lime putty (16 y old) displaya carbonated structure which resemblesLiesegang patterns, which is also observedin historical buildings (Fig. 6). The authors lin-ked the formation of the Liesegang patternwith the microtextural characteristics of agedportlandite crystals.

3.3. Historic lime mortars and plasters

The study of lime mortars and plaster collec-


Fig. 6. Formation of a Liesegang pattern in carbonated lime mortarsprepared with 16 y old lime putty: a) scheme of a mortar prism; b) sec-tion of the prism after 1 y carbonation following phenolphthaleinimpregnation. The red rings correspond to uncarbonated parts, whilethe whiter rings are fully carbonated (modified from Rodríguez-Navarroet al., 2002); c) example of Liesegang pattern following differentialweathering in lime plasters in a historical building (Saint Maria Church,Utrera, Spain) (modified from Hansen et al., 2008).

104

ted from historical buildings has been a cons-tant in recent decades, highlighting their tre-mendous compositional variety, both depen-ding on location and age (Elsen 2006).

The study of such materials, which in manycases present major alteration problems, isjustified for several reasons. First, the study oflime-based materials with different age, prove-nance and uses is of importance from a histo-rical and architectural point of view. On theother hand, their analysis can help identify whatcomponents were historically used in the pro-duction of lime mortars and plasters, and shedlight on the manufacturing technology.

This information is critical if we are to reprodu-ce such materials with a view to their applica-tion in architectural conservation interventions.In this respect, the Venice Charter (1964)recommends the use of materials and techno-logies the closest to those originally used in themaking of the element to be conserved.

The number of published studies on lime-based materials, mostly mortars, used in his-torical buildings from prehistory to practicallyour days, is large, and has steadily increasedover the last two decades.

Analytical techniques such as polarized lightmicroscopy (see an example in Fig. 7), scanningand transmission electron microscopy, x-ray dif-fraction, termogravimetry, differential thermalanalysis, infrared (FTIR) and Raman spectros-copy, x-ray fluorescence, atomic absorptionspectrometry, and calcimetry, among others,have been used to study the composition andmicrostructure of old lime mortars (Moropoulou

et al. 1995; Bakolas et al. 1995a,b, 1998;Riccardi et al. 1998; Franzini et al. 1999;Maravelaki-Kalaitzaki et al. 2003; Elsen 2006;Edwards and Farwell 2008).

Goins (2000) proposed a complete analyticalprotocol for the study of both modern andancient lime mortars. It is not our intention tomake an extensive review of these publica-tions. The interested reader is refereed to theGetty Conservation Project BibliographySeries “Preservation of Lime mortars andplaster” (http://www.getty.edu/conserva-tion/publications_resources) and the reviewby Elsen (2006). Here, however, a fewaspects of the study of old lime mortars willbe reviewed to get an insight into the impor-tance of this field.

An interesting aspect of ancient lime techno-logy refers to the slaking of quicklime mixedwith the aggregate before adding water. This isthe so-called "hot lime" technology(Moropoulou et al. 1996). A variation is called“dry” slaking, which involved the mixing ofquicklime with a wet aggregate (Elsen 2006).Apparently it was a common slaking methodand generated mortars of excellent quality anddurability (Moropoulou et al. 1996). Such aslaking process could have produced alkali-aggregate reactions between lime and silice-ous sand (Armelao et al. 2000), favored by thehigh temperature reached. Margalha et al(2011) have carried out laboratory studies onthe potential benefits "hot lime" technologycan have in the case of air lime mortar mixedwith siliceous sand. Their results, however, donot show that this technology impart any clearadvantage to these mortars. In contrast,Malinovski and Hansen (2011) concluderecommending the use of hot lime technologyfor conservation lime mortars.

A large number of studies, such as those ofBakolas et al. (1995a, 1995b, 1998) onByzantine and post-Bizantine lime mortars, orMoropoulou et al. (1996) show that, in addi-tion to air lime mortar, many of the mortarsof such times included lime mixed with cera-mics, crushed or powdered. In general, itappears that the use of artificial hydrauliclime mortars was much more common afterthe fall of the Roman Empire than previously


Fig. 7. Optical microscopy photomicrograph of a lime mortar from thewalls of the IX c. Castle of Cañete la Real (Malaga, Spain). Plane light.

105

thought (Elsen 2006).

Bakolas et al. (1995b) studied 14th to 18th c.Venetian mortars, and concluded that one ofthe most distinctive features in historic limemortars is the presence of whitish nodules(lumps) with dimensions ranging between afew mm and 1 or 2 cm. These nodules arecomposed of calcite, so they could eithercorrespond to fully carbonated slake limelumps, or uncalcined limestone chips.

The presence of traces of silicates (hydraulicphases), a high porosity and numerous fracturessuggest that these nodules correspond to sla-ked lime lumps that did not dissagregate uponmixing with sand and water (Franzini et al. 1999),possibly due to the lack of aging of the originallime which displayed a low plasticity (Tuncokuand Caner-Saltik 2006). Alternatively, they couldbe the result of the hot lime or dry slaking tech-nology (Bakolas et al. 1995a, 1998). Local over-heating during dry slaking could provoke anagglomeration of the hydrated lime particles,causing the formation of such nodules.

Regarding the formulation of historical limemortars, classic treaties such as those ofVitruvius (30 BC) and Alberti (1452), establis-hed that 1:3 was the optimal binder to aggre-gate ratio. However, many researchers havefound that, in general, Roman and Greek limemortars with considerable age difference hadvery similar textural and compositional charac-teristics, among which it is highlighted that thebinder/aggregate ratio was higher than 1:3,typically 1:2 (Foster 1934; Elsen 2006).

Another important aspect of old lime mortartechnology refers to the use of additives (wedo not consider here those imparting hydrau-lic character). Typically, organic additiveswere dosed in the lime mix to improve themortar properties both in the fresh and har-dened state. Examples of such additives are:egg white, blood of ox, juices of various fruitssuch as figs, keratin and casein, oils fromplants (e.g, linseed oil), and animal fats,among others (Moropoulou et al. 2005).

For instance, both Aztecs and Mayans usednopal juice (Oppunctia ficus indica), which isvery rich in pectin, as an additive in preparing

lime mortars and plasters of high plasticityand capacity to avoid cracking during drying,tradition which is currently preserved in diffe-rent parts of Mexico and the United States(Cárdenas et al. 1998).

Researchers have attempted to reproducetraditional building practices using naturaladditives to improve the performance of limeused in architectural conservation. For exam-ple, Yang et al. (2010) used the ancientChina vernacular sticky rice technology forimproving the performance of new conserva-tion lime-based mortars. The presence ofamylopectin in the rice caused the formationof a matrix of calcite nanocrystals which gran-ted an enormous strength and durability tosuch conservation mortars.

Modern, both natural and artificial, organicadditives have been also tested for improvingthe performance and properties of lime-based conservation mortars. For instance,Izaguirre et al (2010) noted an increase inplasticity, adhesion, durability and mechani-cal strength of lime mortar dosed with potatostarch, plasticizers, air-entraining agents andwater repellents.

3.4. Lime-based conservation mortars

The study of old lime mortars have yielded asignificant amount of information which hasbeen applied for the design of new lime-basedconservation mortars.

To be compatible with old masonry and to playa proper role, restoration mortars must meeta series of requirements, among which thefollowing can be highlighted (Maurenbrecher2004):

• (a) the mortar should not have a higherstrength than stone or brick: a conserva-tion mortar is not "better" the stronger itis;

• (b) the water absorption and water vaporpermeability of the mortar must be of thesame order of magnitude or greater thanthat of the other masonry elements. Thisfavors the accumulation of water (andsalts) in the mortar which may act as asacrificial material;


106

• (c) mortars must have a minimum retractionto prevent crack development;

• (d) should ensure a good contact with bricksor stone. This is favored by a high plasticityand workability;

• (e) in general, its physical properties (ther-mal expansion, color, etc) should ensure ahigh compatibility with the different ele-ments in the old masonry structure, whileensuring that the mortars is sufficientlydurable. This is particularly relevant incases where loss of mortar is so generalthat the survival of the whole masonrystructure is in jeopardy (Fig. 8).

A consensus exists on the need of using con-servation lime mortars with formulations andmanufacturing / application techniques assimilar as possible to the traditional ones. Forexample, it is generally recommended the useof calcitic lime pastes, especially those thathave undergone aging. Faria et al. (2008) notethat resistance to salt crystallization damageis significantly higher in mortars prepared withlime pastes aged for 10 to 16 months if com-pared with mortars prepared with commercialdry hydrate. There is however no consensuson what is the optimal aging time. The goldenrule should be the longer the better. However,from a practical point of view, a period betwe-en 2 and 10 months aging brings about a sig-nificant improvement (Rodríguez-Navarro et al.1998). Ashurst (1980) recommends a mini-mum of three months ageing period.

In addition to slaked lime, conservation mor-tars and plasters include water and aggregate.Unlike in the case of Portland cement, watercontent does not appear to appreciably affect

the mechanical behavior of air lime mortars(Lawrence and Walter 2008). Nonetheless,Arandigoyen et al. (2005) show that increasedwater/lime ratios increase the mortars poro-sity, an effect which could be detrimental inthe presence of ice or soluble salts. It seemsthat depending on the type of lime, water requi-rements vary. A method to overcome such aneffect is to prepare mortars with a standardconsistency (Hansen et al. 2008). A precau-tion to consider is the purity of the water used.Vitruvius recommended the use of high puritysand without fines, discarding the use of seawater due to the harmful effect of salts.

With regards to the most appropriate bin-der/aggregate ratio in conservation mortarsand how it affects the mortar properties,there are mixed results. Classically, the useof a 1/3 (lime/sand) ratio was recommendedas indicated by Vitruvius and Alber ti.However, Lanas and Alvarez (2003) demons-trate that the best physical-mechanical pro-perties are achieved when a binder/aggrega-te ratio of 1:1, or even 2:1, is used. However,other authors such as Cazalla et al. (2002)note that the best mortars (in terms of speedof carbonation and mechanical properties)are those with a 1:4 ratio. Note that Lanasand Alvarez (2003) used hydrated lime pow-der, while Cazalla et al. (2002) got the bestmortars using lime putty aged for 14 years. Itfollows that depending on the type of limeused, and the final application of the conser-vation mortar (joint mortar, repointing, sacrifi-cial layer, or plaster), the binder/aggregateratio has to be gouged.

The aggregate is another important compo-nent of a conservation lime mortar. The aggre-gate is added to reduce costs, to minimize theformation of drying cracks, and to provide pro-per consistency in the fresh state, and adequa-te strength and permeability (porosity) aftersetting. The type of sand is highly variable.Cowper (1927) in his classic report providessome general rules for the selection of propersand (see also Elert et al. 2000).

3.5. Decay and conservation of lime mortarsand plasters

Lime mortars and plasters suffer different


Fig. 8. Ruins of the Castle of Cañete la Real (Malaga, Spain). Thepiece of wall shows extensive loss of joint mortar linking the stoneblocks.

107

damaging processes when exposed both out-doors and indoors. These include chemical(dissolution, sulfation in polluted environ-ments), physical (thermal changes,freeze/thaw, salt crystallization) and biologi-cal (mechanical action of roots, degradationby bacterial action) weathering processes(Ashurst 1990; Furlan 1991; Doehne andPrice 2010). They cause loss of cohesion,increase in porosity, and granular disintegra-tion, among other phenomena.

Such a damage often entails the loss ofmaterial, as it is shown in Fig. 8, so the morecommon type of conservation intervention isto reapply a new mortar or rendering prepa-red according to the guidelines referred to inthe previous section.

However, there are situations in which it isnecessary to consolidate the lime-based histo-rical material. Studies dedicated to the conso-lidation and/or protection of lime mortars arefairly limited when compared with studies dedi-cated to the design and testing of restorationmortars. There are however some studies thatcompare various types of products for the con-solidation of historic lime mortar following anapproach very similar to that followed duringthe consolidation of stone.

For example, Toniolo et al. (2011) have stu-died the effectiveness of different consoli-dants applied on lime mortars. The testedconsolidants were:

• polyethylmetacrylate-methylacrylate copoly-mer (Erabiliz B72);

• tetraethyl orthosilicate (RC 70 - TEOS); • saturated barium hydroxide solution.

Changes in color and mechanical propertieswere evaluated after consolidation. With regardto the appearance of the treated materials, thebest behavior was observed after the applica-tion of barium hydroxide. The highest increasein strength was observed after the applicationof TEOS, while the worst behavior was obser-ved in the case of the barium hydroxide.

Inorganic consolidants such as lime water orbarium hydroxide have been thoroughly appliedon lime plasters. While the classic lime water

treatment seems to be highly ineffective, theBa(OH)2 treatment appears to be very promi-sing (Doehne and Price, 2010). Carbonation ofthis hydroxide leads to the precipitation ofcementing BaCO3. In the presence of sulfates,highly insoluble BaSO4 is also formed whichcontributes to the strengthening of the lime-based material and limits salt damage.

Izaguirre et al. (2009) studied the effects ofsome protective treatments (water repe-llants) applied to air lime mortars, observinga significant improvement of the mortarsresistance to freeze/thaw cycles.

Treatment based on the application of alcoho-lic dispersions of Ca(OH)2 nanoparticles, theso-called “nanolimes”, show some promisingresults as a novel and effective consolidationmethod for lime-based mortars, renderingsand plasters (Baglioni and Giorgi 2006). Yet,much research has to be dedicated to thestudy of novel, effective conservation treat-ments for the in situ consolidation and pro-tection of lime-base building materials.

4. Concluding remarks

Mankind has used gypsum and lime binders fordecorative and structural purposes in historicalbuildings since prehistory. Both binders sharesome interesting features in their processing andsetting (e.g., the gypsum and the lime cycles). Inparticular, their “activation” and settings showsome mechanistic similarities:

• (a) their calcinations/dehydration is a solidstate process which follows a topotacticmechanism, and

• (b) their setting involves two different mecha-nisms, namely (i) crystallization in solution, and (ii) either a solid state reaction or a coupled

dissolution/precipitation process.

Lime is somehow different in that its cycle showsan additional step of hydration, which from amechanistic point of view also involves precipita-tion in solution and another process which maybe either a solid state reaction or a pseudomor-phic tightly coupled dissolution/precipitation.

Research should focus on elucidating which


108

one of the process already mentioned, i.e.,solid state or dissolution/precipitation,actually takes place, and how is the atomicscale mechanism involved in such a process.

It is shown that due to the structural relations-hips between parent and product phases inboth gypsum and lime cycles, the properties ofthe parent phases predetermine the propertiesof the product phases. This has profound con-sequences on the performance of both ancientand modern (conservation) gypsum- and lime-based materials, and emphasizes the fact thateach stage of the two cycles has an effect onthe properties of the final set material.

An overview of the uses of both gypsum andlime as binders in historical buildings, their usein conservation interventions, and their actualdecay and conservation, shows that despitethe numerous research efforts that have takenplace over the last few decades, much is still tobe investigated. For instance, the developmentof novel effective conservation methods for insitu consolidation of both gypsum and lime buil-ding materials is still a challenge.

The study and testing of nanosized crystalline, oreven amorphous, precursors (bassanite, nanoli-mes) for the consolidation of gypsum and limemortars and plasters could be an interesting andpotentially fruitful research path worth to explore.

5. Acknowledgement

This research was financed by the EU InitialTraining Network Delta-Min (Mechanisms ofMineral Replacement Reactions) grant PITN-GA-2008-215360, the Spanish Government(grant MAT2009-11332) and the Junta deAndalucía (research group RNM-179).

Some of the results here presented wereobtained during the development of the LimeMortars and Plasters project, a joint collabo-ration between the Getty ConservationInstitute, the Catholic University of Leuvenand the University of Granada.

Contributions by K. Elert, E. Hansen, K. Van Balen,O. Cizer, E. Ruiz-Agudo, O. Cazalla, E. SebastianPardo and K. Kudlacz are acknowledged.

6. References

Abriel, W., Reisdorf, K. and Pannetier, J. (1990)Dehydration reactions of gypsum: A neutron and X-raydiffraction study. J. Solid State Chem. 85, 23-30.

Adam, C.D. (2003) Atomistic modelling of the hydra-tion of CaSO4. J. Solid State Chem. 174, 141-151.

Adams, J. Kneller, W. and Dollimore, D. (1992)Thermal analysis (TA) of lime- and gypsum-basedmedieval mortars. Thermochim. Acta 211, 93-106.

Adams J., Dollimore D., and Griffiths D.L. (1998) Thermalanalytical investigation of unaltered Ca(OH)2 in datedmortars and plasters. Thermochim. Acta 324, 67-76.

Alberti, J.B. (1452) On the Art of Building in Ten Books(De Re Aedificatoria). Translation by Leach, N., Rykwert,J., and Tavenor, R, 1988. MIT Press, Cambridge.

Alejandre Sánchez, F.J. (2002) Historia,Caracterización y Restauración de Morteros.Universidad de Sevilla, Sevilla.

Allen, E. and Thallon, R. (2011) Fundamentals ofResidential Construction. Wiley, Hoboken, N.J.

Arandigoyen, M., Pérez Bernal, J.L. Bello López, M.A.and Alvarez J.I. (2005) Lime-pastes with differentkneading water: Pore structure and capillary porosity.Appl. Surf. Sci. 252, 1449–1459.

Armelao, L., Bassan, A., Bertoncello, R., Biscontin,G., Daolio, S. and Glisenti, A. (2000) Silica glass withcalcium hydroxide: a surface chemistry approach. J.Cultural Heritage 1, 375-384.

Arredondo, F. (1991) Yesos y Cales. Ed. E.T.S. Ing.Caminos, Madrid.

Ashurst, J. (1990) Mortars for stone buildings. InConservation of Building and Decorative Arts, Vol. 2,J. Ashurst and F.G. Dimes, eds., p. 78-96.Butterworths, London.

Atzeni, C., Farci, A., Floris, D. and Meloni, P. (2004)Effect of aging on rheological properties of lime putty.J. Am. Ceram. Soc., 87, 1764-1766.

Baglioni, P. and Giorgi, R. (2006) Soft and hard nano-materials for restoration and conservation of culturalheritage. Soft Matter 2, 293-303.


109

Bakolas, A., Biscontin, G., Moropoulou, A. and Zendri, E.(1995a) Characterization of the lumps in the mortars ofhistoric masonry. Thermochim. Acta 269/270, 809-816.

Bakolas A, Biscontin G, Contardi V, Francheschi, E.,Moropouolou, A., Palazzi, D. and Zendri, E. (1995b)Thermoanalytical research on traditional mortars inVenice. Thermochim. Acta 269/270, 817-828.

Bakolas, A., Biscontin, G., Moropoulou, A. and Zendri,E. (1998) Characterization of structural Byzantinemortars by thermogravimetric analysis. Thermochim.Acta 321, 151-160.

Beruto, D., Barco, L., Belleri, G. and Searcy, A.W.(1981) Vapor-phase hydration of submicrometer CaOparticles. J. Am. Ceram. Soc. 64, 74-80.

Beruto, D.T. and Botter, R. (2000) Liquid-like H2Oadsorption layers to catalyze the Ca(OH)2/CO2 solid-gas reaction and to form a non-protective solid layerat 20°C. J. Eur. Ceram. Soc. 20, 497-503.

Beruto, D.T., Vecchiattini, R and Giordani, M. (2003)Effect of mixtures of H2O(g) and CO2(g) on the ther-mal half decomposition of dolomite natural stone inhigh CO2 pressure regimes. Thermochim. Acta 405,25-33.

Beruto, D.T., Barberis, F. and Botter, R. (2005)Calcium carbonate binding mechanisms in the settingof calcium and calcium-magnesium putty-limes. J.Cultural Heritage, 6, 253-260

Boynton, R.S. (1980) Chemistry and Technology ofLime and Limestone, Wiley, New York.

Callebaut, K.; Elsen, J.; Van Balen, K.; and Viaene, W.(2001) Nineteenth century hydraulic restoration mor-tars in the Saint Michael's Church (Leuven, Belgium):Natural hydraulic lime or cement? Cement ConcreteRes. 31, 397-403.

Cárdenas, A., Arguelles, W.M. and Goycoolea, F.M.(1998) On the possible role of Opuntia ficus-indicamucilage in lime mortar performance in the protectionof historical buildings. J. Professional Assoc. CactusDevelop. 3, 1-7.

Cazalla, O., Rodríguez-Navarro, C., Sebastian E. andCultrone, G. (2000) Aging of lime putty: effects on tra-ditional lime mortar carbonation, J. Am. Ceram. Soc.83, 1070– 1076.

Charola, A.E., Pühringer, J. and Steiger, M. (2007)Gypsum: a review of its role in the deterioration ofbuilding materials. Environm. Geology 52, 339–352.

Cizer, O., Rodríguez-Navarro, C., Ruiz-Agudo, E., Elsen,J., Van Gemert, D. and Van Balen, K. (2012) Phaseand morphology evolution of calcium carbonate preci-pitated by carbonation of hydrated lime. J. Mater. Sci.47, 6151-6165.

Cotrim, H., Veiga, M.R. and Brito, J. (2008) Freixopalace: Rehabilitation of decorative gypsum plasters.Construction Building Mater. 22, 41-49.

Cowper, A.D. (1927). Lime and lime mortars. DSIRSpecial Report No. 9. Reprinted 1998 by Shaftesbury,Donhead Publishing.

Cultrone, G., Sebastian, E., and Ortega Huertas, M.(2005) Forced and natural carbonation of lime-basedmortars with and without additives: Mineralogical andtextural changes. Cement Concrete Res. 35, 2278-2289.

Davey, N. (1961) A History of Building Materials.London, Phoenix.

Doehne, E. and Price, C.A. (2010) StoneConservation: An Overview of Current Research. GettyConservation Institute, Los Angeles.

Edwards, H.G.M. and Farwell, D.W. (2008) The con-servational heritage of wall paintings and buildings:an FT-Raman spectroscopic study of prehistoric,Roman, mediaeval and Renaissance lime substratesand mortars. J. Raman Spectrosc. 39, 985-992.

Elert, K., Rodríguez-Navarro, C., Sebastian, E., Hansen, E.and Cazalla, O. (2002) Lime mortars for the conservationof historic buildings. Studies Conservation 47, 62-75.

Elsen J. (2006) Microscopy of historic mortars — areview. Cement Concrete Res. 36, 1416–1424.

Faria, P., Henriques, F. and Rato, V. (2008)Comparative evaluation of lime mortars for architech-tural conservation. J. Cultural Heritage 9, 338-346.

Finlayson, B, Mithen, S.J., Najjar, M., Smith, S.,Maricevic, D., Pankhurst, N. and Yeomans, L. (2011)Architecture, sedentism, and social complexity at Pre-Pottery Neolithic A WF16, Southern Jordan. PNAS108, 8183–8188.


110

Foster, W. (1934) Grecian and Roman stucco, mortar,and glass. J. Chem. Edu. 11, 223-225.

Franzini, M., Leoni, L., Lezzerini, M. and Sartori, F.(1999) On the binder of some ancient mortars. Miner.Pretrology 67, 59-69

Freyer, D. and Voigt, W. (2003) Crystallization andphase stability of CaSO4 and CaSO4-based salts.Monatsh. Chem. 134, 693–719.

Furlan, V. (1991) Causes, mechanisms and measure-ment of damage to mortars, bricks and renderings. InScience, technology, and European cultural heritage.Proceedings of the European symposium. Bologna,Italy, 13-16 June 1989, 149-159.

Galwey, A.K. (2000) Structure and order in thermaldehydrations of crystalline solids. Thermochim. Acta355, 181-238.

Gárate Rojas, I. (1994) Las Artes de la Cal (2ªEdición). Ministerio de Cultura-ICBRBC, Madrid.

Gartner, E.M. (2009) Cohesion and expansion inpolycrystalline solids formed by hydration reactions -The case of gypsum plasters. Cement Concrete Res.39, 289–295.

Ghorab, H.Y. Ragai J. and Antar A. (1986) Surface andbulk properties of ancient Egyptian mortars. Part I: X-raydiffraction studies. Cement Concrete Res. 16, 813-822.

Gillott, J.E. (1967) Carbonation of Ca(OH)2 investiga-ted by thermal and X-ray diffraction methods of analy-sis. J. Appl. Chem. 17, 185-189.

Goins, E.S. (2000) A new protocol for the analysis of his-toric cementitious materials: interim report. InInternational RILEM workshop on Historic Mortars:Characteristics and Tests, 71-79, RILEM, Cachan, France

Hansen, E.F., Rodríguez-Navarro, C. and Van Balen, K.(2008) Lime putties and mortars: Insights into funda-mental properties. Studies Conservation 53, 9-23.

Haswell, C.H. (1865) Limes, cements, mortars andconcretes. J. Franklin Institute 49, 361-367; 50, 171-175, 295-302.Houben, H. and Guillaud, H. (1994) EarthConstruction: A Comprehensive Guide. CRATerre-EAG,Intermediate Technology Publications, London.

Hawthorne, F.C., Krivovichev, S.V., and Burns, P.C.(2000), The crystal chemistry of sulfate minerals.Rev. Miner. Geochem. 40, 1-112.

Izaguirre, A., Lanas, J. and Alvarez, J.L. (2009) Effect ofwater-repellent admixtures on the behaviour of aeriallime mortars. Cement Concrete Res. 39, 1095-1104.

Izaguirre, A., Lanas, J. and Alvarez, J.L. (2010) Ageing oflime mortars with admixtures: durability and strengthassessment. Cement Concrete Res. 40, 1081-1095.

Kawiak, T. (1991) Gypsum mortars from a twelfth-cen-tury church in Wislica, Poland. Studies Conservation36, 142-150.

Kingery, W.D., Vandiver, P.B. and Prickett, M. (1988)The beginnings of pyrotechnology, part II: productionand use of lime and gypsum plaster in the pre-potteryneolithic Near East. J. Field Archaeology 15, 219-244.

Lanas, J. and Alvarez J.I. (2003) Masonry repair lime-based mortars: Factors affecting the mechanicalbehavior. Cement Concrete Res. 33, 1867-1876.

Lanzón, M. and García-Ruiz, P.A. (2012) Effect of citricacid on setting inhibition and mechanical propertiesof gypsum building plasters. Construction BuildingMater. 28, 506–511.

Lawrence, R.M., Mays, T.J., Rigby, S.P., Walker, P. andD’Ayala, D. (2007) Effects of carbonation on the porestructure of non-hydraulic lime mortars. CementConcrete Res. 37, 1059-1069.

Lawrence, R.M. and Walker, P. (2008) The impact ofthe water/lime ratio on the structural characteristicsof air lime mortars. In: D´Ayala & Fodde (eds),Structural Analysis of Historic Construction. Taylor &Francis Group, London, 885-889.

Lea, F.M. (1970) The Chemistry of Cement andConcrete, 3rd edn. Edward Arnold Ltd., London.

Le Chatelier, H (1887) Recherches expérimentales surla constitution des mortier hydrauliques. Dunod, Paris.

Lucas, A. and Harris, J.R. (1962) Ancient EgyptianMaterials and Industries. Edward Arnold, London. Luxán, M.P. Dorrego, F. and Laborde, A. (1995)Ancient gypsum mortars from St. Engracia (Zaragoza,Spain): Characterization. Identification of additives andtreatments. Cement Concrete Res. 25, 1765-1775.


111

Luxán, M.P. and Dorrego, F. (1996) Ancient XVI cen-tury mortar from the Dominican Republic: Its charac-teristics, microstructure and additives. CementConcrete Res. 26, 841-849.

Malinowski, R. (1981) Ancient mortars and concre-tes. Durability aspects. Proceedings ICCROMSymposium. Rome, ICCROM, p. 341-349.

Malinowski, E. S. and Hansen T. S. (2011) Hot lime mor-tar in conservation - Repair and replastering of the facadesof Lacko Castle. J. Architec. Conservation 17, 95-118.

Marchese, B. (1980) Non-crystalline calcium hydroxi-de in ancient non-hydraulic lime mortars. CementConcrete Res. 10, 861-864.

Margalha, G., Viega, R., Silva, A.S, and Brito, J. (2011)Traditional methods of mortar preparation: The hot limemix method. Cement Concrete Composit. 33, 796-804.

Maravelaki-Kalaitzaki, P., Bakolas, A. and Moropoulou,A. (2003) Physico-mechanical study of Cretan ancientmortars. Cement Concrete Res. 33, 651-661.

Maurenbrecher, A.H.P. (2004) Mortars for repair oftraditional masonry. Practice Periodical Struct. DesignConstruct. 9, 62-65.

Matsuda, O. and Yamada, H. (1973) Investigation of themanufacture of building materials by carbonation harde-ning of slaked lime. Sekko to Sekkai 125, 170-179.

Melikhov I.V., Rudin V.N. and Vorob'eva L.I. (1991)Non-diffusive topochemical transformation of calciumsulphate hemihydrate into the dehydrate. MendeleevCommun. 1, 33-34.

Mellaart, J. (1967) Çatal Hüyük: A Neolithic town inAnatolia. Thames & Hudson, London.

Mertens, G., Snellings R., Van Balen K., Bicer-Simsir B.,Verlooy P. and Elsen J. (2009) Pozzolanic reactions ofcommon natural zeolites with lime and parameters affec-ting their reactivity. Cement Concrete Res. 39, 233–240.

Moorehead, D.R. (1986) Cementation by the carbonationof Hydrated Lime, Cement Concrete Res. 16, 700–708.Moropoulou A., Bakolas A. and Bisbikou, K. (1995)Characterization of ancient Byzantine and later historicmortars by thermal and X-ray diffraction techniques.Thermochim. Acta 269/270, 779–795.

Moropoulou, A., Tsiourva, T., Bisbikou, K., Bakolas, A.and Zendri, E. (1996) Hot lime technology impartinghigh strength to historic mortars. ConstructionBuilding Mater. 18, 151-159.

Moropoulou, A. Bakolas, A. and Anagnostopoulou, S.(2005) Composite materials in ancient structures.Cement Concrete Compos. 27, 295–300.

Ogino, T., Suzuli, T. and Kawada, K. (1987) The formationand transformation mechanism of calcium carbonate inwater. Geochim. Cosmochim. Acta 51, 2757-2767.

Prasad, P.S.R., Chaintanya, V.K. Prasad, K.S. andRao, N. (2005) Direct formation of the %-CaSO4 phasein dehydration process of gypsum: In situ FTIR study.Am. Miner. 90, 672-678.

Putnis, A. (2009) Mineral replacement reactions. Rev.Miner. Geochem. 70, 87-124.

Ramachandran, V.S., Sereda, P.J. and Feldman, R.F.(1964) Mechanism of hydration of calcium oxide. Nature201, 288-289.

Ray, K.W. and Mathers, F.C. (1928) The colloidal beha-viour of lime. Ind. Eng. Chem. 20, 475-477.

Ridge, M.J. and King, G.A. (1976) A problem with hydra-ted lime. J. Appl. Chem. Biotechnol. 26, 742-746.

Regourd, M., Kerisel, J., Deletie, P., and Haguenauer, B.(1988) Microstructure of mortars from three Egyptianpyramids. Cement Concrete Res. 18, 81-90.

Riccardi, M.P., Duminuco, P., Tomasi, C. and Ferloni, P.(1998) Thermal, microscopic and X-ray diffraction stu-dies on some ancient mortars. Thermochim. Acta 321,207-214.

Rodríguez-Navarro, C. Hansen, E. and Ginell, W.S. (1998)Calcium hydroxide crystal evolution upon aging of limeputty, J. Am. Ceram. Soc. 81, 3032–3034.

Rodríguez-Navarro, C., Cazalla, O., Elert, K. andSebastian, E. (2002) Liesegang pattern development incarbonating traditional lime mortars. Proc. R. Soc. Lond.A 458, 2261-2273.

Rodríguez-Navarro, C., Ruiz-Agudo, E., Ortega-Huertas,M., and Hansen, E. (2005) Nanostructure and irreversi-ble colloidal behavior of Ca(OH)2: Implications in cultural


112

heritage conservation. Langmuir 21, 10948-10957.

Rodríguez-Navarro, C., Ruiz-Agudo, E., Luque, A.,Rodríguez-Navarro, A. B. and Ortega-Huertas, M. (2009)Thermal decomposition of calcite: Mechanisms of forma-tion and textural evolution of CaO nanocrystals. Am.Mineral. 94, 578-593.

Rubio-Domene, R.F. (2006) El material de yeso: comporta-miento y conservación. Cuadernos Restauración 6, 57-68.

Ruiz-Agudo, E. and Rodríguez-Navarro, C. (2010)Microstructure and rheology of lime putty. Langmuir 16,3868-3677.

Sato, T., Beaudoin, J.J., Ramachandran, V.S., Mitchell,L.D. and Tumidajski, P.J. (2007) Thermal decompositionof nanoparticulate Ca(OH)2 – Anomalous effects. Adv.Cement Res. 19, 1-7.

Sayre, E.V. (1976) Deterioration and restoration of plas-ter, concrete and mortar. In Preservation andConservation: Principles and Practices, S. Timmons, ed.,191-201, Preservation Press, Washington, D.C.

Singh, N.B. and Middendorf, B. (2007) Calcium sulphatehemihydrate hydration leading to gypsum crystallization.Progr. Crystal Growth Charact. Mater. 53, 57-77.

Sipple, E,M., Bracconi, P., Dufour, P. and Mutin, J.C.(2001) Electronic microdiffraction study of structuralmodifications resulting from the dehydration of gypsum.Prediction of the microstructure of resulting pseudo-morphs. Solid State Ionics 141–142, 455–461.

Smeaton, J. (1791) Narrative of the building and a des-cription of the construction of the Eddystone Lighthouse.London.

Solberg, A. and Hansen, S. (2001) Dissolution ofCaSO4•1/2H2O and precipitation of CaSO4•2H2O: Akinetic study by synchrotron X-ray powder diffraction.Cement Concrete Res. 31, 641-646.

Teutonico, J.M., McCaig, I., Burns, C. and Ashurst, J.(1994) The Smeaton project: Factors affecting the pro-perties of lime-based mortars. APT Bull. 25, 32-49.

Toniolo, L., Paradisi, A., Goidanich, S. and Pannati, G.(2011) Mechanical behaviour of lime based mortarsafter surface consolidation. Contruct. Building Mater. 25,1553-1559.Triollier, M. and Guilhot B. (1976) The hydration of cal-cium sulphate hemihydrate. Cement Concrete Res. 6,

507-514.

Tuncoku, S.S. and Caner-Saltik, E.N. (2006) Opal-A richadditives in ancient lime mortars. Cement Concrete Res.36, 1836-1893

Van Balen, K. and Van Gemert, D. (1994) Modeling limemortar carbonation. Mater. Struct. 27, 393-398.

Van Balen, K. (2005) Carbonation reaction of lime, kine-tics at ambient temperature. Cement Concrete Res. 35,647-657.

Van Driessche, A.E.S., Benning, L.G., Rodríguez-Blanco,J.D., Ossorio, M., Bots, P. and García-Ruiz, J.M. (2012)The role and implications of bassanite as a stable precur-sor phase to gypsum precipitation. Science 336, 69-72.

Vaniman, D.T. and Chipera, S.J. (2006) Transformationsof Mg- and Ca-sulfate hydrates in Mars regolith. Am.Miner. 91, 1628-1642.

Venice Charter (1964) International Charter for theconservation and restoration of monuments andsites, Venice, 1964,http://www.icomos.org/docs/veni-ce_charter.html).

Vicat, L.J. (1837) A practical and scientific treatise on cal-careous mortars and cements, artificial and natural.John Weale, London.

Vitruvius, P. (30 BC), The Ten Books on Architecture(Traslation by M.H.Morgan 1960). Dover Publications,New York.

Von Landsberg, D. (1992) The history of lime productionand use from early times to the Industrial Revolution.ZKG 45, 199-203.

Wolter, A., Luger, S. and Schaefer, G. (2004) The kineticsof the hydration of quicklime. ZKG Int. 57, 60-68.

Yang, F., Zhang, B. and Ma, Q. (2010) Study of stickyrice_lime mortar technology for the restoration of histori-cal masonry construction. Acc. Chem. Res. 43,936–944.

Zanco, A. (1994) Identification of vaterite, metastableCaCO3 phase, inside the Eremitani Church, in Padua.Sci. Technol. Cultural Heritage 3, 135-137.
