analysis of pigments from roman wall paintings found in vicenza

7/21/2019 Analysis of pigments from Roman wall paintings found in Vicenza

1/8

Analysis of pigments from Roman wall paintings found in

Vicenza

G.A. Mazzocchin a,*, F. Agnoli a, S. Mazzocchinb, I. Colpo b

a Dipartimento di Chimica Fisica, Universita CaFoscari di Venezia, Calle Larga S. Marta 2137, Venezia 30123, Italyb Dipartimento di Scienze dellAntichita, Universita di Padova, Piazza Capitaniato 7, Padova 35139, Italy

Received 18 February 2003; received in revised form 6 May 2003; accepted 23 May 2003

Abstract

The analysis of about 60 samples of wall paintings was carried out using different chemicophysical techniques: optical

microscopy, scanning electron microscopy (SEM) equipped with an EDS microanalysis detector, X-ray powder

diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR). The identified pigments were cinnabar,

hematite, red ochre, celadonite, cuprorivaite (Egyptian blue), yellow ochre, goethite and carbon. Only in one case some

lead white was found instead. In general, the mortar preparation did not correspond to the complex structure suggested

by Vitruvius (De Architectura), but it generally showed a porous layer, with crushed grains under the pigment layer. Incertain cases two superposed pigment layers were found: yellow superimposed on both red and pink, black on pink,

green on black.

# 2003 Elsevier B.V. All rights reserved.

Keywords: Roman pigments; Wall paintings; Physicochemical analysis; EDS; XRD; FTIR

1. Introduction

Paintings and colour remains of the Roman Age

have been discovered and picked up for over a

century. Among the most important discoveries,Pompeii and Rome stand out [1].

In France and Switzerland, a systematic analysis

on the found fragments was performed [2/5]; in

Spain, scientific research on archaeological finds is

sprouting[6,7].

In Italy, there is a lack of a systematic analysis

of the numerous Roman painting finds, but the

research is now increasing.One of the most important questions in the

analysis of ancient paintings is the nature of the

pigments used. The analysis, in fact, gives infor-

mation useful in defining the gamut of pigments

available at a local, regional or even wider scale

and to understand the techniques of colour pre-

paration and application. In addition, through the

study of pigments, it is possible to discover the

lines of communication and trade exchange.

* Corresponding author. Tel.: /39-041-234-8530; fax: /39-

041-234-8594.

E-mail address: [email protected] (G.A.

Mazzocchin).

Talanta 61 (2003) 565/572

www.elsevier.com/locate/talanta

0039-9140/03/$ - see front matter # 2003 Elsevier B.V. All rights reserved.

doi:10.1016/S0039-9140(03)00323-0
mailto:[email protected]:[email protected]:[email protected]:[email protected]


2/8

The aim of the present work is the study of

about 60 fragments of few square centimetres of

wall painting coming from an archaeological

excavation in a Roman Villa in Vicenza (ContraPedemuro, S. Biagio) directed by the Sovrinten-

denza Archeologica per il Veneto by Dr. Luigi

Malnati and Prof. Gian Pietro Brogiolo of the

University of Padova.

The techniques used were optical microscopy,

scanning electron microscopy (SEM) equipped

with an EDS microanalysis detector, X-ray pow-

der diffraction (XRD) and Fourier transform

infrared spectroscopy (FTIR).

The coupling of the used techniques, most of

which non-destructive, suitable for the determina-tion of anionic groups, crystalline phases, structure

and elemental composition allowed us to have a

complete enough characterisation of the used

pigments, materials and techniques.

2. Experimental

2.1. Sampling

The sampling was made by the archaeologicalteam, which collected pieces representative of the

variety and differences of colour. The dimension

of the fragments analysed ranged from 2 to 20/25

cm2, but it is quite hard to have a precise idea of

the dimensions of the whole frescoes because all

the painted walls were found as fragments covered

by mould and rubble.

2.2. Optical microscopy

The samples were observed by means of an

optical microscope Wild-Leitz M8 with a 19.2/

to 256/ zoom. The samples were illuminated

using a movable fibre glass system. This technique

was used specially in the examination of the grains,

crushed calcite or other tiny stones inside mortars.

2.3. Scanning electron microscopy

SEM images were taken using a Jeol (Tokyo,

Japan) JSM 5600 LV equipped with an Oxford

Instruments 6587 EDS microanalysis detector.

The images were taken under low vacuum condi-

tions where samples did not show any charging

effects; in this way, it was possible to avoid the

coating of the samples with a high-conductancethin film (gold or graphite films).

EDS microanalysis was made to obtain infor-

mation on the elemental composition of the

sample.

2.4. X-ray powder diffraction

XRD was used to identify the different crystal-

line phases present in pigments. A Philips XPert

vertical goniometer with Bragg-Brentano geome-

try, connected to a highly stabilised generator, wasused for XRD analysis. Cu Ka Ni-filtered radia-

tion, a graphite monochromator on the diffracted

beam and a proportional counter with pulse height

discriminator were used. Measurements in a 5/608

range were taken with a step size of 0.058 and 2

seconds per point.

2.5. Infrared spectroscopy (FTIR)

Absorption spectra in the IR region were

collected using a Nicolet Magna 75 FTIR spectro-meter. Thirty-two signal-averaged scans were ac-

quired on the samples. Few milligrams of each

sample were diluted in KBr (IR grade, Merck)

pellet with a diameter of about 13 mm.

3. Results and discussion

3.1. Pigments identification

The analysed samples, the stratigraphic unitthey belong to, the identified pigments and the

analytical techniques used are shown in Table 1.

3.2. Black

In Fig. 1, we show a 200/ SEM image

performed on a black sample (S.U. 784). From

this image one can see the homogeneous distribu-

tion of a fine-grained black. A trace, left on the

mortar during its preparation, is also visible.

G.A. Mazzocchin et al. / Talanta 61 (2003) 565 /572566


3/8

XRD analysis does not show the presence of

crystalline phases related to the presence of black

colour, and this suggests the use of amorphouscarbon for the preparation of this pigment. FTIR

analysis showed that the pigment is mainly made

of CaCO3, as demonstrated by the bands centred

at 1435, 875 and 715 cm1 assignable to CO3

group and the band centred at 3450 cm1 related

to the water OH group stretching, while the band

around 1100 cm1 is due to small amounts of

silicates.

The black colour disappears when the sample is

heated at 500 8C in muffle oven, confirming the

use of carbon (coal black, soot) as dyestuff. EDS

analysis (Fig. 2) showed the presence of C and

revealed the absence of P (due for example to bone

black) and of dark metal oxides, which suggest the

use of coal black as a basis for the preparation of

this pigment. The presence of Al, Si and Ca is

related to the mortar below.A sample, belonging to the stratigraphic unit

256, showed the presence of a pink colour laying

on a coal black layer. Both colour layers were

spread before the complete exsiccation of the

support because, looking at the FTIR spectrum,

Table 1

The analysed samples, the stratigraphic unit they belong to, the identified pigments and the analytical techniques used

Colour Number of sam-ples Stratigraphicunit Identified pigments Technique

Black 4 784 Coal black FTIR, SEM, EDS, XRD, muffle

oven test

Pink on black 1 256 Coal black FTIR, SEM, EDS, XRD, muffle

oven test

Grey 8 456, 784 Coal black and bone black FTIR, SEM, EDS, XRD, muffle

oven test

Blue 3 784 Cuprorivaite FTIR, EDS, XRD

Clear blue 2 784 Cuprorivaite FTIR, EDS

Red 10 313, 766, 784 Hematite FTIR, EDS, XRD

Violet 3 226, 784 Hematite FTIR, SEM, EDS, XRD

Cinnabar red 2 784, 766 Mercuric sulphide FTIR, SEM, EDS, XRD

Green 5 766, 784 Cuprorivaite, celadonite, glauconite FTIR, SEM, EDS, XRDIvory 2 766 Calcite, yellow ochre FTIR, SEM, EDS, XRD

Yellow 5 766 Limonite, yellow ochre, goethite FTIR, SEM, EDS, XRD

Stripes, bands and

figures

10 456, 784 Lead whites and other pigments

above cited

SEM, EDS

Fig. 1. SEM image of the black sample.

Fig. 2. EDS spectrum of the black sample.

G.A. Mazzocchin et al. / Talanta 61 (2003) 565 /572 567


4/8

around 2900 cm1 we did not find detectable

quantity of organic matter as binder or dyeing

stuff; the matching of different colour layers is

reported in literature[2].The EDS analysis of thepink colour revealed the presence of large quan-

tities of Al, Si and small quantities of Fe and K,

suggesting the presence of clay minerals together

with CaCO3.

3.3. Grey

Many grey samples, belonging to the S.U. 456,

were found in the excavation site.

EDS analysis (Fig. 3) revealed a small amount

of C, big quantity of Ca and O, significantamounts of P, Al and Si, and traces of Fe, Mg,

K and S. Like the black sample, when the grey

sample was heated in a muffle oven at 500 8C, the

colour tends to disappear and this behaviour is

indicative of the presence of carbon.

Looking at the low-magnification SEM images,

one can see that the colour seems to be spread

smoothly throughout and adherent to the layer

below, while calcite reflecting crystals crop out.

XRD analysis (Fig. 4) revealed the presence of

calcite, with traces of quartz and phosphoruscompounds.

FTIR analysis shows the main presence of

calcite, with traces of Fe oxides (560 and 470

cm1) and silicates (900/1100 cm1).

Hence, we can state that our grey sample is not

constituted of ash containing metal oxides, as

suggested by Bearat [3], but mainly of coal black

with traces of bone black and of superficial calcite

crystals added to confer brilliance to the painting.

3.4. Blue and clear blue

All the examined blue samples were constitutedby Egyptian blue (differently diluted by CaCO3), a

compound produced since the time of ancient

Egypt time by means of a high-temperature (/

800 8C) synthesis starting from siliceous sand, a

copper compound, calcium carbonate and sodium

carbonate as a flux. In the correct proportion,

these chemicals together create a product of

formula CaCuSi4O10, which corresponds to the

cuprorivaite mineral[8,9].

EDS analysis of a deep blue sample is shown in

Fig. 5. XRD analysis confirmed that the blue

sample was constituted by cuprorivaite with tracesof silica and residual calcite. The presence of

calcite, as confirmed by FTIR analysis, could

also be indicative of an incomplete synthesis. The

spectrum also reveals the presence of the Si/O

stretching bands at 1011 and 1051 cm1 [10].

This pigment, which Vitruvius and Pliny defined

caeruleus, was found in the wall paintings of

many Roman sites [1,11,12]. It is used alone or

added to other colours to confer more brilliance or

different hues. In the case of the villa in Vicenza,Fig. 3. EDS spectrum of the grey sample.

Fig. 4. XRD spectrum of the grey sample with the crystallinephases calcite, phosphorous oxides and quartz, and a diffuse

halo typical of amorphous substances.



5/8

the blue pigment was added just to the green made

of celadonite and glauconite.

3.5. Red and violet

Red samples were rather numerous; different

red hues were analysed: an orange red; a brick red;

a brilliant red; a violet red and a wine violet.

The XRD analysis of a wine violet, the EDS

spectrum of the violet red (Fig. 6) and the FTIR

allowed us to verify that the colour is due to the

presence of hematite in different quantities, ac-

cording to the areas of FTIR and EDS peaks and

to the hue of the sample: in the lighter samples

(orange red) a smaller quantity of hematite was

present with respect to the darker ones like the

wine violet which is made of pure hematite [6]. By

diluting hematite with CaCO3 a pink colour is

obtained, while brilliant red hues could be ob-

tained by using different ochres.

3.6. Brilliant red*/cinnabar

A particular attention was paid to the brilliant

red sample. In fact, it is not made of hematite, but

of cinnabar (mercuric sulphide). Though cinnabar

is considered a precious pigment, we did not find

substantial differences in the preparation of themortar below, which was made by a Ca(OH)2layer of about 1/2 mm immediately below the

pigment, followed by Ca(OH)2 with crushed

calcite crystals and then by a mortar with sand

and/or tiny river stones. In one of the examined

sample, a thin yellow pictorial film was found

below the brilliant red. It is worth noting that all

the samples are very well preserved [13]. EDS

spectrum is shown in Fig. 7, while the XRD

spectrum is shown inFig. 8. In both spectra, one

can see the sole presence of cinnabar. The extrac-tion of the possible protective films carried out

using methylene chloride or xylene did not show

any characteristic of oils or preservative pitch by

FTIR analysis.

Fig. 5. EDS spectrum of the blue sample.

Fig. 6. EDS spectrum of the violet red sample. Fig. 7. EDS spectrum of the cinnabar sample.



6/8

3.7. Green

There were a number of green samples of

different hues, from grey green to blue green to

mossy green to grass green.

Some samples are characterised by a double

pictorial layer: a dark grey under a light green

layer.

XRD spectra of a grass green sample pointedout the presence of celadonite and cuprorivaite

besides traces of glauconite. The presence of

celadonite confers to the sample the green colour.

3.8. Ivory

An ivory white sample was made of calcite with

traces of dolomite and iron oxides, as can be

clearly seen from EDS (small peak of Fe) and

FTIR (560 and 490 cm1) spectra. This hue of

colour can be reached also by pinkish and greenpigments which underwent changes as time passed.

3.9. Yellow

The yellow ochre painting pigments analysed

are characterised by a double pictorial layer. We

found a yellow on red and a yellow on a pink

basis. We also got a yellow with plain traces of

coal black on the surface. In Fig. 9, the XRD

spectrum of a yellow sample shows goethite as

main crystalline phase, besides other iron hydro-

xides.

EDS and FTIR analyses confirmed that the

pictorial layer is mainly made of iron oxides and

hydroxides[6]. The absence of As and Pb allowed

us to exclude the use of orange compounds like

Pb3O4 or As2S3 to obtain the yellow colour. Thepresence of Al, Fe, Si and K, revealed by means of

EDS analysis, suggests the use of clay minerals,

which allowed us to define the yellow pigment as

an ochre.

The yellow ochre sample with traces of coal

black put in muffle oven at 500 8C showed the

disappearance of the surface blackish patina and

the persistence of the yellow, which got a darker

hue. Consequently, the coal black can be ascribed

to the residual combustion products due to the

lighting of the env

ironment, or as a consequence ofa fire.

3.10. Non-homogeneous pieces (lines, stripes,

figures)

Together with monochromatic wall painting

pieces, many fragments with white, pink, dark

red, and blue lines and stripes on a grey layer were

found. Some of these pieces are shown in Fig. 10.

Fig. 8. XRD spectrum of the cinnabar sample with the

crystalline phases cinnabar and calcite. Fig. 9. XRD spectrum of the yellow sample with the following

crystalline phases: goethite, quartz, calcite and other iron

hydroxides.



7/8

FTIR analysis of the grey layer showed the

typical bands of the carbonate and of the silicon/

oxygen group. Generally, the grey of the striped

samples had the same characteristics of the mono-

chromatic grey sample. XRD analysis confirmed

that the pigment is mainly made of calcite. The

crystals on the surface confer brightness to the

colour.

The v

iolet and blue stripes of the pieces inFig.10did not show any novelty, since they were made,

respectively, of hematite and cuprorivaite, while

the white line of these pieces revealed the presence

of Pb, due to the use of cerussite (PbCO3), or white

lead (2PbCO3 /Pb(OH)2).

Among the non-homogeneous pieces there was

a blackish grey angular piece (decorated with

white, pink and blue stripes) which, most prob-

ably, was the rounded edge of a wall. The material

below the painting was very hard, perhaps because

accurately worked. SEM, EDS and FTIR analyses

showed that the pigment used was the same blackcoal of the grey monochromatic sample previously

described. Also in this case, the pink line was made

of hematite diluted by calcite, the white was made

of pure calcite and the blue was cuprorivaite.

In this group of samples we also examined a wall

fragment painted with colours from lilac to pink,

with blue strokes of brush and weak yellow stripes

(Fig. 11). The main components of the used

pigments are analogous to the red and blue already

described for the monochromatic samples, except

for the thick pink colour, which is not made of

diluted hematite, but of diluted cinnabar, confirm-

ing the preciousness of the painting.

4. Conclusions

Ev

en if modest with respect to the v

ariety ofcolours found in other archaeological sites like in

Rome or Pompeii [1,11,12], the Vicenza palette is

rich and includes, besides the ordinary pigments,

precious pigments like cinnabar and Egyptian blue

and well-done paintings.

The adding of refracting materials such as

Egyptian blue was used exclusively for the green

pigment, while calcite was used for the grey one;

we also found the use of different combinations of

pictorial layers.

We did not find a wide use of lead as orange

(minium, Pb3O4) or as white (cerussite, PbCO3;white lead, 2PbCO3 /Pb(OH)2), while we found

practically the whole gamut given by hematite

(orange red, red, brick red, violet) and by cinnabar

(brilliant red and pink). We never identified

organic ligands or pigments which could lead one

to think about pictorial techniques different from

frescoes.

We did not find As-based pigments which, on

the contrary, were found in Gallia, in Argentoma-

gus[14].

Fig. 10. Picture of some pieces with white, pink, dark red andblue lines on a grey layer.

Fig. 11. Picture of a wall fragment painted with colours fromlilac to pink and with blue strokes and yellow stripes.



8/8

Different kind of mortars were used for the

walls of this villa, but, in general, the preparation

of these supporting materials was not exactly the

one already suggested[15,16].

However, we think that the white and porous

layer of Ca(OH)2 with crushed calcite crystals of

about 2 mm of length (Fig. 12) was made on

purpose in order to make the layer resistant and at

the same time resilient and porous. Finally, we can

state that the employed analytical techniques are

suitable to give the information required for thecharacterisation of the pigments found.

Acknowledgements

The authors are grateful to Dr. Carlo Bragato

and Mr. Danilo Rudello for helpful assistance.

References

[1] J.A. Chaptal, Sur quelques couleurs trouvees a

Pompeia Annales de Chimie et Physique, in: S.Augusti (Ed.), I colori Pompeiani, studi e

documentazioni I, Vol. 70, 1809, pp. 22/31

(Roma 1967).

[2] H. Bearat, in: Proceedings of the International Workshop

on Roman Wall Paintings, Fribourg, 1997,

pp. 11/34.

[3] H. Bearat, Archaeometry 38 (1) (1996) 81/95.

[4] F. Delamare, E. Darque-Ceretti, J.E. Dietrich, M.C. Gay,

J.M. Monget, Revue dArcheometrie 6 (1982) 71/86.

[5] M. Frizot, Revue dArcheometrie 6 (1982) 47/59.

[6] M.C. Edreira, M.J. Feliu, C. Fernandez-Lorenzo, J.

Martn, Anal. Chim. Acta 434 (2001) 331/345.

[7] M.T. Domenech-Carbo, F. Bosch-Reig, J.V. Gimeno-Adelantado, V. Periz-Martnez, Anal. Chim. Acta 330

(1996) 207/215.

[8] C. Minguzzi, Periodico di Mineralogia 9 (1938) 333.

[9] A. Pabst, Acta Crystallogr. 12 (1959) 733.

[10] P. Mirti, L. Appolonia, A. Casoli, R.P. Ferrari, E.

Laurenti, A. Amisano Canesi, G. Chiari, Spectrochim.

Acta A 51 (1995) 437/446.

[11] G. Bortolaso, V. Massa, Il giardino dipinto nella casa del

bracciale doro a Pompei, in: Exposition Catalogue,

Florence, 6 July/11 August 1991, pp. 13/18.

[12] M. Pancani, M. Seracini, S. Vannucci, Alla ricerca di Iside:

analisi, studi e restauri dellIseo Pompeiano nel museo di

Napoli, Napoli, 1991, pp. 123/132.

[13] V. Daniels, Recent Advances in the Conservation and

Analysis of Artefacts, Summer School Press, London,

1987, pp. 280/282.

[14] B. Guineau, I. Fauduet, J.-M. Biraben, Etude de frag-

ments de couleurs recueillis sur le site dArgentomagus,

Anzeiger der Romisch-Germanischen Kommission des

Deutschen Arrchaologischen Instituts, 1995.

[15] R. Giovanoli, Archaeometry 11 (1969) 53.

[16] A. Barbet, Romana Pictura, Electa Eds, 1998.

Fig. 12. SEM image of crushed calcite grains under the

cinnabar pigment layer.


analysis of pigments from roman wall paintings found in vicenza

Documents