Vibrational Spectroscopy 53 (2010) 274278

Contents lists available at ScienceDirect

Vibrational Spectroscopy
journal homepage: www.elsevier.com/locate/vibspec

Macro-ATR-FT-IR spectroscopic imaging analysis of paint cross-sections

E. Joseph a,1 , C. Ricci b , S.G. Kazarian b, , R. Mazzeo a, , S. Prati a , M. Ioele c
a

Microchemistry and Microscopy Art Diagnostic Laboratory, University of Bologna, via Guaccimanni 42, 48100 Ravenna, Italy
Department of Chemical Engineering, Imperial College London, London, SW7 2AZ, UK
c
Istituto Superiore per la Conservazione e il Restauro, Piazza San Francesco di Paola 9, 00184 Rome, Italy
b

a r t i c l e

i n f o

Article history:
Received 11 March 2010
Received in revised form 19 April 2010
Accepted 19 April 2010
Available online 28 April 2010
Keywords:
Fourier transform infrared spectroscopy
Attenuated total reection
Imaging
FTIR imaging
ATR microscopy
Conservation studies
Painting materials

a b s t r a c t
Macro-attenuated total reection Fourier transform infrared (ATR-FT-IR) spectroscopic imaging is presented as non-destructive analytical methodology with respect to the sample for the study of paint
surfaces. This paper illustrates new opportunities offered by a conventional diamond ATR accessory coupled with a focal plane array detector to obtain chemical images of multi-layered paint cross-sections.
The use of a large internal reection element permits to increase the eld of view of the imaging area
up to 700 m 500 m and to obtain FT-IR images of the complete stratigraphy of the sample. This
approach represents a versatile analytical tool to determine the composition of the inorganic and organic
painting materials and to obtain information on their distribution within the different layers. In the
optical arrangement used, a spatial resolution of approximately 15 m is achieved without the use of
infrared microscope. This study demonstrates the possibilities offered by macro-ATR-FT-IR imaging for
a simple identication of the different compounds present in paint cross-sections before performing
further molecular and elemental analyses which may be time consuming or require particular sample
preparation.
2010 Elsevier B.V. All rights reserved.

1. Introduction
Fourier transform infrared (FT-IR) spectroscopy is commonly
used for the study of the inorganic and organic components of
paintings and represents a useful analytical tool in the understanding of painting techniques, state of conservation and causes of
degradation. The characterisation of the pigmentbinder interactions and degradation processes taking place within the complex
multi-layered paint structures is generally carried out on crosssectioned micro-samples collected from paintings. FT-IR imaging,
which makes use of a focal plane array (FPA) detector, is one of
the few methods, providing both spectral and spatial information
on the characterisation and distribution of the different painting
materials within each individual paint layer constituting a complex
heterogeneous matrix. In particular the combination of attenuated
total reection (ATR) FT-IR spectroscopy and an imaging system
provides an improvement of the spatial resolution due to the high
refractive index of the ATR crystal, according to the Rayleigh cri-

Corresponding author. Tel.: +44 2075945574; fax: +44 2075941989.

Corresponding author. Tel.: +39 0544937150; fax: +39 0544937159.
E-mail addresses: edithjoseph@gmail.com (E. Joseph),
camilla.elena.ricci@googlemail.com (C. Ricci), s.kazarian@imperial.ac.uk
(S.G. Kazarian), rocco.mazzeo@unibo.it (R. Mazzeo), s.prati@unibo.it (S. Prati),
ioele@tiscali.it (M. Ioele).
1
Present address: Muse national suisse, Sammlungszentrum, Lindenmoosstrasse 1, CH-8910 Affoltern a. Albis, Switzerland.
0924-2031/$ see front matter 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.vibspec.2010.04.006

terion [13]. Micro-ATR measurements are performed using an

infrared microscope with an ATR objective. For macro-ATR measurements, a single reection diamond ATR accessory is directly
attached to an FT-IR spectrometer, as is the case presented here. So
far, macro-ATR-FT-IR imaging has been used in different elds, such
as pharmaceutical [46], biomedical [7], forensic [810] or polymer
[11] studies. Only a few examples of micro-ATR imaging related to
cultural artefacts have been reported in the literature: the study of
albumen photographs [12], painting cross-sections [13,14] and the
identication of mineral components in a Maya paint [15]. However, to our knowledge, no examples were reported for the use of
macro-ATR imaging in conservation science and this study represents the rst application of such technique in the eld. Recently
the use of large internal reection elements has been developed and
the possibilities of increasing the imaging area have been demonstrated elsewhere [1,2,6,7,9,10,16,17]. Despite the improvement of
the analysed area achieved with these accessories, problems of contact uniformity between sample and crystal may sometimes arise,
especially in the case of hard or non-easily deformed samples. Due
to its hardness which allows a high pressure to be applied, diamond is very useful material for ATR-FT-IR accessories. Using such
arrangement, reproducible spectra of hard and soft materials are
successfully measured with a uniform contact between sample and
crystal [1,2,6]. In this work, the potential offered by macro-diamond
ATR-FT-IR imaging to the study of paint cross-sections, in terms
of characterisation and localisation of the compounds present, is
illustrated.

E. Joseph et al. / Vibrational Spectroscopy 53 (2010) 274278

2. Materials and methods

2.1. Samples
Three historic samples collected from (a) a piece of gilt decorated
leather (1718th century AD) manufactured in central Italy (sample CD), (b) the blue robe of the Madonna (sample Fe2) of a 14th
century AD Italian polychrome sculpture belonging to a private collection and (c) the red robe of the Madonna (sample MCR8) of the
terracotta altar-piece Assumption of the Virgin between S. Sebastian
and S. Rocco (16th century AD) by Andrea della Robbia, located in
the collegiate church of S. Maria Assunta in Montecassiano (Macerata, Italy). All samples have been rst carefully observed under
a binocular stereomicroscope Leica MZ6. Samples MCR8 and Fe2
have been embedded in a polyester resin support (Serix. Struers),
then cross-sectioned and submitted to wet polishing with conventional methods using silicon carbide papers with grit 120, 400, 800
to 1000.
2.2. Optical microscopy
Dark eld observations of cross-sectioned samples have been
performed using an optical microscope Olympus BX51M equipped
with xed oculars of 10 and objectives with different magnications (5, 10, 20, 50 and 100). Visible and ultraviolet lights
were respectively provided by a 100 W halogen projection lamp
and an Ushio Electric USH102D lamp. Photomicrographs have been
recorded with a digital scanner camera Olympus DP70 directly connected to the microscope. The observations under UV light are
commonly used as a rst tool to qualitatively detect and stratigraphically locate the presence of organic materials (binding media,
varnishes, etc.) as a result of the autouorescence of such materials
when submitted to a UV radiation source.
2.3. ATR-FT-IR spectroscopic imaging
An FPA detector was used to measure FT-IR images with an FTIR spectrometer operating in continuous scan mode together with
a diamond ATR accessory (Golden GateTM , Specac Ltd.) at Imperial College London. The imaging ATR spectrometer is patented by
Varian [18]. In the optical arrangement used, the spatial resolution
was approximately 15 m [2]. Spectra were collected with 8 cm1
spectral resolution in the range 4000900 cm1 and 64 scans. The
FPA detector comprised 4096 small pixels arranged in a 64 64
grid format with each of the pixels measuring an infrared spectrum; thus, 4096 spectra were acquired in a single measurement. A
plot of the distribution of the absorbance of a specic spectral band
corresponding to a particular substance in the imaged area of the
sample produced the chemical image for that particular substance.
The imaging area measured using this accessory and this particular
FPA detector was 700 m 500 m. To ensure that good contact
was established between the cross-sections and the crystal, a drop
of water was added, acting as immersion medium for the infrared
radiation and compensating the micro-irregularities of the surface.
The remaining water absorbance bands present in the spectrum
have been further subtracted. A colour scale from high value (red)
to low value (blue) is employed.
3. Results and discussion
3.1. Macro-ATR-FT-IR imaging of non-embedded fragment
3.1.1. Gilt leather (sample CD)
Leather gilding originates in Spain and in particular from Cordoba but it has also been manufactured in Italy. The typical golden
aspect was achieved by rst an application of a metal leaf (tin,

275

silver or copper) adhered to the leather by the use of glue. The

metal surface was then brushed with egg white in order to avoid
metal oxidation. A nal nishing (mecca) made up of a natural varnish containing yellow organic pigments (e.g. saffron, aloe)
was then applied to achieve the golden colour. In the attempt to
characterise the uppermost varnish layer of sample CD, macro-ATR
measurements have been performed directly on the surface of the
gilt leather fragment. A view of the leather fragment is shown in
Fig. 1a. The FT-IR image created by plotting distribution of the integrated absorbance of the CO fatty ester stretching band between
1270 and 1210 cm1 , which is considered as marker for the presence of a natural resin [19], is presented in Fig. 1b. The spectrum
extracted from the higher intensity area points out to the presence
of free fatty acids as well as different bands attributable to a natural
resin, possibly assignable to shellac (1698, 1453, 1415, 1237, 1167
and 1038 cm1 ) as shown in Fig. 1h with the reference spectrum
of shellac for comparison. The distribution plot of the absorbance
of the integrated C O carbonyl stretching band between 1720 and
1685 cm1 (Fig. 1c) coincides with the CO distribution plot and
conrms the attribution of the carbonyl band to a natural resin, possibly shellac. Vibrational bands assigned to the presence of calcium
oxalates were also identied in the spectrum. The varnish distribution does not appear homogeneous and some degradation products
have been identied in the blue areas of the varnish distribution
plot which overlap with an area where the metal leaf appears
blackened by corrosion (Fig. 1df). In the presence of moisture and
corrosive gases (sulphur dioxide, ozone, nitrogen oxides), different
corrosion products can form depending on the atmospheric conditions. The spectrum extracted from this area indicates here the
presence of calcium oxalates (sym COO 1373, 1318 cm1 , Fig. 1d),
sulphates (asym SO4 2 1198 cm1 , Fig. 1e) and silicate deposition
materials ( SiO 1076, 1035 cm1 , Fig. 1f), as shown in Fig. 1i. All
these compounds are often observed during degradation processes
in indoor environment [2022] and even if the extracted spectrum
is rather complex, it has been possible to identify each of them
due to the presence of characteristic absorbance bands. Moreover,
in the same area a proteinaceous material (amide I 1637, amide
II 1535 cm1 ) has been identied and can probably be assigned
to the presence of egg white originally applied to avoid the oxidation of the metal leaf (Fig. 1g). In particular, the amide I band
is observed as a shoulder on the 1617 cm1 peak, whose broadness towards higher wavenumbers cannot be attributable to the
solely calcium oxalates, as shown in Fig. 1i. It is noted that the relative absorbance of the amide I and II bands do not reect the usual
ngerprint prole of proteinaceous materials. This feature may suggest the contribution of metal carboxylates to the absorbance of the
amide II band, which is nearly as strong as the amide I band. The
metal carboxylates may be formed as degradation products. In this
example, diamond ATR-FT-IR imaging measurements allowed the
advanced characterisation of the original natural resin as well the
degradation products and deposition material in a non-destructive
way towards the sample. This analytical technique can therefore be
used not only for the material characterisation but also, being nondestructive, repeated at different periods of time opening up novel
analytical possibilities in the monitoring of ongoing deterioration
processes and performance of restoration materials.
3.2. Macro-ATR-FT-IR imaging of paint cross-sections in polyester
resin
3.2.1. Blue painted area from Italian polychrome sculpture
(sample Fe2)
Sample Fe2 shows a complex stratigraphy with two different
blue paint layers (thickness of 30 and 20 m respectively for the
light blue layer 3 and the dark blue layer 4) applied over a gold leaf
(layer 2) under which a red gilding ground (layer 1, 10 m thick)

276

E. Joseph et al. / Vibrational Spectroscopy 53 (2010) 274278

Fig. 1. (a) View of the fragment under stereomicroscope; (b) FT-IR image representing the integrated absorbance of the natural resin band between 1270 and 1210 cm1 ;
(c) FT-IR image representing the integrated absorbance of the natural resin band between 1720 and 1685 cm1 ; (d) FT-IR image representing the integrated absorbance of
the oxalate band between 1355 and 1290 cm1 ; (e) FT-IR image representing the integrated absorbance of the sulphate band between 1215 and 1185 cm1 ; (f) FT-IR image
representing the integrated absorbance of the silicate band between 1100 and 1060 cm1 ; (g) FT-IR image representing the integrated absorbance of the amide II band
between 1570 and 1485 cm1 ; (h) FT-IR spectrum extracted from the top central area of b, marked sp1, compared with a reference spectrum of shellac; (i) FT-IR spectrum
extracted from the top right corner of d, marked sp2. The size of the FT-IR images is 700 m 500 m. The gure is available in colour in the online version via Science Direct.

Fig. 2. Detail of a cross-section from the polychrome sculpture (Fe2): (a) visible microscopic image; (b) image of sample under ultraviolet light; (c) FT-IR image created by
plotting the integrated absorbance of the embedding resin band between 1330 and 1200 cm1 ; (d) FT-IR image showing the distribution of the silicate integrated absorbance
between 1050 and 1000 cm1 ; (e) FT-IR image showing the distribution of the azurite integrated absorbance between 970 and 920 cm1 ; (f) FT-IR image showing the
distribution of the carbonate integrated absorbance between 1335 and 1440 cm1 ; (g) FT-IR image showing the distribution of the amide II integrated absorbance between
1580 and 1480 cm1 ; (h) FT-IR image showing the distribution of the triglycerides integrated absorbance between 1765 and 1725 cm1 ; (i) FT-IR spectrum extracted from
the right area of h, marked sp1. The size of the FT-IR images is 700 m 500 m. The gure is available in colour in the online version via Science Direct.

E. Joseph et al. / Vibrational Spectroscopy 53 (2010) 274278

277

Fig. 3. Detail of a cross-section from the terracotta altar-piece (MCR8): (a) visible microscopic image; (b) image of sample under ultraviolet light; (c) FT-IR image representing
the integrated absorbance of the natural resin band between 1330 and 1200 cm1 ; (d) FT-IR image representing the integrated absorbance of the gypsum band between 1175
and 1085 cm1 ; (e) FT-IR image representing the integrated absorbance of the carbonate band between 1440 and 1340 cm1 ; (f) FT-IR image representing the integrated
absorbance of the amide II band between 1570 and 1505 cm1 ; (g) FT-IR image representing the integrated absorbance of the fatty ester band between 1760 and 1720 cm1 ; (h)
FT-IR spectra of different areas of b, extracted from the inner layers 13, marked sp1 and from the top layers 57, marked sp2. The size of the FT-IR images is 700 m 500 m.
The gure is available in colour in the online version via Science Direct.

and a white ground layer 0 are visible (Fig. 2a). Under UV illumination, an uppermost brownish layer (layer 5, 15 m thick), which is
observed with difculties in the reected light microphotograph,
uoresces yellowish and a strong uorescence is also evident in
the white ground and light blue layers (Fig. 2b). Macro-ATR-FT-IR
imaging measurements have been performed and the whole data
cube has been analysed in order to obtain chemical images for
each identied painting material. In order to determine the area of
the cross-section which was analysed, an FT-IR image was created
for the polyester embedding resin by plotting distribution of the
integrated absorbance of the ( CO) band at 1253 cm1 (Fig. 2c).
The integrated absorbance of the SiO stretching band occurring
between 1050 and 1000 cm1 has been used to characterise a silicate based blue pigment (Fig. 2d). Azurite (a natural basic copper
carbonate blue pigment) has been dened by integrating the C O
bending at ca. 950 cm1 (Fig. 2e). Finally a distribution plot of lead
white (a basic lead carbonate), which has been already characterised by the authors in a previous work [23], has been obtained
by integrating the C O asymmetric stretching band between 1440
and 1340 cm1 . The carbonate absorption bands of lead white and
azurite are overlapped, resulting in an inaccurate image for the
localisation of lead white, which also represents the distribution
of azurite (Fig. 2f). On the other hand, the integration of the amide
II band at ca. 1530 cm1 has been used to plot the distribution of
a proteinaceous binder (Fig. 2g) which seems to be present in all
painting layers. The spectra collected from the dark blue outer layer
4 are additionally characterised by a carbonyl band at 1730 cm1 ,
associated with triglycerides, suggested the use of egg tempera as
binding medium (Fig. 2h). The same band is visible in the uppermost
layer 5 and a mixture of siccative oil and natural resin together with
proteinaceous materials seems to have been used as varnish. In fact,
the spectrum extracted from the red area in the layer 5, as shown in
Fig. 2i, points out the presence of triglycerides (1732 cm1 ) together
with some characteristic absorbance bands of a natural resin (1710,
1454, 1413, 1272 cm1 ) as well as the amide I (1635 cm1 ) and
amide II (1534 cm1 ) bands. Some degradation products, such as
oxalates at 1379, 1315 cm1 and deposition materials (silicates
at 1004, 981 cm1 and calcium sulphate at 1159, 1114 cm1 ) are
also present. The thickness of the painting layers in respect to the

dimension of the imaging area does not permit establishing the

exact localisation of the identied components between the different blue layers. It was challenging to identify compounds present
in the area of the sample under the gold leaf (red and white ground
layers) and only the absorption bands of water (used as an immersion medium for improving quality of contact) were present in the
spectrum. It is probably due to material discontinuities within the
different layers, which have been generated during the polishing
procedure used for the cross-section preparation, resulting in a
non-perfect ATR crystal/surface contact. However, even if the interpretation of these ground layers has not been yet resolved here,
this method could certainly be useful for other case studies. The
polishing procedure could also be improved paying attention at
using micromeshTM polishing cloths with no abrasion uid instead
of water [24].
3.2.2. Red painted area from Italian terracotta altar-piece
(sample MCR8)
A complex stratigraphy is visible under the optical microscope
observations of the sample MCR8, which is subdivided in 9 layers
as shown in Fig. 3. A white layer with red-orange inclusions (layer
8, thickness of 65110 m) is laid over a thin orange layer (layer 7,
2050 m thick) and a pink with white and red particles layer (layer
6, thickness ca. 60 m). Below a transparent brownish layer (layer
5, 1535 m thick) is applied over a light pink layer with bright
red inclusions (layer 4, thickness of 1540 m) and a discontinuous thin transparent layer (numbered 3) is visible on the right part
of Fig. 3a. A transparent red layer (layer 2, 2545 m thick) is then
superimposed to a dark orange layer (3580 m thick, layer 1) and
a white layer with some orange particles (layer 0). Under UV illumination, the layers 3 and 5 showed a light blue uorescence and
in all other layers a bright uorescence is evident (Fig. 3b). MacroATR-FT-IR imaging has been performed on the cross-section and
the resulting chemical images are presented in Fig. 3. The distribution of the embedding resin was plotted using the integrated
absorbance of the CO stretching band at ca. 1235 cm1 to demonstrate the area of the cross-section which was studied (Fig. 3c).
Several inorganic compounds have been identied and distribution
plots were obtained by integrating characteristic spectral bands

278

E. Joseph et al. / Vibrational Spectroscopy 53 (2010) 274278

for each component (Fig. 3d and e): gypsum (SO4 2 asymmetric

stretching 11451110 cm1 ) has been located in the layer 0 and carbonates using the CO3 2 asymmetric stretching 14401340 cm1
in the layers 17. However distinction between calcium or lead
carbonate has not been possible since the asymmetric stretching
principal bands are close together and the bending ones (875 and
835 cm1 respectively for calcite and lead white) are absent from
the spectral region investigated (4000900 cm1 ). The SEM-EDX
results (not shown) suggest that red lead (layers 1 and 6), haematite
(layers 6 and 7) and vermilion (layer 2) have been used. It is worth
saying that all these red pigments are inactive in the mid-infrared
region here investigated. It has not been possible to characterise
red lake, which has been further identied in layers 25 by FT-IR
transmission diamond cell on isolated particles, as madder lake.
In fact, this pigment is present in very low concentration and small
particle size in an inorganic matrix much more absorbent and these
conditions are typically under the detection and/or spatial resolution limits of the system used. More interesting is the localisation of
organic substances. The integrated absorbance of amide II band at
ca. 1530 cm1 was used to plot the distribution of a proteinaceous
binding medium (Fig. 3f). Although it is widely distributed among
the different painting layers, the spectra collected from the outer
layers 57 are characterised by a more intense amide II band than
the amide I (1636 cm1 ) band. In the ngerprint prole showed by
proteinaceous molecules, the absorbance of the amide II band is
weaker than that of the amide I as shown in Fig. 3h. Therefore, this
feature may suggest a contribution from metal carboxylates. The
distribution plot of triglycerides using the integrated absorbance
of C O ester asymmetric stretching 17601720 cm1 shows also
the presence of lipidic compounds in all layers numbered from 0
to 7, allowing the attribution to egg tempera as binding medium in
these layers (Fig. 3g).
4. Conclusions
This study highlighted the possibility of performing macroATR-FT-IR spectroscopic imaging using a diamond ATR accessory
on paint cross-sections for the characterisation and localisation
of either inorganic compounds or organic substances present as
painting materials. Macro-ATR spectroscopic measurements can
be performed in a relatively easy way and the only requirement
is to achieve a good contact between the sample and the diamond crystal. In order to obtain contact uniformity, water has been
successfully used as an immersion medium on the cross-sections
embedded in polyester resin. In only few cases where the absorption bands of water were prominent in the spectrum, spectral
subtraction has been employed. The study will be further extended
to the use of chlorinated solvents, such as carbon tetrachloride
or tetrachloroethylene, which have no absorbance bands in the
spectral region investigated (4000900 cm1 ). The same procedure
could also be applied to cross-sections embedded in KBr, transparent to mid-IR radiation because traditional embedding polyester

resin, depending on the amount and penetrability degree of the

binder, does not allow the identication of the organic substances,
in particular for those of lipidic nature with any certainty.
The method presented is a fast qualitative analysis which can
be performed without the use of an infrared microscope, which
is required for micro-ATR measurements, and can represent the
rst non-invasive step with respect to the sample in an analytical
approach that allows further elemental and molecular analyses.
These promising results open up opportunities for future applications which may deal with other conservation issues, such as,
for example, the corrosion behaviour of metallic objects and the
characterisation of the degradation products.
Acknowledgments
CR and SGK thank EPSRC for support (Grant EP/C532678/1). This
research has been partially carried out with the support of the
European Union, within the VI Framework Programme (Contract:
EU-ARTECH, RII3-CT-2004-506171).
References
[1] S.G. Kazarian, K.L.A. Chan, Appl. Spectrosc. 64 (2010) 135A152A.
[2] K.L.A. Chan, S.G. Kazarian, Appl. Spectrosc. 57 (2003) 381389.
[3] A.J. Sommer, L.G. Tisinger, C. Marcott, G.M. Story, Appl. Spectrosc. 55 (2001)
252256.
[4] K.L.A. Chan, S.V. Hammond, S.G. Kazarian, Anal. Chem. 75 (2003) 21402146.
[5] C. Ricci, L. Nyadong, F. Fernandez, P. Newton, S.G. Kazarian, Anal. Bioanal. Chem.
387 (2007) 551559.
[6] K.L.A. Chan, S.G. Kazarian, J. Biomed. Opt. 12 (2007), 044010.
[7] C.S. Colley, S.G. Kazarian, P.D. Weinberg, M.J. Lever, Biopolymers 74 (2004)
328335.
[8] C. Ricci, K.L.A. Chan, S.G. Kazarian, Appl. Spectrosc. 60 (2006) 10131021.
[9] C. Ricci, P. Phiriyavityopas, N. Curum, K.L.A. Chan, S. Jickells, S.G. Kazarian, Appl.
Spectrosc. 61 (2007) 514522.
[10] B.M. Patterson, G.J. Havrilla, C. Marcott, G.M. Story, Appl. Spectrosc. 61 (2007)
11471152.
[11] P. Zhang, D.B. Otts, M.W. Urban, PMSE Preprints 87 (2002) 170171.
[12] C. Ricci, S. Bloxham, S.G. Kazarian, J. Cult. Herit. 8 (2007) 387395.
[13] M. Spring, C. Ricci, D. Peggie, S.G. Kazarian, Anal. Bioanal. Chem. 392 (2008)
3745.
[14] E. Joseph, S. Prati, G. Sciutto, M. Ioele, P. Santopadre, R. Mazzeo, Anal. Bioanal.
Chem. 396 (2010) 899910.
[15] R.A. Goodall, J. Hall, R.J. Sharer, L. Traxler, L. Rintoul, P.M. Fredericks, Appl.
Spectrosc. 62 (2008) 1016.
[16] B.M. Patterson, G.J. Havrilla, Appl. Spectrosc. 60 (2006) 12561266.
[17] K.L.A. Chan, S.G. Kazarian, Appl. Spectrosc. 62 (2008) 10951101.
[18] E.M. Burka, R. Curbelo, ATR imaging spectrometer, US Patent 6,141,100 (2000).
[19] D. Omecinsky, G.W. Carriveau, AIC 10th Annual Meeting Preprints, Milwaukee
(Wisconsin), 1982, pp. 141149.
[20] A.R. Mendoza, F. Corvo, Corros. Sci. 42 (2000) 11231147.
[21] R. Mazzeo, in: P. Tiano, C. Pardini (Eds.), Le patine: genesi, signicato e conservazione, Nardini Editore, Firenze, 2005, pp. 2943.
[22] R. Mazzeo, A.M. Carruba, G. Chiavari, C.P. Presicce, in: I.D. MacLeod, J.M. Theile,
C. Degrigny (Eds.), Metal 2001Proceedings of the ICOM Committee for Conservation Metals Working Group, Santiago, Chile, 2001, pp. 3443.
[23] R. Mazzeo, E. Joseph, S. Prati, A. Millemaggi, Anal. Chim. Acta 599 (2007)
107117.
[24] A.V. Loon, K. Keune, J. Boon, in: C. Parisi, G. Buzzanca, A. Paradisi (Eds.),
Art058th International Conference on the Non Destructive Investigations
and Microanalysis for the Diagnostic and Conservation of the Cultural and
Environmental Heritage, Lecce, 2005, B-089.