Study of the frescoes in Ione știi Govorii wooden church (Romania) using [617732]

Study of the frescoes in Ione știi Govorii wooden church (Romania) using
multi-technique investigations ☆
Ileana Mohanua,⁎, Dan Mohanub, Ioana Gomoiuc, Olimpia-Hinamatsuri Barbud, Roxana-Magdalena Fecheta,
Nicoleta Vlada,G e o r g e t aV o i c ue, Roxana Tru șcăf
aSC Ceprocim SA, 6 Preciziei Blvd, 6-Bucharest 062203, Romania
bNational Arts University, 19 General Budi șteanu, 1-Bucharest 010773, Romania
cInstitute of Biology, Romanian Academy, 296 Spl Independentei, 1-Bucharest 060031, Romania
dNational Museum of Romanian History, Centre of Research and Scienti fic Investigation, 12 Calea Victoriei, 030026 3-Bucharest, Romania
eFaculty of Applied Chemistry and Materials Science Department of Science and Engineering of Oxide Materials and Nanomaterials, University Politehnica of Bucharest, 1-7,Gh. Polizu str.,
1-Bucharest 11061,Romania
fMetav Research Development srl, 31 C.A. Roseti Str., Sector 2, Bucharest 020011, Romania
abstract article info
Article history:
Received 10 August 2015Received in revised form 10 December 2015Accepted 16 December 2015Available online 24 December 2015Romanian wooden churches with a secco ornamentation have Slavic counterparts in Central and Northern
Europe, but the Romanian wooden churches with inner and outer frescoes constitute a unique heritage in thatthey require speci fic conservation techniques. Although there are a number of initiatives to rescue this heritage,
to the best of our knowledge, no one has characterised these frescoes from the wooden churches using a varietyof analytical techniques. This paper aimed to investigate the state of conservation and the composition ofthe frescoes in the wooden church of Ione știi Govorii Commune, Vâlcea County, Oltenia Region, Romania. The
investigations were undertaken to characterise the materials, techniques and degradation products of the frescopaintings on wood and to help answer speci fic questions regarding the state of conservation of these paintings. In
situ observations have revealed an advanced state of degradation of the monument with wide gaps in the frescoand substantial fresco detachment from the wooden wall. The mortar used for the fresco consists of fine-grained
carbonated lime (used as binder), fine-grained silica aggregate, porous fragments of limestone and cellulosic
(plant) fibres. The mass of the lime plaster has preserved mineral fragments of bioclasts pertaining to foraminif-
era and shell fragments. The non-invasive and micro-destructive analyses enabled us to reconstruct the palette
used by the painter. Non-germinated spores of filamentous fungi are found on the surface of the wooden struc-
ture. Furthermore, SEM images have revealed hyphae on cross-sections in the lime plaster ( intonaco ) and at the
surface of the painted layer. Speci fic investigation techniques have revealed the presence of gypsum and calcium
oxalate on the painted surface of the fresco.
© 2015 Elsevier B.V. All rights reserved.Keywords:
Wooden churchFresco paintingLime lumpsATR-FTIRSEM-EDX
1. Introduction
The complexity and dif ficulty of the conservation and restoration
techniques for ecclesiastical monuments made of wood make them
one of the most pressing current issues in the protection of the
European heritage. Certain wooden churches [1]in such countries as
Norway, Poland, and the Ukraine are World Heritage Sites (WHSs). InRomania, in the region of Maramures, eight of the wooden churches
are also on the list of WHSs [2]. Their value is de fined on the one handby architecture and on the other hand by painted decoration. The latter
often involves a complex iconography [3], typically decorating the
interior of the church and, in certain cases, also the outer walls of themonument.
In Romania, two painting techniques have been used for the painted
decoration of wooden churches: secco painting [4]and fresco painting,
applied on various types of lime plaster. For secco painting, thewooden boards were previously prepared with gesso ground [5].A s
M. F. Mecklenburg et al. note [6], gesso can be subjected to greater
humidity changes than most of the other materials, meaning that agesso/wood composite can survive the signi ficant variations in relative
humidity. By contrast, the fresco painting on wooden walls, which hasnot been studied suf ficiently to date, constitutes a special case, in that
fresco painting involves applying a masonry-speci fic technique to wood.
Recent initiatives of in situ saving and conserving the wooden
churches in Romania [3]have offered models of interdisciplinaryMicrochemical Journal 126 (2016) 332 –340
☆Selected papers presented at TECHNART 2015 Conference, Catania (Italy), April 27-30,
2015.
⁎Corresponding author.
E-mail addresses: ileana.mohanu@ceprocim.ro (I. Mohanu),
dan_ileana_m@yahoo.com (D. Mohanu), gomoiu@hotmail.com (I. Gomoiu),
barbuolimpia@gmail.com (O.-H. Barbu), georgeta.voicu@upb.ro (G. Voicu),
truscaroxana@yahoo.com (R. Tru șcă).
http://dx.doi.org/10.1016/j.microc.2015.12.0200026-265X/© 2015 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Microchemical Journal
journal homepage: www.elsevier.com/locate/microc

treatment —i.e., involving both the architecture and the ornamental
painting [7]. However, no systematic policy has yet been able to encom-
pass the entire territory on which such wooden churches are located.
Considered to be less protected monuments or actually derelict
monuments, the fresco wooden churches still remain insuf ficiently
investigated. Located exclusively in the southern part of the country,
they belong to the less privileged area of vernacular architecture andare on the brink of disappearance [8]. The present programme comes
to the rescue of such monuments. This programme is an in-depth casestudy of the technique, materials, and speci fic deterioration of the fresco
decoration in wooden churches to identify adequate materials and tech-nology for reconditioning the fresco plaster.
A representative case for the appearance and evolution of wooden
churches with fresco ornamentation in Romania is Saint Nicolas Church
of Ione știi Govorii. The church belongs to the ecclesiastical monuments
in the regions of Muntenia and Oltenia and was built by village commu-nities between the second half of the 18th century and the first half of
the 19th century [9]. The church is situated on the right bank of the
Olt River, near the road leading from Râmnicu Vâlcea —the capital
town of Vâlcea County —to the town of Dr ăgășani. According to the
1840 census, in 1770, the first founder of the church was Priest George
1
in addition to other community members [10]. Although the church was
originally made of wood following the plan of one of the characteristic
wooden ecclesiastical structures —i.e., rectangular plan with polygonal
altar, nave, narthex, and porch supported by carved columns [11] —
the Church of Ione știi Govorii later became a complex structure in the
category of churches partially “wrapped ”in masonry. According to the
inscription above the entrance to the church, during the first half of
the 19th century, in approximately 1836, Radu Sl ătineanu— the main
founder of the painting— and jupan2Stoica Sârbu together with other
founding members decided to envelop the wooden structure in brick-work. The operation succeeded only for the portico, the narthex, and
the western wall of the nave. The construction of a masonry iconostasis
was added. In approximately 1836, the new masonry surfaces were
covered in fresco, which constituted the first part of the iconographic
programme. A decade later, in approximately 1844, Radu Sl ătineanu
embarked on a mission to complete the iconographic programme
with the painting of the apse.
The object of this investigation is the second part of the iconographic
programme of the apse. Similar to the first part, this part was also un-
dertaken in fresco that was applied directly onto the oak beams and
planks that formed the original church structure.
First, the fresco painting technique had to cause the lime plaster to
adhere to the oak beams and planks —a technique also found in other
similar examples situated in the church proximity [7]. This adherence
was achieved using a cutting axe to roughen the wooden surface. At
the same time, according to the tradition of post-Byzantine painting,
the lime plaster ( intonaco ) was reinforced with fibres (probably hemp
tow). This non-carbonated, compact, and trowel-polished lime plasterwas covered with a painting layer using the Byzantine fresco technique
of hue overlaying and modelling. The pigments were those consecrated
by tradition as being compatible with lime plaster [12], i.e., yellow
ochre, red ochre, green earth, and cinnabar red.
The investigations described in this paper focused on identifying
the materials employed in applying frescoes on wood and possiblecompounds from their subsequent degradation and were conducted
using speci fic analytical techniques. Only a small number of studies
have been devoted to Romanian fresco paintings in wooden churches,and to the best of our knowledge, none was performed using multi-
technique investigations.2. Materials and methods
2.1. Sampling
The analysis in situ with the portable X-ray fluorescence analyser
(PXRF) was performed on thirteen areas ( Table 1 ) of the altar fresco.
The PXRF analyser identi fied the elements in the painting layer in addi-
tion to the elements in the lime plaster. Subsequently, samples were
required for further analytical investigations meant to provide evidence
of the elements in the lime plaster and the painting layer separately. All
samples were analysed by Fourier transform infrared spectroscopy
(FTIR), but a small number underwent analysis by optical microscopy
(OM), scanning electron microscopy coupled with energy dispersive
X-ray analysis (SEM –EDX) and X-ray diffraction (XRD). The wax drops
on the fresco were investigated by SEM.
2.2. Analysis methods2.2.1. In situ analysis of frescoes
Macroscopic observations of the frescoes were performed in situ with
the naked eye and under a magnifying glass in direct and grazing light.
The photographs were taken with a standard Nikon digital camera
(Coolpix s8000 model, Nikon Corp., Japan, China) with 10× magni fication
and high resolution (14.2 megapixels). The wax drops, the lime plaster,and the painting layer were also observed in situ with a Dino-Lite digital
microscope (AM7013MT-FV2W, AnMo E lectronics Co., Taiwan) with ap-
proximately 50× and 200× magni fication. A portable analyser, Innov –X
Alpha Series (Innov –X Systems, Inc., Woburnn, USA), based on energy
dispersive X-ray fluorescence technology, a tungsten anticathode,
Si–PIN detector, 35 kV acceleration voltage, 40 μA, and a 30-second
acquisition time was used for X-ray fluorescence analysis. The Innov –
XPC 1.53 software (Innov –X
Systems, Inc., USA) identi fied the elements
in the painting layer and the lime plaster.
2.2.2. X –ray diffraction (XRD)
X-ray diffraction analysis was used to identify the mineralogical
compounds and the breakdown compounds potentially found only in
the lime plaster. The XRD analysis was conducted using a Shimadzu
XRD 6000 Diffractometer (Shimadzu Corporation, Tokyo, Japan) with
Ni-filtered CuK αradiation ( λ= 1.5406 Å). Diffraction patterns were
taken at 0.02° intervals from 5° to 65°. The powder diffraction file
(PDF) was used for qualitative identi fication. The analysis of the lime
plaster was performed on samples ground to pass through a 90- μm
sieve.
2.2.3. Attenuated total re flectance Fourier transform infrared spectroscopy
(ATR–FTIR)
To obtain compositional data for the lime plaster and the painting
layer to indicate any possible degradation, a Bruker Alpha FTIR Spec-
trometer (Bruker GmbH, Ettlingen, Germany) was used in Attenuated
Total Re flection (ATR) mode —(4 cm−1resolution, 32 scans, and spectra
taken from 4000 to 360 cm−1). The Opus 4.2 (Bruker Optik GmbH,
Germany) software was used for data acquisition. The spectra processed
initially with Opus underwent statistical analysis as follows: smooth
(9 points), baseline correction (rubberband, 64 baseline points), and
normalisation (min –max). Multivariate data processing was performed
by means of Matlab code (an in-house code, Matlab version 7.12,R2011a, The MathWorks, Inc., USA). Prior to the proper data analysis,
the interval 2720 –1820 cm
−1was eliminated, and all of the spectra
were standardised by subtracting the mean and dividing by the stan-dard deviation.
2.2.4. Optical microscopy (OM)
Optical microscopy was conducted with a Carl Zeiss Axio Imager A1m
optical microscope (Carl Zeiss Microscopy GmbH, Munich, Germany)
with a 10× eyepiece, a 10×/0.20 objective lens, 100× magni fication,1Gheorghe in Romanian.
2From župan ǔ(Slavic), meaning ”owner of a stretch of land ”(Șeineanu, Rosetti,
Candrea cf. DER), in http://www.dexonline.ro .333 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

software, and a high-resolution 5 megapixels camera, AxioCam MRc 5
(Carl Zeiss Microscopy GmbH, Munic h, Germany). The analysis was con-
ducted on polished sections and thin sections under polarised light usingparallel nicols and crossed nicols. To cut and polish sections for analysis,
the samples were embedded in Buehler EpoThin resin —an epoxy resin
with refraction index 1.5702 (Buehler United Kingdom, Esslingen amNeckar, Germany). Thin sections were cut out from the embedded sam-
ples using an IsoMet Low Speed Saw cutting machine (Buehler United
Kingdom, Esslingen am Neckar, Germany). Polishing was performed
using a Phoenix Beta Grinder/Polisher (Buehler United Kingdom,
Esslingen am Neckar, Germany). Optical microscopy was effective in
obtaining information on the structure and texture of the mortars and
the type of mineral binder and aggregate [13].
2.2.5. Scanning electron microscopy coupled with energy dispersive X-rayanalysis (SEM-EDX)
Scanning electron microscopy coupled with energy dispersive X-ray
analysis was carried out to highlight the morphology and composition
of the lime plaster and the painting layer and to identify possible degra-
dation and biodegradation products. The analysis was performed using
an Inspect F Quanta analyser with 1.2 nm resolution (FEI-Philips,
Netherlands), equipped with an energy dispersive X-ray (EDX) spec-
trometer with a resolution of 133 eV at MnK. The analysis was
performed using high vacuum and an accelerating voltage of 30 kV.
The samples were coated with a thin layer of gold. The fresco samples
were analysed on cross-sections of the lime plaster and on the surface
of the painting layer. A JSPM-5200 scanning electron microscope
(JOEL, Japan) was used to analyse the wax drops to identify the types
of microorganisms and their arrangement. For image acquisition, the
samples were coated with gold. The SEM images were acquired under
high vacuum by employing an accelerating voltage of 25 kV.
3. Results and discussion
The fresco covering the polygonal apse walls and the lower area of
the vault in Ione știi Govorii church features a series of degradations.These degradations have appeared as a result of the shrinking and
swelling of the wooden beams and planks because of variations in
humidity and ambient temperature and casual movements and
vibrations of the foundation. Examination with the naked eye in direct
and grazing light has indicated that cracking, detachments, and gaps
(Fig. 1 ) are the primary types of degradation.
The fresco is cracked along the spaces between the beams and
planks both on the walls and on the lower area of the vault. In places
where shrinkage has led to cracking networks, the fresco has detached
from the beam and planks. In time, the cracks and detachments have
widened, leading to fresco gaps. The gaps vary in size from several
centimetres wide/long to approximately 30 cm wide and 70 cm long.
Wax drops, whitewashing repairs, graf fiti, and nails were also noticed
in addition to the degradations. The polychrome and secco decorationat the top of the vault features stains caused by moisture in filtration.
The macroscopic inspection of the altar painting has revealed
deposits of adherent wax, soot, and dust. Also visible are small insects,
living or dead, and cobwebs. The microscopic examination of the wax,
painting layer, and wooden support has revealed large mycelium
masses, which form when the water in the wooden substrate is avail-
able for a long time (more than 7 days), and smaller mycelium areas,
which appear when the water in the wooden substrate is available for
shorter intervals of time.
The PXRF technique allows elemental identi fication of chemical
elements with Z higher than 15. For this reason, PXRF can be usedonly for inorganic substances. At the same time, the pigments that
contain the same chemical elements but have another chemical compo-
sition cannot be identi fied with precision. To identify these elements,
other techniques must be employed to complement the data obtainedfrom the PXRF analysis [14]. However, the PXRF analysis has yielded
general information to identify the types of materials.
Thirteen areas of the painting were investigated to cover all of the
colours. Table 1 shows the results. The pigments have been identi fied
on the basis of the speci fic X-ray peak energies of the chemical elements
in the XRF spectra of the investigated areas. The elemental analysisshowed that the pigments are those commonly used at that time [15]:Table 1
The analysis and sampling areas and the results obtained by PXRF (from heavy element spectra) and ratio of intensities of Ca K and Sr K peaks.
Sample code Investigation and sampling area Elements CaK α/SrKα
1I SE wall, St. Nicholas —Red vestment Ca, Fe, Pb, Sr 11
2I SE wall, St. Nicholas —Red vestment Ca, Hg, Fe, Pb, Zn, Sr 10
3I SE wall, St. Nicholas —Green epitrachelion Ca, Fe, Pb, Cr, Ni, Sr 13
4I SE wall, St. Gregory the Theologian —Ochre phelonion Ca, Fe, Pb, Sr, Ag 20
5I SE wall, St. Nicholas —White scroll Ca, Fe, Zn, Sr 14
6I SE wall, St. Nicholas —Ochre right hand Ca, Fe, Hg, Pb, Zn, Sr 17
7I SE wall, St. Nicholas —Black sticharion vestment, right leg upward Ca, Fe, Pb, Sr, Ag 19
8I Southern wall, St. Spyridon —Grey background on the right side Ca, Fe, Sr 15
9I Prothesis (Jesus grapevine —background), red and green Ca, Fe, Pb, As, Sr 14
10I Prothesis (black background), red Ca, Fe, Hg, Pb, Sr 17
11I NE upper corner between St. Peter and St. John Chrysostom —Grey background Ca, Fe, Sr 22
12I NE lower corner between St. Basil the Great and St. Gregory the Theologian —Red background Ca, Fe, Pb, Sr 17
13I SE corner (between St. Nicholas and St. Spyridon) —Red background next to the scroll Ca, Fe, Pb, As, Zn, Sr 10
Fig. 1. Altar fresco degradations: cracking nets (a, b), cracks along the laths (a –c), gaps (a –c), detachments (a –c), whitewashing (c).334 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

pigments based on Fe, determined as red ochre and yellow ochre;
red cinnabar (HgS), which was determined solely on the basis of the
presence of Hg in the XRF spectrum because the characteristic S K
peak overlaps with the lines Hg M and Pb M; red lead (Pb 3O4), which
was identi fied on the basis of the presence of Pb in the orange areas
(its presence con firmed by FTIR analysis); an arsenic-based pigment
(unidenti fied; additional analytical methods are needed to ascertain
what arsenic pigment is present); a Fe-containing green pigment,
which was revealed by the ATR –FTIR analysis to be green earth; a Cr-
based pigment, which could not be determined accurately by PXRF;and a white Zn-based pigment ( Fig. 2 ). Traces of silver on the phelonion
and sticharion vestments (samples 4I and 7I) were identi fied, suggest-
ing a possible decoration with silver in certain areas ( Table 1 ).Table 1
shows that the ratio of calcium/strontium counts in all the paintedareas is between 10 and 22 indicating, according to the literature [16],
calcium carbonate (in our case, from lime binder and limestone aggre-gate) as the source of strontium.
The X-ray diffraction analysis performed only on the lime plaster
in samples 9I, 11I, 12I, 13I revealed the mineralogical compounds ofthe binder, namely, carbonated lime (calcite) and portlandite, and ofthe aggregate, namely, river sand (quartz, potassium feldspar, musco-vite, and plagioclase feldspar as albite and anorthite). Among these
compounds, the XRD analysis revealed calcite as the main component
in the lime plaster. This method revealed no degradation products,
i.e., gypsum or calcium oxalate. The diffractograms in Fig. 3 show that
there are no signi ficant differences between the analysed samples, so
the mortar employed in preparing the lime plaster of the fresco washomogeneous.
In terms of petrography, the lime plaster in samples 9I, 11I, 12I, and
13I consists of the binder and aggregate granules. Macroscopically,
there was marked reinforcement with cellulosic fibres of the plaster
(cellulosic material has been identi fied with FTIR), which are probably —
according to the traditional technique —hemp tow. The thickness of the
lime plaster varies from 2 –3m mt o1 5 –20 mm. On the surface of the
lime plaster, a thin layer of painting is distinguishable.
To obtain more detailed data concerning the structure and addi-
tional data on the mineralogy of the fresco, a pair of samples, namely,
12I and 13I, were analysed by optical microscopy on polished and thin
sections. The samples showed similar characteristics although they
had been taken from different areas: sample 12I from the NE corner
and sample 13I from the SE corner of the altar, indicating a uniform
composition of the fresco. The fresco sections that were analysed
show a resistant porous greyish-white mass. One can distinguish, both
macroscopically and microscopically, white relict lime clasts with sizes
from approximately 0.5 to 5 –7 mm. These relict lime clasts are free
from inclusions, siliceous, or other ( Fig. 4 a). In the literature, these relict
lime clasts are known as lime lumps [17–19]and are frequently found in
historical lime mortars. Based on the data in the literature, Pesce [18]
classi fies the lime lumps in relation to their composition. The lime
lumps may be fragments of underburnt or overburnt limestone [20],
fragments of burnt limestone containing large amounts of siliceous
compounds [17], and fragments of pure calcium carbonate issuing
from the carbonation of the lime paste [17,20] . In the mortar samples
12I and 13I that were investigated, the lime lumps are calcium carbon-
ate. We found that some of the lime lumps contain calcium hydroxide,
which is indicated in the X-ray diffraction diagrams. The presence of
portlandite was also identi fied
by ATR-FTIR analysis performed on
such lime lumps. A hypothesis for the presence of this type of lime
lumps, mentioned also by Bakolas in his paper [17], would be a slight
mixing of the lime putty with the mineral and vegetable aggregates.
Another likely hypothesis would be the use of a lower amount of
water in the mortar [17,19] . In the case of wooden churches, it may
Fig. 3. The diffractograms of altar samples 9I, 11I, 12I, and 13I.
Fig. 2. The PXRF spectra of samples 2I, 3I, and 10I from Ionesti, showing the presence of
mercury-, iron-, lead-, zinc-, and chromium-based pigments.335 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

well have been necessary to use a lower amount of water to obtain a
workability suitable for applying the mortar to the wooden structure.
The aggregate consists of angular and rounded, fine-grained, porous
limestone fragments with sizes ranging from 50 to 300 μm(Fig. 4 a);
rounded subangular quartz fragments with diameters of up to 150 μm
(Fig. 4 a); lamellar fragments of micaceous minerals with diameters of
up to 200 μm(Fig. 4 b), and subangular feldspar fragments with diame-
ters of up to 150 μm(Fig. 4 c).
The binder consists of fine-grained calcite. The mass of the binder
has preserved mineral skeletal parts of bioclastic organisms that appear
as small globular skeletal shapes ( Fig. 4 d) pertaining to foraminifera,
and also shell fragments of larger size ( Fig. 4 d). The presence of bioclasts
and shell fragments suggests that the limestone fragments might have
originated in the Jurassic –Cretaceous limestone deposits, which pertain
to the sedimentary succession in the Getic Blanket located in the
Romanian Southern Carpathians [21,22] . The literature indicates that
these deposits contain bioclasts, biomorphs, pellets, algae, and subordi-nately ooidic bioconstructions [21,22] . The lime plaster also presents
skeletal areas where sparitic calcite deposits may be noticed aroundmicritic calcite. The lime plaster is crossed by cracks, some of which
are already filled with sparitic calcite ( Fig. 4 e). Using optical microscopy,
we estimated that the binder is in a proportion of approximately 75 –
80%, the rest representing the aggregate.
Across the fresco surface, at the upper edge of the section in Fig. 4 f,
the painting thickness presents wide fluctuations, from approximately5μm to approximately 120 μm. The painting layer is well attached
to the lime plaster as a result of binding the pigment in the process of
carbonation of the calcium hydroxide in the lime plaster. The process
is speci fic to the fresco technique [23,24] .
Additional information on the structure of the fresco samples was
obtained by SEM analysis. For this analysis, four samples (9I, 10I, 12I,
and 13I) were selected and were analysed both in cross-sections of
the lime plaster and at the surface of the painting layer. SEM analysis
performed on the cross-sections of the lime plaster revealed a porous
structure with micro-cracks ( Fig. 5 a and c). The sections allow visualisa-
tion of the fibres ( Fig. 5 a) and the thickness of the painting layer ( Fig. 5 a
and b). SEM images in Fig. 5 b and c reveal a microgranular morphology
of the calcitic binder, which validates the observations carried out byoptical microscopy, namely, that the binder is a fine-grained calcite.
The investigations by scanning electron microscopy and energy-
dispersive X-ray spectroscopy indicate that in addition to calcite, thelime plaster includes quartz and silicate minerals. Fig. 6 shows a picture
of
a lamellar fragment of mica with a thickness of approximately 23 μm.
Investigations have revealed deposits of sparitic calcite on the walls of
the pores and cracks in the calcite mass, which have also been indicated
by optical microscopy analysis.
The samples taken from the lime plaster showed only a trace of
sulphur; however, all of the analysed samples from the painting
layer have revealed signi ficant levels of sulphur. Thus, on the strati-
graphic section of sample 12I that includes painting areas, sulphur
Fig. 5. SEM images in cross-sections: overview of samples 9I (a) and 13I (b); detail of the micro-crystalline calcite mass and shrinkage micro-cracks in sample 9I (c).
Fig. 4. Optical microscopy images: sample 13I —Polished section, polarised light (a), thin section, polarised light, cross nicol (b) and parallel nicol (c); sample 12I —Polished section,
polarised light (d, e), and direct light (f) 1 —Relict lime clasts (lime lumps); 2 —porous, fine-grained limestone fragments; 3 —quartz; 4 —mica minerals; 5 —feldspar; 6 —mineral parts of
the bioclastic organism pertaining to foraminifera; 7 —shell fragments; 8 —painting layer.336 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

was identi fied on the super ficial layer ( Fig. 7 ), which is an indication
that the painting presents deposits of gypsum (as a degradation
product). This degradation has also been con firmed by the result of
the ATR –FTIR analysis. The presence of gypsum is highly likely due to
the action of sulphur dioxide present in the air on the calcium carbonate
in the lime plaster. This degradation process has been discussed in
many papers in the literature [25–28]. Sulphur dioxide in the air may
come from burning coal and wood in stoves for house heating, whichis likely, given the location of the church in a rural area. Sulphur dioxidemay also come from burning lignite in the nearby (approximately
20 km) thermal power station, which provides heat and electricity to
the urban area.
The elements found by analysing sections of the painting layer and
endorsed by data obtained from other analytical techniques (PXRF,
ATR-FTIR) may suggest the presence of a natural earth based on iron
oxides, kaolin, silicates, potassium feldspar, and Ti impurities. Thepresence of titanium in the painting layer with Al, Si, K, and Ca may be
an indication that the pigment is natural [29]because titanium is easily
carried by minerals in the form of rutile, TiO
2[30].
On cross-sections, close to the cellulosic fibres, hyphae may be
noticed, which demonstrates that the hyphae belong to the species offungi producing cellulase ( Fig. 8 a). The SEM analyses also revealed
branch-split hyphae sparsely distributed on the surface of the paintinglayer because of settling of organic compounds and environmental
conditions ( Fig. 8 b). An extensive distributi on of mycelium and fruiting
bodies that could be analysed using the FTIR technique was not identi fied.
The small and large wax drops are covered by organic deposits and
microorganisms involved in decomposition. The SEM images showedhyphae nets and developing mycelia involved in breaking down organic
deposits, wax included ( Fig. 9 a). In the course of time, the wax has bro-
ken down, and the pressure subsequently acting on the hyphae has ledto their fragmentation ( Fig. 9 b). The hyphae grow at the surface and
below the surface of wax ( Fig. 9 c). When environmental conditions
are favourable, the mycelium grows, and spore germination takes
place ( Fig. 9 d).
The ATR –FTIR analysis performed on the painting layer and the lime
plaster (multiple analyses on each sample) demonstrated the presence
of calcium carbonate through the bands at approximately 1412, 875,
and 712 cm
−1and gypsum through the bands at 3550, 3400, 1621,
Fig. 7. SEM image and EDX maps of the elemental distribution in sample 12I with red painting 1 —Thin gypsum layer on the painting layer; 2 —painting layer; 3 —the lime plaster/painting
layer interface; 4 —lime plaster.
Fig. 6. The SEM image and the elemental analysis spectrum of a cross-section in sample
10I, including a lamellar fragment of mica.337 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

1112, 672, and 600 cm−1, with the latter being most likely a compound
resulting from degradation. In a number of the spectra that we exam-
ined, the presence of the calcium oxalate monohydrate, whewellite
(bands at 1620, 1321, and 782 cm−1) was also observed in the pigment
samples. Calcium oxalate may be the result of the interaction of the limeplaster with the environment and/or an organic binder. The presence
of these compounds can lead not only to the degradation of the lime
plaster but also to the degradation of the paint layer. Changes occurring
to the painting and caused by the presence of sulphate and oxalate have
been reported in the case of historical monuments and works of art,
where the appearance of the painting has changed visibly [31–33].
The IR analysis of the painting samples also revealed the presence ofkaolin (3689, 3620, 1028, 1006, 913, 779, 532, 468, 426 cm
−1) for the
dark red and yellow samples and red lead (528, 428 cm−1) for the
orange-red samples. With IR spectroscopy, the spectra of the lime plas-
ter have also revealed other components such as quartz (1080, 798, 780,
and 460 cm−1), feldspar (broad bands between 1200 and 900 cm−1,
800 and 700 cm−1), in addition to calcium carbonate. Portlandite(Ca(OH) 2, with band at 3642 cm−1), was identi fied in some of the
lime lumps from the lime plaster. As reinforcement fibres for the lime
plaster, cellulosic (plant) fibres have been identi fied by FTIR and by
optical microscopy (Fig. S1 —Supplementary Materials). The spectral
features (bands at 1202, 1157, 1104, 1051, 1024, and 984 cm−1)a r e
ascribable, according to the literature [34] to cellulose.
In the interpretation of the FTIR data, a chemometric data analysis
method was preferred, i.e., the principal component analysis (PCA), todown-scale the data volume and to convert the original set of variables
into a new group of variables, hereinafter referred to as main compo-
nents, to classify and identify the painting materials, and to highlight
the presence of gypsum in the paint layer. PCA is perhaps one of
the oldest and most commonly used methods of multivariate data anal-
ysis, used both in pigment characterisation [35,36] and in the character-
isation of the historical mortars [37–39]. As seen from the PC1 –PC2
score plot ( Fig. 10 a), all samples including pigment that were analysed
were found to have positive values on the axis of the first principal com-
ponent, PC1, indicating a signi ficant contribution of gypsum to the IR
Fig. 9. SEM images of a drop of wax colonised by fungi: hyphae net (a); detail showing fragmented mycelia following the cracks of the wax (b); branch-split hyphae (c); detail showing
germination of a fungal spore (d).
Fig. 8. SEM image of sample 9I: The microcrystalline calcite mass within the lime plaster, showing the cellulosic fibres and mycelium (a) and the painting layer showing the presence of the
mycelium (b).338 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

spectra ( Fig. 10 b). PC1 described 75% of the variance. The second princi-
pal component (PC2) described 11% of the variance, and its loading
spectrum again included gypsum and calcium carbonate, with negative
bands ( Fig. 10 c). The PC2 loading spectrum also included an intense
positive band of green earth. The principal component analysis andthe cluster analysis (agglomerative clustering) have shown that the
samples containing pigments (grouped in cluster number 2 and in all
other clusters situated on the right part of the PC1 axis) included gyp-
sum, suggesting that gypsum is present as ef florescence. In contrast,
the samples in cluster number 1 including samples from the lime plasterare high in calcium carbonate ( Fig. 10 d, e). The calcium carbonate
contained in samples of pigment from cluster 1 is either from the limemortar or it is a component of the painting layer. The samples out of
clusters 1 and 2 do not give any additional information regardingpigments. They show only different features due to different ratios
between components (gypsum, calcium oxalate, calcium carbonate,
green earth, etc.). The spectrum of a green sample with green earth
(with bands at 3554, 968, 798, 494, 440 cm
−1suggesting the presence
of celadonite [40,41] )i ss h o w ni n Fig. 11 .4. Conclusions
The information obtained and presented in this paper represents a
first and useful step in developing a methodology for the conservation
of frescoes applied to a special type of historical monuments, namely,
wooden churches painted in fresco on wood. The case study was the
Ione știi Govorii wooden church, Vâlcea County, Romania. The use of
analytical techniques allowed the characterisation of the compositionand structure, and visual, macro- and microscopic observation allowed
the assessment of the state of conservation of the frescoes.
The fresco applied to oak beams and planks is similar to the fresco
applied to masonry. The binder of the plaster was lime putty, and
the aggregates were fine river sand and finely crushed limestone.
Reinforcement cellulosic (plant) fibres were added to increase the
cohesion between the mortar components. The lime plaster was foundto contain lime lumps, also including portlandite. The presence of
bioclasts and shell fragments identi fied in the aggregate suggests the
local origin of the latter. The results obtained from the multiple investi-gation techniques are evidence for the use of a rather limited range of
pigments, namely, red and yellow earth, cinnabar, red lead, green
earth, green chromium-based pigment, and white zinc-based pigment,
all of which are speci fic to that time and were frequently used in brick
wall frescoes.
The main degradation consists of cracks and crack nets, massive
detachment of the wooden beams and planks, and gaps of varyingsizes, all caused in particular by the incompatibility between the charac-
teristics of the mineral matter in the fresco and the organic material
(wood planks and beams). In terms of the chemical degradation, partic-
ularly at the surface of the painting layer of the fresco, gypsum and
calcium oxalate were identi fied as products of degradation.
The biodeterioration process was elucidated, and the biodeteriogens
were identi fied as pertaining to microscopic fungi and as present both
on local wax deposits and on the wooden substrate and the surface ofthe painting layer. The fungi formed on the cellulosic (plant) fibres led
to the hydrolysis of the cellulose component which, in turn, has led tolower cohesion at the substrate level. The fungi colonisation process is
not continuous because the relative humidity also varies depending
on rainwater in filtration.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.microc.2015.12.020 .
Fig. 11. Spectrum of a green sample (10I) and reference spectra (from our database) of
gypsum (G), calcium carbonate (C), calcium oxalate with calcite impurity (O), and green
earth (from IRUG database, www.irug.org IMP00311) containing celadonite, glauconite,some calcite, terre-verte (E).
Fig. 10. PC1 (75%) —PC2 (11%) score plot of FTIR data —lime plaster samples are marked with coloured dots and pigment samples are marked with coloured dots and “x”(a); loadings plot of
thefirst two principal components (b); PC1 reveals mainly gypsum in the positive part and calcium carbonate in the negative part of the first principal component whereas PC2 groups
samples with both gypsum and calcium carbonate and samples with green earth (c); comparison between spectra from cluster 1 that contain mainly calcium carbonate (d) and cluster 2that corresponds to the samples that contain calcium carbonate and gypsum (e) (legend: G —gypsum, C —calcium carbonate, E —green earth).339 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340

Acknowledgements
This work was carried out by Partnerships in Priority Areas —PN
II—developed with the support of MEN –UEFISCDI, Romania, Project
No. PN– II–PT–PCCA –2013 –4–1311.
References
[1] I. Godea, Biserici de lemn din Europa, CD Press, Bucure ști, 2008
[2]http://whc.unesco.org/en/list/904
[3] A. Bârc ă, Biserici de lemn din S ălaj, NOI Media Print, Bucure ști, 2010
[4]A. Bucur, The wooden church The Nativity of St. John the Baptist from Gârda de Sus,
Alba County –degradation and conservation, An Extreme Case, in Restitutio,
Conservation Restoration Bulletin, 9, Dimitrie Gusti National Village Museum,Bucharest, 2015 (ISSN 2065-2992).
[5]E. Rogoz, O.H. Barbu, Identi fication of protein content in the gypsum painting
support by FTIR spectroscopy, PCA and neural network, in: A. Macchia, E. Greco,
B.A. Chiarandà, N. Barbabietola (Eds.), Proceedings YOCOCU Contribute and Role
of Youth in Conservation of Cultural Heritage, Italian Association of ConservationScientist, Rome 2011, pp. 329 –336.
[6]M.F. Mecklenburg, C.S. Tumosa, D. Erhardt, Structural response of painted woodsurfaces to changes in ambient relative humidity, Painted Wood: History and
Conservation (Part 6: Scienti fic Research), The Getty Conservation Institute, ISBN:
0-89236-501-3 1998, pp. 464 –483.
[7]D. Mohanu, I. Gomoiu, A. Chiricu ță, The Saving of the Wooden Church of Ur șii na
New Posture: The Inter-Disciplinary Research, Planning and Conservation, Caiete
ARA, 42013 249 –269.
[8] The 60 wooden churches, Funda ția PRO PATRIMONIO România,
http://www.
propatrimonio.org/en/60-wooden-churches , (the last accessed date 29.11.2015)
[9]N. Stoicescu, Bibliogra fial o c a l i t ățilorși monumentelor feudale din România, I –Țara
Româneasc ă, vol. 11970 382.
[10] C. B ălan, Inscrip ții medievale și din epoca modern ăa României, Jude țul istoric
Vâlcea, Ed. Academiei Române, Bucharest, 2005, p. 609 –610.
[11] A. P ănoiu, Din arhitectura lemnului în România, Ed. Tehnic ă,B u c u r e ști, 1977, p. 113.
[12] Dionisie din Furna, Erminia picturii bizantine, Ed. So fia, Bucure ști, 2000, p.54 –58.
[13] B. Middendorf, J.J. Hughes, K. Callebaut, G. Baronoi, I. Papayanni, Investigativemethods for the characterization of historic mortars –Part 1: Mineralogical charac-
terization, RILEM TC 167-COM: “Characterization of Old Mortars with Respect to
their Repair, Mater. Struct. 38 (8) (2005) 761 –769.
[14] O.H. Barbu, M. Georgescu, Image processing and X-Ray fluorescence analysis for pig-
ment identi fication, in CMA4CH Employ of multivariate analysis and chemometrics
in cultural heritage and environment fields, in: Marcovalerio (Ed.), Proceedings
Book 2014, pp. 28 –29.
[15] N. Easthaugh, V. Walsh, T. Chaplin, R. Siddall, The Pigment Compendium: A Dictio-
nary of Historical Pigments, Elsevier Butterworth-Heinemann, 2007.
[16] E. Franceschi, F. Locardi, Strontium, a new marker of the origin of gypsum in culturalheritage? J. Cult. Herit. 15 (5) (2013) 522 –527.
[17] A. Bakolas, G. Moropoulou, E. Zendri, Characterization of the lumps in the mortars
of historic masonry, Thermochim. Acta 269/270 (1995) 809 –816, http://dx.doi.org/
10.1016/0040-6031(95)02573-1 .
[18] G.L.A. Pesce, R.J. Ball, Dating of old lime based mixtures with the “pure lime lumps ”
technique, in: D. Michalska-Nawrocka (Ed.), Radiometric Dating, Publisher InTech,
ISBN: 978-953-51-0596-1 2012, pp. 21 –38,http://dx.doi.org/10.5772/35419 .
[19] G. Borsoi, A. Santos Silva, P. Menezes, A. Candeias, J. Mirão, Chemical, mineralogical
and microstructural characterization of historical mortars from the Roman villa ofPisões, Beja, Portugal, in: J. Válek, C. Groot and J.J. Hughes (Eds.), RILEM Publications
SARL, 2nd Conference on Historic Mortars – HMC 2010 and RILEM TC 203-RHM final
workshop, e-ISBN: 978-2-35158-112-4, 2010, pp. 43-54
[20] J. Elsen, Microscopy of historic mortars –a
review, Cem. Concr. Res. 36 (2006)
1416 –1424.
[21] V. Mutihac, M.I. Stratulat, R.M. Fechet, Geologia României, Ed. Didactica si
Pedagogica, Bucharest, 2004, pp. 169 -176.
[22] O. Dragastan, Getic carbonate platform –Platforma Carbonatica Getica, Stratigra fia
Jurasicului și Cretacicului inferior, Reconstituiri, Paleogeogra fie, Provincii și
Biodiversitate, Editura Universit ății din Bucure ști, 2010 622.[23] A. Hein, I. Karatasios, D. Mourelatos, Byzantine wall painting from Mani (Greece):
microanalytical investigation of pigment and plasters, Anal. Bioanal. Chem. 395(7) (2009) 2061 –2071, http://dx.doi.org/10.1007/s00216-009-2967-6 .
[24] R. Piovesan, C. Mazzoli, L. Maritan, P. Cornale, Fresco and lime-paint: an experimen-
tal study and objective criteria for distinguishing between these painting techniques,
Archaeometry 54 (4) (2012) 723 –736, http://dx.doi.org/10.1111/j.1475-4754.2011.
00647.x .
[25] A. Izaguirre, J. Lanas, J.I. Alvarez, Ageing of lime mortars with admixture: durability
and strength assessment, Cem. Concr. Res. 40 (7) (2010) 1081 –1095, http://dx.doi.org/
10.1016/j.cemconres.2010.02.013 .
[26] G. Cultrone, A. Arizzi, E. Sebastian, C. Rodriguez-Navarro, Sulfatation of calcitic and
dolomitic lime mortars in the presence of diesel particulate matter, Environ. Geol.56 (2008) 741 –752, http://dx.doi.org/10.1007/s00254-008-1379-9 .
[27] S. Martinez-Raminez, G.E. Thompson, Degradation of lime-pozzolan mortar exposed
to dry deposition of SO2 pollutant gas: in fluence of curing temperature, Mater.
Struct. 32 (5) (1999) 377 –382.
[28] C. Giavarini, M.L. Santarelli, R. Natalini, F. Freddi, A non-linear model of sulphatation
of porous stones: numerical simulations and preliminary laboratory assessments,
J. Cult. Herit. 9 (2008) 14 –22,http://dx.doi.org/10.1016/j.culber.2007.12.001 .
[29] M. Gil, M.L. Carvalho, A. Seruya, A.E. Candeias, J. Mirão, I. Queralt, Yellow and red
ochre pigments from southern Portugal: elemental composition and characteriza-
tion by WDXRF and XRD, Nucl. Instrum. Methods Phys. Res. Sect. A 580 (2007)
728 –731, http://dx.doi.org/10.1016/j.nima.2007.05.131 .
[30] M. Gil, M.L. Carvalho, A. Seruya, I. Ribeiro, I. Queralt, A.E. Candeias, et al.,
Limewashing paintings in Alentejo urban heritage: pigment characterization anddifferentiation by WDXRF and XRD, Appl. Phys. A Mater. Sci. Process. 90 (2007)
49–54,http://dx.doi.org/10.1007/s00339-007-4252-x .
[31] O.H. Barbu, D. Mohanu, Studiul transform ării pigmentului miniu în sulfat de plumb,
Stabilirea tratamentului optim de conservare-restaurare, in Colocviul Na țional
de Istorie, Istoria Artei Decorative, Conservare-Restaurare șiR e l a ții Publice, Muzeul
Național Cotroceni, Bucure ști 2009, pp. 113 –12
0.
[32] J.L. Perez-Rodriguez, M.D. Robador, M.A. Centeno, B. Siguenza, A. Duran, Wall paint-
ings studied using Raman spectroscopy: A comparative study between various as-says of cross sections and external layers, Spectrochim. Acta Part A Mol. Biomol.
Spectrosc. 120 (2014) 602 –609, http://dx.doi.org/10.1016/j.saa.2013.10.052 .
[33] S. Mugnaini, A. Bagnoli, P. Bensi, F. Droghini, A. Scala, G. Guasparri, Thirteenth
century wall paintings under the Siena Cathedral (Italy). Mineralogical and petro-graphic study of materials, painting techniques and state of conservation, J. Cult.
Herit. 7 (2006) 171 –185, http://dx.doi.org/10.1016/j.culher.2006.04.002 .
[34] P. Garside, P. Wyeth, Identi fication of cellulosic fibres by FTIR spectroscopy: differen-
tiation of flax and hemp by polarized ATR FTIR, Stud. Conserv. 51 (3) (2006)
205 –211.
[35] F. Rosi, A. Burnstock, K.J. Van den Berg, C. Miliani, B.G. Brunetti, A. Sgamellotti, A
non-invasive XRF study supported by multivariate statistical analysis and re flec-
tance FTIR to assess the composition of modern painting materials, Spectrochim.Acta A Mol. Biomol. Spectrosc. 71 (2009) 1655 –1662, http://dx.doi.org/10.1016/j.
saa.2008.06.011 .
[36] J. Romero-Pastor, C. Cardell, Á. Yebra-Rodríguez, A.B. Rodríguez-Navarro, Validating
chemical and structural changes in painting materials by principal component anal-
ysis of spectroscopic data using internal mineral standards, J. Cult. Herit. 14 (2013)
509 –514, http://dx.doi.org/10.1016/j.culher.2012.11.006 .
[37] L. Rampazzi, A. Pozzi, A. Sansonetti, L. Toniolo, B. Giussani, A chemometric approach
to the characterisation of historical mortars, Cem. Concr. Res. 36 (2006) 1108 –1114,
http://dx.doi.org/10.1016/j.cemconres.2006.02.002 .
[38] A. Moropoulou, K. Polikreti, Principal component analysis in monument conservation:
three application examples, J. Cult. Herit. 10 (2009) 73 –81,http://dx.doi.org/10.1016/
j.culher.2008.03.007 .
[39] I. Ozga, N. Ghedini, C. Giosuè, C. Sabbioni, F. Tittarelli, A. Bonazza, Assessment of air
pollutant sources in the deposit on monuments by multivariate analysis, Sci. Total
Environ. 490 (2014) 776 –784, http://dx.doi.org/10.1016/j.scitotenv.2014.05.084 .
[40] L.M. Moretto, E.F. Orsega, G.A. Mazzocchin, Spectroscopic methods for the analysisof celadonite and glauconite in Roman green wall paintings, J. Cult. Herit. 12 (4)
(2011) 384 –391.
[41] F. Ospitali, D. Bersani, G. Di Lonardo, P.P. Lottici, ‘Green earths ’: vibrational and ele-
mental characterization of glauconites, celadonites and historical pigments,
J. Raman Spectrosc. 39 (2008) 1066 –1073.340 I. Mohanu et al. / Microchemical Journal 126 (2016) 332 –340