A

rieIfhrtwaat©

K

1

tfotarotta

1d

Journal of Cultural Heritage 10 (2009) 379–387

Original article

The scientific approach to the restoration and monitoring of muralpaintings at S. Girolamo Chapel – SS. Annunziata Church in Florence

Sara Bianchin a, Monica Favaro a, Pietro Alessandro Vigato a,∗,Guido Botticelli b, Gioia Germani b, Silvia Botticelli b

a Istituto di Chimica Inorganica e delle Superfici, CNR, Corso Stati Uniti 4, 35127 Padova, Italyb Università Internazionale dell’Arte (UIA), Firenze, Italy

Received 30 January 2008; accepted 5 November 2008

bstract

The scientific approach to the restoration and monitoring of mural paintings, at S. Girolamo Chapel – SS. Annunziata Church in Florence, iseported as a fruitful example of the synergic collaboration between restorers and scientists in the planning and development of conservativenterventions. Before restoration, the painting technique and the state of preservation of the pictorial cycle have been investigated firstly by closexamination of the painted surfaces and then by optical and/or SEM-EDS microscopy and �FT-IR spectroscopy on appropriately selected samples.n particular, the original constituent materials and those belonging to subsequent restorations were characterized together with those originatingrom decay processes. In this mural painting cycle, a peculiar “fresco” technique has been used although an auxiliary binder for pigment distemperas been also employed. The use of this technique, sometimes in an improper way together with uncorrected restoration interventions, are the mainesponsible of the unsatisfactory state of conservation of many painted areas. Preliminary conservation trials and scientific studies were carried outo design the most convenient restoration intervention and to verify the correctness and non-invasive of the necessary operations. Special attentionas devoted to cleaning procedures, continuously monitored by a physico-chemical methodology, mainly based on microinvasive, microscopic

nd spectroscopic investigations, to evaluate the efficiency, advantages and drawbacks of the proposed cleaning procedures and define the mostppropriate ones. Finally, specific decay markers have been recognized, by a comparison of the results obtained from the detached samples withhose deriving from artificially aged models, to be used for a correct future monitoring and maintenance of the wall paintings.

2009 Elsevier Masson SAS. All rights reserved.

ests;

trfirtd

2

eywords: Wall painting; Pictorial technique; State of conservation; Cleaning t

. Research aim

The painted surfaces were affected by a severe darkeninghat compromises the satisfactory overview, comprehension andruition of the paintings inside this chapel. The consequent needf a highly requested intervention gave to the technical and scien-ific equipe, that is, restorers and scientists, the chance to proposend verify a scientific approach to address correctly the wholeestoration procedures and to propose a convincing methodologyf general applicability for mural-painting monitoring and main-

enance. Starting from the knowledge of the pictorial technique,he state of conservation, the identification of past repaintingsnd restorations, it was possible to guide the intervention scien-

∗ Corresponding author.E-mail address: vigato@icis.cnr.it (P.A. Vigato).

arbs

o

296-2074/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved.oi:10.1016/j.culher.2008.11.002

Restoration procedures; Monitoring

ifically, solving the raised problems step by step. A remarkableesult of these joint efforts is the possibility to see and bene-t again from frescoes, in their true chromatism, thanks to theemoval of unaesthetic deposits and materials of past interven-ions, making it possible to perceive the intimate meaning of theifferent scenes.

. Introduction

Restorations of the mural paintings have been carried outt San Girolamo Chapel – SS. Annunziata Church in Flo-ence, Italy. This pictorial cycle, executed by Alessandro Allori

etween 1560 and 1564, is the first relevant work after his studytay in Rome [1].

The frescoes represent The Expulsion from the Gardenf Eden on the vault (Fig. 1), Prophets and Sybils on the

380 S. Bianchin et al. / Journal of Cultural Heritage 10 (2009) 379–387

the va

pPtoTcpBBMVotciFHiprAo

oafc[

tce

ae

3

ctaPfctoOtos4fiaev

tetu

Fig. 1. The paintings on

endentives, Christ’s Histories, that is, Annunciation, Nativity,resentation of Christ to the Temple and Flight into Egypt, on

he sides, The Disputation with the Doctors and the Expulsionf merchants from the Temple on the lateral walls. In particular,he Disputation contains a gallery of portraits of famousitizens, men of letters, doctors and Allori’s friends (i.e., theainters Maso da San Friano, Giovanni Maria Butteri, Agnoloronzino and Jacopo Da Pontormo, the sculpturers Baccioaldini and Bartolomeo Ammannati, the doctors Alessandroenchi and Gherardo Quadri, the politicians and scholars Pier

ettori and Vincenzo Borghini) identifiable by the inscriptionsn the clothes. The Four Evangelists are painted on the sides ofhe altar, while the altar-piece represents the Last Judgement, alear tribute to Michelangelo and his frescoes at Sistina Chapeln Rome, as attested by the inscription “Alexander Allorius civislor. Bronzini Alumnus Inventum Octimi Pictoris Buonarrotaeaec Seculo Pinxit” and by the presence of Buonarroti’s portrait

n the low figure on the left. Fortunately, Allori’s interventionreserved The vision of Saint Jerome by Andrea del Castagno,ediscovered in 1899 in consequence of the detachment ofllori’s Judgement during the conservation work after the floodf Florence in 1966 [2].

Among the others, two main interventions have been carriedut on the paintings: the first operated in 1799 by Santi Pacinis commemorated by a marble slab inserted on the base of theresco that included extensive repaintings, the second one exe-uted during the thirties of the last century by Amedeo Benini3].

Integrated physico-chemical measurements were carried outo describe, at molecular, nano- or micrometric scale, the matericomplexity of the frescoes and the decay processes of the differ-nt materials, addressed to characterize the pictorial technique

Eiflr

ult after the restoration.

nd the conservation state of the paintings and to offer the nec-ssary scientific tools for the restoration.

. Experimental

Before any conservation intervention, the whole paintingycle was carefully and jointly observed by restorers and scien-ists to detect and macroscopically define the more appropriatereas to be investigated by physico-chemical methodologies.ainting fragments and one powder sample were collectedrom the selected areas and analyzed as detached and asross-sections; their macroscopic and microscopic descrip-ion and properties are reported in Table 1. The opticalbservations on the cross-section were carried out by anlympus BX51 microscope, equipped with UV-light source

o identify the stratigraphic sequences and the presence ofrganic material within the layers. The cross-sections wereubsequently investigated by SEM-EDS (Philips Model XL0 LaB6), after metallization through the deposition of a graphitelm. Semi-quantitative elemental analyses were obtained byn EDAX-EDS, Prime X-ray energy dispersive spectrometerquipped with a thin beryllium window, using an acceleratingoltage of 25 KeV [4].

By stereomicroscope observation, the different layers ofhe collected samples have been mechanically separated fromach other and the resulting materials studied by �FT-IR spec-roscopy, carried out either on the crushed layers and on their sol-ble fractions, extracted by appropriate solvents (CH2Cl2, H2O,

tOH, CH3COCH3) to identify the binding media of the paint-

ng layers. The separated layers and the extracted fractions wereattened on a gold surface and their IR spectra were collected ineflection mode by a Nicolet Magna IR560 FT-IR Spectrometer

S. Bianchin et al. / Journal of Cultural Heritage 10 (2009) 379–387 381

Table 1Sample location, identified materials and painting details are reported for the fragments collected from the painted surfaces and analytically characterized.

Label Sample location Identified materials

SA1 04 Disputation with the Doctors/red veil of the personage at the centreof the scene on the left wall of the altar

Pigments: red earth, smalt blue, minium/organic materials: animalglue, waxes/other: gypsum on the surface

SA2 04 Disputation with the Doctors/clear blue veil of the personage nearthe stairs

Pigments: smalt blue, carbon black, green earth, iron oxides/organicmaterials: animal glue, waxes/other: Ca oxalates

SA4 04 Evangelist/yellow veil Pigments: yellow ochre, Pb–Sn yellow, green and red earths/organicmaterials, waxes/other: candle soot deposit

SA5 04 Evangelist/green veil Pigments: green and red earths, realgar, minium/organic materials:animal glue, waxes/other: neoformation gypsum on the painting layers

SA7 04 Eritrean Sybil/green and red drapery Pigments: green and red earth earths/organic materials: animal glue,waxes/other: Ca-oxalates, neoformation gypsum on the plaster andpainting layers

SA8 04 Saint Geremia/red drapery, repainting near the filling Pigments: red earth, quartz/organic materials: animal glue,waxes/other: Ca-oxalates, neoformation gypsum on the plaster andpainting layers

SA9 04 Saint Geremia/red drapery, original painting, next to SA8 04 Pigments: red earth/organic materials: animal glue, waxes/other:Ca-oxalates, neoformation gypsum on the painting layers

SA10 04 Eritrean Sybil/clear green-bluish drapery, below the book Pigments: green earth/organic materials: animal glue, waxes/other:Ca-oxalates

SA11 04 The expulsion from Eden/green meadow, near Adamo’s arm Pigments: green earth, yellow ochre/organic materials: animal glue,waxes/other: Ca-oxalates, neoformation gypsum

SA12 04 Expulsion of the Merchants from the Temple/superficial, brown,incoherent layer

Siccative oil

SA1 05 Frontal subarch/brown decoration Lime priming/pigments: green and red earths/organic materials:animal glue, waxes/other: Ca-oxalates

SA3 05 Evangelist/yellow drapery Pigments: red and green earths/organic materials: animal glue,waxes/other: Ca-oxalates, gypsum

SA4 05 Evangelist/ next to brown background cleaned with water Pigments: green earth, carbon black/organic materials: animal glue,waxes/other: Ca-oxalates

SA5 05 Evangelist/ next to brown background uncleaned, next to SA4 05. Pigments: green and red earth, carbon black/organic materials: animalglue, waxes

SA1 06 Frontal subarch/green decoration Pigments: green and red earths/other: Ca-oxalatesSA2 06 Disputation with the Doctors/blue background Pigments: smalt blue, realgar/organic materials: waxesSA4 06 Expulsion of the Merchants from the Temple/yellow veil of a

personagePigments: yellow ochre/organic materials: waxes

SA5 06 Disputation with the Doctors/greyish background, uncleaned Pigments: carbon black, smalt blue/organic materials: waxes/other:neoformation gypsum on the plaster and on the painting layers

SA6 06 Disputation with the Doctors/greyish background, cleaned withwater, next to SA5 06.

Pigments: carbon black, smalt blue/organic materials: waxes/other:neoformation gypsum on the plaster and on the painting layers

SA7 06 Disputation with the Doctors/greyish background, cleaned with Pigments: carbon black, smalt blue

A erent

at

abltt

4

4

gaw

qvlat

mjdr“t

(NH4)2CO3, next to SA5 06

lime plaster with silicate-carbonate aggregates has been identified on the diff

nd Spectra Teck microscope. The IRUG database was used forhe identification of the material constituents of the samples [5].

Distilled water and aqueous solutions of ammonium carbon-te were used as cleaning agents while aqueous solutions ofarium hydroxide were used as consolidation products. Cel-ulose pulp, sometime mixed with sepiolite to enhance thehixotropic effects, was employed as solution support either forhe cleaning and the consolidation procedures [6,7].

. Results and discussion

.1. The painting technique

The structure of the pictorial cycle clearly refers to Michelan-elo’s frescoes in Sistina Chapel, both in the space divisionnd in the figurative planning and style. Quite probably, Allorias helped by coworkers, as testified by the non-homogeneous

aiib

scenes of the painting cycle.

uality of the paintings, proved by scientific investigations anderified during restoration. Although the general planning fol-ows the fundamental rules of fresco, there are several executionnomalies, hardly ascribable to Allori’s inexperience and mis-akes.

The painting substrate was made by plaster of small oredium dimensions; that is, the giornate, which are quite well

oined with each other and makes their identification and/oristinction difficult sometimes. The preparatory drawing waseproduced on the plaster by the “cartone” method. The use ofspolvero” was proved only on the repeating decoration aroundhe oval window.

The finishing-lime stratum, directly observed on an area

ffected by painting detachments, does not contain any binder,ndicating that it was possibly obtained by a mechanical smooth-ng/flattening of the wet plaster to make more homogeneous andright the painting.

382 S. Bianchin et al. / Journal of Cultural Heritage 10 (2009) 379–387

F detera r; (b)t hs.

islecaiwwpptuwtarp

Fcsb(

absiiocact

dmb

ig. 2. EDS spectra collected on blue painting layer (sample SA2, a) and on and Co indicate that smalt blue has been employed as pigment for the blue layehat neoformation gypsum has formed on a painting layer containing green eart

Optical and SEM-EDS stratigraphic investigations on paint-ng fragments prove a finishing lime-plaster stratum with a fineilicate and carbonate sand, followed by the original paintingayers, mainly red and yellow ochres, green, brown and umberarths. Furthermore, smalt as blue pigment, lead and tin yellow,innabar and minium were used to obtain more brilliant yellownd red tones [8–10] (Fig. 2a). Notwithstanding the pigments arentimately mixed with lime in all the pictorial layers in agreementith the classical fresco technique, the painting layer thickness,hich spans from 5 to 95–100 �m, and the IR identification ofroteinaceous material and calcium oxalate in some pictorial andlaster strata, the latter, quite probably deriving from the reac-ion of lime with the employed organic compounds, prove these of a different painting technique from fresco, in agreementith the non-exclusive use of lime as ligand [11] (Fig. 3). Thus,

he use of an “auxiliary tempera” (i.e., the painter’s intentionalddition of a binder to the pigment and the application of theesulting mixture on a damp plaster) appears to be a convincingroposal [11].

ig. 3. IR spectra of the sample SA1 06, testifying the presence of (a) proteina-eous material (3400–3200 cm−1 N H stretching band, 3100–2800 cm−1 C Htretching bands, 1660–1600 C O stretching band, 1565–1500 cm−1 C N Hending band and 1480–1300 cm−1 C H bending band) and (b) oxalate salts1620 cm−1 and 1320 cm−1).

btibbtw[

gtbsdgtp

4

ws

iorated green layer (sample SA11, b). (a), the detected elements Si, Ca, K, Fe, the detected elements Ca and S, together with Si, Al, Mg, K and Fe, indicate

A careful observation of the painting surfaces reveals originalreas with light finishings whose adhesion to the substrate haseen assured by lime carbonation and other areas whose adhe-ion to the support is extremely compromised: the final results the occurrence of exfoliation and pulverization of the paint-ng layers in some areas and a satisfactory conservation state inthers. Moreover, some portions of the paintings show a deepolour saturation when observed by ranking light and a low waterbsorption during cleaning procedures; this surface behaviour,ompletely different from that of a fresco one, may be ascribedo a fatty consistency of the pictorial materials.

Optical and electronic observations show that many of theseeteriorated original light finishings contain lead white [8–10]ixed with the above mentioned pigments and proteinaceous

inders. The presence of fatty materials has been confirmedy �FT-IR spectra, collected from the CH2Cl2 soluble frac-ions which clearly prove the presence of siccative oil, whichs also partially deteriorated due to hydrolysis of fatty estersy the lead of the corresponding white pigment as confirmedy characteristic absorption band at 1592 cm−1 attributed tohe νasym of the carboxylate groups (Fig. 4a), in agreementith previous investigation on artificially aged painting models

12–14].The scientific data support the restorer observations and sug-

est that Allori might have used the same binding medium forhe whole pictorial cycle, applying the painting layer not car-onated on wet plaster (fresco technique) in some areas and onemi-dried plaster (stanco technique) in others, as shown by theifferent conservation state of the identical substrate of adjacentiornate. It is not possible, however, to ascribe unambiguouslyhe presence of fatty binders to an intentional addition of theainter or to subsequent restoration interventions.

.2. Past restorations and conservation state

Before restoration, the different scenes could be observedith great difficulty and their meaning was highly misunder-

tanable, mainly owing to extensive blackening due to the use

S. Bianchin et al. / Journal of Cultura

Fig. 4. IR spectra of (a) the sample SA8 06 showing the presence of carboxylategroups (1592 cm−1 asymmetric stretching of COO−) probably originating fromsiccative oils’ esters hydrolysis by lead white; (b) the sample SA4 06 showinge1b

adbeadt

acitl

ac(o

sd

ortotasbp

ibttosetgp

tm

sta

4

Fa

sters of fatty acid (3100–2800 cm−1 C H stretching bands of aliphatic groups,750–1725 cm−1 C O stretching band and 1300–1100 cm−1 C O stretchingands) ascribable to unburnt residue of candles.

long the centuries of candles and votive lamps (Fig. 5). Thearkening of wide areas of the cycle is due to a surface black-rown deposit, containing carbonaceous particulate and fattysters, as confirmed by IR analyses; the deposit compositionnd its detection on the painting surfaces strongly suggest iterives from unburnt residues of candles smoke condensed onhe coldest surfaces of the painted vault (Fig. 4b).

Moreover, the close examination of the painted surfacesimed at identifying subsequent restoration interventions, indi-ated that, except some localized retouchings and limitednterventions in areas interested by seepage or other damage,he most relevant intervention at the end of 18th century was theast one extended to the whole pictorial cycle.

The past restorations can be grouped as follow: those with

heavy reinforcement and complete alteration of the original

olour, together with the addition of completely new detailsfor instance, the inkpot and goose quill of The Eritrean Sybiln one of the pendentive of the vault) and those carried out to

boto

ig. 5. The scene The Disputation with the Doctors during restoration: extensive blappears in the upper part after the different cleaning trials.

l Heritage 10 (2009) 379–387 383

olve degradation problems, in consequence of lack of pictorialefinition.

The remarkable variation of the conservation state of theriginal paintings might have caused the need of consistentepaintings and remarkings of the pictorial strata. In particular,he effects of aggressive past cleanings with caustic solvents,bserved on the surfaces, have been proved by scientific inves-igations, that is, by the presence of chlorine, although in a lowmount and sodium proving the use of HCl and NaOH. Theseevere cleanings made necessary repaintings, particularly visi-le in those areas where the painter has uncorrectly worked onartially and totally dry plasters.

The repaintings, carried out in the past to mitigate this unsat-sfactory situation, and to obtain a more adequate view and aetter chromatic equilibrium, involved not only the figures ofhe scenes and the adjacent, false architectonic frames, but alsohe architectonic fillets of the vault. Moreover, the whole surfacef the Chapel was treated with an organic fixative to reinforce ortrenghten the chromatic effects: this, in addition to the dark-ning due to the degradation of constituent materials and tohe candle soot absorption, was severely altered by thermohy-rometric variations with the consequent formation of whitishatinas.

In particular, SEM-EDS and �FT-IR measurements provehat the pictorial layers of the past retouchings were made by

ixing gypsum and natural earths (Fig. 2b).Neoformation gypsum was also found in correspondence of

ome areas, deteriorated by water seepage and salt crystalliza-ion, with the consequent occurrence of blisterings, detachmentsnd pulverizations of the pictorial layers.

.3. Conservative restoration

At the first instance to define the conservation procedures, it

ecame necessary by a careful examination by close inspectionsf the surfaces to define better the original paintings, to highlighthe deterioration pattern, to show the repaintings and fixativesf past interventions and to collect samples for the scientific

ckening hides the painting on the uncleaned lower area, while brilliant colours

384 S. Bianchin et al. / Journal of Cultural Heritage 10 (2009) 379–387

F pplieda

aqpmrt

r

Fda

ig. 6. Cleaning trials with water and ammonium carbonate aqueous solutions after compresses’ removal (right).

nalyses. The removal of deposits or degradation materials isuite complex and delicate; it needs a deep knowledge of the

hysical and chemical properties of these materials and of theost appropriate methodologies for their removal, joined to a

efined deftness, experience and ability for a careful control ofhe cleaning procedures (Fig. 5).

bctf

ig. 7. Cross-sections of the samples SA5 06, SA6 06 and SA7 06 contains a thicisappearance of the deposit stratum and the absence of the characteristic bands of thmmonium carbonate is capable to remove completely the deposit without any altera

by cellulose compresses on a monochrome area (left) and the cleaned surfaces

The intervention was firstly addressed to the mechanicalemoval of the solid particles on the surface and candle soot

y Wishab sponge. Where possible, no preconsolidation wasarried out in order to avoid fixing firmly the candles soot tohe substrate. Careful laying of the fragile exfoliated pictorialragments were carried out with aqueous solution of ammo-

k surface stratum of gypsum, partially removed by poultice with water. Thee sulphate group in the IR spectrum confirms that the subsequent poultice withtion of the below pictorial layer.

S. Bianchin et al. / Journal of Cultural Heritage 10 (2009) 379–387 385

fter (b

nlss

taao

•

•

pbo

segcncto

niecrta

Fig. 8. Some details of the pictorial cycle before (a) and a

ium caseinate on Japanese paper superimposed as a protectiveayer over the paintings, making their adhesion on the underurface possible and postponing the final consolidation after aatisfactory cleaning.

As the cleaning was with water solutions and chemicals,he synergic efforts of restorers and scientists become morend more necessary in order to combine correctly diagnosisnd intervention. The scientific investigations during removalperations were primarily focused:

to monitor the efficiency of the different cleaning proceduresin removing gypsum and other organic and/or inorganic con-taminants from the surface and contemporary;to set up the most appropriate non-expensive, fast and as lessinvasive as possible, scientific tests and methodologies to beproposed to verify the different cleaning procedures.

The scene The Disputation with the Doctors, where a diffuseresence of gypsum and blackening of the surface occurs, haseen selected for cleaning trials with distilled water and aque-us solution of (NH4)2CO3 (Fig. 6). The cross-sections of the

olco

) cleaning of the scenes: Nativity and The Eritrean Sybil.

amples SA5 06, SA6 06 and SA7 06, reported as explanatoryxamples of the cleaning sequence (Fig. 7), show that the thickypsum deposit can be partially removed by a first application ofompress imbibed with water. The next compress with ammo-ium carbonate is capable to remove completely the deposit, asonfirmed by the disappearance of the characteristic bands ofhe sulphate group in the IR spectra, without any modificationf the inner original pictorial layer.

Furthermore, diagnostic investigations confirm the correct-ess of the applied selective cleaning methodology, carried outn well-defined areas. SEM-EDS and �FT-IR prove the pres-nce of gypsum and wax residues before cleaning; the firstompress with water removes gypsum but not the organic mate-ials, the subsequent one with ammonium carbonate removeshe fatty materials; after this procedure only calcium carbonatend silicates have been detected on the surface.

As mentioned above, the first cleaning attempt was carried

ut by distilled water; this operation, owing to the thick surfaceayer of candles soot, was executed by applying cellulose pastompress imbibed with water on the surface with contact timesf about one hour. This favoured the highest swelling of the dirty.

3 ultura

Mtft

itd

wifsc

cciaitawtctcl

otipfcasa

paoi

5

iFcmc

wiit

cswhnrsibG

eot

awaw

tpmcd

cipw

tob

A

t

R

86 S. Bianchin et al. / Journal of C

oreover, owing to the low interpenetration of the original pic-orial layers with the repainting ones, the latter clearly exfoliatedrom the surfaces; a light cleaning with water was almost enougho remove most of the weakly adherent, non-original repaintings.

These procedures showed the original brightness of the paint-ngs together with the remarkable inaccuracies: areas of greatechnical refinement alternate to others neglected and made onry mortar, corrections and rushed paintings.

Many areas, where neoformation gypsum has been identified,ere also treated for about one hour with cellulose compress

mbibed with a saturated solution of ammonium carbonate toacilitate the sulphate removal, as demonstrated by IR mea-urements at different cleaning steps, and shown to favour theonsolidation with barium hydroxide.

The cleaning procedure has made it possible to restore thehromatic equilibrium and to see correctly the whole cycle,omprehensive of some very interesting details, for instance thenscriptions on the clothes of the personages painted in Christmong the Doctors. Moreover, the retouching removal madenteresting interpretative discoveries, for instance the identifica-ion of the Young Christ with the benefactor Bastiano Montautond other member of this family [1]. Also, the Allori’s signatureas found on the Christ among the Doctors and an inscrip-

ion on the S. Matthew. Obviously, on one side cleaning hasontributed to the recovery of the originality of the pictorialext, on the other side it has shown, in its own crudity, the badonservation state and the heavy degradation of the pictorialayers.

After cleaning, consolidation of the surfaces by compressf barium hydroxide was extensively performed. In additiono the elimination of sulphate soluble salts, the latter resultsn the regain of the chromatic equilibrium and the contem-orary increase of the material consistency of the surfaces,avouring recoesion of the pictorial strata. Furthermore, thehemical reactions of barium hydroxide to barium carbon-te contributed to reduce the high porosity of the surfaces,trengthening their capability to stand up to future pollutantsttack.

Finally, the designed pictorial integration, made with lightainting retouching, have mitigated or smoothed the effect ofbrasions and consumptions, with great benefit for the fruitionf the entire cycle. Some explanatory examples of the restorationntervention are reported in Fig. 8.

. Conclusion

The conservation intervention carried out on the mural paint-ngs at San Girolamo Chapel – SS. Annunziata Church inlorence, Italy gave the chance to approach scientifically theonservation issues, that is, diagnosis, planning and develop-ent of conservative interventions, through a fruitful synergic

ollaboration between restorers and scientists.This pictorial cycle, executed by Alessandro Allori and co-

orkers between 1560 and 1564, was interested by two mainnterventions: the first carried out in 1799 by Santi Pacini, thatncluded extensive repaintings, and the last during the thirties ofhe last century by Amedeo Benini.

l Heritage 10 (2009) 379–387

The restorer’s experienced perception of painting surfacesombined the results of the scientific investigations demon-trated that “fresco” painting technique has been used on thehole cycle, although an auxiliary binder for pigment distemperas been also employed. Furthermore, the use of this tech-ique, sometimes in an improper way together with uncorrectedestoration works, are the main responsible of the unsatisfactorytate of conservation of many painted areas, as confirmed by thedentification of the deterioration markers, previously definedy laboratory experiments, on the painted surfaces of the Sanirolamo chapel.The wall paintings were almost completely hidden by an

xtensive blackening due to a thick deposit of unburnt residuesf candles smoke, as demonstrated by IR analyses, ascribed tohe use over the centuries of candles and votive lamps.

Furthermore, it has been proved that a severe cleaning withggressive products, that is, acid and alkaline inorganic products,as responsible of loss of painting details and a treatment with

n organic fixative has been subsequently applied on almost thehole painting to reinforce the chromatic effects.After completion of the diagnosis, the crucial steps of

he cleaning procedures were continuously monitored by ahysico-chemical methodology, mainly based on microinvasiveicroscopic and spectroscopic investigations, to verify the effi-

iency, advantages and drawbacks of the cleaning trials and toefine the most appropriate ones.

The cleaning procedure has made it possible to restore thehromatic equilibrium and to discover original details hidden bymproper retouching. When necessary, the consolidation of weakainting layers have been performed with barium hydroxide inater solution.The final very satisfactory results are the rehabilitation of the

rue chromatic equilibrium in the whole cycle and the validationf scientific approach to the conservation issues that successfullyalances the experience and skills of either restorers or scientists.

cknowledgements

We thank Mr. A. Aguiari for the qualified technical assis-ance.

eferences

[1] E. Pilliod, Studies on the early career of A. Allori, PhD dissertation, Uni-versity of Michigan, Ann Arbor, 1989.

[2] U. Baldini, P. Dal Progetto, Firenze restaura, Catalogo della mostra, San-soni, Firenze, 1972.

[3] P. Tonini, Il Santuario della SS, Annunziata di Firenze, Firenze, 1876.[5] A. Armigliano, V. Valdrè, Microscopia Elettronica a Scansione e

Microanalisi, Lo Scarabeo, Bologna, 1980.[4] R.G. Messerschmidt, M.A. Harthcock (Eds.), Infrared Microspectroscopy,

Pratical Spectroscopy Series, vol. 6, Marcel Dekker Inc, NY, 1988.[6] G. Botticelli, Metodologia di restauro delle pitture murali, Centro di,

Firenze, 1992.[7] G. Botticelli, S. Botticelli, Lezioni di restauro, Centro di, Firenze, 2008.[8] S. Rinaldi, G. Quartullo, A. Milaneschi, R. Pietropaoli, S. Occorsio, F.

Costantini Scala, G. Minunno, C. Virno (Eds.), La Fabbrica dei Colori,Bagatto libri, ristampa, Roma, 1995.

ultura

[

[

[

[

“protusions” in aged traditional oil painting, Polyhedron 22 (2003)3171–3179.

S. Bianchin et al. / Journal of C

[9] R.L. Feller (Ed.), Artist’s Pigment, National Gallery of Art, vol. 1, OxfordUniversity Press, New York, 1986, 109–140.

10] A. Roy (Ed.), Artist’s Pigment, National Gallery of Art, vol. 2, OxfordUniversity Press, New York, 1993, 83–182.

11] C. Alliata di Villafranca, Restauro dei dipinti e tecniche pittoriche–Teoriae procedimenti operativi, Quattrosoli, Palermo, 1997.

12] M.R. Derrick, D. Stulik, J.M. Landry, Infrared spectroscopy in conservationscience–scientific tools for conservation, The Getty conservation Institute,Los Angeles, 1999.

[

l Heritage 10 (2009) 379–387 387

13] M.J. Plater, B. De Silva, T. Gelbrich, M.B. Mursthouse, C.L. Hig-gitt, D.R. Saunders, The characterization of lead fatty acids soaps in

14] R. Arbizzani, U. Casellato, E. Fiorin, L. Nodali, U. Russo, P.A. Vigato,Decay markers for the preventive conservation and mainteinance of paint-ings, J. Cultur. Herit. 5 (2004) 167–182.