O

Aa

HC

a

ARAA

KTSASCPSFD

1

misoAbh

httirrp

1d

Journal of Cultural Heritage 12 (2011) 412–419

riginal article

new approach for conservation treatment ofsilk textile in Islamic Art Museum, Cairo

arby E. Ahmed ∗, Yassin E. Ziddanonservation Dept. Faculty of Archeology, Cairo University, 12613 Orman, Giza, Egypt

r t i c l e i n f o

rticle history:eceived 8 October 2010ccepted 21 February 2011vailable online 1 April 2011

eywords:extileilkdhesive

a b s t r a c t

The paper presents strategies for the conservation of historical textiles in Egypt that have been in uncon-trolled storage and display. The silk textile is highly decorated, multicoloured and dates to the Ottomanperiod, and was exhibited in case # 12014. The textile has various types of deterioration. An examinationand analysis of the textile was undertaken in order to develop a plan of conservation treatment. FTIR wasused to identify the kinds of dyes and organic stains, and XR-D was used to identify mordants and dust.Light microscope and SEM were used to identify the kind of fibers, their condition and surface morphol-ogy. The effects of cleaning materials on the natural dyes were tested. The researcher designed a newmetallic frame support system which has advantages over the wooden frames commonly used in Egypt.

tainsleaningoulticeupportrameyes

This presentation will review the conservation treatment step by step. Poultices were used to removeall the sticking cardboard and adhesive. Old conservation repairs were removed. Separated parts weresupported. Cleaning included mechanical and wet cleaning. New silk fabrics dyed with natural dyes wereused to complete the missing parts. The textile was supported on new linen fabric which was stretchedon a metal frame. The method of exhibition will be discussed. Photographs are included to document theconservation process.

© 2011 Elsevier Masson SAS. All rights reserved.

. Introduction

Archeological textiles in Egyptian Museums are exposed toany challenges such as oscillate relative humidity (RH), chang-

ng temperature, effect of light, effect of air pollution, non standardtorage and display methods, old poor restoration and old typef adhesives such as starch adhesive, animal glue adhesive, andrabic gum adhesive. All these factors cause damage and decay inoth fibers and dyes which results in weakness in fibers and dye,ardness in fibers, separate parts, loss parts, stain and dust [1–5].

The paper aims to present the strategies for the conservation ofistorical textiles in Egypt. It shows new methods in the conserva-ion of historic textiles, at least in Egypt. The paper aims at knowinghe kinds of fibers and dyes, stain, dust, and different damagesn this object through different ways of investigation. The papereports the conservation treatment of the object such as cleaning,emoving the old adhesive and old restorations, completing lossarts, fixing separate parts and making a new display.

∗ Corresponding author.E-mail address: harbyezzeldeen@yahoo.com (H.E. Ahmed).

296-2074/$ – see front matter © 2011 Elsevier Masson SAS. All rights reserved.oi:10.1016/j.culher.2011.02.004

2. Description of object

The textile object dates back to Ottoman age and was shown incases no 12014 in the Islamic Art Museum, Cairo. Its measurementsare 116 × 69.7 cm. Its contains many decorations such as plants dec-oration (flowers and leafs), written decorations inside lamp hangfrom arch, and geometric decorations (columns, arches). It also con-tains different colors such as red, blue, green and yellow. The objectwas kept under glass fixed on a plywood support lined by cottonfabric, a fixing support, and glass covered with a plastic sticker fromthe edges. There are many signs of damage on this object such as oldadhesive, old cardboard stuck on the object’s back, many separateparts from the edges, loss parts, weakened fibers, and brittlenesscombined with old adhesives and other previous repairs, hardness,old restoration error (such as using paper sticker to fix separateparts), staining and dust. Figs. 1 and 2 illustrations the condition ofthe original.

3. Testing and analysis

3.1. Morphological study

The morphology of the surface of the fabrics was investigatedusing a Quanta 200 ESEM FEG from FEI Scanning Electron Micro-

H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419 413

Fig. 1. Textile object dates back to Ottoman age and shows cases no. 12014; one can see a lot of decorations, colors. The object was kept under glass fixing on a plywoodsupport lined by cotton fabric, a fixing support, and glass covered with a plastic sticker from their edges (A). The plywood support lined by cotton fabric (B). The old glassp o, we

sdoSstw

Fc

late over the object was removed by a researcher (C). The object from the back als

cope (SEM). Small samples were taken from the object fromifferent parts and investigated under SEM, to show the qualityf the fibers as well as the damage aspects on these fibers [6–8].

EM Photos of examined Ottoman textile are illustrated in Fig. 3howing the silk fibers that were identified from different parts ofhe object. The fibers are extremely roughened, damaged, brokenith transverse cracking and longitudinal splitting characterized

ig. 2. The figure shows different type of deteriorations such as dust, grease, losses, seardboard support and other previous repairs.

can see the remnants of the old cardboard (D).

by small scratches, small slits and holes. Furthermore, one can seethe dust, dirt and adhesive that covered the fiber as shown in Fig. 4.

3.2. X-ray diffraction analysis

X-ray diffraction of fabrics was carried out with a SIEMENS X-Ray Diffractometer–D 5000, given 40 Kv CU Ka, radiation of 30 mA.

parated parts, weakened fibers, and brittleness combined with old adhesives, old

414 H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419

F at thl

Tad

3

u

F(

ig. 3. The figure shows SEM images of examined Ottoman textile, One can see thongitudinal splitting characterized by small scratches, small slits and holes.

he diffractograms were recorded over 2� = 50 to 300 continuouslyt a scan rate of 20/min. to show the kinds of mordent (Alum) andust (sand) [6,7].

.3. Fourier Transform infrared spectral analysis (FTIR)

FTIR analysis of solid phase samples can be typically performedsing two different methodologies. The first, more traditional and

ig. 4. The figure shows SEM images of examined Ottoman textile, one can see different tC and D).

e fibers are extremely roughened, damaged, broken with transverse cracking and

widely used approach, FTIR–KBr, utilises a pellet of potassium bro-mide (KBr) which is transparent in the mid-infrared region, as asupport medium for the sample. The mixture is usually ground withan agate mortar and pestle and subjected to a pressure of about

10 tonne in an evacuated die. This sinters the mixture and pro-duces a clear transparent disc, which is then placed in the samplechamber of the spectrometer for direct absorption measurements.The second, more recent approach is termed FTIR–ATR. This utilises

ypes of dirt and stain (A and B) Also, we can see old adhesive that covered the fibers

H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419 415

e (A).

amipwaaiwiBitr

TttSbaw

3

w

Fig. 5. Spectra of silk dyed with Safflower dy

n attenuated total reflection (ATR) accessory which operates byeasuring the changes that occur in a totally internally reflected

nfrared beam when the beam comes into contact with a sam-le. An infrared beam is directed onto an optically dense crystalith a high refractive index at an angle greater than the critical

ngle for total internal reflection. This internal reflectance createsn evanescent wave that extends beyond the surface of the crystalnto the sample held in contact with the crystal. This evanescent

ave protrudes only a few microns beyond the crystal surface andnto the sample [9,10]. These FTIR spectra were obtained using aRUKER–FTIR- TENSOR 27 Spectrometer. An expanded spectrum

n the 4000–400 cm 1 range was used for measuring several fac-ors with a spectral resolution of 4 cm 1. Each spectrum was theesult of an average of 16 scans.

We took small samples of different colors and investigated them.hen, we brought all the dyes which give this color and investigatedhem. After that, we compared the original samples with new dyeso find the following: (Red color is Cochineal dye–Yellow color isafflower dye–Blue color is Indigo dye–Green color is a mixtureetween Indigo and Turmeric dye–the adhesive is the Arab gumdhesive) as shown in Fig. 5. Furthermore, no mordants were foundith FTIR analyses.

.4. Testing the stability of dyes

The dry fabric (the object) was softened by spraying distilledater, to counter its extremely dry condition. The next step was to

Spectra of silk dyed with Cochineal dye (A).

test the stability of the colored parts to wet cleaning by immersinga piece of cotton wrapped round a wooden stick into the cleaningsolutions and placing it in contact with the colorful parts of the rib-bons, each color was individually tested. It was found that all thedyes were stable and did not bleed with the wet cleaning solution.The final step was to apply a primary support to the Ottoman textileby placing it between two webbed support fabrics, and stabiliz-ing the fabric by fixing it to the support fabric, using appropriatelythin needles and fine silk thread in order to protect the vulnerablepart of the textile from disintegrating during the different cleaningprocesses [11,12].

4. Removal of the old restoration error

In the beginning, the old glass plate over the object wasremoved. The object in the case showed severe dryness. Hence,the water sprayed over the object to moisten it. Then, the old adhe-sive and old cardboard were removed. There are two methods toremove the old adhesive and the cardboard.

4.1. Traditional method

Wet the upper surface of cardboard by warm water, then leaveit for 10 mins until water penetrates the cardboard and dissolve theadhesive. After that, remove wet parts carefully. This method doesnot remove all the sticking cardboard from the textile.

416 H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419

F utting a paper free acid on the upper surface of cardboard (A). Wetting it by warm water,a Remove the poulticing from the cardboard, and that will remove the sticking cardboarda

4

attauawacswc

5

5

dw

5

tpafiwosc

ig. 6. The figure shows the steps of using poultice to separate adhered cardboard. Pnd leave it for 10 mins to permit water to penetrate and dissolve the adhesive (B).nd adhesive (C and D).

.2. Using poultice to separate adhered cardboard

Application of a poultice locally on the textile within the pastedrea, allows the moisture to migrate vertically through the poul-ice to cardboard. The acid free paper (blotter) is a multilayer paperhat has been especially designed for poulticing. In order to achieven even decomposition of the paste, it is necessary to guaranteendisturbed and homogenous migration of water. Firstly, place ancid paper free on the upper surface of cardboard. Then, wet it byarm water, and leave it for 10 mins to permit water to penetrate

nd dissolve the adhesive. Finally remove the poulticing from theardboard, and that will remove the sticking cardboard and adhe-ive. The use of poultices would appear to be particularly suitablehen a localized treatment is necessary to remove all the sticking

ardboard and adhesive as shown in Fig. 6 [13].

. Cleaning procedure

.1. Mechanical cleaning

Various types of fine brushes are used to remove free dust andirt (i.e., not attached to textile fibers). Aerobic aspiration of dustas done by using a vacuum cleaner [3].

.2. The wet cleaning procedure

This cleaning procedure used water with other detergent agents,o assist the cleaning process. The ratio was one part detergent Syn-eronic N to 100 parts of distilled water. The water was agitated tollow it to penetrate between the fibers to release the dirt particles,or 15 mins. The bath temperature was 30 ◦C. Then a second clean-ng bath with distilled water only was applied for 10 mins again

ith water agitation, and then a third bath with distilled waternly, for 10 mins as shown in Fig. 7. The wet cleaning reduced theoiling, relaxed the fibers, removed the creasing and brightened theolors [3,14,15].

Fig. 7. The figure shows the wet cleaning procedure (A), the drying procedure (B).

6. The drying process

To dry the object without distortions, we experimented witha Japanese tissue sandwich using different tissues (usugami or

H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419 417

F he meo rt onw

giwsettth

7

maapb

hvtit

•••

8

bo

color, which turns blue when the fabric is exposed to the air. After20 mins, the silk is completely blue. Rinse thoroughly with water,but only after the fabric has dried completely. In order to obtain adeeper shade, dip the silk fabric into the liquor again and take it out

ig. 8. The figure shows the steps of fixing the linen support to the metallic frame. Tn the metallic frame (A). Thick cotton thread was used for fixing the linen suppohich were perforated after fixing on the metallic frame (D).

ampi) applied cross-grain or parallel-grain to the textile with vary-ng amounts of moisture and pressure. Consequently, the object

as left uncovered to complete drying at ambient conditions ashown in Fig. 7. In air drying, the process can be shifted towardsvaporation in several ways: by ensuring that the wet textile is set-ed in a place (i.e. workroom) that is sufficiently large in comparisono the size and surface area of the drying textile; and by ensuringhat water vapor is removed as soon as it is formed e.g. with theelp of dehumidifiers and/or effective ventilation [16,17].

. The final support process

After completion of the object drying, the fabric webbing pri-ary support was removed. A new linen support was prepared

nd washed to remove chemical residues and prevent shrinkaget a later time due to humidity changes. Then the new linen sup-ort was ironed to remove creases and was placed inside the shirtetween the outer and the inner layers as shown in Fig. 8.

Wooden frames were used in the past for mounting, but theyave the some disadvantages: wood is easily affected by theariation of moisture and temperature that causes twisting and dis-ortion, and it is affected by biological deterioration (i.e., fungi andnsects). So, the researcher created a new metallic frame which hashe following advantages:

it is not affected by moisture or biological deterioration;it is light and easily moved;it can be coated with different colors.

. Completion and final support of lost parts

The samples that were used to complete the lost parts were dyedy the researcher who used indigo for the blue color and a mixturef indigo and turmeric to get the green color.

tallic frame, one can see the rings were fixed on this perforation for easy stretchingthe metallic frame (B and C). A linen support was used after supporting the edges

8.1. Dying procedure of new parts

8.1.1. Blue color with indigo on silkStir 15 g of indigo powder with 75 ml of warm water in beaker

glass until it forms a paste. In a second vessel, dissolve 30 g of sodain 120 ml of warm water. Pour 60–70 ml of this solution over theindigo paste and stir vigorously. Then add 30 g of sodium dithion-ite and stir again. Add one liter of warm water and stir carefullyuntil the whole is thoroughly mixed. Heat this mixture to 55 ◦C.The liquid should now have a yellowish color. When it is left stand-ing for 20 mins, the color will have turned yellow-green. Add 30 gof sodium dithionite to the solution.

Heats the 1 L dye bath to 55 ◦C. Immerse 30 g of silk fabric inwarm water until the material is thoroughly wet, and then enter itinto the dyeing liquor. Let the dye bath stand still, so that no oxygencan enter into the vat. Keep the silk fabric in the vat for 20 mins, andthen take it out of the vat and squeeze the liquor out thoroughly.When the silk comes out of the vat, it will have a green-yellow

Fig. 9. Temperature time diagram of the one-bath dyeing process.

418 H.E. Ahmed, Y.E. Ziddan / Journal of Cultural Heritage 12 (2011) 412–419

F houldi parts

ad

8

dt

••

Ft

ig. 10. Lost part before and after conservation. One can see the added dyed parts s.e., below the lost part where the dyed parts are slightly larger in size than the lost

fter 20 mins. Repeated dipping and airing makes the dyed shadeeeper and deeper [18–20].

.1.2. Green color with indigo and turmeric dye on silkSilk dyed fabric with indigo was immersed again in Turmeric

ye. The dyeing with Turmeric dye was carried out according tohe following steps:

prepare a 10% dye in water solution;soak the dyes in the distilled water for 24 h to extract the colorfrom the powder;

ig. 11. Another lost part before and after conservation. One can see the added dyed partextile i.e., below the lost part where the dyed parts are slightly larger in size than the los

be between the linen band (used to support the edges) and the body of the textile,.

• heating the extract to the boiling temperature for 2 h with con-tinuous stirring. It may require addition of water to compensatethe evaporated water during the heating process;

• allow the extract to be cooled and then filtered many times to geta clear colored solution.

8.1.2.1. Dyeing procedures. The dyeings were performed by the

exhaustion method using a liquor ratio (LR) of 1:20 (1 g of fab-ric per 20 ml of bath). The dyeing experiments were performed inglass beakers according to the temperature-dyeing diagram givenin Fig. 9. In the experiments mordant (alum), was added as concen-

s should be between the linen band (used to support the edges) and the body of thet parts.

H.E. Ahmed, Y.E. Ziddan / Journal of Cult

Fs

tot1

8

washAtb[

8

wbtl

tmblh

8

wsfi

8

u

[

[

[

[

[

[

[

[

[

[

Annual Conference, UK, 2005, pp. 246–250.

ig. 12. The object after conservation treatment and after fixing it on new linenupport.

rated solution (50 g/l) to give a final dye bath concentration of 2.5r 5 g/L. After dyeing, the unfixed dyestuff was removed by rinsinghree times with cold water (5 min, room temperature [25 ◦C], LR:20.) [18–20].

.2. Completion of the separate parts

The margins were weak and stiff. The stiffness was removed byater and the back of these margins were lined from four sides bylinen band of 5 cm width fixed under the margins including the

eparated edge and the rest of the textile. Fine silky stitches, whichave the same color of the part we wanted to support, were used.t first, a piece of the linen band is fixed under the textile, and then

he separated edge is moved to the textile and fixed to the linenand. Small stitches are used to stitch the edge and textile together3,16].

.3. Completion of the lost parts

The lost parts were completed using silk parts of similar colorhich were slightly lighter. These added dyed parts should be

etween the linen band (used to support the edges) and the body ofhe textile, i.e., below the lost part where the dyed parts are slightlyarger in size than the lost parts.

The added parts have been fixed by fine stitches of silk dyed byhe same color. When the lost part is free, i.e., present at the textile

argin to the outside, the silky parts are fixed at first to the linenand by fine stitches followed by adjustment of the band under the

ost part to be fixed by fine stitches in the back using silk thread thatas the same color of lost part a shown in Figs. 10 and 11 [3,16].

.4. Fixing the linen support to the metallic frame

A linen support was used after supporting the edges whichere perforated. Then, rings were fixed on this perforation for easy

tretching on the metallic frame. Thick cotton thread was used forxing the linen support on the metallic frame (Fig. 8).

.5. The final support process

After completion of the object treatment, tacking stitches weresed with a very fine needle and fine silk thread to fix it into

[

[

ural Heritage 12 (2011) 412–419 419

the object. In the beginning of the final stage, the edges of theobject all around were attached by sewing with a small stitchtechnique (blanket stitch) and afterwards the edges of the miss-ing and vulnerable parts were attached by small stitches. Similarly,sized stitches were used to attach the body of the object. Thesleeves were supported by attaching them to new linen fab-ric. After completing the cleaning process and fixing the object,it could be displayed in a suitable manner according to themuseum requirements. One can see the object in final stage inFig. 12 [21].

Acknowledgement

The authors would like to thank the conservators and archaeol-ogists in Islamic Art Museum, Egypt for their contributions.

References

[1] F. Lemberg, Textile conservation and research, Bern, 1988, pp. 31–55.[2] A. Harby, Treatment and conservation of textile with written decoration applied

to some chosen samples from the textile of the Ottoman Age, Cairo Univ., Egypt(2002) 220–257 (M.A. thesis).

[3] S. Landi, The textile conservator’s manual, 2nd edition, London, 1992, pp.120–123.

[4] J. Coddington, M. Ainsworth, Conservation and art history, Art J. 54 (1995)16–17 (www.jstor.org/stable/777457).

[5] M. Brown, A preservative compound for archaeological materials, Am. Antiq-uity 39 (1974) 469–473.

[6] S. Andrew, Using ultraviolet and infrared techniques in the examination anddocumentation of historic textile, Conservator 18 (1994) 86–94.

[7] G. Nicola, The examination and conservation of tow indian textile, Conservator17 (1993) 102–111.

[8] J. Batcheller, Optical and scanning electron microscopy techniques for the iden-tification of hair fibers from Romano-Egyptian textiles, in: R. Janaway, P. Wyeth(Eds.), Scientific Analysis of Ancient and Historic Textiles: informing preserva-tion, display and interpretation, AHRC Research Center for Textile Conservationand Textile Studies, First Annual Conference, UK, July, 13–15, 2005, pp.51–57.

[9] D. Michele, Fourier Transform infrared spectral analysis of natural resins usedin furniture finishes, J. Am. Inst. Conserv. 28 (1989) 43–56.

10] T. Thompson, M. Gauthier, Meez Islam, The application of a new method ofFourier Transform Infrared Spectroscopy to the analysis of burned bone, J.Archaeol. Sci. 36 (2009) 910–914.

11] O. Brigitte, Fastness to light and washing of direct dyes for cellulosic textile,Studies Conserv. 41 (1982) 97–106.

12] E. Wilson, Color conservation problems of a twentieth century historicdress, preservation of paper and textile of Historic and Artistic, Vol. 2, USA,1981.

13] K, Thompson, Sepiolite poulticing an alternative for cleaning of textile, Conserv.News 52 (1994).

14] S. Haward, An introduction of the wet cleaning of carpets, The InternationalPerspective on Textile Conservation Report, ICOM, Amsterdam, 1994.

15] Y. Moes, Tapestry Cleaning by Aerosol Suction, The International Perspectiveon Textile Conservation Report, ICOM, Amsterdam, 1994, pp. 32–35.

16] A. Balázsy, D. Eastop, Chemical principles of textile conservation. Butterworth-Heinemann, UK, 1998.

17] S. Fletcher, J. Walsh, The treatment of three prints by whistler on fine Japanese,J. Am. Inst. Conserv. 18 (1979) 119–126 (http://aic.stanford.edu/jaic/articles/jaic18-02-005.html).

18] T. Bechtold, A. Turcanu, E. Ganglberger, S. Geissler, Natural dyes in moderntextile dyehouses – how to combine experiences of two centuries to meet thedemands of the future? J. Clean. Product. 11 (2003) 499–509.

19] H. Ahmed, Y. Zidan, K. El-Nagar, Studies on dyeing with cochineal and age-ing of silk dyed fabric, in: R. Janaway, P. Wyeth (Eds.), Scientific analysis ofancient and historic textiles: informing preservation, display and interpreta-tion, AHRC Research Center for Textile Conservation and Textile Studies, First

20] H. Schweppe, Practical information for the identification of dyes on historicaltextile materials, Washington DC, USA, 1988, pp. 1–9.

21] V. Constance, The conservation of tapestries for museum display, Stud. Conserv.15 (1970) 134–153.