Post on 20-Jan-2023
ORIGINAL PAPER
HPLC–DAD–MS analysis of dyes identified in textilesfrom Mount Athos
Dimitrios Mantzouris & Ioannis Karapanagiotis &
Lemonia Valianou & Costas Panayiotou
Received: 9 July 2010 /Revised: 29 November 2010 /Accepted: 4 January 2011 /Published online: 28 January 2011# Springer-Verlag 2011
Abstract Organic colorants contained in 30 textiles (16th toearly 20th century) from the monastery of Simonos Petra(Mount Athos) have been investigated using high-performance liquid chromatography equipped with diode-array detection and mass spectrometry (HPLC–DAD–MS).The components of natural dyes identified in samples treatedby the standard HCl dyestuff extractionmethod were: alizarin,apigenin, butein, carminic acid, chrysoeriol, dcII, dcIV, dcVII,ellagic acid, emodin, fisetin, flavokermesic acid, fustin,genistein, haematein derivative (Hae′), indigotin, indirubin,isoliquiritigenin, isorhamnetin, kaempferide, kaempferol,kermesic acid, luteolin, naringenin, purpurin, quercetin,rhamnazin, rhamnetin, sulfuretin, and type B and type Ccompounds (last two are markers for Caesalpinia trees).Early, semi-synthetic dyes, for example indigo carmine,fuchsin components, and rhodamine B were identified inobjects dated late 19th to early 20th century. A dyestuff
extraction method which involves use of TFA, instead ofHCl, was applied to selected historical samples, showingthat the mild method enables efficient extraction of weld(Reseda luteola L.) and dyer’s broom (Genista tinctoriaL.) glycosides. The marker compound (Hae′) for logwood(Haematoxylum campechianum L.) identification aftertreatment with HCl was investigated by liquid chromatog-raphy coupled to mass spectrometry (LC–MS) in negativeelectrospray ionization (LC–MS-ESI−) mode. LC–MS innegative atmospheric pressure chemical ionization (LC–MS-APCI−) mode was used, probably for the first time, toinvestigate cochineal (Dactylopius coccus Costa) samples.Positive electrospray ionization (LC–MS-ESI+) modewas used for identification of fuchsin components.Detailed HPLC–DAD studies were performed on youngfustic (Cotinus coggygria Scop.) and Persian berries(Rhamnus trees).
Keywords Archeometry . Fine arts . HPLC .
Natural products
Introduction
Natural dyes played a central role in the production oftextiles and embroidered ecclesiastical objects [1, 2]. It isnoteworthy, however, that natural dyestuffs have been usedsince antiquity not only to dye textiles but also as colorantsfor paintings [3], food additives [4], cosmetics [5], andmedicines [5, 6]. Consequently, analysis and characteriza-tion of natural dyes is an important and challenging taskwhich has an impact on different applications of analyticaland bioanalytical chemistry.
The first objective of the work discussed in this articlewas to identify the organic colorants contained in
Published in the special issue Analytical Chemistry for CulturalHeritage with Guest Editors Rocco Mazzeo, Silvia Prati, and AldoRoda.
D. Mantzouris :C. PanayiotouDepartment of Chemical Engineering,Aristotle University of Thessaloniki,54124 Thessaloniki, Greece
D. Mantzouris : I. Karapanagiotis : L. ValianouArt Diagnosis Center, Ormylia Foundation,63071 Ormylia, Greece
I. Karapanagiotis (*)University Ecclesiastical Academy of Thessaloniki,54250 Thessaloniki, Greecee-mail: y.karapanagiotis@aeath.gr
I. Karapanagiotise-mail: g.karapanagiotis@artdiagnosis.gr
Anal Bioanal Chem (2011) 399:3065–3079DOI 10.1007/s00216-011-4665-4
historical textiles (16th to early 20th century) from themonastery of Simonos Petra in Mount Athos (Greece),using high-performance liquid chromatography (HPLC)equipped with a diode array detector (DAD) and massspectrometry (MS). Many textiles of inestimable reli-gious, spiritual, and historical significance have beenpreserved in the monasteries of the Holy Mountain,which has enjoyed autonomous status since Byzantinetimes. The materials which were used to dye theseobjects are to a large extent unknown, because of thevery limited number of relevant chemical studies [7, 8].This report is probably the first integrated effort to identifythe colorants contained in a relatively large number oftextiles from Mount Athos.
The second, and more important, objective of theinvestigation was to carry out chemical studies usinghistorical samples and reference materials to gain betterinsight into the results obtained from the historical objects.Before HPLC analysis, textile samples have to be treated toextract and dissolve the dyestuffs they contain. Treatmentof the dyed samples with methanolic hydrochloric acid(HCl) at elevated temperatures was suggested in the past[9]. This is still a widely adopted approach, which throughthe years, has become the standard dyestuff extractionmethod, because it has been successfully applied by severallaboratories to analyse probably thousands of historicaltextile samples. Because of its reliability the method isselected and applied to treat all samples extracted from thevaluable historical objects included in this study. However,the HCl method has some disadvantages including thedecomposition of glycosides which occurs under harshacidic conditions. Recent investigations have suggested theuse of milder acids for extraction of dyes from textilesubstrates [10–14]. The use of trifluoroacetic acid (TFA),instead of HCl, suggested by our group, seemed to giveoverall good results [14]. This previously devised TFAmethod has been applied to selected historical samples fromMount Athos to show that it is an efficient means ofextraction of weld (Reseda luteola L.) and dyer’s broom(Genista tinctoria L.) glycosides.
Dyestuff analysis using liquid chromatography cou-pled to mass spectrometry (LC–MS) can provide valu-able information [15, 16]. In this work, LC–MS studieswere carried out using reference dyestuffs (identified inthe historical samples) and some historical samples. Inparticular, logwood (Haematoxylum campechianum L.)was analysed using LC–MS in negative electrosprayionization (LC–MS-ESI−) mode to investigate the effectof the HCl extraction method [17]. Positive electrosprayionization (LC–MS-ESI+) mode was used for identifica-tion of fuchsin components, detected in the historicalsamples. Commercial fuchsin reference dyes are rarelycompletely pure; usually they are sold as “basic fuchsin”
which is a mixture of four fuchsin homologues—pararos-aniline, rosaniline, magenta II, and new fuchsin [18].Consequently, HPLC–DAD cannot be used for identifica-tion of these compounds. We show that identification canbe achieved by LC–MS. Cochineal components dcII, dcIVand dcVII are commonly reported in results from analysisof historical samples, including the results reported in thisinvestigation. However, the exact structures of thesecompounds are still unknown, despite the interestinginformation obtained from nuclear magnetic resonance(NMR) [19] and LC–MS-ESI− studies [19, 20]. LC–MS innegative atmospheric pressure chemical ionization (LC–MS-APCI−) mode has been used, probably for the firsttime, to analyse Mexican cochineal (Dactylopius coccusCosta) samples and provide information about the com-pounds dcII, dcIV, and dcVII.
Finally, detailed HPLC–DAD studies were carried outusing young fustic (Cotinus coggygria Scop.) and Persianberries (Rhamnus trees) as reference materials; both dyeswere identified in the historical samples.
Experimental
Materials and solvents
Dyestuff components which were available in pure formand were used as standard materials for identificationpurposes, and the corresponding suppliers, are summarizedin Table 1. Some of the standards were not obtainedcommercially but were previously produced in pure form[8, 21].
The dyestuff materials cochineal (Dactylopius coccusCosta), buckthorn berries (Rhamnus trees), and logwood(Haematoxylum campechianum L.) were purchased fromKremer Pigmente.
The following solvents and chemicals were used forchromatography, preparation of solutions, and sampletreatment processes: type I reagent-grade water withresistivity up to 18.3 MΩ cm−1 and organic content lessthan 5 ppb, produced by use of a Barnstead EASYpurewater-purification system, HPLC-grade methanol (MeOH)and acetonitrile (CH3CN) from J.T. Baker, HPLC-gradedimethyl sulfoxide (DMSO; Sigma–Aldrich), analyticalgrade hydrochloric acid (37%) (Riedel–de Haën), trifluoro-acetic acid of 99% purity (TFA; Riedel–de Haën), and proanalysis (98–100% purity) formic acid (HCOOH; Merck).
Chromatography
The HPLC–DAD system (Thermoquest, Manchester, UK)consisted of a 4000 quaternary HPLC pump, an SCM 3000vacuum degasser, an AS3000 autosampler with column
3066 D. Mantzouris et al.
oven, a Rheodyne 7725i injector with 20-μL sample loop,and a UV 6000LP diode-array detector. Analyses werecarried out with an Altima C18 (Alltech Associates, USA)column (5 μm particle size, 250 mm×3.0 mm) at a stabletemperature of 35 °C. Gradient elution was performed withtwo solvents, H2O–0.1% TFA and CH3CN–0.1% TFA [7,8]. Limits of detection (LODs) obtained for some of thecompounds listed in Table 1 by use of these chromato-graphic conditions have been reported previously [7]. Thewavelengths selected for routine monitoring and dyestuff
identification were: 254, 275, and 288 nm. To improve thequality of the data some samples were re-analysed usingother wavelengths (indicated in the figure captions).
LC–MS was used to analyse cochineal, fuchsin, andlogwood reference materials and some historical sampleswhich contained these dyes. A mass spectrometer(Thermoquest) equipped with ESI and APCI interfacesand a single quadrupole mass filter was used to recordion signals, as follows. ESI negative ionization-mode(ESI−) was used for investigation of logwood and its
Standard Supplier Dye
Carminic acid* Fluka Cochineal
Flavokermesic acid Previously produced [21] Cochineal
Kermesic acid Previously produced [21] Cochineal
Genistein* Fluka Dyer’s broom
Genistein-4′,7-dimethyl ether Extrasynthese Dyer’s broom
Genistin Extrasynthese Dyer’s broom
Luteolin-8-C-glucoside Extrasynthese Dyer’s broom
New Fuchsin* Sigma Fuchsin
Pararosaniline* Sigma Fuchsin
Indigo carmine* Fluka Indigo carmine
Indigotin* Fluka Indigo/Woad
Indirubin Previously produced [21] Indigo/Woad
Haematein Sigma Logwood
Haematoxylin Fluka Logwood
Alizarin* Aldrich Madder
Purpurin* Aldrich Madder
Emodin Acros Organics Persian berries
Isorhamnetin Sigma Persian berries
Kaempferide Extrasynthese Persian berries
Kaempferol Fluka Persian berries
Rhamnazin Extrasynthese Persian berries
Rhamnetin* Fluka Persian berries
Rhodamine B* Fluka Rhodamine B
Ellagic acid* Fluka Tannin products
Luteolin-3′,7-di-O-glucoside Extrasynthese Weld
Apigenin* Fluka Weld/Dyer’s broom
Apigenin-7-O-glucoside Extrasynthese Weld/Dyer’s broom
Chrysoeriol Extrasynthese Weld/Dyer’s broom
Luteolin* Sigma Weld/Dyer’s broom
Luteolin-4′-O-glucoside Extrasynthese Weld/Dyer’s broom
Luteolin-7-O-glucoside Extrasynthese Weld/Dyer’s broom
Butein Previously produced [8] Young fustic
Fisetin* Fluka Young fustic
Fustin Previously produced [8] Young fustic
Isoliquiritigenin Previously produced [8] Young fustic
Naringenin Previously produced [8] Young fustic
Sulfuretin* Wako Young fustic
Quercetin Previously produced [8] Young fustic/Persian berries
Table 1 Standard materials(colouring compounds) used foridentifications purposes. Com-pounds followed by an asterisk(*) are usually detected in largeamounts and are used as markersfor identification of thecorresponding dyes [1, 2]. Stand-ards were purchased from thesuppliers indicated; some com-pounds were previously pro-duced in pure form [8, 21]
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3067
components, haematein, and haematoxylin. ESI positiveionization-mode (ESI+) was selected to study fuchsin,because of its chlorinated components. Finally, APCInegative ionization-mode (APCI−) was used, probablyfor the first time, to study cochineal samples and tocompare the results with previously reported ESI− data[19, 20]. ESI conditions were: nitrogen gas temperature400 °C, capillary potential 4.5 kV, and cone potential30 V. In APCI mode the corresponding conditions were350 °C, 4 kV, and 30 V. MS conditions were selected afterpreliminary tests and were not optimized. Chromatographywas carried out using the same column and conditionsapplied for HPLC–DAD, except for the gradient elutionprogram and solvents, which had to be changed becauseTFA is an aggressive chemical for the MS detector. Forthis reason, an ammonium formate buffer solution (pH 3)was used with CH3CN [22].
Dyestuff extraction methods
The standard HCl extraction method [7] was used to treatthe historical samples. Samples were immersed in 400 μLH2O–MeOH–37% HCl 1:1:2 (v/v), at 100 °C, for 10 min.The solutions were then evaporated at 65 °C under gentlenitrogen flow and finally the dry residues were dissolved inDMSO. The use of DMSO (a relatively good solvent forindigoid dyes) in the last step was the only deviation fromthe original HCl method which suggested use of H2O–MeOH 1:1 (v/v).
Remaining parts of the historical samples, which werefound to contain weld and dyer’s broom according to theresults obtained with the HCl method, were treated with anewly devised dyestuff extraction method which includesthe use of TFA (2 mol L−1) instead of HCl (37%) [14]. Allother materials and conditions used in the milder TFAmethod, for example solvents and treatment temperatures,were the same as with the HCl method [14].
Historical samples
Microsamples (88) were extracted from 30 textiles (mon-astery of Simonos Petra, Mount Athos, Greece), dated fromthe 16th to early 20th century. Images of the objects areavailable on the web page of the Byz-tex-Athos projectwhich is funded by the Getty Foundation [23]. Most of thesamples were silk. Small samples (typically <1 mg) wereused as extracted from the historical objects and treatedwith the HCl method. Samples available in large quantities(>1 mg), were divided into two parts. One part from eachsample was analysed by HPLC using the standard HClmethod. Remaining parts, not treated with HCl, were eitherstored for future investigations (mordant identification) ortreated with the TFA extraction method and subjected to
HPLC. Therefore, the TFA method was applied only onsamples which were found to contain weld and dyer’sbroom according to the results collected with the HClmethod and were available in large quantities.
Results and discussion
Dyes identified in the historical samples
The HPLC results obtained from analysis of the historicalsamples using the HCl extraction method are summarizedin Table 2. Comparison of Table 1, which contains themarker compounds for dyestuff identification in unknownhistorical samples and other secondary components, andTable 2 leads to the conclusion that the dyes detected in theobjects from the monastery of Simonos Petra were: weld,dyer’s broom, cochineal, young fustic, madder, Persian(buckthorn) berries, indigo/woad, fuchsin, indigo carmine,and rhodamine B. Logwood and soluble redwood were alsoidentified in some samples, by use of marker compoundswhich are not included in Table 1 but are described later inthe text. Finally, the presence of tannin sources in somehistorical objects is suggested by the identification ofellagic acid (Table 2). These results (i.e. identified dyes)are discussed next.
Weld and dyer’s broom
As indicated in the Experimental section, all historicalsamples were treated by use of the HCl method. Remainingparts (not treated with HCl) of samples which were found tocontain components of flavonoid dyes, for example weld anddyer’s broom by the HCl extraction method, were treatedwith the milder TFA method to detect glycosides. The resultsobtained by use of the mild extraction method aresummarized in Table 3. This table shows that compoundswhich are extracted with HCl are also extracted with themilder TFA. Consequently, no information is missed if HClis replaced by TFA for sample treatment. Furthermore, theTFA method gives better results, because glycosides whichare actually present in the textiles are not decomposed duringextraction and are thus detected by HPLC.
This is further apparent from Fig. 1 which shows thechromatograms obtained from samples 9.4 (Fig. 1a) and14.1 (Fig. 1b) after treatment with TFA. Weld and dyer’sbroom glycosides are identified in Figs. 1a and 1b,respectively; these were not detected with the HCl method(not shown). It is noteworthy that the identifications ofweld glycosides shown in Fig. 1a are in agreement withresults that we published in a previous report [14]. Thesuccessful extraction of dyer’s broom glycosides with theTFA method (Fig. 1b) is reported for the first time.
3068 D. Mantzouris et al.
Table 2 Compounds identified in the historical samples using the HClmethod. Objects and samples are codified according to the archives ofthe monastery [23]. Samples were of silk nature except for 12.1, 16.1and 16.2 (cotton) and 5.2 (linen). Identified compounds are included in
Table 1 except for dcII, dcIV and dcVII which are components ofcochineal, type B and C compounds which are marker compounds forthe identification of soluble redwood and haematein derivative which isthe marker compound for the identification of logwood
Object Samplecode
Compounds identified
Code Type Date
2 Epitrachelion Early 17th century 2.2 Indigotin, indirubin
2.3 Indigotin, indirubin
2.4 Carminic acid, ellagic acid, dcIV, dcVII, alizarin, purpurin
3 Epitrachelion Mid. 16th century 3.1 Carminic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
3.2 Carminic acid, type B, type C, kermesic acid, flavokermesic acid
4 Epitrachelion 18th century 4.1 Fisetin, sulfuretin
4.3 Ellagic acid
4.4 Carminic acid, type B, ellagic acid, dcIV, type C, dcVII, kermesic acid,flavokermesic acid
4.5 dcII, carminic acid, type B, ellagic acid, dcIV, type C, dcVII, kermesic acid,flavokermesic acid
4.6a Fisetin, sulfuretin, luteolin
4.6b Luteolin, apigenin, chrysoeriol, indigotin, indirubin
5 Epitrachelion Mid. 16th century 5.1 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
5.2 Type B, type C
5.3 Indigotin, indirubin
6 Epitrachelion 17th century 6.1 Carminic acid, type B, type C
6.2 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
6.3 Fustin, fisetin, sulfuretin, quercetin, butein, isoliquiritigenin
8 Epimanikion 16th century 8.1 Carminic acid, type B, ellagic acid, dcIV, type C, dcVII, kermesic acid,flavokermesic acid, indigotin
8.2 Type B, type C
8.3 Indigotin, indirubin
8.4 Luteolin, apigenin, chrysoeriol, indigotin, indirubin
8.5 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid,alizarin, purpurin
9 Epimanikion 16th century 9.1 Carminic acid, type B, type C
9.2 Fustin, fisetin, sulfuretin, quercetin, butein, naringenin, kaempferol,isorhamnetin,rhamnetin, kaempferide, rhamnazin, emodin
9.3 Indigotin, indirubin
9.4 Luteolin, apigenin, chrysoeriol
10α Pair ofepimanikia
Second half of 18thcentury
10a.1 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
10b Second half of 18thcentury
10b.1 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
11 Podea 18th century 11.1 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
11.2 Carminic acid, fustin, ellagic acid, fisetin, sulfuretin
12 Mandyas Early 20th century 12.1 Rhodamine B
12.2 Pararosaniline, rosaniline, magenta II, new fuchsin
13 Podea 18th–19th century 13.1a Luteolin, apigenin, genistein, chrysoeriol
13.1b Luteolin, apigenin, genistein, chrysoeriol, indigotin, indirubin
13.2 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid,indigotin, indirubin
13.3 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
13.4 Luteolin, genistein
14 Podea 18th–19th century 14.1 Luteolin, apigenin, genistein, chrysoeriol, indigotin
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3069
Table 2 (continued)
Object Samplecode
Compounds identified
Code Type Date
14.2 Luteolin, apigenin, genistein, chrysoeriol, indigotin, indirubin
14.3 Indigotin, indirubin
14.4 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
14.5 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
14.6 dcII, carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid
14.7 Type B, type C
15 Belt Late 19th century 15.1 Type B, type C
15.2 Luteolin, apigenin, chrysoeriol
16 Phelonion Late 19th century 16.1 Pararosanilne, rosaniline, magenta II, new fuchsin
16.2 Quercetin, kaempferol
17 Phelonion Late 19th century 17.1 Haematein derivative, ellagic acid, fisetin, sulfuretin, luteolin
17.3 Luteolin, indigotin, indirubin
18 Epitaphios Late 19th century 18.1 Pararosanilne, rosaniline, magenta II, new fuchsin
18.3 Pararosanilne, rosaniline, magenta II, new fuchsin
18.4 Pararosanilne, rosaniline, magenta II, new fuchsin
19a Pair ofepimanikia
19th century 19a.1 Haematein derivative, luteolin, genistein, indigotin, indirubin
19b 19th century 19b.1 Haematein derivative, luteolin, genistein, indigotin, indirubin
22 Antimension 18th century 22.1 Indigotin
23 Podea 1627 23.1 Luteolin, apigenin, genistein, indigotin, indirubin
23.2 Indigotin, indirubin
23.3 Carminic acid
23.5 Carminic acid, sulfuretin, luteolin, apigenin, genistein, chrysoeriol, indigotin
23.6 Carminic acid, luteolin, apigenin, genistein, indigotin, indirubin
24 Polos Late 18th–early19th century
24.1 Type B, type C, fisetin, sulfuretin
24.2 Type B, type C, fisetin, sulfuretin, butein, naringenin, isoliquiritigenin
24.3 Type B, type C, fisetin, sulfuretin
24.4 Ellagic acid, luteolin, apigenin, chrysoeriol
26 Epitrachelion Late 18th–early19th century
26.1 dcII, carminic acid, type B, ellagic acid, dcIV, type C, dcVII, kermesic acid,flavokermesic acid
26.2 Indigotin, indirubin
26.3 Carminic acid, ellagic acid
27 Epitrachelion Late 18th–early19th century
27.1 Sulfuretin
27.3 Fisetin, sulfuretin
27.4 dcII, carminic acid, Type B, ellagic acid, dcIV, type C, dcVII, kermesic acid,flavokermesic acid
28 Sakkos 19th century 28.1 Luteolin, indigotin, indirubin
28.2 Haematein derivative, type B, type C
28.3 Haematein derivative, type B, type C, luteolin
29 Sakkos 19th century 29.1 Indigotin, indirubin
30 Sakkos 19th century 30.1 Indigo carmine, luteolin, apigenin, chrysoeriol
30.2 Luteolin, apigenin, chrysoeriol
30.3 Indigo carmine, luteolin, apigenin, chrysoeriol
30.4 Luteolin, apigenin, chrysoeriol
30.5 Luteolin
30.6 dcII, carminic acid
32 Polos Early 20th century 32.1 Carminic acid, pararosanilne, rosaniline, magenta II, new fuchsin
3070 D. Mantzouris et al.
This investigation has therefore shown that the TFAextraction method developed recently [14], is successful inextracting components of weld and dyer’s broom, includingboth aglycones and glycosides, from silk substrates.Simultaneously, dyestuff components such as carminicacid, sulfuretin, indigotin and indirubin which wereextracted with the HCl method, were also extracted withthe TFA method, as reported in Table 3.
Logwood and soluble redwood
Logwood is an interesting dye both for historians and chemists.Historically, it is important to note that logwood is native toAmerica [2] and therefore it was imported to the Europeancontinent after the discovery of the New World (1492), whichoccurred in the same historical period as the fall of theByzantine Empire (1453). Thus the identification of logwood
Table 2 (continued)
Object Samplecode
Compounds identified
Code Type Date
32.2 Carminic acid, ellagic acid, dcIV, dcVII, pararosanilne, rosaniline, magenta II,kermesic acid, flavokermesic acid, new fuchsin
32.3 Pararosanilne, rosaniline, luteolin, magenta II, apigenin, new fuchsin, chrysoeriol,indigotin, indirubin
33 Polos Early 20th century 33.1 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid, flavokermesic acid,alizarin,purpurin
33.2 Fisetin, sulfuretin
34 Polos Early 20th century 34.1 Carminic acid, ellagic acid, dcIV, dcVII, kermesic acid
34.2 Sulfuretin, luteolin, apigenin, chrysoeriol
Table 3 Compounds identified in selected historical samples usingthe TFA method [14]. Samples included in the table below were silkfibres dyed with weld or dyer’s broom and some other dyes.
Compounds followed by an asterisk (*) were also detected by use ofthe HCl method and are therefore reported in Table 2
Sample code Compounds identified
4.6b Luteolin*, apigenin*, chrysoeriol*, indigotin*, indirubin* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside, luteolin-4′-O-glucoside
8.4 Luteolin*, apigenin*, chrysoeriol*, indigotin*, indirubin* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside, luteolin-4′-O-glucoside
9.4 Luteolin*, apigenin*, chrysoeriol* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside, luteolin-4′-O-glucoside
13.1b Luteolin*, apigenin*, genistein*, chrysoeriol*, indigotin*, indirubin* and:
Luteolin-7-O-glucoside, genistin, genistein-4′,7-dimethyl ether
14.1 Luteolin*, apigenin*, genistein*, chrysoeriol*, indigotin* and:
Luteolin-8-C-glucoside, luteolin-7-O-glucoside, genistin, apigenin-7-O-glucoside, luteolin-4′-O-glucoside, genistein-4′,7-dimethylether
23.5 Carminic acid*, sulfuretin*, luteolin*, apigenin*, genistein*, chrysoeriol*, indigotin* and:
Luteolin-7-O-glucoside, genistin
23.6 Carminic acid*, Luteolin*, apigenin*, genistein*, indigotin*, indirubin* and:
Luteolin-7-O-glucoside, genistin
30.2 Luteolin*, apigenin*, chrysoeriol* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside, luteolin-4′-O-glucoside
30.4 Luteolin*, apigenin*, chrysoeriol* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside, luteolin-4′-O-glucoside
34.2 Sulfuretin*, luteolin*, apigenin*, chrysoeriol* and:
Luteolin-3′,7-di-O-glucoside, luteolin-7-O-glucoside, apigenin-7-O-glucoside
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3071
in textiles from the Simonos Petra monastery, suggests thatthese objects belong to the postbyzantine (and not to thebyzantine) period, a piece of information which is importantfor the historical documentation of the objects. Chemically, itis interesting to note that identification of logwood in ahistorical sample treated with the HCl method is based ondetection of a haematein derivative, symbolized as Hae′ [24].The structure of this marker compound was recentlyelucidated by NMR which showed that Hae′ is 3,4,10-trihydroxy-7H-indeno[2,1-c]chromen-9-one [17]. Hae′ is notdetected in the dyestuff source when this is simply dissolvedin a usual solvent and subjected to HPLC, without acidtreatment. As shown in Fig. 2a, logwood contains haematox-ylin, haematein (the oxidized form of haematoxylin), andseveral other compounds, some of which have been named astype G compounds [24]. The HPLC–DAD–MS data collectedfor standards of haematoxylin and haematein, after dissolutionin DMSO are shown in Figs. 2b and 2c, respectively. The twocompounds were identified in logwood (Fig. 2a). Several typeG compounds, detected in logwood (Fig. 2a) give acharacteristic DAD spectrum, shown in Fig. 2d [24].
Haematoxylin and haematein, however, were not detectedin the historical samples, which apparently had to be treatedwith acid (HCl) to release the dyes from the mordant metal. Inthe historical extracts another compound was detected, asshown in Fig. 3a. The same result was obtained for the rawlogwood source when this was treated with the HCl method
(Fig. 3b). Interestingly, haematein was not detected evenwhen pure haematein standard was analysed by HPLC, aftertreatment with HCl (Fig. 3c). Instead, the same result reportedfor historical samples (Fig. 3a) and logwood raw source(Fig. 3b) was recorded for haematein standard (Fig. 3c). Thecompound detected in Figs. 3a, 3b, and 3c is the Hae′compound, used as a marker for identification of logwood inhistorical extracts. The identification of Hae′ is based on:
1. the characteristic DAD spectrum, shown in Fig. 3d,which is in good agreement with a previously publishedreport [24]; and
2. the MS-ESI− spectrum obtained for this compound.
The latter is provided in Fig. 3e and shows that the major,deprotonated molecular ion recorded in the MS–ESI−
spectrum corresponds to m/z=281. Comparison of this resultwith the MS–ESI− spectrum of haematein, shown in Fig. 2c(m/z=299), suggests that Hae′ is a dehydrohaemateinproduct, which is in agreement with the structure of Hae′revealed previously by NMR [17].
Similar considerations apply to soluble redwood(Caesalpinia trees), as was suspected some years ago[17]. This dye contains brazilin and brazilein (the oxidizedform of brazilin). However, the two compounds are notdetected when soluble redwood is treated with HCl [20,24]; instead, another compound, known as Bra′ (or type Bcompound) is recorded by HPLC [24]. We have recently
12 14 16 18 20 220
20
40
60
80
100
Historical sample 9.4a
b
43
1
2
Rel
ativ
e A
bsor
banc
e
5
76
12 14 16 18 20 22 24 26 28 300
20
40
60
80
100
Historical sample 14.1Indigotin
8
7
6
12,6 14,0 27 30
91
2
3
45R
elat
ive
Abs
orba
nce
Time (min)
Fig. 1 Chromatograms obtained from historical samples 9.4 (a) and14.1 (b), after treatment with the TFA method. Compounds (compo-nents of weld) detected in (a) were: luteolin-3′,7-di-O-glucoside (1),luteolin-7-O-glucoside (2), apigenin-7-O-glucoside (3), luteolin-4′-O-glucoside (4), luteolin (5), apigenin (6), and chrysoeriol (7). Thecompounds (components of dyer’s broom) detected in (b) were:
luteolin-8-C-glucoside (1), luteolin-7-O-glucoside (2), genistin (3),apigenin-7-O-glucoside (4), luteolin-4′-O-glucoside (5), luteolin (6),apigenin and genistein, coeluting (7), chrysoeriol (8), and genistein-4′,7-dimethyl ether (9). Indigotin, reported in (b) was apparentlyoriginated from an indigoid dyeing source. Chromatograms wereacquired at 275 nm for sample 9.4 (a) and 340 nm for sample 14.1 (b)
3072 D. Mantzouris et al.
performed detailed LC–MS-ESI− studies and shown thatthe deprotonated ion of Bra′ corresponds to m/z=265suggesting thus that this is probably a dehydrobrazileinproduct, produced from brazilein (m/z=283) during acid(HCl) treatment [20]. This seems to be similar to thetransformation of haematein to Hae′ (dehydrohaematein),discussed in the previous paragraph. In the same way that
Hae′ is used as a marker compound for logwood, the Bra′compound is used as a marker for the identification ofsoluble redwood in historical samples, treated with HCl [7,20, 24]. The results in Table 2 indicate that the type B (Bra′)compound was present in some samples admixed with thetype C compound, which is also detected in hydrolyzedextracts of soluble redwood [7, 20, 24].
6 8 10 12 14 16 18 200
20
40
60
80
100
m/z
Time (min)
6 8 10 12 14 16 18 200
20
40
60
80
100
Haematein in DMSO
6 8 10 12 14 16 18 200
20
40
60
80
100
Haematoxylin in DMSO216
GGGGG
Haematein
G
Rel
ativ
eA
bsor
banc
eR
elat
ive
Abs
orba
nce
Rel
ativ
e A
bsor
banc
e
Haematoxylin
G
301
Logwood in DMSOa
b
c
d
0
50
100
b. Haematoxylin in DMSO
299
289
m/z
Rel
ativ
e A
bsor
banc
e
Wavelenght (nm)
0
50
100
Rel
ativ
e A
bund
ance
0
50
100445
276R
elat
ive
Abs
orba
nce
Wavelenght (nm)
200 300 400 500 600250 300 350 400
200 300 400 500 600250 300 350 4000
50
100
Rel
ativ
e A
bund
ance
200 300 400 500 6000
20
40
60
80
100
291
253
Rel
ativ
e A
bsor
banc
e
Wavelength (nm)
225
DAD spectrum of type G compounds
Fig. 2 Chromatograms obtainedat 275 nm from (a) logwood rawsource, (b) haematoxylin stan-dard, and (c) haematein standard.Samples were dissolved inDMSO and subjected to HPLC.DAD and MS-ESI− spectra ofhaematoxylin and haematein areshown in (b) and (c), respective-ly; the deprotonated molecularions are present in the MS-ESI−
spectra. DAD spectra of type Gcompounds detected in logwoodare shown in (d) [24]
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3073
Cochineal
The exact structures of the dcII, dcIV, and dcVII compo-nents of cochineal are not known. These compounds,however, are often detected in historical samples (inmixture with carminic acid, of course, the main componentof the red dye), as can be concluded from the results listedin Table 2. A previous NMR study has shown that dcII ismost probably the 7-C-glycoside of flavokermesic acid anddcIV and dcVII must be isomeric with carminic acid [19].
Furthermore, LC–MS-ESI− results have shown that thedeprotonated molecular ion of dcII corresponds to m/z=475and the deprotonated molecular ions of dcIV and dcVII tom/z=491 [19, 20]. We continue our work with the minorcomponents of cochineal [20] by applying LC–MS-APCI− to analyse samples of Dactylopius coccus Costa.To the best of our knowledge this is the first attempt toanalyse cochineal using MS-APCI. The results are shownin Fig. 4 and are in excellent agreement with previous LC–MS-ESI− findings [19, 20].
Time (min)
0
20
40
60
80
100
Rel
ativ
e A
bsor
banc
e
Hae'Historical sample treated with HCl
0
20
40
60
80
100
Wavelength (nm)
0
20
40
60
80
100
Logwood treated with HClR
elat
ive
Abs
orba
nce
6 8 10 12 14 16
0
20
40
60
80
100
Haematein treated with HCl
Rel
ativ
e A
bsor
banc
e
6 8 10 12 14 16
6 8 10 12 14 16
250 300 350 400200 300 400 500 6000
20
40
60
80
100
MS spectrum of Hae'
Rel
ativ
e A
bund
ance
281453
331
275
Rel
ativ
e A
bsor
banc
e
Wavelength (nm)
221
DAD spectrum of Hae'
a
b
c
d e
Fig. 3 Chromatograms obtainedat 450 nm from (a) historicalsample 19b.1, (b) haemateinstandard, and (c) logwood. Sam-ples were treated with the stan-dard HCl method and subjectedto HPLC. The haematein deriva-tive, Hae′, the marker compoundfor logwood identification inhistorical samples, was detectedin all three samples. DAD andMS-ESI− spectra of Hae′ areshown in (d) and (e) respectively.Other compounds detected in thehistorical sample (Table 2) arenot eluted within the time rangeselected in (a) to clearly demon-strate identification of Hae′
3074 D. Mantzouris et al.
Young fustic
Identification of young fustic (Cotinus coggygria Scop.) inhistorical samples is usually based on detection ofsulfuretin and fisetin [2, 7]. As indicated in Table 2, thesetwo compounds were detected in samples extracted fromobjects 4, 6, 9, 11, 17, 23, 24, 27, 33, and 34. However,young fustic contains several other components [8, 25]some of which were found in historical samples 6.3 (fustin,quercetin, butein, isoliquiritigenin), 9.2 (fustin, quercetin,butein, naringenin), and 24.2 (butein, naringenin, isoliquir-itigenin). Butein, fustin, isoliquiritigenin, naringenin, andquercetin have previously been isolated from the heartwoodof Cotinus coggygria Scop. [8]. In our study, the isolatedcompounds were used as standard materials for analysis ofthe historical samples.
It is worthy of note that the identification of young fusticin a relatively large number of objects (10 from 30 total)supports our previous suggestion that the yellow dye mighthad an important role in the dyeing industry of ecclesias-tical textiles [7, 8].
Madder
As indicated in Table 2, alizarin and purpurin were detectedin three samples—2.4, 8.5, and 33.1. These two antraqui-none compounds are characteristic of the Rubiaceae familyof plants. In the Mediterranean area there are two specieswhich are dominant and have been used for dyeing sinceantiquity—Rubia tinctorum L. and Rubia peregrina L. (wildmadder) [1]. Chromatograms obtained from the historical
samples showed that alizarin and purpurin were present incomparable quantities. Furthermore, rubiadin, which isusually present in Rubia peregrina L. in high amounts[26], was not detected in the historical samples. These resultssuggest that the madder source used to dye the historicalsamples was probably Rubia tinctorum L. [20, 26].
Persian berries
“Persian berries” or “buckthorn berries” are general termsthat include several plant species belonging to the genusRhamnus, family Rhamnaceae [1, 2]. The marker com-pound for identification of Persian berries is rhamnetin.This was detected in sample 9.2 (Table 2), with othercomponents of Persian berries which, however, have notbeen investigated in detail [27].
The chromatogram recorded for sample 9.2 is shown inFig. 5a. For comparison, the chromatogram obtained fromthe raw material of buckthorn berries is shown in Fig. 5b.Both sample 9.2 and the reference raw dyestuff source weretreated with the HCl method, before HPLC analysis.Identification of several compounds detected in the chro-matograms of Figs. 5a (historical sample) and 5b (rawsource of buckthorn berries) was achieved, by comparingthem with Fig. 5c, which shows the chromatogram obtainedfrom a standard mixture that contained quercetin (1),kaempferol (2), isorhamnetin (3), rhamnetin (4), kaempferide(5), rhamnazin (6), and emodin (7).
Compounds 1, 2, 3, 4, 6, and 7 have previously beendetected in a sample of Persian berries [27] which,however, was not obtained from Kremer Pigmente, the
020406080
100
m/z
020406080
100
m/z
02040
6080
100
dcVIIdcIVdcII491491
Rel
ativ
e A
bund
ance
m/z
475
300 400 500300 400 500300 400 500
12 14 16 180
10
20
30
dcVIIdcIV
dcII
Rel
ativ
e In
tens
ity
Time (min)
Carminic acidFig. 4 LC–MS-APCI− analysisof cochineal (Dactylopiuscoccus Costa)
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3075
supplier of the sample analysed in Fig. 5b. In this previousinvestigation compound 5, kaempferide (3,5,7-trihydroxy-4′-methoxyflavone), was not detected, although it is contained inPersian berries [2]. Instead, rhamonocitrin (3,4′-5-trihydroxy-7-methoxyflavone) was identified by LC–MS-ESI without,however, the corresponding standard being available [27].Similarly, rhamonocitrin was not available in pure form in ourstudy too. Apparently, the slight disagreement between ouridentifications and the results reported previously [27] mightoriginate from the different Rhanmus samples analysed in thetwo investigations. It must be noted, however, that rhamo-nocitrin and kaempferide have similar structures, shouldhave similar UV–visible spectra, and their separation byHPLC might be difficult.
Indigo and woad
Indigotin, the marker compound for both indigo (Indigoferaspecies and others) and woad (Isatis tinctoria L.), wasidentified in several historical samples, as indicated by the
results listed in Table 2. In some samples, indirubin, whichis also contained in indigoid dyes but in smaller quantitiesthan indigotin [28], was also detected. Both indigo andwoad have been known since antiquity. The two dyescannot be separated by HPLC [2] and therefore it was notpossible to identify the specific dyestuff source used in thetextiles from Mount Athos.
Early/semi-synthetic dyes
Natural dyes, for example those discussed so far, werewidely used until the end of the 19th century, whensynthetic dyes were introduced and gradually replacednatural colorants almost entirely. The term “early syn-thetic dyes” is usually used to describe the syntheticorganic colorants developed in the 2nd half of the 19thcentury [29]. These materials have not been studied indetail [29].
As indicated by the results listed in Table 2, fuchsin andrhodamine B (early synthetic dyes), were detected in
0
20
40
60
80
100
0
20
40
60
80
100
Mixture of standards
Time (min)
7
6
54
3
21
NA
BUSU
Rel
ativ
e A
bsor
banc
eR
elat
ive
Abs
orba
nce
7
6
5
4
321
Historical sample
18 20 22 24 26 28
18 20 22 24 26 28
18 20 22 24 26 280
20
40
60
80
100
32 57
6
4
Rel
ativ
eA
bsor
banc
e
1
Buckthorn berries
a
b
c
Fig. 5 Chromatograms obtainedat 254 nm from (a) historicalsample 9.2, (b) raw material ofbuckthorn berries, and (c) amixture of quercetin (1), kaemp-ferol (2), isorhamnetin (3),rhamnetin (4), kaempferide (5),rhamnazin (6), and emodin (7)standards. Historical sample andraw buckthorn berries were trea-ted with the HCl method beforeHPLC analysis. Other com-pounds, not related to buckthornberries, detected in the historicalsample were: sulfuretin (SU),butein (BU), and naringenin(NA). These three compounds arepresent in young fustic. Othercomponents of young fustic, forexample fustin and fisetin, werealso detected in the historicalsample, as indicated in Table 2.However, they corresponded toretention times <18 min and weretherefore not present in (a)
3076 D. Mantzouris et al.
samples extracted from objects 12, 16, 18, and 32 which aredated late 19th or early 20th century. This is reasonable,because fuchsin and rhodamine B were developed in 1856and 1887, respectively [29]. In samples 30.1 and 30.3 (both
extracted from object 30, dated 19th century) indigocarmine was identified. This is a semi-synthetic dyeproduced in 1740 by treating natural indigo with sulfuricacid [2].
020406080
100
020406080
100
020406080
1004
1
Pararosaniline
4
32
Rel
ativ
e A
bsor
banc
e1
Historical sample
New Fuchsin
432
Rel
ativ
e A
bsor
banc
e
Time (min)
3
2
Rel
ativ
e A
bsor
banc
e
0
20
40
60
80
100
CH3
316
Rel
ativ
e A
bund
ance
m/z
302
0
20
40
60
80
100
H3C
H3C
H3C
H3C
H3C
Rel
ativ
e A
bund
ance
m/z
0
20
40
60
80
100
4: New Fuchsin3: Magenta II
2: Rosaniline
330
Rel
ativ
e A
bund
ance
m/z
288
MS spectra of compounds detected in (a)
1: Pararosaniline
14 16 18 20 22 24 26
14 16 18 20 22 24 26
14 16 18 20 22 24 26
250 300 350 400
250 300 350 400
250 300 350 400
250 300 350 4000
20
40
60
80
100
H2NH2N
H2N H2N
NH2
NH2 NH2
NH2
NH2+
NH2+Cl
-
NH2+Cl
-
NH2+Cl
-
Cl-
Rel
ativ
e A
bund
ance
m/z
a
b
c
d
Fig. 6 Chromatograms obtainedat 540 nm from (a) historicalsample 16.1, (b) pararosanilinestandard (1), and (c) new fuchsinstandard (4). (d) MS-ESI+ spectraof compounds 1, 2, 3, and 4detected in the historical sampleare shown; structures of theidentified compounds areincluded
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3077
Indigo carmine and rhodamine B were available in thepure form (Table 1) and were, therefore, easily identified inthe historical samples. The identification of fuchsin com-ponents in sample 16.1 is illustrated in Fig. 6. Figure 6ashows the chromatogram obtained from the historicalsample. Figures 6b and 6c show the chromatogramsobtained from pararosaniline (CI 42500) and new fuchsin(CI 42520) standards, respectively. Identifications of thesetwo compounds (1 and 4) in the historical sample (Fig. 6a)was confirmed by the mass spectra, shown in Fig. 6d.When LC–MS was operated in positive ESI mode (LC–MS-ESI+) the protonated cations m/z=288 and m/z=330were recorded for compounds 1 (identified as pararosani-line with molar mass 323.5) and 4 (identified as newfuchsin with molar mass 365.5), respectively. According tothe HPLC results in Figs. 6b and 6c, neither standard ispure—they contain other compounds as impurities. LC–MS-ESI+ analysis of the historical sample furnished themass spectra of compounds 2 and 3 (Fig. 6d). Major ionproducts correspond to m/z=302 and m/z=330 for com-pounds 2 and 3, respectively. Considering that the other twocompounds usually detected in fuchsin samples are rosani-line (molar mass 337.5) and magenta II (molar mass 351.5)[18] it can be concluded that ion products 2 and 3correspond to the protonated cations of rosaniline andmagenta II, respectively.
Tannin sources
Identification of ellagic acid in many historical samples(Table 2) is indicative of the use of tannin products duringdyeing. Ellagic acid is contained in several plants andcannot be used as a marker for identification of a particularnatural source.
Conclusions
It has been shown that the TFA method, developed recently[14], enables successful extraction of the components ofweld and dyer’s broom, including aglycones (luteolin,apigenin, chrysoeriol, genistein) and glycosides (apigenin-7-O-glucoside, genistein-4′,7-dimethyl ether, genistin,luteolin-3′,7-di-O-glucoside, luteolin-4′-O-glucoside,luteolin-7-O-glucoside, luteolin-8-C-glucoside) from silksubstrates. This was shown by use of samples taken fromimportant historical textiles from the monastery of SimonosPetra, Mount Athos. Other dyes identified in the historicalextracts by HPLC–DAD–MS were logwood, solubleredwood, cochineal, young fustic, Persian (buckthorn)berries, madder (most probably Rubia tinctorum L. accord-ing to HPLC results), indigo/woad, fuchsin, indigo carmine,and rhodamine B.
LC–MS was used to provide useful information regard-ing some of the dyes detected in the historical samples. Inparticular, LC–MS-ESI− studies showed that the markercompound for logwood identification (Hae′), after treatmentwith HCl, should be a dehydrohaematein product [17]. LC–MS-APCI− was used, probably for the first time, to analysecochineal. The MS-APCI− spectra acquired for the com-pounds dcII, dcIV, and dcVII showed that the major ionproducts (most probably the deprotonated molecular ions)corresponded to m/z=475 (for dcII) and m/z=491 (for bothdcIV and dcVII). These results are in agreement withprevious NMR and LC–MS-ESI studies [19, 20]. Separa-tion and identification of four fuchsin homologues, namelypararosaniline, rosaniline, magenta II, and new fuchsinwere achieved by LC–MS-ESI+.
Detailed HPLC–DAD studies led to identification ofemodin, isorhamnetin, kaempferide, kaempferol, quercetin,rhamnazin, and rhamnetin in Persian berries (Rhamnustrees). Identification of young fustic in the historicalsamples was achieved by detection of sulfuretin andfisetin; secondary components, for example butein,fustin, isoliquiritigenin, naringenin, and quercetin werealso detected in some samples.
Acknowledgements The authors would like to thank Dr C. Karydisfor providing the historical samples. Support by the Getty Foundation(USA), through the project Byz-tex-Athos (www.byztexathos.org) isgratefully acknowledged.
References
1. Cardon D (2007) Natural dyes—sources, tradition, technologyand science. Archetype, London
2. Hofenk-de Graaff JH (2004) The colourful past: origins,chemistry and identification of natural dyestuffs. Archetype,London
3. Karapanagiotis I (2006) Am Lab 38:36–404. Wilkins J, Hill S (2006) Food in the ancient world. Blackwell
publishing, Malden, MA5. Osbourn AE, Lanzotti V (2009) Plant-derived natural products:
synthesis, function and applications. Springer, Berlin HeidelbergNew York
6. Papageorgiou VP, Assimopoulou AN, Couladouros EA, Hep-worth D, Nicolaou KC (1999) Angew Chem Int Ed 38:270–300
7. Karapanagiotis I, Lakka A, Valianou L, Chryssoulakis Y (2008)Microchim Acta 160:477–483
8. Valianou L, Stathopoulou K, Karapanagiotis I, Magiatis P,Pavlidou E, Skaltsounis A-L, Chryssoulakis Y (2009) AnalBioanal Chem 394:871–882
9. Wouters J (1985) Stud Conserv 30:119–12810. Surowiec I, Nowik W, Trojanowicz M (in press) Dyes History
Archaeol 2211. Zhang X, Laursen RA (2005) Anal Chem 77:2022–202512. Sanyova J, Reisse J (2006) J Cult Herit 7:229–23513. Marques R, Sousa MM, Oliveira MC, Melo MJ (2009) J
Chromatogr A 1216:1395–1402
3078 D. Mantzouris et al.
14. Valianou L, Karapanagiotis I, Chryssoulakis Y (2009) AnalBioanal Chem 395:2175–2189
15. Rosenberg E (2008) Anal Bioanal Chem 391:33–5716. Petroviciu I, Albu F,Medvedovici A (2010)Microchem J 95:247–25417. Hulme AN, McNab H, Peggie DA, Quye A, Vanden Berghe I,
Wouters J (2005) ICOM 14th Triennial Meeting The Hague: 12–16 September 2005: Preprints, vol II. Earthscan Publications Ltd.,London, U.K., p 783
18. Lyon H, Schulte E, de Leenheer A, Lewis S, Friemert V, Struck C,Gadsdon D, Allison R, Brunk U, van Liederkerke B, HasselagerE, Horobin R, Husain O, Wittekind D, Zschoch H (1992)Histochem J 24:236–237
19. Peggie DA, Hulme AN, McNab H, Quye A (2008) MicrochimActa 162:371–380
20. Karapanagiotis I, Minopoulou E, Valianou L, Sister Daniilia,Chryssoulakis Y (2009) Anal Chim Acta 647:231–242
21. Valianou L, Wei S, Mubarak MS, Farmakalidou L, KarapanagiotisI, Rosenberg E, Stassinopoulos S (2010) J Archaeolog Sci, Inpress doi:10.1016/j.jas.2010.08.025
22. Assimopoulou AN, Karapanagiotis I, Vasiliou A, Kokkini S,Papageorgiou VP (2006) Biomed Chromatogr 20:1359–1374
23. www.byztexathos.org24. Nowik W (2001) Dyes Hist Archaeol 16/17:129–14425. Antal DS, Schwaiger S, Ellmerer-Müller EP, Stuppner H (2010)
Planta Med 76:1–826. Wouters J (2001) Dyes History Archaeol 16/17:145–15727. Ferreira E, Quye A, McNab H, Hulme A (2003) Dyes History
Archaeol 19:19–2328. Puchalska M, Połeć-Pawlak K, Zadrożna I, Hryszko H, Jarosz M
(2004) J Mass Spectrom 39:1441–144929. Van Bommel MR, Vanden Berghe I, Wallert AM, Boitelle R,
Wouters J (2007) J Chromatogr A 1157:260–272
HPLC–DAD–MS analysis of dyes identified in textiles from Mount Athos 3079