J. Sep. Sci.

2014, 37, 3393–3410 3393

Volodymyr Pauk Review Article

Petr Barták
Karel Lemr
Characterization of natural organic colorants
Faculty of Science, Department
of Analytical Chemistry, Regional in historical and art objects by
Centre of Advanced Technologies
and Materials, Palacký University high-performance liquid chromatography
in Olomouc, Olomouc, Czech
Republic High-performance liquid chromatography plays an important role in analysis of histori-
cal organic colorants. A number of papers have been published in this field over the last
Received June 15, 2014 30 years. Classification of the most commonly used natural dyes and an overview of high-
Revised August 24, 2014 performance liquid chromatography methods with main focus on recent works (2008 to the
Accepted September 9, 2014 beginning of 2014) are provided. The review deals with an entire analytical protocol cov-
ering sample preparation, chromatographic separation, and suitable detection (UV/visible
and fluorescent spectroscopy and mass spectrometric techniques). High-performance liq-
uid chromatography has been successfully used in the complete characterization of some
organic dyestuffs present in historical and art objects. The possibilities and difficulties for
identification of natural sources of historical colorants are also discussed.

Keywords: Anthraquinoids / Flavonoids / Indigoids / Lake pigments / Tannins

DOI 10.1002/jssc.201400650

 Additional supporting information may be found in the online version of this article
at the publisher’s web-site

1 Introduction certain dyes were used in different times or geographic

locations, their identification assists in dating or localizing
The history of the usage of natural colorants in art dates of the origin of historical artifacts. Minor components or
back to the Stone Age [1]. In prehistoric times, plant dyes degradation products of dyes [5–7] can confirm or disprove
were mainly used for body painting and hair dyeing, but the authenticity of the evaluated object of interest.
as recent findings show [2], colors had already been applied
to fibers as long as 30 000 years ago. Natural colorants
also functioned as pigments for murals, frescos, paintings, 1.1 Methods of dye analysis
drawings, and for decoration of other materials and objects
such as wood, leather, sculptures, vessels, and even for The available amount of a sample from an art object is of-
ritual treatment [3]. Up to the 19th century, a “chemical ten extremely small, but information about main and mi-
palette” of organic dyes was formed by natural species and nor components, corresponding degradation products, as
remained nearly unchanged. In 1856, Perkins synthesized well as a quantitative ratio of substances can be required
mauveine [4], and a new era of synthetic organic colorants to reach a suitable conclusion. This issue can be solved by the
was initiated. Afterwards, natural colorants were almost use of modern and highly sensitive and selective analytical
completely replaced by synthetic ones. methods.
The identification of historical dyestuffs provides cru- There are many analytical tools potentially suitable for
cial information for conservators and restorers to choose analysis of art objects [8]. Techniques such as UV/Vis [9, 10],
appropriate technique for preservation and restoration of art IR [9], Raman [3, 11–14], fluorescence [15], X-ray fluores-
objects. The analytical data give useful hints on the dyeing cence [9, 13], or NMR [13] spectroscopy are virtually non-
technique of a craftsman or an artist but also on household destructive, require no or little sample preparation, and most
activity and the culture of particular historical period. Since of them can be performed in situ. Nowadays, multitech-
nique spectroscopic protocols are preferred [16, 17]. Since
Correspondence: Karel Lemr, Faculty of Science, Department of historical dyes are often complex mixtures of organic com-
Analytical Chemistry, Regional Centre of Advanced Technologies pounds, the spectroscopic techniques do have some lim-
and Materials, Palacký University in Olomouc, 17. listopadu 12, itations. UV/Vis spectroscopy is not as selective as other
771 46 Olomouc, Czech Republic techniques and gives limited structural information. Fluores-
E-mail: Karel.Lemr@upol.cz
cence or spectral overlap of binders, pigments, and other com-
Fax: +420585634433
ponents can interfere in Raman and IR spectra. Raman spec-
Abbreviations: MeOH, methanol; MSA, methanesulfonic troscopy can fail to discriminate dyes of very similar chem-
acid; PCA, principal component analysis ical structure. Fluorescence spectroscopy is useful only for


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3394 V. Pauk et al. J. Sep. Sci. 2014, 37, 3393–3410

fluorescent active compounds. X-ray fluorescence requires A dye itself can be a complex mixture of substances. Com-
the presence of metals in a sample and is unsuitable for the pounds containing chromophore groups are called coloring
identification of organic compounds. NMR spectroscopy is an matters or coloring principles. Some dyes can contain dif-
excellent tool but can be insufficient for very small amounts ferent amounts of the same coloring matters. For example,
of sample and is unsuitable for analysis of trace components. the main component of kermes dye is kermesic acid, which
On the contrary, the hyphenation of separation and MS (CE– is also present in cochineal (see Supporting Information
MS, GC–MS, and HPLC–MS) is sufficiently sensitive and Table S1). Moreover, the ratio of coloring matters in the same
selective, provides structural information, and allows identi- dyes derived from different species can vary greatly (Sup-
fication of unknown compounds; however, sampling of an porting Information Table S1). A dyestuff can also contain
object must be performed [18]. Although CE has been suc- minor components characteristic of the particular species,
cessfully used to identify natural dyes in historical objects, its like in case of sea snails. Some of these molluscs (Bolinus
detection limits (even using MS detection) were substantially brandaris L., Stramonita haemastoma L.) produce dyes com-
higher in comparison with HPLC with diode array detection posed entirely of brominated indigo-derivatives, and the oth-
(DAD) and HPLC–MS [19, 20]. GC–MS has not been widely ers (Hexaplex trunculus L.) with some amount of nonbromi-
used in the study of natural dyes, due to the relatively high nated indigo [30, 31] (see Supporting Information Table S2).
molecular mass and polarity of the analytes. Derivatization Dyes can also contain degradation products and mineral im-
of the target analytes is a compulsory requirement for GC– purities useful for identification purposes. Colorants can be
MS [7, 21–24]. classified according to their source, application method, or by
The vast majority of the recent works dealing with iden- color and corresponding chemical structure.
tification of colorants has used HPLC. Starting from 1975 Most natural colorants are of plant origin but animals are
when HPLC provided data for investigation of binding media used for their production too. They are traditionally extracted
in paintings [25] and, later, for dyestuffs [26], it is the most pre- from roots (madder, turmeric, bedstraw), berries (buckthorn),
ferred method for analysis of natural dyes. In 2008, Surowiec flower heads (saffron), bark (oak), leaves (indigo, henna), and
published a paper delineating high-performance separation even from lichens and mushrooms [1]. Besides being avail-
techniques used for analysis of archaeological objects [18]. able, dyes also have to be sufficiently stable. For example,
In that work, HPLC of natural dyes was briefly mentioned. chlorophyll (green parts of plants), carotene (carrot), and beta-
In the same year, Rosenberg reviewed LC–MS methods for nine (red beet) degrade too fast, especially in light. Sometimes
the characterization of historical organic dyestuffs [27]. He the consumption of sources is enormous. The production of
focused more on MS detection rather than chromatographic only 1.4 g of Tyrian purple dye requires 12 000 snails [30].
separation. In 2009, Degano published an extensive review Moreover, the dye itself is not present in the shells. It is de-
on analytical methods used for the analysis of artworks and rived from precursor compounds, also called chromogens,
historical textiles [28].Various approaches using spectromet- present in hypobranchial gland secretions of molluscs by
ric techniques, direct mass spectrometric methods, LC, and means of subsequent fermentation, reduction, and photore-
CE were described. actions [32, 33]. Harvesting the red dyes of cochineal requires
This paper primarily surveys recent analytical protocols planting cacti, the manual collection of insects, and their sub-
for the identification of historical organic dyes by HPLC. De- sequent drying and extraction [34]. The most important his-
liberations on the operational and instrumental features in- torical dyestuffs, their chemical constituents, and application
dicate broad spectrum of information that can be obtained methods are listed in Supporting Information Table S1.
from HPLC data. Three types of dyes are distinguished in historical textile
dyeing: direct, mordant, and vat. Direct dyeing uses soluble
dyes with some affinity to the fibers. It does not require any
2 Classification of organic colorants mordants or metallic salts for fixing. As a result, the products
exhibit poor color-fastness to washing due to weak chemical
Colorants are commonly divided into two groups: dyes and interactions.
pigments [4]. Dyes are generally organic substances at least Most of the natural dyes bind to fibers through the appli-
temporary soluble in the vehicle. Pigments are usually in- cation of a mordant. They become more resistant to light and
organic compounds (metal oxides, sulfides, and other min- washing. Acid mordants have been obtained from tannin-rich
erals) insoluble in the paint medium. Nevertheless, organic sources such as oak gallnuts or bark. Basic mordants are salts
colorants can be used as pigments like indigo in oil paint- of different metals, such as aluminum, iron, copper, tin, or
ings [29] and drawings [1], or cochineal that after treatment zinc, that form coordination complexes between molecules
with mordant, e.g. alum (KAl(SO4 )2 ), and addition of metal of dyes and textile substrate such as wool or silk (Fig. 1). The
salt, such as chalk (CaCO3 ), produces an insoluble Ca–Al– color of the resulting dye is dependent on the mordant used.
colorant complex called the carmine lake pigment [1]. A dye Iron salts produce darker shades while alum intensifies color
penetrates into the substrate, usually fibers of textile, whereas brilliance [1].
a pigment is bonded to the surface of the substrate and is usu- Vat dyes are insoluble in common solvents. They are re-
ally applied as a suspension in a suitable binding medium duced in alkaline medium to form a colorless water-soluble
(oils, proteins, and resins). substance called leuco dye (Fig. 2). Fibers or textiles are


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410 Liquid Chromatography 3395

Extracts are separated on a suitable HPLC column; gradient

elution is usually applied. Various additives to the mobile
phase improve separation. Chromatography is followed by
diode array, mass spectrometer, or other common detection
systems. Methods capable of detection and quantitation
of a number of colorants belonging to different chemical
groups were developed to enable tracing of their natural
sources [37–48]. Essential information on many of the
methods devoted to analysis of organic colorants in historical
and art objects described in the literature is summarized in
Table 1. Some comments on the methods and sample prepa-
ration useful for their adoption are presented in Supporting
Information Tables S3–S5.

3.1 Sample preparation

3.1.1 Sampling

Since the physical integrity of art and historical artifacts has to

be maintained, the amount of sample available is often highly
Figure 1. Formation of wool–mordant–dye complex.
limited. A 0.3 to 1.0 cm long (0.5–1 mg weight) piece of textile
fiber of a certain color is usually taken [6, 40–44, 49–52] and
immersed into a prepared leuco-bath. After oxidation, the dye coloring components are then extracted. In a recent work, a
restores its original color and properties. There are only three polychromic yarn (<50 ␮g) was dissected into single fibers
natural vat dyes: indigo, woad, and Tyrian purple. These dyes (<5 ␮g) under a microscope and each distinct color set was
are problematic for HPLC due to their negligible solubility in analyzed [53].
common mobile phases. A fiber sample can be frozen in liquid nitrogen, ground,
The fundamental property of a dye or a pigment is its and extracted [54]. No additional explanation is given in the pa-
color. Often dyes of a similar color have very similar chem- per but one can expect more complete extraction with this pro-
ical structure (see Supporting Information Table S1 and cedure. For paintings and archaeological artifacts, scrapes of
Figs. S1–S12). Sampling and analysis of historical col- pigments from the paint layers weighing around 0.1–0.5 mg
orants is usually conducted according to their color. Natu- were typically obtained [40, 46, 55]. Recently, the successful
ral dyes are classified into four color groups: blue/purple, analysis of scrapes (<0.1 mg) from the 16th century drawings
red/pink/orange, yellow, and black/brown. There are very few has been reported [56]. However, for sampling of historical
green dyes, and in most cases they are obtained by mixing maps and drawings, a safer procedure was introduced [57]. A
blue and yellow. True purple was too expensive and was im- brush imbibed with the proper solution (0.1 M SDS) was ap-
itated by double-dyeing with blue (indigo) and red (madder). plied directly onto the map. Soluble compounds and particles
The resultant color was called “poor people’s purple” [35]. were transferred to the solvent. The method is advantageous
Sometimes this name refers to purple orchil dye [36]. There- over the use of scalpel, which damages the object, or the drop
fore, what we actually see can be a mixture of a few coloring of a solvent or blotting paper, which can cause stains.
species.
3.1.2 Extraction

3 HPLC protocol Various approaches and their combinations were more or less
successfully utilized for extraction of natural dyestuff com-
During almost three decades of HPLC practice in the field ponents from art and historical samples: HCl (hydrochloric
of natural organic colorants, the following general analytical acid) hydrolysis [5, 6, 10, 37, 40, 41, 43–45, 48, 49, 51, 52, 58–71],
protocol was established. The first step is extraction of EDTA (ethylenediaminetetraacetic acid) method [37, 41, 49,
dyestuff components from the sample of textile or paint. 68], formic acid extraction [37, 39–41, 48, 49, 58, 68, 72], TFA

Figure 2. Reduction of vat dye indigo to

leucoindigo.


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

3396

Table 1. List of HPLC methods

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
V. Pauk et al.

References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

1985 [26] Anthraquinoids (11) 3 M HCl, 100ºC Nucleosil C18; 150 × 4.1 A: water 55; 1.5 UV/Vis
B: MeOH
C: 1% form. acid
1989 [66] Anthraquinoids (insect Water/MeOH/37% HCl (1:1:2), - - - UV/Vis
reds) (11) hot; rediss. water/MeOH (1:1)
1989 [92] Anthraquinoids (madder) (1) MeOH/20% form. acid; rediss. Cosmosil ODS 5C18; 150 × MeOH/0.1 M amm. 24; 1 UV/Vis
20% DMF/MeOH 4.6 acetate/ac.

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Shim-pack CLC ODS(M); 150 acid (100:40:7) (isocratic) TS-MS, SIM
× 4.6
1992 [67] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Spherisorb ODS2, C18; 100 A: water 30; 1.2 UV/Vis
indigo (27) 100ºC; rediss. water/MeOH × 4.6; 3 B: MeOH
(1:1) 10% C: 5% ph. acid
pyridine, 100ºC
1994 [78],1995 Anthraquinoids, shellfish 3 M HCl/MeOH (1:1), 100ºC; Lichrosorb 15537 RP-18, A: water 20; 1 UV/Vis
[79, 80], 2005 [35] purple, indigo (12) rediss. MeOH C18; 150 × 4; 7 B: MeOH
DMF, 150ºC 10% C: 5% ph. acid
1996 [91] Anthraquinoids, flavonoids, 3 M HCl/MeOH, 100ºC; rediss. Spheri-5 ODS, C18; 100 × A: water/0.1% TFA 40; 0.6 UV/Vis
indigo, carthamin, tannins water; MeOH 2.1; 5; 38ºC B: ACN/0.1% TFA
(38)
b)
1997 [77] Shellfish purple (7) DMF, hot Kromasil; 250 × 4.6; 5 A: water 23 ; 1 UV/Vis
B: MeOH
10% C: 5% ph. acid
b)
1999 [62] Anthraquinoids (8); water/MeOH/37% HCl (1:1:2), Purospher RP18e, C18; 125 A: 0.1 M aq. citrate buffer 12 ; 0.6 UV/Vis, 0.6–12 ng
naphthoquinoids (3) 100ºC × 4; 5 B: MeOH
b)
MeOH/0.2 M aq. acetate 10 ; 0.5
buffer (72:25) (isocratic)
b)
2003 [95] Anthraquinoids (insect 10% HCl, 65ºC; ext. n-amyl Zorbax SB-C18; 250 × 4.6; 5 Water/80% MeOH/0.2% ac. 12 ; 0.7 UV/Vis, 0.18–1.71 ␮g/mL;
reds) (4) alcohol, MeOH acid ESI-MS, 30–90 ng/mL
b)
2003 [85] Anthraquinoids, flavonoids, 3 M HCl/EtOH (1:1), 100ºC; Phenomenex Luna, C18; 250 A: water/5% ACN/0.1% TFA 36 UV/Vis
indigo (12) rediss. water/MeOH (1:1) × 4.6; 5; 40ºC B: ACN/0.1% TFA

www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410

Table 1. Continued

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)
J. Sep. Sci. 2014, 37, 3393–3410

b)
2003 [85] Indigoids (4) Warm pyridine A: water/ACN/THF (50:45:5) 18.3
B: THF/ACN (95:5)
2003 [90] Anthraquinoids, flavonoids, 3 M HCl/EtOH (1:1), 100ºC; Zorbax RX-C18; 150 × 2.1; 5; A: water/0.3% form. acid 40; 0.2 UV/Vis; ESI-MS, APCI-MS,
indigoids (13) rediss. water/MeOH (1:1) 30ºC, 20ºC B: ACN MRM
b)
2004 [59] Insoluble reds (15) Water/MeOH/37% HCl (1:1:2), Hypersil BDS C18; 100 × 2.1; A: water 32 ; 0.3 UV/Vis
100ºC; rediss. DMF 3; 30ºC B: ACN
Water/MeOH (1:1), 100ºC C: 1% MSA

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

DMF, 100ºC
2008 [60] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Hypersil BDS C18; 100 × 2.1; A: water 45; 0.3 UV/Vis, 3–24 ppb
hydroxybenzoic acids (20) 100ºC; rediss. DMSO 3; 30ºC B: ACN
10% C: 1% MSA
A: water FLD (ZrO2+ ), 1–226 ppb
B: MeOH
10% C: 1% MSA
2004 [54] Indigoids (10) DMSO/HCl (100:5), 80ºC; ACN Zorbax SB-C18; 150 × 4.6; A: water/0.15% form. acid 35; 0.6 UV/Vis; ESI-MS, SIM
3.5 B: ACN 0.01–0.15 ␮g/mL
2006 [50] Flavonoids, neoflavones, DMSO/37% HCl (100:2), 80ºC; Zorbax SB-C18; 150 × 4.6; A: water/0.15% form. acid 55; 0.5 UV/Vis; ESI-MS, SIM
indigoids (16) MeOH 3.5 B: MeOH
2011 [39] Anthraquinoids, flavonoids, MeOH/form. acid (9:1), 60ºC; Zorbax SB-C18; 150 × 4.6; A: water 27; 0.5 UV/Vis; ESI-MS, SIM
indigoids (29) MeOH/37% HCl (7:1), 60ºC 3.5 B: MeOH
10% C: 1% form. acid
2004 [55] Anthraquinoids, indigoids MeOH/30% HCl (30:1), 60ºC; Genesis 80, C18; 250 × 4.6 A: water/19% ACN/0.1% TFA 40; 1 UV/Vis; FLD
ext. ethyl acetate B: water/50% ACN
2005 [68] Anthraquinoids, flavonoids Water/MeOH/37% HCl (1:1:2), Vydac 214TP52, C4; 250 × A: water 60; 0.2 UV/Vis; ESI-MS
(4); 100ºC; rediss. water/MeOH 2.1; 5 B: ACN
2009 [49] flavonoid glycosides (1:1) 5% C: 1% form. acid
MeOH/form. acid (95:5), 40ºC;
rediss. water/MeOH (1:1)
0.001 M EDTA/ ACN/MeOH
(2:10:88), 60ºC; rediss.
water/MeOH (1:1)
Liquid Chromatography

www.jss-journal.com
3397

3398

Table 1. Continued

Column, stationary phase; Analysis

V. Pauk et al.

Investigated Extraction dimensions (mm); particle time (min); Detection,

References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

2006 [86] Shellfish purple (10) DMSO, 100ºC Symmetry C18; 150 × 3; 5 A: water 30; 0.8 UV/Vis, 50 pg
B: MeOH
2008 [31, 88] 10% C: 5% ph. acid 25; 0.8–2.1 UV/Vis
b)
2006 [81] Shellfish purple (6) DMF, 70ºC XTerra C18, 40ºC; 250 × 3; 5 A: water/0.001% TFA 12 ; 0.4 UV/Vis; APCI-MS, SIM
B: ACN/0.001% TFA
b)
2006 [63] Anthraquinoids, flavonoids, water/MeOH/37% HCl (1:1:2), - - 32.5 UV/Vis; ESI-MS, SIM 0.01

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

indigoids (21) 100ºC; rediss. DMF, ␮g/mL
water/MeOH (1:2)
2006 [61] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Phenomenex Sphericlone, A: 20% MeOH 22.5; 1.2 UV/Vis
indigotin (11) 100ºC; rediss. water/MeOH C18; 150 × 4.6; 5; 25ºC B: MeOH
(1:1); MeOH/DMF (1:1), 100ºC Phenomenex Gemini, C18; 10% C: 5% ph. acid 25; 0.35
150 × 2.1; 5
b)
2006 [57] Anthraquinoids, flavonoids, 0.1 M SDS Phenomenex Luna, NH2 ; 250 A: 40 mM SDS/10 mM ph. 42 ; 0.6 UV/Vis
indigotin (6) × 4.6; 5; 35ºC buffer/0.1% TFA
B: ACN
2007 [5] Anthraquinoids, flavonoids, water/MeOH/37% HCl (1:1:2), Zorbax RX C8; 150 × 2.1; 5 A: water/0.3% form. acid 45; 0.2 UV/Vis, 2–31 ng/mL; ESI-MS,
hydroxybenzoic acids (17) 100ºC; rediss. water/MeOH B: ACN SIM, MRM, 0.5–32 ng/mL
(1:1)
b)
2011 [6] Anthraquinoids, tannins, water/MeOH/37% HCl (1:1:2), Zorbax RX C8; 150 × 2.1; 5; A: water/0.3% form. acid 32 ; 0.2 UV/Vis, 10–70 ng/mL;
flavonoids, hydroxybenzoic 100ºC; rediss. DMSO 30ºC B: ACN APCI-,ESI-MS SIM, MRM
acids, indigotin (15) 0.4–20 ng/mL
2008 [45] Anthraquinoids, flavonoids ACN/MeOH/4 M HF (1:1:2) Alltima RP, C18; 250 × 4.6; 5; A: MeOH 65; 1 UV/Vis
and their glycosides Water/MeOH/12 M HCl 20ºC B: water/5% ACN
(1:1:2), 110ºC; rediss. C: 0.1% TFA/MeOH (1:1)
water/MeOH (1:1) D: ACN
2008 [47] Anthraquinoids, flavonoids, DMF, 80ºC; Alltima HP, C18, 35ºC; 250 × A: water/0.1% TFA 35; 0.5 UV/Vis, 2–29 ng/mL
2009 [46] indigotin, indirubin (12) water/MeOH/37% HCl 1:1:2 3; 5; 33ºC B: ACN/0.1% TFA UV/Vis, ESI-MS
100 ºC; rediss. DMF, DMSO

www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410

Table 1. Continued

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)
J. Sep. Sci. 2014, 37, 3393–3410

2009 [37] +flavonoid glycosides, +Water/MeOH/0.5 M citric Alltima HP, C18, 250 × 3; 5; UV/Vis, ESI-MS
curcuminoids (14) acid (1:1:1), 100ºC; DMSO 35ºC
Water/MeOH/1 M ox. acid
(1:1:1), 100ºC; DMSO
Water/MeOH/2 M TFA (1:1:1),
100ºC; DMSO
5 M form.

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

acid/water/MeOH/0.5 mM
EDTA (2:1:1:4), 100ºC; DMSO
2008 [76] Shellfish purple and DMF Phenomenex Synergi A: water/0.1% form. acid 18; 0.3 UV/Vis; ESI-MS
precursors (12) Hydro-RP, C18; 250 × 4.6; 4 B: ACN
2008 [93] Flavonoids, anthraquinoids, DMF, 140ºC; Phenomenex Luna C18, 100 A: water 27 UV/Vis
indigoids, tannins, orchill water/MeOH/37% HCl (1:1:2), ×2 B: MeOH
100ºC, rediss. DMF 10% C: 5% ph. acid
2009 [38] Indigo, flavonoids, MeOH/acetone/ water/ox. Nucleosil RP-18; 250 × 4.6; A: water/0.3% perchloric 40; 1.7 UV/Vis;
coumarins (12) acid 0.2 M (3:3:4:0.1), 60ºC; 5; 35ºC acid
DMF, 60ºC; rediss. B: MeOH
water/MeOH (1:1)
Polaris C18-A; 150 × 2; 5; A: water/0.8% form. acid 40; 0.3 UV/Vis; ESI-MS
35ºC B: ACN
2010 [58] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Phenomenex Luna, C18; 150 A: water/0.1% TFA 50; 0.5 UV/Vis
indigotin, tannins 100ºC, MeOH/DMF (1:1), 100ºC × 2.1; 5; 35ºC B: ACN
Form. acid/MeOH (5:95), 50ºC,
MeOH/DMF (1:1), 100ºC
2010 [69] Anthraquinoids, tannins (5) Water/MeOH/37% HCl (1:1:2), Nova-Pack C18; 150 × 3.9; 4; A: water/0.1% TFA 45; 0.5 UV/Vis
100ºC; rediss. water/MeOH 30ºC B: ACN/0.1% TFA
2011 [70] Flavonoids, tannins(4) (8) (1:2)
2012 [71] Shellfish purple (3)
2011 [83] DMF, 115ºC Phenomenex Luna, C18; 150 A: water 10; 0.9 UV/Vis; APPI-MS, SIM
× 4.6; 3
Liquid Chromatography

www.jss-journal.com
3399

3400

Table 1. Continued

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
V. Pauk et al.

References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

2012 [75] Shellfish purple (7) DMF, 110ºC B: ACN

2011 [48] Anthraquinoids (insect Form. acid/MeOH (5:95), 60ºC; Zorbax Eclipse Plus, C18; A: water/0.3% perchloric 30; 0.5 UV/Vis; ESI-MS
reds) (6) rediss. water/MeOH/0.3% 150 × 2.1; 5; 35ºC acid
perchloric acid (50:20:30) B: MeOH
Water/MeOH/37% HCl (1:1:2),
60ºC; rediss.
water/MeOH/0.3% perchloric

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

acid (50:20:30)
0.2 M ox.
acid/acetone/MeOH/water
(0.1:3:3:4), 60ºC; rediss.
water/MeOH/0.3% perchloric
acid (50:20:30)
2 M TFA, 60ºC; rediss.
water/MeOH/0.3% perchloric
acid (50:20:30)
b)
2011 [3] Shellfish purple (7) DMF, 60ºC Wakosil II 5C18RS, C18; 250 A: water/0.1% TFA 21 ; 0.4 UV/Vis
× 4.6; 5; 40ºC B: ACN/0.1% TFA
2011 [42] Anthraquinoids, flavonoids, DMSO, 60ºC; HCl/MeOH Wakosil II 5C18RS; 250 × A: water/0.1% TFA 60; 0.8–1 UV/Vis, 10–70 ng/mL
indigo, tannins, carthamin (1:30), 60ºC 4.6; 5 B: ACN/0.1% TFA
(18)
c) b)
2011 [89] Shellfish purple (14) DMSO, 30ºC, 70ºC Alltima , C18; 150 × 2.1; 3; A: water 26 ; 0.3 UV/Vis
70ºC B: ACN
10% C: 1% MSA
2012 [44] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), LiChrosorb-C18; 125 × 4; 5 A: MeOH 35; 1.2 UV/Vis
indigotin (28) 100ºC, rediss. water/MeOH B: MeOH/water (1:9)
(1:1); DMF, 140ºC 10% C: 5% ph. acid
2011 [51] Water/MeOH/37% HCl (1:1:2), Zorbax C18; 150 × 4.6; 5; A: water/0.2% form. acid 18; 0.8 UV/Vis; ESI-MS, SIM, MRM
2012 [44] 105ºC, rediss. water/MeOH 40ºC B: MeOH/ACN (1:1)
2010 [43] (14) (1:1)

www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410

Table 1. Continued

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

c)
2011 [73] Anthraquinoids (30) MeOH/ACN/4 M HF (1:1:2), Interchim Uptisphere NEC , A: water 86; 0.2 UV/Vis
45ºC; rediss. DMSO B: ACN
J. Sep. Sci. 2014, 37, 3393–3410

C18; 150 × 2.0; 5; 30ºC

c)
Hypersil Gold , C18; 150 × 10% C: 1% form. acid 91; 1.0
4.6; 5; 30ºC
2011 [74] Synaptec Caltrex
c)
Resorcinaren Calixarene
Resorcinol; 250 × 4; 5; 30ºC
c)
Pursuit XRs DP , Diphenyl;
250 × 4.6; 5; 30ºC
c)

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Luna Phenyl-Hexyl ; 250 × 86; 0.2
2; 5; 30ºC
2011 [40] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Phenomenex Luna C18(2); A: water/10% MeOH 50; 0.160 UV/Vis; MS
indigo, polyenes (24) 100ºC; all rediss. water/MeOH 100 × 2.0; 3; 25ºC B: MeOH
(1:1) 10% C: 2% form. acid
MeOH/form. acid (95:5), 40ºC;
water/MeOH/37% HCl (1:1:2),
100ºC
MeOH/acetone/ water/0.21 M
ox. acid (30:30:40:1), 60ºC;
water/MeOH/37% HCl (1:1:2),
100ºC
MeOH/acetone/water/HF
(30:30:40:1); water/MeOH/37%
HCl (1:1:2), 100ºC
2011 [41] Anthraquinoids, flavonoids Water/MeOH/37% HCl (1:1:2), Lichrocart Purospher Star A: water/2.5 % ACN/0.5 % 15; 1.0 UV/Vis
and glycosides, indigo, 100ºC RP-18; 250 × 4.6; 5; 30ºC form. acid
brazilwood, logwood (11) B: ACN
Water/MeOH/37% HCl (1:1:2), Fortis C18; 150 × 2.1; 3; 30ºC A: water/0.1 % form. acid 30 UV/Vis ESI-MS
100ºC, MeOH/DMF(1:1), 100ºC B: ACN
Form. acid/MeOH (1:19), 40ºC
0.001 M
Na2 EDTA/ACN/MeOH
(1:5:44), 60ºC
Liquid Chromatography

0.1% Na2 EDTA/water/DMF

(1:1), 100ºC

www.jss-journal.com
3401

3402

Table 1. Continued
V. Pauk et al.

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

2 M ox.
acid/acetone/MeOH/water
(1:30:30:40), 60ºC
Pyridine, 100ºC
DMF, 100ºC
b)

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

2012 [87] Shellfish purple (12) DMSO Hypersil Gold C18; 50 × 2.1; A: water/0.1% form. acid 24 ; 0.3 UV/Vis, HESI-MS
1.9; 40ºC B: ACN/0.1% form. acid
c)
2013 [84] Flavonoids, anthraquinoids, DMF; DMF 90ºC UHPLC BEH Shield RP18 ; A: water/10% MeOH 40; 0.2 UV/Vis, 0.046–0.303 ng
indigotin, curcumin (9) 150 × 2.1; 1.7; 40ºC B: MeOH (except curcumin)
C: water/1% form. acid
HPLC Phenomenex Luna C: 5% ph. acid 0.140–0.321 ng (except
C18; 150 × 2; 3; 35ºC curcumin)
2012 [56] Indigotin, indirubin (2) 6 M HCl/MeOH (50:50), 95ºC; Zorbax Extend-C18 RRHT; 50 A: water/0.1% form. acid 25; 0.8 UV/Vis; ESI with
rediss. MeOH/DMF (1:1) × 2.1; 1.8; 35ºC B: ACN superheated N2
2014 [72] Sawwort flavonoids (18) MeOH/form. acid (9:1), 60ºC; Zorbax SB-Phenyl; 150 × A: water/0.15% form. acid 30; 0.5 UV/Vis; ESI
MeOH/37% HCl (7:1), 60ºC 4.6; 3.5 B: MeOH
c)
2012 [64] Flavonoids, anthraquinoids Water/MeOH (1:1), 2 M HCl Acquity UPLC BEH C18 ; A: water/0.1% form. acid 6; 0.25 UV/Vis
(8) Water/MeOH (1:1), 2 M EDTA 100 × 2.1; 1.7; 30ºC B: ACN/0.1% form. acid
Water/MeOH (1:1), 2 M TFA
Indigotine DMF, 100ºC
THF
2013 [65] Insect anthraquinoids (7) Water/MeOH/37% HCl (1:1:2), Alltima HP C18; 250 × 3; 5; A: water/0.1% TFA 35; 0.5 UV/Vis
100ºC, rediss. DMSO 33ºC B: ACN/0.1% TFA
A: water/amm. form. buffer, ESI, APCI
pH 3
B: ACN

www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410

J. Sep. Sci. 2014, 37, 3393–3410

Table 1. Continued

Column, stationary phase; Analysis

Investigated Extraction dimensions (mm); particle time (min); Detection,
References colorantsa) procedure diameter (␮m); temperature Mobile phase flow LOD
(mL/min)

2013 [53] Shellfish purple (4) DMSO, 150ºC Symmetry C18; 150 × 3; 5 A: water 30; 0.6 UV/Vis

C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

B: MeOH
2012 [52] Anthraquinoids (12) Water/MeOH/37% HCl (1:1:2), Zorbax C18; 150 × 4.6, 5; A: water/0.2% form. acid 20; 0.8 UV/Vis, ESI
105ºC, rediss. water/MeOH 40ºC B: MeOH/ACN 1:1
(1:1)
2014 [94] Flavonols (13) Pyridine/water/1 M ox. acid Vydac C18; 250×2.1; 5 A: water/0.1% form. acid 35 UV/Vis, ESI
(95:95:10), 100ºC B: ACN/0.1% form. acid
2013 [10] Anthraquinoids, flavonoids, Water/MeOH/37% HCl (1:1:2), Synergi C18; 150 × 2.1; 3 A: water/0.15% form. acid 30; 0.2 UV/Vis, ESI
indigotin, brazilein (10) 100ºC; all rediss. in B: ACN/0.15% form. acid
water/MeOH
Water/MeOH/5 M form. acid
(1:1:2), 0.5 ␮M EDTA, 100ºC
DMF, 60ºC

a)
Numbers in parentheses indicate the number of resolved or identified compounds (e.g. coeluting compounds can be distinguished by MS).
b)
The last eluting compound.
c)
The most efficient columns investigated in the study.
Abbreviations: ACN, acetonitrile; APPI, atmospheric pressure photoionization; EtOH, ethanol; FLD, fluorescence detection; HESI, heated-electrospray ionization; MRM, multiple reaction
monitoring; SIM, selected ion monitoring; THF, tetrahydrofuran.
Liquid Chromatography

www.jss-journal.com
3403
3404 V. Pauk et al. J. Sep. Sci. 2014, 37, 3393–3410

(trifluoroacetic acid) extraction [37, 48], oxalic acid extrac- of flavonoid glycosides. For flavonoid dyes on silk, the EDTA
tion [37, 38, 40, 41, 48], HF (hydrofluoric acid) procedure method seems to be more efficient, while formic acid treat-
[40, 45, 73, 74], DMF (dimethylformamide) [3, 35, 41, 44, 46, ment is more useful for the anthraquinoid-type dyes [68]. The
47, 51, 59, 64, 75–84], pyridine [41, 67, 85], DMSO (dimethyl- combination of both formic acid and EDTA is also very effec-
sulfoxide) [31, 42, 50, 53, 54, 86–89], and few others (see Table tive [37]. However, among five examined methods (citric acid,
1). Usually extraction is performed as follows: oxalic acid, formic acid/EDTA, TFA, and HCl methods), TFA
(water/MeOH/2 M TFA, 1:1:1, 100⬚C) was superior for cur-
(i) The sample is treated with a small volume (50–400 ␮L) of cumins, rubiadin, and sulfuretin; the use of TFA was compa-
solution mixture for a time ranging from a few minutes rable to other methods for indigotin and indirubin (according
to a few hours [40] at elevated temperature (40–60⬚C for to S/N) [37].
mild methods, 100⬚C for HCl extraction, or higher for Oxalic acid (0.2 M oxalic acid/acetone/MeOH/water,
DMF or DMSO extraction of indigoids) or, in rare cases, 0.1:3:3:4, 60⬚C) and 2 M TFA treatments are the most effi-
without heating [40, 45]. cient methods for removing cochineal dye from textiles [48].
(ii) The extract is cooled and evaporated to dryness under The oxalic acid method showed better chromatographic res-
vacuum or under nitrogen flow (strong acid must be olution than TFA and other methods and rendered a higher
evaporated due to its corrosive nature, it can damage level of agreement, based on PCA scores, between cochineal-
metal counterparts of an HPLC instrument as well as dyed reference textiles and extracts from Dactylopius coccus
column packing). Costa insect.
(iii) The residue is redissolved in a small volume (typically HF is also an efficient complexing agent for mordant dye
100–400 ␮L) of water/methanol (MeOH) [5,10,38,40,43– metals, especially Al and Fe. A mild extraction by HF was as
45, 49, 51, 52, 61, 63, 66–71, 85, 90], MeOH [35, 78–80, 91], efficient as the classical HCl method for the removal of acid-
DMF [47, 59, 63], MeOH/DMF [56, 58, 61, 92], or DMSO stable components (alizarin, purpurin, kermesic acid) from
[6, 37, 46, 60, 65, 73, 74]. Al-based lakes. Besides, mild conditions allowed analysis of
acid-labile compounds (pseudopurpurin, munjistin, and var-
Filtration or centrifugation may be applied after each of ious glycosides) [45]. The HF method can also be applied
these steps. In any case, undissolved particulates should be for dyes mordanted with other metals (Ba, Sn) and for non-
removed prior to HPLC injection. mordant and synthetic dyes. A disadvantage does exist in the
HCl hydrolysis (water/MeOH/37% HCl, 1:1:2, 100⬚C) is necessity to eliminate excess fluorides either by SPE cleaning
the most widely used method, especially for mordant dyes or by evaporation to dryness and HPLC online clean-up. Ex-
where a strong acid is needed to disrupt dyestuff–metal com- treme care must be taken while working with HF since it is
plexes. Since many dyes, particularly flavonoids, belong to highly caustic. Only properly equipped and certified labora-
O-glycosides, they easily hydrolyze and only parent aglycon tories should be allowed to perform this type of extraction.
can be observed. Information about the original dyestuff is It was shown recently that double extraction of the same
lost (the presence of the glycoside moiety may provide valu- sample with a mild method followed by HCl treatment pro-
able information on the plant source, see Supporting Infor- vided more information than a single-step procedure [39]. The
mation Table S1, yellow). Moreover, the method alone is not first step isolated glycosides, and the second increased yield of
efficient for extraction of indigoid compounds insoluble in aglycones. However, formation of stable O-methylated deriva-
aqueous medium. Therefore, attention was switched to mild tives due to reaction of flavonoid O-glucuronides with MeOH
extraction methods to preserve the glycoside structure and in the presence of HCl should be taken into account [72].
acid-labile dyestuffs such as saffron or turmeric. Nine methods including double extraction were applied
There are several papers explicitly devoted to revealing to different textile samples (wool, cotton, and silk), pigments,
the best extraction method [37, 40, 41, 48, 68]. The methods and paints obtained from various biological sources and
are usually compared according to the following criteria: from crude plant material as well. The method involving HF
(i) HPLC peak areas of extracted compounds [37, 40, 41, 68]; (MeOH/acetone/water/HF, 30:30:40:1, 4 h) followed by tradi-
(ii) number of compounds extracted from substrates [37, 68]; tional HCl hydrolysis was the most efficient for the extraction
or (iii) S/N measured for each compound [37]. Principal com- of acid-labile components and indigoids, except that recover-
ponent analysis (PCA) scores of multivariate component anal- ies of carthamin and rutin were low. The best extraction for
ysis [48] or differences identified by analysis of variance statis- textile combines oxalic acid (MeOH/acetone/water/0.21 M
tics [41] were also applied in some cases. oxalic acid, 30:30:40:1, 60⬚C) and traditional HCl hydrolysis.
EDTA as chelating agent for Al3+ and other cations used Formic acid (MeOH/formic acid, 95:5, 40⬚C) followed by HCl
in dyeing can be efficient to aid in the analysis of mordant hydrolysis is recommended for extraction of indigoids from
dyes [49, 68]. Interestingly, free EDTA is reported to be more textiles [40].
effective than its disodium salt [68]. Upon comparison of In the most common extraction methods, the extract of
the results of a formic acid/MeOH (5:95, v/v, 40⬚C, 30 min) a dyestuff is dried and reconstituted with a water/MeOH
and a 0.001 M EDTA/acetonitrile/MeOH (2:10:88, v/v, 60⬚C, mixture. Improvement using MeOH/DMF was reported.
30 min) extraction, both methods rendered better results than Dissolving of the residue, obtained after acid hydrolysis and
traditional HCl water/MeOH systems and allowed detection evaporation, in MeOH and DMF (1:1, v/v) improved the


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410 Liquid Chromatography 3405

recovery for major flavonoids, anthraquinoids, and some 3.2.1 Columns

other compounds (luteolin, luteolin methyl ether, apigenin,
kaempferol, alizarin, purpurin, maclurin, and ellagic The most separations have been done using columns packed
acid)—even for indigotin (up to sevenfold increase) [61]. A with 3 or 5 ␮m particles. Recently, UHPLC with sub-2 ␮m par-
similar approach (6 M HCl/MeOH, 1:1, 95⬚C; redissolved ticles was also applied [56, 64, 84, 87]. A comparison of HPLC
in MeOH/DMF, 1:1, 90⬚C) was successfully used to extract and UHPLC columns for the separation of flavonoids and
indigoids from Maya blue pigment (strong complex of inor- anthraquinones showed that the latter provided better resolu-
ganic clay and indigo dyestuff) [56]. Contrary to this result, tion, higher peak capacity, and lower LODs, while maintain-
with the exception of brazilein and hematein, no significant ing the same reproducibility (RSD below 1%) [84]. Monolithic
improvement in extraction yields was observed with the columns were tested with standards but not used for analysis
MeOH/DMF step, as it was observed in the other study [41]. of real historical samples [73, 89]. Further improvement in
A modified EDTA method (Na2 EDTA 0.1% in water/DMF, HPLC analysis of mass-limited samples can be achieved by
1:1, 100⬚C) was evaluated to be the most effective procedure reduction of column diameter [61, 91]. The use of a narrow-
for extraction of unknown dyes [41]. Regardless of the extrac- bore column can result in increased peak heights for analyzed
tion method, direct injection of extracts has rendered higher compounds [61].
areas of chromatographic peaks in comparison to procedures
using evaporation and reconstitution with MeOH [39].
Indigoid dyes are efficiently dissolved only in solvents 3.2.2 Stationary phases
such as DMF, pyridine, or DMSO. Nevertheless, indigotin
was not found in a DMF extract; the results from the EDTA Octadecylsilyl (C18 ) phases are the most often used. In some
method (Na2 EDTA 0.1% in water/DMF, 1:1) did not signif- cases, C4 [49,68] or C8 [5,6] phases have been used for analysis
icantly differ from pyridine extraction [41]. Pyridine caused of flavonoid glycosides or hydroxybenzoic acids due to the
double peaks due to its own retention on the column [85]. higher polarity of these compounds. Interestingly, among
DMSO is sometimes considered as advantageous over DMF three stationary phases examined, the resolution of an NH2
and pyridine for its higher boiling point. It is also odorless column was superior to C8 and C18 [57]. It is worth noting,
and nontoxic [73]. DMSO acidified with HCl has been sug- 40 mM SDS was used as an additive in the mobile phase in
gested for extraction of natural dyestuffs when the presence that case.
of indigoids is expected [50,54]. It is also efficient in extraction The most extensive systematic evaluation of station-
of carthamin [42]. Indigoids decompose to a large extent at ary phases and other chromatographic conditions for an-
elevated temperatures. In one study, an increase of the extrac- thraquinoid and indigoid colorants was carried out by Nowik
tion temperature from 30 to 70⬚C induced the loss of about and Bellaistre in [74] [73, 74, 89]. Peak symmetry and separa-
30% for brominated indigotins and 50% for indigotin, and tion capacity (number of critical pairs) were the major criteria.
the content of degradation products, the isatins, was raised Polar stationary phases were not suitable for analysis of an-
about 20% [89]. Solutions of indirubin and isoindigo are rela- thraquinoid compounds [73]. No peaks for molecules with a
tively photostable, but indigotin degraded by 50% after seven carboxylic acid group such as carminic acid, kermesic acid,
days of exposure to daylight. After 30 days, only isatin was flavokermesic acid, and anthraquinone-2-carboxylic acid were
found in the examined solution [54]. observed. The authors suggest a strong or an irreversible in-
Multistep extraction may be considered when the pres- teraction between this group and polar groups of stationary
ence of different dye types is expected. A two-step proce- phases under selected conditions. Chelating anthraquinones
dure was useful for analysis of Hallstatt textiles [93]. First, are sensitive to the presence of metallic impurities in the
indigoid components were extracted with hot DMF and then silica support and tend to lead to the observation of peak tail-
mordant dyes were hydrolyzed by the classical HCl method. ing [74]. Two complementary C18 columns, Uptisphere NEC
The combination of different extraction methods for distinct (Interchim) and Hypersil Gold [73], or three functionalized,
color fibers was also used recently [10]. Red fibers and part of Caltrex Resorcinaren (Synaptec), Pursuit XRs DP (Varian),
orange-pink samples were extracted by the classical HCl pro- and Luna Phenyl-Hexyl (Phenomenex), were recommended
cedure, yellow fibers and another part of orange-pink fibers for real samples analysis [74] (see Table 1 for details).
by formic acid and EDTA, blue fibers were treated with DMF, Retention of slightly soluble shellfish purple components
and then extraction as for yellow fibers was applied. The re- on various C18 stationary phases was related to the maxi-
verse procedure was performed for greens (formic acid and mum available hydrocarbon surface of a stationary phase [89].
EDTA, then DMF). Stronger retention allowed the use of a higher content of or-
ganic modifier in the mobile phase and a higher column
temperature. Both improved analyte solubility and as a con-
3.2 Chromatographic separations sequence, peak symmetry. Optimized analysis with the most
retentive phase, Alltima (Grace), used at 70⬚C, resulted in an
Natural dyestuff components are usually separated in RP almost 400% improvement of solubility compared to a less-
mode on octadecylsilyl (C18 ) stationary phases using MeOH retentive phase used at near-ambient temperature. Finally,
or acetonitrile as an organic modifier (for details, see Table 1). major and trace components were quantified [89].


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3406 V. Pauk et al. J. Sep. Sci. 2014, 37, 3393–3410

3.2.3 Mobile phases 3.3 Detection

Water and MeOH were used in the early days of HPLC of Colorant molecules absorb well in the UV/Vis range. Diode
natural colorants [26,66,67,78–80,92]. Acetonitrile has certain array detectors can support identification of a compound
advantages over MeOH, including lower absorbance in the through the comparison of UV/Vis spectra.
200–275 nm range, lower UV cut-off, and lower back-pressure Fluorescence detectors are more selective than UV/Vis
due to lower viscosity [91]. However, its higher price may be and offer lower LODs [18, 55, 60]. Limited number of com-
disadvantageous when large amounts of analysis is necessary. pounds exhibit natural fluorescence, but derivatization can
Other solvents were evaluated to improve the solubility of be performed. Complexation with metal ions increases fluo-
indigoid dyes but did not receive general recognition, as in rescence of flavonoids. Anthraquinoids and hydroxybenzoic
the case of tetrahydrofuran [85], or were not successful at all, acids in natural dyes and textiles were detected fluorimetri-
like DMSO [89]. Addition of 5% DMSO to acetonitrile did not cally after postcolumn complexation with zirconyl ion [18,60].
improve peak shape [89], and higher content was precarious With optimized fluorescence detection, considerably lower
due to the higher cut-off wavelength (268 nm) and viscosity detection limits were obtained for all complex-forming com-
it imparted. pounds (all flavonoids and dihydroxybenzoic acids), with the
Usually, gradient elution is applied and performed as exception of alizarin [60]. Additionally, bathochromic shifts
follows: Starting from water/organic (95:5) modifier mixture for flavonoids after postcolumn deprotonation or complexa-
with additives and changing to 100 (90)% organic modifier tion with zirconyl facilitated structural recognition of flavones
within 30–60 min, depending on the column diameter, flow or flavonols, respectively [60].
rate, and complexity of analyzed mixture. In one study, a MS has been shown to provide a few considerable
faster gradient from 40 to 90% acetonitrile (acidified with advantages over UV/Vis detection, including better LODs,
0.001% TFA) enabled an acceptable separation of six natu- improved selectivity, and the possibility for structural
ral indigoid colorants within 15 min [81]. In other studies, elucidation of unknown compounds. For some fixed HPLC
a linear gradient (65–100% acetonitrile in 10 min) allowed conditions, representative LODs can be compared:
the efficient separation of the seven indigoid components
of shellfish purple [75, 83]. Work using UHPLC has shown (i) UV/Vis: 0.18–1.71 ␮g/mL; MS: 30–90 ng/mL (anthraq-
decreased analysis times to 6 min [64]. uinoids) [95];
(ii) UV/Vis: 2–31 ng/mL; MS: 0.5–12 ng/mL (anthraq-
uinoids, flavonoids, benzoic acids), 32 ng/mL (quinizarin)
[5];
(iii) UV/Vis: 10–70 ng/mL; MS: 0.4–20 ng/mL (anthraq-
3.2.4 Additives uinoids, flavonoids, benzoic acids, tannins, indigotin) [6].

Addition of acid suppresses dissociation of analytes, most Thermospray ionization [27, 92] and later more robust
of which contain acidic groups, and prevents nonspecific API techniques, ESI [5,6,10,29,37–39,41,43,44,48–52,54,63,
interactions with silanol groups on the stationary phase 65,68,72,76,90,94,95] and APCI [6,65,81,90], have been exten-
silica support [18, 91]. Formic acid [5, 6, 10, 26, 29, 39–41, sively used to facilitate MS detection of colorants. The perfor-
43, 44, 49–52, 54, 56, 58, 64, 68, 70, 72–74, 76, 84, 87, 90, 94], mances of ESI and APCI have been specifically compared [6].
TFA [3, 37, 42, 45–47, 55, 65, 69, 71, 81, 91], and phosphoric ESI is suitable for the detection of all anthraquinoids,
acid [31, 35, 44, 61, 67, 73, 77–80, 84, 86] belong to the most of- hydroxybenzoic acids, and flavonoids investigated. However,
ten used additives. Sometimes methanesulfonic acid (MSA) only indirubin was detected among indigoids. This result
[59, 60, 89], acetic acid [95], acetate buffer [62, 90, 92], formate is in agreement with another work [90]. Carminic acid was
buffer [65], and perchloric acid [38, 48] have been added. The detected in negative ESI mode. Under APCI conditions, all
use of strong acids such as TFA and MSA with acetonitrile flavonoids were ionized both in positive and negative mode
has been shown to lead to improved peak shape, and thus, while anthraquinones were detected in negative mode only.
better resolution of analytes (indigoids) [89]. However, TFA Mono- and dihydroxybenzoic acids provided signals both in
causes significant signal suppression in mass spectrometric the positive and negative ionization modes, whereas gallic
detection. Reducing TFA concentration from 0.1 to 0.001% acid was only observed in the negative mode. Both indigotin
can still provide adequate chromatography without peak de- and indirubin were ionized in negative mode APCI [6]. On the
terioration [81]. TFA also causes corrosion of LC–MS equip- contrary, another study [90] demonstrated that positive-mode
ment [27]. Its concentration should be kept at the reasonable APCI was suitable for indigoids. Recently, atmospheric
minimum or it should be replaced with more suitable additive pressure photoionization in negative mode [75, 83] and
such as MSA or formic acid. MSA is sometimes preferred for thermally assisted ESI in positive mode [56, 87] have also
its lower absorption in the UV region and weak ion-pairing been successfully used for indigoid analysis.
properties [89]. Formic acid is suitable for HPLC–MS and can Collision-induced dissociation after soft ionization pro-
render comparable separation performance as the stronger vided clue for identification of natural organic dyestuffs, but
TFA or perchloric acid choices [64]. it was limited to samples with higher concentration or to pure


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410 Liquid Chromatography 3407

standards in the past [27]. Using a modern high-sensitivity even belonging to the same biological family can differ in
Q-TOF instrument, it was possible to confirm the presence of quantitative colorant profiles, e.g. madder (Rubia sp.),
indigotin and indirubin in 0.1 mg microsamples according to cochineal and kermes (see Supporting Information Table
MS/MS data [56]. Identification can be supported by accurate S1), or shellfish purple (see Supporting Information Table
mass determination. An LTQ Orbitrap instrument with mass S2). Nevertheless, quantitation alone sometimes produces
measurement error <0.5 ppm allowed trustworthy identifi- data from which differences are not obvious (see Support-
cation of four new compounds in extracts from H. trunculus ing Information Table S1, cochineal). Multivariate data
L [87]. The mass error of the MS/MS measurement in most analysis (PCA method) differentiated between species of
cases was about 5 ppm. A considerable amount of informa- cochineal [48]. The investigations involved freshly prepared
tion on MS of natural historical dyes, including ionization and dyestuffs but historical samples represent a more complicated
typical fragmentation patterns of main chemical classes of issue. Even if the exact amount of dye components in the
colorants, is presented in Rosenberg’s review [27]. Since that extract is known, the interpretation is complicated due to the
time, several new papers dealing with LC–MS of flavonoid unknown impact of degradation processes during aging [18].
glucuronides from sawwort [72], plant-derived flavonol Identification of shellfish purple dye origin is the most
conjugates [94], newly elucidated anthraquinoid components difficult task. Some authors identify the dye source (a
of cochineal [65], and, as mentioned above, recently discov- particular shellfish species) according to the ratio of in-
ered indirubin analogues from shellfish purple [87] were digoid components [31, 53, 80, 83]. However, all quantitative
published. Simple MS methods without chromatographic presumptions are made using the relative area of chro-
separation and using direct ESI [96] or laser desorption– matographic peaks for a given wavelength without a proper
ionization MS [97, 98] can confirm the presence of a certain calibration procedure, as it is difficult to obtain standards that
dyestuff. contain all possible isomers. Further, significant diversity
It is worth noting that significantly worse repeatability of dye composition occurred even for the same species. The
(ten times worse SDs) of MS detection in comparison to DAD content of certain compounds in a dye obtained from the
has been reported [5]. It is not an exception that the RSD of H. trunculus L. species collected in different geographical lo-
MS detection results can reach 10% [5, 95]. cations (Spain and Israel) differed noticeably [31] (Supporting
Information Table S2). Sometimes the relative abundances
of dye components can overlap with other species. It is
3.4 Quantitative aspects and identification of hardly possible to distinguish between B. brandaris L. and S.
origin/source haemastoma L. (see Supporting Information Table S2). The
composition of dye depends on the sex of snails [76] and their
One of the most important issues in analysis of natural biological life phase. Probably, ancient dyers were aware of
historical colorants is the identification of their biological these differences since different composition alters the final
source and, thus, their possible geographical origin. A very shade of the colorant. They could also mix all together in
efficient method of analysis has to be developed. This should one dyeing vat without sex- or age-specific selection since
enable extraction, separation, identification, and quantitative all of the used snails produced the desired dye to a greater
estimation of major and trace components from extremely or lesser extent. Actually, mixtures of shells from different
small amounts of an original sample. In addition, without species are often found in coastal deposits [99]. Samples
appropriate biological reference material, the reliable identi- of a colorant taken from a 1000-year-old artifact (textile or
fication of either plant or animal source is impossible. Even painting) most likely will differ from the freshly prepared
its availability is not a guarantee of success. For example, dye. Indigoid bromo-derivatives in leuco-form undergo
although over 200 plants have been analyzed and their dye subsequent photodebromination under UV radiation, thus,
profiles were obtained, some of the textile extracts did not the ratio of mono-, di-, and nonbrominated compounds can
match any of them [49]. There are several dyestuffs that be dependent on the processing of the dye [82]. Therefore, it
contain the same major coloring matters, such as carminic is practically impossible to determine the source of shellfish
acid (cochineal species), alizarin and purpurin (madder and purple dye in case it was produced from mixture of snails
lady’s bedstraw), luteolin and apigenin (weld, sawwort, and or was affected by processing or physical environmental
dyer’s broom; see Supporting Information Table S1), or conditions [3].
bromoindigoids (shellfish species; Supporting Information Another issue in analysis of natural historical colorants
Table S2). Differentiation can be made on the basis of certain is the detection of degradation products of the primary dye
highly specific compounds usually present in very small components. Aging experiments followed by HPLC–MS
amounts. For example, chrysoeriol (3 -O-methylluteolin) is analysis showed that flavonols degrade to form molecules
a key marker for weld, genistein for dyer’s broom, or as it specific to each dye [5]. For example, 3,4-dihydroxybenzoic
was recently found, chlorogenic acid together with luteolin-, acid was detected in aged threads dyed with quercetin,
eriodictyol-, and diosmetin-O-glucuronides is a marker 2,4-dihydroxybenzoic acid in aged morin dyed threads, and
for sawwort. Another approach is to quantify the main 4-hydroxybenzoic acid in aged kaempferol-dyed threads [5].
components of the dye. Dyes obtained from different species Dye components can be traced by identification of their


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3408 V. Pauk et al. J. Sep. Sci. 2014, 37, 3393–3410

degradation products. However, another work contradicts The authors have declared no conflict of interest.
the notion that hydroxybenzoic acids can serve as molecular
markers for ageing of flavonoid dyes since they were found
in the freshly dyed yarns [6]. These acids can be produced 5 References
by the photo-oxidation of the chromophore-containing
molecules and the matrix as well. The authors state that only [1] Bechtold, T., Mussak, R. (Eds.), Handbook of Natural Col-
the presence of 4-hydroxybenzoic acid in yarn extracts is a orants, Wiley, Chichester 2009.
strong indication of ageing, but not di- and trihydroxybenzoic [2] Kvavadze, E., Bar-Yosef, O., Belfer-Cohen, A., Boaretto,
acids [6]. E., Jakeli, N., Matskevich, Z., Meshveliani, T., Science
2009, 325, 1359.
[3] Deviese, T., Ribechini, E., Baraldi, P., Farago-Szekeres,
B., Duday, H., Regert, M., Colombini, M., Anal. Bioanal.
Chem. 2011, 401, 1739–1748.
4 Conclusion and outlook [4] Gregory, P., in: Kirk-Othmer (Ed.), Encyclopedia of Chem-
ical Technology, 5th Edition, Vol. 9, Wiley, Hoboken 2007,
The analysis of historical dyes is an important and demanding pp. 238–300.
task. Investigation of natural colorants can provide valuable [5] Surowiec, I., Szostek, B., Trojanowicz, M., J. Sep. Sci.
cultural and historical information. Natural organic dyes 2007, 30, 2070–2079.
are complex mixtures of compounds belonging to various [6] Degano, I., Biesaga, M., Colombini, M., Trojanowicz, M.,
chemical families. Numerous examples from the literature J. Chromatogr. A 2011, 1218, 5837–5847.
demonstrate that HPLC is a useful technique for their [7] Colombini, M., Andreotti, A., Baraldi, C., Degano, I.,
analysis. Carefully elaborated HPLC separation is typically Lucejko, J., Microchem. J. 2007, 85, 174–182.
followed by DAD or MS detection. It can confirm the pres- [8] Craddock, P. (Ed.), Scientific Investigations of Copies,
ence and/or quantify a certain compound in natural organic Fakes and Forgeries, Butterworth-Heinemann, Oxford
dye or even identify unknown compounds. Besides char- 2009.
acterization of art objects, sources of used dyestuffs can be [9] Miliani, C., Rosi, F., Burnstock, A., Brunetti, B. G., Sgamel-
revealed. However, comparison of freshly prepared colorants lotti, A., Appl. Phys. A 2007, 89, 849–856.
with that from samples, which are hundreds of years old, can [10] Gulmini, M., Idone, A., Diana, E., Gastaldi, D., Vaudan,
be problematic and data have to be interpreted cautiously. D., Aceto, M., Dyes Pigm. 2013, 98, 136–145.
The difficulty of natural dyestuff analysis typically results [11] Vandenabeele, P., Weis, T., Grant, E., Moens, L., Anal.
from the limited availability of samples from historical or art Bioanal. Chem. 2004, 379, 137–142.
objects. Extremely small amounts (usually, the weight of a [12] Vandenabeele, P., Edwards, H., Moens, L. Chem. Rev.,
fiber used for analysis is <1 mg, and in the case of a sample 2007, 107, 675–686.
of paint, it is even smaller) are sampled. Tiny amounts of [13] Castro, K., Pessanha, S., Proietti, N., Princi, E., Capitani,
dyes have to be extracted in a way that preserves the integrity D., Carvalho, M., Madariaga, J., Anal. Bioanal. Chem.
of the analytes and their relative quantities. 2008, 391, 433–441.
Even with all of the improvements that were made in [14] Bruni, S., Guglielmi, V., Pozzi, F., J. Raman Spectrosc.
sample preparation, chromatographic separation, and detec- 2011, 42, 1267–1281.
tion during the last years, analysis of natural organic dyestuffs [15] Clementi, C., Miliani, C., Romani, A., Santamaria, U.,
still remains a challenging and inspiring task. It would be in- Morresi, F., Mlynarska, K., Favaro, G., Spectrochim. Acta
teresting to use HPLC for the evaluation of authenticity of A 2009, 71, 2057–2062.
historical artifacts, especially to investigate aging of natural [16] Aceto, M., Agostino, A., Fenoglio, G., Gulmini, M.,
colorants. The search for new specific components, markers Bianco, V., Pellizzi, E., Spectrochim. Acta A 2012, 91, 352–
359.
of authenticity, and characteristic for synthesized or natural
colorants would be also useful. From the point of view of [17] Amat, A., Miliani, C., Brunetti, B. G., J. Cult. Herit. 2013,
14, 23–30.
detection, taking advantage of new instruments, e.g. mass
spectrometers with high resolution and accuracy, or those [18] Surowiec, I., Microchim. Acta 2008, 162, 289–302.
with ion mobility separation units, is expected since they can [19] Surowiec, I., Pawelec, K., Rezeli, M., Kilar, F., Trojanowicz,
improve identification or overcome isobaric interferences and M., J. Sep. Sci. 2008, 31, 2457–2462.
support isomeric discrimination. [20] Lopez-Montes, A., Garcia, R., Espejo, T., Huertas-Perez,
J., Navalon, A., Vilchez, J., Electrophoresis 2007, 28,
1243–1251.
The authors gratefully acknowledge the support of the Czech [21] Fabbri, D., Chiavari, G., Ling, H., J. Anal. Appl. Pyrol.
Science Foundation (P206/12/1150), the Ministry of Education, 2000, 56, 167–178.
Youth and Sports of the Czech Republic, and the Operational [22] Andreotti, A., Bonaduce, I., Colombini, M., Ribechini, E.,
Program Research and Development for Innovations—European Rapid Commun. Mass Spectrom. 2004, 18, 1213–1220.
Regional Development Fund (project CZ.1.05/2.1.00/03.0058), [23] Chiavari, G., Fabbri, D., Prati, S., Zoppi, A., Chro-
IGA PrF 2014031. matographia 2005, 61, 403–408.


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
J. Sep. Sci. 2014, 37, 3393–3410 Liquid Chromatography 3409

[24] Casas-Catalan, M., Domenech-Carbo, M., Anal. Bioanal. [52] Petroviciu, I., Cretu, I., Vanden Berghe, I., Wouters, J.,
Chem. 2005, 382, 259–268. Medvedovici, A., Albu, F., Creanga, D., e-Preserv. Sci.
[25] Roelofs, W., ICOM Committee for Conservation, 4th Tri- 2012, 9, 90–96.
ennial Meeting, Venice, 1975, pp. 1–21. [53] Koren, Z., Verhecken-Lammens, C., e-Preserv. Sci. 2013,
[26] Wouters, J., Stud. Conserv.1985, 30, 119–128. 10, 27–34.

[27] Rosenberg, E., Anal. Bioanal. Chem. 2008, 391, 33–57. [54] Puchalska, M., Polec-Pawlak, K., Zadrozna I., Hryszko,
H., Jarosz, M., J. Mass Spectrom. 2004, 39, 1441–
[28] Degano, I., Ribechini, E., Modugno, F., Colombini, M. P., 1449.
Appl. Spectrosc. Rev. 2009, 44, 363–410.
[55] Colombini, M., Carmignani, A., Modugno, F., Frezzato, F.,
[29] Eikema Hommes, M., Dissertation, University of Amster- Olchini, A., Brecoulaki, B., Vassilopoulou, V., Karkanas,
dam 2002, pp. 109–166. P., Talanta 2004, 63, 839–848.
[30] Cooksey, C., Molecules 2001, 6, 736–769. [56] Sanz, E., Arteaga, A., Garcia, M. A., Camara, C., Dietz, C.,
[31] Koren, Z., Microchim. Acta 2008, 162, 381–392. J. Archaeol. Sci. 2012, 39, 3516–3523.
[32] Michel, R., McGovern, P., Archeomaterials 1987, 1, 135– [57] Blanc, R., Espejo, T., Lopez-Montes, A., Torres, D.,
143. Crovetto, G., Navalon, A., Vilchez, J. L., J. Chromatogr. A
[33] Benkendorff, K, Bremner, J., Davis, A., Molecules 2001, 2006, 1122, 105–113.
6, 70–78. [58] Sanz Rodriguez, E., Arteaga Rodriguez, A., Garcia Ro-
[34] Greenfield, A., A Perfect Red: Empire, Espionage, and the driguez, M., Egido, M., Camara, C., Bailao, A., Garcia,
Quest for the Color of Desire, HarperCollins Publishers, M., e-Conserv. Mag. 2010, 15, 32–45.
New York 2005. [59] Surowiec, I., Nowik, W., Trojanowicz, M., J. Sep. Sci.
[35] Koren, Z., in: Janaway, R., Wyeth, P. (Eds.), Scientific 2004, 27, 209–216.
Analysis of Ancient and Historic Textiles: Informing, [60] Surowiec, I., Nowik, W., Trojanowicz, M., Microchim.
Preservation, Display and Interpretation, Archetype Pub- Acta 2008, 162, 393–404.
lications, London 2005, pp. 194–201. [61] Surowiec, I., Quye, A., Trojanowicz, M., J. Chromatogr.
[36] Abel, A., in: Best, J. (Ed.), Colour Design: Theories and A 2006, 1112, 209–217.
Applications, Woodhead Publishing, Cambridge 2012,
[62] Novotna, P., Pacakova, V., Bosakova, Z., Stulik, K.,
433–470. J. Chromatogr. A 1999, 863, 235–241.
[37] Valianou, L., Karapanagiotis, I., Chryssoulakis, Y., Anal.
[63] Karapanagiotis, I., Icons: Approaches to Research, Con-
Bioanal. Chem. 2009, 395, 2175–2189.
servation and Ethical Issues, International Meeting,
[38] Marques, R., Sousa, M., Oliveira, M., Melo, M., J. Chro- Athens 2006.
matogr. A 2009, 1216, 1395–1402.
[64] Taujenis, L., Olsauskaite, V., Chemija 2012, 23, 210–215.
[39] Lech, K., Jarosz, M., Anal. Bioanal. Chem. 2011, 399,
[65] Stathopoulou, K., Valianou, L., Skaltsounis, A. L., Kara-
3241–3251.
panagiotis, I., Magiatis, P., Anal. Chim. Acta 2013, 804,
[40] Wouters, J., Grzywacz, C., Claro, A., Stud. Conserv. 2011, 264–272.
56, 231–249.
[66] Wouters, J., Verhecken, A., Stud. Conserv. 1989, 34, 189–
[41] Manhita, A., Ferreira, T., Candeias, A., Barrocas Dias, C., 200.
Anal. Bioanal. Chem. 2011, 400, 1501–1514.
[67] Wouters, J., Rosario-Chirinos, N., J. Am. Inst. Conserv.
[42] Degano, I., Lucejko, J., Colombini, M., J. Cult. Herit. 2011, 1992, 31, 237–255.
12, 295–299.
[68] Zhang, X., Laursen, A., Anal. Chem. 2005, 77, 2022–2025.
[43] Petroviciu, I., Albu, F., Medvedovici, A., Microchem. J.
[69] Deveoglu, O., Torgan, E., Karadag, R., Asian J. Chem.
2010, 95, 247–254.
2010, 22, 7021–7030.
[44] Petroviciu, I., Berghe, I., Cretu, I., Albu, F., Medvedovici,
[70] Deveoglu, O., Sahinbaskan, B., Torgan, E., Karadag, R.,
A., J. Cult. Herit. 2012, 13, 89–97.
Asian J. Chem. 2011, 23, 5441–5446.
[45] Sanyova, J., Microchim. Acta 2008, 162, 361–370. [71] Deveoglu, O., Torgan, E., Karadag, R., J. Liq. Chromatogr.
[46] Karapanagiotis, I., Minopoulou, E., Valianou, L., Daniilia, Relat. Technol. 2012, 35, 331–342.
S., Chryssoulakis, Y., Anal. Chim. Acta 2009, 647, 231–
[72] Lech, K., Witkos, K., Jarosz, M., Anal. Bioanal. Chem.
242. 2014, 406, 3703–3708.
[47] Karapanagiotis, I., Lakka, A., Valianou, L., Chryssoulakis,
[73] Bellaistre, M., Nowik, W., Tchapla, A., Heron, S., J. Chro-
Y., Microchim. Acta 2008, 160, 477–483. matogr. A 2011, 1218, 778–786.
[48] Serrano, A., Sousa, M., Hallett, J., Lopes, J., Oliveira, M.,
[74] Nowik, W., Bellaistre, M., Tchapla, A., Heron, S., J. Chro-
Anal. Bioanal. Chem. 2011, 401, 735–743.
matogr. A 2011, 1218, 3636–3647.
[49] Zhang, X., Laursen, R., Int. J. Mass Spectrom. 2009, 284,
[75] Papanastasiou, M., Allen, N., McMahon, A., Naegel, L.,
108–114.
Edge, M., Protopappas, S., Dyes Pigm. 2012, 92, 1192–
[50] Pawlak, K., Puchalska M., Miszczak, A., Rosloniec, E., 1198.
Jarosz, M., J. Mass Spectrom. 2006, 41, 613–622.
[76] Westley, C., Benkendorff, K., J. Chem. Ecol. 2008, 34,
[51] Petroviciu, I., Creanga, D., Melinte, I., Cretu, I., Medve- 44–56.
dovici, A., Albu, F., Rev. Roum. Chim. 2011, 56, 161–168.


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com
3410 V. Pauk et al. J. Sep. Sci. 2014, 37, 3393–3410

[77] Clark, R., Cooksey, C., J. Soc. Dyers Colour. 1997, 113, [89] Nowik, W., Marcinowska, R., Kusyk, K., Cardon, D., Tro-
316–321. janowicz, M., J. Chromatogr. A 2011, 1218, 1244–1252.
[78] Koren, Z., J. Soc. Dyers Colour. 1994, 110, 273–277. [90] Szostek, B., Orska-Gawrys J., Surowiec, I., Trojanowicz,
[79] Koren, Z., Dyes Hist. Archaeol. 1995, 13, 27–37. M., J. Chromatogr. A 2003, 1012, 179–192.

[80] Koren, Z., Isr. J. Chem. 1995, 35, 117–124. [91] Halpine, S., Stud. Conserv. 1996, 41, 76–94.

[81] Karapanagiotis, I., Am. Lab. 2006, 38, 36–45. [92] Yamaoka, R., Shibayama, N., Yamada, T., Sato, M.,
J. Mass Spectrom. 1989, 37, 249–253.
[82] McGovern, P., Michel, R., Anal. Chem. 1985, 57, 1514A–
1522A. [93] Joosten, I., van Bommel, M., Microchim. Acta 2008, 162,
433–446.
[83] March, R., Papanastasiou, M., McMahon, A., Allen, N.,
J. Mass. Spectrom. 2011, 46, 816–820. [94] Mouri, C., Mozaffarian, V., Zhang, X., Laursen, R., Dyes
Pigm. 2014, 100, 135–141.
[84] Serrano, A., van Bommel, M., Hallett, J., J. Chromatogr.
A 2013, 1318, 102–111. [95] Ackacha, M., Polec-Pawlak, K., Jarosz, M., J. Sep. Sci.
2003, 26, 1028–1034.
[85] Orska-Gawrys, J., Surowiec, I., Kehl, J., Rejniak, H.,
Urbaniak-Walczak, K., Trojanowicz, M., J. Chromatogr. [96] Pauk, V., Havlicek, V., Papouskova, B., Sulovsky, P., Lemr,
A 2003, 989, 239–248. K., J. Mass. Spectrom. 2013, 48, 927–930.
[86] Koren, Z., in: Meijer, L., Guyard, N., Skaltsounis, L., Eisen- [97] Kuckova, S., Hynek, R., Nemec, I., Kodicek, M., Jehlicka,
brand, G. (Eds.), Indirubin, the Red Shade of Indigo, Life J., Surf. Interface Anal. 2012, 44, 963–967.
in Progress Editions, Roscoff 2006, 45–53. [98] Ribechini, E., Perez-Arantegui, J., Colombini, M. P., J.
[87] Surowiec, I., Nowik, W., Moritz, T., Dyes Pigm. 2012, 94, Mass Spectrom. 2013, 48, 384–391.
363–369. [99] Reese, D., Mediterr. Archaeol. Archaeometry 2010, 10,
[88] Koren, Z., Dyes Hist. Archaeol. 2008, 21, 26–35. 113–141.


C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com