Flavonoid Dye: Flavonoid Textile Dyeing with a with Dyeing Textile

Home , Reseda luteola

Alexandre Villela chemistry methods chemistry photo-stability and analytical analytical and photo-stability flavonoid dye: flavonoid Textile dyeing with a with dyeing Textile

Textile dyeing with a flavonoid dye Alexandre Villela 2020 Alexandre Villela was Alexandre Villela born in Rio de Janeiro, Brazil in 1975. He studied chemistry in Brazil and the Netherlands: BSc at the Federal University of Rio de Janeiro, MSc at the Federal University of Santa Catarina, and PhD University. at Wageningen Alexandre is member of the International Society for Ethnopharmacology, and can be reached through alexandre.villela@ naturalproductschemistry. com. Propositions

1. There are limitations to the yellow colour of wool dyed with weld. (this thesis)

2. Using short UHPLC columns on conventional HPLC systems is an economical way of modernising HPLC-based analyses. (this thesis)

3. People and planet are better taken care of through organic agriculture than through conventional agriculture. (based on Reganold JP, Wachter JM. Nat. Plants 2016)

4. The release of genetically engineered organisms in natural, rural and urban areas is an ecocide.

5. The pursuit of excellence as human being must precede that of professional excellence.

6. Pride is fruit of short-sightedness.

Propositions belonging to the thesis entitled

Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods

Alexandre Villela Wageningen, 15 April 2020

Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods

Alexandre Villela

Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods

Alexandre Villela

Thesis committee

Promotor Prof. Dr J.T. Zuilhof Professor of Organic Chemistry Wageningen University & Research

Co-promotor Dr T.A. van Beek Assistant professor, Laboratory of Organic Chemistry Wageningen University & Research

Other members Thesis Prof. Dr J.H. Bitter, Wageningen University & Research submitted in fulfilment of the requirements for the degree of doctor Prof. Dr R. Verpoorte, Leiden University at Wageningen University Prof. Dr P.J. Schoenmakers, University of Amsterdam by the authority of the Rector Magnificus, Prof. Dr M.H.M. Eppink, Wageningen University & Research Prof. Dr A.P.J. Mol, in the presence of the Thesis Committee appointed by the Academic Board This research was conducted under the auspices of the graduate school VLAG to be defended in public (Advanced studies in Food Technology, Agrobiotechnology, Nutrition and Health on Wednesday 15 April 2020 Sciences). at 1:30 p.m. in the Aula.

Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods

Alexandre Villela

Thesis committee

Promotor Prof. Dr J.T. Zuilhof Professor of Organic Chemistry Wageningen University & Research

Co-promotor Dr T.A. van Beek Assistant professor, Laboratory of Organic Chemistry Wageningen University & Research

In memory of my grandmother Pérola (Voca)

Alexandre Villela Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods, 168 pages.

PhD thesis, Wageningen University, Wageningen, the Netherlands (2020) With references, with summary in English

ISBN 978-94-6395-334-4 DOI-link https://doi.org/10.18174/516631

In memory of my grandmother Pérola (Voca)

Alexandre Villela Textile dyeing with a flavonoid dye: photo-stability and analytical chemistry methods, 168 pages.

PhD thesis, Wageningen University, Wageningen, the Netherlands (2020) With references, with summary in English

ISBN 978-94-6395-334-4 DOI-link https://doi.org/10.18174/516631 Table of contents

Chapter 1. General introduction 9

Chapter 2. LC–UV method for the quantitation of the main flavonoids of weld 17 Chapter 3. Spectrophotometric comparison of the content of chlorophylls in weld 33

Chapter 4. Photo-stability of the dye of weld in presence of aluminium ions 43

Chapter 5. Analysis of a natural dye: an experiment for analytical organic chemistry 69

Chapter 6. General discussion 77

Summary 87

Acknowledgements 91

Publications and training activities 95

Appendix A. Chapter 2: Supplementary information 99

Appendix B. Chapter 3: Supplementary information 103

Appendix C. Chapter 4: Supplementary information 109

Appendix D. Chapter 5: Supplementary information 147

Table of contents

Chapter 1. General introduction 9

Chapter 2. LC–UV method for the quantitation of the main flavonoids of weld 17 Chapter 3. Spectrophotometric comparison of the content of chlorophylls in weld 33

Chapter 4. Photo-stability of the dye of weld in presence of aluminium ions 43

Chapter 5. Analysis of a natural dye: an experiment for analytical organic chemistry 69

Chapter 6. General discussion 77

Summary 87

Acknowledgements 91

Publications and training activities 95

Appendix A. Chapter 2: Supplementary information 99

Appendix B. Chapter 3: Supplementary information 103

Appendix C. Chapter 4: Supplementary information 109

Appendix D. Chapter 5: Supplementary information 147

Chapter 1

General introduction

Chapter 1

General introduction

1.1. Introduction Coloured textiles have been used by mankind throughout times [1]. Dyes have been obtained from different natural sources, including plants, molluscs and insects [1] and were valuable in the past [2]. Although the use of natural dyes experienced a fast decline from the 1850s with the first synthetic dyes entering the market [3, 4], there has been a renewed interest. Currently, this is due to innovation-related, economical, personal, and ethical reasons [5]. Figure 1.1 depicts yarns dyed with natural dyes in various colours.

The contents of this chapter are, to a large extent, part of the following papers:

Villela A, van der Klift EJC, Mattheussens ESGM, Derksen GCH, Zuilhof H, van Beek TA. Fast chromatographic separation for the quantitation of the main flavone dyes in Reseda luteola (weld). J Chromatogr A 2011; 1218(47): 8544–50

Villela A, Derksen GCH, Zuilhof H, van Beek TA. Spectrophotometric comparison of the content of chlorophylls in weld (Reseda luteola). Anal Methods 2011; 3(6): 1424–7

Villela A, van Vuuren MSA, Willemen HM, Derksen GCH, van Beek TA. Photo-stability of a flavonoid dye in presence of aluminium ions. Dyes Pigments 2019; 162: 222–31

Villela A, Derksen GCH, van Beek TA. Analysis of a natural yellow dye: an experiment for analytical organic chemistry. J Chem Educ 2014; 91(4): 566–9

Figure 1.1. Yarns of wool dyed with natural dyes.1

In addition, identification of dyestuff of historical textiles is of historical and conservational interest [2]. Figure 1.2 depicts a live mannequin wearing a costume dyed with a natural dye.

1 Dyed by Judy Hardman (United Kingdom); picture taken in La Rochelle/France (2011).

10 11

10 Educanalytical Chem 91(4): 2014; 566–9 chemistry. organic J for experiment an dye: yellow a natural of Analysis TA. Beek van GCH, Derksen A, Villela Dyesflavonoid ions. 222–31 Pigments dyeof162: presence 2019; aluminium in HM, Derksen Willemen BeekVillela GCH, A, van van MSA, Vuuren TA. content of weld chlorophyllsin ( Zu GCH, Derksen A, Villela Chromatogr(weld). 1218(47):2011; 8544–50 J A Reseda in luteola of the dyes mainflavone theFast quantitation separationfor chromatographic Mattheussens ESGM,ZuilhofBeek DerksenEJC, GCH, H, TA. vanVillela A, van Klift der The content

s , to a large extent, part of thepapers: following extent, alarge to , are chapter this of ilhof H,ilhof TA. van Spectrophotometric Beek comparison of the Reseda luteola Reseda ). ). Anal Methods 2011; 3(6): 1424–7 Anal 3(6): Methods2011; - Photo

stabilitya of depicts this is due is this innovationto - from different and natural molluscs insects plants, including sources, the first synthetic dyes entering market the [ [ past the [ times throughout mankind by have used been Coloured textiles 1.1. Introduction 1 interest [ In identification of of dyestuffis historical textiles addition, of historical and conservational 1.1.YarnsFigure of wool dyed natural with dye

Dyed by Hardman Judy (United Kingdom); picture Lataken in Rochelle/France (2011). yarns naturalwith colours. dyes dyedvarious in 2] 2] . Figure a a naturalcostume livemannequin dyed 1.2depicts wearing adye. with . Although dye natural the use of related, [ and economical, personal, ethical reasons 3, 4] s. s

1 experienced a fast decline from the 1850s with the with 1850s fast from decline experienced a

, there has been a renewed interest. a renewed been has there , 1] . D

[ 1 yes have been obtained obtained been have yes ]

and were valuable in in valuable were and

5 ]

. Currently, Currently, Figure 1.1 11

Chapter 1 Figure 1.2. Valérie Fischbach acting as a live mannequin OH 3' wearing a costume in Incroyables et Merveilleuses style 2' OH 2 4' dyed with weld (Reseda luteola L.). 8 1 8a 1' HO O 5' 7 2 6' 6 4a 3 5 4

OH O

Figure 1.3. Structure of the flavone luteolin (lut), the aglycone of the main flavonoids of weld (Reseda luteola L.).

RP-HPLC–UV (reversed phase high-performance liquid chromatography with

spectrophotometric detection in the ultraviolet range of the electromagnetic spectrum) is a very

suitable technique for the separation and detection of these non-volatile moderately polar UV-

active analytes, and is available in most laboratories. Various assays have been published for

the quantitation of flavonoids in weld, by different analytical techniques. RP-HPLC–DAD (RP-

HPLC–diode-array detector) has been used to quantify the aglycones quercetin, luteolin, and

apigenin in weld after simultaneous extraction-hydrolysis of the glycosides [10]. Other groups

used RP-HPLC–UV to quantify only luteolin [11, 12]. Cristea et al. [13] extracted weld with

methanol–water 8:2 at room temperature and under reflux and, afterwards, analysed the three

main flavones by the same technique with external standardisation. Moiteiro et al. [14] also

used methanol–water 8:2, but extracted the flavonoids from weld by 10 min sonication at room

′ temperature. In addition to ldg, lmg, and lut, they also quantified luteolin-4 -O-glucoside. Since there was no simple, validated RP-HPLC–UV method for the quantitation of ldg, lmg, and lut in samples of weld, the development of such a method—and its validation for the purpose— was aimed at. The method that was developed is described in chapter 2. It should be noted, however, that the use of natural dyes for textiles is not synonymous with According to the industrial partner of the project, one of the problems to overcome was a environmentally friendly dyeing [5]. Different factors must be addressed for it to be an greenish hue that accompanies the yellow colour after dyeing alum (dodecahydrate aluminium ecologically sound industrial activity [6]. Naturally, techno-economic aspects of production potassium sulphate) pre-treated textiles with weld. Such a greenish hue would be undesirable, also require attention [7]. and it was hypothesised that chlorophylls a and b are potential sources of it. Thus, for the Weld (Reseda luteola L.) used to be an important vegetable source of dye for textiles in purposes of raw material characterisation, the development of a simple analytical method for Europe and the Mediterranean area [1]. The main colouring compounds of the plant belong to the quantitation of chlorophylls in samples of weld was aimed at. The method that was developed envisages to compare the content of porphyrin ring-containing pigments in different the class of flavonoids [1]. The flavone luteolin (lut) and two of its glycosyl conjugates, lut-7- weld plant materials, and is described in . O-glucoside (lut monoglucoside, lmg) and lut-7,3ʹ-O-diglucoside (lut diglucoside, ldg), are chapter 3 Weld has been used to dye both animal fibres—such as wool—and vegetable fibres, after the main flavonoids of the plant [8] (Fig. 1.3). Seven other minor ones include the aglycones 3 apigenin and chrysoeriol, together with mono and diglycosides of all three aglycones [8, 9]. they have been treated (mordanted ) with alum in facultative combination with the dye-assistant tartar [1]. Alum-premordanted wool dyed with weld leads to yellow colours and, generally, the yellow colours of textiles dyed with natural dyes are poorly resistant to light [1]. Considering today’s requirement of photo-resistance (light-fastness) of the colours of dyed textiles, this is

2 Les Incroyables et Merveilleuses was a French fashion trend of the end of the 18th century; made by 3 The word mordant comes from the Latin word mordere, which means “to bite”. Metal salts have been students preparing for a degree in arts and crafts at lycée polyvalent Jules Verne de Sartrouville—a high referred to as mordants in relation to a classical model of binding of dyes to fibres involving the metals school in France—together with their teachers; picture taken in La Rochelle/France (2011). [1].

12 13

school in France— students preparing for a degree the mainflavonoids of the plant[ apigenin and chrysoeriol, together with mono and diglycosidesaglycones togetherallthreeapigenin and mono andwith of [ chrysoeriol, 12 2 O also requireattention [ environmentally dyeing [ friendly It be should noted, however, that the use is ofdyes natural not textiles for synonymous with flavonoids flavonoids of class the Europe and the ecologically sound industrial activity [ Les -

glucoside ( Weld ( Weld Incroyables et Merveilleuse et Incroyables Reseda luteola Reseda lut Mediterranean area area Mediterranean

together with their teachers; monoglucoside, 7] [ 1] .

. The( luteolin flavone L.) used to be an important vegetable source of dye for textiles in in textiles for dye vegetable of source an important be to L.) used

in artsin andcrafts at s

was a wea Figure 1.2. dyed weld with ( 8] lmg 5] [

ring costume a 1] (Fig. 1.3). . French fashion trend of the end of the 18th century; made by ) and lut and ) . Different factors m factors Different 6] The the belong plant maincolouringcompounds of to . - techno Naturally, picture taken La in Rochelle/France (2011) Valérie Fischbach acting Fischbach Valérie lycée polyvalent

- Seven other minor ones include the aglycones include other the aglycones ones minor Seven 7,3ʹ Reseda luteola lut

) and two of its glycosyl conjugates,glycosyl lut its of ) and two in in - O Incroyableset Merveilleuse - diglucoside ( ust beust addressed forbe to an it

Jules Verne de Sartrouville— economic aspects of production aspects production of economic L.) . 2

as a l as lut

diglucoside, diglucoside, ive mannequin s style .

ldg 8, 9]

a high high a ), are are ), . - 7 - methanol used RP apigenin in weld after simultaneo HPLC - today’srequirement of photo main flavones standardisation. technique Moiteiroet bythe same external al. with analytical RP different techniques. weld, in by flavonoids the of quantitation analytes active suitable technique f RP flavonoids of weld ( luteolin ( 1.3.Figure 3 yellow colours of textiles dyed with natural dyesare poorlyresistant to light [ tartar RP validated simple, no was there Intemperature. to addition ldg used methanol spectrophotometric potassium sulphate)potassium pre greeni method that was The developed at. described is aimed was samplesdevelopmentin of weld, for validation ofmethod—and the such its the a purpose the characteris material raw of purposes and was it hypothesis weld plant materials developed they have been treated (mordanted they been have [1] referred to asmordants relin HO The word mordant comes from word mordere Latin the . According thepartner to industrial of one the project, of the problems Weld

HPLC of weld ofquantitation chlorophyllsin samples weld of 7 6 [ sh hue that accompanies the the accompanies that hue sh –diode 1] 5

- lut OH . 8 has been used to been dyehas animal both HPLC –water Alum –UV envisages envisages ), ), the aglycone the of main 4a 8a Structure of flavone the - array detector) has been used quantify to the quercetin, aglycones and luteolin, detector) array water 8:2, but extrac but –water 8:2, O O , and available is in most laboratories 1 –UV quantify to [ luteolin only -

premordanted wooldyedyellow weld with leads to and, generally, colours the 8:2 at room temperature and under reflux and, afterwards,analysed temperatureand atreflux the8:2 room three under 4 (reversed phase high phase (reversed Reseda luteola L.).

or the separationandof these detection the non or detection detection 3 2 1' , and 2' to compare the content of porphyrin of the content compareto ed that chlorophylls that ed - OH treated textiles with weld. 6' described in described is 3' ation a to classical model bindingof dyes of fibresto involving metals the 5' 4' in , resistance (light resistance OH lmg

the the us extraction us - 3

HPLC ) with alum yellow colourafter dyeing alum aluminium (dodecahydrate , and lut ultraviolet

ted the flavonoids from weld by 10 min sonication at room fromted weldsonication at the room flavonoids by10min ation, chapter 3 chapter –UV method for the quantitation of ldg of quantitation for the method –UV the development of a analytical simple method for , they quantified- luteolin also a fibres - performance liquid chromatography with chromatographyperformance with liquid and and - - range of the electromagnetic sp electromagnetic the range of in in fastness) of the colours of the of dyed this textiles, is fastness) hydrolysis [ of the glycosides 11, facultative such as wool —such b

. 12] Such agreenish beSuch undesirable hue would , which means bite”. “to Metal salts have been are . Various assays have been published for beenassays for published Various have . Cristea Cristea . was was potential source ring chapter 2 chapter in combination with the dye combination with aimed at aimed - in different different in containing pigments - et al. volatile moderately polar UV and vegetable fibres, after after fibres, vegetable —and

[ . 13 . The The

4′ ] s

extracted weld with - of it. was overcometo was - O HPLC 10 method - ectrum glucoside. 1 ]

. Other groups ] , lmg . Thus, Considering DAD –DAD

) - [ that was was that assistant , and lut , and is a very very a is 14] for the the for Since

(RP also also — 13 a - - ,

Chapter 1 also the case for the dye of weld. Therefore, the work reported in chapter 4 aimed at finding 1.2. References out whether:4 3+ [1] Cardon D. Natural dyes – sources, tradition, technology and science. 1st ed. Archetype • a progressive decrease of [Al ] would progressively increase the photo-stability of the Publications: London; 2007. dye of weld; • the endogenous glycosidase of weld should be inactivated before extracting the [2] Ferreira ESB, Hulme AN, McNab H, Quye A. The natural constituents of historical textile flavonoids of the plant in order to obtain the most photo-stable dye. dyes. Chem Soc Rev 2004; 33(6): 329–36. [3] Cardon D. Colours in civilizations of the world and natural colorants: history under tension. Because of its wide applicability, high-performance liquid chromatography (HPLC) is one of In: Bechtold T, Mussak R, editors. Handbook of natural colorants, John Wiley & Sons: the analytical techniques with which the undergraduate student of life sciences should be Chichester; 2009, p. 21–6. acquainted. This is also true for the internal standard method [15], which is important in [4] Abel A. The history of dyes and pigments: from natural dyes to high performance pigments. quantitative instrumental analysis [16, 17]. The experiment described in chapter 5 aims at In: Best J, editor. Colour design – theories and applications, Woodhead Publishing: Cambridge; familiarising students with both. This is accomplished using the real-world application of 2012, p. 433–70. natural dyes for textiles. Weld, the plant used in the experiment, and wool dyed by students with its extract are depicted in Figure 1.4. [5] Mussak RAM, Bechtold T. Natural colorants in textile dyeing. In: Bechtold T, Mussak RAM, editors. Handbook of natural colorants, John Wiley & Sons: Chichester; 2009, p. 315–37. [6] Ganglberger E. Environmental aspects and sustainability. In: Bechtold T, Mussak R, editors. Handbook of natural colorants, John Wiley & Sons: Chichester; 2009, p. 353–66. [7] Geissler S. Economic aspects of natural dyes. In: Bechtold T, Mussak RAM, editors. Handbook of natural colorants, John Wiley & Sons: Chichester; 2009, p. 367–84. [8] Marques R, Sousa MM, Oliveira MC, Melo MJ. Characterization of weld (Reseda luteola L.) and spurge flax (Daphne gnidium L.) by high-performance liquid chromatography–diode array detection–mass spectrometry in Arraiolos historical textiles. J Chromatogr A 2009; 1216(9): 1395–402. [9] Peggie DA, Hulme AN, McNab H, Quye A. Towards the identification of characteristic minor components from textiles dyed with weld (Reseda luteola L.) and those dyed with Mexican cochineal (Dactylopius coccus Costa). Microchim Acta 2008; 162(3-4): 371–80. [10] Hartl A, Vogl CR. Faser-und Färbepflanzen aus Ökologischem Landbau. Anbauversuche Färbepflanzen. Endbericht zur Projekterweiterung Teil B (Ertragsleistung und Farbstoffgehalt Figure 1.4. Patch of weld (Reseda luteola L.) and alum- von Färber-Resede (Reseda luteola L.), Färberkamille (Anthemis tinctoria L.) und pretreated wool dyed with extract of weld. Färberknöterich (Polygonum tinctorium Ait.) auf Biobetrieben in Niederösterreich im Vergleich mit praxis-und handelsüblichen Warenpartien). Project L 1043/96, Ministry of Agriculture, Forestry, Environment and Water Management and Environment & Ministry for Education, Science and Research: Vienna; 2000. [11] Cerrato A, De Santis D, Moresi M. Production of luteolin extracts from Reseda luteola and assessment of their dyeing properties. J Sci Food Agric 2002; 82(10): 1189–99.

[12] Angelini LG, Bertoli A, Rolandelli S, Pistelli L. Agronomic potential of Reseda luteola L. as new crop for natural dyes in textiles production. Ind Crops Prod 2003; 17(3): 199–207. [13] Cristea D, Bareau I, Vilarem G. Identification and quantitative HPLC analysis of the main flavonoids present in weld (Reseda luteola L.). Dyes Pigments 2003; 57(3): 267–72.

[14] Moiteiro C, Gaspar H, Rodrigues AI, Lopes JF, Carnide V. HPLC quantification of dye flavonoids in Reseda luteola L. from Portugal. J Sep Sci 2008; 31(21): 3683–7.

4 The rationale for these aims is presented in section 4.1.

14 15

with its extract are depicted dyed bystudents wool the and in experiment, the plantused Weld, textiles. natural for dyes familiarising quantitative instrumental analysis [ method[ standard internal for true the also acquainted. is This tech analytical the Because wide of applicability,high its 14 4 pretreated wool Figure whether:out weld. of dye the for case the also The rationale for these aims is presented in section 4.1. presented in is aimsrationale these for The

• • flavonoids of theorder plantin photo- obtainthe to most the weld; dye of a progressive decrease of decrease a progressive 1.4. endogenous glycosidase of weld should be inactivated before extracting the the extracting before inactivated weldglycosidase be endogenous should of

Patch of weld ( weld of Patch 4

students with both. This is accomplis is This both. with students dyed with extract of weld. extract of with dyed

niques with which with the undergraduateniques sciences of student life be should L.) and alum and L.) luteola Reseda in Figure in 1.4.

[Al

T 3+ herefore, t herefore, 16, ] wouldprogressively]

- performance liquid chromatographyperformance liquid one(HPLC) is of 17]

in in described experiment . The he work reportedhe i hed using the real -

increase the- photo increase stable dye stable n 15] chapter 4 chapter , which is important in - . world application ofworld

chapter 5 chapter

aimed at finding stability of the

aims at aims [8] Marques [8] Wiley Chichester;Handbook 2009,p.367–84. John & ofcolorants, natural Sons: In: Geissler Economic S. Bechtold editors. RAM, [7] aspects T, Mussak ofdyes. natural Wiley Chichester;Handbook 2009,p.353–66. John & ofcolorants, natural Sons: editors. Mussak R, In: Bechtold T, sustainability. Ganglberger and aspects Environmental [6] E. editors. Handbook Wileyp. 315–37. ofcolorants,Chichester; natural 2009, John Sons: & Becht Mussak RAM, [5] 2012, p.433–70. flavonoids present( weld in Cristea[13] D, BareauVilaremI,Identification G. and quantitative HPLC analy Ind production.as new 199–207. for crop dyes textiles 17(3): natural in 2003; Prod Crops Ange [12] Foodassessment Sci ofAgric their 82(10): dyeing J 2002; properties. 1189–99. ( L.)andflax spurge editor. In:Best Colour J, design – Abel[4] A. The history dyes natural of from anddyes pigments: high to performance pigments. Chichester; 2009,p.21–6. In: Bechtold editors. T, Handbook Mussak R, of natural colorants, Wiley John & Sons tension. under historyand colorants: of natural the world civilizations Cardon in D. Colours [3] dyes. 33(6): 329–36. SocRev Chem 2004; constituentsof historical natural The textile A. Quye H, McNab AN, Hulme ESB, Ferreira [2] London;Publications: 2007. Cardon[1] D. Natural dyes – 1.2. References flavonoids Reseda in luteola GasparMoiteiro C, [14] CarnideLopes H, RodriguesJF, V. HPLC of dye quantification AI, extractsfrom of luteolin Cerrato D,[11] Moresi Production Santis M. A, De Research:Education, Scienceand 2000. Vienna; Agriculture, and Water Forestry, Management Environment and E Vergleich mitpraxis ( cochineal Mexican dyed ( textiles weld componentswith minor from characteristic of identification the Towards A. Quye H, McNab AN, Hulme DA, Peggie [9] 1216(9): 1395–402. detection array Färberknöterich ( Färberknöterich von Färber Projekterweiterung zur Färbepflanzen. undFarbstoffgehalt Endbericht Teil B (Ertragsleistung Hartl[10] A, Vogl Faser CR. lini LG,liniBertoli A, Rolandelli Pistelli S, L.Agronomic luteola potential Reseda of -

R, Sousa MM, Oliveira MC, Melo MJ. Characterizationweld ( of MJ. Oliveira Melo MC, MM, Sousa R, Resede ( Resede –mass spectrometryArraiolos in historical textiles. J Chromatogr A 2009;

Polygonum tinctorium Ait.) auf tinctorium BiobetriebenPolygonum Niederösterreich in im Dactylopius coccusDactylopius Daphne gnidium - und handelsüblichenL Warenpartien). Project Ministry 1043/96, of Reseda luteola luteola Reseda old T. Naturalold dyeing. colorants textile in In: T, Mussak RAM, Bechtold Reseda luteola L. 31(21): 2008; SepSci 3683–7. from J Portugal. - und Färbepflanzen ÖkologischemLandbau.und aus Anbauversuche sources, tradition, technology tradition, sources, ed.1st Archetype and science. theories and applications, Woodhead Publishing:theoriesapplications, Cambridge; and

L.) by high- Costa). Microchim Acta Costa). Microchim 2008; L.), Färberkamille ( L.). 267–72. Pigments 57(3): 2003; Dyes

performance liquid chromatography performance liquid Reseda luteola Ant nvironment & Ministrynvironment for hemis tinctoria L.) and those dyed with those dyedL.) with and 162(3- Reseda luteola 4): 371–80. Reseda luteola Reseda sis of the main main the sis of L.) und –diode –diode

and and

15 L. L. :

Chapter 1 [15] Magee JA, Herd AC. Internal standard calculations in chromatography. J Chem Educ 1999; 76(2): 252. Chapter 2

[16] Barrows RD. Quantitative comparison of three standardization methods using a one-way

ANOVA for multiple mean comparisons. J Chem Educ 2007; 84(5): 839–41.

[17] Harvey D. Analytical Chemistry 2.0. http://www.asdlib.org/onlineArticles/ecourseware/Analytical%20Chemistry%202.0/Text_File s.html [accessed August 2013].

LC–UV method for the quantitation of the main flavonoids of weld

The content of this chapter is largely that of the following paper: Villela A, van der Klift EJC, Mattheussens ESGM, Derksen GCH, Zuilhof H, van Beek TA. Fast chromatographic separation for the quantitation of the main flavone dyes in Reseda luteola (weld). J Chromatogr A 2011; 1218(47): 8544–50.

[15] Magee JA, Herd AC. Internal standard calculations in chromatography. J Chem Educ 1999; 76(2): 252. Chapter 2

[16] Barrows RD. Quantitative comparison of three standardization methods using a one-way

ANOVA for multiple mean comparisons. J Chem Educ 2007; 84(5): 839–41.

[17] Harvey D. Analytical Chemistry 2.0. http://www.asdlib.org/onlineArticles/ecourseware/Analytical%20Chemistry%202.0/Text_File s.html [accessed August 2013].

LC–UV method for the quantitation of the main flavonoids of weld

2.1. Introduction In view of the expectation of having to analyse hundreds of samples in the course of the academic and industrial dye research of which this project is part, developing a method suitable for the simultaneous analysis of the three main flavones of the plant with as little manpower input as possible and using standard equipment was aimed at. According to the industrial partner of the project, ldg, lmg, and lut constitute about 80% of the total flavonoid content of R. luteola and the ratio between them and the minor flavonoids is fairly constant. Thus, quantifying only these three flavones, while being informative, keeps the LC-based method simple and cheap. A validated RP-HPLC–UV method for the quantitation of ldg, lmg, and lut that satisfies the demand of the project in terms of low manpower input is reported in this chapter. Moreover, A simple validated industrially-usable quantitative method to assess the flavone content of R. the chromatographic step on a 250 mm × 4.6 mm 5 µm-particle size HPLC column was speeded luteola samples is described in this chapter. The flavones were overnight-extracted from the up and made low in solvent consumption by using a 50 mm × 3.0 mm 1.8 µm-particle size dried and ground aerial parts of the plant at room temperature via maceration with methanol– UHPLC column, while still using a conventional HPLC system. water 8:2. Afterwards, they were quantified through internal standardisation against chrysin by RP-HPLC–UV at 345 nm. The efficiency of the one-step extraction was 95%. The limits of detection (LOD) and quantitation (LOQ) were ≤1 ng and ≤3 ng, respectively, providing ample 2.2. Material and methods sensitivity for the purpose. The precision expressed as relative standard deviation of the entire 2.2.1. Plant material, chemicals and reagents method was <6.5% for the combined content of luteolin-7,3'-O-diglucoside, luteolin-7-O- Five plants were withdrawn from a bunch of dried Reseda luteola L. (cultivar code: H) plants. glucoside, and luteolin. The average absolute recovery (accuracy) at three spiking levels was The plants used were grown in the North of the Netherlands and were harvested on August 102% (range: 98–107%) and the relative recovery ranged from 99 to 102%. The separation was 2007. A voucher specimen was deposited at the Wageningen branch of the National Herbarium initially carried out on a traditional 250 mm × 4.6 mm 5 µm-particle size HPLC column (80 of the Netherlands (Gen. Foulkesweg 37, 6703 BL Wageningen, The Netherlands) and min run time, 35.9 mL MeOH). It was then speeded up by the use of a 50 mm × 3.0 mm 1.8 identified as W. Olsder s. n. (WAG, barcodes WAG0248296–WAG0248298). Drying took µm-particle size UHPLC column (5 min run time, 1.4 mL MeCN), while still using a place by leaving the plants outdoors during daytime and indoors during the nights, for ~10 days. conventional HPLC system. Whereas the retention times on the UHPLC column were relatively The aerial parts of the plants were ground with a cutting mill (Retsch GmbH, type SM1, less reproducible, cross-validation showed that the quantitation of luteolin-7,3'-O-diglucoside, Haan/Germany; 0.25 mm sieve). This combined sample is henceforth referred to as R. luteola luteolin-7-O-glucoside, and luteolin was not statistically significantly different, with H, and was stored in the dark. comparable precision. The method is more sensitive when using the UHPLC column than when Methanol (“HPLC grade” or similar), DMSO (99.9% for spectroscopy, Acros Organics), using the HPLC column. The analytical method described meets the demand for a very small and acetonitrile (HPLC grade or p.a.) were used without further purification. Ultrapure water manpower input per sample and uses standard laboratory equipment. Using a UHPLC column was prepared with an EasyPure UV system (Barnstead/Thermolyne, Dubuque/USA) or on a conventional HPLC system is a way of modernising HPLC-based phytochemical analyses purchased from Merck (Lichrosolv water for chromatography). Demineralised water (Ph. Eur.) inexpensively. was purchased from VWR or prepared with a Seradest SD 2000 ion-exchanger (Seral Erich Alhäuser GmbH, Ransbach-Baumbach/Germany). Formic acid was from Acros Organics (98+%) and VWR (AnalaR Normapur 99–100%), ammonium formiate was from Aldrich, Fluka or VWR, EDTA (dihydrate, either di- or tetrasodium) was from either Aldrich or VWR, and luteolin (lut) used for spiking was from Sigma (≥98%, TLC grade). Standard compounds luteolin-7,3'-di-O-glucoside (luteolin diglucoside, ldg), luteolin-7-O-glucoside (luteolin monoglucoside, lmg), and luteolin (lut) were all of analytical control grade (Extrasynthese). Their purity was determined by 1H-NMR in DMSO-d6 (99.5+ atom % D, Aldrich) with a 400 MHz NMR spectrometer (Bruker, Avance III, Fällanden/Switzerland) as 88, 94, and 98 and 94% for ldg, lmg, and lut (Extrasynthese and Sigma, respectively), respectively. Maleic acid (puriss., ≥99.0 % by HPLC, Fluka) was used as internal standard for the analyses by NMR [1]. Chrysin (97% grade) was from Aldrich. G24 Environmental and Innova 4080 Incubator Shakers (New Brunswick Scientific Co. Inc., Edison/USA) were used. A 15-stirring points

18 19

min run time, 35.9 102%and (range: 98–107%)the r glucoside initially inexpensively. on column a UHPLC equipment. Using laboratory standard and uses sample manpower per input small avery for demand the meets described method analytical using The column. the HPLC more methodis sensitivewhen The thecolumnthan UHPLC when using precision. comparable luteolin less conventional µm method was < sensitivity for the purpose. detection (LOD) (LOQ) and quantitation were ≤ RP chrysin by against standardisation internal through quantified were they Afterwards, water8:2. ma temperature via at room the plant of groundaerial parts dried and luteola 18 A

simple validated industrially a - - HPLC particle size reproducible, cross conventional

- samples 7- carried out on a traditional 250 a on out carried , –UV at 345nm. The efficiency one of the O and The luteolin. a -

glucoside HPLC system HPLC

6.5%

is described in this chapter this in described is

UHPLC column (5 min run column(5UHPLC min 1.4mLtime, MeCN) HPLC systemHPLC a modernising way of is HPLC mL MeOH)

- the combined luteolin for content of , - validation showed thatthe of quantitation validation and . tion of thetion entire devia standard relative as expressed precision The Whereas verage absolute recovery (accuracy) at three spi three at (accuracy) recovery absolute verage luteolin - . usable quantitative method to assess the flavone content of content flavone assess methodto the quantitative usable It elative recovery ranged 102% 99to elative from recovery

was then speeded then was

the the

was was mm r etentiontimes on the UHPLC column were relatively . The flavones overnight were The × not not

4.6 mm 5µm mm 4.6 1 ng≤ and statistically

up - step extraction was 95%. was extraction step by of a50 the use 3 ng, respectively, providing ample 7,3'- -

size particle - d base significantly different, O luteolin - - luteolin diglucoside, ceration with methanol with ceration phytochemical analyses analyses phytochemical , while . - - 7,3'

The separation was was separation The HPLC column (80 column (80 HPLC extracted from the the from extracted mm king levels was was levels king - O

× The limits of - still using a

diglucoside 3.0 mm 1.83.0 mm

7- with with

O R. R. – - , 94% for ldg for 94% partner of ldg the project, t inpu for theanalysis simultaneous of the three mainflavones of the plantwith as little manpower projectwhich part this is of research dye industrial and academic course of the the in of samples hundreds analyse to expectation of having the of In view Shakers (NewBrunswick Scientific Inc., Co. Edison/USA) Chrysin (puriss., ≥ LC thequantifying keeps flavones, three informative, these only while being R. luteola 2.1. Introduction III, spectrometer NMR 88, 94,and (Bruker,MHz Fällanden/Switzerland) as Avance dark.H, and the in was stored luteola as R. referredsample henceforth to is combined sieve). Haan/Germany; This 0.25mm wereplants groundGmbH, a typeThe with cutting(Retsch aerial parts ofmill SM1, the during the nig and indoors outdoorsduring daytime plants place the byleaving identified Olsder as n.(WAG, W. s. barcodes–WAG0248298). WAG0248296 FoulkeswegNetherlands) BLof and the Netherlands (Gen. Wageningen, 37,6703 The 2007. The used plants weregrown the North in of the Netherlands and were harvested onAugust Reseda luteola from ofFivewithdrawn a dried bunch were plants reagentsmaterial,2.2.1. Plant and chemicals t demand of the project terms in of lowmanpower input reportedis this in chapter. cheap. and simple Alhäuser GmbH, Ransbach GmbH, Alhäuser or preparedSeradestwas VWR - purchased a2000ion with SD from (Lichro Merck from purchased Dubuque/USA) or (Barnstead/Thermolyne, EasyPure system an UV was with prepared further without grade Ultrapurewerewater p.a.)purification. used and (HPLC or acetonitrile methods 2.2. Material and column UHPLC consumption solvent up andlowin made Their Their monoglucoside, luteolin and ( luteolin di (dihydrate,either EDTA Fluka VWR, or Norma(98+%) (AnalaR and VWR he step chromatographic A v Methanol (“HPLC grade” or si or grade” (“HPLC Methanol purity purity was was A voucher specimen as and possible usingequipment standardwas aimed at. alidated alidated - 7,3' (97% grade) was fromgrade) Aldrich.(97% was G24 Fluka) was used as internal standard internal as used was Fluka) HPLC, 99.0 % by and the ratio between themand the flavonoids minor is fairly constant. - , was determined by by determined was di lut lmg - RP O , while ) spiking used for was lmg - ,

- e ( glucosid HPLC and ) , and luteolin ( luteolin and still using still a conventional lut UV –UV on a 250 mm × mm on a 250 ,

(Extrasynthese and Sigma, respectively), respec respectively), Sigma, and (Extrasynthese lmg

- Baumbach/Germany). Formic acid was from Acros Organics Acros from was acid Formic Baumbach/Germany).

luteolin solv watersolv water chromatography). for Demineralised Eur.) (Ph. method for the quantitation of ldg , and lut deposited at the Wageningen branch of the Nat the of branch Wageningen the at deposited 1 H - milar), NMR NMR lut was formiate from ammonium pur Aldrich, 99–100%),

from Sigma (≥ Sigma from ) were all of analytical control grade (Extrasynthese). (Extrasynthese). grade control analytical of all were )

diglucoside, constitute aboutconstitute 80% of theflavonoid total content of 4.6 mm 5 µm 5 4.6 mm - DMSO in DMSO (99.9% for (99.9% DMSO Organics) spectroscopy, Acros - by by or tetrasodium) was from either Aldrich or VWR, or Aldrich either from was or tetrasodium) using

HPLC system HPLC Environmental and Innova 4080 Incubator InnovaEnvironmental and 4080

d6 ( a - 98%, TLC grade).98%, compounds TLC Standard ldg particle size HPLC column 50 mm× 50 99.5+ atom % D, Aldrich) with a 400 aAldrich) 400 with % D, atom 99.5+ ), luteolin

were used were . , developing ,

for for 3.0 mm 1.8 µm 3.0 mm lmg According to the industrial theAccording to industrial L. (cultivar code: H) plants. L. H) code: plants. (cultivar by NMR NMR by analyses the - , and lut 7- exchanger (Seral Erich Erich (Seral exchanger O . - A glucoside (luteolin tively. Maleic acid acid Maleic tively.

15- a method suitable a methodsuitable hts, for ~10 days. ~10 days. for hts,

ional Herbarium that satisfies the - s based method based method tirring points -

Drying took was was particle size Moreover, Moreover, speeded speeded 98 Thus, Thus, and and [ 1] 19 , .

Chapter 2 magnetic stirrer (Variomag Poly 15, H+P Labortechnik, Munich/Germany) was used. The column, which now acted as the mixer and connected damper and injector; and (ii) shortening sonication bath used was from Elma, type T 700 (Germany). of the tubing from the injection system to the column and from the column to the detector. The injected volume was 2 µL. The column temperature was 35 oC (measured with a digital 2.2.2. Sample preparation thermometer, model 3150, PeakTech). The DAD scan range was 245–500 nm and detection (200.0 ± 0.9) mg of dried R. luteola H were weighed (Mettler AE 260 balance) in 50 mL was carried out at 345 nm (bandwidth 4 nm), having 470 nm (bandwidth 40 nm) as reference Erlenmeyer flasks. A magnetic stir bar and 20 mL of methanol–water 8:2 (v/v) were added to λ. The following solvents were used as eluent: solvent A was the same as for the HPLC column, each flask, after which they were closed with a rubber stopper. The flasks were placed either whereas solvent B was acetonitrile. The flow was 0.90 mL min−1 and the following linear on a 15-stirring point magnetic stirrer or on individual magnetic stirrers. Samples were extracted gradient was used: 85–45% A (0–4.00 min), 45–85% A (4.00–4.01 min) and 85% A, isocratic for 16 h at room temperature, at 300 rpm. Then, 5.00 mL of chrysin stock solution were added (4.01–5.00 min). Injections were made every 5 min. to the Erlenmeyer flask and stirring continued for another 10–14 min. Subsequently, an aliquot of the solution was filtered via a 0.45 µm syringe filter (Minisart RC4, Sartorius Stedim Biotech, 2.2.5. Calibration curves, linearity and determination of relative response factors (RRFs) Goettingen/Germany)/polypropylene syringe (with rubber stopper) to a glass HPLC For the HPLC column, the stock solutions were prepared as follows: (i) 125 mg of chrysin autosampler vial, discarding the first drops. (internal standard, i.s.; MW: 254.24 g mol−1) were dissolved in methanol (0.5 mg mL−1; 2.0 mM); (ii) 2.0 mg of ldg (MW: 610.52 g mol−1) were accurately weighed and dissolved in 2.2.3. LC I (HPLC column) methanol–dimethyl sulfoxide 7:3 (v/v); (iii) 3.6 mg of lmg (MW: 448.38 g mol−1) were HPLC analyses in Wageningen were performed on a Waters system consisting of a binary pump accurately weighed and dissolved in methanol–dimethyl sulfoxide 7:3 (v/v); and (iv) 2.4 mg of (1525 µ), a photodiode array detector (DAD) (996) and an autosampler (717 plus), and lut (MW: 286.24 g mol−1) were accurately weighed and dissolved in methanol. In the cases of equipped with a column oven model 5CH. An Alltima 250 mm × 4.6 mm C18 5 µm-particle i.s. and lut, the dissolution was assisted by the use of a shaker at room temperature. Generally, size column was used, in combination with a C18 guard column. The volumes of the syringe the exposure of the solutions of the flavones to light was minimised. As 10 mL volumetric and loop installed were 250 and 200 µL, respectively. The system was controlled by Waters flasks were used for ldg, lmg, and lut, the concentration of the stock solutions were 176 µg Empower 2 software. The draw rate of the autosampler was 1 µL sec−1 and the injected volume, mL−1 (0.29 mM), 341 µg mL−1 (0.76 mM), and 233 µg mL−1 (0.81 mM), respectively (taking 20 µL. Mixtures of methanol–water and methanol only were used as needle wash solvent (the into account the purity determined by NMR). The i.s. stock solution was stored at 4 oC. The latter is preferred for keeping sample carryover below 1%). One of the relative recovery dilutions were made by adding 1.00 mL of the i.s. solution and different volumes of each of the experiments was carried out in Steenbergen, on an analogous system (see section SI A.1.1, stock solutions to 5 mL volumetric flasks. Five concentrations for ldg and lmg, and seven for Appendix A). At both locations, a pH 3 buffer of 160 mM formic acid, 40 mM ammonium lut (for which a 10x diluted stock solution was additionally required) were prepared by the use formiate and 0.04 mM of EDTA in water (A), and methanol (B) were used as solvents. The of volumetric pipettes, ranging as follows: (i) ldg: 7–141 µg mL−1; (ii) lmg: 14–205 µg mL−1; flow was 1.00 mL min−1 and the following linear gradient was used: 85–60% A (0–35 min), and (iii) lut: 2–116 µg mL−1. This was done separately for each of the flavones. In all cases: (i) 60–36% A (35–47 min), 36–23% A (47–60 min), 23–0% A (60–65 min), 0% A (65–70 min). only methanol was added to the mark and (ii) each solution was injected once. For the UHPLC Subsequent injections could be made after 10 min of re-equilibration. The column was kept at column, the stock and diluted solutions were prepared as above, except that 2.4 mg of ldg, 3.7 40 oC. The DAD was set to scan from 245 to 500 nm (Wageningen) and 200 to 595 nm mg of lmg, and 2.3 mg of lut were weighed (thus, taking into account the purity determined by (Steenbergen). The peak areas were based on the absorbance at 345 nm. LC–DAD–MS for peak NMR, the concentrations of the stock solutions were: 215 µg mL−1 or 0.35 mM, 340 µg mL−1 purity checking was performed in Wageningen with the same column on a Thermo Scientific or 0.76 mM, and 230 µg mL−1 or 0.80 mM, respectively) and the concentrations of the diluted system consisting of an Eppendorf CH-30 column oven, a Finnigan Surveyor autosampler plus, solutions ranged as follows: (i) ldg: 9–150 µg mL−1; (ii) lmg: 14–238 µg mL−1; and (iii) lut: 2– MS pump plus, DAD plus, and a Finnigan LXQ linear ion trap MS with ESI probe working in 115 µg mL−1. In all cases, each solution was injected once. (+)-ionisation mode in combination with Xcalibur software (version 2.0.5). Conditions were basically the same as above, except that solvent A contained no EDTA. About 10% of the flow 2.2.6. Evaluation of the extraction efficiency was directed to the MS. R. luteola H was extracted according to the normal procedure (above). However, the work-up was done differently. The extract was filtered under reduced pressure by the use of an ordinary 2.2.4. LC II (UHPLC column) 60o filtration funnel and a suction flask. The filtrate was collected in a test tube placed inside Analyses were performed on an Agilent system consisting of a binary pump (Agilent 1200 the flask. It was then transferred to a 25 mL volumetric flask and methanol–water 8:2 (v/v) was Series), a diode array detector (DAD), an autosampler, and a column oven (all Hewlett-Packard, added to the mark. The solution was syringe-filtered and analysed by HPLC. No i.s. solution Series 1100). An Agilent XDB-C18 50 mm × 3.0 mm 1.8 µm-particle size column without was added as the experiment was run on a relative basis and the results were calculated based guard column was used. The system was controlled by HP ChemStation for LC 3D (Rev. on the sum of the areas of the ldg, lmg, and lut peaks. The filter paper was transferred back to A.06.03) software. The following modifications were done to reduce the internal volume of the the Erlenmeyer flask. The powder of the plant material was removed from the filter paper with system: (i) the mixer was replaced by a 7.5 mm × 3.0 mm Alltima C18 (5 µm-particle size) pre- the aid of 20 mL of freshly added methanol–water 8:2 (v/v). The same extraction and work-up

20 21

system: (i) the mixer was A.06.03) software. The m following guard Series 1100). An Agilent XDB Agilent An Series 1100). min 1.00mL flow was formiate Series), a diode array detector (DAD), an autosampler an (DAD), detector array a diode Series), Agilent syst onan performed Analyses were 2.2.4. LC II ( MS. the to directed was EDTA no contained A solvent that except above, as same the basically (+) plus, pump MS DAD plus, and a FinniganLXQ linear ESI trapwith ion MS working probe in CH an Eppendorf consistingsystem of purity Wageningen checkingwasin the same performed ona column Thermo with Scientific (Steenbergen).areas peakwere ontheat absorbance345nm. The based A Appendix experiment recovery relative the of One 1%). below carryover sample keeping for preferred is latter 40 of20 µL. methanol Mixtures 20 Subsequent could injections be made after 10m 60–36% A (35 36–23%–47 min), A(60 A (47 23–0% –60 min), sonication bath Munich/Germany) Labortechnik, H+P 15, Poly (Variomag stirrer magnetic Erlenme ( 2.2.2. Sample preparation draw rate of the autosampler was 1 µL 1 was autosampler the of rate Empower draw 2software. The trolled con was system 250 and by respectively. The 200µL, and were installed loop Waters size equipped a with columnoven 250 An model5CH. Alltima plus),and (717 autosampler (996) an and (DAD) array detector photodiode (1525 a µ), pump abinary of consisting system a Waters on performed were Wageningen in analyses HPLC ( I 2.2.3. LC autosampler the first drops. vial,discarding to rubberGoettingen/Germany)/polypropylene (with stopper) syringe syringe 0.45 µm via a filtered was of the solution t to were added chrysin stock solution of mL 5.00 rpm. Then, at 300 temperature, at room for 16h on a 15- were placedeither flasks The rubber stopper. closed a with were they which after flask, each 200.0 - he Erlenmeyer stirring flask and continuedanother for Subsequently, 10–14min. aliquot an o ionisation mode in ionisation C. C. column was used, in a combinationwith C18 guard column.The of volumes the syringe column was used. The system was controlled by HP ChemStation for LC Thecolumn was by for system used. controlled ChemStation HP was 3D (Rev. ± 0.9) ± The DAD was set to scan to Thewas from DAD set 245 stirring point magnetic stirrer or on individual magnetic stirrers. Samples were extracted yerflasks. A magnetic stirand barof 20mL methanol and of 0.04mM EDTA water in (A) s

) HPLC was carried out in Steenbergen,an analogous (see in system on out carried was .

UHPLC column mg of dried of mg At both locationsAt both used was from used was column) −

combination with Xcaliburcombination with software 1

replaced by replaced 60% A (0 A 85–60% was gradient used: linear following and the

R. luteola

water and methanol only were used as needle wash solvent ( solvent wash needle as used were only methanol and –water Elma, )

- , C18 50 C18 3 buffer of 160 mM formi mM 160 of buffer 3 a pH odifications type T type

a 7.5 - 30 column oven, a Finnigan Surveyor autosampl Surveyor Finnigan oven, a column 30 H were weighed (Mettler AE 260 balance) in 50 mL mL 50 in balance) 260 AE (Mettler weighed were H mm mm

00 (Germany). 700 em consisting of a binary pump (Agilent 1200 (Agilent 1200 pump aconsisting binary of em × , to to ×

3.0 mm 1.8µm mm 3.0 m and in of re in of to reduce to the internal volumeof done thewere filter (Minisart RC4, Sartorius Stedim Biotech,

3.0 mm Alltima3.0 C18 (5 µm- 500 nm (Wageningen)500 nm 200 and , and and ethanol (B - equilibration. The columnwas kept at a column oven (all Hewlett (all oven column a

mm (version 2.0.5) (version water 8:2 –water

70 min). 0%–65 min), A (65–70 min). - sec × size particle ) were used as) solvents were used

c acid, 40 mM ammonium ammonium 40 mM c acid, 4.6 mm 5µm C18 − 1 and the injected volume, . About . LC

MS for peak peak for –DADMS ( ) were added to to added were ) v/v

. Conditions were. Conditions section section particle si particle

a glass HPLC HPLC a glass

% of the flow the 10% of column without column without was used was to to 35 min), –35 min), - SI A.1 SI Packard, Packard, - 595 nm 595 nm ze particle particle er plus, plus, er . ) pre . The The The the the .1, - flasks were used for used were flasks min 0.90mL was acetonitrile. flow B The whereas was solvent λ. reference as 40nm) (bandwidth nm 470 nm), having (bandwidth 4 at 345nm was out carried PeakTech). thermometer, 3150, model the aid of 20 mL of fr the aid of 20mL paper with filter from the removed was plantmaterial the of powder The flask.the Erlenmeyer ofon the sum the areas of the ldg b arelative on run was experiment the as added was and the andi.s. lut methanol in dissolved and weighed accurately For curves,2.2.5. Calibration linearity and determination of relative response factors (RRFs) Injections were(4.01–5.00 min). every 5min. made (0 A gradient used: 85–45% was 3 was temperature column injectedµL. volumewas The 2 theto detector. The column and from the the column to injection system of from the tubing the column, which now acted as the mixer and connected damper and injector was syringe The mark. solution added the to Itthe flask. was then a transferred to 25mL volumetric flask and methanol 60 ordinary an of use the by pressure reduced under filtered was extract The differently. done was R. luteola efficiency the extraction of 2.2.6. Evaluation 115 µg mL ranged ( solutions as follows: or NMR lmg of mg only lut 5mL to stockvolumetric solutions flasks.concentrations Fivefor ldg weredilutions made byadding 1.00mL the i.s. of accountinto the mL lut methanol mM) 254.24gMW: mol (internal ; i.s. standard, column, of volumetric pipettes, ranging as follows: ( The were solvents following e used as o 0.76 mM

− stock solution diluted a 10x which (for (MW: 286.24g(MW: mol exposure of the solutions ofexposure of the the solutions flavones to light was minimised filtration funnel and a suction flask. The filtrate was collected in a test tu ( the the 1

; iii) methanol was added to the mark the to added was methanol , ( 341 µg mL µg 0.29 mM),341 ( the concentrationswere: of the stockmL solutions 215µg ii) HPLC HPLC lut the stock and diluted solutions wereasexceptthe above, stock and solutions diluted prepared –dimethyl sulfoxide7:3 , H a shaker at room temperature. room at , the dissolution wasof ashaker assisted bythe use 2.0 mg of2.0 mg ldg − : 2 and 2.3mg lut of , 1 . In all cases, each solution was injected once. injected was solution each cases, In all . was extracted according to the normal procedure procedure normal the to according extracted was and –116 µg mL column,

purity byNMR) determined 230 µg mL

ldg eshly added methanol added eshly − 1 the stock solutions werepreparedthe stock solutions as follows ,

) lmg − (MW: 610.52g(MW: mol were accurately weighed and dissolved in methanol. In the case andmethanol. in accurately dissolved were weighed 1 . This was. This done separate −

i) i) − were weighed (thus,taking account into the purity by determined 1 , 1 or or ldg and ( 0.76 mM),and 4.01 min) and 85% A, isocratic 85% A, isocratic and A (4.00 –4.01 min) 45–85% –4.00 min), , 0.80 mM, respectively)0.80 mM, and : 9 lmg ( lut v/v); v/v); –150 µg mL

and (ii) each solution was injectedsolution and once. (ii) each , , the concentration of the stock solutions were of the, thestock solutions concentration 176µg luent: solvent A solvent wasluent: 245 was range scan DAD The and ( was additionally required additionally was iii) 3.6 mg of lmg − water 8:2 ( 8:2 –water i) i) 1 lut - ) were dissolved in methanol (0.5 mL mg methanol in dissolved ) were . The i.s. –dimethyl sulf i.s. No HPLC. by analysed and filtered ldg

− peaks. The filter paper was transferred back to to back transferred was paper filter The peaks. 1 233 µg mL µg 233 ) − : 7 solution and differentsolution of volumes each the of 1 were accurately weighed and dissolved in in dissolved and weighed accurately were ; –141 µg mL ly for each of the flavones. In all cases: (i) (i) cases: In all flavones. the of each for ly ( asis and the results were calculated based based calculated were results the and asis ii) stock solution was stock solution stored at 4 v/v). The same extraction andextraction work- v/v). The same lmg

−

: 14–238µg: mL oxide 7:3 7:3 oxide t 1 he same as for the HPLC column, column, HPLC the for as same he the concentrations of the dilutedthe concentrations ( 5 (

0.81 mM), respectively (taking 0.81 mM),respectively(taking above (MW: 448.38g(MW: mol − o 1 − C (m C ; 1 − 1 ( or or ) were prepared prepared were )

ii) and the linear following ) ( . However,- the work

: . As 10mL volumetric 0.35 mM, 340µg0.35 mM, mL detection and nm detection –500 v/v); andv/v); and and

easured with a digital adigital with easured lmg that ( i) 125 mg ofi) chrysin water 8:2 ( 8:2 –water ; and; (ii) shortening lmg : 14–205µg: mL

− 2.4 mg of ldg 1 ; and; ( be placed inside inside placed be For the , ( and seven for and for seven iv) 2.4 mg of of mg iv) 2.4 Generally, Generally, iii) by the use use by the − v/v) was v/v) was UHPLC UHPLC solution solution o 1 C. TheC. ) lut − 1 were were ; 2.0 2.0 ; , 3.7 , 3.7 : 2 s

up up up up 21 − of of − 1 – 1 ;

Chapter 2 procedures were carried out twice more. This experiment was carried out thrice. The extraction Thus, the spiking levels were: 40–61% (low), 103–119% (medium), and 131–149% (high), times of each cycle were: 1st 16 h; 2nd 24–26 h; and 3rd 53–56 h. roughly corresponding to 50, 110, and 140% of the flavonoid content that is present in the plant material. The recovery (in %) was calculated as [3]: 2.2.7. Precision of the entire method {[(amount found) – (amount in plant)] / amount spiked} × 100% For the HPLC column, three series of seven replicates of the usual extraction were carried out (sample = R. luteola H). They occurred on days 1, 3 and 5. Those of days 1 and 5 were carried 2.2.9. Stability of samples ready for analysis out with a 15-stirring point magnetic stirrer, whereas those of day 3, by the use of the individual All extracts of series 1 of the HPLC column precision experiment (above) were collected in magnetic stirrers. For the UHPLC column, eight replicates of the usual extraction were carried four HPLC vials. One set of vials was loaded for analysis in the HPLC shortly after the work- out (sample = R. luteola H). After storage of samples at 4 ºC or −20 ºC, they were analysed on up procedure. The second set of vials was also placed in the auto-sampler (room temperature days 5, 5, and 6. and dark), but analysed 24 hrs after the previous series. The third set was placed at 4 oC, and the fourth, at −20 oC. They were analysed 72 h and 15 days after the first set of vials, 2.2.8. Recovery of the method respectively. 2.2.8.1. Relative recovery experiments, including evaluation of the stability of the compounds under the extraction conditions 2.2.10. Data/statistical analysis The relative recovery experiment was conducted similarly to that described earlier [2]. In short: Values of the results presented were generally rounded up according to their uncertainties pre-purified samples were subjected anew to the entire analytical procedure. In this case, a 10- (standard deviation values), expressed with one significant digit. Unless stated otherwise, stirring points magnetic stirrer (RO 10 P, IKA Werke, Staufen/Germany) was used. The areas statistical analyses were carried out using the Mann–Whitney U-test (α = 5%, two-sided test) of the peaks before and after the analytical procedure, including extraction, were compared. [4]. Also, extracts were filtered through filter paper Whatman 42 and brought to a final volume of 25.0 mL with methanol–water 8:2 (v/v). In each case, the area of the peaks of ldg, lmg, and lut of syringe-filtered and non-syringe-filtered (above) samples were compared (n = 3). The 2.3. Results and discussion experiments were carried out with three different amounts of plant material: ∼100, 200, and 300 2.3.1. Extraction mg. As the assay was expected to be used in an industrial setting for hundreds of samples during many years, the amount of labour per sample was of eminent importance. Earlier, a slow 2.2.8.2. Absolute recovery by spiking experiment (accuracy) maceration-type extraction with introduction of i.s. before sampling followed by RP-HPLC Known amounts of each flavonoid of interest were added to the plant material (spiking). Then, proved itself very labour-efficient in the analysis of 750 Taxus samples [5, 6]. A similar the extraction and work-up proceeded as usual. Due to the poor solubility of the glycosides approach (overnight maceration with stirring) was evaluated for R. luteola flavones. (lmg and ldg) in methanol, water, and their mixtures, this required a specific approach. Each Next, the solvent for extraction had to be selected. Cristea et al. [7] compared different flavonoid standard was accurately weighed and transferred to a round-bottom flask with solvents for the extraction of ldg, lmg, and lut from the aerial parts of R. luteola, with methanol– methanol, which was then evaporated in vacuo. Methanol–water 8:2 (v/v) was added to the water 8:2 (v/v) giving the highest yield. The same authors observed 10–30% lower extraction residue. This was followed by simultaneous agitation and sonication during 10 min. This yields of the flavonoids after extraction with methanol–water 8:2 (v/v) for 4 h at room solution was analysed (n = 3) with the internal standard method to determine the absolute temperature than after 15 min of refluxing [7]. This is consistent with our preliminary amount of each flavone present. Next, 20.00 mL of the solution was added to 200 mg of plant observations using a solvent–sample ratio of 20, with which—even after ~18 h of extraction at material, followed by standard extraction, work-up, and analysis (above). The amounts of ldg, room temperature—lower yield of the flavones were obtained than after 15 min of reflux. lmg, and lut present in 200 mg of R. luteola H were 707 µg (average) (3.3% = relative standard However, when using a solvent–sample ratio of 100, the yields were observed not to be deviation), 1,298 µg (3.9%), and 195 µg (5.4%), respectively (n = 12). The spiking was carried statistically significantly different (Mann-Whitney U-test,  = 10%). This fits with the out at three levels, as follows: low (two triplicates), medium (triplicate) and high (triplicate). observation by Cerrato et al. [8] that the yield of lut, upon extraction with methanol at room The added absolute amounts of ldg, lmg, and lut in the spiking experiments were: temperature, reaches an equilibrium after ~16 h. Considering that the aim was to develop a (i) 392 µg (0.7%), 618 µg (0.2%), and 78.1 µg (0.6%), respectively, in the first triplicate of method requiring as little as possible manpower input per sample, the overnight extraction at level low; room temperature is clearly advantageous over the 15 min extraction by reflux. The extraction, (ii) 341 µg (0.2%), 623 µg (0.6%), and 118.2 µg (0.5%), respectively, in the second when carried out overnight and using a solvent–sample ratio of 100, gave 95% efficiency. A triplicate of level low; second and a third extraction yielded 4 and 1%, respectively. A fourth extraction step with (iii) 726 µg (0.5%), 1,369 µg (0.4%), and 232.7 µg (0.4%), respectively, in the level medium; additional sonication yielded no additional flavones. Performing a single extraction in this (iv) 929 µg (0.4%), 1,923 µg (0.2%) and 290.8 µg (0.3%), respectively, in the level high. manner saves time, chemicals, and manpower input. As over 99% of the flavonoids are extracted in the first two cycles and sr (relative standard deviation) values are low, it is an option

22 23

Also, extracts were filtered through filter paper Whatman 42 and brought to a final volumeof a to 42andWhatman brought filter paper through filtered extractswereAlso, compared. were extraction, including procedure, analytical the after and before peaks the of stirring points magne 22 of ldg Theabsolute added amounts lmg material, followed work extraction, bystandard amount of eachpresent. 20.00mL Next, was flavone of added the solution 200mg to of plant absolute the determine methodto standard internal the with = was 3) (n analysed solution residue. was This followed Methanol vacuo in . evaporated then was which methanol, round a to transferred and weighed accurately was standard flavonoid ( the and extraction work- (spiking). Then, plantmaterial the to added flavonoid were of interest each Known of amounts experiment recovery (accuracy)2.2.8.2. Absolute spiking by mg. experiments were carriedout with three different amounts of plant mate syringe of methanol with 25.0 mL pre [ earlier described that to similarly conducted was experiment recovery relative The compounds under the extraction conditions 2.2.8.1. Relative recoveryexperiment 2.2.8. Recovery the method of days 5, (sampleR. luteola out = magnetic stirrers. a with out 15- carried and 5were 3and daysThey 5.Those 1 days of occurred 1, on H). =R. luteola (sample For the HPLC column, method entire the of Precision 2.2.7. 1st were: cycle each of times extraction The thrice. out carried was experiment This more. twice out carried were procedures vels le three at out spiking carried was 12). The (n = respectively (5.4%), 195µg and (3.9%), deviation), 1,298µg lmg (iii) (iv) (ii) (i) -

purified samples were subjected anew to the entire analytical procedure. In this case, a 10- a case, In this procedure. analytical entire the to anew subjected were samples purified ,

and and and triplicate of level low; level low; 392µg (0.7%), 618µg (0.2%)

929 µg(0.2%) 1,923µg high. (0.4%), the respectively, and (0.3%), level in 290.8µg (0.4%) 726 µg 1,369µg (0.5%), (0.6%) µg (0.2%),341 µg 623 lut ldg and 6. - filtered and non- andfiltered

R. luteola of mg R. 200 in present specific approach. Each Each approach. ) in methanol,a water, specific required mixtures,this and their stirring point m point stirring

UHPLC column For UHPLC the , as follows , as The areas areas used. The was IKA Staufen/Germany) Werke, (RO 10P, tic stirrer – three series of seven replicates of the usual extraction were carried out out carried were extraction usual the of replicates seven of series three water 8:2 ( 8:2 water H). After storage of samples at 4 ºC or −20 ºC, they were analysed theywere ºC, on −20 or at 4ºC samples H). of Afterstorage up proceededDue the asto poor usual.solubility of the

16 h; 2nd 24–26 h; and 2nd24–26h; 16h; 3rd 53–56h. agnetic stirrer, whereas those of day 3, by the use of the individual individual of the use 3, by the day those of whereas agnetic stirrer, syringe by simultaneous agitation andby sonication during agitation simultaneous This 10min. : low (two triplicates), medium (triplicate) andhigh(triplicate).

, lmg v/v). In each case, the area of the peaks of ldg the peaks of area Ineach case, the v/v). - , and 78.1µg (0.6%), respectively, the in first triplicate of filtered (

, and 232.7 µg (0.4%), respectively, in the level medium; respectively,the (0.4%), in level 232.7µg and , and lut ,

s eight replicates of the usual extraction were carried carried were extraction usual the of replicates , eight and 118.2 µg (0.5%), respectively, in the second respectively, the(0.5%), second in µg and 118.2 ,

H were 707 µg (average) (3.3% = relative stand relative = (3.3% (average) µg 707 were H including

above - up, up,

in thein spiking were: experiments and analysis ( analysis and ) samples were compared (n = 3). The The = 3). (n compared were samples ) evaluation of the the of stability water 8:2 ( 8:2 –water

above rial: ). The of amounts ). ldg v/v) was added was to v/v) - ∼ bottom flask with 100, 200, , lmg

2] glycosides glycosides . In. short: , and 300 and 300 and ard ard the the lut

water 8:2 8:2 water observations using a solvent of refluxingtemperature15 min [ than after after wit extraction yieldsflavonoids of the extracted in the first two cycles and s and cycles two first the in extracted chemicals time, saves manner yieldedadditional sonication noadditionalflavones. this in extraction Performing a single second and a anda using solvent when overnight out carried room temperature clearly advantageous is over the as l method requiring afterequilibrium hes an ~16that h.Considering reac temperature, observation by al. et Cerrato statistically significantly different ( ro forsolvents the of extraction ldg maceration years,many the amountof labour per sample was ofrtance. eminent impo [ V 2.2.10. Data/s respectively. However, evaluatedapproach stirring) with was (overnight flavones. R. luteola maceration for proved itselfvery labour As the assay was 2.3.1. Extraction 2.3. Results and discussion Mann the using out carried were analyses statistical (standard deviation values −20 at fourth, the Thus, thewere:–119% spiking (medium) (low), levels 103 40–61% All extracts of samples2.2.9. Stability ready analysis for spiked} amount / plant)] in found) ×{[(amount –(amount 100% ( recovery The material. roughly corresponding and dark),24 hrs analysed after but the previous s up procedure. The set of second was vials also placed the in auto - four HPLC of was Onevials vials. loaded set for analysisthe in HPLCafter shortly the work - 4] eir un eir upaccording th generallyrounded to presented were results the of alues om temperature . Next, the for solvent Next, extraction had be to selected. Cristea et al. ( when using a solvent when a using v/v) giving the highest yield. authorsgiving same the The highest v/v) -

type extraction with introduction of of introduction with extraction type of series 1 of the HPLC column HPLC the of 1 series of

tatistical analysis third extraction yielded extraction third respectively. 4and 1%, step with A fourth extraction

lower yield of the flavones were obtained than after 15 min of reflux. 15 min yieldafter were obtained than flavones of—lower the expected to be used in an industrial setting for hundreds of samples during of setting samples be during to usedfor an in expected industrial hundreds o C. They were analysed 72 h analysed TheyC. were 72 ittle as possible manpower input per sample, the overnight extraction at at extraction perthe overnight input sample, manpower as possible ittle to in % in

50, 110, - efficient in the analysis of 750 Taxus samples of 750Taxus efficient the in analysis [ ), expressed one with significant digit. Unless stated otherwise, even after ~18 h ~18 –sample ratioafter of which— 20,with even

)

was calculated as calculated was [ 8] , ,

lmg and manpower input. As over 99% of the flavonoids are are flavonoids of the 99% As over input. and manpower the 100,the –sample ratio of thatyield the of and 140% of the flavonoid content that is present in the plant and flavonoid 140% content present of thatis the the in plant r

, Mann (relative standard deviation) values deviation) standard (relative and lut - Whitney

from the aerial parts of R. luteola of parts aerial the from

precision experiment ( experiment precision

h methanol [ sample ratio of 100, gave 95% efficiency ratio of 100, –sample 7] 3] i.s. . This is consistent with our preliminary consistentwith is . This our preliminary lut eries. The third set was placed at 4 4 at placed was set third The eries. 15 min : –Whitney

and 15 days after first the set vials, of , uponextraction methanol with at room before sampling followed byRP before sampling U - test, extraction byreflux. extraction water 8:2 ( 8:2 –water

30% lower extraction extraction lower observed 10–30% were yields  α U

- = 10%) test (α = (α test sampler (room temperature the above ,

[

149% (high), and 131–149% (high), v/v) for 4h for v/v) 7]

aim was to develop a develop to was aim are low,itis an are option . This fits with. the

observed %, two 5%, compared different different compared ) were collected in in collected were ) 5, , with methanol, with of extraction at at extraction of

Earlier, a slow The extraction, 6 ] . A similar - sided test) certainties certainties not to be be to not at room at room o - C HPLC HPLC , and .

23 A –

Chapter 2 to correct the initial outcome by multiplying it by a factor of 100/95 to arrive at the true flavone content. In the tables, uncorrected values are presented. The manual labour input of the whole analytical procedure is minimal, consisting of weighing the previously ground plant material, adding solvent and stir bar, placing the flask on a magnetic stirrer, adding the i.s. solution, and syringe-filtering the sample into the HPLC vial.

2.3.2. Chromatography The validation was performed on a traditional HPLC column. The solvents used with this 5 µm- particle size RP-HPLC 250 mm × 4.6 mm column were: (i) an aqueous buffer of pH 3 (solvent A) [9], and (ii) methanol (solvent B). The detection wavelength was selected to be 345 nm in view of the absorbance maxima of ldg (340 nm), lmg (350 nm), and lut (350 nm). No sample preparation other than filtration proved necessary (Fig. 1, top). The total run time, including equilibration, was 80 min, and 35.9 mL of methanol were needed. From the results of the precision experiment, retention times were observed to be precise, varying at most 0.4% over a period of 5 days. Within one day, the sr values of the retention times varied from 0.004 to 0.05%. The use of a 1.8 µm-particle size UHPLC 50 mm × 3.0 mm column saves time and solvents. The UHPLC column could be run with a flow of 0.9 mL min−1 on a conventional HPLC system after minor adaption of the hardware and tubing, and changing the B solvent to acetonitrile (Fig. 1, bottom). This shortened the run time to 5 min, and only 1.4 mL of acetonitrile was needed. The resolution between ldg and lmg (Rs = 5.4) was lower than on the HPLC column (Rs = 9.2), but still more than sufficient for reliable integrations. Although in absolute terms the retention time variation was lower on the UHPLC column, relatively (i.e., precision), they were observed to be 25 to 180× (i.s. and ldg cases, respectively) higher than with the 5 µm-particle size HPLC column. However, never did this influence the quantitation (Table 2.4). The higher relative retention time variation is likely caused by working near the maximum pressure of the HPLC pump (400 bar) and, additionally, by the much lower absolute retention times (ldg: varying from 0.85 to 0.93 min, and varying from 33.67 to 33.81 min for the 1.8 and 5 µm-particle size columns, respectively). This was supported by performing some trial analyses on a true UHPLC system. After increasing the flow to 1.4 mL min−1 and adapting the gradient (max. pressure 525 bar), the run time was shortened to 2 min and the retention time variability was reduced to 0.05–0.73% (data not shown). Figure 2.1. Top: HPLC profile of R. luteola H extract on 5 µm-particle size 250 mm × 4.6 mm HPLC column. Column temperature was 40 ºC and injected volume was 20 µL. Identity of the main peaks is indicated: luteolin-7,3′-di-O-glucoside (luteolin diglucoside, ldg), luteolin-7-O-glucoside (luteolin monoglucoside, lmg), luteolin (lut) and chrysin (internal standard, i.s.). Bottom: Analysis of R. luteola H on 1.8 µm-particle size 50 mm × 3.0 mm UHPLC column. Column temperature was 35 ºC and injected volume was 2 µL. Top and bottom: Detection at 345 nm.

24 25

min fl the increasing After system. UHPLC a true on analyses trial some performing the for min (Tablequantitation 2.4). and the retention 0.05–0.73% variability wasto time reduced a pressurenear the maximum of the HPLC (400 pump bar) and, by additionally, the much lower higher ( relatively 1, bottom theafter hardwareandadaption tubing, of minor 24 but still more than suffic The resolutionbetween ldg Thecolumn could UHPLCbe run mL a with flow of min 0.9 ldg of maxima absorbance the of view A) particle size T 2.3.2. Chromatography syringe adding placing and solvent bar, the stir flask ona magnetic analytical consisting procedure minimal, of is weighing the previouslyground plantma content. to correct initial the outcome by multiplyingfactora itby of 100/95 arriveat to the true flavo equilibration (Fig.).preparationfiltration proved 1,top other The total necessary than time run period of 5 days. Within one day, the s period Within of 5days. a 0.4% at over most precise, varying be to were times observed retention precision experiment, bsolute retention times ( he validationwas performed ona traditional HPLC The column. solvents used with

Although in absolute terms the r absolute terms Although in The use of a [ − 9] 1

, and adapting the gradient (max. pressure 525 bar), the run time was shortened to 2 min 2min to shortened run was time bar),the pressure 525 andgradient adapting (max. the

than with thewith 5µm than - and filtering the In of the input whole theare presented. tables,labour The uncorrectedvalues manual ). ). i.e. This shortenedThis the 5 run to time , min (ii) (ii)

RP , , 1.8 and 5 µm was, 80min precision) 1.8 µm - methanol HPLC 250 HPLC sample into the HPLC vial. - size particle

relative relative higher The ient for reliable integrations.

, ldg (solvent B) they were were they -

particle size mm and and 35.9 mL of methanol were needed. From the results of the the results of the wereand needed. ofFrom 35.9mL methanol : varying: 0.85to, from 0.93min - particle size lmg ×

4.6 mm columnwere:4.6 mm etention time variation wasUHPLC lower oncolumn, the UHPLC 50

observed to be 25 to 180× 25to be observed to ( . The wavelength detection was be selected to 345nm in r R

values retention the of time s (340 nm), lmg

HPLC column. HoweverHPLC column. = 5.4) was lower than onthe HPLC column

retention time variation is likely columns, respectively). was This supported

mm and changing the B solvent to acetonitrile (Fig. (Fig. acetonitrile to B solvent the changing and and only 1.4mL wasacetonitrile needed. of ×

3.0 m (350 (i) (i) an aqueous buffer of pH 3 aqueous of an buffer st nm) m columnm saves time and solvents − 1 (data shown) not irrer, i.s. adding the and varying 33.81 33.67to from

on a conventional HPLC system ( i.s. , and s ,

never did this influence this never did the varied from 0.004 to 0.05%. 0.004to from varied and lut lut ldg (350 nm). (350

cases, respectively) respectively) cases, caused by working caused byworking .

ow to 1.4 mL ow 1.4mL to solution, solution, , No sample sample No ( th including including R is (solvent (solvent s

= 9.2)

5 µm 5 terial, and and by by ne ne - . ,

and injectedand volume was luteola monoglucoside, Figure is indicated: column. C

H on 1.8 µm 2.1. olumn temperature was 40 Top luteolin : lmg HPLC profile of - - particle size 50 ), 7,3′

luteolin ( 2µL. - di - O - glucoside ( glucoside Top lut R. luteola R. and and ) and) chrysin (

mm mm

ºC and ºC bottom × luteolin

3.0 mm UHPLC column. C

H extract onH extract injected volume was : D

diglucoside, diglucoside, etection at 345 nm. 345 etection at internal standard, i.s.

5 µm - particle size particle ldg ), luteolin), - 20 µL. Identity of mainthe peaks olumn temperature was 35 ). B ).

250 7 ottom - O mm - glucoside (luteolinglucoside : Analysis: of R. ×

4.6 mm HPLC

ºC ºC

Chapter 2 2.3.3. Accuracy 2.3.4. Identification and peak purity The accuracy of the method was determined by an absolute recovery (spiking) experiment at The identity of the three main flavonoids of R. luteola was confirmed to be ldg, lmg, and lut three levels. Namely, by adding roughly 50%, 100%, and 140% of what is expected in the plant by having identical retention times and on-line UV–vis absorption spectra to those of authentic (Table 2.1). The sr values of the results are equivalent to those of the usual procedure (Table reference substances, and on-line mass spectrometric experiments (see section SI A.1.2). 2.4), except for the low spiking level. The average recovery was not statistically different from Peak purity analysis (see section SI A.1.2) by LC–MS and LC–UV showed there is no 100% (assuming normal distribution of the results and considering 2s), indicating good significant co-elution of minor compounds absorbing at 345 nm of R. luteola and the ldg, lmg, accuracy. To check for any constant zero-order losses due to adsorption on glass and filters lut, and i.s. peaks (see section SI A.2.1) on either the HPLC or the UHPLC column. Thus, the which cannot be detected with spiking experiments, relative recovery experiments at three peak areas on both columns at 345 nm are representative for ldg, lmg, lut, and i.s. different concentrations were carried out. The stability of the compounds under the extraction conditions was simultaneously evaluated (Table 2.1). At all three concentrations, the relative 2.3.5. Calibration curves and limits of detection and quantitation recovery of the three flavonoids was not statistically different from 100% (assuming normal Internal standardisation was the quantitation method selected. Out of four potential internal distribution of the results and considering 2s), showing that—as expected—the extraction standards [3',4'-dihydroxyflavone, lut tetramethylether, chrysin (5,7-dihydroxyflavone) and procedure is mild and that any losses due to sample manipulation, filtration, or adsorption to acacetin (apigenin-4'-methylether)], chrysin was selected based on stability issues, price glass are negligible. considerations (€ 0.01 assay−1), and retention time, in spite of its relatively low absorbance at 345 nm. The relative response factors (RRFs, slope of Ax/Ai.s. plotted vs. wx/wi.s.) of ldg, lmg, and Table 2.1. Relative and absolute recoveries a of ldg, lmg, and lut (n = 3; sample = R. luteola H). lut were determined and the linearity of the concentration of the solutions vs. area of the peaks Relative (%) b,c Absolute (%) b,d was checked. The RRFs were 1.10 (ldg), 1.51 (lmg), and 2.37 (lut) with the HPLC column Level ldg lmg lut ldg lmg lut and 1.13 (ldg), 1.58 (lmg), and 2.29 (lut) with the UHPLC column. The areas of the peaks were observed to linearly correlate with the concentrations of the solutions over the range of (106 ± 15), concentrations studied (Table 2.2). These ranges are satisfactory for the quantitation of the Low (102 ± 2), 2 (99 ± 2), 2 (100 ± 3), 3 (99 ± 8), 8 (103 ± 7), 7 14 compounds of interest at the concentrations in which they are present in R. luteola H and other samples.

(101.0 ± (99.7 ± 0.2), Medium (101 ± 1), 1 (98 ± 6), 6 (101 ± 6), 6 (107 ± 8), 7 0.7), 0.7 0.2 Table 2.2. HPLC calibration curves and linearity of ldg, lmg, and lut. Ldg and lmg: Five point-curves; Lut: Seven point-curves. a a HPLC column UHPLC column Compound [x] −1 2 −1 2 (98.7 ± 0.9), (99.9 ± 0.2), [x] (µg mL ) vs. Ax (R ) [x] (µg mL ) vs. Ax (R ) High (101 ± 2), 2 (99 ± 1), 1 (101 ± 1), 1 (102 ± 4), 4 1.0 0.2 ldg y = 46,517x + 12,410 (0.9995) y = 4.9919x + 0.1505 (0.9998) lmg y = 65,346x − 19,906 (0.9999) y = 7.044x + 7.8141 (0.9998) a Relative recovery is based on subjecting an extract corresponding to 100, 200, or 300 mg of plant material lut y = 101,383x + 89,417 (0.9998) y = 10.249x − 2.0982 (0.9992) anew to the entire analytical procedure; absolute recovery is based on three spiking experiments at 50, 110, a ldg, lmg, and lut: Each case, n = 1. and 140%. b Expressed as (average ± standard deviation), sr (relative standard deviation). c The (relative) recovery of ldg, lmg, and lut due to the syringe-filtration also ranged between 99 and 102%, The limits of detection (LOD) and quantitation (LOQ) of the method were determined based with sr values ranging between 0.1 and 2.6%. on the slopes of the calibration curves and the s of the baseline noise of analyses of analytical d Low level: n = 6. blanks, and on the signal-to-noise ratio of injections of highly diluted solutions (Table 2.3). For more details on the determination of the LOD and LOQ values, see Appendix A. Assuming normal distribution of the results and considering both the s values of the results and, in this case, the uncertainty of the LOD values estimated based on the signal-to-noise ratio, the LOD values using the HPLC column are the same, regardless the procedure used for their estimation. Similarly, using the UHPLC column, except for ldg, the LOD values are also the same. Moreover, the method is more sensitive when using the UHPLC column than when using the HPLC column (Table 2.3).

26 27

glass are negligible. procedure is mild and any that losses due to sample manipulation, filtration considering and 2s of the results distribution was notrecovery flavonoids three statistically of the different from 100% (assuming normal wasevaluated (Table conditions simultaneously T out. carried were concentrations different relative experiments, spiking with detected which be cannot accuracy. To check for any constant zero any constant for check To accuracy. and2s considering results of the distribution 100% (assuming normal different from recovery statisticallyaverage not The was level. the2.4), lowspiking exceptfor 26 d c b a Table (Table levels three at experiment (spiking) recovery absolute by an determined was method the of accuracy The 2.3.3. Accuracy and 140% anew the to analytical entire absolute procedure; recovery on based three is with with

T Level Medium Low High Relative recovery is based on subjecting an extract corresponding to 100, 200, or 300 mg of plantmaterial L E he he ow level:ow n =6. xpressed deviation), as (average s standard ± s (relative) (relative)

r 2.

values ranging between 0.1 and 2.6%. 1 2.1). .

Relative recoveries absolute and

. N The The Relative (%) 0.7), 0.7 0.7), (101.0 ± ldg (102 ± 2), 2 (101 ± 2), 2 recovery of ldg amely

s r

values of thevalues results are thos to equivalent

, by addingroughly 100% by 50%,

b,c lmg 0.2 0.2),(99.7 ± (99 ± 2), 2 (99 (99 ± 1), 1 (99 , lmg

, and and

lut

of of due to the syringe the to due

lut (101 ± 1), 1 (100 ± 3), 3 (101 ± 1), 1 r ldg

(

relative standard deviation standard relative - , order losses duelosses adsorption to order on glass and filters lmg he stability under of the compoundsextraction , ), and and

). 2.1). showing that showing

lut , - and 140% filtration also ranged between 99and 102%,

ldg Absolute (%) (98 ± (99 ± 8), 8 (99 1.0 (98.7 ± 0.9), (n sample =3; luteola =R. At allconcentrations three

6), 6 6), e of the usual procedure (Table

as expected —as ). of what is expected in the in plant what expected is of recovery experiments at three three at experiments recovery spiking spiking

b,d lmg (101 ± 6), 6 (103 ± 7), 7 0.2 (99.9 ± 0.2),

experiments at 50, 110 ), indicating good ), indicating

, H). the extraction extraction —the or adsorption to or adsorption to 14 (106 ± 15), lut (107 ± 8), 7 (102 ± 4), 4

the relative

, R. luteola main flavonoids of the three Theof identity 345 nm. 345 nm. h by 2.3.4. Identification and peak purity HPLC column (Table2.3). Moreover, more the methodis sensitivewhen using thecolumnthan UHPLC when using the Similarly, u estimation. their for used dure proce the regardless same, the are column HPLC the using values estimated LOD values the of uncertainty the case, the both considering and theresults of normal distribution LOQ A LODFor values, ofand Appendix onthe the see determination more details blanks, on the slopesof the calibration curves and the s T a Table checked was lut substances reference and 1.13( co significant samples. samples. compounds of interest at the concentrationswhich in theyare R. luteola in present concentrations studied of range over of the the were solutions the with concentrations correlate linearly observedto standards Internal was standardisation themethodselected. quantitation potential Outof four internal limits curves detectionquantitation 2.3.5. Calibration and of and ldg for atarepeak 345nm resentative areas columns rep onboth lut acacetin (apigenin acacetin considerations (€ 0.01assay S

ldg lut lmg ldg Compound even point even determined based of detectionhe (LOD) limits based and (LOQ) quantitation of the methodwere determined

, The r The Peak purity analysis ( analysis purity Peak were determined determined were

and aving identical retention times

lmg 2. 2 . and onthe signal , elative response factors (RRFs, slope of A (RRFs, factors response elative i.s. HPLC calibration curves and linearity of ldg of linearity and curves HPLC calibration and and - ldg [ curves. [ 3',4' x peaks ( peaks sing the UHPLC column, except for ldg the column,exceptfor UHPLC sing ] . lut

- ), 1.58( The RRFs wereThe RRFs 1.10( and the ldg and luteola the R. of at 345nm compoundsminor absorbing ofelution : - dihydroxyflavone, E

y 65 y = 46 y = [ HPLC column n = 1. n= case, ach x = 101 - ] (µg mL] (µg s 4' SI A.2 SI ee section , and on- , and and the linearity vs. of of the thesolutions concentration - lmg methylether) , , 346x 517x + 12 (Table (Table , 383x + 89 - s ) to SI A.1 SI section ee − , 1 − ) - − and 2.29(

noise ratio of of injections highly (Table solutions diluted 2.3). vs 1 line mass spectrometric experiments ( experiments spectrometric mass line 19 a )

). These ranges are satisfactory ranges for These the2.2). quantitation of the . A , , , 906 (0.9999) 410 (0.9995) and retention time, in spite of its relatively absorbance low at , 417 (0.9998) x

(R ] and lut , chrysin was selected, chrysin was onstability based issue ldg 2 the .1)the HPLC oneither or the )

lut on- tetramethylether,chrysin (5,7 ), 1.51( ) with the UHPLC column. The areascolumn. The UHPLCthe peaks ) of the with line UV

.2 , lmg ) by LC by )

analytical the baselineof noise of analyses of analytical lmg , based onthe signal based –vis and and y = 10.249x [ UHPLC column x y = 7.044x + 7.8141 (0.9998) y = 4.9919x + 0.1505 (0.9998) x /A ) ] (µg mL] (µg , lut and 2.37( i.s.

e LOD values are also the same. same. the also are LOD values e , th MS and –MS absorption was confirmed to be ldg be to confirmed was .

plotted Ldg s −

1 and and values of the results and, this in − )

vs 2.0982 (0.9992) , lmg lut lmg

. A vs. a LC

spectra spectra SI A.1 SI section see UHPLC column. Thus, the column.Thus,the UHPLC x )

: F w (R –UV with the HPLC column with , lut -

- x to ive point ive 2 dihydroxyflavone) and dihydroxyflavone) and /w )

- noise ratio, the LOD LOD ratio, the noise , i.s.

to those ofto authentic showed there is no no is there showed and i.s. )

area of the peaks peaks the of area of of

- curves; curves; ldg , lmg H and other . , Assuming lmg L .2). , ut s, price price s, and : , ,

lmg and and lut 27 ,

Chapter 2 Table 2.3. LOD and LOQ values of ldg, lmg, and lut using both HPLC and UHPLC columns, based on: (i) 2.3.7. Quantitative results and precision the slopes of the calibration curves and the s of the baseline noise of analyses of analytical blanks; and (ii) The concentrations of ldg, lmg, and lut in R. luteola H, as determined on the 5 µm-particle size based on the signal-to-noise ratio. column, were 3.5 mg g−1, 6.5 mg g−1, and 1.0 mg g−1, respectively. For the 1.8 µm-particle size −1 −1 −1 Compound LOD ± s (ng) LOQ ± s (ng) LOD (ng) column, these figures were 3.6 mg g , 6.5 mg g , and 1.0 mg g (Table 2.4). These results HPLC column (injection volume of 20 µL) are not statistically significantly different (ldg). This indicates that the HPLC system copes well i (n = 5) ii (n = 1) with the small UHPLC injection volumes (2 µL). Furthermore, major advantages of the use of ldg 1.1 ± 0.3 3 ± 1 0.9 the UHPLC column are, of course, the 16× faster analysis and the 25× lower solvent lmg 0.7 ± 0.2 2.2 ± 0.7 0.8 consumption. lut 0.25 ± 0.08 0.8 ± 0.2 0.4 The precision of the entire method on the HPLC column was evaluated in terms of UHPLC column (injection volume of 2 µL) repeatability and intermediate precision. The sr values for the whole analytical procedure were ~5% with the HPLC column, except for series 3 (Table 2.4). Still, the larger sr values of series i (n = 6) ii (n = 1) 3 did not influence the average outcome. The type of stirrer used (15-point, which heats itself 0.29 ± 0.05 0.9 ± 0.1 0.1 ldg up by 3 oC during operation, for series 1 and 3 vs. individual ones for series 2) does not seem 0.21 ± 0.04 0.6 ± 0.1 0.2 lmg to have any influence on the outcome, which suggests robustness of the method regarding small lut 0.14 ± 0.02 0.41 ± 0.07 0.1 temperature variations. The sr values of the quantitations carried out with the UHPLC column

were 7% (ldg), 7% (lmg), and 10% (lut) (Table 2.4). These values are comparable to those of

the analyses of the same samples carried out with the HPLC column, 7% (ldg), 7% (lmg), and The LOD and LOQ values of this method are low compared with those of earlier methods. 11% (lut) (footnote d, Table 2.4). Kočevar et al. and Plazonić et al. [11] [10] reported sr values Plazonić et al. and Kočevar et al. [10, 11] determined LOD and LOQ values for lmg and lut for the repeatability and intermediate precision of the areas of lmg and lut peaks ranging from that are 6–20× higher than the values reported here, when using the HPLC column, and 20– 0.9 to 4.3%, respectively, when analysing standard solutions. When analysing extracts, Kočevar 50× higher than the values reported here, when using the UHPLC column. Gaspar et al. [12] et al. [11] observed slightly higher sr values, up to 5.4% for lmg. Gaspar et al. [12] evaluated determined LOD values <7 ng and LOQ values ≤11 ng for ldg, lmg, and lut, which are 6–12× the repeatability of the whole analytical procedure using two samples of R. luteola of extreme (LOD) and 4–10× (LOQ) higher than the values reported here, using the HPLC column, and (high and low) contents of flavonoids. The observed average sr values were 5.1% (ldg), 3.9% are 20–60× (LOD) and 10–20× (LOQ) higher than the values reported here, using the UHPLC (lmg), and 4.9% (lut). The sr values reported here are slightly higher than those of the column. The baselines of the UV detectors used in the work reported here were highly stable. aforementioned works. Possible reasons include the different experimental designs and data On average, the s values of the baseline noise over large time intervals were only 0.04 ± 0.01 processing. Still, the proposed method determines the combined content of ldg, lmg, and lut mAU (HPLC column) and 0.09 ± 0.01 mAU (UHPLC column) (see Appendix A for details). with sr <6.5%, which is fine for the intended purpose. Baselines in earlier studies were possibly less stable. In addition, different methods were used to estimate those limits, as Plazonić et al. [10] used the residual s of the regression line and slope, and Kočevar et al. and Gaspar et al. [11, 12] based it on the signal-to-noise ratio.

2.3.6. Stability of samples The stability of the samples ready for HPLC analysis was evaluated to verify whether the residence time in the autosampler would influence the outcome. This could, for example, be the case when analysing a large sample set. Moreover, it was evaluated whether the samples

o − o are stable upon storage at 4 C for 3 days and at 20 C for 15 days. Upon storage of the samples ready for HPLC analysis, the area of the chrysin peak was not statistically significantly different than that of the samples analysed shortly after the work-up procedure, indicating that neither precipitation of the i.s. nor sample concentration took place. Assuming normal distribution of the results and considering the s, the amounts of the flavones remained the same in all analysed sets of vials. Thus, the compounds in samples are stable upon storage under the aforementioned conditions.

28 29

in all analysed sets of via of sets analysed all in and the s ofconsidering the results distribution neither of the precipitation i.s. signal the on based the slopes of the calibration curves and the s aforementioned conditions. different than that peak chrysin the of area the analysis, HPLC for ready samples 6 are that Table 28 areat 4 stable uponstorage wheth evaluated was it Moreover, set. sample a large analysing when case the for be example, could, This the outcome. residence thewouldinfluence in autosampler time the whether verify to evaluated was analysis HPLC for ready samples the of stability The of samples2.3.6. Stability slope, and Kočevar On average, the s column. The the baselines of 50× Plazonić compared with methodare low LOQ ofLOD this values Theand to estimate those limits Baselines earlierIn in were studies lessstable. different possibly addition, weremethods used mAU (HPLC 0.01mAU 0.09± (UHPLC column) and c 20 are 4 (LOD) and determined LOD values <7 ng and LOQ values ≤ lut lmg ldg Compound lut lmg ldg

higher than the values reported here, when using the UHPLC column. Gaspar et al. column. UHPLC when the using reported here, values than the higher 2. –60 3 . LOD valuand LOQ et al. –20× ×

(LOD) and 10 –10

and Kočevar Kočevar and higher than the valu the than higher - to × 0.14 ± 0.020.14 ± 0.040.21 ± 0.050.29 ± i UHPLC column (injectionvolume of 2µL) 0.080.25 ± ± 0.2 0.7 ± 0.3 1.1 5) = (n i HPLC column (injectionvolume of 20µL) LOD ±

(n = 6) = (n values of the baseline large noise over intervals time were only 0.01 0.04± -

after the work shortly the after analysed samples the of noise ratio. noise (LOQ) higher than the values reported here, using the HPLC higher column,and the reported than here, theusing (LOQ) values et al. s ,

( and Gaspar et al. ls. Thus ls. as es of es ng –20×

) Plazonić Plazonić

et al.

o the work reported usedhere in detectors UV ldg C for 3 days and at − and forC at 3days here, using the UHPLC UHPLC the using the values reported here, (LOQ) higher than , , lmg

the compounds in samples are stable upon storage under the storage the under upon samples are stable the compounds in nor sample concentration place. took [ 10, es reported here, when using the HPLC column,and 20– HPLC the using here, when es reported ,

et al. of the baseline noise of analyses of analytical and and 11] 0.8 ± 0.2 0.8 ± 0.7 2.2 ±1 3 LOQ ± 0.41 ± 0.070.41 ± ± 0.1 0.6 ± 0.1 0.9 lut

[

determined LOD and LOQ values for [ 11, 12]

10] using both HPLC and UHPLC columns, based on: (i)

, t

s he amounts of the flavones remained the same same remained the flavones of amounts the he used the residualused s the

( 11 ng for ldg ng

d it onthe signal d it base )

o for details). details). A for (seeolumn) Appendix C for 15 days. Upon storage of the for storageC 15days. Upon of the was not statistically significantly , lmg - up procedure 0.1 0.2 0.1 0.4 0.8 0.9 LOD ( ii (n = 1) = ii (n ii (n = 1) = ii (n

of the regression line and and regressionline of the

those of earlier methods. of earlier methods. those and ng

- ) lut to

were highly stable. stable. highly were Assuming normal Assuming normal - , which are 6 are which , blanks; and (ii) (ii) blanks; and noise ratio. , indicating that er the samples samples the er lmg

and and –12×

[ 12] lut

11% ( the with out carried samples same the of analyses the ( 7% were up by 3 influence not 3 did the averageof Thestirrer (15- type outcome. used s The precision. intermediate and repeatability consumption. the UHPLC thewith small (2 volumes µL). Furthermore, injection major advantages of the use of (high and low) contents of average flavonoids. Thecontents(high s and observed low) the repeatability using analytical twosamples ofwhole of procedure R. luteola the ~5% the with HPLC are column, these figures were 3.6 The concentrations of ldg, lmg precision results2.3.7. Quantitative and of ldg content the combined methoddetermines the proposed processing. Still, aforementioned reasons works. Possible include the different experimental designs and data ( et al. 0.9 to 4.3%, respectively, when analysing standard solutions. When analysing extracts, Kočevar repeatability the for and intermediate precision s temperature The variations. robustness of the haveto outcome, method any which influence onthe suggests regarding small g column, were 3.5mg with s lmg The precision of the The of precision not statistically significantly different ( UHPLC )

, [ lut r < 11] and 4.9% ( o 6.5% ) C

ldg (footnote observed slightly higher s

, during operation

), , which is fine for the intended purpose. for, which fine the is intended purpose. column are, of course, the 16×

7% (lmg 7% lut , Table 2.4). Table d,

column, except for series 3 (Table 2.4). Still,the (Table larger 3 series for s except column, ). ). − 1 ) , 6.5mg g

The The , the the methodon entire and r for series 1 and 3 and 1 series for

values of the quantitations carried out with the U carried the with out the quantitations of values , 10% ( r

and mg g mg values reported values Kočevar Kočevar − 1

lut lut , r lut −

and values, up to 5.4% for lmg 5.4% for values, upto 1 mg g , 6.5mg in in )

(Table (Table 1.0 mg g 1.0 mg ldg R. luteola

et al. ) . This indicates that the HPLC system copes well well copes HPLCsystem the that indicates This . r

of the areas of lmg of areas the of values for the whole analytical procedure were were procedure analytical whole the for values vs.

and Plazonić 2.

faster analysis and the 25 the and analysis faster − HPLC here than those areof slightly higher the 1

− 4). individual ones for series 2) does not seem 2) does onesseem not forindividual series , 1 , respectively , H and 1.0 mg g

HPLC are values These , as determined onthe 5µm

column was evaluated in terms of of evaluated terms in was column

column, 7% ( et al. r

values . Gaspar Gaspar . . For the. For 1.8µm −

and [

11 (Table (Table point, point,

] lut were were comparable to those of comparable those of to

[ 10 et al.

ldg peaks ranging from from ranging peaks ] 2.4). These results

which heats itself 5.1% ( reported × r )

, values of series series of values

HPLC 7% (lmg 7% [ lower solvent , 12] - lmg - particle size particle size size particle of extreme extreme of ldg

evaluated evaluated , s

column column r ), 3.9% and and

values values ) , and lut 29

Chapter 2 Table 2.4. Precision of the entire method, and that of the analyses using the UHPLC column, for the quantitation of ldg, lmg, and lut in Reseda luteola (sample = R. luteola H). 2.5. Supplementary material HPLC column UHPLC column Appendix A contains information supplementary to that in this chapter. Sections material and Series Compound −1 a,b −1 c,d Concentration (mg g ) (sr in %) Concentration (mg g ) (sr in %) methods, results and discussion, and reference are available. ldg 3.6 (5.3) 3.6 (7.1) 1 lmg 6.7 (5.2) 6.5 (7.2) lut 1.01 (5.8) 1.0 (10.4) 2.6. References ldg 3.5 (3.8) 3.6 (6.2) [1] van Beek TA, van Veldhuizen A, Lelyveld GP, Piron I. Quantitation of bilobalide and 2 lmg 6.5 (3.6) 6.5 (6.5) ginkgolides A, B, C and J by means of nuclear magnetic resonance spectroscopy. Phytochem lut 1.00 (4.9) 1.01 (9.0) Anal 1993; 4(6): 261–8. ldg 3.5 (6.1) 3.6 (8.8) [2] van Beek TA, Scheeren HA, Rantio T, Melger WC, Lelyveld GP. Determination of 3 lmg 6.4 (6.5) 6.5 (8.5) ginkgolides and bilobalide in Ginkgo biloba leaves and phytopharmaceuticals. J Chromatogr A lut 1.01 (8.5) 1.0 (11.5) 1991; 543(2): 375–87. a Extractions were carried out on three non-consecutive days (each day, n = 7). b [3] Gao X-Y, Jiang Y, Lu J-Q, Tu P-F. One single standard substance for the determination of Series 3: s of the quantitation of ldg, lmg, and lut were 0.2, 0.4 and 0.09, respectively; thus, the sr of their multiple anthraquinone derivatives in rhubarb using high-performance liquid chromatography– combined amounts was 6.3%. diode array detection. J Chromatogr A 2009; 1216(11): 2118–23. c Extractions were carried out once (n = 8); then, stored samples were analysed in three series (see section 2.2.7). [4] Massart DL, Vandeginste BGM, Buydens LMC, De Jong S, Lewi PJ, Smeyers-Verbeke J. d −1 −1 Analysis of same samples (n = 8) using the HPLC column: Amount = 3.6 mg g , sr = 6.8% (ldg); 6.5 mg g , Handbook of chemometrics and qualimetrics: part A. Elsevier Science: Amsterdam; 1997. −1 7.2% (lmg); and 1.0 mg g , 10.7% (lut); for that, the following RRFs (determined anew for HPLC column, [5] van Rozendaal ELM, Kurstjens SJL, van Beek TA, van den Berg RG. Chemotaxonomy of details not presented) were used: 1.09 (ldg); 1.51 (lmg); and 2.28 (lut). Taxus. Phytochemistry 1999; 52(3): 427–33.

[6] van Rozendaal ELM, Lelyveld GP, van Beek TA. A simplified method for the determination 2.4. Conclusion of taxanes in yew needles by reversed-phase (C18) high pressure liquid chromatography. The analytical method reported here is suitable for the quantitation of ldg, lmg, and lut in the Phytochem Anal 1997; 8(6): 286–93. aerial parts of R. luteola. According to the extensive validation (extraction efficiency, recovery, [7] Cristea D, Bareau I, Vilarem G. Identification and quantitative HPLC analysis of the main accuracy, precision, sensitivity, peak purity, sample stability), this is accomplished accurately. flavonoids present in weld (Reseda luteola L.). Dyes Pigments 2003; 57(3): 267–72. The precision reported (<6.5%) is adequate for the purpose. The industrial partner of the project [8] Cerrato A, De Santis D, Moresi M. Production of luteolin extracts from Reseda luteola and reported, in daily practice, a standard deviation <4% when analysing over 120 R. luteola assessment of their dyeing properties. J Sci Food Agric 2002; 82(10): 1189–99. samples using this method. The method meets the need for very little manpower input per sample and uses standard laboratory equipment. Although the analytical method was developed [9] Derksen GCH, Niederlander HAG, van Beek TA. Analysis of anthraquinones in Rubia for the quantitation of flavones in R. luteola, it is expected that, after additional validation, it tinctorum L. by liquid chromatography coupled with diode-array UV and mass spectrometric could be applied on other flavone-rich plant materials too. detection. J Chromatogr A 2002; 978(1–2): 119–27. Two types of columns were used: For the validation, a traditional 4.6 mm column with 5 µm [10] Plazonić A, Bucar F, Maleš Ž, Mornar A, Nigović B, Kujundžić N. Identification and particles; then, the chromatographic step was speeded-up by the use of a short UHPLC column, quantification of flavonoids and phenolic acids in burr parsley (Caucalis platycarpos L.), using while still using a conventional HPLC system. The use of the UHPLC column reduced the run high-performance liquid chromatography with diode array detection and electrospray ionization time by a factor of 16 and the organic solvent consumption by a factor of 26. Retention times mass spectrometry. Molecules 2009; 14: 2466–90. were relatively less reproducible with the UHPLC column than with the HPLC column. This −1 [11] Kočevar N, Glavač I, Injac R, Kreft S. Comparison of capillary electrophoresis and high could be remedied by either lowering the flow to 0.6 mL min or using a true UHPLC pump. performance liquid chromatography for determination of flavonoids in Achillea millefolium. J Cross-validation showed that the quantitation of ldg, lmg, and lut was not statistically Pharm Biomed Anal 2008; 46(3): 609–14. significantly different, with comparable precision. Moreover, the method is more sensitive with the UHPLC column than with the HPLC column. Using a UHPLC column on a conventional [12] Gaspar H, Moiteiro C, Turkman A, Coutinho J, Carnide V. Influence of soil fertility on HPLC system is, thus, a way of modernising HPLC-based (phytochemical) analyses dye flavonoids production in weld (Reseda luteola L.) accessions from Portugal. J Sep Sci 2009; inexpensively. 32(23–24): 4234–40.

30 31

2.2.7 Cross while particles Table 30 inexpensively HPLC system is the UHPLC columnt significantly different, could HPLC col the columnthan UHPLC with the with were reproducible relatively less the organic and bytime factor of 16 a c for theof quantitation flavones, R. luteola in it is expected that, after additional validation, it sample and useslaboratory standard was equipment.the analytical Although developed method Th method. this samples using accuracy, sensitivity,purity, peak stabi precision, sample recovery, efficiency, (extraction validation extensive the to . R. luteola of According parts aerial The suitable for analyticalhere is the of quantitation method reported ldg 2.4. Conclusion d c b a reported The precision reported (< detailswere not presented) ( used: 1.09 combined amountswas 6.3%. 7.2% ( of of

oul E 2 1 3 Series A S E

ldg xtractionswere out carried on three non eries 3: s 3: eries xtractionswere carried out once (n = 8); then, samples stored were analysed in series three (see section nalysis of same samples (n = 8) using the HPLC column: column: HPLC the using 8) = (n samples same of nalysis Two types of columns werecolumns usedTwo: types of ). d be flavone applied onother , 2.

still using a conventional lmg - lmg be remedied lowering byeither the flow 0.6mL to min

validation showed thattheof quantitation ldg validation showed 4 . , was speeded then,; the chromatographic step was P ); and g 1.0 mg , recision of method, the entire in daily practice

and and of the quantitation of of quantitation the of ldg lut lmg ldg lut lmg ldg lut lmg Compound . lut

in in , thus, a, thus, way modernising of HPLC Reseda luteola Reseda han w han −

with comparable Moreover precision. with 1 , 10.7%, (

6.5%) adequate is the purpose. for The partner industrial of the project 3.5 (6.1) 1.00 (4.9) 6.5 (3.6) 3.5 (3.8) 1.01 (5.8) 6.7 (5.2) 3.6 (5.3) g (mg Concentration 1.01 (8.5) 6.4 (6.5) HPLC column , a standard< deviation ith ldg

the HPLC column.U HPLC system. lut

e method meets the need for very little manpower input per per manpower input little very thefor need meets e method , (sample luteola =R. ldg lmg

- ); for that, the following RRFs (determined anew for HPLC column, and that of the analyses using the UHPLC column, UHPLC the using analyses the of that and rich plant materials too.

); 1.51 - consecutive days (each days day,consecutive n =7).

, F and and or the validation

solve lut ( lmg −

nt consumption by a factor of 26. by aconsumption factor of nt 1 were 0. ) ); );

a, The use of the The b and and

(

s H). r

A 2, 0.4 and 0.09, respectively;t in %) in 2.28 mount = 3.6

4% sing a UHPLC column on sing a column UHPLC - up by the use of a short UHPLC column, UHPLC aup bythecolumn, useshort of , (

a traditional 4.6 mmcolumn with 5 µm lut

, when lity) ).

lmg

- 1.0 (11.5) 6.5 (8.5) 3.6 (8.8) 1.01 (9.0) 6.5 (6.5) 3.6 (6.2) 1.0 (10.4) 6.5 (7.2) 3.6 (7.1) g (mg Concentration UHPLC column , based based

the methodmore is sensitive w , − UHPLC UHPLC

, mg g mg tely. accura accomplished is this 1 analysi

or using a true UHPLC pump. and −

1 , (

phytochemical s lut r

ng = reduced the run run the reduced column

6.8% (

was was

over 120R. luteola over ,

lmg for the quantitation the quantitation for hus, the s the hus, − ldg 1 not not Retention times )

, a c, ); 6.5 and d

conventional ( s statistically ) r

umn. This This umn. r

in %) in

lut of their their of mg g mg analyses

in the the in

− 1 , ith

Appendix Appendix material2.5. Supplementary 32(23–24): 4234–40. high and Identification N. Kujundžić and phenoli flavonoids quantification of B, Nigović A, Mornar Ž, Maleš F, Bucar A, Plazonić [10] ChromatogrA 978(1–2):detection. 2002; 119–27. J tinctorum Rubia anthraquinones in ofBeek TA. Analysis van Derksen HAG, GCH, Niederlander [9] Food Sci Agric 82(10): J 2002; 1189–99. properties. dyeing their of assessment extracts from of luteolin Production D, Moresi M. Santis A, De Cerrato [8] flavonoids present ( weld in [7] Cristea D, BareauI,Vilarem Identification G. and quantitative HPLCanalysis of main the Phytochem 8(6): 1997; 286–93. Anal by reversed yew needles in taxanes of ABeek simplified TA. Lelyveld van van GP, method for Rozendaal ELM, the determination[6] Taxus van Rozendaal E [5] andHandbook A. 1997. Elsevier qualimetrics: part of Amsterdam; chemometrics Science: SmeyersLewi Massart De DL,BuydensLMC, BGM,S, Jong PJ, [4] Vandeginste diodeChromatogr array J A 1216(11): detection. 2009; 2118–23. anthraquinonemultiple derivatives rhubarb in - high using Gao[3] X 543(2):1991; 375–87. ginkgolides and Ginkgo bilobalidein bi van[2] BeekLelyveld HA, T, TA,Melger Rantio Determination Scheeren GP. WC, of Anal 4(6): 1993; 261–8. byJ B,ginkgolides ofspectroscopy. resonance and nuclear means C magnetic A, Phytochem van[1] BeekLelyveldVeldhuizen A, I. TA, Piron van GP, Quantita 2.6. References methods dye flavonoids production in weld in dye ( flavonoids production of Comparison S. Kreft R, Injac I, Glavač N, Kočevar [11] mass spectrometry. 2466–90. 14: 2009; Molecules [12] Gaspar H, Moiteiro C, Turkman A, Coutinho J, Carnide V. Influence of soil fertility of soil Influence Carnide V. J, Coutinho TurkmanC, A, Gaspar H,[12] Moiteiro Pharm 46(3): Biomed 609–14. Anal 2008; performance chromatographyfor of liquid Achillea flavonoids in determination millefolium - performance liquid chromatographyperformance array liquid electrospray detection diode and with ionization . Phytochemistry 1999; 52(3):. Phytochemistry427–33. 1999; , results discussion and

- A L Y, Jiang Y, Lu J Y, Jiang Y, . by diode chromatography coupled liquid with

material and and chapter. material this contains thatin informationto supplementary Sections

LM, Kurstjens SJL, van Beek TA, van den Berg RG. Chemotaxonomy of Chemotaxonomy of Berg RG. Beek den TA,LM, van van Kurstjens SJL,

- Reseda luteola Q, Tu P

, and reference Reseda luteola - F. F. loba One standard single determination for substance of the c acids in burr parsley ( parsley burr in c acids - phase (C18) high pressure liquid chromatography. high (C18) pressurephase liquid L.). 267–72. Pigments 57(3): 2003; Dyes leaves and phytopharmaceuticals. available. are L.)accessions from2009; SepSci Portugal. J performance liquid chromatograpperformancehy liquid

- capillary and high electrophoresis array UV and mass spectrometric spectrometric mass and UV array

Caucalis platycarpos Caucalis tion of bilobalideandtion and Reseda and luteola J ChromatogrJ A - Verbeke J. J. Verbeke L.), using L.), using

on on 31 . J . J –

Chapter 2

Chapter 3

Spectrophotometric comparison of the content of chlorophylls in weld

The content of this chapter is largely that of the following paper: Villela A, Derksen GCH, Zuilhof H, van Beek TA. Spectrophotometric comparison of the content of chlorophylls in weld (Reseda luteola). Anal Methods 2011; 3(6): 1424–7.

Chapter 3

Spectrophotometric comparison of the content of chlorophylls in weld

3.1. Introduction, material and methods, results and discussion, and conclusion The greenish hue to which reference is made in the introductory chapter of this thesis could be caused by one of the two factors elaborated upon in this paragraph. First, it could be an intrinsic property of the use of the flavone-based dye; the colours of textiles dyed with mordant dyes are influenced by the metal salt used as mordant [1]. Second, chlorophylls a and b (the main and intensely green pigments of terrestrial plants) are present in the aerial parts of weld and it was hypothesised that they could bind to metal-treated fibres due to Lewis acid– Lewis base interactions. Based on this hypothesis, quantitation of chlorophylls in weld samples was of interest as part of the characterization of the raw material for production of the dye-extract. With this as background, the development of a simple analytical method to compare the relative amounts of chlorophylls and their breakdown products in different weld samples was aimed at. Because of the expected large number of samples, low manpower input was of paramount importance. A first consideration in such a development is the observation that chlorophylls are unstable and can be broken down, e.g., by enzymatic activity, light exposure or under acidic conditions [2, 3]. As harvested weld is dried under fairly rough conditions, it not only contains chlorophylls a and b (chls a and b), but also their breakdown products, including pheophytins (magnesium atom replaced by two hydrogen atoms), chlorophyllides (no phytyl chain) and pheophorbides (magnesium atom replaced by two hydrogen atoms and no phytyl chain) [4, 5]. To minimally modify the composition of chls and breakdown products of dried An analytical method for the comparison of the content of chlorophylls and their structurally weld plant material, precautions should be taken during the analysis. The extraction needs to similar breakdown products in weld is described. be fast, carried out in dim light, and use cold solvents of neutral pH [2, 3, 5] The quantitation step either immediately follows the extract preparation [2, 3] or the compounds should be stable in the extract [5]. Chls can be detected by spectrophotometry, spectrofluorimetry or mass spectrometry [4, 6]. Whereas spectrofluorimetry is more sensitive than spectrophotometry, quenching of fluorescence takes place in highly concentrated solutions [7] and spectrophotometry (without a preceding HPLC step) is simpler and faster to use. In addition, it is expected that only chls and some of their breakdown products in weld absorb light at wavelengths >630 nm. Chls a and b, chlorophyllides a and b, pheophytins a and b, and pheophorbides a and b display two main absorption bands: The blue one (Soret), at wavelengths <500 nm and the red one, between 640 and 670 nm [5]. Despite the main blue bands being more intense than red ones, the latter must be used due to increased selectivity. The red bands of chls and chlorophyllides are centered at nearly the same wavelength [5, 8]. Those of pheophytins and pheophorbides are 4 to 11 nm red-shifted and the molar absorption coefficients are about 40% smaller relative to chls [5, 8]. Thus, the red absorption band of extracts of weld (Figure SI B.1, appendix B) comprises the red bands of multiple compounds. The absorbance at its maximum was used to compare different samples.1

1 Although not done during this work, it could be tested whether weld samples containing different chls/chlorophyllides-to-“pheopigments” ratios could be compared after conversion of chls and chlorophyllides to corresponding “pheopigments” using HCl (see section “Pheophytinization under

Defined Conditions”—and two preceding sections—of ref. 3 for elaboration on the conversion of chls to pheophytins using the acid).

34 35

34 similar breakdown products An

analytical method fo r the comparison of the content of the comparison r in weld in is described. is

of chlorophylls and theirand of chlorophylls

structurally b 3.1. Introduction, m 1 was used compare to B appendix smaller relative pheophorbides nm 670 640and one, between extract the in stable contain conditions unstable samples dye was samples Lewis weld mordant dyes flavone use of the property the intrinsic of by one caused T chlorophyllides ones, display main two and>630 someproducts ofnm. weldlight theirin absorb at breakdown wavelengths step) HPLC a preceding fluorescence step be chain was ofimportance. paramount compare Whereas Whereas weld and chain) pheophytins atom (magnesium to pheophytinsto using the acid). chl Defined Conditions” chlorophyllides corresponding to “pheopigments” using se HCl (see Although not during done work, this could it testedbe whether weld samples containing different he green he (the main A first consideration in such a A in first consideration Chl Chl s/chlorophyllides fast - extract. With this as background as , this extract. With

either ) and plant material,

the latter must , s detected be can s [4, 5] [4, s base base

carried out in dim light carried dim in out was aimed at aimed was

a broken down, broken be can and chlorophylls the the it was hypothesised that hypothesised was it ish

spectrofluorimetry

and

) pheophorbides ( pheophorbides [2, 3] [2, immediately follow and intensely green intensely and of interest of partas of characterization the of the raw material for production the of . The absorbance at its m its at absorbance compounds. The multiple of bands the red comprises could thesis be this chapter of introductory referencethe in made is hue which to relative relative

interactions takes place place takes T used as salt mordant used metal the by influenced are

are are b, chlorophyllides a o minimally of thechl modify composition

to to of the are centered at at centered are

harvested harvested As . absor chl - 4 to 11 nm red nm - 11 4 to and —and two preceding sections to [5] different samples different precautions should be should taken precautions - aterial and methods, and r aterial amount . s “pheopigments” ratios could compared be conversion after chl of

Because of the expected large number of samples, low manpower input lowmanpower input of expected samples, Because number large of the . ption bands: ption two factors two [5, 8] be be

by sp by . and is is s highlyin solution concentrated Based of quantitation chlorophylls hypothesis, onthis weld in use simpler and faster and simpler magnesium atom replaced by two hydrogen atoms hydrogen two by replaced atom magnesium

s . b

ectrophotometry, Thus, the red absorption band of extracts of weld

of chlorophylls theirin breakdown and products d

is more sensitive than spectrophotometry sensitive more is ,

( [5] s replaced by two hydrogen two by replaced nearly nearly pigments ofpigments terrestrial chl and

due to the extract preparation extract the

weld weld shifted .

they could bind to metal to bind they could T e.g., s Despite elaborated upon in this in paragraphelaborated upon . and use use a he blue one blue he development is the observation that c that the observation is development . the same wavelength same the 1 and

fairly under fairly dried is by by the development of a the development cold solvent b, pheophytins a increased selectivity increased and

enzymatic activity enzymatic b) the the esults and discussion, and c

- the the of ref. 3 —of ref. , to to ; dye based

spectrofluori bu bands main blue bands use (Soret) mo duringanalysis the t also . In addition, it is expected that only chl thatonly In expected is . it addition, lar lar s ofpH s neutral

plants)

are are absorption coefficients for elaboration on the conversion of at wavelengths < wavelengths , at

their breakdown products, including their breakdown [2, 3] [2,

[7]

and s ands breakdown products of atoms the coloursthetextiles dyed of with -

metry treated fibres due to Lewis acid Lewis due to fibres treated [5, 8] [5,

and [1] , are are . rough conditions rough b, and pheophorbides a

light exposure or or The The simple ction “Pheophytinizationunder ) being . Second ,

present spectrophotometry . Those of pheophytins . Those or mass spectrometry mass or chlorophyllides the compounds should be should the compounds

[2, 3, 5] 3, [2, . T red bands

he he more intense thanmore intense analytical analytical First, itcould

onclusion in the aerial parts of of parts aerial the in , c 500 nm and500 nm extraction extraction hlorophylls a hlorophylls

, The q The

hlorophylls or under acidic or under acidic

quenching of of quenching ( different different and no phytyl and nophytyl

F , it of of igure igure

about uantitation meth (no phytyl (no phytyl

chl not only only not (without (without need aximum the red red the SI B SI

s and

be anbe and od s and and s [4, 6] [4, dried weld weld 40%

chl s and and and and red are are

35 to to to to .1, .1, – b s s

Chapter 3 The method for analysing of chls and their breakdown products in weld was evaluated. Table 3.1. Absorbance of extracts of weld samples due to chlorophylls and their structurally similar breakdown Simultaneous extraction of 20 samples followed by sequential absorbance measurements was products obtained with different solvents and methods of extraction, using three types of cuvettes. a,b observed to be unsuitable, as the absorbance of identical samples gradually increased going c Plant Extraction procedure; sample treatment and d Absorbance (AU) Entry n Solvent e from sample 1 to sample 20 (appendix B). Absorbances of the extracts of samples that were material series of measurements and sr (%) o put to steep (soak) for 2 h (4 C) after dynamic extraction with acetone were higher than of 10 mm-pathlength quartz cuvette those in which the measurements immediately followed the extraction (entries 1 through 4 of 1 B 4 3 min vortexing; 5 min centrifugation Acetone 0.146 (2.6) Table 3.1), indicating an on-going extraction. The 5 min centrifugation step after the 2 B 4 3 min vortexing; 5 min centrifugation; 2 h (4 oC) Acetone 0.161 (2.3) extraction also increased the yield (appendix B). Thus, for a suitable comparison of the 3 B 4 10 min sonication Acetone 0.162 (1.2) content of chls and breakdown products in weld samples, the contact time between sample 4 B 4 10 min sonication; 2 h (4 oC) Acetone 0.177 (3.0) 2 and solvent has to remain constant. Centrifugation followed by transfer of the supernatant to 5 B 4 16 h shaking Acetone 0.241 (0.7) the cuvette without filtration was unsuitable. Results suggested scattering of light due to 6 B 4 16 h shaking; 2 h (4 oC) Acetone 0.227 (1.5) 3 turbidity of the extracts (data not shown), indicating the need for sample filtration. 7 B 3 3 min vortexing Acetone 0.125 (4.9)

The yield of the 3 min vortexing extraction with acetone at room temperature is about 50% MeOH–H2O 8 B 3 3 min vortexing 0.150 (6.0) of what can be extracted after 24 h with this—for chlorophyll—not very efficient solvent (9:1) (entries 5 and 7). Regarding the choice of solvent: Absolute ethanol extracts about 75% of 9 B 2 3 min vortexing Abs EtOH 0.163 (0.5) 4 what the superior DMF extracts but is more efficient than acetone (entries 7, 9, and 12). In 10 B 3 3 min vortexing EtOH 96% 0.168 (0.7) addition, among the tested extraction procedures, vortexing is the fastest and probably mildest, 11 B 3 3 min vortexing MeOH 0.183 (4.8) and ethanol has the advantage of low toxicity and being environmentally friendly. Finally, the 12 B 3 3 min vortexing DMF 0.215 (0.6) presence of water in the solvent leads to formation of hydroxylated allomers [11]. Also, water 2 mm-pathlength plastic cuvette in the solvent leads to a red-shift of the main absorption bands of chls and reduction of their 13 A 1 3 min vortexing; day 1 Abs EtOH 0.022 5 molar absorption coefficients [3]. Thus, the final analytical procedure consists of a 3 min 14 A 3 3 min vortexing; day 7 (morning) Abs EtOH 0.023 (3.5) extraction with absolute ethanol by vortexing at room temperature, followed by syringe- 15 A 4 3 min vortexing; day 7 (afternoon) Abs EtOH 0.021 (2.5) 6 filtration of the extract and absorbance measurement. 16 E 1 3 min vortexing; day 1 Abs EtOH 0.128 17 E 4 3 min vortexing; day 7 (morning) Abs EtOH 0.129 (3.3) 2 No difference in extraction efficiency among weld samples from different plant parts is expected. 18 E 4 3 min vortexing; day 7 (afternoon) Abs EtOH 0.120 (1.4) This is indicated by the linearity of the relation seen ahead (Figure 3.1 and preceding discussion). 19 F 1 3 min vortexing; day 1 Abs EtOH 0.424 3 Optical clarity of solutions was checked at 700 nm [9]. 4 Toxicity of DMF has hampered its recommendation for routine work [5]. DMF is also viscous and 20 F 4 3 min vortexing; day 7 (morning) Abs EtOH 0.420 (5.0) has a “faint amine odour” [10]. 21 F 3 3 min vortexing; day 7 (afternoon) Abs EtOH 0.427 (2.8) 5 Ethanol absolute is: (i) 3× more expensive than ethanol 96%, but still ∼10% cheaper than methanol 22 G 1 3 min vortexing; day 1 Abs EtOH 0.982 and (ii) highly hygroscopic [10]. 23 G 4 3 min vortexing; day 7 (morning) Abs EtOH 0.936 (4.5) 6 Final analytical procedure: (125.0 ± 0.9) mg of sample are weighed in centrifuge tube (10 mL, Oak 24 G 4 3 min vortexing; day 7 (afternoon) Abs EtOH 0.931 (2.4) Ridge, PPCO, Thermo Scientific, Rochester/USA); 5 mL of absolute ethanol added via a 10 mL 25 H 1 3 min vortexing; day 1 Abs EtOH 1.048 −1 graduated cylinder; 3 min vortexing (MS2 Minishaker, IKA Works—Wilmington/USA; 2,500 min ) 26 H 4 3 min vortexing; day 7 (morning) Abs EtOH 0.976 (7.4) of each sample individually; syringe-filtration (5 mL PP syringe; syringe-filter, PTFE, 13 mm, 0.45 27 H 4 3 min vortexing; day 7 (afternoon) Abs EtOH 0.957 (5.8) µm, Grace Davison Discovery Science) of extracts into either a 2 mm-pathlength disposable plastic cuvette (Eppendorf) or a 10 mm-pathlength disposable plastic cuvette (Greiner Bio-One); use of 1 mm-pathlength quartz cuvette disposable plastic cuvette caps; absorbance of subsequent extracts was measured at 3 min intervals 28 A 3 3 min vortexing; day 2 (n = 1) and day 9 (n = 2) DMF 0.015 (2.6) (Cary 100 Scan UV–Visible spectrophotometer—Varian/Australia; double beam mode; λs: 665 nm 29 E 3 3 min vortexing; day 2 (n = 1) and day 9 (n = 2) DMF 0.108 (2.1) and 750 nm; spectral bandwidth: 1.5 nm; signal averaging time: 1 s; light beam: 10 mm high, starting 3 min vortexing; day 2 (n = 1), day 9 (n = 2), and 30 F 4 DMF 0.363 (7.3) at 15 mm from bottom of cuvette holder). Note: (i) Centrifuge tube: held in place during extraction day 9 (n = 1) with supports and clamps. (ii) First drops of syringe-filtered solutions: discarded to prevent possible 31 G 3 3 min vortexing; day 2 (n = 1) and day 9 (n = 2) DMF 0.695 (1.1) changes in concentration. (iii) The 2 mm-pathlength plastic cuvette needs to be lifted up with the lever 32 H 3 3 min vortexing; day 2 (n = 1) and day 9 (n = 2) DMF 0.765 (0.2) of the cuvette holder of the spectrophotometer and to be filled over the 500 µL-mark, as its optical a surface is only 3 mm high and starts 7 mm from the base. (iv) Procedures were conducted under λ of detection: Entries 1 through 12 and 28 through 32, absorbance at 664 nm minus absorbance at 750 nm (664 reduced light. (v) Possibility of further simplification by using an automatic pipette for the addition of − 750 nm); entries 13 through 27, 665 − 750 nm. Note: 1) Subtraction of absorbance at 750 nm needed due to the solvent—most likely with a concomitant gain in precision—and a vortex mixer equipped with a tube holder.

36 37

tube holder.tube solventthe reduced light. (v) Possibility further simplification of by using automatic an pipette for the addition o onlysurface is 3mm starts and 7mm high base. (iv) from Procedures the conductedwere under theof cuvette holder of the spectrophotomet concentration.changes in 2mm (iii)The with supports and clamps. First(ii) drops of syringe at at 15 mm from bottom of cuvette holder 750and nm; spectral bandwidth: 1.5 nm; signal averaging time: 1 s; light beam: 10 mm high, starting and (ii)and highly hygroscopic gradu Ridge, PPCO, Thermo Rochester/USA); Scientific, 5 mL ethanol absolute of via 10 added a mL ( disposable plastic cuvette caps; absorbance of subsequent extracts was measured 3 min at intervals cuvette (Eppendorf) or a 10 mm (Eppendorf) 10 a cuvette or µm, Grace Davison Discovery Science) of extracts into either a2 mm eachof sample individually; syringe has and has solvent content of chl molar absorption coefficient molar absorption the leadsin solvent a to presence of the water in what DMF the superior 7) and 5 (entries aftercted this 24hwith extra be can what of yield the increased also extraction T those steep to put (soak) for 2h( and ethanol has the advantage of ad the cuvettewithout filtration was 36 6 5 4 3 2 absorbance measurement and filtration extract of the extraction with absolute turbidity the of extracts shown) not (data from sample sample 1to observed unsuitable be to S has a “faint amine odour This is indicated by the linearity of the relation seen ahead (Figure 3.1 and preceding discussion). Final analytical procedure: ( Ethanol absolute 3× (i) is: more than expensive ethanol 96%, but still ∼ Toxicity DMF hampered of has recommendation its routine workfor [5] Optical clarity of solutions was checked 700 at nm [ No difference extractionin efficiency among weld samples from different plant parts is expected. Cary 100 Scan UV imultaneous extraction of 20 samples followed by sequential absorbance measurements was by absorbance sequential of extraction imultaneous 20samples followed able dition The The The methodfor ated cylinder; in 3.1) yield

, among the tested extraction procedures, vortexing is the fastest and probably mildest, and vortexing the fastest procedures,is extraction , among tested the whi —most likely with a concomitant gain in precision

ch ch

of the 3 the of indicating weld ands breakdown products weld in the the to to .

3 min Regarding the choice of solvent: A choiceof solvent: the Regarding remain remain –

measurements immediately followed the extraction analysi Visible spectro extraction min vortexing extraction ”

4 [ a solvent vortexing (MS2 Minishaker, IKA Works IKA Minishaker, (MS2 vortexing 10] extracts but is more effic red [ n on- 10] , as , 20 ethanol constant ng 125.0 . -

4 shift of the main absorption bands ofshift thechl mainabsorption s . (

appendix B appendix t of of o [3] of identical samples samples identical of absorbance he C going

- leads formation to of hydroxylated allomer

pathlength disposable plastic cuvette (Greiner Bio

) ± low toxicity chl .

- photometer 5

by vortexing after after . filtration (5 mL PP syringe; syringe

0.9 Thus, 2 suggested suggested unsuitable. Results Centri s andproductss their breakdown ( - pathlengthplastic cuvette needs be to liftedwithup the lever ) appendix B appendix ) extraction . Note: (i) Centrifuge tube: held in place during extracti mg sample of are weighed in centrifuge tube dynamic , 3 ) er and and er to filled be over the 500µL indicating the the . fugation followed bytransfer A —

bsorbance and and final analytical final with acetone acetone with Varian/Australia; double beam mode; —

. 9] - at roomat temperature extraction extraction filtered solutions: discarded to prevent possible for chlorophyll ). ). being environmentally friendly . The The samples, the contact time bet time contact the samples, ient than acetone (entries 7, 9, and 12). 7, 9,and (entries ient than acetone . the need for sample filtration. Thus, 6

bsolute ethanol extracts about 75% of 75% of ethanolbsolute extractsabout s 5

of samples of that we extracts the of — min centrifugation step for for is about 50% about 50% is temperature room at of of than higher were acetone with and a vortex a and mixer equipped with a

—Wilmington/USA; 2,500 min procedure procedure a not very—not efficient solvent suitable comparison of the comparison ofsuitable the - pathlength disposable plastic scattering light of due to gradually increased going going increased gradually 10% cheaper10% than methanol -

. DMF viscous also is and filter, PTFE, 13mm, PTFE, filter, 0.45 , followed by syringe by followed , in s ands reduction

(entries (entries

weld of the supernatantof to c onsists - s mark, optical asits [11]

w 1 ween sample sample ween as - . Also

. through One); use of (10 mL, Oak

Finally of

s 65 nm 665 λs: evaluated. evaluated.

after the the after a of their of their , 3 water water

, 4 min the the − on on of of In In re 1 - f )

products 10 9 8 7 6 5 4 3 2 1 10 mm Table 3.1. a 32 31 30 29 28 mm 1 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 mm 2 12 11 − 750 nm); entries 13 through 27, 665 − 750 nm. 750 − 665 27, through 13 entries nm); 750 − Entry

λ of detection: E detection: of λ

- - pathlength quartz cuvette cuvette plastic pathlength - pathlength quartz cuvette quartz pathlength obtained B F A A B B B B B B B B B B B material H G E A H H H G G G F F F E E E A Plant Plant

Absorbance

ntries 1through 12and 28through 32, absorbance at 664nmminus absorbance at 750nm (664

with with different of extracts of weld samples due to chlorophylls and their structurally similar breakdown breakdown similar structurally their and chlorophylls to due weld samples of extracts of 3 4 3 3 3 3 4 4 1 4 4 1 3 4 1 4 4 1 4 3 1 3 3 3 2 3 4 4 4 4 4 4 n

3 Extraction; procedure 3 3 day (n 9 =1 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 16 16 10 10 5 vortexing; min 3 3 seriesmeasurements of 3 min min min vortexing min vortexing min vortexing min vortexing min minvortexing; day 7(afternoon) (morning) 7 day vortexing; min 1 day vortexing; min minvortexing; day 7(afternoon) (morning) 7 day vortexing; min 1 day vortexing; min minvortexing; day 7(afternoon) (morning) 7 day vortexing; min 1 day vortexing; min minvortexing; day 7(afternoon) (morning) 7 day vortexing; min 1 day vortexing; min minvortexing; day 7(afternoon) (morning) 7 day vortexing; min min vortexing min vortexing min vortexing min vortexing min vortexing min vortex min ; day 2(n = 1), day vortexing min 9(n = 2),

solvent h shaking h shaking min sonication min

vortexing vortexing; day 1 day vortexing; sonication s

and method and ) ; ing; 5

2 h (4

; day 2(n = 1) and day 9(n = 2) ; day 2(n = 1) and day 9(n = 2) ; day 2(n = 1) and day 9(n = 2) ; day 2(n = 1) and day 9(n = 2) ; 2h (4

min centrifugation min centrifugation min o C)

Note: Note: c

s sample treatmen

of extraction of o C) 1) Subtraction of absorbance at 750nmneeded due to

; 2h (4

using three type three using t

and

C) and and

– MeOH Acetone Acetone Acetone Acetone Acetone Acetone Solvent DMF DMF DMF DMF DMF EtOH Abs Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs EtOH Abs DMF MeOH 96% EtOH EtOH Abs (9:1) Acetone s

of cuvette of EtOH

H d

s .

a,b 0.363 (7.3) 0.227 0.241 0.177 ( 0.16 0.16 0.146 (2.6) and 0.765 (0.2) 0.695 (1.1) 0.108 0.015 (2.6) 0.957 (5.8) 0.976 (7.4) 1.048 0.931 (2.4) 0.936 (4.5) 0.982 0.427 (2.8) 0.420 (5.0) 0.424 0.120 (1.4) 0.129 (3.3) 0.128 0.021 (2.5) 0.023 (3.5) 0.022 0.2 0.18 0.1 0.16 0.150 0.12 Absorbance (AU)

15 (0.6 68 (0.7 s 2 1 3 3 5 r

( (1.5) ( (1. (2. (2.1) ( (0. ( (%) 6.0 0.7 3.0 4.8 4.9 2 3 5 37

) ) ) ) ) ) ) ) ) ) e

Chapter 3 large cuvette-to-cuvette variation of 2 mm-pathlength plastic cuvettes (range of readings at 750 nm: −53–89 extracts of sample H is ≤1.0 AU (entries 25 through 27),8 the method covers the entire range mAU); 2) Red-absorption maxima: Abs EtOH, 665 nm; other solvents, 664 nm. of concentrations of chls in weld samples. b A, B, E, F, G, and H are the codes of (dried and ground–sieved) weld samples in increasing concentration of Chls are situated inside the thylakoids of the chloroplasts of plant tissues, bound to chls and breakdown products. Note: 1) Differences among samples include cultivar and plant parts used;7 2) apoproteins through different types of bonding, including Lewis acid–Lewis base interactions Results of analyses of samples C and D are displayed in appendix B. [2, 5]. DMF is a better solvent than absolute ethanol for the extraction of chls and their c Always at room temperature and under reduced light. breakdown products (entries 9 and 12). Furthermore, there is only ∼10% difference between d MeOH = methanol, Abs EtOH = absolute ethanol, and DMF = N,N-dimethylformamide. the absorbance of the extracts of weld in acetone obtained by shaking overnight (entry 5), e 9 Average absorbance (sr = relative standard deviation). assumed to fully extract the chls and their breakdown products, and that of the extracts in DMF obtained by vortexing for 3 min (entry 12). Thus, to ascertain whether extraction with absolute ethanol was selective regarding the “types” of chls and plant parts being extracted, The repeatability {“precision under the same operating conditions over a short interval of the extraction of different samples of weld using absolute ethanol was compared with that time” [12]} of the method was assessed based on the relative standard deviation (sr) of using DMF (Figure 3.1). analyses of weld samples A, E, F, G, and H. Literature data from a study on the extraction of pigments from phytoplankton showed sonication in DMF to be a suitable reference method, as that method displayed excellent sr values (0.4–6.0%, depending on the pigment) and fulfils nearly all criteria for a suitable extraction technique, including precision [5]. The sr values of the assay presented here are typically 2–5%, and in all cases <7.5% (entries 13 through 27). Thus, the method displays good repeatability, more than meeting the demands for the intended purpose. The results of the two series of measurements of each weld sample were compared (entries 14 vs. 15, 17 vs. 18, 20 vs. 21, 23 vs. 24, and 26 vs. 27). In the cases of samples F, G, and H, the difference was ≤2%. This is smaller than that observed for samples A (13%) and E (7%). However, the absolute differences in absorbance of these samples are very low (A: 2 mAU and E: 9 mAU). The sr values are higher than those of samples F, G, and H due to the low absolute absorbance of the extracts of samples A and E. If increased precision is required for samples with absorbance ≤190 mAU, 10 mm-pathlength plastic cuvettes should be used (appendix B). Based on the analysis of 81 weld samples, with the highest absorbance being 1.165 AU (results not shown), sample H is expected to be near the upper practically occurring limit of concentration of chls and breakdown products in weld samples. Because the absorbance of the

7 Dried weld samples: A: Part of the plants: stems; cultivar: G; place of growth: Zeeland, SW of the Netherlands; harvested in: 2008; pore size of sieve used during grinding–sieving: 0.25 mm;

B: aerial parts; H; Groningen, N of the Netherlands; 2007; plants were placed outdoors during daytimes and indoors during night times for ∼10 days; 0.25 mm; C: stems; G; Zeeland; 2009; 0.25, 0.50, or 0.75 mm. Note: grown in another field as A; D: reproductive parts, leaves, and upper parts of the stems; rest, as B; E: leaves and reproductive parts; G; Zeeland; 2008; 0.25 mm; F: leaves and reproductive parts; G; Noord Brabant, S of the Netherlands; 2010; 0.25 mm; G: leaves and reproductive parts; G; Zeeland; 2010; 0.2 mm; 8 If sample H would be ground–sieved using a 0.25 mm sieve, the absorbance of its extracts would be H: leaves; G; Noord Brabant; 2010; 0.2 mm. Note: harvested 3 weeks earlier than F. lower than that seen in Table 3.1 (see also entries 10–12 of Table SI B.1, appendix B). Note: Samples A, C, and E though H were supplied by the industrial partner of the project. The 9 Extraction does not seem to continue upon steeping (2 h/4 oC) (entries 5 and 6). This is not the case general drying procedure used by them was: (root-removed) plants were placed indoors (roughly 30 oC; for the other extraction procedures (entries 1 through 4). An extraction efficiency experiment would be dark) for 2–2.5 weeks, on average. needed to check this assumption.

38 39

chl R dark) for 2 and However, the absolute differences in absorbance of these samples samples these of absorbance in differences absolute the However, the criteria all for a suitablenearly extraction technique, including precision time” { repeatability The e d c b mAU); cuvette large 38 7 concentration (results shown) not mm ( 10 mAU, ≤190 absorbance with samples absolute absorbance 14 intended purpose. T are here presented assay the th as pigments weld of analyses Note: Samples A general drying procedure used by them was: (root-

Dried weld Dried samples: appendix B appendix

Always light.and reduced under at temperature room absolute ethanol methanol,MeOH = Abs =absolute EtOH G F E D C daytimes indoors and during night times B 2008;in: pore size sieve of duringused grinding H A A esults analyses of of C samples Average ( absorbance hus s and breakdown products. Note: 1) D : leaves: and reproductive parts; G; Noord Brabant, S : leavesand reproductive parts; : aerialparts; , : reproductive: parts, leaves, and upper parts of the stems; rest, as Zeeland; 2009; 0.25, 0.50,or 0.75mm. field another grownin as 2009; Note: : stems; Zeeland; G; : Part of the plants: stems; cultivar: G; place of growth: Zeeland, SW the Based onthe The results of the two series of measurements of each weld sample were compared were sample weld each of measurements of series two the of results The : leaves: and reproductive parts; G; Zeeland; 2010

: : leaves; G;Noord Brabant; 2010; 0.2mm. harvested Note: 3weeks earlier than vs. difference was ≤ was difference B E at displays , the method displays ,

: 9 mAU 2) Red 2)

E 15, 17 [1 method displayed , 2] F –2.5 weeks, on average. from phytoplankton showed s from phytoplankton , } - G to - ). absorptionmaxima: Abs EtOH, 665nm; other solvents, 664nm. , - of the method was assessed

vs.

cuvette variation of 2 of variation cuvette of of and and ) . T

, chl H; Groningen, Netherlands; N the of 2007; plants were placed outdoors during 18, 20 analys samples

H C he he ,

s , and onditions over a over shortinterval of conditions operating same the under “precision s ands breakdown products

are

r 2%. s of the extracts of samples samples of extracts the of = relative standard deviation). standard relative = ample ample s r the codes of (dried and ground – i

vs. value s

This is smaller than that observed for samples

excellent A of 81weld samplesof

21, 23vs. though and and , H E typically 2 typically s are higher are s good

, D is is

G; ZeelandG; F

are displayedare appendix in B mm- , expected expected

G H s ifferences samples cultivar among include and plant parts used rep r

ahegh lsi cvte (ag o raig a 70 m −53 nm: 750 at readings of (range cuvettes plastic pathlength values (0.4–6.0%, depending onthe pigment) and fulfils 24, were supplied by industrial the partner the of project. The for ∼ for and eatability –5% reference asuitable be onication DMF in to and 26

; 2008mm; 0.25; , H than those of samples than those

10 days; 0.25mm; and DMF =N,N and DMF to be be to based on the relative standard deviation onthe ( relative standard based . , –sieving: 0.25mm

Literature data from a

and in all cases < cases all in and removed) were plants placed indoors (roughly 30 , in A with ; 0.2 mm; 0.2; -

, vs. weld pathlength be should used plastic cuvettes near near and sieved) more than meet than more

of the Netherlands;

27) t he he

E the the samples . .

- weld samples in increasing concentration of of concentration in increasing weld samples

In the cases of samples F samples of cases In the being highest being absorbance If increased precision is required is If for increased precision dimethylformamide.

upper upper

B ; .

; Because the absorbance of the of absorbance the Because 7.5%

practically occurring occurring practically F

of theNetherlands; harvested , ing

are very low ( study on the extraction of of extraction the on study G

2010 (entries 13through(entries 27) ,

and and for the the for demands the A

[5] ; 0.25; mm;

(13%) (13%) H F . . The

A due the to low ;

and

, s A r

1.165 AU 1.165 AU G

: 2 mAU : 2 mAU values values ,

method, E limit of (entries (entries and

s ( 7% r ) of ) of ; – 7 H o 89 89 of of 2) 2) C; C; ) . , .

needed checkto this assumption. the other extractionfor procedures (entries 1 through An 4). extraction efficiency experiment wouldbe assumed to fully extract the chl the extract fully to assumed extract the of absorbance the 5] [2, Lewis acid including bonding, apoproteins types of through different breakdown 9and (entries products of concentrations DMF DMF H sample of extracts 9 8 using the extraction of different samples of absolute lower thanthat seen Extraction does seem not upon continue to steeping h/4 (2 If sample H Chl DMF DMF . obtained by for 3 vortexing

s are situated inside the thylakoids of the chloroplasts of plant tissues, bound to chloroplasts of boundto ares the planttissues, thylakoids the of situated inside DMF a is better ethanol was ethanol (

F would ground be –sieved using 0.25 mm a absorbance sieve, the of extractsits would be igure of chl of

in in Table 3.1 (see also entries 10 3.1).

1.0 AU (entries 25 through 27) ≤1.0 AU 25 (entries is selective s

solvent thansolvent absolute weld samples. weld s

regarding the “types” of chl of “types” the regarding of of s ands their breakdown products weld weld min ). 12). weld weld (entry 12). in acetone obtained by shaking overnight acetonein obtained by 5), (entry shaking Furthermore, there only is

using –12 of SITable B.1, appendix B). ethanol

Thus, absolute o C) (entries 5 and 6). This not is casethe , for the of extraction chl to to 8

t he he s ascertain whether extraction whether with ascertain ethanol was compared compared was ethanol being extracted being and plantparts method , 9 and that of the extract the of that and ∼ – 10% difference between between difference 10% Lewis base interactions base Lewis c overs

the the entire s and their ands their with

range range

s that

39 in in ,

Chapter 3 3.2. Supplementary material Appendix B contains information supplementary to that in this chapter.

3.3. References [1] Ferreira ESB, Hulme AN, McNab H, Quye A. The natural constituents of historical textile dyes. Chem Soc Rev 2004; 33(6): 329–36.

[2] Harborne JB. Phytochemical methods – a guide to modern techniques of plant analysis. 2nd ed. Chapman and Hall: London/New York; 1984. [3] Lichtenthaler HK. Chlorolphylls and carotenoids: pigments of photosynthetic biomembranes. In: Packer L, Douce R, editors. Plant cell membranes, Academic Press: San Diego/New York/etc.; 1987, p. 350–82. [4] Holden M. Chlorophylls. In: Goodwin TW, editor. Chemistry and biochemistry of plant pigments. 2nd ed, Academic Press: London; 1976, p. 1–37.

[5] Jeffrey SW, Mantoura RFC, Wright SW, editors. Phytoplankton pigments in oceanography: guidelines to modern methods. 1st ed. UNESCO Publishing: Paris; 1997. [6] Schoefs B. Chlorophyll and carotenoid analysis in food products. Properties of the pigments and methods of analysis. Trends Food Sci Technol 2002; 13: 361–71. Figure 3.1. Correlation between absorbance of ethanolic and DMF extracts of [7] In vitro determination of chlorophyll a and pheophytin a in marine and freshwater algae weld. Average absorbances are plotted; error bars = 1× standard deviation. by fluorescence. National Exposure Research Laboratory — USEPA: Cincinnati; 1997. Ethanolic extracts: 2 mm-pathlength plastic cuvette; n = 8 (A and F) and n = 9 (E, G, and H). DMF extracts: 1 mm-pathlength quartz cuvette; n = 3 (A, E, G, [8] White RC, Jones ID, Gibbs E. Determination of chlorophylls, chlorophyllides, and H) and n = 4 (F). pheophytins, and pheophorbides in plant material. J Food Sci 1963; 28(4): 431–6. [9] Vernon LP. Spectrophotometric determination of chlorophylls and pheophytins in plant extracts. Anal Chem 1960; 32(9): 1144–50. The linear relation between the extracted amounts of chls and breakdown products by [10] O’Neil MJ, Smith A, Heckelman PE, Obenchain Jr. JR, Gallipeau JAR, D’Arecca MA, absolute ethanol and by DMF indicates that the extraction with absolute ethanol is not Budavari S, editors. The merck index – an encyclopedia of chemicals, drugs and biologicals. selective and, thus, that absolute ethanol can be suitably used as solvent for the comparative 13th ed. Merck and Co.: Whitehouse Station; 2001. method described here. Additionally, it suggests that the ethanolic solution is not saturated throughout the possible range of concentration of chls and breakdown products in weld [11] Woolley PS, Moir AJ, Hester RE, Keely BJ. A comparative study of the allomerization samples. reaction of chlorophyll a and bacteriochlorophyll a. J Chem Soc, Perkin Trans 2 1998; (8): The method at hand rapidly compares the total absorbance of porphyrin ring-containing 1833–9. pigments (chls and their structurally similar breakdown products) in different weld samples. It [12] Validation of analytical procedures: text and methodology Q2(R1), comprising Q2A is expected to minimally modify the composition of chls and breakdown products of samples (1994) and Q2B (1996). International Conference on Harmonisation of Technical and displays acceptable precision. Analysis of forty samples—including data processing— Requirements for Registration of Pharmaceuticals for Human Use; 2005. could be carried out in one working day, requiring a total amount of 200 mL of ethanol (5 mL per sample). Moreover, the method covers the entire range of occurring concentrations, uses ethanol as a mild, non-toxic extraction solvent—that is not selective regarding the “types” of chls and plant tissues being extracted—and is simple. Finally, although developed for the comparison of the chlorophyll content in weld, it is expected that—after additional validation—it could be applied on other plant materials too.

40 41

Ethanolic extracts mm - : 2 weld 3.1. Figure validation comparison of the chlorophyll content weld in chl as ethanol per sample per could and display pigments ( samples. 40 is expected to minimally modify the composition of chl rangethroughout of the chl of possible concentration method described here selective absolute chl of amounts extracted the between relation linear The ( H and E , compares The compares methodat hand rapidly s and plant tissues being extracted tissues plant and s G . , ) and n = 4( = n and ) Average absorbance be carried out i be out carried and

and ethanol and indicates byDMF —it c H chl a mild, non- a mild, ) Correlation between absorbance of s . ). DMF). extracts:

, Moreover, Moreover, s acceptable acceptable thus, that absolute ethanol can be suitably used as solvent for the comparative comparative for the suitably as solvent used can absolute ethanol be that thus, and their st productsructurally breakdown similar ould beould applied onother plantmaterials too. F ) .

n one working day n one . toxic toxic pathlength the method the Additionally, it suggests that the ethanolic solution is not saturated not is Additionally, thatthe suggests ethanolic it solution s precision. Analysi

are plotted

1 mm - 1 extraction extraction pathlength quartz cuvette pathlength plastic cover ; error; bars —and is solvent , requiring a total amount of 200 a amountof , requiringtotal cuvette s

the total that thethat s ethanolic and DMF extract the entire range of occurring concentrations occurring range of entire the of forty samples forty of not selective regarding the “types” of of “types” the regarding —that selective is not ; n ; 8 =

= simple 1×

, it is expected that expected is it , absorbance absolute extraction with absolute

( standard deviation A . s and ; n =3( n ; s ands breakdown product

Finally, developed although for the and breakdown of samples products F s ands breakdown products ) and ) including data—including processing A weld samples weld different ) in of porphyrin , E

n = n = , s of of s G 9 . ,

mL mL after additional additional —after

of of

ring ethanol ethanol - containing s in

( is not not is

, 5 weld weld uses uses

mL mL . — by by It

2nd ed. Chapman a Harborne[2] Phytochemical – JB. methods dyes. 33(6): 329–36. SocRev Chem 2004; Ferreira[1] ESB, HulmeAN, McNab QuyeThe A. H, natural constituentsof historical textile References 3.3. B chapter.Appendix this contains thatin informationto supplementary material3.2. Supplementary Requirements of Pharma for Registration (1994)International and Q2B (1996). Conference of Technical onHarmonisation Validation of[12] analytical and methodology procedures: text comprising Q2(R1), Q2A 1833–9. reaction chlorophyll a of HesterWoolley[11] Keely MoirAJ, RE, A comparativeBJ. PS, the allomerization of study 2001. ed.Station; 13th Whitehouse Merckand Co.: Budavari editors. T S, MA, D’Arecca JAR, Gallipeau JR, Obenchain Jr. PE, Smith A, Heckelman O’Neil MJ, [10] extracts. Anal 32(9): 1960; Chem 1144–50. pheophytinsand plant in LP. of Vernon chlorophylls determination [9] Spectrophotometric Food 28(4): 1963; Sci 431–6. J plantmaterial. pheophytins, in and pheophorbides RC, White ID,[8] Jones of E.Gibbs Determination chlorophylls,chlorophyllides, 1997. Cincinnati; —USEPA: Laboratory Research Exposure National by fluorescence. [7] pigmentsFood analysis. Technol and Sci 361–71. of methods 13: 2002; Trends the Properties of products. analysis food in Schoefs carotenoid B. and Chlorophyll [6] modernguidelines ed. to oceanography: 1st methods. U pigments editors. Phytoplankton in SW, Wright RFC, Mantoura Jeffrey[5] SW, London;pigments. p.1–37. Press: 2nded, Academic 1976, Holden M. In:Chlorophylls.[4] editor. Chemistry Goodwin TW, plant and biochemistry of York/ Diego/New San Press: Academic membranes, cell Plant editors. R, L, Douce Packer In: biomembranes. of photosynthetic pigments Lichtenthalercarotenoids: Chlorolphylls and HK. [3] In vitro

determination of chlorophyll a

etc. nd Hall: London/New 1984. York; nd Hall: ; 1987, p.350–82. 1987, ; he merck index index merck he and bacteriochlorophy

– an encyclopedia of chemicals, drugs and biologicals. ceuticals for 2005. Human Use; ceuticals and pheophytina a guide modern to techniques of plantanalysis. ll ll . J Chem Soc,Perkin Chem Trans (8): 21998; a. J NESCO Publishing: Paris; 1997. Paris; Publishing: NESCO i n marine and freshwater algae freshwateralgae and n marine 41

Chapter 3

Chapter 4

Photo-stability of the dye of weld in presence of aluminium ions

The content of this chapter is largely that of the following paper: Villela A, van Vuuren MSA, Willemen HM, Derksen GCH, van Beek TA. Photo-stability of a flavonoid dye in presence of aluminium ions. Dyes Pigments 2019; 162: 222–31

Chapter 4

Photo-stability of the dye of weld in presence of aluminium ions

4.1. Introduction Varying the metal ion mordant changes the colour of the dyed textile and has a major influence on its light-fastness [1, 2]. Increasing concentrations of aluminium ion lead to intensified colours of solutions of lut and the flavonols—3-hydroxyflavones—kaempferol and quercetin, as seen by the change of their UV–vis absorption spectra [3-5]. However, members of the flavones and flavonols classes of flavonoids have been reported to display diminished photo- stability in the presence of Al3+ [6, 7]. Thus, the influence of increasing concentrations of aluminium ion on the photo-stability of the dye of weld was studied to address the issue of whether a compromise between the obtainable colours and their photo-resistance would be a way to meet today’s requirement of light-fastness of the colours of textiles. The photo-stability of lut in presence of Al3+ at different lut–Al3+ ratios was studied in solution. Experiments using The main colouring compounds of the dye plant weld (Reseda luteola L.) are the flavones extracts of weld to dye wool premordanted with increasing quantities of aluminium salts were luteolin (lut), lut-7-O-glucoside and lut-7,3ʹ-O-diglucoside. Alum (an aluminium salt)- also carried out. premordanted wool dyed with weld leads to yellow colours that are of low resistance to Glycosylation may confer stability to flavonoid aglycones [8], [9] (chapter 42). For example, light. The photo-stability of lut in aerated methanol–water 8:2 (v/v) solution upon the antioxidant activity of flavonoids is diminished by glycosylation [8], [10] (chapter 5). Weld irradiation with light above 300 nm was studied at different lut–Al3+ ratios. Experiments contains an enzyme that breaks the glycosidic bonds of its flavone glucosides1 [11]. Lut, lmg using extracts of weld to dye wool premordanted with increasing quantities of aluminium and ldg are obtained if this glycosidase is inactivated before extracting the flavonoids. In salts were also carried out. The photo-stability of lut in the polar protic solvent and the contrast, only lut is obtained if the enzyme is not inactivated. Thus, there is an issue of whether photo-resistance (light-fastness) of the colour of weld-dyed wool decrease with increasing the glycosidase should be inactivated before extracting the flavonoids of weld in order to obtain concentrations of aluminium ions. Thus, the lower the [Al3+] used for mordanting the the most photo-stable dye. Experiments were carried out in the context of such an issue. The wool, the more light-fast its colour. Lowering the [Al3+] appears to have no negative relative photo-stability of lut, lmg and ldg in solution was studied, and results of the colours of influence on the wash-fastness of the colour. As the gain in light-fastness by the use of alum-mordanted wool dyed with them are presented. low [Al3+] to premordant the wool is not extensive, however, this does not seem to be a way to meet today’s requirement of light-fastness of the colours of dyed textiles by itself. Nevertheless, it may be part of a broader strategy to address the need for increased light- fastness of the colour of wool dyed with weld. Implementation of this approach by dyers is expected to clarify whether it results in benefits for textile dyeing practice.

1 Although there might be enzymes—and not only one enzyme—breaking the glycosidic bonds, the singular form of the word is used in this thesis.

44 45

irradiation lightwas with 300nm above at studied different lut light. premordanted wool dyed with we with premordanted wooldyed is expected to clarify whether itresults in benefits for textile dyeing practice. fastness of the [Al low luteolin ( Nevertheless, it m it Nevertheless, way to meet today’s requirement of light wash the on influence 44 wool, the more light also were salts of increasingwith alumini dye woolpremordanted quantities using weld to extractsof plantweld of the dye compounds The maincolouring concentrations of aluminium ions. Thus,the lower ions. the [Al concentrations of aluminium photo-

Th resistance ( resistance 3+ e- photo lut ] to premordant to however, extensive, this] not the woolis ) , colour of wool dyed colour of weld. with lut carried out. carried - light- stability of ay 7- O be - - fast its colour. its fast - glucoside fastness of thefastness colour. fastness) part of abroader of part

The- photo lut

of the colour of and

ld leads to yellow colours that leadsld to in aerated methanol aerated in lut Lowering the [Al the Lowering stability lut of - - strategy to address the need for increased light increased for need the address to strategy fastness of thefastness coloursby of dyed textiles itself. 7,3ʹ -

O As the gain light in As the Implementation approach of this by dyers

- weld diglucoside - ( dyed wool decrease with increasingdyed wooldecrease with

Reseda luteola Reseda in thein polar protic solvent and the water 8:2 (v/v) solution upon (v/v)–water solution 8:2 3+ ] appears] nonegative have to . 3+ Alum ] used for] mordanting the –Al

are are -

fastness by the use of by usefastness of the does not seem to does not 3+

(analuminium salt) ratios. of low of

L.) L.) are are Experiments Experiments

resista the the

flavone nce be a be um um

to to s - -

colour stability theof in presence Al flavones and of flavonols classes flavonoidsbeen havephoto- diminished reporteddisplay to UV their of change by the seen as contains t on Varying 4.1. Introduction 1 and out. carried also extracts of weld to dye premordanted wool with increasing quantities of aluminium salts were way to meet today’s requirement of light- whether a and compromise their between photo- the obtainable colours onthe ion photoaluminium - the the of alum relative- photo - photo the most contrast, only lut singular form word the of is used in thesis.this he ant he Although might there enzymes be lut G

glycosidase should be inactivated before extractin before inactivated be should glycosidase its ldg lycosylation lycosylation -

mordanted wool

in presencein of Al s ioxidant activity of flavonoi activity of ioxidant light-

of solutions of lut of solutions

the the enzyme an are obtained are metal ion mordant chang fastness stability lut of stable dye. stable

may confer stabilitymay confer aglycones to flavonoid is obt is

that [1, 2] [1,

dyed with them presented. are i 3+ ained if the enzyme is not inactivated. enzyme the if ained f this glycosidasef is inactivated before

breaks at different lut different at

and . Experiments were carried out in the context of such an issue. an of such context in the out carried were Experiments , Increasing concentrations of al of concentrations Increasing stability the of dye of weld was address to studied

lmg the the 3+

the the

and not only not enzyme—and one bonds,—breaking glycosidic the the

[6, 7][6, ds ds and and flavonols glycosidic of its flavone bonds vis absorption spectra–vis es is is ldg

diminished diminished –Al theof colour the dyed textile . fastness of thefastness colour Thus,

in solution was studied, and was studied, solution in

3+ —3-

ratio the influence of increasing concentrations of of concentrations increasing of the influence hydroxy wass solution. in studied by glycosylation by

g the flavonoids of weld in order to obtain order to in gflavonoids of weld the f lavone

[8] [3 uminium ion Thus, there is an is Thus, there issueof s -

, 5]

of textiles. s [9] extractin kaempferol an —kaempferol .

However, m glucosides

[8] (chapter 42). (chapter

and the the of results , [10] resistance resistance

has a major influence influence amajor has g

T lead to intensified Experiments using using Experiments the flavonoids.

he- photo (chapter 5) (chapter 1 [11] embers of the

For example, example, For

d quercetin, the issue o issue the would be a . colours of Lut whether whether stability . Weld Weld ,

lmg T 45 he he In In f

Chapter 4 4.2. Material and methods over time was quantified by RP-HPLC–UV using peak areas of the 350 nm traces (lut and lmg) The material and methods of the work is summarised in this section. Readers interested in and of the 340 nm traces (ldg), and calibration lines. Except for solution lut0.20, the [flavone] further procedures and details are referred to section SI C.1 (appendix C). of all samples—t0 (t = 0 h) through t24 (t = 24 h)—of each irradiated solution was comprised within the range of concentrations used for construction of the calibration lines. A straight line 4.2.1. Effect of different concentrations of Al3+ and glycosylation pattern of lut (aglycone, mathematical model was fit through the experimental data {[flavone] (mM) vs. time (min)}. monoglucoside, and diglucoside) on its photo-stability in solution This regression analysis was carried out for each of the solutions irradiated during experiments The following chemical compounds were used for the preparation of stock and working 1a–2b and special irradiation experiment via the least-squares method. 3+ solutions of the flavones, Al , and HNO3 used for construction of the calibration lines and The photodecomposition of the flavones was described as being of zero-order, in which the preparation of the solutions used for the irradiation experiments described: Lut (96% by NMR; rate of reaction equals the rate constant (d[flavone]/dt = k). As the integrated form of the zero- Indofine Chemical Company, Hillsborough/USA); lmg (93% by NMR; Extrasynthese, order rate law is [flavone] = kt + [flavone]0, the k values were obtained from the slope of the Genay/France); ldg (86% by NMR; Extrasynthese, Genay/France); aluminium nitrate equations (k = −slope). The relative rates of photodecomposition were calculated through the nonahydrate (99.997%; Aldrich Chemistry, St. Louis/USA); nitric acid (65%, extra pure; Merck, formula kj /ki, in which i was lut in lut0.11, lut0.10, or lut0.05, and j was lut, lmg, or ldg in other Darmstadt/Germany). solutions. Notes: 3+ Solutions of lut, lmg, and ldg—with and without Al —were prepared in aerated methanol– • Solution lut0.20 was used for comparison of the rates of photodecomposition of lut in water 8:2 (v/v). Nitric acid was added to all volumetric flasks, except to those containing lut lut0.20, lut0.11 and lut0.05. For this purpose, a procedure similar to that outlined above was and Al3+ in a ratio 1:10, because the lowering of the pH was due to the Lewis acid Al3+ in this carried out. Peak areas vs. time were used instead of concentrations vs. time, however, case (rather than to HNO3 or a combination of both). The apparent pH of all irradiated solutions and the slopes of the equations of the best-fit lines (y = ax + b) were compared; − was 3.6; the term apparent pH is used as the solvent of the solutions was not water, but the • All irradiated solutions contained NO3 , originating from HNO3, Al(NO3)3, or from both − methanol–water mixture. Stirred solutions were irradiated at 32 °C in glass cuvettes using a of them. As the oxidation state of the N-atom of the ion is +5, NO3 may be reduced in phosphor-coated low pressure mercury vapour lamp, which led to irradiation only with light different systems/under different conditions [12]. One such condition could be the above 300 nm, most importantly in the 300–420 nm range. The photodecomposition of lut, lmg, irradiation with light in the 300–420 nm range used in the irradiation experiments reported − and ldg—decrease of [flavone] over time—was monitored by RP-HPLC–UV. here, even with the [NO3 ] being ≤3 mM. It does not mean that the ion has necessarily Solutions used for the irradiation experiments were prepared on the day the experiments contributed to the photodecomposition of the flavones in the experiments of this study, started, and transferred to the glass cuvettes (3.5 mL; 10 mm light path) in which they were however. Reasons for this are: − irradiated. This was done without using Pasteur pipettes to prevent possible contamination. o If reduction of NO3 took place during the irradiation experiments, methanol may have Slots for four cuvettes were made on the cover plate of a magnetic stirrer. The distance from acted as reducing agent instead of the flavones [12]; methanol was present in solution the front window of the cuvettes to the centre of the lamp was 2.02 cm. Pictures of the set-up in quantities far larger than those of the flavones, as it was part of the solvent mixture; are seen in Figure SI C.1. Calibration solutions and solutions used for irradiation experiments o The relative rates of decomposition of the flavones (Table 4.2) do not match the − were kept in the dark at room temperature not less than 60 min before use. This was motivated relative [NO3 ] of the solutions: 1.4 vs. 1.2 (lut0.10–Al0.02), 3.2 vs. 1.9 (lut0.10–Al0.10), 3+ by the fact that flavonoid–Al complexation does not reach equilibrium immediately. 3.8 vs. 7.8 (lut0.10–Al1.00), 1.4 vs. 1.0 (lmg0.05), 0.2 vs. 1.0 (ldg0.05), and 0.6 vs. 1.4 Observations reported were made during five irradiation experiments. They are referred to (ldg0.05–Al0.05); − as 1a and 1b (1st irradiation replicate), 2a and 2b (2nd irradiation replicate), and special o The photo-stability of lut and ldg in solution (in presence of NO3 ) decreases with 3+ irradiation experiment. During the first four irradiation experiments, the eight solutions—lut0.10, increasing [Al ] (Table 4.2). This matches the results of the experiments in which lut0.10–Al0.02, lut0.10–Al0.10, lut0.10–Al1.00, lut0.05, lmg0.05, ldg0.05, and ldg0.05–Al0.05—were wool was dyed with extracts of weld (experimental details in sections SI C.1.13 and irradiated in duplicate. Solutions lut0.20 and lut0.10–Al0.99 were irradiated during the special SI C.1.15): The light-fastness of the colours of dyed wool decreases with increasing irradiation experiment, with a filter blocking radiation of wavelengths shorter than ~420 nm quantities of Al3+ used to premordant the wool (Tables 4.4 and 4.5); (420 nm cut-off filter) being placed in front of the cuvette containing lut0.10–Al0.99. Notes: In preliminary experiments, the photodecomposition of lut in nitric acid-acidified 3+ o • All solutions—flavones, Al , and HNO3—were prepared anew for use in irradiation aerated methanol–water 8:2 (v/v) solution was observed to be 35 to 55% slower than experiments 2a and 2b, and special irradiation experiment. in non-acidified solution (section SI C.2.2). 3+ − • Code of solutions used: flavone[flavone]–Al[Al], in which [flavone] = [flavone]0, Al = Al , Additionally, lmg and ldg appear to be stable upon overnight heating in presence of NO3 and concentrations are expressed in mM. at a concentration of 0.35 mM (sections SI C.1.12 and SI C.2.10). During the irradiation experiments, 0.05 mL aliquots of the solutions were collected at t = 1.0 h, 2.0 h, 3.0 h, 4.0 h, 8.0 h, and 24.0 h, as irradiations started on one day (day 1) and finished Spectrophotometric analyses were carried out using quartz cuvettes. On day 1, the remainders on the following one (day 2). During the special irradiation experiment, aliquots of the solutions of the solutions used for irradiation were diluted 5× with aerated methanol–water 8:2 (v/v). On were collected at t = 2.0 h, 4.0 h, 8.0 h, and 24.0 h. The photodecomposition of the flavones

46 47

werecollectedat = t 2.0h,4.08.0h, 1.0 h,2.0 3.0h, 4.0h,8.0 h, ( Genay/France); on the following one ( on the one following irradiation experiment,with a irradiated in duplicate. irradiation experiment. During the first four irradiation experiments, the e as 1a and 1b(1stirradiation water (v/v). 8:2 Darmstadt/Germany). Louis/USA);Aldrichnonahydrate nitric acid Chemistry,(99.997%; (65%, St. pure; extra Merck, Hillsborough/USA); Indofine Company, Chemical described experiments : irradiation the used for the solutions preparation of of solutions the flavones, Al working stock and of the preparation for used chemical compounds were The following monoglucoside, and diglucoside) on its photo 46 lut g the to started, and transferred and ldg above most 300nm, phosphor methanol was the 3.6; term apparent used pH is as the solvent of the ( case and Al C.1 SI section to referred are details and procedures further section. Readers this ofin the in The work interested and summarised materialis methods methods and 4.2. Material S irradiated different 4.2.1. Effect of Al flavonoid– that fact by the C. SI Figure in seen are cm the lampwascentrecuvettes of2.02 the to ofthe the front window kept in the in darktemperature before than 60min at kept room were less not use 420 nm cut420 nm lots for four for lots 0.10 During the irradiation experiments • O S of lut Solutions • olutions olutions bs rather than HNO to –Al and concentrations are expressed mM. in All solutions Code of solutions used: flavoneCode of solutions special and experiment 2a 2b,and irradiation experiments 3+ ervations reported were made during five irradiation experiments. They are referred to to referred are They experiments. irradiation five during made were reported ervations —decrease over time of [flavone]

in ain ratio 1:10, because the lowering of the pH was due the to Lewis acid Al - –water mixture Pasteur was. This doneusing without Pasteur 0.02 coated lowpressure which mercury ledirradiation to lamp, vapour onlywith light - , ) filter off were prepared on the day day the on prepared used for were experiments the irradiation

cuvettes cuvettes lut

Nitric acid was added to all volumetric flasks ldg 0.10 , —flavones, Al

lmg importantly in300–420 the –Al

day 2 ). day

being placed in front of the cuvette beingfront placedin of thecontaining lut (86% by NMR; Extrasynthese, Genay/France); aluminium nitrate

were made on the cover plate of of plate cover the on made were 1 , and , and 0.10 Solutions Solutions . 3 .

concentrations of Al of concentrations or a combination of both) Stirred solutions were irradiatedat 3 Calibration solutions and usedCalibration solutions for experiments solutions irradiation ,

3+ During s the lut ldg 3+ and 24.0h, as

complexation does reach not immediately.equilibrium , and H filter ),2b (2nd 2a irradiation and replicate lass cuvettes (3.5 mL; cuvettes lass 0.10 with andAl—with without 3+ lut –Al , and HNO blocking radiation of wavelengths shorter than ~420 nm than of shorter wavelengths radiation blocking 0.20 [flavone] , 1.00 NO 0.05 mL a

pecial irradiation experiment irradiation pecial and 24.0h. —was and , 3

lut –Al irradiations started onone day ( used for construction of the calibration and lines lut 0.05 3+ [Al] 3 nm nm - monitored by RP stability solution in were prepared anew for use in irradiationin use for anew prepared —were 0.10 and glycosylation lut pattern and of , liquots , in which [flavone] = which, in [flavone] [flavone] pipettes to preventpipettes to possible contamination. . r lmg –Al ange The The The

3+ lmg lmg 10 0.99 0.05

were prepared in aerated m aerated in prepared —were . a magnetic stirrer. apparent pH of of the solutions

The photodecomposition of lut The photodecomposition photodecomposition of the flavones ofphotodecomposition the flavones

, mm ( were irradiated during the special special the during irradiated were (93% by NMR; Extrasynthese, Extrasynthese, NMR; by (93% ldg

appendix C appendix solutions was water, not solutions the but , except to those to containinglut except , light path 2 . 0.05 - °C in glass cuvettes using a glass cuvettes a using in °C HPLC , and ldg

, aliquots of the solutions ofaliquots the solutions ), and special ), and special replicate . all 0.10 ). UV. –UV. ight solutions

Pictures of the set the of Pictures ) . were coll were

Lut in which they were whichin they were This was motivated

day 1 ) andday finished irradiated solutions –Al The dista 0.05 the

(96% by NMR; NMR; by (96% 0.99 –Al

. Note 0 experiment , Al =Al Al , ected at t = t at ected (aglycone, 0.05 nce from from nce 3+ ethanol

—lut were —were

s: in this ,

lmg - 0.10 up up 3+ – s , ,

, 1a This mathematical model was fit throughexperimental the data the rangewithin concentrations used forof A oflines. thestraight calibration construction line of all samples and of thetraces 340nm ( solutions. Note solutions. =kt [flavone] is law rate order rate of reaction equals thet rate constant (d[flavone]/d by RP over quantified time was usedof for the solutions irradiation anal Spectrophotometric formula equations ( –2b and special irradiation experiment via experiment the–2b andleast irradiation special T • • he photodecomposition of the flavones was described as being of zer regression analysis was carried for out irradiated each of the solutions during experiments Additionally, o o are:however. for this Reasons contributed the to photodecomposition of the flavones the in experiments study, of this [NO the with even here, irradiation with light in the 300 at a concentration of 0.35 mM (sectionsat C. amM SI concentration of 0.35 o o different systems of them.As thestate oxidation of the N lut Solution and t carried P out. All irradiatedAll NO contained solutions k 0.20 The- photo ( If of reduction NO in non- in methanol aerated In theexperiments, photodecomposition preliminary of lut wool was dyed with extractsweld of (experimental details in sections SI [Al increasing [NO relative match the donot (Table4.2) of the flavones decomposition The of rates relative far quantities in larger than those of the flavones, was as it part of the mixture; solvent agen reducing acted as quantities ofquantities Al SI 3.8 ldg j / he slopesof k k , lut C.

i vs. , = −slope) = 0.05 —t i was in was which i 1.15): The light - s lut

0.11 –Al acidified (sectionC. SI solution : 7.8 (

0.20 and lut (t (t eak areas areas eak lmg 0.05 stability of lut

= lut 3 was used for comparison of for used theof rates photodecomposition lut was − ); . The were rates relative through. the of photodecomposition calculated the equations of the thebest equations ] of the solutions:1.4vs. ] 0 h) 3+ /under different/under [12] conditions

and 0.10 3+ uartz cuvettes quartz using out carried were yses ] (Table (Table ] water 8:2 (v/v)–water 8:2 solution was observe used to premordant 4.4and 4.5); the wool (Tables 0.05 ldg –Al 3

− through t ldg

took place during the irradiation experiments, methanol experiments, may placetook irradiation during the have . lut vs. ) 3 For a similar purpose, procedure this to that outlined above was , 1.00 fastness of the coloursincreasing decreaseswool with of of dyed fastness -

− t instead of the flavones of the[12] instead t

+ [flavone] HPLC appear to be stable upon overnight heating in presence NO in overnight of heating stableappear upon be to and calibration lines ] being ≤3 mM. It does not mean not does It mM. ≤3 being ]

in in time were used instead of concentrations vs. of concentrations weretime used instead ), 1.4vs. 4.2). matches This the results of the experiments which in

were diluted 5× with 5× diluted were and lut

420 nm range–420 nm the in irradiation reported experiments used 24 UV using peak areas of the 350 nm traces ( the 350nm areaspeak of –UV using 0.11

(t =24 (t ldg ,

lut 0 3

1.0 ( , the − in solution (in solution in presence of NO , originating HNO from 0.10 -

atom of the ion is +5,the is NO ion atom of h) - 2.2).

fit lines (y = ax +b) = ax (y lines fit lmg , or 1.2 ( k comprised was comprised —of each irradiated solution

values were the obtainedslope of from the 1.12 and C. SI

0.05 . lut - lut squares method. squares method. Except for solution for Except = ), 0.2vs. 0.05

k 0.10 aerated methanol aerated . One such condition could be the ). As the integrated form of the zero the of form integrated the As ). , and j was lut

–Al ; methanol; was solution in present [flavone] (mM {[flavone] 0.02 d to bed to 55% 35to slower than

1.0 ( 2.10). ), 3.2vs. that the ion has necessarily has necessarily ion the that , 1 . On day were compared; were 3 , Al(NO ,

ldg in nitric acid

3 , o- 0.05 − water 8:2 (v/v). 8:2 –water lut lmg

order, order, may be reduced in in reduced be may 1.9 ( 3 )

), and 0.6vs. − 0.20 ) decreases with with decreases ) 3 vs.

, or ldg ) time the remainders remainders the 3 , t lut , or from both , or from both

time (min) lut in whichin he [flavone] he [flavone] C. 0.10 , -

however, however, and and acidified 1.13

–Al in in other lmg 0.10

and and 1.4 On On the the

47 in in }. 3 ), ), − ) -

Chapter 4 day 2, t24 solutions were diluted and analysed in the same way. In the case of lut0.20, t0 and t24 4.2.2.2. Experiment using aluminium sulphate and tartaric acid solutions were diluted 10×. Ready-to-dye wool was pretreated with an aqueous solution of a textile auxiliary agent aiming Calibration solutions and samples used for irradiation experiments 1a–special irradiation at protecting the fibre from fibre-to-fibre and fibre-to-metal action. The wool was then experiment were analysed by RP-HPLC–UV undiluted. Analyses were carried out on a Waters mordanted with an aqueous (tap water) solution of aluminium sulphate and tartaric acid at 95 system, equipped with C18 5 µm-particle size column. Further details of the HPLC system and °C during 1.75 h. This procedure was carried out with 0.2, 2, 10 and 25 g L−1 solutions of column used are described in publication by Villela et al. [13] (p. 8545, method using the HPLC aluminium sulphate tetradecahydrate, with the concentration of tartaric acid being 1.3 g L−1 in column). See the referred publication also for information on: Eluent, its flow rate and all cases. Each piece of mordanted wool was dyed with an aqueous (tap water) solution of a composition (solvents and gradient); volume of injection; and temperature of the column oven. commercial extract of weld [2.5% (w/w) extract–wool] at 100 °C during 1.75 h. The dyed wool The photodiode array detector scan range was 245–500 nm. was rinsed at 95°C using detergent and, then, rinsed further with tap water at the same temperature. L*, a*, and b* quantities (CIELAB colour space) were measured for each piece 3+ 4.2.2. Effect of increasing quantities of Al bound to mordanted wool on the photo- of dyed wool, and their chroma (C*ab) and hue angle (hab) were calculated as above. The light- stability of the dye of weld fastness of the colours was determined at TO2C (University College Ghent, Belgium) according The following material/chemical compounds were used as wool, mordants, and source of dye: to the ISO 105–B02 norm. Also in this case, the scale 1 (poor)–8 (excellent) was used for Ready-to-dye wool [Kova Wool Sateen White (part number W110); Whaleys (Bradford), reporting the results. Bradford/United Kingdom]; alum (puriss. p.a. grade; Sigma-Aldrich, Steinheim/Germany); aluminium sulphate tetradecahydrate (ViVoChem, Almelo/The Netherlands); tartaric acid (Brenntag, Dordrecht/The Netherlands); sample of the aerial parts of dried and ground weld {described by Villela et al. [13]}; commercial extract of weld (Rubia Yellow; Rubia Pigmenta Naturalia, later Rubia Natural Colours, Steenbergen/The Netherlands).

4.2.2.1. Experiment using alum Ready-to-dye wool was mordanted with a 10 g L−1 aqueous (deionised water) solution of alum [KAl(SO4)2∙12H2O] through heating to 90+ °C during 0.5–1 h, followed by a 1 h-period at ~95 °C [14]. Essentially the same procedure was also carried out with deionised water only (no alum; blank-mordanting), and with 2 g L−1 aqueous solution of alum. In each case, four of the obtained ~5 × 5 cm-pieces of mordanted wool were dyed with a 96% ethanol–deionised water 3:1 (v/v) extract of the aerial parts of weld in deionised water at 80 °C for 15 min [14]. L*, a* and b* quantities (CIELAB colour space) were measured for each piece of dyed wool, and their 2 2 1/2 chroma [C*ab = (a* + b* ) ] and hue angle [hab = arctan(b*/a*)] were calculated [15]. The data were processed as follows: Standard deviation values were expressed with one significant digit, and average values were rounded up accordingly. The light-fastness of the colours was determined according to the ISO 105–B02 norm. Two pieces of dyed wool were irradiated in a weathering tester in each of the three cases (dyed blank- mordanted wool, and dyed wool that had been mordanted with 2 g L−1 and 10 g L−1 aqueous solutions of alum). The scale 1 (poor)–8 (excellent)—with 3 being the acceptable lower limit— 2 2 2 1/2 was used for reporting the results [16]. Colour difference {ΔE*ab = [(ΔL*) + (Δa*) + (Δb*) ] } values were calculated from the average values of the L*, a*, and b* quantities (CIELAB colour space) measured before and after irradiation [15]. The wash-fastness of the colours of the pieces of dyed wool was determined according to the ISO 105–C06 norm at 40 °C for 30 min. Two pieces of wool were used—with only one of them being washed (thus, n = 1; one pair of samples)—in each of the three cases. Afterwards, both pieces were visually compared and L*, a*, and b* quantities (CIELAB colour space) were measured. The scale 1 (poor)–5 (excellent)—with 3–4 being the acceptable lower limit—was used for reporting the results [16].

48 49

of dyed accordingISO woolwas norm determined the at to 105–C06 40 for °C 30min. alum 4.2.2.1. Experiment using Naturalia, Steenbergen/The Colours, Natural Rubia Netherlands). later * quantities ( a*, and b*quantities samples) used were wool of pieces ) space were values alum; °C Ready et al. Villela by {described (Brenntag, Netherlands); Dordrecht/The sample of the aerial and of ground dried parts weld aluminium sulphate tetradecahydrate(ViVoChem, Almelo/The Netherlands); tartaricacid (excellent) and 3:1 (v/v)extract of aerial the partsweld of deionised in water Bradford/United Kingdom]; Ready and of source dye: wool,mordants, used as compoundsThe were material/chemical following the dyeof weld of stability information E on: publicationalso for referred column). See the 48 of alum) solutions cases three the of each in tester weathering a in irradiated were wool dyed of pieces [ chroma obtained ~5 × 5cm [ increasing4.2.2. Effect of quantities of Al The photodiodearray range was detector scan 245 (s composition publicationby in etcolumn used al. are described Villela 5µm C18 equippedsystem, with experiment weresolutions 10× diluted 2,t day was used for results reporting the digit,were androunded values average upaccordingly. S follows: as processed were data mordanted wool KAl(SO

Calibration solutions andCalibration solutions samples used for1a experiments irradiation The light [14] * quantities (CIELAB colour space) were measured for each piece of dyed wool dyed of piece each for measured were space) colour (CIELAB quantities b* - - bl measur to to . Essentially the same procedure was also carried out with deionised water only water deionised with out carried also was procedure same the Essentially . 24 ank in each of the three cases. Afterwards, both pieces were visually compared piecesboth and three L each were—in of visually the cases.Afterwards, C - - 4

dye dye wool was mordanted awith g 10 L dye wool [Kova Wool Sateen White (part number W110); Whaleys Sateen(part White (Bradford), Wool numberwool [Kova W110); dye ) solutions —with 3 * 2

∙12H were analysed by RP analysed were - - ab calculated from the average values of the L the of values average the from calculated mordanting), and 2gL with fastness of the colours was accordingfastness determined t of the

ed before and after irradiation after and before ed = ( olvents andolvents gradient); volumeof injection; and temperature the columnoven. of , 2 O a* mordanted with 2 g L g mordanted 2 with been had woolthat dyed and . The ] the –4 being the - were dilute 2 pieces ofwoolwerepieces ethanol mordanted dyed a with 96% through heating 90+ to

scale 1(poor) b * CIELAB colour space colour CIELAB 2 . with only one only—with [13] ) 1/2

alum (puriss. p.a. ] and hue[ ] angle d and analysed the in same way. } acceptable results —was the lower reporting used for limit

- - ; commercial extract of weld (Rubia Yellow; Rubia Pigmenta (Rubia weld of extract commercial ; [16]

HPLC tandard deviation values wereexpressed one with significant particle s –8 ( . Colour differenceColour {Δ –UV Analyses undiluted. ona out were carried Waters excellent) − 1 ize ize

aqueous solution of alum.In solution aqueous each case, four of the

[15] °C during°C 0.5–1h,followed a by 1h - of them being washed them of column. Further details of the HPLC system and and system HPLC of the Further details column. 3+ h .

ab − - Sigma grade; The wash The bound to mordanted wool on the photo 1 ) were measured were ) –500 nm. –500 nm.

with 3being—with the acceptable lower limit— = arctan( aqueous *,

[13] a*, and b*quantities ( - E fastness of the colours of the pieces pieces colours of the of the fastness solution of alum alum of solution water) (deionised b*/ *

(p. 8545,methodusing the ab at 80 °C for 15 min for at 80°C o the ISOo the 105– Aldrich, Steinheim/Germany *)] were calculated were a*)]

= [(Δ = In the case of lut of case In the . The (th L luent, its flow rate and −

*) 1 us, n = 1; one pair of of pair one n = 1; us,

and 10 g L 10g and 2

+ (Δ +

–special irradiation sc –deionised a*) ale 1 (poor)ale 1 CIELAB colour B02 norm. Two B02 norm. Two period at ~95 period at ~95

0.20 2 (dyed (dyed

[14] + (Δ + −

, t , and their 1 [15]

aqueous aqueous 0 .

b*) and t blank- L HPLC water water

. The *, [16] Two Two 2 (no (no ] – a* 1/2 *, *, 24 ); ); 5 - .

} reporting the results. all cases.mordanted a with pieceofwoolwas Each dyed to the ISO theto 105–B02 norm fastness was of the colours of dyed wool tempe was ( [2.5% weld of extract commercial L g being 1.3 acid tartaric concentration of the sulphatewith tetradecahydrate,aluminium 4.2.2.2. Experiment using alum 4.2.2.2. Experiment using Ready at protecting the fibrefrom fibre °C °C mordanted an (tap with aqueous water) solution during 1.75 h. This procedure was carried out with 0.2, 2, 10 and 25 gL 0.2,2,10and with 25 procedureduring carried was out 1.75 h.This rinsed . rature - to - dye wool was pretreated with an aqueous solution of a textile auxil textile of a solution aqueous anpretreated with woolwas dye

further rinsedand, further usingat then, 95°C detergent , an , L *, d their chroma ( chroma d their *, and a*, and

* quantities ( b* quantities determined at TO2C(UniversityBelgium) College Ghent, at determined . Also Also C inium sulphate and tartaric acid in this casein this * w/w) extract w/w) ab - to ) and hue angle ( angle hue and ) - fibre and fibre and fibre CIELAB colour space) CIELAB , the scale 1(poor) wool] at 100 °C during 1.75 h. The dyed wool duringwool ath.The 100°C dyed 1.75 –wool] of aluminium sulphate and tartaric acid at 95 h - ab to ) were calculated as above. The light The above. as calculated were ) - The action. metal n solution of a solution water) (tap aqueous

were measured for each piece piece each for measured were with water tap 8 (excellent) was used f used was (excellent) –8

iaryagent aiming wool was then wool was then − 1

at the same same the at solutions of solutions

according according − 1

49 or or in in -

Chapter 4

50 c b a

After the dyed wool had been Mordanting solution a W Al sulphate KAl(SO salt Aluminium Table Table ater hardness classification: DH4 classification: hardness ater 2 (SO 4. 4 ) 1

4 tetradecahydrate) 3 ) . ∙ 2

14H Conditionsmaterial and of themordanting and dyeing procedures of the experiments described in section 4. ∙

Table 4.1. Conditions and material of the mordanting and dyeing procedures12H of the experiments described in section 4.2.2: Temperature, time, aluminium salt, and water. 2 O

2 O (alum) O Mordanting Dyeing Water lso lso (aluminium (aluminium

Aluminium salt contained Rinsing after Rinsing after

T (°C) Time (h) T (°C ) Time (h) Mordanting Dyeing

a rinsed mordanting dyeing . tartaric acid. tartaric

using

KAl(SO4)2∙12H2O (alum) ~95 ~1.4 80 0.2 deionised deionised deionised deionised

T ( Mordanting 95 ~95 detergent

°C

)

Al2(SO4)3∙14H2O (aluminium . 95 1.8 100 1.8 tap water b tap water b tap water b tap water b,c sulphate tetradecahydrate) a Time (h) Time 1.8 ~1.4

Mordanting solution also contained tartaric acid. b Water hardness classification: DH4.

c After the dyed wool had been rinsed using detergent.

T ( Dyeing 100 80

°C

)

Time (h) Time 1.8 0.2

Mordanting Water tap watertap deionised

50 mordanting Rinsing deionised tap watertap

2.2: Temperature, time, aluminiumsalt,water. and after after

Dyeing deionised tap watertap

Villela et al. min 15 for °C 80 at dyed above were A blank - of the pieces of dyed wool Later, these samples these Later, chroma ( chroma L follows: as calculated flavones were the in dyeing baths 2.3. Effect of the4.2.3. Effect glycosylation of pattern of lut the colours of colours the on The following chemicalcompounds and material were used for dyeing the alum

of wool w of wool piece the of cm ldg Genay/France); Extrasynthese, NMR; wool: described by Villela Villela by sample {described ofground thedried aerialweld parts and *, • • • • Single pieces of the wool mordanted with a 10 g L g 10 a with mordanted wool the Single of pieces a*, and b*

dyeing after Rinsing deionised tap watertap Extract of weld: T due low[ to Individual A 96% ethanol 11.5 µmol and ground weld. with those ofwith the blank- Lut without C

* lmg (96%Indofine by lmg NMR; Chemical Hillsborough/USA); Company, ab [14]

) and hue angle ( b,c

quantities (

of lut of - flavones: U (Sup wool alum lmg

water 3:1 (v/v) extract of 420 mg of mg a (v/v) extract of 420 –water 3:1 were analysed by RP analysed were ] in the] sample; removal removal after sampled were baths dyeing The similarly. out carried was , lmg portingInformation, page 2)

- he areas of lut of areas he mordanted wool —while still warm and stirring as CIELAB colour space) colour CIELAB , or ldg , or used sing calibrationsing curves,with that of lmg without

h ab ) were calcul were ) in each of the four cases four the of each in ;

according to procedures adapted from that described by thatdescribed by from adapted procedures to —according - wool , lmg dyedit with - . HPLC

, and ldg (86% by NMR; Extrasynthese, Genay/France); Genay/France); Extrasynthese, NMR; by (86% ated as above. as ated

(aglycone, monoglucoside, and diglucoside) –UV (

were measured for the dyed wool dyed the for measured were —with: peaks of peaks the wool

−

as followed by—followed 5× with dilution DMSO. 1

aqueous solution of alum described alumdescribed of solution aqueous . above)

For this sample of the aerial parts of dried dried of parts aerial the of sample leftover . The percentagesof leftover , having to be extrapolated extrapolated be to having only part of a about 5 - dyeing were compared compared were dyeing et al. et [13] - }. mordanted mordanted , and their , and their

(9 5% by by 5% × 51

50 c b a

After the dyed wool had been Mordanting solution a W Al sulphate Table Table KAl(SO salt Aluminium ater hardness classification: DH4 classification: hardness ater 2 (SO 4. 4 ) 1

4 tetradecahydrate) 3 ) . ∙ 2

14H Conditionsmaterial and of themordanting and dyeing procedures of the experiments described in section 4. ∙ 12H 2 O

2 O (alum) O

lso lso (aluminium (aluminium contained contained

a rinsed

. tartaric acid. tartaric

using

T ( Mordanting 95 ~95 detergent

°C

)

Time (h) Time 1.8 ~1.4

T ( Dyeing 100 80

°C

)

Time (h) Time 1.8 0.2

Mordanting Water tap watertap deionised

mordanting Rinsing deionised tap watertap 2.2: 2.2: Temperature, time, aluminiumsalt,water. and after after

Dyeing deionised tap watertap

follows: as calculated flavones were the in dyeing baths samples these Later, of the pieces of dyed wool A blank- of wool w of wool piece the of cm Villela et al. min 15 for °C 80 at dyed above were of colours the on the4.2.3. Effect glycosylation of pattern of lut ( chroma L Villela by sample {described ofground thedried aerialweld parts and ldg Genay/France); Extrasynthese, NMR; The following chemicalcompounds and material were used for dyeing the alum wool: wool: *, • • • • Single pieces of the wool mordanted with a 10 g L g 10 a with mordanted wool the Single of pieces a*, and b*

dyeing after Rinsing deionised tap watertap due low[ to and ground weld. Extract of weld: T Individual A 96% ethanol 11.5 µmol with those ofwith the blank- Lut without C

* lmg (96%Indofine by lmg NMR; Chemical Hillsborough/USA); Company, ab [14]

) and hue angle ( b,c

quantities (

of lut of - flavones: U (Sup wool alum lmg

water 3:1 (v/v) extract of 420 mg of mg a (v/v) extract of 420 –water 3:1 were analysed by RP analysed were ] in the] sample; was carried out similarly. The dyeing baths were sampled after removal removal after sampled were baths dyeing The similarly. out carried was , lmg portingInformation, page 2)

- he areas of lut of areas he mordanted wool —while still warm and stirring as CIELAB colour space) colour CIELAB , or ldg , or used sing calibrationsing curves,with that of lmg without

h ab ) were calcul were ) in each of the four cases four the of each in ;

according to procedures adapted from that described by thatdescribed by from adapted procedures to —according - wool , lmg dyedit with - . HPLC

, and ldg (86% by NMR; Extrasynthese, Genay/France); Genay/France); Extrasynthese, NMR; by (86% ated as above. as ated

(aglycone, monoglucoside, and diglucoside) –UV (

were measured for the dyed wool dyed the for measured were —with: peaks of peaks the wool

−

as followed by—followed 5× with dilution DMSO. 1

aqueous solution of alum described alumdescribed of solution aqueous . above)

(9 5% by by 5% × 51

Chapter 4 4.3. Theory After photon absorption, lut loses some of the energy gained by planarization of its structure. This section provides theoretical background for the discussion in the chapter. Its first paragraph Although the B-ring is 18° out of the plane of rings A–C in the S0, this decreases to only 1° in consists of a brief overview of photophysical processes often taking place upon irradiation of the S1 [18]. Charge transfer from the B-ring towards rings C and A could be the reason for such organic molecules in solution, based on the book by Gilbert and Baggott [17]. planarization [19]. After that, lut is expected to undergo tautomerism, or excited state Organic molecules may be excited from the electronic ground singlet state (S0) to electronic intramolecular proton transfer (ESIPT). excited states upon irradiation in the portion(s) of the UV–vis range of the electromagnetic The main processes through which lut returns to S0 are nonradiative, as its fluorescence −4 spectrum in which they absorb light. This excitation may be to the lowest vibrational level of quantum yield (ɸf) is <10 in methanol [5]. The reason for this may be a high rate of S1→S0 4 −1 4 −1 the first electronic excited singlet state (S1). If the excitation is to a vibrationally excited level IC. As the S0–S1 energy gap of lut is 2.8 10 cm —i.e., larger than 2.5 10 cm —the efficiency of S1, this excess vibrational energy is quickly dissipated through vibrational relaxation (VR) of the transition could be due to an acceleration by charge transfer from the B-ring, ESIPT, or by collisions of the excited molecule with solvent molecules. This leads to heat generation. both of them [18, 19]. After the IC, VR leads to the lowest vibrational level of S0 [17]. Molecules that are excited to higher electronic excited singlet states may reach the lowest Solvent participates in the proton transfer at the S0 in methanol–water 8:2—the solvent used vibrational level of S1 through the nonradiative transition internal conversion (IC)—e.g., for evaluating the photo-stability of lut in the study reported here—as it is a polar protic solvent S2→S1—followed by VR. Then, the molecules may return to S0 with or without emission of [20]. This may be the case at the S1 too. Polar protic solvents lessen ESIPT [21, 22]. Thus, as light. If this transition is nonradiative, it takes place through S1→S0 IC followed by VR. If it is the proton transfer at the S1 might be mediated by the solvent—and possibly also by component radiative, it is referred to as fluorescence. From S1, the organic molecules may also proceed to partners of lut dimers/aggregates [21]—it is henceforth referred to as excited state proton the first electronic excited triplet state (T1) through intersystem crossing (ISC). Also from the transfer (ESPT). vibrationless level of T1, they may return to S0 with or without emission of light. If the transition is nonradiative, it takes place, again, through ISC—this time T1→S0—followed by VR. If it is radiative, it is referred to as phosphorescence. These transitions can be schematically visualised 4.4. Results and discussion in a Jablonski diagram.2 4.4.1. Effect of different concentrations of Al3+ on the photo-stability of the dye of weld Depending on the structure, flavones and flavonols undergo tautomerism. In the case of the 4.4.1.1. Photo-stability of lut in solution as a function of different concentrations of Al3+ flavone lut, this takes place through proton transfer at the 5-hydroxyl and 4-carbonyl groups The photo-stability of lut in aerated methanol–water 8:2 (v/v) solution upon its irradiation with (pseudo-carboxyl group) (Fig. 4.1). Amat et al. studied photophysical processes taking place light above 300 nm at different t0 concentrations—0.20 mM (lut0.20), 0.11 mM (lut0.11) and 0.05 upon irradiation of lut [18]. In the electronic singlet ground state (S0), the O5-bound tautomer mM (lut0.05)—was compared. Straight lines were fit through the experimental data; the slopes was calculated to be energetically favourable over the O4-bound tautomer. In the first electronic of the equations of the best-fit lines range from −1.8 10−14 to −2.3 10−14 AU. These, and the singlet excited state (S1), it is the opposite. number of molecules decomposed in 24 h of irradiation—evidenced by the change in peak area—are very similar (Fig. SI C.11). This suggests that the rate of photodecomposition of lut is independent of its concentration, following a zero-order rate law under the conditions used OH OH in this study. OH OH In the same set of experiments, the photo-stability of lut in presence of Al3+ at different lut– 3+ B Al ratios—5:1 (lut0.10–Al0.02), 1:1 (lut0.10–Al0.10), and 1:10 (lut0.10–Al1.00)—was studied in HO O HO O duplicate. The experimental data were treated as above, with the photodecomposition of lut A C being described by the zero-order integrated rate law ([lut] = kt + [lut]0). As the rate of 5 4 5 4 photodecomposition is equal to the rate constant (k) in zero-order reactions (d[lut]/dt = k), the O O O O relative rates of the photodecomposition of lut could be obtained. The photo-stability of lut in H H solution decreased with increasing [Al3+], with the relative rates of its photodecomposition O5-bound tautomer O4-bound tautomer being 1.5 (lut0.10–Al0.02), 3.0 (lut0.10–Al0.10), and 4.0 (lut0.10–Al1.00) (Fig. 4.2 and Table 4.2).

Figure 4.1. Tautomerism of lut, through proton transfer at the 5-hydroxyl and 4-carbonyl groups (pseudo-carboxyl group).

2 Singlet (S) and triplet (T) refer to the electronic state of molecules, which relates to the spin of their electrons [17]. There is no need for further elaboration on them for the discussion in this thesis.

52 53

excited state ( state singlet excited (pseudo- upon irradiation of lut flavone be to calculated was S M by collisions of the excited molecule with solvent molecules of ISC through again, place, nonradiative,is takes it vibrationl vibrational states d excite 52 2 Figure ain Jablonski These radiative, phosphorescence. as to itis referred the From fluorescence. as radiative, to referred itis light.If th spect This 4.3. Theory 4- the the molecules organic of consists elect HO Singlet (S) and triplet (T) refer to the electronic state of molecules, which relates to the spin of their 2 carbonyl groupscarbonyl (pseudo- olecules olecules →S S Depending onthe structure, f

may be may molecules Organic first electronic excited singlet state first electronic excited triplet state (T 1 rons section provides theoretical the in chapter. discussion background for the , rum in whichrum in light. theyexcitationabsorb This may be the to lowest vibrational 1 O5-bound this excessthis vibrational energy —followed Then, by the molecules VR. 5 4. O A lut

1. carboxyl group) (Fig. group) carboxyl 4.1). [17]. ess is

H T , this takes placethrough takes this , that often processes often of photophysical a overview brief transiti on is nonradiative, takes it placet level of level automerism

level of T of level C O O

diagram. There is no There need for further elaboration onthem for the discussion thisin thesis upon irradiation in the portion(s) of the UV 4 are excited to higher electronic higher to excited are tautomer in solution, based onthe solution, in

energeti S S OH B

1 1 1 [18]

2 ) , they may return to S to return may , they of of

, itis through the nonradiative carboxyl group). lut . OH cally favou cally In the , through proton transfer, through the opposite

excited from the electronic the ground from state singlet excited lavo electronic electronic

nes nes Amat is is at the 5 the proton transfer at

rable over t over rable quickly ( HO S and flavonols

1 1 . ) thro

). If). excitation the is to a et et bookbyand Ba Gilbert 0 O4-bound singlet al. wit 5 O dissipated dissipated ugh intersystem crossing (ISC) S mayreturn to S studied h or without emission of light. light. of emission h or without —this time 1 excited H , the organic also proceed molecules to may he he

at theat 5 transition internaltransition conversion (IC) transitions hrough S O O ground state O4

undergo 4 tautomer -

photophysical processes t bound tautomer singlet state - through (VR) relaxation vibrational hydroxyl and vis range of the electromagnetic electromagnetic the range of –vis

- OH T hydroxyl and 4- hydroxyl and 1 taking place upon irradiation of of uponirradiation taking place . This leads. This heat to generation →S 1 be can tautomerism →S 0

( with or without emission of emission orwith without 0 S OH ggott [17]

0 0 IC vibrational followed by VR. If it is If is —followed it VR. by ) , the O5 s schemat followed byIf VR. itis

may reach . I n the the of case In the . the the

Its . ( -

carbonyl groups ically visualised S bound tautomer ly . If transition the

0 first paragraph paragraph first first electronic Also Also ) to electronic

excited excited aking place aking place

the the from the from . level of level

—e.g. lowest lowest level , .

t light abovedifferent at 300nm ofboth them of the transition yieldquantum ( planarization S the The- photo Al of concentrations different of Effect 4.4.1. 4. Results and discussion trans lut of partners S the proton transfer the at IC intramolecular (ESIPT). proton transfer After being 1.5( solution equal is thephotodecomposition to rate constant ( zero by the described being duplicate Al number in of molecules decomposed ofof the thebest equations ( mM 4.4.1.1. Photo- [20] for the- photo evaluating Although relative rates of the photodecomposition of lut relative photodecomposition of rates the — area in this study independentis of its concentration, following a zero- . 3+ S processes main The In the same set of experiments, t experiments, of set same the In As the S the As .

olvent

fer (ESPT). 1 ratios

This This lut

photon absorption, photon [18] are 0.05 d decrease . the the verysimilar (Fig. SI C. . The experimental data were treated as above, with the photodecomposition above, of lut as treated were data experimental The m lut —5:1 —5:1 stability participate

) C 0 . —w ay –S [18, 19] [18,

har 0.10 B [19]

stability of lut of stability 1 - be the case at case the be

dimers/aggregates ring is 18° out of the plane of ringsA of of out the plane 18° ring is

by char acceleration due an could to be ge transfer from the B the from ge transfer –Al energygap of lut as compared. Straight were lines fit through compared. as the experimental data f ( ) lut .

is of of with 0.02 After that After 0.10

the the . After s <10 lut

in the proton transfer at the S proton transfer the in ), 3.0 ( ), 3.0 – through

stability of lut

increasing increasing Al lut in aerated methanol aerated in − 1 4 might be

0.02 -

in methanolin

fit lines loses some of the energy gained by planarization of its structure of its byplanarization energygained the loses some of -

in different afunction Al solutionas of of concentrations lut the the order integrated rate law ([ law rate integrated order ), ( 1:1 , 0 IC

which 0.10 lut concentrations

11). S is 2 is

VR l , VR he- photo 1 [21] –Al

[ too. too. mediated by solvent the is expected to expected is range Al .8 10

lut This suggests thatThis suggests -

24 h in the study reported here the study in C and C A rings towards ring 0.10

—it 3+ lut

eads 0.10 [5] P ], ], 4 ), and 4.0( olar protic lessen solvents

cm from −1.8 10 −1.8 from returns S to –Al . with with of irradiation is h is stability to t to The water 8:2 (v/v) solution upon (v/v) 8:2 solution its irradiation–water with −

could be obtained.could 3+ 1 — 0.10 —i.e. excited state proton excited as to referred enceforth he lowest vibrationalhe

the on the photo reason for thi for reason k 0.20 mM ( 0.20 mM ) ) in zero ) in , 0

and and

of of lut relative rates of its of rates relative in methanolin under the conditions usedthe conditions under law rate order undergo , larger than 2.510 C inS –C the 0 lut ge tra ge 0.10

the rate of photodecomposition of lut the rate photodecomposition of − 1:10 1:10 are nonradiative evidenced by the change in peak peak in change the by —evidenced 14

lut –Al in in -

and possibly a —and possibly order reactions (d[ order reactions to −2.3 10 −2.3 to lut nsfer from the B the from nsfer ] =kt ] presence of Al of presence ( 1.00 - tautomerism s may be a high rate of rate high a be may s 0.20 lut the dye of weld weld of dye the of stability as it is a polar itis a—as protic solvent 0 water 8:2 –water (Fig. 4.2 ) (Fig.

, this decreases to only 1° in decreases, this 1°in only to ), 0.11mM ( T 0.10 could be the reason for such - photo he

+ [ ESIPT –Al le − 14 4 vel of vel lut

1.00 cm

, as i as , photodecomposition photodecomposition and the and the These, AU. ] and Table 4.2) Table and , the solvent used —the used solvent

0 ) 3+ − [21, 22][21, lso lso ). was —was stability of lut 1 or excited state state excited or

S - the efficiency efficiency —the

at different lut different at lut lut ring, or ESIPT, ts ts As As 0 by by [17] 0.11 fluorescence fluorescence ]/d ; component the rate of of rate the studied in in studied the sl the ) and 0.05 . t . Thus =

S k 1 ), the →S opes opes 3+ . ,

as as in in

– 0 .

Chapter 4 Table 4.2. Relative rate of photodecomposition of lut in solution as a function of different concentrations of Al3+ and glycosylation pattern (aglycone, monoglucoside, and diglucoside). Flavone decomposed in Rate constant, k Relative rate of Solution a r2 b 24 h of irradiation (%) (10−3 mol L−1 min−1) photodecomposition c 1st irradiation replicate

−6 lut0.11 7 0.98 4.9 10 1

−6 lut0.10–Al0.02 9 0.95 6.4 10 1.3

−5 lut0.10–Al0.10 21 1.00 1.5 10 3.1

−5 lut0.11–Al1.05 24 1.00 1.7 10 3.6

−6 lut0.05 12 0.99 4.4 10 1

3+ −6 Figure 4.2. Effect of increasing [Al ] on the photo-stability of lut in solution. Note: lmg0.05 16 0.99 5.7 10 1.3 3+ Subscripts of code of solutions denote [lut]0 and [Al ], in mM; Solvent: aerated methanol– 2 d −7 water 8:2 (v/v); r = quality of description of the photodecomposition of lut by the zero- ldg0.04 1 0.48 3.5 10 0.1 order integrated rate law ([lut] = kt + [lut]0); Duration of irradiation of solutions = 24.0 h d −6 (1,440 min); Data from 1st irradiation replicate. ldg0.05–Al0.05 8 0.69 2.9 10 0.6

2nd irradiation replicate

−6 lut0.10 6 1.00 4.3 10 1

−6 lut0.10–Al0.02 10 0.98 6.7 10 1.6

−5 lut0.10–Al0.10 22 0.99 1.4 10 3.3

−5 lut0.10–Al0.99 26 1.00 1.8 10 4.0

−6 lut0.05 11 0.97 3.9 10 1

−6 lmg0.05 17 0.99 5.8 10 1.5

d −6 ldg0.05 4 0.62 1.3 10 0.3

d −6 ldg0.05–Al0.05 9 0.46 2.3 10 0.6

a 3+ Subscripts of code of solutions denote [flavone]0 and [Al ], in mM; Solvent: aerated methanol–water 8:2 (v/v); b r2 = quality of description of the photodecomposition of the flavones by the zero-order integrated rate law c ([flavone] = kt + [flavone]0); Relative rates of photodecomposition = kj/ki—as d[flavone]/dt = k in zero-order d reactions—in which i is lut in lut0.11, lut0.10, or lut0.05, and j is lut, lmg, or ldg in other solutions; The photodecomposition of ldg over time—with and without Al3+—is plotted in Figures SI C.12–15.

54 55

order integrated rate law([ Subscripts 54 Figure water 8:2 (v/v); (1,440 min)(1,440 4.2.

of codeof solutions of denote [ ; Data; from 1 Effect of increasing [ r 2

= quality= description of of the photodecomposition of lut st irradiation lut ] = ] =

Al + [ replicate. 3+ lut lut ] on the photoon the ] ] 0 0

); Duration of irradiation of s and [Al and

3+ ], in mM; Solvent: aerated methanol – - stability of lut

in solution.in Note: olutions

by the zero-

= 24.0 h

reactions ([flavone] =kt b a

photodecomposition of ldg

solutions denote [flavone] Subscripts denote of code of solutions lut lut lut lut 1 Solution Al Table ldg ldg lut lut 2 ldg ldg lut lut lmg lut lmg lut r nd st 2

0.05 0.10 0.10 0.10 0.10 0.05 0.11 0.10 0.10 0.11 irradiation = quality of description of the of quality= description of 0.05 0.05 0.05 0.04 irradiation 0.05 0.05

andglycosylation pattern (aglycone,monoglucoside, and diglucoside) – – – – – –

– – 4. d d

Al Al Al Al Al Al

Al Al in —in 2

a 1.05 0.10 0.02 0.99 0.10 0.02 .

0.05 0.05

Relative rate of photodecomposition of lut

d d wh

replicate replicate [flavone] ich ich i 11 6 12 7 8 9 24 h of irradiation (%)

Flavone decomposed in 26 22 10 24 21 9 4 1 17 16

lut lut 0

); ); over time c in in

Relative rates lut

0.11 photodecomposition — ,

with and without Al without and with lut 0.10

of of , or photodecomposition 0.9 1.00 0.99 0.98 0. 0.46

r 1.00 0.99 0.98 1.00 1.00 0.95 0.62 0.48 0.99 0.99 0 2

and [Al and

69 b lut

0.05

in solution as a function of different concentrations of of concentrations different of function a as solution in , and j is j and , 3+ by by flavones the of ], in mM in ], 3+ —is plotted 3.9 4. 4.4 4.9 (

2.9 2.3 Rate constant, Rate 1.8 1. 6.7 1.7 1.5 6.4 10 1.3 3.5 5.8 5.7 10 3 4

−

10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 3 ; lut

mol L mol S − − − − − − − − − − − − − − − − = 5 6 6 6 6 6 6 5 5 6 6 6 7 6 6 5 olvent ,

k lmg j /

in Fig k − i —as d[flavone]/d 1 . , or :

the min

a k erated methanolerated

ures

zero − ldg 1 )

SI SI

- in other solutions other in order integratedorder rate law C. 3. 1.3 1 0.6 0.3 1.5 1 4 3. 1.6 1 0. 0.1 1. 1 3 photodecomposition Relative rate

.0 .6 1 3 6 3 12–

t 15. = – water 8:2 (v/v); (v/v); water 8:2

in zero of of ; - d order order

The

Chapter 4 3+ 3+ 3+ 3:1 An explanation for the decrease of the photo-stability of with increasing [Al ] is suggested. i) without Al ii) lut–Al lut + As discussed in the Theory section, the main processes through which electronic excited state OH lut returns to S0 are nonradiative, possibly through an S1→S0 IC accelerated by charge transfer OH 3 3+ from the B-ring towards rings C and A, ESPT, or by both of them. In lut–Al complexes, the OH HO O OH charge transfer from the B-ring of lut after electronic excitation might diminish. Furthermore, OH OH the possibility of ESPT at the pseudo-carboxyl group decreases with increasing [Al3+], up to HO O HO O 3+ O O the point in which all lut molecules are forming a complex with Al and ESPT becomes Al + impossible. These two possibilities are elaborated upon in the remaining of this section. O O The dihedral angle between the plane of ring B and that of rings A–C decreases upon OH O OH O complexation with Al3+ at the pseudo-carboxyl group [4, 23]. This planarization is linked to O OH increased electronic delocalization [4, 23], which is consistent with the lowest UV–vis absorption bands of lut–Al3+ complexes being red-shifted relative to that of lut [Fig. SI C.7, HO OH and [5] (Fig. 7)]. Thus, the charge transfer from the B-ring of lut in lut–Al3+ complexes after electronic excitation might be less pronounced than that of free lut. 3+ 1:2 3+ 1:10 Complexation at the pseudo-carboxyl group is favoured over that at the 3ʹ- and 4ʹ-hydroxyl iii) lut–Al iv) lut–Al groups (catechol group) in a neutral environment [4]. In an acidic environment—methanol– 3+ 2+ 3+ water 9:1 acidified to an apparent pH of 2.5—flavonol–Al complexation at the catechol group OH 2+ Al OH OH O did not even take place [4]. The environment in which the experiments reported in this section OH O were carried out was also acidic (methanol–water 8:2 at an apparent pH of 3.6). Consequently, HO O 3+ HO O complexation between lut and Al is expected to have taken place primarily—if not solely— HO O + at lut’s pseudo-carboxyl group. Therefore, the decrease of the photo-stability of lut with 3+ O O increasing [Al ] is also consistent with a corresponding decreasing possibility of ESPT at the Al O O O O Al pseudo-carboxyl group [6, 7]. Al This can be understood by knowing which complexes are formed at different lut–Al3+ ratios.4 Amat et al. studied it theoretically, with experimental data in methanol obtained by Figure 4.3. Lut–Al3+ complexes formed at different lut–Al3+ ratios in methanol [23]. Favaro et al. as benchmark [5, 23]. The following—also represented in Figure 4.3—is based on their report: • At a lut–Al3+ 3:1 ratio, lut is both free and as the lut–Al3+ 2:1 complex involving the Water in the solvent seems to be unfavourable for lut–Al3+complexation. Because of this, the pseudo-carboxyl group; main species in methanol–water 8:2 solution at a lut–Al3+ 1:10 ratio is expected to be the same • At a lut–Al3+ 1:2 ratio, lut is largely present in the form of the lut–Al3+ 1:1 complex, also as that in methanol at a lut–Al3+ 1:2 ratio {Table SI C.1 and ensuing discussion, [5] (Fig. 7), involving the pseudo-carboxyl group; and [23]}. Therefore, the decrease of the photo-stability of lut with increasing [Al3+] reported • At a lut–Al3+ 1:10 ratio, a mixture of the complexes lut–Al3+ 1:2 involving the pseudo- here is consistent with ESPT being: carboxyl group and the catechol group, and lut–Al3+ 1:1 (as in the previous bullet point) • Possible at a lut–Al3+ 5:1 ratio, due to the main species in solution being the lut–Al3+ 2:1 is likely. complex and free lut, as represented in situation ii of Figure 4.3; • Still possible at a lut–Al3+ 1:1 ratio—but to a much smaller extent than in the previous case—due to the main species in solution being the same as in the previous bullet point, but with an increased number of lut molecules forming a complex with Al3+; • Impossible at a lut–Al3+ 1:10 ratio, due to the main species in solution being the lut–Al3+ 1:1 complex represented in situation iii of Figure 4.3.

The reported fluorescence in a situation in which all lut molecules are forming a complex with 3 Deprotonation at the 7-OH group of lut could also play a role in acceleration of the S1→S0 IC. This is 3+ Al evidences a process for the S1→S0 transition other than accelerated IC. Fluorescence is further elaborated upon in section 4.4.2.1. nearly the only process through which the lut–Al3+ complex(es) release(s) absorbed energy, as 4 The rationale presented for the decreasing possibility of ESPT at the pseudo-carboxyl group with increasing [Al3+] is qualitative on the possible lut–Al3+ complexes present in solution, not involving the ɸf ≈ 1 in methanol [5]. Thus, there is a change of the preferred process through which lut 3+ formation constants of the complexes. releases absorbed energy at high [Al ]. As charge transfer from the B-ring of lut after

56 57

formation constants of the complexes.constants the formation of further elaborated upon in section 4.4.2.1. increasing [Al absorption bands increased electronic delocalization electronic increased pseudo- which in the all point Favaro [Al increasing at groups group) (catechol lut As discussed the in Theory 56 4 3 ratios lut between complexation w out carried were even not did take place 2.5—flavonol waterapparent pH of an to acidified 9:1 excitation electronic and complexation with Al impossible B the from transfer charge B the from A the possibility at the pseudo ESPT - of th Deprotonation at the 7- The rationale pres rationale The n eir report eir lut Th • • • Complexation at the pseudo- Complexation The returns S to explanation explanation

[5] . is is ’s ’s is likely pseudo- carboxyl group and the At At At involving 4

Amat

et al. dihedral angle between the plane of ring B and that of rings A rings of that and B ring of plane the between angle dihedral carboxyl groupcarboxyl 7] [6, (

can be understood bycan understood knowing be pseudo- Fig. 7) a a a lut - . These two possibilities are. These section. twopossibilities elaborated thethis upon in remaining of lut : lut ring towards rings C and A, ESPT, or or A, ESPT, andrings C ring towards

as b as 3+ –Al carboxyl group;carboxyl . –Al –Al et al. 3

] ] is qualitative onthe possible lut 0 + the pseudo- the ] th for ] . are possibly nonradiative, carboxyl groupcarboxyl . enchmark 3+ 3+ is e charge transfer from the B the from transfer e charge Thus, th 3+ of of ented for the the for ented as also acidic ( acidic also as

1:2 ratio,1:2 lut

also also 1:10 ratio might lut lesspronounced thatof be free than studied studied 3:1 ratio lut e 3+

OH groupOH of decrease of the photo- the of decrease

lut [4] –Al

ESPT a possibilityconsistent with ESPT corresponding decreasing of at the pseudo- the at in in

and Al [5, 23] [5, ring ring 3+ molecules molecules carboxyl group;

The environment in which thereported experiments in this section a section it , .

catechol

neutral environment complexes complexes lut theoretically a decreasing possibility of ESPT at the pseudo- carboxyl group is favoured over that at the 3ʹ thatat the over group favoured carboxyl is

of of

is mixture of the complexes lut

methanol

3+ . T Therefore, the decrease of the photo the of decrease the Therefore,

is is largely largely lut lut

state excited electronic which through processes main the ,

is expected to have taken place primarily place taken have to expected is he following both both [4, 23] [4, couldplay also role acceleration in a of the

after after group, are are carboxyl groupcarboxyl increasing[Al decreases with carboxyl group being- red which free present present through through –water 8:2 at–water 8:2 an apparent pH of 3.6) form ,

electronic excitation might diminish. Furthermore, stability lut of , with –Al and

which consistent is and ingcomplex a with Al

3+ also represented i represented —also complexes by by

lut in the form thein experimental

complexes a as as [4] shifted relative to th n both –Al –Al S the the . - 1 [4, 23] [4, ring ring In an environment acidic →S 3+

3+ of them of l

1:1 1:1

with increas with complexation

0 of of

present solution,in not involving the

are formed at at formed are

–Al of the lut the of IC . This planarization is linked –Al lut (as (as

data data accelerated by charge transfer transfer by charge accelerated 3+ .

3+ 3 in in .

in in

2:1 2:1 In In

with lowest the UV 1:2 1:2 n Figure the previous point bullet lut in in lut –Al 3+ ing ing complex involving the methanol obtained by –Al involving at -

– stability of lut at the catechol group and ESPT becomes becomes ESPT and

3+ Al C decreases upon upon decreases –C of carboxyl groupcarboxyl with [Al 3+ different 1:1 1:1 - 3+

—

is based on on based 4.3—is

lut S complexes complexes and 4ʹ and 3+ . Consequently

1 complexes, thecomplexes, if not solely if not →S ] complex

—methanol [ is is Fig the the 0 suggested IC. This is is This IC. - 3+ hydroxyl hydroxyl . SI C. SI . lut pseudo- ] ,

at at –Al up to up to , after after with with –vis –vis also the the — to to 7, 3+ – ) , .

Al a methanol in at that as T and ɸ only process the nearly is consistenthere with main W Figure s release f he he

ater the in solvent s 3+ 1 ≈ • • • HO HO iii i ) [23] reported fluorescence reported ) without rocess aprocess evidences s Still Still of Figure represented iiiof complex situation 1:1 in I lut free and complex an increased number of number increased an with but being—due solution thein to mainspecies case the same point, bullet as the in previous P lut pecies in methanol in pecies mpossibleat a lut 4.3. Lut ossible at a at ossible in methanolin

–Al }. absorbed energy absorbed OH O ossible at alut at possible 3+ Therefore, Therefore, Al Al

3+ 1:2 O O O O –Al 3+ lut

complexes lut formed different at OH [5] OH –Al the decrease o decrease the eems to be to unfavoureems ESPT being ESPT through lut which the . –Al lut change of the preferred process through which lut process preferred the of Thus, there a is change OH OH water 8:2 –water 3+

for for

all all which in ain situation of Figure 4.3; Figure represented iiof , assituation in –Al at [Al high

–Al 5:1 ratio, due to the mainspecies being solution in the lut 3+ the the

1:10 ratio,1:10 due to 2+ 3+ 3+

1:1 ratio S Table SIC. ratio1:2 {Table : 1

ii iv HO HO

→S solution solution f the- photo ) ) lut lut lut 3+ 0 –Al –Al

] transition O OH molecules formingcomplex a with Al molecules . a to —but much 3+ HO able for able 3+ As As Al

3:1 at alut at 1:10 –Al O O the B the from transfer charge the being species the solution in main the stability lut of O O O 3+

Al ( release complex(es) –Al lut accelerated other than accelerated –Al lut O O O O –Al 3+

4.3. , ensuing1 and discussion 3+ molecules are forminga complex with ratios Al O OH

3+ 1:10 ratio is smaller extent than the in previous complexation OH

3+ in methanolin

with increasing [Al + OH HO + +

expected to expected s HO

) . O [23]

absorbed energy Because of this of Because IC Al - ring of lut of ring . . O O

OH F luorescence luorescence 3+ ;

[5] be be

3+ O O ] OH –Al

the same same the lut reported (Fig. 7)

–Al 3+ OH after after OH , t

, as as , 2:1 2:1 57 he he 3+ is ,

OH 2+

Chapter 4 electronic excitation may diminish and the possibility of ESPT decreases, the possible accelerated S1→S0 transition via IC could be progressively replaced by an S1→S0 transition via fluorescence. The rate constant of accelerated S1→S0 transition via IC (kIC) of lut in methanol–water 8:2 might be of the order of magnitude of 1011 s−1, by analogy to that of the O-3-glycosylated flavonol rutin in methanol, also a polar protic solvent [19]. Fluorescence lifetimes are typically −9 of the order of magnitude of 10 s and the measured fluorescence lifetime (τf) is the reciprocal −1 of the rate constant of fluorescence (kf) if ɸf = 1, i.e., τf = kf [17] (chapter 4). Thus, the electronic excited state lifetime of free lut might be about two orders of magnitude shorter than that of lut molecules engaged in complex formation with Al3+. Since molecules are more reactive in the electronic excited state than in the electronic ground state, this would account for the decrease of the photo-stability of the compound with increasing [Al3+].

4.4.1.2. Effect of mordanting wool with different quantities of Al3+ before its dyeing with extracts of weld on the colour and on its light-fastness Two experiments using extracts of weld to dye wool premordanted with increasing quantities 3+ of Al were carried out. In one of them wool was mordanted using alum [KAl(SO4)2∙12H2O], and in the other, aluminium sulphate and tartaric acid. L*, a*, and b* quantities (CIELAB colour space) were measured for the pieces of dyed wool, and their light-fastness was determined. As CIELAB chroma (C*ab) and hue angle (hab) approximately correlate to chroma and hue [15], the C*ab quantities are henceforth referred to as the chroma of the colours, and the hab quantities, as their hue. Figure 4.4. Weld-dyed wool premordanted with varying quantities The dyed wool becomes more colourful {saturated [24]}—displays higher chroma—with of Al3+ (experiment using alum). Pieces of dyed wool in columns, increasing [Al3+], with a plateau seeming to be reached at some point (Fig. 4.4 and Table 4.3). from right to left (n = 4, each case): Wool that had been This takes place at nearly constant lightness (L*) and hue (Table 4.3 and Table SI C.4). It is premordanted with 10 g L−1 and 2 g L−1 aqueous solutions of alum, also worth noting that the colour of the dyed blank-mordanted wool—in which the wool had and blank-mordanted wool (wool that had been pretreated as in the been pretreated as in the other cases, except that the mordanting bath had no alum—was pale, other cases, except that the mordanting bath had no alum). with a chroma value of only 15 (Fig. 4.4 and Table SI C.4). To some extent, the results of colourfulness contrast with those of an earlier report by Vinod et al. (2010), as the chroma of alum-premordanted silk—also a proteinaceous fibre—dyed with a flavonoid dye remained Table 4.3. Quantitation of the colours of weld-dyed wool samples premordanted with varying quantities of Al3+ 3+ constant with increasing [Al ] [25]. The K/S value increased, however, suggesting increased (experiment using aluminium sulphate and tartaric acid): Measured (L*, a* and b*) and calculated (C*ab and hab) dye uptake [26]. Thus, the increase in colourfulness of the dyed wool with increasing [Al3+] quantities using the CIELAB colour space (each case, n = 1). reported here could be—at least in part—due to larger quantities of dye in the fibre. [Aluminium sulphate Chroma Hue angle tetradecahydrate] in L* a* b* (C*ab) (hab; degrees) mordanting solution (g L−1) a

0.2 74.4 −5.6 31.4 31.9 100.1

2 76.8 –9.3 46.4 47.3 101.3

10 78.7 –11.1 53.9 55.0 101.6

25 78.0 –10.6 52.7 53.7 101.3

a [Tartaric acid] = 1.3 g L−1, all cases.

58 59

constant with increasing[Al alum colourfulness with been also ( lightness constant nearly at place takes This h the that of of flavonol methanol, in rutin and hue Al of Al of differentwith wool quantities mordanting 4.4.1.2. Effect of fluorescence. 58 here reported dye uptake increasing determined. woo dyed of pieces the for measured were space) colour alum and the in other, e Two the its on on lightcolourextracts weld and of - decrease of the the photo for gro electronic the in than state excited electronic the in reactive electronic excited state lifetime of of the order of 10 of magnitude might may excitation electronic accelerated accelerated The The T the rate rate the rate he the worth noting thatthe ab - 3+ a the mordanting bath the that except cases, other the in as pretreated proteinaceous fibre a proteinaceous —also premordanted silk extracts of weld dye to of pre wool usingxperiments extracts

quantities, as quantities, their hue. be of the order of magnitude of 10

chroma chroma d ou carried were lut

[15] yed more woolbecomes colourful

[ Al S constant CIELAB chroma ( chroma CIELAB As molecules molecules [26] constant , t 1

could be →S contrast with those of an earlier report by Vinod report earlier an of those with contrast 3+ he he value value (Fig. a with ], plateau (Fig. at reached some point seeming be to . Thus, 0 C

transition v * of accelerated S accelerated of ab fluorescence of of —

complex engaged complex in inium sulphateinium and tartaric acid. L t. Int. mordante of themwoolwas one quantities are henceforth are quantities the increase of colourfulness the in dyed increasing woolwith [Al

at least in part in least at only

colour also a polar protic solvent diminish 3+ stability the of co 15 ( ] ia ia −

[25] 9 IC

s and the meas of- the dyed blank free free C Fig. Fig.

c . *

ould ould 1 ab The K/S value increased, value K/S The ( →S and k lut ) and hue angle ( the fibre the in —due larger of to quantities dye 4.4 f )

0 be progressively replaced by a by replaced progressively be if if might

transition 11 the possibility decreases, of ESPT the possible and

ɸ saturated {saturated

formation s f L

mpound with mpound − = 1 1 *) and hue*) (Table 4.3 Table Table - ured , be be fastness referred as to the chroma of the colours, and by by , a bout bout mordanted wool

i.e.

fluorescence fluorescence via via [19] the thatof the to analogy dyed with a flavonoid dye remained dyea remained flavonoid —dyed with SI SI h , [24] with ab

increasing quantities quantities mordanted increasing with IC τ C. t .

wo f ) approximately correlate to chroma chroma to correlate approximately ) F *,

4) = increasing [Al increasing }—displays higher chroma d using alum d using ( luorescence luorescence k

Al * quantitiesa*, and b* ( . orders of magnitude of orders IC k l

, and the f To some extent, t To some extent, however, suggestinghowever, increase et al. ) 3+ − und state, this w this und state, 1 of of .

lifetime [17] Since Since lut 3+

the which—in the (2010) had

SI SI Table and

before in in (chapter Thus,the 4). n lifetime ir light ir

[ S molecules are more more are molecules no methanol KAl(SO (τ 3+ 1 4.4 →S , as the chroma of f ] ) O . alum

is its dy its - and 3- 0

s - the the

fastness transition via

ould accountould of of results he glycosylated glycosylated

are , pale —was shorter than 4 . Table Table

water 8:2 8:2 –water ) reciprocal 2 e C. wool ∙12H CIELAB CIELAB typically ing with 4). —

4.3 with with 2 It is was was had O 3+ ) ] d ] . ,

and and premordanted with 10 g L from right a Al of Figure other ca

[ 2 0.2 ( solution mordanting tetradecahydrate in ] sulphate [Aluminium quantitiesusingcolour space (eachCIELAB case, n the =1). (experiment using aluminiumsulphate and tartaric Measured acid): ( Table 10 25 T

artaric acid

- blank 3+

4. (experiment using alum). Pieces of dyed wool in columns, 4.4. ses, except that the except that ses, 3 .

mordanted woolmordanted ( Quantitation of the colours of weld of colours the of Quantitation Weld to left ]

= 1.3 gL 1.3 - dyed woolpremordanted with varying quantities

(n = 4, each case): g L − − 1 , all cases. all , 1 ) − 1 a

mordanting bath had

and 2gand L wool 76.8 74.4 78.7 78.0 L *

that had been

− 1

aqueous alum solutionsof

- dyedwool samples premordantedwith varying quantities ofAl W – − – – a ool ool 9.3 11.1 10.6 * 5.6

pretreated as in the

no alum

that had been

) .

46.4 31.4 53.9 52.7 b *

*, *, , a * and b * and *) and calculated (C and *) 47.3 31.9 55.0 53.7 ( Chroma C * ab

)

( Hue angle 101.3 100.1 101.6 101.3 h ab * ; degrees) ab

and and h 59 ab 3+

)

Chapter 4 The light-fastness of the colours of the wool in both experiments was below 3, the acceptable Table 4.5. Light-fastness of the colours of weld-dyed wool premordanted with varying quantities of Al3+ lower limit using the scale 1 (poor)–8 (excellent) [16]. However, that of all samples decreased (experiment using aluminium sulphate and tartaric acid; samples of Table 4.3; each case, n = 1). 3+ with increasing [Al ]. In the experiment in which alum was used, the colour difference values [Aluminium sulphate tetradecahydrate] in mordanting Light-fastness of the colourb calculated from the L*, a*, and b* quantities before and after irradiation increased at higher solution (g L−1)a [Al3+] (Table 4.4). In the experiment in which aluminium sulphate and tartaric acid were used, the light-fastness of the colours of the samples themselves decreased with increasing [Al3+] 0.2 1–2 (Table 4.5). This contrasts with earlier reports in which the colours of Al3+-mordanted proteinaceous fibres dyed with flavonoid dyes have been reported as more or equally light-fast 2 <1 / 1 as those of non-mordanted fibres [16, 25, 27]. The wash-fastness of the colours of the samples of the experiment in which alum was used 10 <1 was also determined. On the scale 1 (poor)–5 (excellent), wash-fastness values of 3–4 are the acceptable lower limit [16]. The wash-fastness of all colours either matched or surpassed this 25 <1 limit (Table SI C.5). Although replicate samples would be needed for increased accuracy of the a [Tartaric acid] = 1.3 g L−1, all cases. description of the phenomenon, the colour of dyed blank-mordanted wool (no alum) appears to b Scale: 1 (poor)–8 (excellent) [16]. be more wash-fast than that of dyed Al3+-mordanted wool. This would contrast with earlier reports too. The colours of Al3+-mordanted proteinaceous fibres dyed with flavonoid dyes have been reported as more wash- or water-fast than those of non-mordanted fibres [16, 25, 27]. The two experiments in which wool was mordanted with different quantities of Al3+ before its dyeing differed in a number of ways. Several of these differences are displayed in Table 4.1. The short heating time of the dyeing process of the alum-experiment is likely to have been Table 4.4. Light-fastness of the colours of weld-dyed wool premordanted with varying quantities of Al3+ positive for the wool in terms of its physical properties. This is due to the loss of soluble proteins (experiment using alum; samples of Figure 4.4), and CIELAB colour difference (ΔE*ab) values calculated from the wool upon high-temperature heating, which increases with the duration of the process from the L*, a*, and b* quantities before and after irradiation. [28]. Nevertheless, long, high-temperature heating is required for the dye molecules to arrive −1 a b [Alum] in mordanting solution (g L ) Light-fastness of the colour ΔE*ab at the regions within the wool fibres where interaction of the dyes with the proteins is thermodynamically favoured [29]. Thus, the long, high-temperature heating of the dyeing Blank (no alum) <1 / 1 4.78 process of the aluminium sulphate/tartaric acid-experiment is expected to have been positive in terms of the quality of the dyeing. A potential negative effect of the short, lower-temperature 2 <1 4.90 heating used in the alum-experiment would be low wash-fastness of the colour, as the dye molecules—not at the regions of thermodynamically favoured interactions yet—could have 10 <1 6.25 been extracted from the fibre with relative ease [29]. As mentioned above, however, the wash- a Each case, n = 2; scale: 1 (poor)–8 (excellent) [16]. fastness of the obtained colours either matched or surpassed the acceptable lower limit. b Calculated from average values; in each case, n = 4 (before irradiation) and n = 2 (after irradiation). In spite of these and other differences between the two experiments, the light-fastness of the 3+ colours of weld-dyed wool decreased with increasing [Al ] in both cases. As mentioned earlier in this section, more dye may have been taken up by wool mordanted with larger quantities of 3+ Al . Light-fastness of the colour generally increases with the increase of the quantity of dye in 3+ the fibre [30, 31]. If this is the case also for the system dye of weld–Al –wool, the detrimental 3+ effect of Al on the light-fastness of the colours is even larger than that seen in Tables 4.4 and 4.5. A similar explanation to that suggested in the previous section for the decrease of the photo- 3+ stability of lut with increasing [Al ] is given here. That is, the decrease of the photo-stability 3+ of the flavones—main components of the dye of weld {[14] (p.7 of SI)}—with increasing [Al ] may be due to a change of their preferred process for releasing absorbed energy. The S1→S0 transition could increasingly proceed via fluorescence rather than via IC accelerated by a photophysical process(es). This could lead to longer electronic excited state lifetimes of the flavones engaged in complex formation with Al3+ than those of the free flavones, accounting for the decrease of the photo-stability of the compounds. However, the picture here differs from

60 61

as those of non- proteinaceous flavonoid with dyes dyed fibres been have reportedor as equally more light (Table [Al wash more as reported been acceptable lower limit calculated lower limitusing the 1(poor) scale 60 b a reports The too. Al colours of limit was the lightwith increasing [Al The light be more wash more be blank - thedyed colourdescription of of the phenomenon,

Each case, n s =2; 2 10 Blank [A the from ( Table Cal samples of Figure 4. Figure of experiment using samples alum;

The

lum 3+ culated fromculated average values; each n in irradiation) (before and case, =4 n (after =2 irradiation). also also ] (Table (Table (Table ] (no alum 4.

wash in in ). 4.5). 4 L - . determined decreased themselves samples the of colours the of fastness

mordanting solution *, fastness of the the of fastness Light from the L the from a SI SI - *, and and *, fastness This contrast This 4.4) ) -

- C. fast than thatof dyedfast Al fastness of colours the weld of mordanted fibres mordanted cale: (poor 1 5) . b In the experiment in which aluminium sulphate and tartaric acid were used, were acid tartaric and sulphate aluminium which in In experiment the 3+ . * quantities * quantities . Although replicate samples would be needed for increased accuracy of increased for be the needed would replicate Although samples , the colour difference values values difference In]. colour which the in experiment alumwas the used,

*, On On of the samples of the experiment the of samples the colours of the of [16] * quantities before andaftera*, and before b*quantities irradiation colours of the woolcolours of the the

) ( . s with earlier reports in whichof earlier in with Als the reports colours – - g L T or water or

8 (excellent) [16] 3+ scale 1(poor) before andbefore after irradiation he wash he − - 1 mordanted proteinaceous dye fibres ) [16, 25,27] [16,

–8 - - (excellent) fast than those offast non- than those fastness of 4 3+ ) , and, CIELAB colour difference ( - - dyed wool . –5 mordanted wool.This

<1 /1 <1 Light <1 <1

. in

(

excellent acceptable 3,the was acceptable experiments both below -

fastness of the colour the of fastness [16] all .

premordanted with varying quantities of of quantities varying with premordanted

colours . However, ) mordanted wool (no alum) appears to wool (nomordanted alum) appears to , wash , mordanted fibres mordanted either matched or surpassed or t matched either - fastness th would contr a samples decreased atall of samples

d with flavonoid dyes flavonoid d with have

in which alum was used alumwas used which in Δ

with increasing [Al increasing with E values of values *

ab increased )

4.78 Δ 4.90 6.25 values values [16, 25,27] [16, E ast * ab

with earlier - 3–4 calculated calculated mordanted mordanted

at higher at higher are the the are Al - . 3+

fast fast his 3+

]

Al larger of quantities with wool mordanted by up taken may been have dye section, more this in - photo for of the the decrease The which in twoexperiments woolwasdifferent of with quantities mordanted Al b a the fibre the of the flavones for process preferred their of achange to due may be stability of lut Al of effect photophysical process( couldtransition increasingly proceedrather via fluorescence than via 4.5. the of terms process ofaluminium the sulphate/tartaric acid favoured thermodynamically at thewhere regionsthe woolfibres within [28] from the wooluponhigh of proteins soluble the loss due is to This properties. physical of its terms woolin forpositive the alum the of process dyeing the of time heating short The dy flavones engaged in complex formation Al with complex engaged flavones in heating the in alum used colour surpassed eitherfastnesscolours the ofmatched acceptable the obtained or lower limit. [29] ease relative with fibre the from extracted been molecules

[Tartaric acid] = acid] [Tartaric 1.3 g L 2 0.2 solution [A ( Table 25 10 Scale: (poor) 1 e

e 3+ xperiment using alum A similar explanation In spite of the

luminium sulphate tetradecahydrate sulphate luminium ing differed in a number of ways. S ways. of a number differed in ing

. . L Nevertheless s 4. ight

of of

( 5 [30, 31] [30, g L .

not at—not the regions of weld - Light 3+ fastness of the colour colour the of fastness − 1 quality of the dyeing.

) – on the light a

main components of the dye—main components of weld

8 (excellent)8 [16] with increasing [Al - - fastness of thefastness ofweld colours of dyed wool wool dyed se and other se . I , f long, highlong, - this is the inium sulphate and tartaric acid tartaric and sulphate inium − 1 , all cases. all , es - decrease of the photo the of that decrease to the section suggested for the in previous - ) temperature heating, w heating, temperature - fastness . experiment experiment d decrease This could lead to longer electronicexcited state lifetimes of the

stability of the compounds stability of the . differences between the two experiments, t experiments, two the between differences

case [29] temperature heating is requiredfor molecules the dye to arrive

] generally generally

in mordanting mordanting in 3+ of the colour the of . thermodynamically favouredthermodynamically interactions also for the system dye of weld dye of system the for also A potentialnegative of the effect short,lower ] Thus

with increasing [Al is w

everal everal given ould beould , thehigh long, increases with the increase of the increase the with increases - dyed wool here - are are differences these of

; s experiment is exp is experiment interaction samples of of samples

hich 3+ is even larger even is low wash low

. than those of the fre of the than those Light <1 /1 <1 1 <1 <1 . As mentionedhowever, above, wash- the decrease of the photo the of That the decrease is, –

{ increases w increases 2

[14]

premordanted with varying quantities of of quantities varying with premordanted releasing absorbed energy absorbed releasing . However, the picture here differs from from . However, here the differs picture -

fastness of the colour the of fastness 3+ Table Table - ]

temperature heating of the dyeing dyeing the of heating temperature

(p.7 of SI) - -

in both cases both in fastness of the colour, as the dye dye the as colour, the of fastness experiment is likely to have been been have to likely is experiment is the proteins is with dyes the of

than th 4. 3 ith the ected have to in positive been ; each case, –Al }— at 3+

displayed displayed

e flavones, accounting accounting flavones, e duration of he light –wool seen . with increasingwith [Al As mentionedearlier IC b

n = 1 n =

quantity of dye in in dye of quantity and 4.4and Tables in ted by a by ted accelera yet , - the detrimental ). fastness of the the fastness of —c

. The S in Table 4.1. Table in - temperature temperature 3+

the process the ould have have ould

before before - stability 1 →S Al 3+ 61 its its 3+ - 0 ]

Chapter 4 that of lut in solution due to additional geometrical constraints and changes in electronic distribution. These relate to the chemical structures of the flavones themselves—which differ from that of lut, with most of them containing sugar moieties—and to their interaction with the proteins of wool. Van der Waals forces are one of the important ways through which dye molecules interact with the proteins of wool [29]. Naturally, interactions such as this influence the electronic distribution of the flavones.

4.4.2. Effect of glycosylation pattern of lut (aglycone, monoglucoside, and diglucoside) on its photo-stability in solution and colours of alum-mordanted wool dyed with it 4.4.2.1. Photo-stability of lut in solution as a function of its glycosylation pattern The photo-stability of lut in solution was studied in duplicate as a function of its glycosylation pattern: Aglycone, lut-7-O-glucoside (lut monoglucoside, lmg) and lut-7,3ʹ-O-diglucoside (ldg). This was done in the same set of the experiments discussed in section 4.4.1.1, in aerated methanol–water 8:2 (v/v) solution with light above 300 nm. Also the experimental data were treated as done for lut in solution in the presence of Al3+, with the relative rates of photodecomposition being obtained. Whereas lmg photodecomposed 1.4× faster than lut, the photodecomposition of ldg was much slower (Table 4.2 and Fig. 4.5). Although this increased Figure 4.5. Effect of glycosylation pattern—aglycone, monoglucoside, and diglucoside— reactivity of lmg upon irradiation relative to lut was also observed by Smith et al. [6], it seems on the photo-stability of lut in solution. Note: Subscripts of code of solutions denote counter-intuitive at first, due to the possibility of glycosylation to confer stability to flavonoid 2 [flavone]0, in mM; Solvent: aerated methanol–water 8:2 (v/v); r = quality of description aglycones [8], [9] (chapter 42), [10] (chapter 5). Possible explanations for the effect of the of the photodecomposition of the flavones by the zero-order integrated rate law ([flavone] glycosylation pattern on the photo-stability of lut in solution are elaborated upon in this section. nd = kt + [flavone]0); Duration of irradiation of solutions = 24.0 h (1,440 min); Data from 2

irradiation replicate.

The possibility of lut to increase the electronic density of the A-ring through deprotonation at

the 7-OH group may account for the difference between its rate of photodecomposition and that

of lmg. The electronic distributions of lut and apigenin—3′-deoxy lut—are similar [18]. The

S0 → S1 transition of apigenin is accompanied by a shift of electron density from ring A to ring

C [18]. Upon irradiation with light above 300 nm, lut mainly undergoes an S0 → S1 transition

[18] (Table 4.2 and Fig. 4.3). Thus, in the experiments reported here, lut is expected to have

undergone a change in electronic distribution equivalent to that of apigenin; importantly for this

discussion, a lowering of the electron density of ring A.

With a pKa < 9, the 7-OH is one of the two most acidic groups of lut [5, 32]. The tendency

for deprotonation at this hydroxyl group should only increase in the S1 because of the change

in electronic distribution [22]. Thus, such a deprotonation could be a way for excited state lut

to compensate part of the loss of electron density of ring A. This is impossible in the case of

lmg because it has an O-glucose moiety instead of a hydroxyl group at C-7. Such an

impossibility could lead to a change in the preferred process through which absorbed energy is

released, causing the electronic excited state lifetime of lmg to be longer than that of lut and,

in this way, account for the reduced photo-stability of lmg.

Stabilization of the transition state of the formation of the flavonoxyl radical of lut by

intramolecular H-bond could explain the difference between the rates of photodecomposition of lut and ldg. The flavonoxyl radical of quercetin—a lut-like flavonoid having an additional hydroxyl group at C-3—is formed through high-energy irradiation (radiolysis) [33]. Flavonoxyl

62 63

62 beingphotodecomposition obtained. lut for done as treated methanol ( lut of that from distribution. These relate to the chemical s that of lut the electronic distribution ofthe electronicflavones distribution molecules interact with proteins of wool glycosylation pattern aglycones ldg of photodecomposition The its glycosylation pattern of of 4.4.2. Effect counter lmg of reactivity pattern 4.4.2.1. Photo ldg photo photo- ) the the set of same the in was. This done : - intuitive A - solution (v/v) –water 8:2

stability in solution of and colours glycone, glycone,

[8] stability lut of in solution due to additional geometrical constraints and changes in electronic changes in and geometrical constraints additional to due solution in , - of lut stability of , with most of them containing sugar moieties [9]

. Van der Waals forces are one of t of one are forces Waals der Van . at first at upon irradiation relativeupon irradiation to lut

(chapter lut

on the- photo - the proteins of wool , 7

due the of to possibility - in solution was in studied solution in duplicate asof aits glycosylation function O in in solution in was much slower much was

- 42), 42), glucoside ( in solution itsglycosylation afunction as pattern of [10] stability of lut with light above 300nm lightwith above Whereas Whereas

(chapter 5) (chapter lut .

lut the the tructures oftructures the flavones themselves

in aerated aerated 4.4.1.1,in discussed section experiments in [29] monoglucoside

Fig. 4.5) Fig. 4.2and (Table on diglucoside) on monoglucoside, and (aglycone, alum lmg presence of A of presence

was also observed by Smith et al. also observedbySmith was

. Naturally, such as interactions t

are solution in . glycosylation

P photodecomposed - s explanation ossible mordanted wool he ways important through which dye , . Also the experimental data were were data experimental the Also . —and to theirinteraction with the lmg l 3+ elaborated upon to , with the relative rates of )

confer and . 1.4 dyedwith it Although this lut

stability to flavonoid × for th for -

faster than lut than faster 7,3ʹ - —which differ e i O

n this section

his influence his effect of t of effect - [6] diglucoside

increased increased

, it seems seems , the he he .

irradiation in this way, released of of intramolecular H impossibility lmg C the hydroxyl group at C to in for deprotonation hydroxylgroup at this should of alowering discussion, undergo ne [18] The [flavone] photoon the - 4.5.EffectFigure of glycosylation pattern— of S = photodecompositionzero- the of bythe flavones the of 0

[18] electronic distribution compensate compensate lut → S → lmg Stabilization of the transition state of the formation of the flavonoxyl radical of lut W

+ [flavone]

possibility of lut ( because because ith a pK OH group

Table Table and . .

1 Upon irradiation

the the causing , The The 0 transition , in mM; , in

replicate. ldg a account

Fig. 4.2 and Fig. change in electron in change electronic electronic stability of lut a c .

0 it has an an has it a change achange to lead ould < 9, t < ); Duration of irradiation of s of irradiation of Duration ); part of the of part T may account force the differen account may he flavonoxyl radical of quercetin of radical flavonoxyl he -

bond Solvent: aerated methanol

of apigenin - for the for he 7 he - through high formed 3—is

electronic excited state lifetime of l to to distribution could could - increase the electronring A of density with lightwith above 300nm OH 4.3) [22] O

loss of loss

reduced - in solution. Note: Subscripts of code solutions of denote glucose glucose . is oneis of the two . difference the explain

is accompanied by by accompanied is Thus Th ic distribution equivalent

us, us, the the electron density electron s , in the, in experiments reported here photo-

in of of such a de moiety electronic density electronic

the preferred process through which absorbed through process energy preferred the lut aglycone,monoglucoside, diglucoside and olutions 24.0 = h (1,440 min) –water 8:2 (v/v);

stability and instead of a hydroxyl group ata C hydroxyl of instead protonation protonation

most acidic groups of lut between between apigenin energy irradiation (radiolysis) only , a

order integrated rate law ([flavone] of of

lut —a shiftelectro of of lmg of between the rates of photodecomposition theof rates between ring

increase increase mainly undergoes S an .

lut

of the A its to that o —3′ mg

. - A r excited state lut state excited afor could way be rate

like flavonoid 2

. -

= quality of description description of quality = deoxy deoxy to This in thein S of photodecomposition

- f be be ring ring apigenin n

is impossible in the case of

long densi lut ; Data from 2 through deprotonation at through deprotonation at , 1 —are lut

er than thatofer lut than

because of the change change the of because [5, 32][5, ring ring from ty having a ; importantly for this

have have to expected is

similar 0 [33]

→ S → . — The nd - . n additional n additional 7.

Flavonoxyl 1

[18] transition tendency A to ring Such an Such an and that .

and The

by by 63

is is ,

Chapter 4 radicals are formed under milder conditions too, since phenoxyl radicals are formed via flavones in the dyeing baths and rinsing waters were calculated with RP-HPLC–UV calibration irradiation of phenols in the UV–vis range of the electromagnetic spectrum [34]. Kozlowski et curves after their y-intercepts had been set to zero. A picture of the dyed pieces of wool of the al. calculated the unpaired electron of the radical species to be delocalized, with C-2 bearing first experiment is seen in Figure 4.6, and the percentages of leftover flavones after dyeing from high spin density if the radical was to be formed at the 3-OH group [33]. Those authors expect both experiments are listed in Table 4.6. molecular oxygen to react with the flavonoxyl radical of quercetin precisely at C-2, before the formed peroxyl radical reacts as the decomposition of the flavonoid continues. Thus, the formation of the flavonoxyl radical is expected to be the first step of the photo-oxidative degradation—referred to as photodecomposition throughout this chapter—of lut and its glycosyl conjugates too. In flavones that have a catechol group, the energy of the flavonoxyl radical is lowered by intramolecular H-bond involving the O-atom bearing the unpaired electron and the neighbouring hydroxyl group [33, 35], [10] (chapter 5). This is why the flavonoxyl radical of lut is expected to be formed at the catechol group; with that formed at the 4′-OH group having been calculated to be the most stable [35]. The H-bond at 3′-OH/4′-O∙ may stabilize not only the radical, but also the transition state of its formation. Such a stabilization could not take place in the case of ldg since it has an O-glucose group instead of a hydroxyl group at C-3′. A lower Figure 4.6. Pieces of alum-mordanted wool dyed with the individual flavones—lut, lmg, or ldg—and energy of the transition state of the formation of the flavonoxyl radical of lut would cause the with an extract of weld (from left to right). activation energy barrier for the formation of the radical of lut to be smaller than that for the formation of the radical of ldg, resulting in a faster formation of the radical of lut. 3+ The rate of photodecomposition of ldg in presence of Al in aerated methanol–water 8:2 Table 4.6. Leftover flavones after dyeing Al3+-mordanted wool with lut, lmg, ldg, or an extract of weld (each case, (v/v) solution was also obtained through the set of the experiments discussed in section 4.4.1.1; n = 1). thus, with light above 300 nm. As seen in Table 4.2, ldg engaged in complex formation with a 3+ Leftover flavones in dyeing baths (%) Al (ldg0.05–Al0.05) photodecomposed at least 2× faster than free ldg. By analogy to lut, the Dye Dyeing conditions main species in solution at a ldg–Al3+ 1:1 ratio are expected to be the ldg–Al3+ 2:1 complex and Lut Lmg Ldg 3+ free ldg (Fig. 4.3 and ensuing discussion). As previously discussed, lut in presence of Al also Lut, lmg, or ldg 80 °C, 15 min b 5 16 63 photodecomposes faster than free lut. Similarly to lut, the faster photodecomposition of ldg in 3+ presence of Al could have resulted from a smaller possibility of accelerated S1→S0 transition Extract of weld 80 °C, 15 min b 6 18 66 via IC due to diminished charge transfer from the B-ring towards rings C and A, diminished possibility of ESPT, or both of them being diminished. This possible change in the process of Lut, lmg, or ldg 100 ºC, 1 h 5 13 58 the S1→S0 transition of part of the ldg molecules could have led to their electronic excited state lifetime being longer, accounting for the decreased photo-stability of ldg in presence of Al3+. a In dyeing baths and rinsing waters, in experiment in which dyeing processes were carried out at 100 ºC for 1 h. b Dyeing processes of samples of Fig. 4.6. 4.4.2.2. Colours of alum-mordanted wool dyed with lut as a function of the glycosylation pattern of the compound Two experiments were carried out in which Al3+-mordanted wool was dyed with different dyes. The colours of alum-mordanted wool dyed with lut or lmg seem to be similar, but different In one of them, alum-mordanted wool was dyed at 80 °C for 15 min. The percentages of leftover from those in which either an extract of weld or ldg are used (Fig. 4.6 and Table SI C.6). The flavones in the dyeing baths were calculated after analysis by RP-HPLC–UV. This experiment wool dyed with the latter two dyes was much less colourful—chroma 60 and 30, respectively— consists of the following two parts: than that dyed with either lut or lmg (chroma 80–85). • Wool was dyed with the individual flavones. In each case, a piece of mordanted wool was In the case of ldg, this is possibly due to the quantity of flavone binding to the wool. Table dyed with lut, lmg, or ldg; 4.6 lists the percentages of leftover flavones in the dyeing baths. The uptake of lut, lmg, and 3+ • A piece of mordanted wool was dyed with an extract of weld (thus, with all three flavones ldg by the Al -mordanted wool is estimated from them. The data suggest lut to be extensively simultaneously). taken up by the mordanted wool, as well as most lmg, but this to be the case only for one-third In the other experiment—not further described—wool mordanted with aluminium sulphate and of ldg molecules. The lack of a free catechol group could be a reason for a limited binding of tartaric acid was dyed at 100 °C for 1 h with the individual flavones. Thus, also in this case, ldg to the mordanted wool. This could also be due to steric hindrance—bulkiness of the glucose pieces of mordanted wool were dyed with lut, lmg, or ldg. Then, the percentages of leftover moiety—and solubility in the dyeing bath (ldg is more hydrophilic than both lmg and lut),

64 65

Two were experiments carried which in out Al pattern glycosyl conjugates too conjugates glycosyl s piece the possibility or ofboth ESPT, via Al of presence for the for barrier energy activation radical of flavonoxyl the formation state of ofenergy the the transition of been degradation as to photodecomposition —referred formation of the flavonoxyl expected radicalbe is to o f thereacts flavonoidas continues. theformed decomposition radical peroxyl at at dyed was acid tartaric In the other 4.4.2.2. photodecomposes faster than free lut of case the in lut neighbouringgroup hydroxyl intramolecular H at C theprecisely with flavonoxyl radical ofmolecular react quercetin to oxygen 64 of consists flavones the in dyeing baths In one of them lifetime free main species solution in a (v/v) solution formation ldg of the radical of the radical w high if density the spin al. irrad r Al thus, since too, conditions milder under formed are adicals 3+ formed at formed be to expected is I • • The of rate photodecomposition

calculated the unpaired electron of the radical species to be delocalized, with C with delocalized, be to species radical the of electron unpaired the calculated radical S n IC

l iation ofiation phenols 1 ( dg calculated calculated with lightwith nm above 300 flavones flavones →S simultaneously A piece mordanted of wool dyed with W ldg

of mordanted wool w of wool mordanted due to diminished charge transfer from the B from the charge transfer diminished to due

being of the of compound Colour wool was was wasool In wool dyed flavones.each the case,with individual a pieceof mordanted ( Fig. Fig. 0.05 0 the also the , but

thetwo parts: following transition

experiment –Al ldg 4.3 longer was was , 3+ that a s 0.05

to be to lut lum

could has an an has since it of of and) ensuing discussion - also also

) photodecomposed bond involving the O bond , lmg , the acatechol the group , have alum -

, accounting for, accounting the dec ) of the partldg of mordanted

in the in UV . the mostthe stable

have have obtained not further—not described transition state of its formation. . , or ldg t - 100 mordanted dyed with lut wool a

were calculated were result l ere of thembeing diminished. dg °C for°C 1h [33, 35] [33, , resulting formationof lut of a in the faster radical . the catechol group; catechol the

–Al As seen in Table Table in seen As ; of the electromagnetic spectrum electromagnetic the of range –vis through wool was dyed for at The 80°C 15min. percentages of leftover O dyed with as

an extract of weld of extract an with dyed was ed - to be formed at the 3 the at formed be to

glucose group instead of a h 3+ mation of the radical of lut ofmation the

of of from from . Similarly to lut

1:1 ratio

[35] molecules molecules , ldg at least 2 least at [10] the set of thediscussed experiments section in 4.4.1.1; . Thus, a flavones.individual Thus, thewith a . As previously lut discussed,

in presence of Al of presence in lut T -

smaller after analysis by RP analysis after - d photo rease (chapter 5) (chapter unpaired electron unpaired the bearing atom he he 3+

—wool mordanted aluminium with , be to expected are by lowered by radical is flavonoxyl the energyof - lmg H mordanted wool was dyed with differentwith mordanted dyed woolwas dyes. × could could 4.2, -

wi bond at 3 faster than free ldg free than faster

, or ldg or , possibility throughout chapter this - th , theof faster ldg photodecomposition diminished diminished and A, C rings ring towards ldg

Such a stabilization c have have that formed 4′ at the . T - This why is the first step of the- photo OH group [33]

his his stability ldg of engaged in complex formation withengaged complex in

henoxyl radicals are formedphenoxyl via as a function of of afunction as . led to their to led ′ Then, the percentages of leftover - 3+ possible possible OH/

of of ydroxylgroup at C

in in to be to - the the accelerated S accelerated HPLC

4′ ( aerated aerated thus, with all three flavones allthree with thus, flavones - l O∙ dg

of of radical flavonoxyl the change in smaller .

in in

electronic excited state By analogy to lut to analogy By may –Al Those authors expect . Those authors expect –UV of A of presence in presence of Al of presence methanol lut

3+ [34] ould ould -

the the of —of OH group OH having . This experiment. This stabilize

2:1 complex2:1 and th would lso lso 1 →S

et et . Kozlowski an thatfor the the the glycosylation not not - .

2, before the 2, before the sulphate and in this case - lut 3′. 0 water 8:2 8:2 –water - process process

take place place take Thus, 2 bearing transition cause the the cause oxidative

and the the and not only only not A lower A and its and its 3+ l

, also also 3+ the the the the in in . of ,

moiety ldg of taken up by wool,as the mordanted well as most both experiments are4.6. listed in Table first experiments is seen in Figure percentage 4.6,and the their after curves The colours ofThe colours F flavones the in dyeing baths ldg 4.6 than that dyed with either l colourful wool dyed withtwo latter less the dyes much was b a with an extract of weld (from left to right) from which those in either

Lut In In Extract of weld of Extract Lut Dye 1). n = Table Dyeing of samples processes Fig. of 4. igure igure ldg In the case of ldg of case In the

dyeing baths and rinsingwaters theto mordanted wool. lists the percentagesof dyeing leftoverflavonesbaths. lists the in by , ,

lmg lmg

molecules. and solubility in the dyeing bath ( —and in solubility the the 4.6. 6 , or ldg , or , or ldg , or .

Leftover flavones after dyeing

Al Pieces of alum 3+

- mordanted wool mordanted alum

y The lack of a free catechol group couldreasonforfree beThe of catechol a a lack - intercepts intercepts 100 80 °C, 15min 80 °C, 15min Dyeing conditions Dyeing , t - dyed dyed mordanted wool his his ºC, 1h ºC, - mordanted wool dyed is possiblyis ut Th an

and rinsing waters or l or

, in e in , is . had zero set to been extract of weld or ldg or weld of extract

could b b is estimated mg

6 xperiment in whichxperiment inwere out dyeing carried at 100 processes ºC for 1 h .

(c of the to quantitydue of 3+ bulkiness of the glucose of the —bulkiness hindrance steric also be due to . hroma

- 5 Lut 6 5 in dyeing baths (%) baths dyeing in flavones Leftover mordanted wool

with ldg from th from 80–85). with the individual flavones

were calculated were

is more hydrophilic than both lmg both moreis hydrophilic than lmg l ut

A picture of the dyed pieces of wool of the the woolof dyed of pieces ofA picture the

(Fig. (Fig. used are , but this, but to or or

with with . l

Th of of mg lut , respectively 30, 60and —chroma 13 Lmg 18 16 e flavone binding to the wool. bindingflavone to leftover flavones after dyeing from from dyeing after flavones leftover

data seem lmg

with

be the case only case for one the be , suggest ldg The The

to to 4.6

RP , or an or , extractweld of (each be uptake uptake - and Table

a HPLC —

similar lut a lut

limited binding of

, extensively beto extensively –UV of of lmg 58 Ldg 66 63 ,

lut

but differ but SI , or ldg , or

calibration ,

C. lmg and and ). The 6). The Table Table — - , third third lut and and and ent — 65 case, ), ), .

Chapter 4 equivalently to suggestion by Mouri et al. when reporting a preferential binding of flavonol [7] Malathy R, Rajendran M. Study on photochemically and chemically generated singlet aglycones to silk relative to that of the glycosides [36]. oxygen quenching rate constant of quercetin. Int J Green Herb Chem 2013; 2(3): 631–40. [8] Heim KE, Tagliaferro AR, Bobilya DJ. Flavonoid antioxidants: chemistry, metabolism and structure-activity relationships. J Nutr Biochem 2002; 13: 572–84. 4.5. Conclusion The photo-stability of lut in solution—polar protic solvent—and the light-fastness of the colour [9] Clayden J, Greeves N, Warren S. Organic chemistry. 2nd ed. Oxford University Press: New of weld-dyed wool decrease with increasing [Al3+]. Thus, the lower the [Al3+] used for York; 2012. mordanting the wool, the more light-fast its colour. Lowering the [Al3+] appears to have no [10] Larson RA. Naturally ocurring antioxidants. 1st ed. Lewis Publishers: Boca Raton/New negative influence on the wash-fastness of the colour. As the gain in light-fastness by the use York; 1997. 3+ of low [Al ] to premordant the wool is limited, however, this does not seem to be a way to [11] Zhang X, Cardon D, Cabrera JL, Laursen R. The role of glycosides in the light-stabilization meet today’s requirement of light-fastness of the colours of dyed textiles by itself. Nevertheless, of 3-hydroxyflavone (flavonol) dyes as revealed by HPLC. Microchim Acta 2010; 169(3–4): it may be part of a broader strategy to address the need for increased light-fastness of the colour 327–34. of wool dyed with weld. Implementation of this approach by dyers is expected to clarify whether it results in benefits for textile dyeing practice. On one hand, lowering the [Al3+] to [12] Fanning JC. The chemical reduction of nitrate in aqueous solution. Coord Chem Rev 2000; mordant the wool has the drawback of limiting the obtainable yellow colour, as its saturation 199(1): 159–79. increases with increasing [Al3+]. On the other hand, reduction of the quantity of mordant should [13] Villela A, van der Klift EJC, Mattheussens ESGM, Derksen GCH, Zuilhof H, van Beek increase the needed environmental friendliness of the dyeing of textiles with natural dyes, and TA. Fast chromatographic separation for the quantitation of the main flavone dyes in Reseda have financial benefits. luteola (weld). J Chromatogr A 2011; 1218(47): 8544–50.

[14] Villela A, Derksen GCH, van Beek TA. Analysis of a natural yellow dye: an experiment

for analytical organic chemistry. J Chem Educ 2014; 91(4): 566–9. 4.6. Supplementary material Appendix C contains information supplementary to that in this chapter. Sections material and [15] Goodman TM. International standards for colour. In: Best J, editor. Colour design – methods, results and discussion, author contributions and references are available. theories and applications, Woodhead Publishing/The Textile Institute: Oxford/Cambridge/etc.; Reference is made in this chapter only to part of the information in appendix C. Thus, readers 2012, p. 177–218. are referred to it for further details on the work. [16] Bechtold T, Mahmud-Ali A, Mussak R. Natural dyes from food processing wastes. In: Waldron K, editor. Handbook of waste management and co-product recovery in food processing, CRC Press/Woodhead Publishing: Boca Raton/Boston/etc.; 2007, p. 502–33. 4.7. References [17] Gilbert A, Baggott J. Essentials of molecular photochemistry. 1st ed. Blackwell Science: [1] Duffield PA. Dyeing wool with acid and mordant dyes. In: Lewis DM, Rippon JA, editors. Oxford; 1991. The coloration of wool and other keratin fibres, John Wiley & Sons/Society of Dyers and [18] Amat A, Clementi C, De Angelis F, Sgamellotti A, Fantacci S. Absorption and emission Colourists: Electronic format; 2013, p. 205–28. of the apigenin and luteolin flavonoids: a TDDFT investigation. J Phys Chem A 2009; 113(52): [2] Crews PC. The influence of mordant on the lightfastness of yellow natural dyes. J Am Inst 15118–26. Conserv 1982; 21(2): 43–58. [19] Bondarev SL, Knyukshto VN, Tikhomirov SA, Buganov OV. Mechanism for highly [3] Deng H, van Berkel GJ. Electrospray mass spectrometry and UV/visible spectrophotometry efficient non-radiative deactivation of electronic excitation in rutin. J Appl Spectrosc 2016; studies of aluminum(III)-flavonoid complexes. J Mass Spectrom 1998; 33: 1080–7. 82(6): 929–35. [4] Cornard JP, Merlin JC. Comparison of the chelating power of hydroxyflavones. J Mol Struc [20] Smail K, Tchouar N, Barj M, Marekha B, Idrissi A. Luteolin organic solvent interactions. 2003; 651–653: 381–7. A molecular dynamics simulation analysis. J Mol Liq 2015; 212: 503–8. [5] Favaro G, Clementi C, Romani A, Vickackaite V. Acidichromism and ionochromism of [21] Smith GJ, Markham RK. Tautomerism of flavonol glucosides: relevance to plant UV luteolin and apigenin, the main components of the naturally occurring yellow weld: a protection and flower color. J Photochem Photobiol, A 1998; 118(2): 99–105. spectrophotometric and fluorimetric study. J Fluoresc 2007; 17(6): 707–14. [22] Huvaere K, Skibsted LH. Flavonoids protecting food and beverages against light. J Sci [6] Smith GJ, Thomsen SJ, Markham KR, Andary C, Cardon D. The photostabilities of Food Agr 2015; 95(1): 20–35. naturally occurring 5-hydroxyflavones, flavonols, their glycosides and their aluminium complexes. J Photochem Photobiol, A 2000; 136(1–2): 87–91.

66 67

of low [Al low of mordanting the wool The coloration of wooland other keratin Wiley John fibres, & S editors. JA, Rippon In:Lewis DM, Duffield dyes. Dyeing PA. acidwool with [1] and mordant 4.7. References work. the on details further for it to are referred it may negative wash- influence onthe 66 and spectrometry mass Electrospray GJ. Berkel van DengH, [3] Conserv 21(2): 1982; 43 The Crews influence[2] PC. ofyellow AmInst mordant on the lightfastness of dyes. J natural Electronic 2013,p.205–28. Colourists: format; chapter this in made is Reference methods C Appendix 4.6. S benefits financial have in mordant the wool of wooldyed weld with of light requirement today’s meet weld of The- photo 4.5. Conclusion aglycones to silk relativeto equivalently suggestion to complexes. J A136(1 2000; Photochem Photobiol, J complexes. ofstudies aluminum(III) [Al increasing with increases whether it results in benefits for textile dyeing practice. On naturally occurring 5 occurring naturally GJ, Thomsen MarkhamSmith SJ, [6] Anda KR, Fluorescspectrophotometric 17(6): study. 2007; 707–14. fluorimetric J and yellow and luteolin apigenin, the thenaturally a maincomponents of occurring weld: i A, Roman C, Vickackaite Favaro Clementi Acidichromism[5] V. G, andionochromism of 381–7. 651–653: 2003; Cornard Merlin[4] JP, of Comparison theJC. chelating Mol Struc power hydroxyflavones. J of c the rease upplementary upplementary

broader a broader of part be - decrease dyed wooldecrease , 3+ stability of lut results and discussion resultsand contains information supplementary ] to pre to ] needed needed

has th has mordant the wool is limited the woolis mordant environmental friendlenvironmental , material . the more - hydroxyflavones, flavonols, their glycosides and their aluminium their their aluminium glycosides and flavonols, hydroxyflavones, –58. - e drawbackyellow of the limiting obtainable saturation as its colour,

flavonoid complexes. J Mass 1080–7. flavonoid J Spectrom 33: complexes. 1998; . —polar protic solution solventin strategy to addre Implementation approachclarify to expected is of this bydyers t by Mo by 3+ hat of the glycosides

] with increasing [Al . On the r other hand,

fastness of thefastness colour. light- -

fastness of the colours of dyed tex dyed of colours the of fastness only only uri uri , et al. fast its colour. author contributions in appendix information appendix in of the part to iness of the dyeing of textiles with natural dyes, and and dyes, with natural textiles of dyeing the of iness ss the need for increased light increased for need the ss

when reporting a preferential bindingreporting of a preferential when

, [36] to that in this chapter. however, –2): 87–91. 3+ ry C, Cardon D. The photostabilities of ry Cardon photostabilities C, D. The eduction of the quantity of mor the quantity eduction of ] Lowering the [Al the Lowering . . Thus

A s the gainlight in the s —and this

, the lower [Al the lower , the

one hand, and

UV/visible spectrophotometry spectrophotometry UV/visible to does seem not the light- tiles references ons/Society of and Dyers S

3+ by itself. Nevertheless, Nevertheless, by itself. lowering the [Al - ections material and ections ] appears to have no no have to appears ] fastness of the co the of fastness fastness -

fastness by the use use by the fastness C

. are Thus, rea 3+ of the colour be a way to to way a be dant ] available. available. used for flavonol flavonol sh 3+ ould ould ders ders lour lour ] to to ] of 3 of Laursen R Cabrera JL, Zhang D, Cardon [11] X, York; 1997. Boca Raton/New Lewis Publishers: ed. Larson 1st antioxidants. ocurring [10] Naturally RA. York; 2012. New University Oxford Press: chemistry. 2nded. ClaydenWarren Organic N, S. Greeves[9] J, structure Heim KE, Tagliaferro Flavonoid Bobilya AR, chemistry,[8] antioxidants: DJ. and metabolism Intoxygen of quercetin.2(3): Green J constant Herb631–40. rate quenching 2013; Chem and Malathyge chemically Study M. onphotochemically [7] Rajendran R, Food Agr 95(1 2015; Huvaere[22] LH. K, Skibsted protecting Flavonoids food andagainst beverages light. Sci J PhotochemPhotobiol,protection A118(2): J and 1998; 99–105. flowercolor. GJ, Smith [21] analysis. Mol A molecularLiq J 503–8. dynamics 212: simulation 2015; Luteolin interactions. A. Idrissi solvent organic K,B, TchouarSmail Barj Marekha M, [20] N, 82(6): 929–35. non efficient Bondarev VN,Buganov forOV. SA, Tikhomirov Mechanism highly [19] Knyukshto SL, 15118–26. of the apigenin andflavonoids:a luteolin TDDFT Phys investigation.A113(52): 2009; J Chem [18] Amat A, Clementi C, De Angelis F, Sgamellotti A, 1991. Oxford; Gilbert[17] A, Baggott Essentials J. molecular of photochemistry. Blackwell 1st ed. Science: Boca Press/Woodhead Publishing: processing, Raton/Boston/ CRC K,Waldron of editor. waste Handbook managementand co- an theories Goodman TM.InternationalIn:[15] editor. design Colour Best standards colour. J, for – Educ Chem 91(4):for 2014; 566–9. chemistry. organic analytical J experiment an dye: yellow a natural of VillelaBeekGCH, TA. van Analysis A,[14] Derksen luteola TA. Villela A,[13] van der Mattheussens EJC, Klift DerksenESGM, GCH, Zuilhof H, van Beek 199(1): 159–79. Fanning[12] The chemicalreduction aqueousCoordin of JC. nitratesolution. Rev Chem 2000; 327–34. [16] Bechtold[16] T, Mahmud- 2012, p.177–218. - Reseda Reseda dyes in of the mainflavone quantitation the separationfor chromatographic Fast hydroxyflavone dyes (flavonol) by 169(3 2010; –4): asHPLC.Acta revealed Microchim J Chromatogr A 2011; 1218(47): 8544–50. Chromatogr 8544–50. J A1218(47): 2011; (weld). - activity relationships. J Nutr Biochem 572–84. J 13: 2002; activity relationships. d applications, Woodhead Publishing/The Textile Institute: Woodhead Publishing/The Oxford/Cambridge/ Textile applications, d - radiative deactivation radiative Markham Tautomerism RK. of flavonol glucosides: relevance plantUV to ): 20 ): –35. Ali A,NaturalAli Mussak R. dyesIn: food processing from wastes.

of electronic excitation in rutin. rutin. in excitation ApplSpectroscJ of electronic 2016; . The role of glycosides light the in of role . The

Fantacci Absorptionand S. emission product recove product etc. ; 2007, p.502–33. 2007, ;

nerated singlet singlet nerated - stabilization ry food in etc. 67 ;

Chapter 4 [23] Amat A, Clementi C, Miliani C, Romani A, Sgamellotti A, Fantacci S. Complexation of apigenin and luteolin in weld lake: a DFT/TDDFT investigation. Phys Chem Chem Phys 2010; Chapter 5 12(25): 6672–84.

[24] Hanson AR. What is colour? In: Best J, editor. Colour design – theories and applications,

Woodhead Publishing/The Textile Institute: Oxford/Cambridge/etc.; 2012, p. 3–23.

[25] Vinod KN, Puttaswamy, Gowda KN, Sudhakar R. Extraction, identification and adsorption-kinetic studies of a natural color component from G. sepium. Nat Sci 2010; 2(5): 469–75. [26] Kanazawa H, Mori J. Relation between dye uptake and K/S value in the dyeing of natural fiber. Sci Rep Fukushima Univ 1995; 57: 17–24.

[27] Vankar PS, Shanker R, Wijayapala S. Dyeing of cotton, wool and silk with extract of

Allium cepa. Pigm Resin Technol 2009; 38(4): 242–7.

[28] Rippon JA. The structure of wool. In: Lewis DM, Rippon JA, editors. The coloration of wool and other keratin fibres, John Wiley & Sons/Society of Dyers and Colourists: Electronic format; 2013, p. 1–42. [29] Rippon JA. The chemical and physical basis for wool dyeing. In: Lewis DM, Rippon JA, editors. The coloration of wool and other keratin fibres, John Wiley & Sons/Society of Dyers and Colourists: Electronic format; 2013, p. 43–74. Analysis of a natural dye: an experiment for [30] Cristea D, Vilarem G. Improving light fastness of natural dyes on cotton yarn. Dyes Pigments 2006; 70(3): 238–45. analytical organic chemistry [31] Giles CH. The fading of colouring matters. J Appl Chem 1965; 15: 541–50. [32] Ramešová Š, Sokolová R, Degano I, Bulíčková J, Žabka J, Gál M. On the stability of the bioactive flavonoids quercetin and luteolin under oxygen-free conditions. Anal Bioanal Chem 2012; 402(2): 975–82.

[33] Kozlowski D, Marsal P, Steel M, Mokrini R, Duroux JL, Lazzaroni R, Trouillas P. Theoretical investigation of the formation of a new series of antioxidant depsides from the radiolysis of flavonoid compounds. Radiat Res 2007; 168(2): 243–52. [34] Rayne S, Forest K, Friesen KJ. Mechanistic aspects regarding the direct aqueous environmental photochemistry of phenol and its simple halogenated derivatives. A review. Environ Int 2009; 35(2): 425–37. [35] Leopoldini M, Russo N, Toscano M. The molecular basis of working mechanism of natural polyphenolic antioxidants. Food Chem 2011; 125(2): 288–306.

[36] Mouri C, Mozaffarian V, Zhang X, Laursen R. Characterization of flavonols in plants used for textile dyeing and the significance of flavonol conjugates. Dyes Pigments 2014; 100: 135– 41.

The content of this chapter is largely that of the following paper: Villela A, Derksen GCH, van Beek TA. Analysis of a natural yellow dye: an experiment for analytical organic chemistry. J Chem Educ 2014; 91(4): 566–9.

[23] Amat A, Clementi C, Miliani C, Romani A, Sgamellotti A, Fantacci S. Complexation of apigenin and luteolin in weld lake: a DFT/TDDFT investigation. Phys Chem Chem Phys 2010; Chapter 5 12(25): 6672–84.

[24] Hanson AR. What is colour? In: Best J, editor. Colour design – theories and applications,

Woodhead Publishing/The Textile Institute: Oxford/Cambridge/etc.; 2012, p. 3–23.

[27] Vankar PS, Shanker R, Wijayapala S. Dyeing of cotton, wool and silk with extract of

Allium cepa. Pigm Resin Technol 2009; 38(4): 242–7.

[36] Mouri C, Mozaffarian V, Zhang X, Laursen R. Characterization of flavonols in plants used for textile dyeing and the significance of flavonol conjugates. Dyes Pigments 2014; 100: 135– 41.

5.1. Introduction Undergraduate students of molecular life sciences and biotechnology programs of Wageningen University take the course Analytical Methods in Organic Chemistry (AMOC) at the beginning of the second year. AMOC has both theoretical and practical components. Its practical component is divided into spectroscopy and chromatography. During the latter, students perform experiments individually and become acquainted with a range of techniques used for the analysis of organic compounds. The experiment reported here was used in 2011, 2012, and 2013.

5.2. Description of the experiment 5.2.1. Introduction The compounds of weld that are responsible for the yellow colour of dyed alum-treated wool belong to the class of flavonoids [1]. The three main compounds are 2-(3,4-dihydroxyphenyl)- 5,7-dihydroxy-4H-1-benzopyran-4-one (luteolin, lut), lut-7-O-glucoside (lut monoglucoside, This experiment exposes second-year undergraduate students taking a course in analytical lmg), and lut-7,3′-O-diglucoside (lut diglucoside, ldg) [2, 3]. The compound used as internal organic chemistry to high performance liquid chromatography (HPLC) and quantitative standard (i.s.) for their quantitation in weld is 5,7-dihydroxy-2-phenyl-4H-1-benzopyran-4-one analysis using the internal standard method. This is accomplished using the real-world (chrysin) [4]. The structures of these four compounds are depicted in Figure 5.1. application of natural dyes for textiles. The extracted flavonoids of the plant weld are responsible for the yellow colour of the dyed wool. Dried and ground weld is extracted for OH dyeing wool and quantifying the plant’s three main flavonoids. The students also mimic the HO OH OH OH OH OH work of chemists investigating historical textiles by carrying out a small scale extraction of the HO OH HO O O OH OH 3' 4' dyed wool. Twenty-one students carried out the experiment and their samples were analysed O ldg 2' OH O lmg OH lut OH i.s. HO 1 HO 8 7 2 1' using either a traditional 5 µm-particle size HPLC column or a modern 1.8 µm-particle size O O 5' O O HO O HO O 6' 4 ultrahigh-pressure liquid chromatography (UHPLC) column mounted in a conventional HPLC 6 3 5 system. OH O OH O OH O OH O

Figure 5.1. Structures of the three main flavonoids of weld and the compound used as internal standard (i.s.) for their quantitation in weld. Their identities are luteolin-7,3′-O-diglucoside (luteolin diglucoside, ldg), luteolin-7-O-glucoside (luteolin monoglucoside, lmg), luteolin (lut), and chrysin (i.s.).

Since the first use of the experiment, the protocol of the experiment has undergone modifications. Both student and teaching assistants’ feedback contributed to this. Students who carried out the experiment in 2012 and 2013 were busy with its practical part for ~4 h. An updated version of the protocol of the experiment handed out to students is available in Appendix D.

70 71

system. ultrahigh 70 using dye work of chemists investigating historical extraction textiles by scale carrying out asmall Th dyeing wooland quantifying the plant’s flavonoids. three The student main organic chemistry to responsibleyellow for the colour application analysis using standard the internal method. is d

experiment experiment wool. either

- pressure liquid chromatography liquid pressure

Twenty of naturalof a traditional 5 exposes exposes - one

high performance liquid chromatography liquid performance high

dyes for textiles students carriedstudents the out experiment second µ m - particle column size HPLC -

year undergraduate students taking a course in analytical analytical in acourse taking students year undergraduate of the dyed Dried and wool. ground extracted is weld for .

( UHPLC The extracted flavonoids of the plant weld are are plantweld the The flavonoids of extracted using the real isThis the accomplished using ) column mounted in a columnmountedin conventional HPLC

or and

a modern modern t heir samples were analysed were samples heir (HPLC) (HPLC) 1.8 1.8 s also mimic and µ m - particle size quantitative - world of the of the

the the

( diglucoside, 5.2. Description of the experiment the of Description 5.2. 2013. 5,7- of flavonoids belong class the to yellow colour that areresponsible the weld for The compounds of 5.2.1. Introduction the analysis of organic compounds experiments perform s component dividedinto is Itsyear. components. has theoretical and practical both practical of theAMOC second University take the course Analytical MethodsOrganic course in the Chemistry take (AMOC) University Undergraduate of students molecular life sciencesand of biotechnology programs Wageningen 5.1. Introduction and teaching modifications. Both student Since standard ( Figure lmg carried 2012 the in out experiment D Appendix in available is handed students to out experiment the protocol of updated of the version standard (chrysin) [ (chrysin)

HO HO i.s. ). dihydroxy ) OH O

6 and O 7 the first use of the experiment, the protocol of the experiment hasexperiment undergone of the the the experiment, protocol the first use of 5. 5 OH OH 8

ldg ( i.s. 1. ) i.s. 4] lut O O 4 1 Structures of the main three flavonoids weld of the and ) for their quantitation for their ) in weld. ldg . . The structures arefour of Figure these5.1. compounds in depicted 2 - 3 1' 2' for their quantitation in weld is 5,7- - 7,3′ ), luteolin ), 4H HO O 6' 3' - 4' - 5' O 1 OH O OH - -

benzopyran diglucoside ( individually OH - OH 7- O - glucosidemonoglucoside, (luteolin HO HO pectroscopy and [ - 4

1] O OH and lut . The experiment reported here was used was here reported experiment The . - one lut (luteolin, O . T

and 2013 OH OH

lmg diglucoside, diglucoside, become become he three main compounds are 2- main three he Their identities are O O assistants’ feedback contributed to this. Students w Students this. to contributed feedback assistants’ OH were busy with its practical part for acquainted with a range of techniques used for for ofused techniques a range acquainted with c OH hromatography. During the latter, students the latter,hromatography. students During dihydroxy ldg ), ), HO ) lut [ 2,

- luteolin OH 7- 3] - lut 2 O . - O O lmg phenyl - The compound used as glucoside ( - ), ), compound as used internal 7,3′ OH

of dyed alum dyed of luteolin ( luteolin - - O OH 4H (3,4 - diglucoside ( diglucoside - 1- lut - in 2011,2012,and in HO dihydroxyphenyl) lut benzopyran

at the beginning monoglucoside, ) , OH and chrysin i.s. - treated wool wool treated O O luteolin luteolin ~

4 h. internal internal - 4-

one An An 71 ho -

Chapter 5 5.2.2. Practical part 5.2.3. Data processing The students carry out two extractions of dried and ground weld simultaneously: One extract is With the chromatogram of the weld sample containing peak areas and retention times, the used for dyeing wool and the other is used for quantifying the plant’s three main flavonoids. students start by assigning the peaks corresponding to ldg, lmg, lut, and i.s. on the basis of the For the wool dyeing, 1.5 g of the dried and ground weld is placed in an Erlenmeyer flask with retention times. Then, using the equation that relates peak area to the quantity of compound in 30 mL of 96% ethanol–water 3:1, and stoppered. The mixture is sonicated for 10 min. After a sample and predetermined relative response factors (RRFs), the students calculate the filtration through a folded filter paper to a round-bottom flask, the ~20 mL of 96% alcohol is concentration of the three main flavonoids in the weld sample. This is done using the following removed by a rotatory evaporator. The dyeing bath is prepared by transferring the alcohol-free equation: extract from the round-bottom flask to a 150 mL beaker using 4 × 15 mL portions of deionized water. To quantify the plant’s three main flavonoids, 200 mg of the dried and ground weld is A w . . w = placed in a 50 mL Erlenmeyer flask and is extracted and sonicated as described above, except A x. . RRF� � for using only 20 mL of 96% ethanol–water 3:1. After sonication, a magnetic stir bar is added x � � ∙ to the Erlenmeyer flask. This is followed by addition of 5.00 mL of a methanolic solution of in which Ax is the peak area of compound x; Ai.s. is the peak area of i.s.; wi.s. is the weight of the i.s. of known concentration using a volumetric pipette (see Appendix D for the possibility i.s.; RRF is the relative response factor; and wx is the weight of compound x. Because the peak of replacing methanol by 96% alcohol), and 10 min-stirring. After filtration of the solution via areas are obtained from the chromatogram, the quantity (weight) of added i.s. is known, and the a syringe filter, the sample is ready for HPLC analysis. RRFs are given, students can calculate the quantity of compound x (ldg, lmg, or lut) in the The students proceed by soaking four ready-made ~5 × 5 cm pieces of wool (pretreated with sample. Then, the concentration of compound x in the weld sample can be calculated. aluminum potassium sulfate dodecahydrate, alum) in water at 50 °C. The dyeing bath is heated In addition, mimicking the work of chemists investigating historical textiles, the students to 80 °C, and the four pieces of wool are added to it. The dyeing step lasts 15 min, after which identify the dye source of the freshly dyed piece of wool. This is done by assigning the peaks the dyed wool is rinsed with water. A small piece of a single thread is removed from the dyed corresponding to ldg, lmg, and lut using chromatograms and UV–vis absorption spectra of wool and the dye is extracted by placing it in 300 µL of methanol–water–methanoic acid authentic standards. They also compare the chromatogram of the dyed wool’s extract sample (formic acid) 80:15:5 in a microcentrifuge tube of 2 mL, and heating via a water bath at 60 °C with that of the weld’s extract sample. for 30 min. Sample is ready for HPLC analysis after filtration of the cooled solution via a The work of the students is concluded by answering a few questions that include separation syringe filter. The samples from both parts of the experiment are given to one of the teaching principles of RP-HPLC analysis and the relation between structure and UV (345 nm) assistants so that they can be analysed. This is done by reversed phase (RP)-HPLC–UV at 345 absorbance of weld’s three main flavonoids. Solutions to the assignments and answers to the nm in either 80 min at 40 °C (Alltima 250 × 4.6 mm C18 5 µm-particle size HPLC column) or questions of the protocol handed out to students, including derivation of the equation above, 5 min at 35 °C [Eclipse XDB-C18 50 × 3.0 mm 1.8 µm-particle size ultrahigh-pressure liquid are available in Appendix D. chromatography (UHPLC) column]. In both cases columns are mounted in conventional HPLC 5.3. Hazards systems. The gradient elution separation is carried out with an aqueous buffer pH 3 and either Basic organic chemistry laboratory safety procedures [5] should be observed when carrying out methanol (HPLC column) or methyl cyanide (acetonitrile) (UHPLC column). The the experiment. Erlenmeyer flasks should not be clamped during sonication, as they may break. chromatograms with the data needed are given to the students for data processing. Although There is the risk of implosion (glass breakage) during operation of rotatory evaporators. 96% this procedure was typically followed, involvement of the students in running the analyses alcohol is flammable [6]. Exposure to methanol (flammable, toxic) and formic acid (corrosive) would be preferable; e.g., by accompanying the teaching assistant in preparing and starting the must be minimized [6]. Sample filtration via a syringe filter has the risk of liquid spillage due sequence of analyses and receiving a brief explanation of HPLC. Even though all students to an improper connection between both parts. The microcentrifuge tube containing the participate in a demonstration on HPLC at the beginning of the chromatography part of AMOC, methanol–water–formic acid extract should be cooled prior to opening because of the pressure the additional exposure to the instrument could be beneficial to the students’ learning. In build-up during the heating step. The support team should pay special attention: Appendix D, different parts of the protocol are discussed and material and methods for the • during the preparation of the methanol–water–formic acid solution and HPLC solvent A preparation of the experiment are detailed. (aqueous buffer pH 3) because of handling of formic acid, when the use of protective gloves is recommended [7]; • so that vapours originating from HPLC solvent B (methanol or acetonitrile, toxic) bottle and HPLC waste container are properly exhausted [6].

72 73

detailed. are experiment the of preparation D, differentAppendix parts of the protocol are discus learning.be beneficial the to students’ thecouldto instrument the additionalexposure participate on HPLC a in demonstration at the beginning of the chromatography part of AMOC, systems. chromatography removed sequence of analyses and receiving a brief explanation of HPLC. of explanation brief a receiving and analyses of sequence would be preferable; by e.g., accompanying the teaching assistant in preparing and starting the th Although processing. data for the students given to are data needed the chromatograms with methanol (HPLC column) or methyl cyanide (acetonitrile)(UHPLC column). The 5 min 72 innm either 80 min assistants for 30min. at 60°C bath (formic a in a water microcentrifuge 80:15:5 2mL, acid) tube of and heating via dye wool andis the the dyed rinsed water.wool is with water. extract filtration 30 mL For wooldyeing the for mainflavonoids. three the other quantifying usedtheusedwool and plant’s is for dyeing The carry students 5.2.2. Practical p sy 80°C to aluminum a syringe filter, the for using only 20mL96% ethanol of except above, described as sonicated and extracted is and flask Erlenmeyer mL a50 in placed ing replac of the theto Erlenmeyer followed is This by flask. of 5.00mL addition of a methanolic is procedureis typically was of running the in followed,students involvement the analyses ringe filter The student i.s.

at 35 °C [ at 35°C

To quantify the plant’s three main flavonoids, 200 mg of the dried and ground weld is is ground weld and dried of 200mg the flavonoids, main three plant’s To quantify the of known concentration using a from from of , and the four pieces woolareof added The it. to dyeing after 15min, step lasts which

by a a by

The gradient elution separation is carried out with an aqueous buffer pH 3 and either aneither separation with aqueous out gradient 3and carried bufferThe is elution pH through a folded filter paper through a folded

so thatso the 96% otassium sulfate dodecahydrate, a dodecahydrate, sulfate potassium

methanol by 96% alcohol) S the .

ample is ready for HPLC analysis HPLC ready for is ample rotatory ev rotatory teaching the teaching one of given to are experiment the of parts both from samples The s

ethanol

proceed by soaking soaking by proceed

Eclipse Eclipse round- (UHPLC) column] (UHPLC) art

sample is ready for HPLC analysis HPLC for ready is sample extractions of dried of extractions two out at 40 ( °C , y be can

1.5 extracted by placing300 itin bottom flaskbottom to a 150mL – XDB water 3:1 water aporator. g

of the dried and ground weld is placed in an Erlenmeyer flask with flask with Erlenmeyer an in weld placed is ground and dried the of

analysed - Alltima 250×Alltima 4.6mm 5µm C18 C18 50 × 3.0 mm 1.8 µm mm × 3.0 50 C18 , and stoppered. dyeing bath is prepared is bytransf bath dyeing The . A four four In are casescolumns both mountedin conventional HPLC water 3:1. –water

. removed from removed the dyed is thread asingle of piece small to a round to , and 10min-

This is done is This volumetric pipette ready

lum -

and and made made

beaker beaker - stir bar stir added is amagnetic sonication, After )

bottom flask

after filtration of the cooled solution via a in water at 50 °C. The dyeing bath is heated heated is bath dyeing The °C. 50 at water in The mixture is sonicated is The mixture for 10min ground

by by stirring ~5 × 5 cm pieces of wool (pretreated of pieces with cm ×5 ~5 . µ

reversed phase (RP) phase reversed sed and material and for methods the L of methanol L of - using particle size

weld (see (see . After filtration of the solution via

4 , - Appendix D Appendix

particle size particle the the × simultaneously:

deionized of deionized portions 15 mL ~

Even th Even

20 mL of 96% alcohol is 96% of 20 mL ultrahigh –water errin - HPLC column)

HPLC for the possibility g the g the ough all students oughall students methanoic acid acid –methanoic - pressure liquid liquid pressure One extract is is extract One alcohol –UV solution ofsolution

. at 345 A - free fter fter or or In In

corresponding ldg to w equation: concentration of the three m the calculate students the (RRFs), factors response relative predetermined and a sample retention times. identif concen the Then, sample. compound of quantity calculategiven, can the areRRFs students areas are obtainedchromatogram, from the the(weight) quantity of i.s. added is the RRF ; i.s. in student the times, and retention areas containing peak the chromatogram sample of the weld With 5.2.3. Data p principles of RP authentic standards. absorbance with that of the weld build- methanol between an to The improper containing microcentrifuge parts. the connection both tube must be minimized [ alcohol is flammable [ 96% evaporators. rotatory of operation during (glass breakage) implosion riskof the There is Erlenmeyer experiment. the Basic 5.3. Hazards D. Appendix in available are above, equationincluding of the handed derivation of questions students, the to protocol out x wh In a In • Th • = ich e the workstudent of waste container are properly exhausted properly are container waste and HPLC so so recommended is gloves ( during

The s up during step. The the heating organic chemistry laboratory safety procedures safety laboratory chemistry organic freshly dyed freshly the of source dye y the aqueous A ddition, mimicking the workchemists of investigating historical textiles, the students s A start �

that x .

–water � A . ∙

x of of

RRF w is the the the by by vapours rocessing

� weld’s mainflavonoids three relative response factor; and w and factor; response relative . �

buffer buffer – Then, using the equation that relates peak area to the quantity of compound in area peak of the to quantity equation thatrelates Then, the using . the the of preparation

assign formic acid extract should be cooled prio cooled be should extract acid formic -

the the and analysis HPLC peak areapeak of compound ’s extract ’s dyed dyed the of chromatogram the compare also They 6

, HPLC originating from HPLC ] 6] pH 3 pH . lmg ing

Sample filtrationa via syringe filter has risk the liquid of spillage due .

tration of compound x Exposure to methanol to (flammable,Exposure and formic toxic) aci the peaks corresponding ldg theto peaks s , and lut

) ain flavonoids in the weld sample. flasks should not beflasks not clamped should during as they sonication, may break. a few questions concludedfew is byansweringa

[ because of because sample. 7] ;

methanol

using chromatograms

upport team should pay should upport team

handling acid, of formic whenof protective the use

x piece of wool. piece . –water ; A relation betweenrelation UV and structure Solutions to the to assignments Solutions and answers the to solvent (methanol B or acetonitrile, toxic) bottle x is in the weld sample can be calculated. be can sample weld the in i.s. the

is the formic acid –formic

weight of compound x [ 5 ]

[

; w i.s. of area peak should be observed 6] , lmg r to opening because of the pressure the becausepressure of r opening to This is done by assigning the peaks assigning done the is bypeaks This and UV .

special This isThis done using , lut

solution and HPLC solvent A A andsolvent HPLC solution onthe bas , and i.s. vis absorption spectra–vis of x

a ( ldg that include separation thatinclude ttention wool , lmg Because the peak peak the . Because i.s. when is known, andis the ’s

is the

: , or lut or ,

extract extract

the following the following d (corrosive) d (corrosive) carrying out out carrying

(345 nm) weight of

is of the is of the ) in the the in ) sample 73

Chapter 5 5.4. Results and discussion difficulty with it. Additional observations on the students’ learning regarding the didactic aims In 2011, 12 students carried out the experiment. Their samples were analysed using a traditional and background of the experiment are described in Appendix D. 25 cm-long, 5 µm-particle size RP-HPLC column. Five and four students carried out the experiment in 2012 and 2013, respectively.1 Their samples were analysed using a modern 5 cm-long, 1.8 µm-particle size RP-UHPLC column mounted in a conventional HPLC system. 5.5. Summary Results obtained by a student in 2011 and another student in 2012 are depicted in Figure 5.2. Students are exposed to analysis by HPLC and quantitative analysis using the internal standard The peaks of the main compounds in the chromatograms are assigned. The compounds method by carrying out this experiment. This is accomplished using the real-world application responsible for the minor peaks also belong to the class of flavonoids. Their identities are listed of natural dyes for textiles, which is a current topic in chemistry. in Appendix D. The signal-to-noise ratios of the chromatograms of the dyed wool’s extract samples are much smaller than those of the chromatograms of the weld’s extract samples because only small pieces of single threads of dyed wool are analysed, resulting in small 5.6. Supplementary material quantities of flavonoids in the samples. This mimics of the work of chemists investigating Appendix D contains information supplementary to that in this chapter. The following material historical textiles, in which it is crucial to minimize damage to artifacts. is available: • Updated version of the protocol of the experiment handed out to students; • Discussion of different parts of the protocol; • Material and methods for the preparation of the experiment; • Instructor notes; • Additional observations on the students’ learning based on 2011 and 2012 reports; • Updated version of the inventory of specialized material handled by students; • Example data sets of the HPLC analysis of weld sample.

5.7. References [1] Cardon D. Natural dyes – sources, tradition, technology and science. 1st ed. Archetype Publications: London; 2007. [2] Cristea D, Bareau I, Vilarem G. Identification and quantitative HPLC analysis of the main flavonoids present in weld (Reseda luteola L.). Dyes Pigments 2003; 57(3): 267–72. [3] Marques R, Sousa MM, Oliveira MC, Melo MJ. Characterization of weld (Reseda luteola L.) and spurge flax (Daphne gnidium L.) by high-performance liquid chromatography–diode array detection–mass spectrometry in Arraiolos historical textiles. J Chromatogr A 2009; 1216(9): 1395–402. [4] Villela A, van der Klift EJC, Mattheussens ESGM, Derksen GCH, Zuilhof H, van Beek TA. Figure 5.2. Chromatograms of the extracts of weld and dyed wool prepared by two students. Fast chromatographic separation for the quantitation of the main flavone dyes in Reseda luteola All cases: 345 nm traces. Peaks of the main compounds are assigned. (weld). J Chromatogr A 2011; 1218(47): 8544–50.

[5] Vogel AI. Vogel's textbook of practical organic chemistry. 5th ed [revised by Furniss BS,

Hannaford AJ, Smith PWG, Tatchell AR]. Longman Scientific & Technical: Essex; 1989. In 2011, 10 students submitted a report of the experiment. In 2012, this was done by the five students. On the basis of those reports, it was observed that nearly all students understood how [6] O’Neil MJ, Smith A, Heckelman PE, Obenchain Jr. JR, Gallipeau JAR, D’Arecca MA, separation by reversed phase chromatography works. Furthermore, half of the students knew Budavari S, editors. The merck index – an encyclopedia of chemicals, drugs and biologicals. how to use the internal standard method for quantitative analysis, and one-third of them had 13th ed. Merck and Co.: Whitehouse Station; 2001. [7] Safety data sheet of formic acid – revision 08-Jun-2012, Acros Organics: http://www.acros.com; 2012 [accessed August 2013]. 1 In addition to the 21 students who carried out the experiment in 2011–2013, the experiment was also carried out by 10 students in 2014.

74 75

All cases: 345 nm traces. Peaks ofthe main compounds experiment in 2012 and 2013, respectively. 2012and in experiment historical textiles, in which itis crucial to minimize damage to artifacts. ofquantities flavonoidsthe in samples. resultingdyed smallwool arein analysed, ofbecause pieces threads single only small extract weld’s the of chromatograms the of those than smaller much are samples 25 cm in Appendix D. Appendix in responsible The the compounds in chromatograms arecompoundsTheassigned. peaks main of the Figure 5.2. in 2012are depicted in student another 2011and in student obtained bya Results 74 1 how use to the internal standard separation by phase reversed chromatography Furthermore, works. the knew halfstudents of students In a report submitted 2011,10students of the experiment Figure cm In 2011,12 discussion 5.4. Results and carried out by 10students in 2014. In theto 21 addition students who carried out the experiment 2011– in - long, 1.8 - 5.2. long, 5 . On the basisof those reports, was it observed that nearly understoodall students how

Chromatograms of the extracts of weld and dyed wool prepared by stude for the peaks minor also belong the to class of flavonoids. Their are identities listed µ m µ The signal The - experiment. the out carried nts m particle size RP - particle size RP -

to - noise ratios of the chromatogramsnoise ratios of the wool’s of extract the dyed

method for quantitative analysis, and one - UHPLC aHPLCcolumn mountedin conventional system. - HPLC column.Five and four carried students the out This mimics of the work of chemists investigating 1 Their samples were analysed us analysed were samples Their

Their samples were analysed were samples Their are assigned. . In. was 2012,this done by the five

2013, the experiment2013, the

two students -

third ofthird themhad using a traditional ausing traditional ing a modern 5 ing modern a

samples samples .

was also is available chapter. D this contains thatin informationto Appendix supplementary material Supplementary 5.6. of natural dyesfor textiles, which is a current topic in chemistry experiment this out method bycarrying areStudents analysis to exposed by analysis and HPLC quantitative internal using standard the 5.5. Summary and background of the experiment difficulty Additional it. with observations onthe students’ learning regarding the didactic aims [accessedhttp://www.acros.com; 2012 2013]. August – acid formic of sheet data Safety [7] 2001. ed. Station; 13th Whitehouse Merckand Co.: index merck Budavari editors. The S, MA, D’Arecca JAR, Gallipeau JR, Jr. SmithA, Heckelman Obenchain O’Neil PE, MJ, [6] Hann of practical Vogel[5] chemistry. organic textbook AI.byFurniss Vogel's edBS, 5th [revised Chromatogr(weld). 1218(47):2011; 8544–50. J A Reseda in luteola of the dyes mainflavone theFast quantitation separationfor chromatographic Zuilhof H, Mattheussens Derksen GCH, ESGM, Villela EJC, Klift A,[4] van der 1216(9): 1395–402. detection array L.) and spurge f ( weld of Characterization MJ. Oliveira Melo MarquesMC, MM, Sousa R, [3] flavonoids present ( weld in I, Bareau D, Cristea [2] London;Publications: 2007. Cardon[1] D. Natural dyes – References 5.7. • • • • • • • aford AJ, Smith Longman 1989. Smith PWG, Tatchell Scientific Essex; &aford AJ, AR]. Technical: Updated the protocol version of Discussion of Discussion Example data sets of the HPLC analysis of weld sample. weld of analysis HPLC the of sets data Example observationsAdditional onthebased students’ learning on2011 and 2012 Instructor notes the preparationMaterial of the andexperiment; for methods Updated the i version of :

–mass spectrometryArraiolos in historical textiles.Chromatogr J A 2009;

lax ( lax different parts of differentthe protocol parts Daphne gnidium ;

VilaremIdentification G. and quantitative HPLCanalysis of main the Reseda luteola nventory

sources, tradition, technology tradition, sources, ed.1st Archetype and science. are described in Appendix D. Appendix are described in of the experiment the of

– L.) by high- of specialized of materialhandled bystudent . This is accomplished using the real the using accomplished is . This an encyclopedia of chemicals, drugs and biologicals. L.). 267–72. Pigments 57(3): 2003; Dyes ;

revision 08- performance liquid chromatography performance liquid

handed out

. Jun-

to students to 2012, Acros Organics: The The following following - world application world application ; Reseda luteola Reseda

reports van Beek TA. TA. Beek van s ;

material ; –diode –diode

Chapter 5

Chapter 6

General discussion

Chapter 6

General discussion

6.1. Introduction The work reported in this thesis describes concepts and analytical chemistry methods relevant for the use of natural dyes for textiles. The contents of chapters 2, 3, and 4 refer to different aspects of textile dyeing technology. They all relate to the use of the dyestuff of weld (Reseda luteola L.). The analytical method for the analysis of the main colouring compounds of the plant—described in chapter 2—was applied to research (chapter 4) and education in chemistry (chapter 5). In this chapter, the works presented in chapters 2–5 are summarised, elements of what they report are discussed, and future prospects of dyeing textiles with natural dyes are briefly elaborated upon.

6.2. Chapter 2 A validated RP-HPLC–UV method for the quantitation of ldg, lmg, and lut via internal standardisation in samples of weld is described in chapter 2. The method is simple, requiring very little manpower input per sample. With the exception of the 15-stirring point magnetic stirrer—that, less conveniently, could be replaced by individual magnetic stirrers—only standard laboratory equipment is needed. The analytical method proved fit-for-purpose based on its validation in terms of extraction efficiency, stability and losses of the compounds, accuracy, precision, and sensitivity. As the method was developed using only one batch of weld (R. luteola H) plant material, it has the potential drawback of not being suitable for analysing samples of weld having a concentration of the three flavones much lower or much higher than that of R. luteola H. The industrial partner of the project analysed 120+ batches of weld plant material. No samples were encountered with such low concentrations of ldg, lmg, and lut that they could not be analysed using the method as described in chapter 2 (that is, samples leading to injected quantities smaller than the smallest injected quantities of the calibration curves). However, there were samples having concentrations of the flavones that were too high. For the analyses of such samples, the industrial partner of the project modified the method by simply weighing 100 mg—instead of 200 mg—of sample. Such an adaptation of the method is expected to lead to: • no major change in extraction efficiency even with the solvent–sample ratio increasing to 200, as it is already high (95%) with a solvent–sample ratio of 100; • no change in accuracy; neither in terms of stability of the flavones or losses due to adsorption on glass and filters (the relative recovery experiments were also carried out using 100 mg of plant material), or in absolute terms (as there is even less plant material to which the flavones could bind); • no major loss of precision as, if a 0.1 mg-least division balance is used, the weighing error remains <1%. The validity of this approach would need testing. This could be done, for example, by comparing the results of the analyses of a small number of replicates of 100 and 200 mg of a few samples. After the publication of a validated RP-HPLC–DAD method for the quantitation of ldg, lmg, apigenin-7-O-glucoside, luteolin-4'-O-glucoside, lut and apigenin in samples of weld by Gaspar et al. [1], the chromatographic separation using a 5 µm-particle size RP-HPLC column was speeded up by using a short sub-2 µm-particle size (RP-UHPLC) column. This

78 79

78 standard laboratory equipment is needed. is equipment standard laboratory conveniently,—that, less could be replacedstirrer by magnetic individual stirrers very little standardisation RP validated A 2 6.2. Chapter elaborated briefly upon. are dyes natural discussed are report they what of elements has the potentialdrawback of being not suita on its chemistry ( plant concentration of the three flavones much lowermuch higher or than thatof R. luteola precision accuracy, there wer there partnerindustrial s project of analysed the batche 120+ luteola column was speeded up bycolumn was a speeded upusing short Gaspar samples. few a and of 200mg 100 of replicates a number of small of s analyse the of results the comparing The validity of this to: 100 mg of such samples, the industrial partner of the project modified the methodby simplyweighing quantities smaller than the smallest injected quantities of the calibration curves as described the method analysed using with encountered were technology. dyeing textile of aspects for lmg describes thesis Thereported work this in 6.1. Introduction

As the method was developed using only one batch usingAs the methodwasone only developed After publication the a validated of RP • • • the use of natural dyes for textiles. The content textiles. for dyes natural of use the , apigenin- chapter 2 chapter in —described error error no ); bind whichto thecould flavones material using 100mg of plant adsorption no change accuracy; neither in terms in of stability the flavones of or with (95%) high 200,asto already itis extraction in change major no

val —instead of 200mg L.) et al. major e

idation manpower input . samples having concentration having samples remain chapter 5 chapter T

he analytical method for the analysis of the method main colouring compoundsof analytical he

[1] is is of weld samples in loss - O HPLC on glass and filters in efficiency terms of extraction , usingthe chromatographic a separation 5µm s - , and <1%. glucoside, luteolin of precision as,of precision

). approach approach

– In this chapter,In this the such sensitivity UV method for the quantitation of ldg quantitation UV the methodfor

—of sample . sample per

low concentrations w was applied to research ( research to applied —was ould ould .

), or in absolute terms ( ), or absolute in terms efficiency even with the with even efficiency - a0.1mg if Theyall relate to (the relative recovery experiments were also carried out out carried also were experiments recovery relative (the described - 4' testing. need

. - With the exception ofexception the 15- the With s O S

The The in in methods methods chemistry analytical and concepts sub- of the flavones

uch an adaptati an uch - - works works glucoside, HPLC asolvent chapter 2 chapter , and future of, and prospects dyeing with textiles analytical method proved fit methodproved analytical 2 µm

in in analysing analysing for ble

least division balance the weighingdivision used, least is of of presented chapter 2 chapter DAD –DAD s , -

ldg of of particle size (RP size particle stability and losses of thestability compounds, and losses –

This couldforThis be by done, example, lut the use of the dyestuff of weld of the dyestuff of the use of weld of sample ratio of 100;

c , ( hapters 2 hapters lmg weld of

samples leading to leading that samples is, on of the methodis method for the quantitation of ldg quantitation the method for . high too were that and apigenin in samples of weld by by ofand weld apigenin samples in

as there is there as in in . , and The method is simple, requiring requiring simple, methodis The solvent

chapter 4 chapter ( chapters 2 chapters R. luteolaR.

, plant material . samples of weld of samples , lut lmg , and 4 3, and - sample ratio increasing increasing ratio –sample

particle size RP size particle -

even less plant material material plant less even UHPLC) UHPLC) stirring point magnetic that that , and , and

) and education in in education ) and H) –5 -

they they - for

refer are are plant material , it For t For expected to expected lut

purpose base purpose loss column.

could not be could be not summarised, ) N

via internal he analyses analyses he . However, However, . to different different to o samples samples o

es due to to due es having a

( relevant relevant injected Reseda Reseda - H —only —only HPLC HPLC .

This This lead The The the the 79 d ,

Chapter 6 was done while still using a conventional HPLC system, after minor hardware adaptation. The would only be suggestive. This is because the chroma of the greenish hue could be due to method using the UHPLC column proved fit-for-purpose through comparison of accuracy and something else. For example, it could be correlated to the total content of flavonoids. precision of the quantitation of ldg, lmg, and lut in R. luteola H with those achieved by using the HPLC column. Large gains in analysis time and eluent consumption were obtained. Thus, the use of short UHPLC columns on conventional HPLC systems is an economical way of 6.4. Chapter 4 modernising HPLC-based analyses. The effect of different concentrations of aluminium ion on the photo-stability of the dye of weld, and the relative photo-stability of lut, lmg and ldg in solution are reported in chapter 4. Results of the colours obtained by dyeing alum-mordanted wool with the three flavones are 6.3. Chapter 3 also presented. An analytical method for the comparison of the content of porphyrin ring-containing The observed order of photo-stability of the flavones in solution was ldg >> lut > lmg. pigments—chlorophylls (chls) and their structurally similar breakdown products—in different However, alum-mordanted wool dyed with ldg (or with an extract of weld) was much less samples of weld is described in chapter 3. The method: colourful than that dyed with either lut or lmg. At least in part, this could be due to a reduced • uses ethanol as a mild, non-toxic extraction solvent (that is not selective regarding the uptake of ldg by Al3+-mordanted wool relative to that of lut or lmg. These observations “types” of chls and plant tissues being extracted); should support answering the question whether it is preferable for the endogenous glycosidase • covers the entire range of occurring concentrations; of weld to be inactivated before extraction of the flavonoids in order to obtain the most photo- • displays acceptable precision; stable dye, but are not conclusive. A dye containing flavonoid glycosides and aglycones is • is simple. obtained if this glycosidase is inactivated before extracting the flavonoids of the plant. Conversely, a dye containing only the aglycones is obtained if the enzyme is not inactivated Nevertheless, the manpower input required by the method is not low. One working day was before the extraction process. Comparison of the light-fastness of the colour of Al3+- needed for the analyses and data processing of 40 samples. It is expected, however, that more mordanted wool dyed with a glycosidase-hydrolysed extract of weld with that of Al3+- samples can be analysed in the same time through implementation of the following: mordanted wool dyed with an extract in which the glycosidase has been inactivated before • weighing accurately ~125 mg instead of (125.0 ± 0.9) mg, with a correction factor (125 extracting the flavonoids is expected to be decisive. divided by amount weighed) being used in the calculations. This would speed up the The light-fastness of the colour of weld-dyed wool decreased with increasing [Al3+] used weighing of the samples; for premordanting the wool. As the gain in light-fastness by the use of low [Al3+] was limited, • having 60 centrifuge tubes available instead of 20. This would speed up the analysis of this cannot be a way to meet today’s requirement of light-fastness of the colours of dyed the samples, as no tubes would need to be washed during the data-acquisition day. textiles by itself. In chapter 4, reference is made to a broader strategy of which using low [Al3+] for premordanting the wool is but one part in order to address the need for increased Footnote 1 of chapter 3 relates to a limitation of the method. Comparison of the content of light-fastness of the colour of wool dyed with weld. Such a strategy—largely based on porphyrin ring-containing pigments in samples of weld having different chls/chlorophyllides- information in different publications [5-8]—could also include: to-“pheopigments” ratios is hampered, as the optical molar absorption coefficients of • Using slightly more dyestuff than one would normally use; pheophytins and pheophorbides are ~40% smaller than those of the corresponding chls and • Using the preferable approach regarding the use of a dyestuff that is flavonoid chlorophyllides [2, 3]. Possibly, this can be fixed by conversion of chls and chlorophyllides to glycoside-rich or flavonoid glycoside-poor. This relates to the optional inactivation of the corresponding “pheopigments” through addition of one drop of a 25% aqueous HCl the endogenous glycosidase before extracting the flavonoids of weld; solution to 5 mL of an extract of the pigments, as outlined by Lichtenthaler [4]. The • Adding antioxidants—such as gallic acid—to the dyed textile, preferably, with the conversion is expected to be rather fast, as chls are generally fully converted to pheophytins antioxidants being obtained from agricultural/food processing residues (as an example, in ≤1 min [4]. Thus, one could test whether the same procedure described in chapter 3 can be gallic acid and many other polyhydroxyphenols can be obtained from pistachio green used with the extraction solvent being HCl-acidified ethanol instead of absolute ethanol. hull).1 One could test the usefulness of the method reported in chapter 3 for the intended purposes and the hypothesis that chlorophylls a and b are the sources of the greenish hue that 1 The antioxidants could also be generated through the photodecomposition of part of the components accompanies the yellow colour after dyeing alum pre-treated textiles with weld. This could be of the dyestuff. For example, Al3+-mordanted wool could be dyed with both outer scales of onions done by seeing whether there is a relation between the optical absorbance values—due to chls (onion skins) and weld. This could lead to a characteristic change in the yellow colour over time as, 3+ and their structurally similar breakdown products—of extracts of samples of weld and the upon exposition to light, textiles dyed with onion skins appear to fade [9, 10] and Al -mordanted wool dyed with weld darkens and becomes redder [11] (Fig. 6). Lightening of the fabric would be chroma (saturation) of the greenish hue of alum-mordanted wool dyed with the same plant expected to precede its darkening/reddening since the flavonol quercetin is the main flavonoid materials. However, although many samples would be needed, the presence of such a relation aglycone of onion skins {[10, 12] and data not shown} and quercetin appears to be less stable upon exposition to light than luteolin, the main flavonoid aglycone of weld [9]. The photo-oxidative

80 81

materials. chroma (saturation) of the (saturation) ofchroma the and their structurally breakdown similar products between relation a is there whether seeing by done p alum dyeing after colour yellow the accompanies in purposes used conversion solution solution a of one drop addition the correspondingthrough “pheopigments” 80 chlorophyllides samples 3 of weld is describedchapter in pigments done was An analytical methodfor the the comparison content of 3 6.3. Chapter modernising HPLC the use of shortUHPLC on columns pheophytins than are~40% smaller and pheophorbides to the HPLC column. precision method using the UHPLC porphyrin ring Nevertheless F can samples needed the for ootnote 1 - ≤ “pheopigments” ratios is hampered is ratios “pheopigments” One could t test • • • • • • 1 min [4] 1 min with the being solvent extraction HCl displays acceptable precision acceptable displays ran entire the covers of “types” amil as ethanol uses divided by is simple. having samples; the of weighing mg ~125 accurately weighing the samples, as no tubes would needthe samples, asbe to notubes washed

—chlorophylls (

while still using to 5 mL of an and the hypothesis thatchlorophyllsa the hypothesis and

of the quantitation of ldg the quantitation of However, However, is expected to be rath be to expected is of of be analysed in the same time through time same the in analysed be , . 60 the the

- T chapter 3 chapter analyses data and processingsamples. of 40

containing pigments in containing pigments

hus, hus, [2, 3] [2, cen chl amount weighed) be required by the method the by required manpower input - L s and plant tissues being extracted being tissues plant and s t based analyses. based he he rifuge t rifuge one although eluent eluent and time analysis in gains arge . Possibly, t usefulness usefulness

could chl relates to a to relates ge of occurring occurring ge of a conventional system a HPLC column proved fit extract d, non d, greenish hue of alum greenish of hue ubes s) and breakdown similar products their structurally m

test any samples would be needed be would samples any

his his available instead of 20. This wouldspeed This available upthe analysis of 20. instead of er fast, as chl as fast, er -

toxic extraction solvent (

; , of the pigments chapter 3 chapter in described procedure same the whether

of the method lmg instead of instead can

limitation of the method.

ing ing conventional samples of weld having different chl

, and , as the optical, as the be fixed be . T concentrations; - the calculationsused the in acidified ethanolacidified ethanol instead of absolute - he method: for lut

(125.0 ± 0.9) mg,(125.0 ± s ares generally fully converted to pheophytins - and and

purpos in in by ofconversion chl - implementation following: of the plant plant e same th with wool dyed mordanted

of extracts of samples of weld of samples of extracts —of re the the

R. luteola , as outlined byLichtenthaler [4] chapter 3 reported chapter in HPLC systems is an economical way of of way economical an is systems HPLC b - ); treated textiles with weld. s source the are optical optical

, during

after minor hardware adaptation. hardware minor after

those of through c molar absorption coefficients of consumption consumption It expected is is not low. not is that is not selective regarding the ,

absorbance values absorbance H - data the

presence the of porphyrin ring C with th with

ompari the corresponding chl . accuracy and and of accuracy omparison This would speed up the speed upthe would This

s ands chlorophyllides to of the greenish hue that of the a

O ose acquisition correction factor ( factor correction son son were obtained. were , however, that more ne working day ne 25%

achieved achieved s/chlorophyllides of such a relation arelation such of for the intended the intended for of of

the content of This couldThis be in different different —in aqueous aqueous due to chl —due to - day.

containing containing

by using by using and the

. can

. Thus, Thus, s

HCl HCl The The was was The The 125 and and

be be s - exposition to light than luteolin, aglycone skins { onion of precededarkening/reddeningmain its to expected quercetin flavonoid flavonol is the since the information different in [5 publications [Al textiles by itself. of light requirement meet today’s cannotato way this be glycosidase awith mordanted wooldyed stable dye, conclusive are not but would 1 light- gain As the light in wool. for the premordanting extracting is the flavonoids before inactivated been glycosidase has which the an extract in with dyedmordanted wool process. extraction the before Conversely of the plant. flavonoids the extracting glycosidase before inactivated is obtained if this weld of supportanswer should weld The 4 6.4. Chapter something else. wool dye R also However, alum uptake of ldg eitherwith lut dyedcolourful than that upon exposition to textiles light, dyed onion with skins appear fade [9, to (onion skins) and weld. This could lead a characteristic to chang of the dyestuff. For example, Al The The obtained by dyeing by colours obtained ofesults the • • • l The - photo of The observedorder 3+ presented effect , ] for one but the premordanting woolis ] fastness of thewooldyed colour weld. with antioxidants could also generated be through the photodecomposition of part Using slightly more Using glycoside Using the preferableuse approach of flavonoid regardinga is the dyestuff that hull ga beingagricultural/ from antioxidants obtained residues processing food Adding antioxidants endogenous the and the relative- photo only be

to llic acid ight d with weld darkens and becomes redder )

. be inactivated before extracti before be inactivated 1 of , a dye containing only the aglycones is obtained if the enzyme is not inactivated inactivated not is obtainedis aglycones if the enzyme containingthe , aonly dye

- fastness colour of the weld- . different different

by Al by -

- suggestive. For example, it could be it correlatedForflavonoids. thecontent to total example, of rich glycoside or flavonoid

mordanted wooldyedmordanted w

polyhydroxyphenols and polyhydroxyphenols many other In In chapter 4 chapter 3+ before extracting the flavonoids of weld of flavonoids the extracting before glycosidase ing - concentrations of aluminium ion onthe- ion photo concentrations of aluminium mordanted that to woolrelative of lut [10,

the the for preferable the whether question is it dye such as —such

expected to be decisive. be to expected This is because the chroma of the the of chroma the because is This stability of lut 12] 12] stuff than one wouldnormallystuff use; 3+ broader a broader to made is reference ,

- and dataand not shown} the mainthe flavonoid aglycone weld of mordanted wool could be dyed both with outer scales onionsof Comparison of the light of Comparison stability of the flavones in solution was ldg was solution in flavones the stability of . Aglycosides dye andcontaining agl flavonoid

gallic acid gallic -

or or on 8] ith —could lmg

, of of alum - -

lmg dyed wooldecreased increasing with [Al poor hydrolysed extract of weld with thatofwith Al extract of weld hydrolysed ldg ldg the flavonoids in order in - photo the obtain flavonoids to most

. part in order to address the need for increased need increased thefor address orderto in part At leastAt in part, this could be -

- [11] and and fastness by the use of low [Al low of bythe use fastness —to dyed the textile . T ( mordanted wool an extract o extract or an with also include

his relates to the optional and quercetin appears to stable be less upon ldg (Fig. Lightening 6). of the fabric would be can be pistachio green obtained from

chapter 4 chapter in solutionin reported are Such astrategy Such - fastnes e in the yellow the e in colour over time as, - fastness of the co the of fastness :

greenish hue greenish

strategy of which using low low using which of strategy or or with the flavones with three s of the colour of Al of colour the of s lmg endogenous [9] f weld) , 10] 10] stability of the dyeof . . preferably, largely based on on based —largely ; These observations

The photo- The and Aland

of the components theof

due to a reduced reduced a due to could be due to to be due could (a

3+ was much less less much was inact >> >> s an example, example, an s lours of dyed dyed of lours ] was] limited, 3+ glycosidase lut -

ivation mordanted ycones is with the with the oxidative 3+

> ] u lmg

sed sed are

3+ 3+ 81 of of - - . .

Chapter 6 The use of such a strategy limits the range of colours that can be obtained upon dyeing textiles lmg of the chromatogram of the extract of the dyed wool relative to that of the extract of weld with weld. However, each of its elements could lead to a small gain in light-fastness of the are expected to be due to the following reasons: colour. Combined, such gains might be a way of meeting today’s requirement of the property. • lower uptake of the diglycosides relative to the monoglycosides and of the monoglycosides relative to the aglycones by the Al3+-mordanted wool (this is in agreement with the data discussed in section 4.4.2.2); 6.5. Chapter 5 • worse extraction of the diglycosides relative to the monoglycosides and of the Chapter 5 reports a work on education in chemistry at undergraduate level. Students of monoglycosides relative to the aglycones from the Al3+-mordanted wool {the latter is in molecular life sciences and biotechnology of Wageningen University had the opportunity to line with results published by Willemen et al. [11] (left panel of Fig. 5 and preceding deepen their knowledge on HPLC and quantitative analysis using the internal standard discussion)}. method while working on a “natural dyes for textiles”-themed experiment. Their assignment As the difference between the chromatograms in terms of relative quantities of the flavonoids was to quantify ldg, lmg, and lut in a sample of weld, and to dye alum-mordanted wool with is small, the reported procedure is suitable for the intended purposes. It may also be used for an extract of the same sample. Based on 15 reports by students, it was observed that nearly all research purposes, as done—in an adapted form—by Willemen et al. [11]. of them understood how separation by reversed phase chromatography works. Furthermore, half of the students knew how to use the internal standard method for quantitative analysis, and one-third of them had difficulty with it. 6.6. Concluding remarks Students also mimicked the work of chemists investigating historical textiles. For that, they The work reported in this thesis is focused on the use of dyestuff of weld (R. luteola L.) for had the assignment of performing a small-scale extraction of the dyed wool. After analysis of textile dyeing. Obtaining dyes from crops cultivated as dye plants may not be environmentally the sample by HPLC, each student was asked to compare the retention times and UV– friendly [15]. Advantageously from the environmental impact point of view, natural dyes for absorption spectra of the three main flavonoids of the chromatogram of the wool sample with large-scale use can be obtained from sources such as residues of food production [16, 17]. those of pre-analysed authentic standards of ldg, lmg and lut, and to compare the Therefore, the large-scale use of weld as a dye plant might be something of the past. If that is chromatographic profile of the wool sample with that of the weld sample. the case, then the methods described in chapters 2 and 3 could be used for analysing other The dyeing bath was prepared with deionised water in this experiment and in part of the sources of dyes after adaptation and additional validation. On what concerns the experiments reported in chapter 4. Although it is expected that the extract of weld with which chromatographic separation of flavonoids, further improvement might be achieved by the use the bath was prepared contained ions, it could have been preferable to use a bath richer in ions. of a column packed with a sub-2 µm superficially porous stationary phase on UHPLC Osmotic pressure is developed as water enters the wool fibre, and electrolytes in the dyeing instrumentation [18]. bath lead to a decrease in the difference between the pH inside the fibres and that of the bath More research on understanding the photodecomposition of natural dyes and on improving [13]. The pH equilibration relate to the flow of ions through the dyeing bath–fibre interface. If the light-fastness of the colours of textiles dyed with many such dyes is needed [19]. the presence of electrolytes in the dyeing bath also influences the flow of the dye molecules Therefore, the discussion in chapter 4 and above is expected to contribute to ongoing or through the dyeing bath–fibre interface, it might be preferable to prepare the dyeing bath with future works in the field of textile dyeing with natural dyes. The experiment for use in electrolytes-rich water. education in chemistry elaborated upon in chapter 5 may be used as is, as well as the A procedure for extracting dye molecules from the wool was developed for the experiment procedure for extracting dye molecules from wool described in the same chapter. The latter described in this chapter. A small piece of a single thread of dyed wool is extracted with may also be used for research purposes. Finally, cultivation of dye plants—weld included— methanol–water–formic acid 80:15:5 at 60 °C for 30 min. The development of this procedure remains important for production of goods at a handicraft level in relation to biological was based on: diversity and cultural conservation. • the fact that methanol–water 8:2 is a good solvent for the extraction of the flavonoids of weld {Cristea et al. reported one of the works using this solvent for the purpose; they observed it to be the most efficient among five different solvents [14]}; • the work by Zhang et al., in which samples of dyed textile fabrics were extracted with methanol–formic acid 95:5 at 60 °C for 30 min [9]. The extraction conditions are mild, with the chromatogram of the extract of the dyed wool being very similar to that of the extract of weld (Fig. 5.2). The larger ratios lmg–ldg and lut–

degradation of quercetin leads to at least one compound that displays antioxidant activity, 3,4- dihydroxybenzoic acid (protocatechuic acid) [10].

82 83

those of pre threeabsorption spectrawool sample the chromatogram theof mainflavonoids of the of chromatographic profile the sample bath contain prepared was bath the experiments Osmotic pressureOsmotic developed is enter as water with weld. 82 dihydroxybenzoic acid (protocatechuic acid) [10] quercetin degradation leads of at leastto one compound that bath through the dyeing the flow of influencesthe also bath the electrolytesin dyeing the of presence [13] perform had the assignmentof and one half howuse standard of to the theknew internal students analysis, methodfor quantitative works. Furthermore, phase chromatography separationbyreversed how of themunderstood the of extract an was quantify to a “ on method while working deepen the opportunity University biotechnology had molecular to Wageningen and life of sciences 5 Chapter 5 6.5. Chapter colour Thea use ofstrategy such methanol chapter described this in being The are conditions t extraction with mild, basedwas on: electrolytes T Students • A • . he dyeing bath was prepared deionised with water in this experiment

lead procedu The pH equilibration relate the to flow of ions methanol the work byZhang et al. observed itto be the efficient most methanol that fact the w (Fig. (Fig. weld of very extract similar of the to that . C

eld eld - their knowledge on HPLC and onHPLC knowledge their third of difficulty them had with it.

ombined, such gains might away be –water to

Cristea {Cristea However, each of its

chapter 4 reported chapter in also rich water. rich reports oneducation a work chemistry in at undergraduatelevel. by by a decrease in the difference between the pH inside pH the between difference the in a decrease re for extracting dye molecules dye extracting for re -

analysed authenticanalysed standardsldg of

–formic acid 95:5 at 60 °C for–formic at [9] 95:5 60°C 30 min acid

– m HP ldg same sample. same formic acid 80:15:5 imick LC , et al. lmg ,

– each student was student each ed of the woolsample with that of sample the weld , and lut , it interface fibre .

limits reported reported

A small piece the work of chemists investigating historical textiles water 8:2 –water ed natural dyes for textiles

ing ing nearly all all thatnearly was it observed by reports Based students, on15 ,

the range of of range the samples of dye of whichin samples ions, ions,

elements in aweld,in sample and dye of to . Although itis expected that the extractof weld with which a one of the works using this solvent for solvent worksone this the purpose using of the small- at 60° in in richer bath a use to preferable have been could it

is ais good for solvent the extra among five among

might be of of scale extraction extraction scale

quantitative analysis using extract of the dyed the wool of he chromatogram of the extract

could lead to asked asked C

. from the wool from that can be obtained be can that colours of meeting of today’s dyed dyed of thread a single

s for. 30min

the woolfibre,the to to

preferable to prepare the dyeing bath with bath dyeing the prepare to preferable th diffe compare compare , 5.2). ” r ough the dyeing bath - themed lmg . d rent [14] solvents a small gain

T of the dyed wool The larger ratio larger The textile fabric

was developed for the experiment experiment the for developed was he development of this procedure and displays antioxidant activity, 3,4-

the the and electrolytes in the dyeing dyeing the in electrolytes and nt experime the fibres the requirement of the property requirement

alum lut retention times and times UV retention ction of the flavonoids of ,

wool

in light- - and compare to the the internalthe standard s ted wool with mordanted woolwith .

upon dyeing textiles were extracted with extracted were and thatof the bath s };

–fibre . and of the part in lmg Their assignment assignment Their is

fter fter fastness . dye molecules extracted with –ldg For that

Students of interface. analysi

and

of the the of ; ,

with they they they they lut ions. s

of of If If – – .

diversitycultural conservation. and remain purposesmayforalso be research used procedure chemistryeducation in future works infield the of textile dyeing natural with dyes. Th the light- instrumentation [18] of As chromatographic separationof fla adapt after dyes of sources 2 chapters in described methods the then case, the - large the Therefore, - large is to expected are lmg friendly textile dyeing. ( of dyestuff weld the focused use is of thesis on The reportedthis in work remarks 6.6. Concluding purposes research small, erefore More research More • • a column the

of theof chromatogram of the extract of the dyed wool weld relative of to that the of extract scale use can be obtai be can use scale discussion) theto aglyconesmonoglycosides relative worse of extraction the diglycosides the to monoglycosides relative and of the agreementdiscussed section the in with data4.4.2.2); the to aglyconesmonoglycosides relative lower uptakediglycosides of the the to monoglycosidesand relative of the line with

s difference difference

[15] important the fastness colours of of d textiles the iscussion , the discussion for extracting dye molecules extracting for

. reported reported

Advantageously from of impact the view, point environmental natural dyes for O packed with a with packed

be due to be results published by Willemen et al. btainingdyes from }.

, as done on understanding the photodecomposition of natural dyes and oni of naturaland dyes photodecomposition understanding the on relative quantities of the flavonoids quantities relative of the terms chromatograms in between for production offor production . scale use of weld as ady as weld of use scale

procedure procedure elaborated upon the following reasons:

—in an—in adapted form in ned 17] from such sources [16, as residuesfood production of

chapter 4 chapter sub- the concerns validation.Onthe what additional ation and is vonoids, vonoids,

2 µm superficially porous stationary phase on UHPLC UHPLC phase stationary on porous superficially 2 µm sui crops cultivated as dye plants may not be environmentally environmentally maybe not crops as dye cultivated plants table for the intended purposes fortable the goods

from wool and above is expectedand contribute is to above ongoing to or Finally, c in in further improvement

e plant mightIf beofe somethingthe plant past. thatis chapter 5 chapter at a handicraft level in relation to biological level handicraft a at

— from from yed yed

by Willemen [11] al. et ultivation of dyeultivation plants by the Al the by

described in the same chapter same the in described

and the Al the with [11]

may be used as is, as well as the the as well as is, as used be may 3

(left panel of F of panel (left

many many 3+ could be used for analysing other analysing other for used could be - mordanted wool 3+

- might mord

for for experiment The such . I anted (this wool is in

be achieved bythe use t dyes dyes may . ig. 5 and preceding preceding and 5 ig.

—weld included— R. luteola

is also also

{

the latter is in needed needed . The latter be be mproving mproving use L.) for use in in use d [19]

for 83 . .

Chapter 6 6.7. References and Colourists: Electronic format (doi: 10.1002/9781118625118.ch7); 2013, p. 205–28. [1] Gaspar H, Moiteiro C, Turkman A, Coutinho J, Carnide V. Influence of soil fertility on [16] Ganglberger E. Environmental aspects and sustainability. In: Bechtold T, Mussak R, dye flavonoids production in weld (Reseda luteola L.) accessions from Portugal. J Sep Sci editors. Handbook of natural colorants, John Wiley & Sons: Chichester; 2009, p. 353–66. 2009; 32(23–24): 4234–40. [17] Bechtold T, Mahmud-Ali A, Mussak R. Natural dyes from food processing wastes. In: [2] Jeffrey SW, Mantoura RFC, Wright SW, editors. Phytoplankton pigments in Waldron K, editor. Handbook of waste management and co-product recovery in food oceanography: guidelines to modern methods. 1st ed. UNESCO Publishing: Paris; 1997. processing, CRC Press/Woodhead Publishing: Boca Raton/Boston/etc.; 2007, p. 502–33. [3] White RC, Jones ID, Gibbs E. Determination of chlorophylls, chlorophyllides, [18] de Villiers A, Venter P, Pasch H. Recent advances and trends in the liquid- pheophytins, and pheophorbides in plant material. J Food Sci 1963; 28(4): 431–6. chromatography–mass spectrometry analysis of flavonoids. J Chromatogr A 2016; 1430: 16– [4] Lichtenthaler HK. Chlorolphylls and carotenoids: pigments of photosynthetic 78. biomembranes. In: Packer L, Douce R, editors. Plant cell membranes, Academic Press: San [19] Shahid-ul-Islam, Sun G. Thermodynamics, kinetics, and multifunctional finishing of Diego/New York/etc.; 1987, p. 350–82. textile materials with colorants extracted from natural renewable sources. ACS Sustainable [5] Giles CH. The fading of colouring matters. J Appl Chem 1965; 15: 541–50. Chem Eng 2017; 5(9): 7451–66. [6] Cristea D, Vilarem G. Improving light fastness of natural dyes on cotton yarn. Dyes Pigments 2006; 70(3): 238–45. [7] Mouri C, Mozaffarian V, Zhang X, Laursen R. Characterization of flavonols in plants used for textile dyeing and the significance of flavonol conjugates. Dyes Pigments 2014; 100: 135–41.

[8] Seifzadeh N, Sahari MA, Barzegar M, Gavlighi HA, Calani L, Del Rio D, Galaverna G. Evaluation of polyphenolic compounds in membrane concentrated pistachio hull extract. Food Chem 2019; 277: 398–406. [9] Zhang X, Cardon D, Cabrera JL, Laursen R. The role of glycosides in the light- stabilization of 3-hydroxyflavone (flavonol) dyes as revealed by HPLC. Microchim Acta 2010; 169(3–4): 327–34.

[10] Ferreira ESB, Quye A, McNab H, Hulme AN. Photo-oxidation products of quercetin and morin as markers for the characterisation of natural flavonoid yellow dyes in ancient textiles. Dyes History Archaeol 2002; 18: 63–72. [11] Willemen H, van den Meijdenberg GJP, van Beek TA, Derksen GCH. Comparison of madder (Rubia tinctorum L.) and weld (Reseda luteola L.) total extracts and their individual dye compounds with regard to their dyeing behaviour, colour, and stability towards light. Color Technol 2019; 135(1): 40–7. [12] Takahama U, Hirota S. Deglucosidation of quercetin glucosides to the aglycone and formation of antifungal agents by peroxidase-dependent oxidation of quercetin on browning of onion scales. Plant Cell Physiol 2000; 41(9): 1021–9.

[13] Rippon JA. The chemical and physical basis for wool dyeing. In: Lewis DM, Rippon JA, editors. The coloration of wool and other keratin fibres, John Wiley & Sons/Society of Dyers and Colourists: Electronic format (doi: 10.1002/9781118625118.ch2); 2013, p. 43–74. [14] Cristea D, Bareau I, Vilarem G. Identification and quantitative HPLC analysis of the main flavonoids present in weld (Reseda luteola L.). Dyes Pigments 2003; 57(3): 267–72. [15] Duffield PA. Dyeing wool with acid and mordant dyes. In: Lewis DM, Rippon JA, editors. The coloration of wool and other keratin fibres, John Wiley & Sons/Society of Dyers

84 85

84 editors. The coloration of wooland Wiley otherfibres, keratin John & Sons/Society Dyers of Duffield[15] Dyeing PA. acid woolwith and mordant dyes.In:Lewis DM, Rippon JA, weldmain flavonoids present( in Cristea D,[14] BareauI,Identification G. and Vilarem quantitative HPLC analysis of the p.43–74. Electronicand 2013, Colourists: format 10.1002/9781118625118.ch2); (doi: editors. The coloration of wooland Wiley otherfibres, keratin John The Rippon chemical[13] JA. and physicalIn:Lewis basisfor DM, Rippon wooldyeing. JA, of scales. Physiol onion Cell Plant 41(9): 2000; 1021–9. formation of antifungal agents by peroxidase Takahama[12] Deglucosidation S. U, quercetinglucosides Hirota of the to aglycone and TechnolColor 135(1): 2019; 40–7. dye ( madder Willemen[11] H, van den van Meijdenberg Beek of Comparison Derksen GJP, GCH. TA, 63–72. 18: Dyes HistoryArchaeol 2002; flavo natural of characterisation the for markers as morin Ferreira[10] ESB,A, McNab Quye Photo- H, Hulme AN. 169(3–4):2010; 327–34. stabilization of 3 glycosides the in light role Theof Laursen R. ZhangD, Cabrera JL, [9] X, Cardon 398–406. 277: 2019; Chem extract. Food membrane compoundsconcentratedhull in Evaluation of polyphenolic pistachio Seifzadeh N,[8] Sahari MA, B 135–41. Pigments 100: 2014; conjugates. flavonol Dyes the significance dyeingof and textile used for Mozaffarian MouriC, [7] CharacterizationLaursen V, R. Zhang of X, flavonolsplants in Pigments 70(3): 2006; 238–45. Dyes yarn cotton Improving ofdyes. natural on Cristeafastness G. light D,[6] Vilarem Giles 541– 15: 1965; ApplChem Theof CH. colouring[5] fading matters. J York/ Diego/New biomembranes. In: Packer L, of photosynthetic pigments Lichtenthalercarotenoids: Chlorolphylls and HK. [4] Food 28(4): 1963; Sci 431–6. J plantmaterial. pheophytins, in and pheophorbides ID,RC, Jones of E.Gibbs Determination chlorophylls,chlorophyllides, White [3] modern1997. guidelines ed. to Paris; oceanography: 1st methods. UNESCOPublishing: pigments editors. Phytoplankton in SW, Wright RFC, Mantoura Jeffrey[2] SW, 32(23–24):2009; 4234–40. dye production weld in ( flavonoids Gaspar[1] Turkman H, MoiteiroC, A,Influence CoutinhoCarnide J, V. fertility of soil on 6.7. References compounds with regardcompounds with their to dyeing colour, behaviour, and stability towards light. Rubia tinctorum Rubia

- etc. Microchim Acta Acta Microchim hydroxyflavone asbyHPLC. revealed dyes (flavonol) ; 1987, p.350–82. 1987, ;

L.) and weld ( Douce R, editors. Plant cell membranes, Academic Press: San San Press: Academic membranes, cell Plant editors. R, Douce arzegar M, Gavlighi HA, Calani L, Del Rio D, Galaverna G. G. Galaverna D, Rio L, Del Calani HA, Gavlighi M, arzegar Reseda luteola Reseda Reseda luteola luteola Reseda - depe

L.). Dyes ndent oxidation of quercetinndent oxidation onbrowning L.) SepSciPortugal. accessions from J L.) total extracts and their individual L.)and their individual extracts total noid yellownoid dyes in ancient textiles. Pigments 2003; 57(3): 267–72. Pigments 2003; oxidation productsoxidation of quercetin and

& Sons/Society of Dyers Dyers of & Sons/Society 50. - EngChem 5(9): 2017; 7451–66. Sustainable ACS textile materials sources. with colorants renewable extracted natural from - Shahid [19] 78. chromatography de- Villiers the in liquid and Pasch[18] A,trends advances VenterH. P, Recent Boca Press/Woodhead Publishing: processing, Raton/Boston/ CRC K,Waldron of editor. waste Handbook managementand co - [17] Bechtold[17] T, Mahmud- editors. Handbook Wileyp.353–66. ofcolorants,Chichester; natural 2009, John Sons: & Ganglberger[16] aspects sustainability. E. Environmental In: and T, Mussak R, Bechtold Electronicand Colourists: format 10.1002/9 (doi:

ul - Islam, G. kinetics, Thermodynamics, and finishing Sun multifunctional of mass spectrometry analysis of flavonoids. J Chromatogr J 16– spectrometryA1430: 2016; of flavonoids. –mass analysis Ali A,NatAli Mussak R. 781118625118.ch7); 2013, p.205–28. 781118625118.ch7); 2013, ural dyesIn: food processing from wastes. product recovery food in product etc. ; 2007, p.502–33. 2007, ; 85

Chapter 6

Summary

Coloured textiles have been used by mankind throughout times. The use of natural dyes plant material having different chls/chlorophyllides-to-“pheopigments” ratios is hampered due experienced decline when synthetic dyes entered the market. Although dyeing textiles with to differences in optical molar absorption coefficients. One could test if this limitation can be natural dyes is not synonymous with environmentally friendly dyeing, there has been a renewed overcome by conversion of chls and chlorophyllides to the corresponding “pheopigments” by interest in natural dyes. Weld (Reseda luteola L.) used to be an important vegetable source of using HCl-acidified ethanol—instead of absolute ethanol—as extraction solvent. dye for textiles in Europe and the Mediterranean area. Alum (an aluminium salt)-premordanted Alum-premordanted wool dyed with weld leads to yellow colours that are of low resistance wool dyed with weld leads to yellow colours, with the main colouring compounds of the plant to light. The work described in chapter 4 was carried out in the context of such an issue. The belonging to the class of the flavonoids. The flavone luteolin (lut) and two of its glycosyl effect of different concentrations of aluminium ion on the photo-stability of the dye of weld, conjugates—lut-7-O-glucoside (lut monoglucoside, lmg) and lut-7,3ʹ-O-diglucoside (lut and the relative photo-stability of lut, lmg and ldg in solution are reported. Results of the diglucoside, ldg)—are the main flavonoids of the plant. The research described in this thesis is colours obtained by dyeing alum-mordanted wool with the three flavones are also presented. part of an academic and industrial project focusing on the use of dyestuff of weld for textile On one hand, the observed order of photo-stability of the flavones in solution was ldg >> lut dyeing. > lmg. On the other hand, alum-mordanted wool dyed with ldg (or with an extract of weld) was A reversed phase-high-performance liquid chromatography with UV detection (RP-HPLC– much less colourful than that dyed with either lut or lmg. The light-fastness of the colour of UV) analytical method was developed and validated for the quantitation of ldg, lmg, and lut weld-dyed wool decreased with increasing [Al3+] used for premordanting the wool. As the gain via internal standardisation in weld plant material. This was done in connection with the in light-fastness by the use of low [Al3+] was limited, this cannot be a way to meet today’s expectation of having to analyse hundreds of samples in the course of the dye research of which requirement of light-fastness of the colours of dyed textiles by itself. Nevertheless, it could be this project is part. The method—described in chapter 2—is simple, requiring very little part of a broader strategy to address the need for increased light-fastness of the colour of wool manpower input per sample. With the exception of the 15-stirring point magnetic stirrer—that, dyed with weld. less conveniently, could be replaced by individual magnetic stirrers—only standard laboratory Undergraduate students of life sciences should be acquainted with HPLC and internal equipment is needed. The analytical method proved fit-for-purpose based on its validation in standard method for the quantitation of compounds. Students of molecular life sciences and terms of extraction efficiency, stability and losses of the compounds, accuracy, precision, and biotechnology of Wageningen University had the opportunity to deepen their knowledge on sensitivity. Weld plant material having high concentrations of the flavones could be analysed both while working on a “natural dyes for textiles”-themed experiment. This experiment is through a slightly modified version of the method, using 100 mg instead of 200 mg of sample. reported in chapter 5. Analyses of a small number of replicates of 100 and 200 mg of a few samples of the plant are Their assignment was to quantify ldg, lmg, and lut in a batch of weld plant material, and to expected to be sufficient to test the validity of this approach. dye alum-mordanted wool with an extract of the same batch. Based on 15 reports by students, The analytical method was validated using a 5 µm-particle size RP-HPLC column. Then, it was observed that nearly all of them understood how separation by reversed phase the chromatographic separation was speeded up by using a sub-2 µm-particle size (RP-UHPLC) chromatography works. Furthermore, half of the students knew how to use the internal standard column. This was done while still using a conventional HPLC system, after minor hardware method for quantitative analysis, and one-third of them had difficulty with it. adaptation. The method using the UHPLC column proved fit-for-purpose through comparison Students also mimicked the work of chemists investigating historical textiles. For that, they of accuracy and precision of the quantitation of ldg, lmg, and lut with those achieved by using had the assignment of performing a small-scale extraction of the dyed wool. After analysis of the HPLC column. Large gains in analysis time and eluent consumption were obtained in this the sample by HPLC, each student was asked to compare the retention times and UV– way. absorption spectra of the three main flavonoids of the chromatogram of the wool sample with According to the industrial partner of the project, one of the problems to overcome was a those of pre-analysed authentic standards of ldg, lmg and lut, and to compare the greenish hue that accompanies the yellow colour after dyeing alum-premordanted textiles with chromatographic profile of the wool sample with that of the weld sample. weld. Such a greenish hue would be undesirable, and it was hypothesised that chlorophylls a The procedure for extracting dye molecules from the wool used by the students was and b are potential sources of it. In relation to this, an analytical method was developed for the developed for the experiment described in this chapter. A small piece of a single thread of dyed comparison of the content of porphyrin ring-containing pigments—chlorophylls (chls) and their wool is extracted with methanol–water–formic acid 80:15:5 at 60 °C for 30 min. These structurally similar breakdown products—in different samples of weld. The method, that is conditions are mild, with the chromatogram of the extract of the dyed wool being very similar described in chapter 3: to that of an extract of weld. Thus, the procedure is suitable for the intended purposes, and was i. uses ethanol as a mild, non-toxic extraction solvent (that is not selective regarding the used—in an adapted form—for research purposes. “types” of chls and plant tissues being extracted); The work reported in this thesis is focused on the use of dyestuff of weld (R. luteola L.) for ii. covers the entire range of occurring concentrations; textile dyeing. From an environmental impact point of view, obtaining natural dyes for large- iii. displays acceptable precision; scale use from sources such as residues of food production might be advantageous over iv. is simple. obtaining dyes from crops cultivated as dye plants. Although the large-scale use of weld as a The method has the drawback that the manpower input required is not low and displays a dye plant might be something of the past, what is reported in this thesis remains relevant either limitation. Comparison of the content of porphyrin ring-containing pigments in batches of weld “as is” (chapters 4 and 5) or after adaptation/additional validation for analysing compounds in

88 89

88 89 other biological material (chapters 2 and 3). Finally, cultivation of dye plants—weld included—remains important for production of goods at a handicraft level in relation to biological diversity and cultural conservation.

Acknowledgements

90 other biological material (chapters 2 and 3). Finally, cultivation of dye plants—weld included—remains important for production of goods at a handicraft level in relation to biological diversity and cultural conservation.

Acknowledgements

I feel happy and grateful for having arrived at the stage of the PhD work in which I now am. The work is nearly complete. That is, the reading copy of the thesis has been submitted, the moment of defending it and graduating is at sight. As the final bits and pieces of the thesis are put together, it is my pleasure to write this very section. I mention few names, but I would like you to feel thanked for your contribution to this journey of mine regardless of its extent. If we have not met during this journey—or have not met yet—you can have an idea of those who were important to it by reading this section. I am thankful for those who took care of me as I grew up and supported me on the journey. This is not limited to but importantly, includes my family and friends. I would like to particularly mention Dani—a good friend—with whom I walked a good length of the journey, lived abundance and experienced scarcity; Ciro—my father here on Earth—in whom I had an example of endurance in adversity; and Josely—my mother—who did a PhD herself over the last years, in whom I had unyielding encouragement to keep studying. I am grateful to all those who—directly or indirectly—contributed to making the beginning of this PhD journey possible. This includes teachers and supervisors during my education in chemistry, internships, MSc thesis work, and jobs. Different people contributed to the work of this thesis. I am grateful to all of them for the input. The political, conceptual and technical contribution to my PhD work of Eric, Monique, Hendra and Dorien (Rubia Natural Colours), and Elbert, Jacob, Barend, Han and Teris The future has a long history (Laboratory of Organic Chemistry) was of particular importance. I grew professionally Stadszegel, Zwolle/The Netherlands considerably through working under the supervision of Han and Teris. Both of them are committed professionals, and the knowledge of Teris on analytical organic chemistry is vast. My professional development during the large time span of working on this thesis also encompasses that which took place during activities that ran in parallel to it. Different people contributed to my growth as I worked as teaching assistant at Wageningen University and analytical chemist at Entomology. I am grateful to the input of all of them too. Working with Erik and Frank (Laboratory of Organic Chemistry), and Marcel (Laboratory of Entomology) was encouraging and inspiring. The interaction with different colleagues with whom I worked during the professional activities mentioned above was also important to me. Some became friends. I grew personally through these interactions, and am grateful for our time together. I was involved on a number of activities besides work. Two particularly important ones were the Dutch language course, and the tennis trainings and competitions. I am thankful for the time I had together with many people during those activities. Mates and those who served as organizers and teachers—professionally or as volunteers—contributed substantially to my development. My spiritual development was of extreme importance to me. It was supported and facilitated by a number of people and organizations. I am thankful to all of them. In particular, participating in various activities organized by the International Christian Fellowship and the Student Chaplaincy—and the community that remained after its official end—was joyful, instructive, and of spiritual significance. God—my father in Heaven—did great things to me. All what He did allowed me to be where I now am, including being at this point of concluding the PhD work. I am very thankful to Him for the nearness, for everything.

92 93

During the whole process, I have seen many people much less often that I would have liked to. I would be glad to see many of them—both those I knew before this very journey started and others I got to know along the way—much more often in the future. With Fefo and Leka—my brother and sister—and many others, I look forward to—to varying degrees— share many adventures to come.

Publications and training activities

PhD work-related publications: full-text papers in peer-reviewed Overview of completed training activities journals If no geographic reference is made, it is because the activity took place in Wageningen or its Villela A, van Vuuren MSA, Willemen HM, Derksen GCH, van Beek TA. Photo-stability of a surroundings. flavonoid dye in presence of aluminium ions. Dyes Pigments 2019; 162: 222–31 Discipline-specific activities Villela A, Derksen GCH, van Beek TA. Analysis of a natural yellow dye: an experiment for KNCV (Royal Netherlands Chemical Society) Annual Spring Meeting: analytical organic chemistry. J Chem Educ 2014; 91(4): 566–9 • Advances in hyphenated MS; 17 Apr 2008 • New developments in separation methods; 20 Apr 2012 Villela A, van der Klift EJC, Mattheussens ESGM, Derksen GCH, Zuilhof H, van Beek TA. Fast chromatographic separation for the quantitation of the main flavone dyes in Reseda luteola 7th Joint Meeting of AFERP, ASP, GA, PSE & SIF on Natural Products with pharmaceutical, (weld). J Chromatogr A 2011; 1218(47): 8544–50 nutraceutical, cosmetic and agrochemical interest (Athens/Greece); 04–07 Aug 2008

Villela A, Derksen GCH, Zuilhof H, van Beek TA. Spectrophotometric comparison of the Annual NWO Meeting on Analytical Chemistry (Lunteren/The Netherlands): content of chlorophylls in weld (Reseda luteola). Anal Methods 2011; 3(6): 1424–7 • 2008 edition; 03–04 Nov 2008 • 2009 edition; 02–03 Nov 2009 • 2010 edition; 01–02 Nov 2010

Annual NWO Meeting on Organic Chemistry (Lunteren/The Netherlands); 20–21 Oct 2009

Advanced Organic Chemistry; 2009–2010 (VLAG and Laboratory of Organic Chemistry, WUR)

International Symposium and Exhibition on Natural Dyes (La Rochelle/France); 25–29 Apr 2011 (ARRDHOR CRITT HORTICOLE and CIHAM UMR 5648)

46th International Symposium on Essential Oils (Lublin/Poland); 13–16 Sep 2015 (Medical University of Lublin, Department of Pharmacognosy with Medicinal Plant Unit of Medical University of Lublin, and Polish Academy of Sciences)

19th Congress of the International Society for Ethnopharmacology (Dresden/Germany); 12–14 Jun 2019

Advances in Plant and Food Metabolomics (3rd WURomics Symposium); 12 Dec 2019 (EPS and WUR)

General courses Information Literacy for PhD Including Introduction to EndNote; 27–28 May 2008 (Library. WUR)

20th VLAG PhD Week (Bergeijk/The Netherlands); 27–30 Oct 2008 (VLAG)

Training “Study Skills”; 10 Oct 2009 (Purple Monkey and WUR)

96 97