g model article in press · please cite this article in press as: m. clément chami, et al.,...

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/328503953

Purification of two valepotriates from Centranthus ruber by centrifugal

partition chromatography: From analytical to preparative scale

Article in Journal of Chromatography A · October 2018

DOI: 10.1016/j.chroma.2018.10.044

CITATIONS

2READS

258

7 authors, including:

Some of the authors of this publication are also working on these related projects:

Sustainable Tomato Production: plant defense enhancement, development of new biopesticides and optimization of environmental, water and chemical inputs

(STomP) View project

Tocols View project

Bouju Elodie

Extrasynthese

14 PUBLICATIONS 106 CITATIONS

SEE PROFILE

Rene De Vaumas

Extrasynthese


SEE PROFILE

Francis Hadji-Minaglou

Botanicert


SEE PROFILE

Xavier Fernandez

Université Côte d'Azur

170 PUBLICATIONS 2,930 CITATIONS

SEE PROFILE

All content following this page was uploaded by Thomas Michel on 04 November 2018.

The user has requested enhancement of the downloaded file.

https://www.researchgate.net/publication/328503953_Purification_of_two_valepotriates_from_Centranthus_ruber_by_centrifugal_partition_chromatography_From_analytical_to_preparative_scale?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_2&_esc=publicationCoverPdfhttps://www.researchgate.net/publication/328503953_Purification_of_two_valepotriates_from_Centranthus_ruber_by_centrifugal_partition_chromatography_From_analytical_to_preparative_scale?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_3&_esc=publicationCoverPdfhttps://www.researchgate.net/project/Sustainable-Tomato-Production-plant-defense-enhancement-development-of-new-biopesticides-and-optimization-of-environmental-water-and-chemical-inputs-STomP?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_9&_esc=publicationCoverPdfhttps://www.researchgate.net/project/Tocols?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_9&_esc=publicationCoverPdfhttps://www.researchgate.net/?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_1&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Bouju-Elodie?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Bouju-Elodie?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/institution/Extrasynthese?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_6&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Bouju-Elodie?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Rene-De-Vaumas?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Rene-De-Vaumas?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/institution/Extrasynthese?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_6&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Rene-De-Vaumas?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Francis-Hadji-Minaglou?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Francis-Hadji-Minaglou?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Francis-Hadji-Minaglou?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Xavier-Fernandez-10?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_4&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Xavier-Fernandez-10?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_5&_esc=publicationCoverPdfhttps://www.researchgate.net/institution/Universite-Cote-dAzur?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_6&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Xavier-Fernandez-10?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_7&_esc=publicationCoverPdfhttps://www.researchgate.net/profile/Thomas-Michel-3?enrichId=rgreq-f424fade622d1495f0296238ed285d79-XXX&enrichSource=Y292ZXJQYWdlOzMyODUwMzk1MztBUzo2ODkzNTMzOTUyODYwMThAMTU0MTM2NjA2NTgyMQ%3D%3D&el=1_x_10&_esc=publicationCoverPdf

C

Pp

MFa

b

c

d

a

ARRAA

KDSSCVC

1

dIobfi[

a“

t

h0

ARTICLE IN PRESSG ModelHROMA-359768; No. of Pages 8Journal of Chromatography A, xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Journal of Chromatography A

journa l homepage: www.e lsev ier .com/ locate /chroma

urification of two valepotriates from Centranthus ruber by centrifugalartition chromatography: From analytical to preparative scale

élissa Clément Chami a,b, Elodie Bouju c, Céline Lequemener d, René de Vaumas c,rancis Hadji-Minaglou b, Xavier Fernandez a,∗, Thomas Michel a,∗

Université Côte d’Azur, CNRS, Institut de Chimie de Nice, UMR 7272, Nice, FranceBotaniCert, 4 traverse Dupont, 06130 Grasse, FranceExtrasynthèse, rue Jacquard, 69730 Genay, FranceGilson Purification, 22 rue Bourseul, 56890 Saint-Avé, France

r t i c l e i n f o

rticle history:eceived 28 August 2018eceived in revised form 17 October 2018ccepted 22 October 2018vailable online xxx

eywords:ietary supplementecondary metabolites isolationcale upentrifugal partition chromatographyalepotriates

a b s t r a c t

Considering chemical complexity of plant crude extracts, purification of natural products is a rate limitingprocess to identify new compounds as well as to obtain standard references for quantitative or qualitativepurposes. In the present work, a centrifugal partition chromatography (CPC) method was developed toisolate and produce high quality reference standards of valtrate and 7-homovaltrate from Centranthusruber L. roots. These two compounds are controversial aglycon iridioids regulated by the legislation onplant-based dietary supplements. A new biphasic solvent system suitable for CPC separation of valepo-triates was developed. It was composed of methanol/hexane/water (5/5/0.8, v/v/v). It yielded a partitioncoefficient near 1 and a theoretical selectivity of 1.3 between both targeted compounds. Optimizationof CPC experimental parameters at the analytical scale (50 mL- and 100 mL-column capacity) enabledcompounds’ separation with a flow rate of 8 mL/min at 2500 rpm. Then a scale up from a 100 mL-columncapacity to a 1000 mL-column capacity has been studied using the “free-space between peaks” concept.

entranthus ruber L. It allowed an injected quantity 16 times higher in comparison to the maximal loading capacity of the100 mL-column. Both valtrate and 7-homovaltrate were recovered in one single step with a purity over97%. Further MS and NMR characterization allowed to confirm unambiguously the compounds’ struc-tures. The highly efficient CPC separation developed in this work provides valepotriates in amountssuitable for further study and strong bases for future industrial development.

© 2018 Elsevier B.V. All rights reserved.

. Introduction

Plant biodiversity is a source of naturally occurring active ingre-ients of increasing interest for both consumers and industries [1].

ts valorisation by diverse sectors like cosmetics, pharmaceuticsr agri-food is reflected by the plurality of commercialized plant-ased products such as dietary supplements, drugs, or functionaloods. In order to guarantee consumers’ safety, their manufactur-ng processes require strict specifications before commercialization2].

Please cite this article in press as: M. Clément Chami, et al., Purificapartition chromatography: From analytical to preparative scale, J. Chro

Legislation of plant-based dietary supplements is different from country to another: for instance in France, a new decree calledPlants decree” has entered into force on January 1st, 2015 [3]. It

∗ Corresponding authors.E-mail addresses: [email protected] (X. Fernandez),

[email protected] (T. Michel).

ttps://doi.org/10.1016/j.chroma.2018.10.044021-9673/© 2018 Elsevier B.V. All rights reserved.

describes the appropriate ways of use of 534 plants addressed tomanufacturers: therefore, each plant has to be identified throughdefined criteria like compliance to pharmacopoeia specifications,etc., and for 210 of these plants, a complementary dosage ofcritical compounds has to be completed. This new regulation fol-lows international trend to frame dietary supplement markets:on February 20, 2017 Belgian authorities increased the number ofplants approved in food supplements and set conditions to theiremployments. In Italy, extended positive list of plants associated toenforced assays have also been adopted via a decree dated March27, 2014. These three examples are the first results of a harmoniza-tion program called “BELFRIT project” initiated between Belgium,France and Italy to facilitate plants-based food supplements mar-keting and to ensure consumers’ safety [4].

tion of two valepotriates from Centranthus ruber by centrifugalmatogr. A (2018), https://doi.org/10.1016/j.chroma.2018.10.044

New regulatory requirements concerning the dosage of specificsecondary metabolites present two challenges. Firstly, numerouscritical compounds contained in plants that have to be monitoredare not yet described in major Pharmacopoeias such as the Euro-

https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044http://www.sciencedirect.com/science/journal/00219673http://www.elsevier.com/locate/chromamailto:[email protected]:[email protected]://doi.org/10.1016/j.chroma.2018.10.044

ING ModelC2 hrom

p(tt

icapceaiv

ac[fiEaaioapcppel

etTCtthcswvtacaw

2

2

mfw

2

aMM

ARTICLEHROMA-359768; No. of Pages 8 M. Clément Chami et al. / J. C

ean (Ph. Eur.), United State (USP), and Japanese PharmacopoeiasJP) which implies new analytical methods development. Secondly,his lack of standardized methods has maintained access difficultieso analytical standards of critical compounds [5].

It is notably the case of molecules present in Valeriana offic-nalis L. and V. jatamansi J. These plants are used since manyenturies for the treatment of sleep disorder or anxiety [6,7] andccording to the “Plants decree” they are allowed in dietary sup-lement formulations if they do not contain critical compoundsalled valepotriates Valepotriates is the contraction of valeriana-poxy-triesters [8]. These cyclopentan-c-pyran monoterpenoidsre secondary metabolites present in plants belonging to the fam-ly Caprifoliaceae [9] such as Centranthus ruber L. also called “redalerian” or “spur valerian”, an endemic Mediterranean species.

Despite many studies on these iridoids, their effects on humanre still not clearly defined but their mutagenic, cytotoxic andarcinogenic activities have been distinctively assessed in vitro10,11]. Therefore, valepotriates may be defined as undesirable innal products. Even though V. officinalis is described in both Ph.ur. and USP, associated monographs do not report valepotriatessays. This gap results in a limited number of commercially avail-ble analytical standards of valepotriates and, at the same time,n bottleneck limitations for the dosage of these compounds. Tour knowledge, one-step industrial scale isolation of valepotri-tes has never been reported yet. The only cases of valepotriates’urifications involved multi step strategies including open columnhromatography fractionation, followed by semi preparative highressure liquid chromatography (HPLC) and/or preparative higherformance thin layer chromatography (HPTLC) [11–14]. Thus,fficient recovery of valepotriates is currently considered as a chal-enging task.

In this context, centrifugal partition chromatography (CPC) wasvaluated at analytical and preparative scales for isolation of isoval-rate [1], valtrate [2] and 7-homovaltrate [3] from C. ruber (Fig.1).hanks to its principle based on liquid-liquid chromatography,PC theoretically meets natural products’ isolation requirements:he absence of solid stationary phase avoids irreversible adsorp-ion, the flexibility of biphasic solvent system composition allowsigh adaptability of the chromatographic system without furtheronsumables investments and finally, considering the absence ofupport, CPC affords high injection capacity [15]. First, extractionas optimised to ensure selective and quantitative extraction of

alepotriates from C. ruber roots. The CPC separation of Centran-hus valepotriates were optimised on a 50 mL-column. Afterwards,

scale up was achieved between a 100 mL-column and a 1000 mL-olumn using the “free space between peaks” approach [16]. Themount of extracted valepotriates as well as purity monitoringere evaluated using UPLC-PDA.

. Experimental section

.1. Chemicals

Dichloromethane (DCM), acetone (Ac), n-hexane (n-Hex),ethanol (MeOH), acetonitrile (ACN), water and formic acid used

or crude sample preparation and chromatographic separationsere of analytical grade (Biosolve, Dieuze, France).

.2. Plant material and extraction process


C. ruber L. roots were collected in Grasse (France) on April 2017nd identified by pharmacist and ethnobotanist Dr. Francis Hadji-inaglou. A voucher specimen of the plant was deposited at theuseum of Natural History of Nice (NICE-D-4386).

PRESSatogr. A xxx (2018) xxx–xxx

Fresh C. ruber roots (2 kg) were dried at ambient temperature for3 days and reduced into powder with a ball mill MM 400 (Retsch,Haan, Germany). Ultrasound assisted extraction of dried powder(250 g) was achieved for 10 min with 2000 mL of DCM. The prepa-ration was centrifugated 4 min at 914 g using a Digitor 21 equippedwith a RT 191 rotor (Ortho Alresa, Madrid, Spain) the supernatantwas filtrated using 0.2 �m reticulated cellulose (RC) membrane fil-ters (Phenomenex, Torrance, California, USA) and evaporated underreduced pressure to constitute the crude extract (12 g) used for CPCseparation.

Several extraction solvents were tested in order to obtain themost concentrated valepotriates extract. Solvents were comparedin triplicate as follows: 15 mL of solvent were added to 100 mg ofC. ruber roots powder. Each sample was sonicated for 10 min andfiltered through 0.2 �m RC filters. Filtrates were evaporated undernitrogen flow. Yields were calculated before dilution in MeOH. Allsamples were analysed by UPLC-PDA using the method described inSection 2.3 and concentration was adjusted until peak areas of val-trate and 7-homovaltrate were within the range of the calibrationcurve.

2.3. UPLC-PDA-ESI-HRMS analysis

Sample analyses were performed on an UPLC Acquity systemcoupled to a photodiode array detector PDA, e� Acquity and aXEVO-G2QTOF instrument (Waters, Milford, Massachusetts, USA).Chromatographic and spectrometric conditions were adapted frompublished methods [17,18]. Separations were achieved on anAcquity UPLC C18 column (Waters, 1.7 �m, 100 × 2.1 mm) at 25 ◦Cwith a flow rate of 0.500 mL/min. The mobile phase consisting ofwater (solvent A) and ACN (solvent B), both acidified with 0.1%formic acid, was used in multistep gradient mode. The gradientwas operated as follows: 0 min, 40% B; 5 min, 60% B; 10 min, 95%B; final isocratic step for 2 min at 95% B. The sample manager wasthermostated at 15 ◦C, and the injection volume was set at 1 �L.The HRMS data were acquired over a mass range of 100–1500m/z. ESI conditions operated in positive mode were set as follows:source temperature: 150 ◦C, desolvation temperature: 500 ◦C; cap-illary voltage: 3 kV and cone voltage: 20 V. Nitrogen was used ascone (35 L/Hr) and desolvatation gas (1000 L/Hr). Lockspray flowrate was set at 20 �L/min and lockspray capillary voltage at 2.5 KV.

2.4. External standard calibration

In order to calculate total content of valepotriates duringthe extraction optimization step as well as to quantify iso-lated valepotriates, external calibration curves of valtrate and7-homovaltrate were realized. For this purpose, “reference stan-dards” were produced during analytical scale method developmenton the SCPC-100 apparatus. A series of highly purified valepotri-ates of 37.5, 75, 200, 400, 600 �g/mL were prepared in methanoland analysed in triplicate by UPLC-PDA using the method previ-ously described (Section 1.3). Calibration curves were constructedbased on the average area of valtrate and 7-homovaltrate for eachconcentration (cf. Supplementary material).

2.5. CPC equipment

Both analytical and preparative scale CPC separations wereperformed on Armen Instruments: a SCPC-50 apparatus (columncapacity: 50 mL), equipped with a binary pump, a Spot CPC sys-tem, a LS-5600 fraction collector coupled with a Flash 065 DAD 600


UV/vis detector (Armen Instrument, Saint Avé, France) scanningwavelengths from 200 nm to 600, and a SCPC-100 + 1000 appara-tus (column capacity: 100 mL and 1000 mL) coupled with a SpotPrep II system and an evaporative light scattering detector (ELSD).

https://doi.org/10.1016/j.chroma.2018.10.044

ARTICLE IN PRESSG ModelCHROMA-359768; No. of Pages 8M. Clément Chami et al. / J. Chromatogr. A xxx (2018) xxx–xxx 3

ucture of valepotriates.

A(

2

wKd5ldcwul1cc

2

aapf

pat3tTpf

2

rsirr

16J(1

Table 1Rate of extracted valepotriates depending on extraction solvent (relative standarddeviation (RSD): ratio of the standard deviation to the mean).

Solvent Extraction yield (%) Valepotriates content (%) RSD (%)

MeOH 24.4 8 1.7DCM 3.1 70 1.1Acetone 4.6 44 2.3ACN 3.9 50 1.2

Fig. 1. Typical str

ll the instruments were operated thanks to Armen Glider softwareArmen Instrument, Saint-Avé, France).

.6. Selection of the appropriate two-phase solvent system

Several biphasic solvent systems were tested. Selection criteriaere based on partition coefficient (Kd) results and selectivity (� =

2/K1)) of each system for compounds of interest. Kd and � wereetermined as follows: biphasic solvent systems were prepared in0 mL centrifugal tubes. Each tube was shaken and allowed to equi-

ibrate. If equilibration time exceeded 30 s, the system tested wasiscarded. Otherwise 2 mL of each phase were placed in a 10 mLentrifugal tube and 2 mg of crude extract were added. Test tubesere vigorously shaken, and compounds were allowed to divide

p in each phase. Upper and lower phases of each tube were ana-ysed by UPLC-PDA with the method previously described (Section.3). Area ratios of valtrate and 7-homovaltrate in each phase werealculated to obtain the Kd of each compound. Kd ratios of eachompound were calculated to determine � of each solvent system.

.7. CPC separation procedure

MeOH, n-Hex and water (5:5:0.8, v/v/v) were added in a sep-ratory funnel, vigorously shaken and allow to equilibrate. Uppernd lower phases were collected to constitute mobile/stationaryhase and crude extract of C. ruber was dissolved in upper phase

or injection.CPC separations were performed in ascending mode. The lower

hase was first loaded into the CPC column to constitute the station-ry phase. It was kept in place by the centrifugal force generated byhe rotor (2500 rpm, 3–30 mL/min for SCPC-50/100 and 1200 rpm,0 mL/min for SCPC-1000). Upper phase was then pumped throughhe stationary phase until equilibrium of the system was reached.he latter was observed when system pressure stabilized and upperhase was collected in vessel used for stationary phase retention

actor (Sf) calculation.

.8. H and 13C RMN of isolated compounds

1H NMR and 13C NMR spectra (cf. Supplementary material) wereecorded in CDCl3 at 25 ◦C on a 400 MHz Bruker® Avance NMRpectrometer. NMR Fourier transform, integration and peak pick-ng were done with NMR Notebook software. Chemical shifts wereeported in ppm relative to CDCl3, while coupling constants (J) wereeported in Hertz (Hz).

[2] : � 1H NMR (400 MHz, CDCl3): 6.67 (s, 1H, H-2), 5.96 (d,H, J = 10.2, H-9), 5.84 (t, 1H, J = 2.8, H-5), 5.35 (d, 1H, J = 2.9, H-


), 4.68 (m, 2H, H-1), 3.41 (dd, 1H, J = 2.8–10.1, H-8), 2.94 (dd, 2H, = 4.9–48.3, H-10), 2.21 (m, 2H, H-14), 2.16 (m, 2H, H-19), 2.09m, 2H, H-15, H-20), 2.04 (s, 3H, H-12), 0.96 (d, 6H, J = 6.5, H-16,H-7), 0.92 (d, 3H, J = 6.6, H-21, H-22); � 13C NMR (400 MHz, CDCl3):

172.69 (C-18), 171.10 (C-11), 170.55 (C-13), 148.7 (C-2), 141.14 (C-4), 118.86 (C-5), 108.54 (C-3), 92.76 (C-9), 83.26 (C-6), 64.39 (C-7),61.07 (C-1), 48.10 (C-10), 43.60 (C-19), 43.27 (C-14), 43.22 (C-8),26.05 (C-15), 25.83 (C-20), 22.56 (C-22,C-21, C-17), 22.45 (C-16),21.18 (C-12).

[3]: � 1H NMR (400 MHz, CDCl3): 6.68 (s, 1H, H-2), 5.96 (d,1H, J = 10.1, H-9), 5.84 (t, 1H, J = 2.7, H-5), 5.35 (d, 1H, J = 2.9, H-6), 4.68 (m, 2H, H-1), 3.41 (dd, 1H, J = 2.5–10.3, H-8), 2.94 (dd, 2H,J = 4.8–47.2, H-10), 2.29 (m, 1H, Ha-19), 2.21 (m, 2H, H-14), 2.09(m, 2H, H-15, Hb-19), 2.04 (s, 3H, H-12), 1.84 (m, 1H, H-20), 1.26(m, 2H, H-21), 0.96 (d, 6H, J = 6.5, H-16,H-17), 0.88 (m, 6H, H-22,H-23); � 13C NMR (400 MHz, CDCl3): 172.93 (C-18), 171.11 (C-11),170.56 (C-13), 148.7 (C-2), 141.13 (C-4), 118.86 (C-5), 108.53 (C-3), 92.78 (C-9), 83.26 (C-6), 64.38 (C-7), 61.07 (C-1), 48.10 (C-10),43.25 (C-14), 43.20 (C-8), 41.63 (C-19), 32.21 (C-20), 29.51 (C-21),25.81 (C-15), 22.56 (C-17), 22.43 (C-16), 21.17 (C-12), 19.43 (C-23),11.48 (C-22).

3. Results and discussion

3.1. Extraction procedure

Plant extracts are complex matrix containing a lot of com-pounds. As shown in Fig. 2, the UPLC analysis of the methanolic rootextract indicates that belowground part of C. ruber contain severalcompounds, including hydrophilic (

ARTICLE IN PRESSG ModelCHROMA-359768; No. of Pages 84 M. Clément Chami et al. / J. Chromatogr. A xxx (2018) xxx–xxx

Fig. 2. UPLC-PDA chromatogram of a methanolic extract of C. ruber roots, highlighted yellow zone for most apolar compounds (For interpretation of the references to colourin this figure legend, the reader is referred to the web version of this article.).

Table 2Partition coefficient in some biphasic solvent systems and related selectivity for isovaltrate [1], valtrate [2] and homovaltrate [3].

System Solvent system (v/v/v)Kd �

[1] [2] [3] [1]/ [2] [1]/ [3] [2]/ [3]

1 n-Hex/MeOH (7:5) 5.34 4.97 4.12 1.07 1.19 1.202.51.82.6

3

bsstobui

rbtwtpitdpwtm

3

3

ptcats52a

2 n-Hex/MeOH/H2O (5:5:0.5) 3.20 3 n-Hex/MeOH/H2O (5:5:0.8) 2.30 4 n-Hex/MeOH/H2O (5:5:1.5) 2.57

.2. Solvent system selection

Prior to start CPC method development, it is crucial to select theest solvent system able to separate valeriana’s iridoids. Herein,uitable solvent system was chosen according to partition andeparation factors (Kd and �). Due to the apolar character of valepo-riates it was decided to customize a solvent system based on a mixf hexane and methanol. Some water was added to consolidate theiphasic system. Different ratios of these three solvents were eval-ated. Table 2 summarized Kd and � factors of target compounds

n the different solvent systems.System 1 was discarded because of its high Kd for which long

un time could be foreseen. Systems 2 and 4 were not chosenecause of their low selectivity regarding valtrate [2] and homoval-rate [3], predicting the coelution of those compounds. System 3,hich depicted acceptable Kd and selectivity, has been selected

o develop the analytical scale CPC method. Mobile and stationaryhases were determined according to Ito’s recommendation which

ndicates that if an adequate Kd greater or equal to 1 is obtained,he numerator should be considered as the mobile phase and theenominator as the stationary phase [20]. So, in our case, the upperhase will constitute the mobile phase and the stationary phaseill be the lower phase. Such configuration is preferable because

he evaporation of the collected fractions will be simplified as theobile phase is predominantly constituted of n-hexane.

.3. Analytical development methodology in CPC

.3.1. Determination of the laboratory scale conditionsBesides selection of biphasic solvent system, the CPC separation

rocess is also influenced by the rotation speed and the flow-rate ofhe mobile phase. In addition, the “Free space between peaks” con-ept is only applicable if resolution between peaks is greater than 1t the laboratory scale [16]. Therefore, chromatographic optimiza-ion has been performed on a SCPC-50 apparatus. The rotor rotation


peed varied from 1000 rpm to 2500 rpm at a fixed flow rate of mL/min. It is worthwhile to note that rotation speed greater than500 rpm induced pressures close to the limits of the system fixedt 50 bars.

0 2.53 1.28 1.26 1.015 1.45 1.24 1.59 1.287 3.02 1.04 1.18 1.13

Compounds separation was evaluated in terms of resolutivepower (Rs). Nevertheless, resolution between peaks could not becalculated on the basis of acquired chromatograms because half-height peak width of 3 was not visible due to peaks overlapping.To overcome this issue, fractograms have been designed usingUPLC-PDA analyses of each fraction. As shown on the fractogramsrepresented on Fig. 3, variation of rotation speed does not allowdistinct separation of 1 and 2. However, an increase from 1000 rpmto 2500 rpm induces higher resolution between 2 and 3 (Rs at1000 rpm: 0.44; Rs at 2000 rpm: 0.66; Rs at 2500 rpm: 0.69).

The effect of flow rate variations (from 3 to 30 mL/min) wasthen tested at 2500 rpm. Any tested conditions afforded resolu-tion between peaks superior to 1. Furthermore, increasing flow rateinduced a decreased resolution, and was correlated to a reductionof the stationary phase retention (Fig. d in supplementary data). Ourresults confirm data published by Faure’s group [21] who under-lined that the amount of liquid stationary phase retained in thecolumn decreases linearly as the flow rate of the mobile phaseincreases.

As column capacity of 50 mL could not afford Rs>1 between2 and 3, a separation on a 100-mL column has been performedin order to confirm that this issue is correlated to an equipmentlimitation. Rotation speed was set at 2500 rpm and flow rate at8 mL/min. The latter has been chosen because it represented areasonable compromise between resolution and run time. Herein,resolution between 7-homovaltrate and valtrate reached 1.13 using100 mL-column. Increasing column capacity significantly decreasecoelution between targeted analytes. At this point, the methodcalled “free space between peaks” could be applied for a scale-upto a 1000 mL-column.

3.3.2. Free space between peaks method descriptionLinear scale up on CPC instruments often underestimate CPC

large volumes load capacities. The method called “free-spacebetween peaks” described by Bouju E. et al (2015) proposes to cal-


culate a scale up transfer factor (Fsu) that corresponds to the freespace volume ratio (�V2/�V1) between two peaks thanks to differ-ent injections on both laboratory and preparative scale instruments(Fig. 4).



Fig. 3. Influence of rotation speed on resolution between homovaltrate [3], valtrate [2] and isovaltrate [1]. Data obtained on small scale rotor (SCPC-50) at fixed flow rate(5 mL/min).

e “free

mpSrwtwtat

o

�

Fig. 4. Illustration of the parameters used to determine th

Firstly, optimum chromatographic conditions have to be deter-ined on lab scale column in order to obtain a resolution between

eaks greater than 1 on the analytical injection. In our case, theCPC-100 rotor has been selected; a flow rate of 8 mL/min and aotation speed of 2500 rpm have been chosen. Then, two injectionsere performed: a small amount injection called “analytical injec-

ion”, and a high amount injection called “preparative injection”hich should represent the maximal load that can be separated on

he small rotor (Qmax.inj1). Criteria for both injections are resolutionnd retention factor of the stationary phase which should remainhe same.

The first �V1 value is calculated based on the analytical injectionn the lab scale column as follows:


V1 = �Vr1 − 2�A1 − 2�B1 (1)

space between peaks” volume on an analytical injection.

After that, an analytical injection is performed on the large-scalecolumn (operating conditions are set in order to work close to thelimit of pressure) in order to calculate �V2:

�V2 = �Vr2 − 2�A2 − 2�B2 (2)

Afterwards Fsu can be determined and applied on the small-scaleinstrument maximal load quantity to estimate the maximal loadcapacity of the large-scale instrument (Qmax.inj2).

Fsu = �V2�V1

(3)

Qmax.inj2 = Fsu × Qmax.inj1 (4)


The estimated maximal quantity is then injected on the large-scale rotor which is in our case a SCPC-1000.

To underline the importance of resolution between peaks, Eqs.(5) and (6) illustrate that if the resolution is lower or equal to 1, than


ARTICLE IN PRESSG ModelCHROMA-359768; No. of Pages 86 M. Clément Chami et al. / J. Chromatogr. A xxx (2018) xxx–xxx

nalytical and preparative injections on SCPC-1000.

tc

R

�

3

siaaetfmia

tocvwT3Ftipo

wtgptia

Table 3Determination of the Fsu factor using f̈ree space between peaks" parameters.

Analytical scale (100mL column capacity)

Preparative scale (1000mL column capacity)

Retention time [3] 22,18 min 47,62 minRetention time [2] 28,8 min 60,68 min2� [3] 2,55 min = 20,4 mL 4,73 min = 141,9 mL2� [2] 3,31 min = 26,48 mL 5,03 min = 150,9 mLVr [3] 177,44 mL 1428,6 mLVr [2] 230,40 mL 1820,4 mL�Vr 52,96 mL 391,8 mL

�V1=52,96-20,4-26,48=6,08mL

�V2=391,8-141,9-150,9=99mL

Fsu 16,3

Fig. 5. Overlay of ELSD chromatograms for a

he “free space between peaks concept” cannot be applied becausealculated volume would be negative or equal to 0.

s = �Vr2 (�A + �B)

(5)

V = �Vr − 2�A − 2�B = 2 (�A + �B) × (Rs − 1) (6)

.3.3. Scale up from a 100 mL-column to a 1000 mL-columnIn order to determine the maximal injected quantity on the

mall rotor (Qmax.inj1), 50 mg, 100 mg and 200 mg of extracts werenjected in 1 mL of mobile phase (Experimental chromatogramsvailable in Fig. e supplementary data). The resolution between 2nd 3 was around 1.1 for the 50 mg and the 100 mg injections. How-ver, stationary phase retention factor and pressure changed forhe 200 mg injection: stationary phase retention factor decreasedrom 71% to 68% and pressure raised from 67 bars to 75 bars (maxi-

al pressure on the SCPC-100 is fixed at 80 bars). Therefore, 50 mgnjection was considered as the “analytical injection” and 200 mgs the “preparative injection” also considered as Qmax.inj1.

Then, a small quantity of extract (500 mg) was injected onhe SCPC-1000 rotor which constitutes the “analytical injection”n the large volume column. For this purpose, chromatographiconditions were not optimized due to predictable high solventolumes consummation, so classical chromatographic conditionsere used on the large-scale CPC rotor (30 mL/min – 1200 rpm).

hanks to column capacity enhancement, resolution between 2 and significantly increased and reached a value around 1.4 (Fig. 5).urthermore, acquired ELSD chromatograms allowed Fsu calcula-ion (Table 3). So even if volume ratio between the two columnss 10, the method “free space between peaks” establishes that it isossible to inject 16.3 times more products which represents 3.26 gf extract (Qmax.inj2).

However, preparative injection on the 1000 mL rotor calculatedith Fsu factor is supposed to produce the same separation as

he preparative injection on the 100 mL-column: so a separationain is not expected. Retention time shifts between analytical and


reparative injections are the result of the charge effect caused byhe 20 m L-preparative injection (Fig. 5). The latter also induced anmportant stationary phase loss, which decreased to 50%, whichttest that the scale up reached the working limits of the SCPC-1000.

To highlight CPC performances Table 4 exhibits isolated quan-tities and compounds’ purities. 95% was set as the threshold toconsider the molecule as purified. Maximal reached purity is alsoindicated. CPC purification of 2 and 3 afforded compounds withpurity higher than 95% in a single step. From a single injection of3260 mg of crude extract, it was able to recover 1726 mg of 2 and196 mg of 3, which represent recovery yields around 90% for bothcompounds.

3.3.4. NMR structural elucidationStructures of valtrate and homovaltrate were elucidated using

NMR data and by comparison with the literature. Even thoughthe structure of valtrate has extensively been described [22], thestructure of homovaltrate may be subject to discussion since sev-eral structures are proposed [8,17,23]. In our case, the structureof 7-homovaltrate (Fig. 6) was confirmed thanks to the absenceof the doublet at � 0.92 ppm in the 1H NMR spectra which indi-cates that the isovaleryl group (C-13 to C-17) is positioned onthe C-9 and not on the C-6. Also, 1H -1H COSY and 1H -13CHMBC experiences confirmed the positioning of the sequence ofthe �-methylvaleryl group (C-18 to C-23) on the C-6 and partial


conformation is suggested based on 1H -1H NOESY spectrum (cf.Supplementary material).



Table 4Quantitative analysis of isolated valtrate and 7-homovaltrate on small scale (SCPC-100) and large scale (SCPC-1000) rotors.

4

rttehfco

u“itrttdw

bniq

A

dCr

[

[

[

[

[

[

[

[

Fig. 6. Structure and partial conformation of isolated 7-homovaltrate [3].

. Conclusion

In this work, one-step isolation strategy was set-up for efficientecovery of two structurally close valepotriates from C. ruber: val-rate and 7-homovaltrate. To the best of our knowledge, this ishe first report using CPC to purify valepotriates from plant crudextract. In less than one hour, the developed method providesighly purified compounds (>95%). Recovery yields reached 90%

or both compounds. Furthermore, compounds were structurallyharacterized by NMR enabling the confirmation of 7-homovaltrateccurrence in C. ruber.

The separation was firstly optimised on a 50 mL-capacity col-mn and then scaled-up on a 1000 mL-capacity column using thefree space between peaks” approach. Increasing column capac-ty enhances injection capacity (16.3 times more in comparison tohe maximal load capacity on the 100 mL-column) and improvesesolving power between targeted compounds. Nevertheless, dueo its isomer nature, isovaltrate could not be separated from val-rate. To overcome this issue, CPC could be performed in multiual-mode. Isovaltrate could be also purified from V. officinalis,here it is a major valepotriate [24], using the developed method.

Compared to conventional chromatographic strategies, CPC cane considered as an effective technique for purification of regulatedatural products at laboratory and preparative scales. It produces

n one step high-quality reference standards directly available foruality control.

cknowledgements


Authors greatly acknowledge the French Association Nationalee la Recherche et de la Technologie (ANRT) for supporting theIFRE partnership Botanicert-University of Nice Sophia Antipolis,eference number 2015/0222. Authors would like to thank the

[

European Research Institute on Natural Ingredients (ERINI) forHRMS analyses, Dr Marc Gaysinski for RMN interpretation supportand Olivier Gerriet from the Muséum d’Histoire Naturelle de Nicefor the voucher deposit at the herbarium. M.C.C. also thanks F. Fer-reira and A. Decarlis for their technical help in solvent system andextraction studies.

Appendix A. Supplementary data

Supplementary material related to this article can be found, inthe online version, at doi:https://doi.org/10.1016/j.chroma.2018.10.044.

References

[1] K.A. Shipkowski, et al., Naturally complex: perspectives and challengesassociated with botanical dietary supplement safety assessment, Food Chem.Toxicol. 118 (2018) 963–971, http://dx.doi.org/10.1016/j.fct.2018.04.007.

[2] A.C. Brown, An overview of herb and dietary supplement efficacy, safety andgovernment regulations in the United States with suggested improvements.Part 1 of 5 series, Food Chem. Toxicol. 107 (2017) 449–471.

[3] Arrêté du 24 juin 2014 établissant la liste des plantes, autres que leschampignons, autorisées dans les compléments alimentaires et les conditionsde leur emploi, J. Off. Républ. Franç aise (2014), http://dx.doi.org/10.1051/dmbd/120108.

[4] X. Fernandez, M. Clément Chami, L. Loffredo, Compléments alimentaires àbase de plantes - Évolutions réglementaires, Tech. l’ingénieur 33 (2016).

[5] X. Fernandez, M. Clément Chami, L. Loffredo, Compléments alimentaires àbase de plantes - Approche analytique, Tech. l’ingénieur 33 (2016).

[6] L. Mineo, et al., Valeriana officinalis Root Extract Modulates CorticalExcitatory Circuits in Humans, Neuropsychobiology 75 (2017) 46–51.

[7] E. Toolika, N. Bhat, S. Shetty, A comparative clinical study on the effect ofTagara (Valeriana wallichii DC.) and Jatamansi (Nardostachys jatamansi DC.)in the management of Anidra (primary insomnia), AYU (An Int. Q. J. Res.Ayurveda) 36 (2015) 46.

[8] R. Bos, H.J. Woerdenbag, N. Pras, Determination of valepotriates, J.Chromatogr. A 967 (2002) 131–146.

[9] L.J. El-Naggar, J.L. Beal, Iridoids. A review, J. Nat. Prod. 43 (1980) 649–707.10] Y.-Q. Wang, et al., Design, synthesis and biological evaluation of substituted

11H-benzo[a]carbazole-5-carboxamides as novel antitumor agents, Eur. J.Med. Chem. 46 (2011) 5878–5884.

11] S. Lin, et al., Minor valepotriates from Valeriana jatamansi and theircytotoxicity against metastatic prostate cancer cells, Planta Med. 81 (2015)56–61.

12] S. Lin, et al., Three minor valepotriate isomers from Valeriana jatamansi andtheir cytotoxicity, J. Asian Nat. Prod. Res. 19 (2017) 15–21.

13] P.C. Wang, et al., Iridoids and sesquiterpenoids from the roots of Valerianaofficinalis, J. Nat. Prod. 72 (2009) 1682–1685.

14] N. Fuzzati, et al., Isolation of antifungal valepotriates from Valeriana capenseand the search for valepotriates in crude valerianaceae extracts, Phytochem.Anal. 7 (1996) 76–85.

15] T. Michel, E. Destandau, C. Elfakir, On-line hyphenation of centrifugalpartition chromatography and high pressure liquid chromatography for thefractionation of flavonoids from hippophaë rhamnoides l. berries, J.Chromatogr. A 1218 (2011) 6173–6178.

16] E. Bouju, A. Berthod, K. Faure, Scale-up in centrifugal partitionchromatography: the ‘free-space between peaks’ method, J. Chromatogr. A1409 (2015) 70–78.

17] C. Wang, et al., Flexible and powerful strategy for qualitative and quantitativeanalysis of valepotriates in Valeriana jatamansi Jones using high-performance


liquid chromatography with linear ion trap Orbitrap mass spectrometry, J.Sep. Sci. 40 (2017) 1906–1919.

18] S. Lin, et al., Mass spectrometric profiling of valepotriates possessing variousacyloxy groups from Valeriana jatamansi, J. Mass Spectrom. 50 (2015)1294–1304.

https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044https://doi.org/10.1016/j.chroma.2018.10.044dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007dx.doi.org/10.1016/j.fct.2018.04.007http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0010dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108dx.doi.org/10.1051/dmbd/120108http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0020http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0025http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0030http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0035http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0040http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0045http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0050http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0055http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0060http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0065http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0070http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0075http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0080http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0085http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0090

ING ModelC8 hrom

[

[

[

[

[23] Franfois Gränicher, Philippe Christen, I. Kapétanidis, Production ofvalepotriates by hairy root cultures of Centranthus ruber DC, Plant Cell Rep.14 (1995) 294–298.

[24] G.B. Mahady, H.H.S. Fong, N.R. Farnsworth, Botanical Dietary Supplements:

ARTICLEHROMA-359768; No. of Pages 8 M. Clément Chami et al. / J. C

19] K. Hostettmann, J.-L. Wolfender, C. Terreaux, Modern screening techniquesfor plant extracts, Pharm. Biol. 39 (2001) 18–32.

20] Y. Ito, Golden rules and pitfalls in selecting optimum conditions forhigh-speed counter-current chromatography, J. Chromatogr. A 1065 (2005)


145–168.21] N. Fumat, A. Berthod, K. Faure, Effect of operating parameters on a centrifugal

partition chromatography separation, J. Chromatogr. A 1474 (2016) 47–58.22] A.M. Doyle, et al., Nature’s sedative: isolation and structural elucidation of

valtrate from Centranthus ruber, J. Chem. Educ. 81 (2004) 1486.

View publication statsView publication stats

PRESSatogr. A xxx (2018) xxx–xxx


Quality, Safety and Efficacy, Swets & Zeitlinger Publishers, 2001.

https://doi.org/10.1016/j.chroma.2018.10.044http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0095http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0100http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0105http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0110http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0115http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120http://refhub.elsevier.com/S0021-9673(18)31316-5/sbref0120https://www.researchgate.net/publication/328503953

Purification of two valepotriates from Centranthus ruber by centrifugal partition chromatography: From analytical to prepa...1 Introduction2 Experimental section2.1 Chemicals2.2 Plant material and extraction process2.3 UPLC-PDA-ESI-HRMS analysis2.4 External standard calibration2.5 CPC equipment2.6 Selection of the appropriate two-phase solvent system2.7 CPC separation procedure2.8 H and 13C RMN of isolated compounds

3 Results and discussion3.1 Extraction procedure3.2 Solvent system selection3.3 Analytical development methodology in CPC3.3.1 Determination of the laboratory scale conditions3.3.2 Free space between peaks method description3.3.3 Scale up from a 100 mL-column to a 1000 mL-column3.3.4 NMR structural elucidation

4 ConclusionAcknowledgementsAppendix A Supplementary dataReferences

g model article in press · please cite this article in press as: m. clément chami, et al.,...

Documents