Journal of Chromatography B, 812 (2004) 23–33

Review

Extraction methods and chemical standardization of botanicalsand herbal preparations

Eng Shi Ong

Applied Science School, Temasek Polytechnic, 21 Tampinese Avenue 1, Singapore 529757, Republic of Singapore

Received 4 April 2004; accepted 20 July 2004Available online 11 September 2004

Abstract

Botanicals and herbal preparations are medicinal preparations, containing a single or two or more medicinal plants. The focus of this reviewpaper is on the analytical methodologies, which included the combination of sample preparation tools and chromatographic techniques for thechemical standardization of marker compounds or active ingredients in botanicals and herbal preparations. The common problems and keychallenges in the chemical standardization of botanicals and herbal preparations were discussed. As sample preparation is the most importants strength andw plications ofc , weaknessa racts, usingH©

K

C

. . . 23. . 25

2728

. . 29. . . 32. . 32. .

1

t

inalac-l

lantsthe

1d

tep in the development of analytical methods for the analysis of constituents present in botanicals and herbal preparations, theeakness of different extraction techniques are discussed. For the analysis of compounds present in the plant extracts, the apommon chromatographic techniques, such as HPLC, CE, HRGC/MS, HPLC/MS and HPLC/MS/MS are discussed. The strengthnd applicability of various separation tools are stated. Procedures for the identification of marker or active compounds in plant extPLC/MS, were proposed. Finally, the effects of batch-to-batch variation of the medicinal plants are investigated and discussed.2004 Elsevier B.V. All rights reserved.

eywords: Sample preparation; Extraction; Botanicals; Herbal preparations

ontents

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. Preparation of sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. Analysis of botanicals and herbal preparation by liquid chromatography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. Analysis of botanicals and herbal preparations by capillary electrophoresis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. Hyphenation procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 32

. Introduction

Botanicals and herbal preparations are medicinal prepara-ions, containing a single plant or a mixture of two or more

E-mail address:esong@tp.edu.sg.

different types of medicinal plants. Monographs of medicplants can be found in the United States Pharmacopei[1],Chinese Pharmacopeia[2], WHO monographs for mediinal plants [3,4], Japanese Pharmacopeia[5] and HerbaMedicine (Expanded Commission E Monographs)[6]. Otherinformation on the chemical substances in medicinal pand their pharmacological properties can be found in

570-0232/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.jchromb.2004.07.041

24 E.S. Ong / J. Chromatogr. B 812 (2004) 23–33

Table 1Some of the common herbal products, their uses and side effects[3,6]

Botanicals Common medicinal uses Side effects

Aloe vera Short-term treatment of occasional constipation Abdominal spasms and pain may occur after even single dose.Overdose can lead to colicky abdominal spasms and pain, aswell as the formation of thin, watery stools.Overdose will result in electrolyte imbalance.

Echinacea To support and promote the natural powers of resistance of thebody, especially in infectious conditions, such as influenza and coldin the nose and throat.

Chills, short-term fever reaction and nausea and vomiting mayoccur.

External uses include promotion of wound healing and treatment ofinflammatory skin conditions.

St. John’s Wort For psychovegetative disturbances, depressive moods, anxiety andnervous unrest

Photosensitization is possible, especially in fair skin individuals.

Gingko biloba For symptomatic treatment of disturbed performance in organicbrain syndrome within the regimen of a therapeutic concepts, withthe following principal symptoms: memory deficients, disturbancesin concentration, depressive emotional conditions, dizziness,tinnitus and headache.

Very seldom, cases of stomach or intestinal upset, headaches orskin allergic skin reaction.

Galic As a support to dietary measures at elevated levels of lipids in theblood and as a measure for age dependant vascular changes.

In rare instances, there may be gastrointestinal symptoms,changes to the flora of the intestine, or allergic reaction.

Ginseng A tonic for invigoration and fortification times of fatigue anddebility or declining capacity for work and concentration.

None

Liquorice root For bronchitis, peptic ulcer, chronic gastritis, rheumatism andadrenocorticoid insufficiency.

On prolong use and higher doses, sodium and water retentionand potassium loss may occur, accompanied by hypertension,edema, hypokalemia and in rare cases, myoglobinuria.

books cited[7–10]. A list of commercially available products,containing medicinal plants, their common medicinal usesand side effects are tabulated inTable 1. The major com-pound types in Chinese medicinal plants include alkaloids,saponins, flavonoids, anthraquinones, terpenoids, coumarins,lignans, polysaccharides, polypeptides and proteins.

Botanicals, which show promising results included theuse of Ginkgo biloba in which the gingkolides have an-tioxidant, neuroprotective and chlolinergic activities relevantto Alzeimer’s disease[11–13]. The in vivo neuromodula-tory effects ofG. biloba was studied, using a product thatis a standardized leaf extract (EGb761) reported to contain24% flavone glycosides (primarily composed of quercetin,kaempferol and isorhamnetin)[13]. The inhibitory effects ofcancer cell proliferation and anti-tumor effects of the total ex-tracts fromScutellariae radixwere well studied. In an in vivoexperiment, significant inhibition of tumor growth had beenobserved.Scutellaria baicalensisinhibited cyclooxygenase-2 (COX-2) expression and a 66% reduction in tumor masswas observed in the nude mice.S. baicalensisselectively andeffectively inhibits cancer cell growth in vitro and in vivoand can be an effective chemotherapeutic agent for head andneck squamous cell carcinoma. It was proposed that the inhi-bition of PGE2 synthesis via suppression of COX-2 expres-sion may be responsible for its anticancer activity[14,15].T bs,s ed tot hropat kingw of a

manufacturing error, one of the medicinal plants in the pillswas replaced withAristolochia fangchiwhere the main com-ponents are aristolochic acids[17]. In Singapore, undeclareddrugs, such as chlorpheniramine, paracetamol and others canbe found in health supplements and herbal preparations[18].

At the same time, works with some botanicals have pro-gressed to clinical trials. A multicentric, open, prospective,observational and non-randomized clinical trial with 190 pa-tients for the efficacy and safety of a standardized phytoe-strogen preparation derived fromGlycine max(L.) Merr inClimacteric symptomatology was carried out[19]. A ran-domized controlled trial with 428 patients on the effectsof standardizedHypericum perforatum(St. John Wort) inmajor depressive disorder was reported[20]. A Phase II clin-ical trial on botanical explored standardized green tea’s an-tineoplastic effects in patients with androgen independentprostate carcinoma[21]. Other Phase II and III clinical trialssponsored by National Center for complementary and alter-native medicine (NCCAM) are progressing at the moment(http://nccam.nih.gov).

For botanicals and herbal preparations, there is a need toapproach scientific proof and clinical validation with chem-ical standardization, biological assays, animal models andclinical trials. Quality assurance of botanicals and herbalpreparations is the prerequisite of credible clinical trials. Ac-c tatesF -rv rposeo ons.F ater

he pharmacological basis of ‘Yang-invigoration’ in heruch asHerba epimedii, Rhizoma drynariaeand others weriscussed[16]. On the other hand, problems with regard

he usage of herbs were encountered. Chinese herb nephy or end stage renal failure was reported in patients taeight reducing pills containing Chinese herbs, because

-

ording to the draft guidelines stated by the United Sood and Drug Administration (USFDA)[22] and The Euopean Agency for the Evaluation of Medicinal Products[23],arious aspects of analysis must be performed for the puf certification of botanical drugs and herbal preparatior medicinal plants, the tests include identification, w

E.S. Ong / J. Chromatogr. B 812 (2004) 23–33 25

content, chemical assay of active ingredients, inorganic impu-rities (toxic metals), microbial limits, mycotoxins, pesticidesand others. For herbal preparations, in addition to the testsmentioned above, other tests include disintegration, dissolu-tion, hardness/friability and uniformity of dosage unit shouldbe available. However, due to the unique nature of botanicalsand herbal preparations, the USFDA found it appropriate toapply regulatory policies that was different from those ap-plied to synthetic, semi-synthetic, or otherwise highly puri-fied or chemically modified drugs. Clinical studies may beinitiated at phase II for common botanicals. The continu-ing drive to optimize health in today’s society has created ahuge market for dietary supplements where there is a growingneed to validate methods of analysis in botanicals and herbalpreparation[24].

The focus of this review paper is on the analytical method-ologies for the chemical standardization of marker com-pounds or active constituents in botanicals and herbal prepa-rations. According to draft guidelines stated by the USFDA,a marker compound is a chemical constituent of a botanicalraw material, drug substance, or drug product that is usedfor identification and/or quality control purposes, especiallywhen the active constituents are not known or identified. Theactive constituent is responsible for the intended pharmaco-logical activity or therapeutic effects. Chemical standardiza-t pico say)f Thek s forb arkero owne allyl certi-fi thoda thep opeds ger-p Witht ablet -to-b Duet sim-p ionsf ora-t lysiso n in-v anda ired.

s ofm toolsf asisi itiesa useT f tar-g ill bed iples

of these modern sample preparation tools, chromatographictechniques and hyphenated procedures, can be found in thereview articles and other papers cited in the respective sec-tions.

2. Preparation of sample

Sample preparation is the most important step in the devel-opment of analytical methods for the analysis of botanicalsand herbal preparations. Review papers on sample prepara-tion techniques for the extraction plant materials are available[25,26]. A brief summary of the experimental conditions forthe various methods of extraction stated in this paper can befound inTable 2. The basic operation included steps, such aspre-washing, drying of plant materials or freeze drying, grind-ing to obtain a homogenous sample and often improving thekinetics of analyte extraction. For the monograph stated inthe pharmacopeia, methods, such as sonication, heating un-der reflux, Soxhlet extraction and others are commonly used[1,2,5]. However, such methods can be time consuming, re-quired the use of large amount of organic solvent and mayhave lower extraction efficiencies. As the target compoundsmay be non-polar to polar and thermally labile, the suitabilityof the methods of extraction must be considered. The otherp t thea canta g oft ict vali-d cov-e diedi dif-f tiono venw rkerc tingc d ando

anics sam-p xtrac-t ler-a tion( ytesp en-e cantr in).O eed,M mp-t db tan-s ne I)f[ usedc used

ion often involves chemical identification by spectroscor chromatographic fingerprint and chemical assay (as

or active constituents or marker compounds if available.ey challenges in the development of analytical methodotanicals and herbal preparations are: (1) analysis of mr active compounds in a complex and sometimes unknnvironment, (2) target analytes may be polar and therm

abile, (3) lack of chemical reference substances anded reference materials, (4) selection of extraction mend (5) batch-to-batch variation of the composition oflant materials obtained. The analytical methods develhould be stability indicating, be used for chemical finrinting and assaying of marker or active compounds.

he chemical fingerprint obtained, the method should beo perform composition analysis and monitor the batchatch variation of the plant materials obtained for use.

o the number of parameters required, methods that arele, rapid and environmentally friendly are attractive opt

or regulatory laboratories, testing laboratories and labories supporting manufacturing activities. For the anaf botanicals and herbal preparations, a holistic solutioolving the combination of the methods of extractionnalytical techniques, such as separation tools are requ

In this article, recent developments and applicationodern sample preparation techniques and separation

or final analysis in plant extracts are reviewed. Emphs placed on the brief description of the unique capabilnd advantages and disadvantages of the approacheshe common problems and challenges in the analysis oet components in botanicals and herbal preparations wiscussed. More detailed description of the basic princ

d.

roblem for the selection of methods of extraction is thactive or marker compounds are present naturally, signifinalyte–matrix interaction will be present, hence spikin

he target compounds into the plant matrix will not mimhe real environment. Depending on how the method isated, it may be possible to have a method with high rery but lacks accuracy. From the various botanicals stu

n our laboratory, different methods of extraction andering conditions may often be required for the extracf marker compounds from different plant materials. Eith the same technique of extraction, for different maompounds in different plant materials, different operaonditions, such as the solvent use, temperature appliethers may be required.

In the move to reduce or eliminate the use of orgolvent and improve the extraction processes, newerle preparation methods, such as microwave assisted e

ion (MAE), supercritical fluid extraction (SFE) and acceted solvent extraction (ASE) or pressurized liquid extracPLE) have been introduced for the extraction of analresent in plant materials. Using MAE, the microwavergy is used for solution heating and results in signifieduction of extraction time (usually in less than 30 mther than having the advantage of high extraction spAE also enables a significant reduction in the consu

ion of organic solvents[25–27]. A novel MAE method haeen developed for the extraction and determination ofhinones (tanshinone IIA, cryptotanshinone and tanshinorom the root ofSalvia miltiorrhizawith analysis by HPLC28,29]. Other methods, such as the use of SFE thatarbon dioxide and some form of modifiers had been

26 E.S. Ong / J. Chromatogr. B 812 (2004) 23–33

Tabl

e2

Abr

ief

sum

mar

yof

the

expe

rimen

talc

ondi

tions

for

vario

usm

etho

dsof

extr

actio

n,su

chas

soni

catio

n,S

oxhl

etex

trac

tion,

MA

E,

supe

rcrit

ical

extr

act

ion,

AS

E,

PLE

,P

HW

Ean

dsu

rfac

tant

assi

sted

PH

WE

for

med

icin

alpl

ants

Son

icat

ion

Sox

hlet

extr

actio

nM

icro

wav

eas

sist

edex

trac

tion

(MA

E)

Sup

ercr

itica

lFlu

idex

trac

tion

(SF

E)

Acc

eler

ated

Sol

vent

extr

actio

n,st

atic

(AS

E)

Pre

ssur

ized

Liqu

idE

xtra

ctio

n,dy

nam

ic(P

LE)

Pre

ssur

ized

Hot

Wat

erE

xtra

ctio

n(P

HW

E)

Sur

fact

anta

ssis

ted

PH

WE

Com

mon

Sol

vent

sus

edM

etha

nol,

etha

nol,

orm

ixtu

reof

alco

hola

ndw

ater

Met

hano

l,et

hano

l,or

mix

ture

ofal

coho

land

wat

er

Met

hano

l,et

hano

l,or

mix

ture

ofal

coho

land

wat

er

Car

bon

diox

ide

orca

rbon

diox

ide

with

mod

ifier

s,su

chas

met

hano

l

Met

hano

lM

etha

nol

Wat

eror

wat

erw

ith10

–30

%et

hano

l

Wat

erw

ithsu

rfac

tant

s,su

chas

Trit

onX

100

orS

DS

Tem

pera

ture

(◦ C)

Can

behe

ated

Dep

endi

ngon

solv

entu

sed

80–1

5040

–100

80–2

0080

–200

80–3

0080

–200

Pre

ssur

eap

plie

dN

AN

AD

epen

ding

onif

itis

clos

edor

open

edve

ssel

extr

actio

n

250–

450

atm

100

bar

10–2

0ba

r10

–50

bar

10–2

0ba

r

Tim

ere

quire

d1

h3–

18h

10–4

0m

in30

–100

min

20–4

0m

in20

–40

min

40–5

0m

in40

–50

min

Volu

me

ofso

lven

treq

uire

d(m

l)50

–100

150–

200

20–5

0N

A20

–40

20–3

040

–45

40–4

5R

efs.

[25,

26]

[25,

26]

[25–

29]

[25,

26,3

0,31

][2

5,26

,32]

[34–

40]

[38,

41–4

6][4

5,48

]

in the extraction of compounds present in medicinal plants[25,26,30,31].

In the move to reduce the use of solvent and time in themethods of extraction, our laboratory have developed ana-lytical methods for botanicals and herbal preparations usingPLE. Methods using PLE, have been commonly used for theextraction of persistent organic pollutants in environmentalsamples. The technique uses organic solvent at elevated tem-perature and pressure, which drastically improves the speedof the extraction process. The parameters that have a sig-nificant effect on the extraction efficiencies of marker com-pounds in botanicals and herbal preparations by PLE are thetime for extraction/volume of solvent, applied temperatureand the nature of solvent used. The time for extraction was setat 20 min as it was found that a significant portion of the tar-get analytes would be extracted. The pressure was observedto have little effect on the extraction efficiency as it was ap-plied to keep the solvent in the liquid phase. Application ofPLE, using commercially available systems based on staticextraction to medicinal plants, was reported[32,33]. Our lab-oratory have reported the used of a laboratory made dynamicPLE system for the extraction of aristolochic acids, berberine,strychnine, ginsenosides, glycyrrhizin, baicalein and othersin medicinal plants and herbal preparations[34–38]. It wasobserved that the extraction efficiency of marker compoundsi para-b s fort et ex-t e theh ionr ture,t aalsf n so-l pted.T oids,s i-o s hadb

pliedp 00H cur-r pera-t d hotw f es-s ped[ 125a fra-gr opedf ithfi ro-m ort om-pim

n botanicals and herbal preparations by PLE was comle or higher than Soxhlet extraction. The main reason

he enhanced performance when using PLE over Soxhlraction and other conventional methods of extraction arigher solubility of analytes in solvent and higher diffusate as a result of higher temperature. At higher temperahe strong solute–matrix interaction caused by van der Worces, hydrogen bonding and dipole attractions betweeute molecules and active sites on the matrix were disruhe used of pressurized solvent extraction of polar steruch as withanolides from the leaves ofLochroma gesneridesand cocaine and benzylecgonine from coca leaveeen reported[39,40].

Water has similar properties as methanol at an apressure of 50 bar and a temperature between 150 and 2◦C.owever, it was observed that a number naturally oc

ing substances may decompose, using the stated temure. To reduce the use of organic solvent, pressurizeater extraction (PHWE) was used for the extraction oential oil components from plant materials was develo41,42]. Sub-critical water under pressure and betweennd 175◦C had been used to rapidly extract oxygenatedrance and favor compounds from rosemary[43]. A labo-atory made system using sub-critical water was develor the extraction of iridoid glycosides in plant matrix wnal determination by micellar-electrokinetic capillary chatography[44]. In our laboratory, PHWE was applied f

he extraction of thermally labile and reasonably polar conents, such as berberine inCoptidis rhizoma, glycyrrhizin

n Radix glycyrrhizae(liquorice), baicalein inS. radixandarker compounds inRadix codonopsis pilosula[38,45]. An

E.S. Ong / J. Chromatogr. B 812 (2004) 23–33 27

instrumental set-up similar to PLE can be used for PHWE.The parameters that would have a significant effect on theextraction efficiencies of marker compounds in botanicals,using PHWE, are the applied temperature, pressure and per-centage of organic modifiers. The non-significant effect ofthe pressure was observed.

However, due to the heat sensitivity and unique propertiesof certain marker compounds in botanicals, the extractionefficiencies of PHWE may not be as comparable to conven-tional methods, such as Soxhlet extraction. For components,such as tanshinone I and IIA inS. miltiorrhiza, PLE usingmethanol was found to give comparable or higher extractionefficiencies compared to PHWE with reference to Soxhletextraction[46]. For the analysis of marker compounds inR.codonopsis pilosula(DangShen), it was observed that PHWEgive lower extraction efficiencies compared to Soxhlet extrac-tion [45]. To improve the extraction efficiencies of methodsusing PHWE, surfactants, such as Triton X-100 and sodiumdodecyl sulfate (SDS) may be added into the water to enhancethe solubility of target compounds present in the botanicals.In this model of extraction, a high dilution of the extract wasproduced as higher extractant volume was used. With sur-factant assisted PHWE in a dynamic mode, the presence ofsurfactant, such as Triton X-100 in the water was found toenhance the solubility of the target compounds and pushedt se toc sam-p ona erbalm hanola Tri-t thee ec per-a 0 to2 undst , 12E-d ene-8 sp ritonX o thew us-i -i ctiono odels ex-t h fort priort

3l

ionr and

based on the number papers cited in this review for the anal-ysis of marker compounds that are thermally labile in botan-icals and herbal preparations. In most of the papers cited, thereversed octadecyl silica (C-18) is most commonly used. Inthe course of our experiments, we found that columns withsmaller inner diameter, such as 1.0 or 2.1 mm i.d. were wellsuited to the analysis of components present in botanicals. Forcolumns with smaller inner diameter, it was observed that thesystem precision for the retention time and peak area/heightwere comparable to analytical columns with 4.6 mm i.d.[46].Most important of all, methods using columns with smallerinner diameter and the right mobile phase can be readilyadopted to mass spectrometry. The most common mode ofdetection remains to be ultraviolet detection. Applications ofchromatographic techniques to medicinal plants and Chinesetraditional medicines are outlined[49]. Methods using gra-dient elution HPLC with reversed phase columns had beenapplied for the analysis of multiple constituents present inmedicinal plants and herbal preparations[50–52]. Gradientelution HPLC with ultraviolet detection, using a C18 reversedphase column had been used to profile components presentin C. rhizoma, Radix aristolochiae, ginseng,R. glycyrrhizae(liquorice),S. radix, R. codonopsis pilosulaandS. miltior-rhiza [34,36,38,45,46]. The advantages of liquid chromatog-raphy include its high reproducibility, good linear range, easeo on-s r, fort witht fteno com-p yb ch asl tiveL

vel-o ns int is ofm -d nts ly-s ap rod-u vari-a tiong odw idesi esultsdos dt rvedt ointo ncesi

lantm uid

he target analytes in the sample matrix in the mobile phaompleteness with the fresh liquid pumped through thele continuously[45]. Hence, micelle-mediated extractind preconcentration of ginsenosides from Chinese hedicine had been introduced. When compared to metnd water, an aqueous surfactant solution containing 10%

on X-100 yielded faster kinetics and higher recovery forxtraction of various ginsenosides[47]. A method, using thombination of surfactant and PWHE with an applied temture below the boiling point and lower pressure from 10 bars was developed for the analysis of marker compo

hat are reasonably hydrophobic, such as tetradeca-4Eiene-8,10-diyne-1,6,7-triol and tetradeca-4E, 12E-di-10-diyne-1,6,7-triol-O-�-d-glucoside in R. codonopsiilosula(DangShen). The type of surfactants, such as T-100, sodium dodecyl sulfate (SDS) and others added tater will affect the extraction efficiencies of the methods

ng surfactant assisted PHWE[45]. Similarly, PLE with nononic surfactants solution had been applied for the extraf ginsenosides from the roots of American ginseng as molid samples. The combination of PLE and cloud pointraction was shown to be a new and effective approache rapid sample pre-concentration of herbal materialso analysis by HPLC[48].

. Analysis of botanicals and herbal preparation byiquid chromatography

Liquid chromatography with a isocratic/gradient elutemain to be the method of choice in the pharmacopeia

f automation and its ability to analyze the number of ctituents in botanicals and herbal preparation. Howevehe analysis of marker compounds in herbal preparationswo or more medicinal plants, co-eluting peaks were observed in the chromatograms obtained due to thelexity of the matrix[54]. The complexity of matrix mae reduced with additional sample preparation steps, su

iquid-liquid partitioning, solid phase extraction, preparaC and TLC fractionation.

One of the challenges for the analytical methods deped was to study the effects of batch-to-batch variatio

he medicinal plants. It was reported that HPLC analysarker compounds inPlatycodi radixof Playcodon graniforumshowed variation in the composition from differeources[55]. With HPLC methods for simultaneous anais of cichoric acids and alkamides inEchinacea purpurelant and products, a selection of international herbal pcts available on the Danish market show surprisingble quality, not necessary reflecting the product informaiven on the labels[56]. A reversed phase HPLC methith ultraviolet detection was used to identify ginsenos

n products classified as health supplements where the rid not reflect what is given on the label[36]. Similarly, inne product labeled as containing purePanax ginseng, no gin-enosides could be detected[8]. For ginseng, the variety anhe method of cultivation are very important. It was obsehat various species of ginseng, from a morphological pf view, are similar, but as far as their content of substa

s concerned they are not[8].To investigate the batch-to-batch variations of the p

aterials obtained, a method using PLE with liq

28 E.S. Ong / J. Chromatogr. B 812 (2004) 23–33

Table 3Comparison ofRadix scutellariaeextracted from different batches of medic-inal plants by PLE at 120◦C with HPLC[31]

Batch No. Amount of Baicalein (mg/g) EC50

1 20.65± 0.20 18.13± 0.752 21.47± 0.28 17.63± 1.653 12.80± 0.16 11.88± 0.254 19.34± 0.16 17.00± 0.58

For the determination of EC50, sample stock solutions were diluted to fivedifferent concentrations in ethanol. One ml of a 0.3 mM DPPH ethanol so-lution was added to 2.5 ml of sample solutions of different concentrationand allowed to react at room temperature. After 30 min the absorbent valueswere measured at 518 nm, using a Jasco V-530 LSE 1335 UV Spectropho-tometer (Japan) and converted into the percentage antioxidant activity (AA),using the following formula: AA% = 100− {[(Abssample− Absblank) ×100]/Abscontrol}. Ethanol (1.0 ml) plus plant extract solution (2.5 ml) wasused as a blank. DPPH solution (1.0 ml; 0.3 mM) plus ethanol (2.5 ml) wasused as a negative control. The positive controls were those using the stan-dard solutions. EC50 represents the level where 50% of the radical werescavenged by test sample.

chromatography with ultraviolet detection[38] was used toassay the content of baicalein inS. radixfrom four differentsources labeled as Batch 1–4 inTable 3. As shown inTable 3,batch-to-batch variations were observed with the amount ofbaicalein varied from 12.8 to 21.47 mg/g. A chromatographicfingerprint of various components present inS. radix wasshown inFig. 1. A similar chemical fingerprint as inFig. 1was obtained for all the four different batches ofS. radixpro-filed. On top of the chemical fingerprint, the antioxidant activ-ity, using the 1,1-diphenyl-2-picryl-hydrazyl (DPPH) radicalphotometric assay in the process guided by its discoloration,was measured. However, the level of baicalein and the an-tioxidant activities was found to vary in the four differentbatches ofS. radix. The anti-oxidant activities measured were

FM alc forer col-u ent

found to correlate with the amount baicalein and other com-ponents present inS. radix. Hence, the assayed of marker oractive compounds together with chemical fingerprinting, us-ing HPLC, will be able to provide further information aboutthe quality of the botanicals and herbal preparations.

4. Analysis of botanicals and herbal preparations bycapillary electrophoresis

Capillary electrophoresis (CE) proved to be a powerful al-ternative to HPLC in the analysis of polar and thermally labilecompounds. Reviews on the analysis of natural medicinesor natural products in complex matrix by CE are well re-ported[56,57]. Many publications showed that all aspects ofCE, such as capillary zone electrophoresis (CZE), micellar-electrokinetic capillary chromatography (MEKC) and cap-illary isoelectric focusing (cIEF) have been used for theseparation of natural products. The separation in CZE is basedon the differences in the electrophoretic mobilities resultingin different velocities of migration of ionic species in the elec-trophoretic buffer in the capillary. For MEKC, the main sepa-ration mechanism is based on solute partitioning between themicellar phase and the solution phase. Factors that are knownto affect separation in CZE and MEKC include pH of runningb andt

minet phe-n ds inp dsu s inR -i nR ob n offM bea say-i cinalp

ob-s inalp lysiso uctst CM)o f then ara-t ob-t plexe fort ion,g hi-ne ribedf and

ig. 1. HPLC Chromatogram obtained for baicalein inS. radixby PLE.obile phase of (A) 25 mM NaH2PO4 at pH 2.5 and (B) acetonitrile. Initi

ondition was set at 30% of B, gradient up to 100% B in 15 min beeturning to initial condition for 10 min. C-18 reversed phase HPLCmn (4.6 mm i.d.× 150 mm length, 5�. Detection was at 254 nm. Ov

emperature was at 40◦C and flow rate was 1.0 ml/min.

uffer, ionic strength, applied voltage and concentrationype of micelle added.

From the review articles, CE had been used to deterhe amount of catechin and others in tea composition,olic acids in coffee samples and flavonoids and alkaloilant materials[56,57]. In our laboratory, analytical methosing CE, included the determination of aristolochic acid. aristolochiae, strychnine inStrychnos nux-vomica, berber

ne in Rhizoma coptidis(Huang-lian) and glycyrrhizin i. glycyrrhizae(liquorice) [35,37,53]. CZE was found te a method for the rapid separation and determinatio

our phenylpropanoid glycosides fromT. chamaedrys[58].ethods using CZE and MEKC have been proved topowerful technique in the chemical profiling and as

ng of marker compounds present in extracts from medilants.

As herbal preparations are products that were oftenerved to contain two more different types of mediclants, it had always been a challenge for the anaf target compounds in herbal preparations and prod

hat are classified as, Traditional Chinese Medicine (Tr herbal remedies. It is certain that the presence oumber of medicinal plants used in the herbal prep

ions will increase the complexity of the sample matrixained. In the analysis of marker compounds in comnvironment, methods using MEKC had been applied

he determination of icariin, rhein, chrysophanol, physclycyrrhetic acid and glycyrrhizic acid in traditional Cese herbal preparations[59]. The use of MEKC withlectrochemical or ultraviolet detection had been desc

or the determination of hesperidin, naringin, puerarin

E.S. Ong / J. Chromatogr. B 812 (2004) 23–33 29

daidzein in medicinal preparations[60,61]. To differentiatebetween medicinal plant, such asS. radix from Astragaliradix, CE with electrochemical detection was employed[62].

For the determination of berberine, aristolochic acids Iand II in herbal preparations, we reported that it was possibleto assay the amount of marker or active compounds with asingle step extraction followed by CZE. This was due to thefact that separation by CZE is mainly based on the differentelectrophoretic mobility of charged solutes in solution in re-lation to their different molecular weight, size and charge ata given pH. It is also possible for both cations and anionsto be separated in a single run which provides specific se-lectivity for the separation method. It was observed that itcan be difficult to assay the amount of marker compoundsin herbal preparations using HPLC as co-eluting peaks wereoften present. With CZE and under the conditions used foraristolochic acids, it was found that most of the positivelycharged and neutral species migrate close or with the electro-osmotic flow (EOF). The amount of aristolochic acid I andII present in herbal preparations, that was in a complex en-vironment, could be determined with a single step extractionfollowed by CZE[35,53].

The advantages of methods using CE for botanicalsand herbal preparations are: (1) the high selectivity anda on-m sam-p letee iono au-t tec-t outu cificp en-s ther th-o ever,t eaka aredt ail-u dards anol.T in-s non-c thei nol,1 s andss orts[

usingC nces,s vesti-t inedtp will

be more difficult for compounds that are essentially non-polarpresent in some medicinal plants.

5. Hyphenation procedures

The use of chromatographic separation with mass spec-trometry for the chemical characterization and compositionanalysis of botanicals has been growing rapidly in the re-cent years. Reviews on the use of mass spectrometry andhigh-performance liquid chromatography mass spectrome-try (HPLC/MS) on botanicals had been reported[68,69].The use of hyphenated techniques, such as high resolu-tion gas chromatography mass spectrometry (HRGC/MS),high performance liquid chromatography/mass spectrome-try (HPLC/MS), liquid chromatography tandem mass spec-trometry (HPLC/MS/MS) and tandem mass spectrometry(MS/MS) to perform on line composition and structural anal-yses provide rich information that is unsurpassed by othertechniques.

The use of HRGC/MS, remains the method of choice forthe analysis of volatile and semi-volatile components, suchas essential oil and others in botanicals and herbal prepa-rations. With high resolution separation with capillary col-umn coupling with mass spectrometry using electron impacti ine,p healths easily[ sucha -t asss ea mi-v ractsf ith-o

cals,H rolea undst ulest ratesa n inH mo-s yzer,s -of –fl ed.W for-m ever,m hata pH,s rdert

eenu one I,t[ -

bility to analyze target compounds in complex envirent, such as herbal preparations without additionalle preparation steps, (2) significant reduction or complimination of organic solvent for the electrolyte solutr running buffer, (3) good linear range, (4) ease of

omation and (5) speed of analysis. Using ultraviolet deion, on-column detection was often used in CE. Withsing bubble capillary columns and applying any spere-concentration techniques, this will result in lower sitivity compared to methods using HPLC. To improveeproducibility of migration time and peak area with meds using CE, internal standards are often added. How

he system precision for the relative migration time and prea will always be higher in methods using CE comp

o HPLC. It was observed in our laboratory that run fres were encountered in CZE and MEKC where stanolutions were prepared in high percentage of methhe effect was due to a disruption of the conductivityide the capillary caused either by the presence of aonductivity plug or the result of the out-gassing innjection zone. As the sample was extracted with metha0% water in methanol was added to standard solutionample extracts to counter run failures[35,37,53]. Our ob-ervations on run failure, was consistent with other rep63,64].

Despite of some the weakness described for methodsE, the enantioseparation of naturally occurring substauch as atropine and flavonoids, such as medicarpin andone in plant materials by CE with chiral selectors remao be the method of choice[65–67]. From our works withlant materials, the applicability of methods using CE,

onization (EI), undeclared drugs, such as chlorpheniramaracetamol and others that may be present or added inupplements and herbal preparations can be detected18]. Other than undeclared drugs, marker compounds,s tetrahydrompalmatine inCorydalis yanhusuocan be iden

ified using HRGC/MS with the chromatograms and the mpectra as shown inFig. 2. The power of HRGC/MS is thbility to rapidly detect compounds that are volatile or seolatile present in a complex environment, such as extrom herbal preparations using a single step extraction wut any additional clean-up steps.

For the analysis of components present in botaniPLC/MS has been playing a increasingly significants the technique is capable of characterizing compo

hat are thermally labile, ranging from small polar moleco macromolecule, such as peptides/proteins, carbohydnd nucleic acids. The most common mode of ionizatioPLC/MS included electrospray ionization (ESI) and atpheric pressure chemical ionization (APCI). Mass analuch as single quadruple, triple quadruple, ion-trap, timeight, quadruple time-of-flight (Q-TOF) and others are usith tandem mass spectrometry, additional structural ination can be obtained for the target compounds. Howethods using HPLC/MS is still limited to conditions tre suitable for MS operations. There are restrictions onolvent choice, solvent additives and flow rate for LC in-oo achieve optimal sensitivity.

In our laboratory and other report, HPLC/MS had bsed to characterize ginsenosides in ginseng, tanshin

anshinone II and cryptotanshinone present inS. miltiorrhiza46,68]and marker compounds present inR. codonopsis pilo

30 E.S. Ong / J. Chromatogr. B 812 (2004) 23–33

Fig. 2. A Total ion chromatograms (TIC) obtained for marker compounds, such as tetrahydrompalmatine inC. yanhusuowith HRGC/MS and 2B) mass spectrawith electron impact (EI) ionization obtained for tetrahydropalmatine. Injector: splitless at 250◦C. Carrier: helium. Oven: initial at 80–300◦C at 15◦C/min.Detector: Mass selective detector, 280◦C, full scan: 40–600 amu. Column: HP5, 30 m× 0.25 mm i.d., 0.25�m.

sula[45]. The presence of non-volatile components, such astanshinone I, tanshinone II and cryptotanshinone inS. miltior-rhizacan be easily identified by positive ion ESI/MS. Directsample introduction by ESI or separation with reversed phasecolumn resulted in the abundant of [M + H]+, [M + Na]+ and[M + M + Na]+ for tanshinone I, tanshinone II and cryptotan-shinone[46,68]. For compounds, such as tetradeca-4E, 12E-diene-8,10-diyne-1,6,7-triol and tetradeca-4E, 12E-diene-8-10-diyne-1,6,7-triol-O-�-d-glucoside in R. codonopsispilosula(DangShen), [M + Na]+ was observed to be the mostabundant ion[45]. With ESI/MS, there is a great tendency forcertain compounds in medicinal plants to form adducts, suchas [M + Na]+ or [M + K] +.

Methods using HPLC/ESI/MS had been applied for theidentification components present inBetula pubescenand

Peschiera fuschiaefolia. [70,71]. For the characterizationof hydrolysable tannins from leaves ofB. pubescen, themain peak in the mass spectra of an individual hydrolysabletannins in the negative ion ESIMS, was the deprotonatedmolecule [M − H]− [70]. The alkaloids, such as voa-camine, voacamidine, vobasine and voachalotine present inP. fuschiaefoliawere determined using external standardcalibration with acceptable precision and accuracy. The pres-ence of [M + H]+, [M + 2H]2+ and other adduct ions wereobserved in the positive electospray ionization mode[71].Using APCI interface with mass analyzer, such as ion trapor triple quadruple instruments, main constituents present inRhodiola roseaextracts and tropane alkaloids fromErythrox-ylum vacciniifoliumcan be characterized[72,73]. With in-ternal standard calibration, constitutents, such as salidroside,

E.S. Ong / J. Chromatogr. B 812 (2004) 23–33 31

Fig. 3. A Total ion chromatogram (TIC) obtained for glycyrrhizin in themethanolic extractR. glycyrrhizaeusing HPLC/ESI/MS and 3B MS/MSof glycyrrhizin with precursor ionm/z: 839.1 amu. The gradient elution,using a mobile phase consisting of (A) 0.1% formic acid in water and (B)0.1% formic acid in acetonitrile. The initial condition was set at 20% of B,gradient up to 100% in 20 min and returning to initial condition for 6 min.Oven temperature was set at 40◦C and flow rate was set at 0.2 ml/min. For allexperiments, 5�l of standards and sample extract were injected. The columnwas a reversed phase C18 Luna, 150 mm× 2.0 mm, 5�m (Phenomenex,USA). The ESI-MS was acquired in the positive ion mode. The scanningmass range was from 100 to 1000 amu. The heated capillary temperature wasmaintained at 350◦C, the drying gas and nebulizer nitrogen gas flow rateswere 10 l/min and 50 psi, respectively. Data was acquired using automatedMS/MS.

rosavin and benzyl-O-�-glucopyranoside inR. roseaextractswas determined[73]. Similarly, the content of ginsenosides,such as Rb1, Rc and Re in plant extracts fromP. ginsengandPanax quinquefoliuswas determined by HPLC/ESI/MS withexternal standard calibration. The [M + H]+ and [M + Na]+ions were observed for ginsenosides standards (Rb1, Rb2, Rc,Rd, Re, Rf and Rg1)[74]. In our earlier work, it was observedthat it would be difficult to differentiate betweenP. ginsengandP. quinquefoliususing reversed phase HPLC with ultra-violet detection[36]. With HPLC/ESI/MS, the presence andratio of ginsenosides Rf and 24-(R)-pseudoginsenosides F11in the plant extracts can be used to differentiate betweenP.ginsengandP. quinquefolius[75,76].

From Figs. 3A and 4A, the total ion chromatograms(TIC) of glycyrrhizin and baicalein in the methanolic ex-tracts fromR. glycyrrhizaeandS. radixwere obtained, usingHPLC/ESI/MS. With tandem mass spectrometry (MS2), sig-nificant fragmentation pattern was observed for glycyrrhizinand baicalein as inFigs. 3B and 4B. In most instances, signif-icant fragmentation patterns can be obtained for a number ofcompounds in plant materials with MS2. However, for certaincompounds, such as baicalin inS. radix, significant fragmen-tation pattern was not observed with MS2 as inFig. 4C. Asbaicalin and baicalein differs from each other by one side

Fig. 4. A Total ion chromatogram (TIC) obtained for baicalein in themethanolic extracts fromS. radixusing LC/ESI/MS and 4B) MS/MS ofbaicalein with precursor ionm/z: 270.7 amu, 4C) MS/MS spectra of baicalinwith precursor ionm/z: 446.6 amu. Conditions used were the same as inFig. 3.

chain, MS/MS of balcalin with precursor ionm/z: 470.7 gavethe molecular ion of baicalein as inFig. 4C. With MS3, thefragmentation pattern for baicalein as inFig. 4B was observedfor baicalin. The current example showed the ability of MS2

and MS3 in the structural elucidation of naturally occurringcompounds in botanicals. For certain compounds, MS2 maynot be able to provide as much information compared to MS3.

The strength of HPLC/MS and HPLC/MSn remained inits ability to determine the presence of target compounds incomplex environment, such as in herbal preparations. Thiswas seen in the evaluation of major active components inSt. John’s Wort dietary supplements with HPLC/ESI/MS.The content of the target components, such as rutin, hy-perosides, quercetin and others were determined using in-ternal standard calibration. The presence of [M + H]+ and[M + Na]+ ions were observed for mass spectra obtainedfor rutin and hyperosides[77]. As a result of end stage

32 E.S. Ong / J. Chromatogr. B 812 (2004) 23–33

renal failure reported in patients taking weight reducing pillscontaining Chinese herbs that was found to contain aris-tolochic acids, the presence of aristolochic acids in a numberof herbal remedies was screened and detected with meth-ods using HPLC/ESI/MS and HPLC/APCI/MS[78]. Othergroups have also used HPLC/MS to confirm the presence ofaristolochic acids in Chinese phytomedicines and dietary sup-plements used as slimming regiments[79,80]. Using selectiveion monitoring (SIM) or extracted ion chromatogram at a par-ticularm/z, the mass analyzer can become a highly selectivedetector for the detection of target components in complexenvironment, such as herbal preparations. An HPLC/ESI/MSmethod had been developed for an estimation of the saponincontent in food supplement containingTribulus terrestis L.[81]. In conjunction with an AOAC task group on dietarysupplement, a method using liquid chromatography/tandemmass spectrometry was validated for measurement of six ma-jor alkaloids in rawEphedra sinicaherb,Ephedraextractsand health supplement products[82]. In both methods, in-ternal standard was added to determine the content of targetcompounds present.

For the identification of substances present in botanicalsand herbal preparations using HRGC/MS or HPLC/MS, theauthors’ find that the following conditions are useful whenstandards are available. These include: (1) a suspect peakh timeo ectraf( darda ts ana-l s tot om-p gherc nal-y thata Foro se-l 1.0oo ol-u asss ratea ucha eth-o Thec usedi ited.F ddi-t Ref.[ ntainh lyzeri asss ofa do

6. Conclusions

Chemical analysis of extracts from plant material willplay a central role in the development and modernization ofChinese medicines, botanicals and herbal preparations. Themethod used is required to identify the active or marker com-pounds, composition analysis and fingerprinting purposes.For the chemical standardization of botanicals and herbalpreparations, the method involved sample preparation pro-cedures, such as techniques of extraction and other analyt-ical techniques, such as the separation tools. However, dueto complexity of the plant materials obtained, different meth-ods using different methods of extraction and separation toolsmay be needed. It is certain that methods that are simple, rapidand environmentally friendly will be preferred and is likely toplay an important role in the effort in providing high qualityproducts to consumers worldwide. Just as for pharmaceuti-cal products, a well-validated method will help in monitoringthe stability of the botanicals and herbal preparations over aperiod of time. It is likely that the combination of chemi-cal standardization with biological assay will provide furtherknowledge about therapeutic effects of the medicinal plant.The use of validated methods in the chemical standardizationof botanicals and herbal preparations will enhance the qual-ity of the products, assist in pharmacological studies, performc nceb

A

ofN ni-v

R

, NF002., The

ub-

ub-

iety of

Ex-ncil,

s onntific,

har-

rcourt

as to show retention time similar to average retentionf the pure standard or control sample and (2) mass sp

or the suspect peaks has to show relative abundance± 10%arithmetic difference) of the relative abundance of stannalyzed that day[83]. With HPLC/MS, applying the righeparation, with the right ionization interface and mass

yzer, significant information can be obtained with regardhe target compounds. However, for the quantitation of counds in plant materials, the system precision will be hiompared to HPLC with ultraviolet detection. For the asis of target compounds in botanical extracts, systemsllowed for gradient elution HPLC are often required.n-line HPLC/MS, the internal diameter of the column

ected will be an important consideration. The used ofr 2.1 mm i.d. columns with flow rate at 100–200�l/minr lower is optimal for most ESI/MS operation. Post cmn splitter is often needed to reduce LC flow into the mpectrometer when using 4.6 mm i.d. columns with flowt 1.0 ml/min. On top of that, the common buffers, ss phosphate buffer and others used in most HPLC mds cannot be adopted for methods using HPLC/MS.hoice of buffers and ion pairing reagents that can ben the proposed methods using HPLC/MS are more limor further information of the suitable of solvents and a

ives in HPLC/MS, the readers can refer to an excellent84]. From our experiences, samples that are found to coigh salt content will create problems for the mass ana

n most HPLC/MS system. Finally, interpretations of mpectra from ESI/MS will be more tricky in the presencedduct ions, such as [M + Na]+, [M + 2H]2+, dimmers anthers.

redible clinical trials and propel the move towards evideased medicine.

cknowledgements

The author would like to thank A/P Swee Ngin Tanational Institute of Education, Nanyang Technological Uersity for reading the manuscript.

eferences

[1] United States Pharmacopeia and National Formulary, USP 2519, United States Pharmacopeial Convention Inc., Rockville, 2

[2] Pharmacopoeia of the People’s Republic of China, English ed.Pharmacopeia Commission of PRC, Beijing, 2000.

[3] WHO Monographs on Selected Medicinal Plants, vol. 1, WHO Plications, Geneva, 1999.

[4] WHO Monographs on Selected Medicinal Plants, vol. 2, WHO Plications, Geneva, 2002.

[5] The Japanese Pharmacopeia, fourteenth ed., JP XIII, The SocJapanese Pharmacopeia, Japan, 2001.

[6] M. Blumenthal, A. Goldberg, J. Brinckmann, Herbal Medicine,panded Commission E Monographs, American Botanical CouAustin, 2000.

[7] N.J. Brown, D.E. Brown, Chemicals from Plants, PerspectivePlant Secondary Products, Imperial College Press, World ScieLondon, 1999.

[8] S.J. Cutler, H.G. Gutler, Biologically Active Natural Products: Pmaceuticals, CRC Press, Florida, 1999.

[9] W.C. Evan, Trease and Evans Pharmacognosy, fifteenth ed., HaPublisher, London, 2002.

E.S. Ong / J. Chromatogr. B 812 (2004) 23–33 33

[10] B.G. Katzung, Basic and Clinical Pharmacology, eighth ed., McGrawHill, New York, 2001 (Chapter 65).

[11] E.K. Perry, A.T. Pickering, W.W. Wang, P.J. Houghton, N.S. Perry,J. Pharm Pharmacol. 51 (1999) 517.

[12] D. Pratico, N. Delanty, Am. J. Med. 109 (2000) 577.[13] C.M.H. Watanabe, S. Wolffram, P. Ader, G. Rimbach, L. Packer, J.J.

Maguire, P.G. Schultz, K. Gohil, PNAS 98 (2001) 6577.[14] S. Ikemoto, K. Sugimura, N. Yoshida, R. Yasumoto, S. Wada, K.

Yamamoto, T. Kishimoto, Urology 55 (2000) 951.[15] D.Y. Zhang, J. Wu, F. Ye, S. Jiang, J. Yi, W. Zhang, H. Wei, M.

Sung, W. Wang, X. Li, Cancer Res. 63 (2003) 4037.[16] K.M. Ko, D.H.F. Mak, P.Y. Chiu, M.K.T. Poon, Trends Pharmacol.

Sci. 25 (2004) 3.[17] J.L. Nortier, M.C. Martinez, H.H. Schmeiser, V.M. Arlt, C.A. Bieler,

M. Petein, M.F. Depierreux, L. De Pauw, D. Abramowicz, P. Vereer-straeten, J.L. Vanherweghem, N. Engl. J. Med. 342 (2000) 1686.

[18] H.L. Koh, S.O. Woo, Drug Saf. 23 (5) (2000) 351.[19] A. Albert, C. Altabre, F. Baro, et al., Phytomedicine 9 (2002) 85.[20] Hypericum Depression Trial Study Group, et al., JAMA 287 (2002)

1807.[21] A. Jatoi, N. Ellison, P.A. Burch, et al., Cancer 97 (2003) 1442.[22] Guidance for Industry, Botanical Drug Products, draft guidance,

CDER/USFDA, 2000.[23] CPMP/QWP/2820/00 (EMEA/CVMP/815/00), Note for guidance on

specifications: test procedures and acceptance criteria for herbaldrugs, herbal drug preparations and herbal medicinal products,EMEA, 2001.

[24] R.C. Thompson, S. Morris, LCGC 19 (2001) 1142–1149.[25] C.W. Huie, Anal. Bioanal. Chem. 373 (2002) 23.[26] B. Zygmunt, J. Namiesnik, J. Chromatogr. Sci. 41 (2003) 109.[[[[ nal.

[[ 99)

[ 501

[ 57.[[ 02)

[[[ Sep.

[ 001)

[ astro,

[ hro-

[ od

[ r. A

[[[ r. A

[ ogr.

[49] F. Li, S. Sun, J. Wang, D. Wang, Biomed. Chromatogr. 12 (1998)78.

[50] S. Dheu, K. Li, J. High Resol. Chromatogr. 21 (1998) 569.[51] S. Li, S.S.K. Chan, G. Lin, L. Ling, R. Yan, H.S. Chung, Y.K. Tam,

Planta Medica 69 (2003) 445.[52] X. Zhang, M. Yu, J. Chen, J. Chromatogr. Sci. 41 (2003) 241.[53] E.S. Ong, S.O. Woo, Electrophoresis 22 (2001) 2236.[54] T. Saeki, K. Koike, T. Nikaido, Planta Medica 65 (1999) 428.[55] P. Molgaard, S. Johnsen, P. Christensen, C. Cornett, J. Agric. Food

Chem. 51 (2003) 6922.[56] H.J. Issaq, Electrophoresis 18 (1997) 2438.[57] H.J. Issaq, Electrophoresis 20 (2000) 3190.[58] B. Avula, R.B. Manyam, E. Bedir, I.A. Khan, Chromatographia 58

(2003) 751.[59] H.T. Liu, K.T. Wang, H.Y. Zhang, X.G. Chen, Z.D. Hu, Analyst 125

(2000) 1083.[60] Y. Cao, C. Lou, X. Zhang, Q. Chu, Y. Fang, J. Ye, Anal. Chim.

Acta 452 (2002) 123.[61] S.G. Zhao, X.G. Chen, Z.D. Hu, Chromatographia 57 (2003) 593.[62] G. Chen, X. Ying, J. Ye, Analyst 125 (2000) 815.[63] M.A. Kelly, K.D. Altria, B.J. Clark, J. High Resol. Chromatogr. 21

(1998) 193.[64] M.A. Kelly, K.D. Altria, B.J. Clark, J. High Resol. Chromatogr. 22

(1999) 55.[65] S. Tahara, A. Okayama, Y. Kitada, T. Watanabe, H. Nakazwaw, K.

Kakehi, Y. Hisamatu, J. Chromatogr. A 848 (1999) 465.[66] D.J. Allen, J. Gray, N.L. Paiva, J.T. Smith, Electrophoresis 21 (2000)

2051.[67] L. Mateus, S. Cherkaoui, P. Christen, J.L. Veuthey, J. Chromatogr.

A 868 (2000) 285.[ 002)

[[ a, J.

[ 02)

[ 003)

[ oz,03)

[ . 71

[ ng,

[ .X.

[ .P.

[ un.

[ l. 41

[ 002)

[ nti,

[ d, J.

[[ men-

SA,

27] V. Camel, Trends Anal. Chem. 19 (2000) 229.28] X. Pan, G. Niu, H. Liu, J. Chromatogr. A 922 (2001) 371.29] X.J. Pan, G.G. Niu, H.Z. Liu, Biochem. Eng. J. 12 (2002) 71.30] M.D. Luque de Castro, M.M. Jimmenez-Carmona, Trends A

Chem. 19 (2000) 223.31] Q. Liu, C.M. Wai, Talanta 53 (2001) 771.32] B. Benthin, H. Danz, M. Hamburger, J. Chromatogr. A 937 (19

211.33] J.D. Denery, K. Dragull, C.S. Tang, Q.X. Li, Anal. Chem. Acta

(2004) 175.34] E.S. Ong, S.O. Woo, Y.L. Yong, J. Chromatogr. A 904 (2000)35] E.S. Ong, S.N. binte Apandi, Electrophoresis 22 (2001) 2723.36] H.K. Lee, H.L. Koh, E.S. Ong, S.O. Woo, J. Sep. Sci. 25 (20

160.37] E.S. Ong, J. Sep. Sci. 25 (2002) 825.38] E.S. Ong, S.M. Len, Anal. Chim. Acta 482 (2003) 81.39] A. Brachet, S. Rudaz, L. Mateus, P. Christen, J.L. Veuthey, J.

Sci 24 (2001) 865.40] B. Kaufmann, P. Christen, J.L. Veuthey, Chromatographia 54 (2

394.41] V. Fernandez-Perez, M.M. Jimenez-Carmona, M.D. Luque de C

Analyst 125 (2000) 481.42] M.M. Jimenez-Carmona, J.L. Ubera, M.D. Luque de Castro, J. C

matogr. A 855 (1999) 625.43] A. Basile, M.M. Jimenez-Carmona, A.A. Clifford, J. Agric. Fo

Chem. 46 (1998) 5205.44] J. Suomi, H. Siren, K. Hartonen, M.L. Riekkola, J. Chromatog

868 (2000) 73.45] E.S. Ong, S.M. Len, J. Sep. Sci. 26 (2003) 1533.46] E.S. Ong, S.M. Len, J. Chromatogr. Sci. 42 (2004) 211.47] Q. Fang, H.W. Teung, H.W. Leung, C.W. Huie, J. Chromatog

904 (2000) 47.48] M.P.K. Choi, K.K.C. Chan, H.W. Leung, C.W. Huie, J. Chromat

A 983 (2003) 153.

68] Z. Cai, F.S.C. Lee, X.R. Wang, W.J. Yu, J. Mass Spectrom. 37 (21013.

69] J. Sherma, J. AOAC Int. 86 (2003) 873.70] J.P. Salminen, V. Ossipov, J. Loponen, E. Haukioja, K. Pihlaj

Chromatogr. A 864 (1999) 283.71] F. Lepine, S. Milot, L. Zamir, R. Morel, J. Mass Spectrom. 37 (20

216.72] A. Tolonen, A. Hohtola, J. Jalonen, J. Mass Spectrom. 38 (2

845.73] B. Zanolari, J.L. Wofender, D. Guilet, A. Marston, E.F. Queir

M.Q. Paulo, K. Hostettmann, J. Chromatogr. A 1020 (2075.

74] X. Wang, T. Sakuma, E. Asafu-Adiaye, G.K. Shiu, Anal. Chem(1999) 1579.

75] W. Li, C. Gu, H. Zhang, D.V.C. Awang, J.F. Fitzloff, H.H.S. FoR.B. Breemen, Anal. Chem. 72 (2000) 517.

76] T.W.D. Chan, P.P.H. But, S.W. Cheng, I.M.Y. Kwok, F.W. Lau, HXu, Anal. Chem. 72 (2000) 1281.

77] F.L. Liu, C.Y.W. Ang, T.M. Heinze, J.D. Rankin, R.D. Beger, JFreeman, J.O. Lay Jr., J. Chromatogr. A 888 (2000) 85.

78] G.C. Kite, M.A. Yule, C. Leon, M.S.J. Simmonds, Rapid CommMass Spectrom. 16 (2002) 585.

79] J.R. Ioset, G.E. Raoelison, K. Hostettmann, Food Chem. Toxico(2003) 29.

80] J.R. Ioset, G.E. Raoelison, K. Hostettmann, Planta Medica 68 (2856.

81] N. Mulinacci, P. Vignolini, G. la Marca, G. Pieraccini, M. InnoceF.F. Vincieri, Chromatographia 57 (2003) 581.

82] D. Sullivan, J. Wehrmann, J. Schmitz, R. Crowley, J. EberharAOAC Int. 86 (2003) 471.

83] D.N. Heller, Anal. Chem. 72 (2000) 2711.84] R.B. Cole, Electrospray Ionization Mass Spectrometry, Funda

tals, Instrumentation and Applications, Wiley-Interscience, NY, U1997.