119

Evaluation of Artemisia annua L. clean-up methods for artemisinin quantification by

HPLC
Celeghini, R. M. S.; Silva, A. P.; Sousa, I. M. O.; Foglio, M. A.
Centro de Pesquisas Qumicas, Biolgicas e Agrcolas da Universidade Estadual de Campinas (CPQBA/UNICAMP),
Campinas, SP, Brazil. E-mail: foglioma@cpqba.unicamp.br

ABSTRACT: Evaluation of Artemisia annua L. clean-up methods for artemisinin quantification by HPLC.
Artemisinin, an endoperoxide sesquiterpene lactone isolated from Artemisia annua L. (Asteraceae), has received
considerable attention due to its antimalarial activity. The limiting step in sample preparation is result of large
amounts of chlorophyll encountered in the crude extract. In this work we describe several sample preparation
methods using solid-phase extraction (SPE) and compare them with conventional liquid-liquid extraction method.
The aim of this work was to define the optimum conditions for a clean-up procedure of plant extract. Silica, and
Florisil cartridges were evaluated as solid supports for SPE and treatment with lead acetate for liquid-liquid
extraction. The data presented herein suggest that the use of Florisil in analysis clean-up procedures is a low
cost, rapid and efficient method with high yield recoveries.
Key words: Artemisia annua, artemisinin, clean up, High Performance Liquid Chromatography

INTRODUCTION
Artemisinin (1), an endoperoxide
sesquiterpene lactone isolated from Artemisia annua
L. (Asteraceae), has received considerable attention
due to its antimalarial activity. Sesquiterpene lactones
are common in most Asteraceae tribes, with more
than 4000 known structures. Several methods have
been reported for the measurement of artemisinin and
its main derivatives in plant material and biological
fluids (Edwards, 1994). However, most of them are
either not sufficiently sensitive and do not offer reliable
results, or are difficult to apply in routine analyses
(Christen & Veuthey, 2001). Therefore, new methods
for the determination of these compounds, such as
supercritical fluid extraction and chromatography,
pressurized solvent extraction, extraction, highperformance liquid chromatography coupled to mass
spectrometry scattering detection (Wang et. al., 2005).
A diversity of bioactivities has been reported
for this class of compounds, such as antiinflammatory, antitumoral, antiulcerogenic, cytotoxic,
diuretic and cardiotonic, among others. Previously
we demonstrated that the resulting enriched
sesquiterpene fraction from Artemisia annua L. crude
ethanolic extract inhibited the ulcerative lesion index
in all experimental models tested, in rats. The results
mentioned therein suggested that the antiulcerogenic
properties were afforded by cytoprotective
mechanisms as result of active principles that increase
the gastric mucous prostaglandin level (Foglio et. al,
2002).
A variety of new analytical methods for the
Recebido para publicao em agosto/2004
Aceito para publicao em julho/2006

determination of artemisinin and derivatives have

appeared during the last decade. New extraction
processes use mild operating conditions to avoid the
degradation of the analytes. They allow obtaining clean
extracts from complex matrices. In this context, SFE,
PSE and MAE are interesting alternatives to classical
solid-liquid extraction under reflux. Several
methodologies have been applied to the analysis of
artemisinin and its derivatives. Techniques such as
HPLC-ELSD, HPLC-EC, HPLC-MS, GC-MS, CE-UV
or SFC-ELSD have been developed with the goal of
analyzing metabolites in a faster, better, cheaper and
more efficient manner (Hussen et. al. 2006). Many
important herbal drugs (e.g. ginkgo, black cohosh,
soy) contain bioactive compounds possessing no
chromophore, thus detection techniques other UV are
needed for their identification and standardization.
Evaporative light scattering is one of the possible
alternatives, but despite the fact that this technique
has been developed more than 30 years ago, it is still
considered to be an exotic detection alternative, with
few applications at hand (Ganzera&Stupper 2005). The
refractive index detector is one of the least sensitive
LC detectors. It is very sensitive to changes in ambient
temperature, pressure changes, flow-rate changes and
can not be used for gradient elution. Despite these
many disadvantages, this detector is extremely useful
for detecting those compounds that are nonionic, do
not adsorb in the UV, and do not fluoresce, as
artemisinin.
Analysis of artemisinin [1] is difficult because
the compound is unstable and other compounds in
the crude plant extracts interfere in its detection. Due
to the lack of chromophores, artemisinin is suitable
for HPLC with UV detection only after derivatization.

Rev. Bras. Pl. Med., Botucatu, v.8, n.esp., p.119-122, 2006.

120

This research is focused on validation of clean up

methods for analytical analysis (HPLC-IR) the
antimalarial, artemisinin, identified in Artemisia annua
L. for the purpose of quality control.

(3 x 5 mL), extracted with chloroform (3 x 10 mL),

dried over MgSO4, filtered and concentrated almost
to dryness under vacuum and diluted to a final volume
(10 mL) for analysis by HPLC-IR.
Slica Clean Up Procedure

[1] Artemisinin

The limiting step in sample preparation is

result of large amounts of chlorophyll encountered in
the crude extract. In this work we describe several
sample preparation methods using solid-phase
extraction (SPE) and compare them with conventional
liquid-liquid extraction method. The aim of this work
was to define the optimum conditions for a clean-up
procedure of plant extract. Silica, and Florisil
cartridges were evaluated as solid supports for SPE
and treatment with lead acetate for liquid-liquid
extraction.
MATERIAL AND METHOD
Plant material and fractionation
Artemisia annua L. leaves (hybrid CPQBA 2/
39 x PL5) were collected from the experimental field
of CPQBA/UNICAMP. Voucher specimen is deposited
at CPQBA/UNICAMP under registration number 229.
This material was allowed to dry under air circulation
(40C) and grinded for use. The plant material was
extracted using Ultra Turrax mixture during 2 min ,
6000 rpm at room temperature (25oC) with three
portions of 5 mL methanol for 2 min . The samples
were filtered on Buchner funnels fitted with glass frit
(porosity 4 mm), concentrated under reduced pressure
providing the crude extract (yield 20%). The crude
extracts were submitted to clean-up procedures with
Florisol as described by Manirakiza et al. (2000),
with lead acetate as described by Lonergan et al.
(1992) and with silicagel.
Clean-up procedure with 10% lead acetate
solution (Lonergan et al., 1992)
A 10% lead acetate solution (2 mL) was
added to the crude methanolic extract (1g) dissolved
in ethanol (0.5 ml). This mixture stood at room
temperature for one hour and filtered. The filtrate was
centrifugated at 10000 rpm, washed with H2O: CH3CN

Slica (2g) were added to a 6 mL cartridge

used with vacuum Manifold. The cartridge was
conditioned with hexane (6 mL) and not allowed to
dry. Small air chambers were removed from the
adsorbent by gentle tapping and the concentrated
extract (1 mL) was then applied to the column. The
analytes were eluted with n-hexane-ethyl acetate (4:1,
4 mL), filtered and concentrated almost to dryness
under vacuum and diluted to a final volume (10 mL)
for analysis by HPLC-IR.
Florisil Clean Up Procedure (Manirakiza et al.
2000)
Florisil (500 mg,170 m, 80A) cartridge (6
mL volume) used in conjunction with vacuum Manifold.
The cartridge was conditioned with n-hexane (6 mL)
and not allowed to dry. Small air chambers were
removed from the adsorbent by gentle tapping and
the concentrated extract (1 mL) was then applied to
the column. The analytes were eluted with
dichloromethane (4 mL), filtered and concentrated
almost to dryness under vacuum and diluted to a final
volume (10 mL) for analysis by HPLC-IR.
Chromatographic methodologies
A modular Waters system comprised of a
Waters 515 pump, a column oven, a Waters 2414
refraction index detector and an Empower- Waters
workstation was utilized. A Phenomenex LC-CN
column (25mm x 4mm x 5m) was employed, at 35C.
Separations were made in the isocratic mode, using
methanol: water (60:40v/v) mL/min at a flow rate of 1
mL/min with 20L injection volume ; detector and
column temperature were 35 C.
Quantitative analysis
Determination of the content of the artemisinin
in plant material was performed by external standard
method. Stock solutions of the standards of 550, 750,
1000, 1250 and 1500 g/mL were used. Triplicate
determination were carried out.
Recovery
Each sample was spiked with artemisinin
standard in three different concentrations levels and
subject to extraction and clean up procedure

Rev. Bras. Pl. Med., Botucatu, v.8, n.esp., p.119-122, 2006.

121
FIGURE 2. Standards
mixture HPLC
separation using a
refractive index detector
with Zorbax SB- CN
column (150mm
x4.6mm x 5m), eluted
with methanol: water
(60:40v/v), flow-rate 1.0
mL/min, sample
injection 20L and
detector and column
temperature were 35
C.

according to method under test. The percent

recovery of added artemisinin was calculated
comparing peak areas of the resulting solutions
with reference standard artemisinin solutions at the
same concentration.
RESULT AND DISCUSSION
The extraction of natural products is essential
not only as an evaluation tool for raw materials, but
also for the quality control of products. In fact, whatever
the analytical method used, an extraction procedure
of the plant material is required. Artemisinin is stable
in neutral solvents heated up to 150C. However,
traditional methods of extraction can be both time
consuming and labour intensive, creating delays in
the flow of information from the analysis laboratory
to the field or product line. This is why in a plant
development project it is important to have simple,
rapid and specific extraction and analytical
procedures, which allow the quantitative
determination of the analytes and possibly of their
precursors (Christen & Veuthey 2001).
Preliminary studies were carried out in

order to develop a suitable clean-up method for the

purification of crude methanolic Artemisia annua
extracts obtained using Ultra Turrax mixture. For
this purpose, sorbents such as Florisil, silica gel
and lead acetate were studied.
The crude extract was filtered on a column
containing silica or Florisil that was preconditioned
with appropriate solvents. The silicagel clean up
method was discontinued as result of the large
amounts of solvent required for separations of
artemisinin and chlorophyll with low efficiency.
Therefore this procedure was not considered
suitable for sample procedures prior to analysis.
Table 1 show that yields recoveries with
standard Florisil and clean up with lead acetate.
The low recoveries obtained with lead acetate
clean up method are a result of low artemisinin
yield recovery efficiency. The recovery of both
methods was expressed as the percentagem
recovery of standard artemisinin added to
sample. The results obtained with Florisil
indicated good accuracy and, consequentially,
an agreement between the theorical value and
the real value of concentration.

TABLE 1. Clean-up methods results of percent recovery of artemisinin reference standard added to Artemisia annua L.

Rev. Bras. Pl. Med., Botucatu, v.8, n.esp., p.119-122, 2006.

122

With Florisil more polar interfering materials

were retained, larger amounts of impurities were
trapped at the top of the cartridge avoiding pigment
contamination compared to silica clean up procedure,
that retain less polar material showing less efficiency.

Since recoveries achieved with Florisil were higher

than those obtained with lead acetate, and the volume
of solvents required for the elution of the analytes were
low, fully activated Florisil was selected as the cleanup sorbent for all subsequent studies (Figure 1).

FIGURE 1. Chromatogram of A. annua crude extract after clean-up procedure with Florisil cartridge obtained by High
Pressure Liquid Chromatography (HPLC) separation using a refractive index detector, A Phenomenex LC-CN column
(25mm x 4mm x 5m) eluted with methanol: water (60:40v/v), flow-rate 1.0 mL/min, sample injection 20L and detector
and column temperature maintined at 35 C.

CONCLUSION
The data presented herein suggest hat the
use of Florisil in analysis clean-up procedures is a
low cost, rapid and efficient method with high yield
recoveries for large number of sample analysis.
The results obtained were highly reproducible
and in good agreement with the certified values
demonstrating the suitability of the developed
analytical method for the extraction and clean-up of
Artemisia annua L.

ACKNOWLEDGMENT: FAPESP for research grant

REFERENCE
CHRISTEN, P.; VEUTHEY, J.L.. New Trends in Extraction,
Identification and Quantification of Artemisinin and
its Derivatives. Current Medicinal Chemistry. v.8,
n.15, p. 1827-1839,.2001.
EDWARDS, G. Measurement of Artemisinin and its
Derivates in Biological- Fluids. The Royal Society of
Tropical Medicine and Hygiene, v. 88, supl.1, p. 3739,. 1994.
FOGLIO, M.A. et al. Antiulcerogenic activity of some

sesquiterpene lactones isolated from Artemisia

annua L. Planta Medica, v.68,p.515-518, 2002.
GANZERA,M. et al Simultaneous determination of
Ephedra sinica and Citrus aurantium var. amara
alkaloids by ion-pair chromatography. Talanta, v.66,
n.4: p.889-894 ,2005.
HOSTETTMANN, K. et al.. Princpios ativos de plantas
superiores. So Carlos: Editora Edufscar, 2003. p
59-100.
HUSSEN, A. et al. Development of a pressurized liquid
extraction and clean-up procedure for the
determination of -endosulfan, -endosulfan and
endosulfan sulfate in aged contaminated Ethiopian
soils. Journal of Chromatography A, v.27, p. 202210, 2006.
LONERGAN, G. et al. Isolation NMR studies
andbiological activities of onopordopicrin from
Centaurea sochifolia. Journal of Natural Products.
v.55, p 225-228, 1992.
MANIRAKIZA, P. et al. Single step clean-up and GC-MS
quantification of organochlorine pesticide residues
in spice powder. Chromatographia, v. 52, n 11/12, p
787-790, 2000.
WANG,M.F. et al. Analysis of artemisinin in Artemisia
annua L. by LC-MS with selected ion monitoring.
Journal of Agricultural and Food Chemistry. v.53,
n.18, p. 7010-7013. 2005.

Rev. Bras. Pl. Med., Botucatu, v.8, n.esp., p.119-122, 2006.