Vous êtes sur la page 1sur 8

J.

of Supercritical Fluids 57 (2011) 129136

Contents lists available at ScienceDirect

The Journal of Supercritical Fluids


journal homepage: www.elsevier.com/locate/supflu

Supercritical CO2 extraction of Helichrysum italicum: Inuence of CO2 density and


moisture content of plant material
Jasna Ivanovic a , Mihailo Ristic b , Dejan Skala a,
a
b

University of Belgrade, Faculty of Technology and Metallurgy, Karnegijeva 4, 11120 Belgrade, Serbia
Institute for Medical Plant Research Dr Josif Pancic, Tadeusa Koscuska 1, 11000 Belgrade, Serbia

a r t i c l e

i n f o

Article history:
Received 21 December 2010
Received in revised form 13 February 2011
Accepted 14 February 2011
Keywords:
Helichrysum italicum
Supercritical extraction
Kinetics
Mathematical modelling
Process optimization
Moisture content inuence

a b s t r a c t
Kinetics and selectivity of supercritical carbon dioxide (SC CO2 ) extraction of Helichrysum italicum owers
were analyzed at pressures in the range of 1020 MPa and temperatures of 40 C and 60 C (density of SC
CO2 from 290 to 841 kg/m3 ) and also at 10 MPa and 40 C using owers with different moisture contents
(10.5% and 28.4%). Increased moisture content of H. italicum owers resulted in enchased solubility of
solute enabling decrease of SC CO2 consumption necessary for achieving desired extraction yield. The
most abundant compounds in the supercritical extracts are sesquiterpenes and waxes while monoterpenes and sesquiterpenes are the main constituents of essential oil obtained by hydrodistillation. The
optimal set of working parameters with respect to extraction yield, SC CO2 consumption and chemical
composition of extract were dened related to moisture content of raw material and SC CO2 density.
2011 Elsevier B.V. All rights reserved.

1. Introduction
The genus Helichrysum belonging to Asteraceae family include
approximately 600 species widespread all over the world. It is represented in the Mediterranean area by nearly 25 native species [1].
Helichrysum italicum (commonly known as curry plant, Italian everlasting or immortelle) is widespread along the coast and on the
islands of the Adriatic Sea. It is a dwarf aromatic shrub with yellow
owers, up to 5070 cm high, growing on dry cliffs and sandy soil
[2]. Metabolites isolated from H. italicum and especially its volatile
oil have been found to display many biological properties, such as
the anti-inammatory [3,4], antiallergic, antibacterial, antifungal
[510], antioxidant [3,1113] and antiviral [4,14].
The essential oil of Helichrysum species for commercial applications is traditionally obtained by steam and hydrodistillation
[1,2,6,10,1520]. Since high temperature during distillation process can induce thermal degradation, hydrolysis and solubilization
of some compounds some authors suggest low temperature processing techniques for isolation of bioactive principles from H.
italicum such as extraction with water [2123], organic solvents
[3,79,11,13,14,2124] and supercritical carbon dioxide (SC CO2 )
[12,17,18]. Process of extraction with SC CO2 is optimally performed at mild pressure (910 MPa) and temperature (4050 C)
in order to isolate essential oil rich extracts [17,18,25]. In the

Corresponding author. Tel.: +381 113370473; fax: +381 113370473.


E-mail address: skala@tmf.bg.ac.rs (D. Skala).
0896-8446/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.supu.2011.02.013

available literature there is limited information about SFE of


Helichrysum species [12,17,18]. Marongiu et al. [18] isolated H.
italicum extract by SC CO2 extraction (9 MPa and 50 C for 2 h; density of CO2 is 285 kg/m3 ) with a fractional separation technique
and found that extract contained mainly neryl acetate (11.2%),
-curcumene (8.1%), nerol (6.1%), ar-tumerone (5.4%), cis-dihydrooccidentalol (4.4%), neryl propionate (4.2%), ar-curcumene (3.8%)
and (E)-tagetone (3.0%). Oil obtained by fractional SFE had similar
chemical composition as oil obtained by hydrodistillation but they
differed quantitatively. Poli et al. [12] isolated antioxidant extracts
from H. italicum with SC CO2 at higher pressure (26 MPa; 50 C;
843 kg/m3 ) during 5 h of extraction.
Considering previous reports on SFE of Helichrysum species
[12,17,18], the main goal of the present study was to investigate
inuence of wide range of SC CO2 densities (290841 kg/m3 ) on the
kinetics of supercritical extraction of H. italicum and to determine
the chemical composition of obtained extracts.
Besides tuning pressure and temperature, process of supercritical uid extraction (SFE) is commonly optimized by the addition of
small amount of an entrainer or co-solvent (polar compounds, usually low molecular weight alcohols such as ethanol and methanol)
to improve both selectivity and solubility of SC CO2 [25].
Up to now a number of research groups reported that extraction
yield and SC CO2 selectivity to certain compounds can be differently
affected by moisture level in plant material SFE [2633] which was
explained by chemical composition and/or structures of the herbaceous matrices used for given studies. Inuence of moisture on
oil mass transfer and oil solubility is reported to be negligible if

130

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

moisture is in the range between 3% and 12% [27] (or for some
plant material up to 18% [26]). Normal moisture content of tobacco
had to be increased by up to 25% to affect nicotine extraction, presoaking of coffee beans aids the removal of caffeine and also to
increase the extraction of vanilla from vanilla beans [32]. Varying
bed moisture levels can differently affect mass transfer of solute in
the solid phase [33]. In the case of Origanum vulgare L. discontinuous addition of water led to increase in the extraction degree of
essential oils and reduction in the yield of the co-extracted waxy
material [33]. They observed that extraction was enhanced due to
changes in the transport properties of the solvent and a decrease
in intraparticle resistance of the solute inside the particles. Five
times higher predicted effective diffusivity for less brous fresh
ginger from the water-rich new season compared to the value for
the freeze-dried ginger samples previously reported was explained
by the structure of the matrix [32]. Increased moisture level in
oilseeds and paprika samples had negative effect on extraction efciency and time of extraction necessary for obtaining desired yield
of extract [26,28], or increased the loss of volatile compounds [29].
Yoda et al. [31] reported different inuence 510% moisture in the
Stevia rebaudiana on extraction yield at supercritical and subcritical
conditions.
Up to now there are no data about inuence of the moisture content in the plant material on kinetics of SFE of H. italicum and on
chemical composition of the supercritical extracts. Therefore, the
goal of this study was also to analyze the inuence of moisture content of H. italicum owers on the kinetics of SFE and, to analyze the
chemical composition of the obtained extracts isolated at 10 MPa
and 40 C (CO2 = 630 kg/m3 ).
In this paper, Sovovs mass transfer model [28,3437] was
used for simulation of experimentally obtained extraction curves
in order to analyze the inuence of the SC CO2 density and moisture
content in the plant material on the solubility of extracts and mass
transfer parameters.

2. Materials and methods


2.1. Plant material
Dried owers of H. italicum used in this study were collected
during September and October in 2008 (owering stage) from the
region Konavle (Croatia) and the voucher specimen (no. 16282)
was deposited at Herbarium of Institute of Botany and Botanical
Garden Jevremovac, Faculty of Biology, University of Belgrade (Belgrade, Serbia). Flowers were dried in the air protected against direct
sunlight. In order to obtain owers with increased moisture content, one part of owers was sprayed with distilled water and left
in the plastic bag overnight. The moisture contents of the air-dried
and wetted owers determined by Karl Fischer volumetric titration
were 10.5% and 28.4% (w/w), respectively. Flowers were milled in
a blender for 60 s and immediately subjected to supercritical CO2
extraction or hydrodistillation. The average particle size of milled
owers was 0.40 mm.
Commercial carbon dioxide (99% purity) supplied by Tehnogas (Messer-Tehnogas, Serbia) was used for SFE, and chloroform
(GC purity, SigmaAldrich, Germany) was used for dissolution of
supercritical extracts prior to GC/FID/MS analyses.

2.2. Hydrodistillation
The milled owers of H. italicum (50 g) were distilled using
1000 ml distilled water according to Clevenger standard method
(4 h) and volatile essential oil was collected. The obtained essential
oil was kept in a sealed vial at 4 C.

2.3. Supercritical extraction


Extractions from air-dried and wetted H. italicum owers were
carried out in the Autoclave Engineers Screening System previously
described in Ref. [38] at 40 C and 60 C and 10, 15 and 20 MPa (CO2
density in the range 290841 kg/m3 ). The mass of 30 g of the plant
material was used for the extraction while the ow rate of CO2
was 0.52 kg/h (SD = 0.011) in all performed runs. Extraction yields
were determined after total consumption of 850 g of CO2 which
correspond to 35.6 4.0 g of CO2 used per 1 g of H. italicum owers.
The extraction time was 1.5 0.13 h.
Fractional extraction with SC CO2 was also performed in order
to investigate dynamics of extraction of the compounds present
in H. italicum owers. A differential quantity of extract was collected in order to monitor the changes of the chemical composition
during 60 min of extraction with respect to CO2 consumption
(9.3131.65 kg CO2 /kgplant material ), CO2 density (290781 kg/m3 )
and moisture content (10.5% and 28.4%, w/w). All experiments were
carried out in duplicate or triplicate and the average experimental
yields were within the experimental error of about 25%.
2.4. Sovovs mass transfer model [35,36]
The semi-empirical mass transfer mathematical model proposed by Sovovs that had been previously described in detail
in the literature [35,36] was used to simulate SFE curves of H.
italicum owers. The experimentally obtained overall extraction
curves were presented as extraction yield (%, w/w) versus specic amount of consumed SC CO2 (g CO2 /gdry plant material ). Sovovs
model separates the extraction curve into three parts or periods:
the rst period represents the extraction of easily accessible solute
which is linear function of consumed supercritical solvent where
the slope of straight line dene the oil solubility; the rate of extraction decreases rapidly in the second period; and, the extraction
rate is very slow and almost constant in the last and third period of
extraction.
Mathematical expression of extraction process, i.e. the relationship between the yield of extract (y), and consumed solvent,
is based on two sets of parameters. The rst set of parameters
(co-called contextual parameters) includes extraction conditions and characteristics of plant material used for extraction or
is a function of them: pressure (p), temperature (t), CO2 density (CO2 ), viscosity (CO2 ) [39], CO2 ow rate (qCO2 ), supercial
velocity (u), mass of non-extractable material (extract-free) (mSI ),
initial mass of extractable material, relative to mass of nonextractable material (xo ), density of plant material (s = ms /Vs ),
bed porosity (), interfacial area (a0 ) [36], binary diffusion coefcient (D12 ) [40], mass transfer coefcients in the lm (kf , Z = f(kf ))
[35,36,40]. According to Sovovs model complex mixture of H.
italicum supercritical extracts was represented by one so called
pseudo component (-selinene) for the calculation of binary diffusion coefcient and mathematical modelling. For prediction of
extraction curve one need additional number of so called model
parameters which are, in principle, functionally related to the
contextual parameters. Model parameters are proposed and
estimated to t the experimental data with the mathematical
model. Such parameters are solubility of solute (yr ), coefcients
of mass transfer in solid (plant) matrix (ks , ks,eff [41], W = f(ks ))
and xk initial mass of extractable material in intact cells (glandular trichomes [15]), relative to mass of non-extractable material
[35,36]. Recent microscope studies of ower heads and leaves of
H. italicum ssp. microphyllum indicated that glandular trichomes
secreting essential oils are of one single morphological type
[1,15].
The results of mathematical simulation are shown as a percentage of average absolute relative deviation (AARD), which is dened

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

as deviation of the yield calculated by model and those experimentally determined [34].

131

Table 1
Chemical composition of the essential oil of H. italicum owers.
Constituents

RIa

RT/MSb

932
1020
1020
1024
1095
1135
1165
1174
1186
1227
1359

7.3
8.53
10.5
10.6
13.03
13.9
15.27
15.68
16.17
16.97
21.56

1373
1374
1405
1410
1411
1417
1481
1479
1489
1498
1500
1505
1513
1522
1582
1600
1644
1649

21
21.83
23.22
23.43
23.56
23.69
24.92
25.78
25.85
25.86
26.21
26.44
26.61
26.87
28.21
28.91
30.24
30.5

1100

13.17
15.41

2056

40.72

2.5. Analytical procedures


2.5.1. Gas chromatography (GC/FID)
Gas chromatography analysis of the extracts was carried out
on a HP-5890 Series II GC apparatus [Hewlett-Packard, Waldbronn
(Germany)], equipped with splitsplitless injector and automatic
liquid sampler, attached to HP-5 column (25 m 0.32 mm, 0.52 m
lm thickness) and tted to ame ionization detector (FID). Carrier
gas ow rate (H2 ) was 1 ml/min, split ratio 1:30, injector temperature was 250 C, detector temperature 300 C, while column
temperature was linearly programmed from 40 to 260 C (at rate of
4 C/min), and then kept isothermally at 260 C for 10 min. Solutions
of samples in chloroform were consecutively injected in amount
of 1 l. Area percent reports, obtained as result of standard processing of chromatograms, were used as base for the quantication
analysis.
2.5.2. Gas chromatography/mass spectrometry (GC/MS)
The same analytical conditions as those mentioned for GC/FID
were employed for GC/MS analysis, along with column HP-5MS
(30 m 0.25 mm, 0.25 m lm thickness), using HP G 1800C Series
II GCD system [Hewlett-Packard, Palo Alto, CA (USA)]. Helium was
used as carrier gas. Transfer line was heated at 260 C. Mass spectra were acquired in EI mode (70 eV); in m/z range 40450. The
amount of 0.2 l of sample solution in chloroform was injected. The
components of the oil were identied by comparison of their mass
spectra to those from Wiley 275 and NIST/NBS libraries, using different search engines. The experimental values for retention indices
were determined by the use of calibrated Automated Mass Spectral
Deconvolution and Identication System Software (AMDIS ver. 2.1)
[42], compared to those from available literature (Adams [43]) and
used as additional tool to approve MS ndings.

Monoterpenes
-Pinene
p-Cymene
Limonene
1,8-Cineole
Linalool
trans-Pinocarveol
Endoborneol
Terpinene-4-ol
-Terpineol
Nerol
Neryl acetate
Sesquiterpenes
-Ylangene
-Copaene
Italicene
-Cedrene
cis--Bergamotene
trans--Caryophyllene
-Curcumene
ar-Curcumene
-Selinene
-Selinene
Murolene
-Bisabolene
-Cadinene
-Cadinene
Caryophyllene oxide
Viridiorol
-Murolool
-Eudesmol
Hydrocarbon aldehydes, ketones and esters
Nonyl aldehyde
2-Methyl, 2-butenoic acid, isopentyl ester
Diterpenes
Manool
ID (%)
Number of constituents
a

3. Results and discussion

3.1. The yield and chemical composition of the essential oil

First of all, the content and chemical prole of H. italicum


essential oil has been investigated. The yield of the essential oil
isolated from H. italicum owers by hydrodistillation in the Clevenger apparatus after 4 h was 0.23%, w/w (on dry weight basis).
The H. italicum essential oil comprised mostly monoterpene and
sesquiterpene compounds (72%) (Table 1). The most abundant
compounds were -curcumene (12.4%), -selinene (9.94%), trans-caryophyllene (6.91%), -selinene (5.91%), italicene (4.62%) and
ar-curcumene (4.01%). According to previously published data
the most abundant compounds in the essential oil of H. italicum
owers originated from Corsica, Croatia and Tuscan islands were
neryl acetate, nerol, ar-curcumene, -curcumene and trans-caryophyllene [1,6,15,16,18,19] which differ in quantity due to
origin and period of collection of plant material.

EOc
% (w/w)d
10.38
2.8
0.28
0.64
0.22
0.42
0.49
0.3
0.43
0.49
0.21
3.1
61.3
0.52
3.61
4.62
0.96
2.19
6.91
12.4
4.01
9.94
5.91
1.15
1.21
1.36
0.85
0.63
3.01
0.74
1.07
1.03
0.35
0.29
4.22
4.22
76.8
46

Kovats index.
Retention time (min) of corresponding constituent obtained by GC/MS.
Essential oil.
Percentage >0.2% (w/w).

the same plot the extraction curves simulated by Sovovs model


are also presented.
As expected, pressure increase generally resulted in a higher
yield of extract as well as by faster rate of extraction due to
positive effect of increased density on solvent power of the SC

3.2. Kinetic study and mathematical modelling of supercritical


extraction
In order to investigate inuence of different SC CO2 densities on
the kinetics of H. italicum supercritical extractions were performed
at pressures of 10, 15 and 20 MPa and temperatures of 40 C and
60 C, which dene SC CO2 density from 290 to 841 kg/m3 . Inuence of different humidities of H. italicum ower heads (10.5% and
28.4%) on the kinetic of SFE at 10 MPa and 40 C was also studied.
Extraction yields (on dry weight basis) were plotted against SC CO2
consumption (per 1 g of dry plant material) and shown in Fig. 1. At

Fig. 1. Extraction yield as a function of the specic amount of SC CO2 for SFE at
different p and t (IVI, Table 2); solid line presents the yield calculated according to
Sovovs model.

132

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

CO2 [25]. As might be seen from Fig. 1 extractions realized at


40 C (15 MPa and 20 MPa) and 60 C (20 MPa) had very similar
extraction rate and yield due to relatively small differences of SC
CO2 density (725840 kg/m3 ). However, change of extraction rate
and total yield for experiments performed at 10 MPa/60 C and at
10 MPa/40 C is more pronounced due to larger decrease of SC CO2
density with temperature increase at same pressure (from 630 to
290 kg/m3 ). Finally, according to the results of this experiments
it seems that combination of pressure and temperature which is
dened by the value of SC CO2 density around 700 kg/m3 can be
dened as optimal for achieving a high rate of extraction and about
4% (w/w) yield of H. italicum extract.
Experimental results indicated that higher moisture content in
the plant material resulted in the increase of the extraction rate
at the beginning of extraction as well by larger yield of extraction
(Fig. 1). The moisture content in the H. italicum owers was about
three times larger (28.4%) compared to the measured moisture of
dry owers (10.4%). Evidently, humidication of H. italicum owers
by water presoaking led to increase of extraction rate for 40% and
consequently to 25% less SC CO2 consumption for achieving same
extraction yield at same pressure and temperature (2.45%, w/w
at 10 MPa; 40 C; CO2 = 630 kg/m3 ). It was previously suggested
[26,32] that this relative increase of the extraction yield from plant
material with increased water content might be related to following
factors:
(a) the dissolution of water into the supercritical phase enhance
solubility of solute through an entrainer type mechanism,
and/or
(b) the water-swelling of plant tissue inuence the internal or solid
mass transfer resistance and movement of solute to the particle
surface.
In order to support observed effects of SC CO2 density and presence of the increased moisture content in the starting plant material
on the extraction kinetics, the results of mathematical simulation of
extraction process according to model proposed by Sovov [35,36]
were used:
(a) to analyze specic parameters which give information related
to phase equilibrium data (solubility, yr ) and the mass transfer
rates in CO2 and solid phases (parameters Z = f(kf ) and W = f(ks ),
respectively),
(b) to explain qualitative changes of extracts compositions and
their inuence on the kinetic data according to GC/FID and
GC/MS analyses.
Experimentally obtained yields of extraction versus SC CO2 consumption per 1 g of dried plant material and that calculated by
Sovovs are presented in Fig. 1. Sovovs model described experimental data with satisfactory high accuracy (AARD in the range of
1.35.6%). The contextual and model parameters are presented in
Table 2.
Using Sovovs model, the solubility of the extract in SC CO2 (yr )
at given conditions (p, t) was estimated as the slope of the rst
part of the extraction curve. Total yield was xed as the asymptotic
value for large extraction times (t ). The mass transfer coefcient of solute in the solid phase (ks ) and the mass of inaccessible
solute per mass of solute-free solid phase (xk ) were simultaneously
optimized in order to minimize deviation of calculated extraction
yields from experimentally obtained data. Parameters Z and W are
directly proportional to coefcients of the mass transfer of solute
in the lm (kf ) and solid phase (ks ), respectively.
As can be seen from the results presented in Table 2, increase
of SC CO2 density led to increase of diffusion rate in the CO2 lm
described by parameter Z (directly proportional to kf and ) due to

increase of SC CO2 density especially above 630 kg/m3 . On the other


hand, model parameter related to diffusion of solute in the solid
phase (ks , ks,eff , W) increases with SC CO2 density from 290 kg/m3
to 630 kg/m3 and above 630 kg/m3 it is decreased. The following
explanation could be given: a higher amount of solute in SC CO2
dened by increase of extracts solubility (yr ) at densities higher
than 630 kg/m3 also needs more time for diffusion of extract from
solid matrix to supercritical phase.
According to results presented in Table 2 increased moisture
content (for about three times compared to dried owers) in the
herbaceous matrix leads to ve times higher solubility (yr ) and
somewhat higher mass transfer resistance in the solid phase (W) in
the presence of higher amount of water in the plant material.
The last conclusion indicated that more pronounced increase of
the rate of H. italicum owers extraction within the rst period of
extraction (Fig. 1) is result of the inuence of solubility increase
rather than swelling effect of the plant matrix to the overall extraction rate in the second period of extraction. Waters enhanced
solubility in the SC CO2 under pressure [26] might cause enhancing the solubility of solute through an entrainer type mechanism
as previously suggested in literature [32]. Under the experimental conditions (10 MPa and 40 C), 0.039 g H2 O would be expected
to dissolve in SC CO2 (in the volume of extractor of 120 cm3 ) at
equilibrium [44]. This is 1.15% and 1.24% of water presented in
the owers containing 28.4% and 10.5%, respectively. On the other
hand, water content in the isolated extracts from both, dried and
water pre-soaked owers was constant 6.4 1.0% (with respect to
mass of extract). Water content in the exhausted material was 6.6%
for dried owers with initial moisture content of 10.5% and 20.4%
for the water pre-soaked owers with initial moisture content of
28.4%. According to these observations following conclusions have
been made:
(a) the increase of extraction yield for SFE from plant material with
increased water content is not result of water co-extraction
with solute;
(b) about two times higher extraction of water with SC CO2 from
material containing 28.4% of water is result of larger driving
force (a difference of initial moisture content of plant material
and moisture in the CO2 phase at equilibrium at same extraction
conditions);
(c) A higher moisture content of residue after SFE of plant material
(20.4%) containing three times higher initial amount of water
(28.4%) is result of limited rate of diffusion of water into SC CO2
phase and same extraction time used for all experiments (1.5 h).
Somewhat higher increase of the mass transfer in the solid plant
material with increased moisture content at same extraction conditions (Table 2) might be result of the existence of co-called water
bridging effect. Such an effect occurs in the presence of increased
moisture content (1020% and more) in the porous solids when
water exceeds the quantity necessary to ll the internal pore volume and starts to bridge between particles and increases path
length controlling the mass transfer process in the porous solid
matrix [45]. Further work might consider SEM analysis and/or optical detection of plant tissue of the dry and pre-soaked H. italicum
ower heads exposed to SC CO2 in a high-pressure view chamber
in order to verify this assumption.
3.3. Chemical composition of supercritical extracts
Increase of SC CO2 density resulted in both quantitative and
qualitative changes of composition of the extracts isolated from
H. italicum owers. As expected, the highest percentage of lighter
compounds (mono- and sesquiterpenes) has been found in the

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

133

Table 2
Contextual and model parameters for SFE from H. italicum owers.
Contextual parameters

Units

Temperature, t
Pressure, p
Density of CO2 , CO2
Viscosity of CO2 , CO2 105
Moisture content
Bed porosity,
Mass ow rate, qCO2 104
Supercial velocity, u 104
Specic interfacial area, a0 103
Total extraction yield, ytot a
Binary diffusion coefcient, D12 108
Mass transfer coefcient in the CO2 phase, kf 104
Sovovs model parameters
Solubility, yr 103
Mass transfer coefcient in the solid phase, ks 107
Mass of inaccessible solute per mass of solute-free solid phase, xk 102
Mass transfer coefcient in the solid phase, ks,eff 109 b
Z
W

AARD

Sample
I

a
b

C
MPa
kg/m3
Pa s
%, w/w

kg/s
m/s
m/m3
%, w/wa
m2 /s
m/s
kgoil /kgCO2
m/s

m/s

III

IV

VI

40
10
630
4.68
10.5
0.45
1.81
4.02
0.856
2.45
2.19
0.712

II
40
10
630
4.68
28.4
0.45
1.5
3.33
0.904
2.78
2.19
0.609

40
15
781
6.7
10.5
0.45
1.28
2.84
1.133
4.25
2.19
0.446

40
20
841
7.84
10.5
0.45
1.19
2.64
1.252
4.1
2.19
0.388

60
10
290
2.41
10.5
0.45
1.67
3.71
0.629
1.37
2.19
0.928

60
20
725
5.89
10.5
0.45
1.18
2.62
1.087
3.79
2.19
0.445

2
5
2.2
11.00
72.2
0.113

10
4
2.75
11.00
62.70
0.094

20
2.3
4.2
9.66
102.9
0.095

25
1.8
4
7.20
114.8
0.089

1.5
3.2
1.36
4.35
34.86
0.058

20
2.5
3.7
9.25
109.7
0.108

5.42

5.21

1.80

5.56

1.30

3.40

On dry weight basis.


Previously dened in Ref. [41] as ks,eff = ks xk .

extracts isolated by single stage extraction at lowest density of SC


CO2 (290 kg/m3 ). Set of following process parameters: 9 MPa and
50 C (CO2 = 285 kg/m3 ) has been previously identied as optimal
for isolation of oxygenated terpenes from H. italicum owers which
are found to give more relevant contribution to fragrance than nonoxygenated terpenes and other compounds [18]. This study also
show that increase of SC CO2 density favored isolation of waxes and
oxygenated hydrocarbons, whereby percentage of monoterpenes
and sesquiterpenes in extract was decreased (Table 3).
Fractional extraction with SC CO2 has been performed in the
density range of 290781 kg/m3 in order to determine dynamics of isolation of particular type of compounds during extraction.
Isolation of compounds which belong to specic families (M:
monoterpenes; S: sesquiterpenes; E: esters; W: waxes; O: others;
and, ni: not identied) is dened as percents of extracted compounds (grams per 100 g of dry plant material) per 1 kg of SC CO2
consumed (Fig. 2). Fractions of extract were taken within extraction
period necessary for achieving 70100% of total extraction yield at
some working conditions. As can be seen from Fig. 4, the increased

Fig. 2. Inuence of SC CO2 density on isolation of different types of compounds (M:


monoterpenes; S: sesquiterpenes; E: esters; W: waxes; O: other).

amount of all types of compounds per 1 kg consumed SC CO2 is


detected by increasing pressures of extraction due to decrease of
SC CO2 selectivity, especially in the case of SC CO2 densities above
630 kg/m.
The results of the investigation of the moisture content in the
starting plant material are presented in Figs. 3 and 4. The chemical
analysis of supercritical extracts isolated from owers with different humidities at 10 MPa and 40 C (CO2 = 630 kg/m3 ) showed
increase of percentage of lighter compounds (monoterpenes and
sesquiterpenes) in the extract and at the same time decrease of percentage of the waxy compounds isolated from wet owers (28.4%
of H2 O) compared to the chemical composition of extract isolated
from dried owers (10.5% of H2 O). This indicates that higher value
of moisture in the vegetable matrix modies solvent power of SC
CO2 and increases its selectivity to lighter compounds (more polar
compounds) (Fig. 3).
On the other side, quantitative analysis revealed that presence
of increased moisture in the vegetable matrix enables isolation
of higher amount of all specied compound families (Fig. 4). It
might be expected due to increase of solute solubility in SC CO2

Fig. 3. Inuence of moisture content of different classes of compounds in the


extracts (M: monoterpenes; S: sesquiterpenes; E: esters; W: waxes; O: other; ni:
non-identied).

134

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

Table 3
Chemical composition of the supercritical extracts of H. italicum owers.
Components

RIa

RT/MSb

II

III

VI

5.483
1.279
0.305
0.918
1.702
21.278
1.157
1.617
0.46
0.498
3.04
4.308
1.347
0.374
0.498
0.58
0.914
0.486
0.996
0.821

0.151
1.595
0.481
0.798
2.431
0.242
1.337
0.531
22.457
0.348
22.109
28.000
2.228
3.116
8.176
0.851
10.392
1.527

4.305
1.015
0.398
0.959
0.441
23.643
1.136
1.673
0.251
0.557
3.463
4.939
1.675
0.488
0.543
0.555
1.325
0.394
0.708
0.684
0.297
0.111
1.335
0.461
0.732
3.613
0.39
1.878
0.778
18.738
0.404
18.334
25.448
2.192
3.022
7.792
0.73
8.517
1.109

5.312
0.0762
2.495
1.044
0.573
26.422
0.827
1.333
0.397
0.488
3.287
4.846
1.437
0.578
0.622
0.62
0.733
0.639
0.851
1.058
0.628
0.367
2.191
0.664
0.912
1.15

21.397
0.179
21.217
27.357
1.789
2.826
8.048
0.767
10.459
1.679

0.565
0.193
20.668
0.681
19.986
23.069
1.821
3.794
7.718
0.617
7.371
0.638

5.317
2.56
0.135
0.94
0.151
21.929
1.23
1.586
0.291
0.576
2.984
4.558
1.511
0.373
0.49
0.561
1.365
0.401
0.912
0.668
0.315
0.707
0.634
0.416
0.739
1.431
0.629
0.349
0.15
20.393
0.312
20.081
27.039
2.034
2.807
7.435
1.086
9.972
1.677

76.921
59

75.746
65

76.677
59

76.622
59

76.108
58

%, w/w
Monoterpenes
-Pinene
Pugenol
Neryl acetate
Geranyl tiglate
Sesquiterpenes
-Copaene
Italicene
trans--Caryophyllene
9-Epi-(E)-caryophyllene
ar-curcumene
-Selinene
-Selinene
-Bisabolene
-Cadinene
-Cadinene
Caryophyllene oxide
Cedrol
Globulol
Rosifoliol
Muurola-4,10(14)-dien-1-beta-ol
-Eudesmol
Eudesm-7(11)-en-4-ol
Xanthorrhizol
(Z)-Lanceol
Hydrocarbon acids and esters
2.5-Dimethyl-2-hexanold
Methyl oleate
Methyl linoleate
Other
Dihidroangelicin(2 ,3 ,3 -trimethyl-2 ,3 -)c
Mw = 242274d
Waxes
Pentacosane
Heptacosane
Nonacosane
Triacontane
Untriacontane
Dotriacontane

932
1356
1359
1696

7.04
21.42
21.56
31.26

1374
1405
1417
1464
1479
1489
1498
1505
1513
1522
1582
1599
1590
1598
1630
1649
1652
1760
1767

21.83
22.7
23.019
24.94
25.21
25.28
25.54
25.98
26.11
26.4
28.11
28.56
28.75
28.85
29.43
30.03
30.16
33.67
33.9

2082
2095

17.37
41.13
41.27

1893
19312203

36.23

2500
2700
2900
3000
3100
3200

51.75
55.38
59.33
61.04
63.22
68.58

ID%
Number of constituents
a
b
c
d

3.478
0.479
0.33
0.897
0.914
20.557
1.034
1.361
0.219
0.456
3.027
4.218
1.45
0.369
0.464
0.502
0.811
0.4
0.927
0.633

0.206
1.135
0.446
0.696
1.163
0.519
0.162

Kovats index.
Retention time of corresponding constituent obtained by GC/MS.
Only the most abundant components from different classes of compounds are presented.
Tentative identication (MS).

Fig. 4. Inuence of moisture content on isolation of different types of compounds


(M: monoterpenes; S: sesquiterpenes; E: esters; W: waxes; O: other; ni: nonidentied).

swelling effect in presence of moisture is conrmed to be negligible.


Contents of particular group of compounds per 1 kg CO2 consumed which were obtained by GC/MS analysis of fractions
collected during fractionation were used to calculate extraction
selectivity of volatile compounds in SC CO2 (mono- and sesquiterpenes) as a function of solvent density and moisture of H. italicum
owers (Fig. 5). Selectivity of monoterpene and sesquiterpene
compounds, SM+S , was calculated taking into the account the
content of other characteristic groups of compounds which are
dened as: (a) esters; (b) waxes; (c) others; and (d) non-identied
compounds.
Due to adverse effect of SC CO2 density on its selectivity above
densities of 630 kg/m3 much larger amount of both high and low
molecular weight compounds is extracted at the initial period of
extraction (Fig. 5).
The presence of moisture modied solubility power increasing its selectivity to more polar compounds (sesquiterpenes) and
decreasing its afnity to non-polar waxy compounds. Due to fast
exhaustion of volatiles at the beginning of extraction process SM+S
decreases within the later stages of extraction (during exhaustion

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

135

4. Conclusion
Present study provides new data on supercritical extraction of
H. italicum owers in the range of SC CO2 densities from 290 to
0.840 kg/m3 . It was shown that increase of SC CO2 density resulted
in the faster extraction and higher extraction yields due to positive
effect of higher density on solubility of extract in the SC CO2 . On the
other hand, increasing SC CO2 density above 630 kg/m3 had adverse
effect on selectivity toward terpenes extraction.
Increased moisture content positively inuences solubility of
lighter and oxygenated compounds in SC CO2 enabling a higher rate
of extraction and thus, the reduction of SC CO2 consumption. For the
same time of extraction, the increased moisture of H. italicum owers changes selectivity of SC CO2 to aroma compounds particularly
to the higher content of sesquiterpenes in supercritical extract.
Results obtained using raw material with different moisture
contents and SC CO2 density was used for dening optimal working
parameters with respect to extraction yield, time of extraction, SC
CO2 consumption and chemical composition of extracts.
Acknowledgements
Authors thank to Dr Sandra Glisic for valuable assistance during
experimental work.
Financial support of this work by the Serbian Ministry of Science,
Project No. 142073, is gratefully acknowledged.
References

Fig. 5. Inuence of (a) SC CO2 density and (b) moisture content in the plant material
on selectivity of extraction to mono- and sesquiterpene compounds.

of plant material) due to isolation of higher amount of less polar


and higher molecular weight compounds primarily waxes (Fig. 5).
According to experimental results the maximal selectivity SM+S
but with relatively small yield of extract might be obviously
achieved within rst fraction at 10 MPa and 40 C or 60 C (90%
or 55%, respectively) using dried owers containing 10.5% H2 O.
A larger yield and good average selectivity SM+S of extract (about
40% as sum of fractions I, II and III,) was determined performing
extraction at 10 MPa and 40 C with wet owers containing 28.4%
H2 O.
Finally, taking into account results of chemical analysis and yield
of extract, the optimal working parameters in relation to SC CO2
consumption necessary for obtaining desired quality of extract (30%
of terpenes) are:

(a) 10 MPa and 40 C (CO2 = 630 kg/m3 );


(b) increased moisture of plant material compared to the dry
ower (this study indicated that moisture of 28.4% gave larger
yield of extract than moisture of 10.5% H2 O);
(c) extraction time of 40 min (consumption of 16.3 kg of SC CO2 per
kg of dry plant material) at 0.52 kg/h mass ow rate of SC CO2 ;
(d) extraction yield of 2.1%, w/w (wet H. italicum owers) which
corresponds to approximately 90% of total yield from dried
owers and larger consumption of SC CO2 .

[1] I. Morone-Fortunato, C. Montemurro, C. Ruta, R. Perrini, W. Sabetta, A. Blanco, E.


Lorusso, P. Avato, Essential oils, genetic relationships and in vitro establishment
of Helichrysum italicum (Roth) G. Don ssp. italicum from wild Mediterranean
germplasm, Industrial Crops and Products 32 (2010) 639649.
[2] J. Mastelic, O. Politeo, I. Jerkovic, Contribution to the analysis of the essential
oil of Helichrysum italicum (Roth) G. Don.-determination of ester bonded acids
and phenols, Molecules 13 (2008) 795803.

[3] A. Sala, M.C. Recio, G.R. Schinella, S. Mnez,


R.M. Giner, N.M. Cerd, J.L. Ros,
Assessment of the anti inammatory activity and free radical scavenger activity
of tiliroside, European J. Pharmacology 461 (2003) 5361.
[4] G. Appendino, M. Ottino, N. Marquez, F. Bianchi, A. Giana, M. Ballero, O. Sterner,
B.L. Fiebich, E. Munoz, Arzanol, an anti-inammatory and anti-HIV-1 phloroglucinol -pyrone from Helichrysum italicum ssp. microphyllum, J. Natural Products
70 (2007) 608612.
[5] V. Voinchet, A.M. Giraud-Robert, Utilisation de lhuile essentielle dhelichryse
italienne et de lhuile vgtale de rose musque aprs intervention de chirurgie
plastique rparatrice et esthtique, Phytothrapie 2 (2007) 6772.
[6] J. Mastelic, O. Politeo, N. Radosevic, Composition and antimicrobial activity
of Helichrysum italicum essential oil and its terpene and terpenoid fractions,
Chemistry of Natural Compounds 41 (2005) 3539.
[7] A. Nostro, G. Bisignano, M.A. Cannatelli, G. Crisa, M.P. Germano, V. Alonzo,
Effects of Helichrysum italicum extract on growth and enzymatic activity of
Staphylococcus aureus, International J. Antimicrobial Agents 17 (2001) 517520.
[8] A. Nostro, M.A. Cannatelli, G. Crisa, A.D. Musolino, F. Procopio, V. Alonzo,
Modication of hydrophobicity, in vitro adherence and cellular aggregation of
Streptococcus mutans by Helichrysum italicum extract, Letters in Applied Microbiology 38 (2004) 423427.
[9] A. Nostro, M.A. Cannatelli, A.D. Musolino, F. Procopio, V. Alonzo, Helichrysum
italicum extract interferes with the production of enterotoxins by Staphylococcus aureus, Letters in Applied Microbiology 35 (2002) 181184.
[10] A. Angioni, A. Barra, M. Arlorio, J.D. Coisson, M.T. Russo, F.M. Pirisi, M. Satta, P.
Cabras, Chemical composition, plant genetic differences, and antifungal activity
of the essential oil of Helichrysum italicum G. Don ssp. microphyllum (Willd)
Nym, J. Agricultural and Food Chemistry 51 (2003) 10301034.
[11] A. Rosa, M. Deiana, A. Atzeri, G. Corona, A. Incani, M.P. Melis, G. Appendino,
M. Assunta Dess, Evaluation of the antioxidant and cytotoxic activity of
arzanol, a prenylated -pyrone-phloroglucinol etherodimer from Helichrysum italicum subsp. Microphyllum, Chemico-Biological Interactions 165 (2007)
117126.
[12] F. Poli, M. Muzzoli, G. Sacchetti, G. Tassinato, R. Lazzarin, A. Bruni, Antioxidant
activity of supercritical CO2 extracts of Helichrysum italicum, Pharmaceutical
Biology 41 (2003) 379383.
[13] G.R. Schinella, H.A. Tournier, J.M. Prieto, P. Mordujovich de Buschiazzoa, J.L.
Ros, Antioxidant activity of anti-inammatory plant extracts, Life Sciences 70
(2002) 10231033.
[14] A. Nostro, M.A. Cannatelli, A. Marino, I. Picerno, F.C. Pizzimenti, M.E. Scoglio, P.
Spataro, Evaluation of antiherpesvirus-1 and genotoxic activities of Helichrysum italicum extract, New Microbiologica 26 (2003) 125128.

136

J. Ivanovic et al. / J. of Supercritical Fluids 57 (2011) 129136

[15] R. Perrini, I. Morone-Fortunato, E. Lorusso, P. Avato, Glands, essential oils and


in vitro establishment of Helichrysum italicum (Roth) G. Don ssp. italicum, Industrial Crops and Products 29 (2009) 395403.
[16] J. Paolini, J.M. Desjobert, J. Costa, A.F. Bernardini, C.B. Castellini, P.L. Cioni, G.
Flamini, I. Morelli, Composition of essential oils of Helichrysum italicum (Roth)
G. Don l subsp. italicum from Tuscan archipelago islands, Flavour Fragrance J.
21 (2006) 805808.
[17] B. Marongiu, A. Piras, S. Porcedda, Comparative analysis of the oil and supercritical CO2 extract of Artemisia arborescens L. and Helichrysum splendidum (Thunb.)
Less., Natural Product Research 20 (2006) 421428.
[18] B. Marongiu, A. Piras, E. Desogus, S. Porcedda, Analysis of the volatile concentrate of the leaves and owers of Helichrysum italicum (Roth) Don ssp.
microphyllum (Willd.) Nyman (Asteraceae) by supercritical uid extraction and
their essential oils, J. Essential Oil Research 15 (2003) 120126.
[19] A. Bianchini, P. Tomi, J. Costa, A.F. Bernardini, Composition of Helichrysum
italicum (Roth) G. Don l. subsp. italicum essential oils from Corsica (France),
Flavour Fragrance J. 16 (2001) 3034.
[20] V. Roussis, M. Tsoukatou, P.V. Petrakis, I. Chinou, M. Skoula, B. Jeffrey, Harborne, Volatile constituents of four Helichrysum species growing in Greece,
Biochemical Systematics and Ecology 28 (2000) 163175.
[21] M. Aslan, D.D. Orhan, N. Orhan, E. Sezik, E. Yesilada, In vivo antidiabetic
and antioxidant potential of Helichrysum plicatum ssp. plicatum capitulums
in streptozotocin-induced-diabetic rats, J. Ethnopharmacology 109 (2007)
5459.
[22] M. Bonanomi, B. Silvestrini, C. Ghiara, V. Mercati, Water insoluble extract, process for preparing the same and uses thereof, WO 2007/083190 (2004).
[23] E. Lemberkovics, E. Czinner, K. Szentmihalyi, A. Balazsa, E. Szokea, Comparative evaluation of Helichrysi os herbal extracts as dietary sources of
plant polyphenols, and macro- and microelements, Food Chemistry 78 (2002)
119127.
[24] A.J. Al-Rehaily, O.A. Albishi, M.M. El-Olemy, J.S. Mossa, Flavonoids and
terpenoids from Helichrysum forskahlii, Phytochemistry 69 (2008) 1910
1914.
[25] E. Reverchon, I. De Marco, Supercritical uid extraction and fractionation of
natural matter, J. Supercritical Fluids 38 (2006) 146166.
[26] R. Eggers, Supercritical uid extraction of oilseeds/lipids, in: Supercritical Fluid
Technology in Oil and Lipid Chemistry, AOCS Press, Champaign, IL, USA, 1996,
pp. 3560.
[27] J.W. Goodrum, M.B. Kilgo, Peanut oil extraction with SC-CO2 : solubility and
kinetic functions, Transactions of the ASAE 30 (1987) 18651868.
[28] B. Nagy, Modelling of Supercritical Fluid Extraction, Ph.D. Thesis, Budapest
University of Technology and Economics Faculty of Chemical and Bioengineering Gyrgy Olh Doctoral School, 2010 (http://www.omikk.bme.hu/
collections/phd/Vegyeszmernoki es Biomernoki Kar/2010/).
[29] H.P. He, H. Corke, J.G. CAI, Supercritical carbon dioxide extraction of oil and
squalene from Amaranthus grain, J. Agricultural and Food Chemistry 51 (2003)
79217925.

R.S. Mohamed, P. Mazzafera, Supercritical carbon dioxide


[30] M.D.A. Saldana,
extraction of methylxanthines from mat tea leaves, Brazilian J. Chemical Engineering 17 (2000) 251259.
[31] S.K. Yoda, M.O.M. Marques, A.J. Petenate, M.A.A. Meireles, Supercritical uid
extraction from Stevia rebaudiana Bertoni using CO2 and CO2 + water: extraction kinetics and identication of extracted components, J. Food Engineering
57 (2003) 125134.
[32] S. Balachandran, S.E. Kentish, R. Mawson, The effects of both preparation
method and season on the supercritical extraction of ginger, Separation and
Purication Technology 48 (2006) 94105.
[33] G. Leeke, F. Gaspar, R. Santos, Inuence of water on the extraction of essential oils from a model herb using supercritical carbon dioxide, Industrial &
Engineering Chemical Research 41 (2002) 20332039.
[34] D. Hrncic, M. Mernik, M. Knez Hrncic, Z . Knez, Fitting Sovovas mass transfer
model using an evolutionary algorithm and differential evolution, International
J. Innovative Computing and Applications 2 (2010) 237243.
[35] H. Sovov, Rate of vegetable oil extraction with supercritical CO2 -modelling of
extraction curves, Chemical Engineering Science 49 (1994) 409414.
[36] H. Sovov, J. Kucera, J. Jez, Rate of the vegetable oil extraction with supercritical CO2 -II. Extraction of grape oil, Chemical Engineering Science 49 (1994)
415420.
[37] J. Martnez, J.M. Martnez, Fitting the Sovovs supercritical uid extraction
model by means of a global optimization tool, Computers and Chemical Engineering 32 (2008) 17351745.
[38] I. Zizovic, M. Stamenic, A. Orlovic, D. Skala, Supercritical carbon dioxide extraction of essential oils from plants with secretory ducts: mathematical modelling
on the micro-scale, J. Supercritical Fluids 39 (2007) 338346.
[39] J.A. Jossi, L.I. Stiel, G. Thodos, The viscosity of pure substances in the dense
gaseous and liquid phases, AIChE J. 8 (1962) 5963.
[40] O.J. Catchpole, Measurement and correlation of binary diffusion coefcients
in near critical uids, Industrial & Engineering Chemistry Research 33 (1994)
18281837.
[41] J. Ivanovic, I. Zizovic, M. Ristic, M. Stamenic, D. Skala, The analysis of simultaneous clove/oregano and clove/thyme supercritical extraction, J. Supercritical
Fluids 55 (2011) 983991.
[42] Automated Mass Spectral Deconvolution and Identication System software
(AMDIS ver. 2.1), National Institute of Standards and Technology (NIST), Standard Reference Data Program, Gaithersburg, MD, USA, 2005.
[43] R.P. Adams, Identication of Essential Oil Components by Gas Chromatography/Mass Spectrometry, 4th ed., Allured Publishing Corporation, Carol Stream,
IL, USA, 2007.
[44] R. Wiebe, The binary system carbon dioxidewater under pressure, Chemical
Reviews 29 (1941) 475481.
[45] T.J. Smyth, R.G. Zytner, W.H. Stiver, Inuence of water on the supercritical uid
extraction of naphthalene from soil, J. Hazardous Materials 67 (1999) 183196.