SEPARATION SCIENCE AND ENGINEERING

Chinese Journal of Chemical Engineering, 17(3) 359365 (2009)

Proposed Models for Subcritical Water Extraction of Essential Oils
M. Khajenoori, A. Haghighi Asl
*
and F. Hormozi
Chemical Engineering Department, Faculty of Engineering, Semnan University, P. O. Box: 35195-363, Semnan, Iran
Abstract Mechanisms that control the extraction rate of essential oil from Zataria multiflora Boiss. (Z. multiflora)
with subcritical water (SW) were studied. The extraction curves at different solvent flow rates were used to deter-
mine whether the extractions were limited primarily by the near equilibrium partitioning of the analyte between the
matrix and solvent (i.e. partitioning thermodynamics) or by the rates of analyte desorption from the matrix (i.e. ki-
netics). Four simple models have been applied to describe the extraction profiles obtained with SW: (1) a model
based solely on the thermodynamic distribution coefficient K
D
, which assumes that analyte desorption from the ma-
trix is rapid compared to elution; (2) one-site kinetic model, which assumes that the extraction rate is limited by the
analyte desorption rate from the matrix, and is not limited by the thermodynamic (K
D
) partitioning that occurs dur-
ing elution; (3) two-site kinetic model and (4) external mass transfer resistance model. For SW extraction, the
thermodynamic elution of analytes from the matrix was the prevailing mechanism as evidenced by the fact that ex-
traction rates increased proportionally with the SW flow rate. This was also confirmed by the fact that simple re-
moval calculations based on determined K
D
(for major essential oil compounds) gave good fits to experimental data
for flow rates from 1 to 4 mlmin
1
. The results suggested that the overall extraction mechanism was influenced by
solute partitioning equilibrium with external mass transfer through liquid film.
Keywords essential oils, Zataria multiflora Boiss., subcritical water extraction, mechanism
1 INTRODUCTION
Continuous subcritical water extraction (SCWE)
is a technique with water as extractant, at temperatures
between 100 and 374C and pressure high enough to
maintain the liquid state [1]. Under such conditions,
the intermolecular hydrogen bonds of water are bro-
ken, causing water polarity to decrease. As a result,
water becomes a more effective solvent for several
organic compounds. The review on extraction of
medical botanicals with subcritical solvents has re-
cently been available [2].
Zataria multiflora Boiss. (Z. multiflora) belongs
to the family Labiatae and it is native to Iran. Z. mul-
tiflora is used traditionally in food, especially in yo-
ghurt flavouring. There are also commercial pharma-
ceuticals with formulae based on Z. multiflora essen-
tial oil [3]. Thymol and carvacrol are two major con-
stituents in most essential oils, including oils used in
variety of drugs. In our previous study, we have
shown the feasibility of extracting thymol and car-
vacrol from Z. multiflora with subcritical water (SW)
and the effects of various factors such as temperature
and flow rate on extraction efficiency of this com-
pound were determined [4]. In the present work, the
extraction mechanism of thymol and carvacrol from Z.
multiflora using SCWE is studied. The extraction
curves at different water flow rates are used to deter-
mine whether the extractions are limited primarily by
the near equilibrium partitioning of the analyte be-
tween the matrix and solvent (i.e. partitioning ther-
modynamics) or by the rates of analyte desorption
from the matrix (i.e. kinetics).
Four simple models are employed to test the data.
One model is attempted to predict the extraction rates
based on the thermodynamic distribution coefficient
(K
D
), and the other model is tried to predict the extrac-
tion rates using a one-site exponential kinetic model
(k). The third model is a two-site exponential kinetic
model, which attempts to predict the extraction rates
using a fast and a slow kinetic rate constant, and the
forth model is the thermodynamic partition with ex-
ternal mass transfer model [5].
The purpose of this paper is to elucidate the
mechanisms controlling the extraction rates achieved
with SCWE at different flow rates and same tempera-
ture 150C, with the mean particle size 0.5 mm and
pressure 2 MPa. The relative importance of the diffu-
sion and external mass transfer step are determined
during SCWE by varying the extraction flow rate.
2 MATERIALS AND METHODS
2.1 Materials
Leaves of Z. multiflora were collected (Shiraz,
Iran) in May 2006. The moisture content of the leaves
was 5% (dry basis). The samples were ground by
grinder and screened by standard sieves immediately
prior to extraction in order to avoid losses of volatiles.
Two replications of the extraction and analysis proce-
dure were performed for each run. Thymol and car-
vacrol (from Roth, Germany) were used as internal
standard. NaCl, Na
2
SO
4
and n-pentene (from Merck,
Germany) were used as demulsifier, drying agent and
extractant respectively, in the liquid-liquid extraction
step of the aqueous extracts. HPLC grade hexane (Al-
drich Chemical Co., USA) was used as diluting sol-
vent before gas chromatography (GC). The doubly
distilled, de-gassed water purified through a Milli-Q
Received 2008-10-27, accepted 2009-03-02.
* To whom correspondence should be addressed. E-mail: ahaghighi@semnan.ac.ir
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 360
de-ionizing unit (Millipore, Bedford, MA, USA) was
used as the extractant.
2.2 Subcritical water extraction system
The subcritical water extractions were carried out
in a laboratory-built apparatus [Iranian Research Or-
ganization for Science and Technology (IRSOT), Te-
hran, Iran] shown in Fig. 1.
De-ionized water filled into a stainless steel feed
tank was first purged for 2 h with N
2
to remove dis-
solved O
2
. A Dosapro Milton Roy (H9 series, USA)
high pressure pump was used to deliver the extractant
water through the system at a constant flow rate of 2
and 4 mlmin
1
. The water was preheated in a
stainless steel coil (3 m3 mm i.d.). The extractor was
a stainless steel cylindrical extraction chamber (103
mm16 mm i.d.). The solid bed inside the extractor
was fixed with ring screws at both ends in order to
permit the circulation of the water through it. The
main body of the extractor was closed with screw caps
at both ends. The heating system (up to 200C) was a
fan-equipped temperature-controlled oven (Teb Azma,
Tehran, Iran). A double pipe heat exchanger (240 cm
2
heat transfer surface area) was used to cool the extract
immediately after coming out from the oven to a tem-
perature close to 20C. The cooling medium was wa-
ter (15C, 3 Lmin
1
flow rate). A 50 cm length
stainless steel tube (1 mm i.d.) was used before a
pressure regulator. All parts in contact with the ex-
tractant water were stainless steel 316.
2.3 Subcritical water extractions
For all subcritical water extractions, the extractor
was filled with 4.0 g Z. multiflora. The extractor was
assembled in the oven and pressurized by closing the
valves 4, 6 and the end line regulator, and opening
valve 5. Then, valve 4 was opened and the water was
pumped until the system reached 2 MPa again. After
that, the pump was turned off, valve 5 was closed,
valves 4 and 6 were fully opened, and the oven was
brought up to the required temperature, a process that
was required 20 min. Then, the pressure regulator was
opened, the pump was turned on, and the flow rate
was adjusted at the desired rate. Regarding to the se-
lected flow rate and a void volume between the ex-
tractor and collection vessel (60 ml), in the all runs,
around 20 ml of the water coming out of the system
was discarded and the extraction process was sup-
posed to be started at that time (extraction time 0 = ).
Using GC analysis, it was proved that it was free from
any components. After collecting the required volume
of extract, a liquid-liquid extraction step using n-pentene
was carried out. The volumetric ratio of n-pentene to
extract was 12 in all experiments and extractions
were completed by two equal volume of solvent in
two steps. Around 8 g NaCl was added to facilitate
breaking of the emulsion. Essential oil was stored in 1
ml HPLC grade hexane before GC analysis. For the
kinetic experiments, the collection vial was replaced at
appropriate time intervals.
3 ANALYSIS
The GC-flame ionization detection (FID) analy-
sis were performed using a Varian Model CP-3800 gas
chromatograph equipped with a 60 m CP Sil 8 CB
fused silica column (0.32 mm i.d., 0.25 m film thick-
ness). An injection volume of 1.0 l of the hexane ex-
tracts was injected using auto sampler. The oven tem-
perature program was a 3Cmin
1
temperature ramp
from 50 to 230C. The carrier gas was nitrogen
Figure 1 Schematic diagram of subcritical water extraction system
1pressure regulator; 2needle valve (NV); 3micro filter; 4burette; 5pump; 6check valve; 7relief valve; 8pressure
indicator; 9oven; 10fan; 11temperature indicator controller; 12extraction cell; 13temperature indicator
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 361
(99.999%, Roham Gas Co., Tehran, Iran). The column
head pressure was 70 kPa. The detector and injector
temperatures were 250C and 230C, respectively.
GC-mass spectrometry (MS) analysis was con-
ducted on a Varian Saturn Model 3400. GC-MS sys-
tem equipped with a DB-5 fused silica column (30
m0.25 mm, film thickness 0.25 m) and interfaced
with a Varian ion trap detector. The GC conditions
were the followings: increase of oven temperature
from 40 to 200C at 4Cmin
1
, injector and transfer
line temperature 210 and 220C respectively, helium
as carrier gas with a flow rate 40 mlmin
1
, and split-
ting ratio 113. The detector temperature was main-
tained at 240C. The MS conditions were the follow-
ings: ionization energy 70 eV, mass range 40400 amu,
and scan mode EI. The percent of composition of the
identified components was calculated from the GC
peak area without considering response factors. The
components were identified by comparing their reten-
tion time and mass spectrum with those of pure refer-
ence components. Mass spectra were also compared
with those in the NIST (National Institute of Standards
and Technology), WILEY5 and TERPENOIDES mass
spectra libraries and our own created library.
4 MASS TRANSFER MODELS
Four main mass transfer steps are generally in-
volved: (1) diffusion of solute through a stagnant liq-
uid film around the solid plant particles; (2) diffusion
of solvent into solid particles through the pores; (3)
diffusion of the dissolved solute from within particles
to the particle surface through the pores; (4) removal
by partition from the particle surface into the bulk
solvent [6]. The effect of step (1) is typically small and
often neglected. Although the diffusion of the dis-
solved solute within the solid is usually the rate limit-
ing step for most botanicals [7, 8], partitioning of sol-
ute between the solid matrix and solvent have been
reported as the rate-limiting mechanism for subcritical
water extraction of essential oil from savory [5].
The relative importance of these steps can be de-
termined by the plots of the amount of compound ex-
tracted versus solvent flow rates and solvent volume.
For example, if the rate of extraction is controlled by
intra-particle diffusion or kinetic desorption, the in-
crease in flow rate of bulk fluid would have little ef-
fect on extraction rate. On the other hand, if the ex-
traction is controlled by external film transfer diffu-
sion, extraction rates increase with solvent flow rate.
In the case where the extraction rate is controlled by
thermodynamic partitioning, doubling the flow rate of
bulk fluid would double the extraction rate, while the
curves of extraction efficiency versus the volume of
water passed for all flow rates would overlap. In this
study, four models will be considered and used to fit
with the experimental data. These include (1) parti-
tioning coefficient model, (2) one-site, (3) two-site
desorption models and (4) thermodynamic partition
with external mass transfer model.
5 MODELING OF SCWE OF BIOACTIVES
FROM PLANT MATERIALS
5.1 Thermodynamic model
The simple thermodynamic model is based on a
single distribution coefficient defined as K
D
= (con-
centration of analyte in the matrix)/(concentration of
analyte in the extraction fluid) at equilibrium [9]. For
this model, it is assumed that the rates of the initial
desorption step and subsequent fluid-matrix partition-
ing are rapid, and thus do not significantly affect the
extraction rate. Essentially, the mass of analyte in each
unit mass of extraction fluid and the mass of analyte
remaining in the matrix at that period in the extraction
time is calculated for the entire extraction time based
on the K
D
value determined for each compound.
Therefore, if the thermodynamic model applies to a
certain extraction, the shape of an extraction curve
would be defined by:
( )
a
0 b a
D 0 0
b a
1
1
S
S S S
K m S S
V V
| |

|
\ .
= +
(
+
(

(1)
where S
a
is the cumulative mass of the analyte extracted
after certain amount of volume V
a
(mgg
1
based on
dry sample), S
b
the cumulative mass of the analyte
extracted after certain amount of volume V
b
(mgg
1
,
based on dry sample), S
0
the total initial mass of ana-
lyte in the matrix (mgg
1
, based on sample), S
b
/S
0
and
S
a
/S
0
are the cumulative fraction of the analyte extracted
by the fluid of the volume V
b
and V
a
(ml), respectively,
K
D
is the distribution coefficient, concentration in
matrix/concentration in fluid, is the density of ex-
traction fluid under given condition (mgml
1
), and m
is the mass of the extracted sample (mg dry sample).
5.2 Kinetic mode
5.2.1 One-site kinetic desorption model
One-site kinetic desorption model describes the
extractions that are controlled by intra-particle diffu-
sion. This occurs when the flow of fluid is fast enough
for the concentration of a particular solute to be well
below its thermodynamically controlled limit. The
one-site kinetic model was derived based on the mass
transfer model that is analogous to the hot ball heat
transfer model [10, 11]. The assumptions are that the
compound is initially uniformly distributed within the
matrix and that, as soon as extraction begins, the con-
centration of compound at the matrix surfaces is zero
(corresponding to no solubility limitation). For a
spherical matrix of uniform size, the solution for the
ratio of the mass, S
r
, of the compound that remains in
the matrix sphere after extraction time, t, to that of the
initial mass of extractable compound, S
0
, is given as:
( )
r 2 2 2
e 2 2
1 0
6 1
exp /
n
S
D n t r
S n

=
= t
t

(2)
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 362
in which n is an integer and D
e
is the effective diffu-
sion coefficient of the compound in the material of the
sphere (m
2
s
1
).
The curve for the above solution tends to become
linear at longer time (generally after t0.5t
c
), and ln
(S
r
/S
0
) is given approximately by
( )
r 0 0
ln / 0.4977 / S S t t = (3)
where t
0
is the initial time and t
c
is a characteristic
time (min), defined as:
2 2
c e
/ t r D = t (4)
An alternative form of Eq. (3), or so called a
one-site kinetic desorption model, can be written for
the ratio of mass of analyte removed after time t to the
initial mass S
0
, as given by:
0
1 e
kt t
S
S

= (5)
in which S
t
is the mass of the analyte removed by the
extraction fluid after time t (mgg
1
dry sample), S
0
is
the total initial mass of analyte in the matrix (mgg
1
dry sample), S
t
/S
0
is the fraction of the solute extracted
after time t, and k is a first order rate constant de-
scribing the extraction (min
1
).
5.2.2 Two-site kinetic desorption model
Two-site kinetic model is a simple modification
of the one-site kinetic desorption model, which de-
scribes the extraction occurring from the fast and
slow part [5]. In such case, a certain fraction (F) of
the analyte desorbs at a fast rate defined by k
1
, and the
remaining fraction (1 F) desorbs at a slower rate
defined by k
2
. The model has the following form:
( ) ( )
2
1
0
1
e
e 1
t k t
k t
S
F F
S

= (

(6)
The two site kinetic model does not include sol-
vent volume, but relies solely on extraction time.
Therefore, doubling the extractant flow rate should
have little effect on the extraction efficiency when
plotted as a function of time. On the contrary, the
thermodynamic model is only dependent on the vol-
ume of extractant used. Therefore, the extraction rate
can be varied by changing the flow rate. Hence, the
mechanism of thermodynamic elution and diffusion
kinetics can be compared simply by changing the flow
rate in SCWE. If the concentration of bioactive com-
pounds in the extract increases proportionally with the
flow rate at given extraction time when the solute
concentration is plotted versus extraction time, the
extraction mechanism can be explained by the ther-
modynamic model. However, if an increase in flow
rate has no significant effect on the extraction of the
bioactive compounds, with the other extraction parame-
ters being kept constant, the extraction mechanism can
be modeled by the two site kinetic model [12, 5]. The
mechanism of control and hence the model valid for
SCWE may be different depending on the raw mate-
rial, the target analyte and extraction conditions.
5.3 Thermodynamic partition with external mass
transfer resistance model
This model describes the extraction controlled by
external mass transfer, whose rate is described by re-
sistance type model of the following form:
( )
s
e p s D
/
C
k a C K C
t
c
( =

c
(7)
in which C is the fluid phase concentration (molm
3
),
C
s
is the solid phase concentration (molm
3
), k
e
is the
external mass transfer coefficient (mmin
1
), and a
p
is
the specific surface area of particles (m
2
m
3
) [13]. If
the concentration of the solute in the bulk fluid is as-
sumed small and the ratio of solute concentration in
the liquid to that at the surface of solid matrix is de-
scribed by partitioning equilibrium, K
D
, the solution of
Eq. (7) for the solute concentration in the solid matrix,
C
s
, becomes:
( ) e p D s 0
/ exp k a t K C C = (8)
Equation (8) can be rewritten as the ratio of the
mass of diffusing solute leaving the sample to the ini-
tial mass of solute in the sample, S
t
/S
0
, as given by the
following equation.
( ) e p D 0
/ 1 exp
t
k a t K S S = (9)
Because a
p
is difficult to be measured accurately,
a
p
and k
e
are usually determined together as k
e
a
p
,
which is called overall volumetric mass transfer coef-
ficient. The factors that influence the value of k
e
a
p
include the water flow rate through the extractor and
the size and shape of plant sample.
6 RESULTS AND DISCUSSION
Figure 2 shows extraction curves generated from
SCWE of thymol and carvacrol compounds of Z. mul-
tiflora (at a flow rate of 2 mlmin
1
). While it is
tempting, based on these plots, to assign the thermo-
dynamic K
D
model to the SCWE, we cannot make the
interpretation based only on the results in Fig. 2. The
reason is that, without knowledge of the effect of flow
rate, the relative importance of the desorption kinetics
Figure 2 Comparison of SCWE profiles for major repre-
sentative essential oil compounds from Z. multiflora
(flow rate of 2 mlmin
1
, temperature 150C, mean particle
sizes 0.5 mm, pressure 2 MPa)
thymol;carvacrol
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 363
and the extraction curves for SCWE could be de-
scribed by a single site kinetic model, as well as the
single K
D
model proposed above.
6.1 Effect of flow rate
Based on the discussion above, the importance of
K
D
and desorption kinetics was determined by com-
paring the effects of changing flow rate on the extrac-
tion rate of the same samples (Fig. 3). As can be seen,
the rate of essential oil extraction was faster at the
higher flow rate. It is in accordance with the previous
work. It means that the mass transfer of essential oil
components from the surface of the solid phase into
the water phase regulated most of the extraction proc-
ess. Increase of flow rate resulted in increase of super-
ficial velocity and thus, quicker mass transfer [12]. In
practice, the best flow rate must be chosen with two
important factors, extraction time and final extract
concentration. Shorter extraction time and more con-
centrated final extract will be preferable.
(a) Thymol
(b) Carvacrol
Figure 3 Effect of extraction fluid flow rate on SCWE of
thymol and carvacrol from Z. multiflora
(temperature 150C, mean particle sizes 0.5 mm, pressure 2 MPa)
Q/mlmin
1
:1;2;3;4
6.2 Partitioning coefficient (K
D
) model
The model Eq. (1) and the experimental data
from all volumetric flow rate plots were used to de-
termine the K
D
value by minimizing the errors be-
tween the measured data and the K
D
model using
Matlab curve fitting solver. The values of K
D
are
shown in Table 1 for different flow rates. It was dem-
onstrated that individual essential oil compounds have
a range of K
D
values from4 to250 [5]. The K
D
model agreed reasonably with the experimental data
(the average error 8%9%). Nevertheless, if the ex-
traction is strictly controlled by partitioning equilib-
rium, K
D
values for all flow rates must be equal. The
deviation found was possibly due to the existence of
external film transfer resistance, whose model would
be discussed later.
In addition, when the K
D
model was applied to
the SCWE of thymol and carvacrol from Z. multiflora,
the calculated extraction curves and the experimental
curves also show good agreement in Fig. 4. Also K
D
values of thymol and carvacrol were nearly similar,
because they were structural isomer and had similar
behaviour in extraction.
(a) Thymol
exp, Q= 4 mlmin
1
, K
D
= 2;exp, Q= 2 mlmin
1
,
K
D
= 80;exp, Q= 1 mlmin
1
, K
D
= 80
(b) Carvacrol
exp, Q= 4 mlmin
1
, K
D
= 2;exp, Q= 2 mlmin
1
,
K
D
= 70;exp, Q= 1 mlmin
1
, K
D
= 70
Figure 4 Experimental data and K
D
model for different
flow rates of SCWE of thymol and carvacrol from Z. multi-
flora
(temperature 150C, mean particle sizes 0.5 mm, pressure 2 MPa)
The effect of different values of the thermody-
namic distribution coefficient (K
D
) on extraction rates
(with a constant flow rate of 2 mlmin
1
) for thymol
extraction is shown in Fig. 5. As expected, a higher K
D
(stronger competition of the matrix versus the fluid for
the solute) yields slower extraction rates. Based on a
comparison of Fig. 5 with the experimental data shown
Table 1 K
D
values of partitioning coefficient model for
different volumetric flow rates
K
D Flow rate/
mlmin
1
Thymol Carvacrol
1 80 70
2 80 70
4 2 2
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 364
in Figs. 3 and 4, it appears that the K
D
model shows a
general extraction curve shape which is the typical
behavior of SCWE.
6.3 One-site kinetic desorption model
Matlab curve fitting solver was used to determine
the desorption rate constant, k, from the data for all
flow rates. The values are show in Table 2. As men-
tioned, the kinetic desorption model does not include a
factor describing extraction flow rate, k should be the
same value for all flow rates if the model is said to fit
the experimental data. However, this was not the case
(Table 2, the average error 3%17%). The kinetic de-
sorption rate increased for the volumetric flow rate of
1 to 4 mlmin
1
. This indicated that the kinetic de-
sorption model may not be suitable for describing the
data at different flow rates of Z. multiflora.
Table 2 Values of k for one-site kinetic desorption model
for different volumetric flow rates
k/min
1
Flow rate/
mlmin
1
Thymol Carvacrol
1 0.0025 0.0028
2 0.0042 0.0039
4 0.0157 0.0157
6.4 Two-site kinetic desorption model
For the two-site kinetic desorption model, the
values of k
1
and k
2
were determined by fitting the ex-
perimental data with the two-site kinetic desorption
models by minimizing the errors between the data and
the model results. In the two-site model, the extraction
rate should not be dependent on the flow rate. The k
1
and k
2
values shown in Tables 3 and 4 demonstrated
that the extraction rates were not completely inde-
pendent of flow rate (the average error 11%20%).
Table 3 k
1
and k
2
values of two-site kinetic desorption
model for thymol at different flow rates
Flow rate/mlmin
1
k
1
/min
1
k
2
/min
1
Mole fraction F
1 0.0088 0.0015 0.21
2 0.0152 0.0026 0.28
4 0.0770 0.0083 0.27
Table 4 k
1
and k
2
values of two-site kinetic desorption
model for carvacrol at different flow rates
Flow rate/mlmin
1
k
1
/min
1
k
2
/min
1
Mole fraction F
1 0.0101 0.0017 0.21
2 0.0747 0.0088 0.42
4 0.0469 0.0082 0.27
6.5 Thermodynamic partition with external mass
transfer model
The values for the model parameters, K
D
and k
e
a
p
in Eq. (9) determined by Matlab curve fitting solver
from the experimental data obtained at 150C are
summarized in Tables 5 and 6 for different mass flow
rates (Q
m
, mgmin
1
). Linear regression of the plot
between ln(k
e
a
p
) and lnQ gives the following correla-
tion for k
e
a
p
and Q:
for thymol
0.2078
e p m
6.5748 k a Q = (10)
for carvacrol
0.6017
e p m
0.1605 k a Q = . (11)
Table 5 Parameters K
D
and k
e
a
p
for external mass transfer
model of SCWE of thymol
Flow rate/
mlmin
1
Mass flow rate
Q
m
/mgmin
1
Parameter K
D
Parameter
k
e
a
p
/min
1
1 938 80 26.700
2 1876 80 32.7975
4 3752 2 1.300
Table 6 Parameters K
D
and k
e
a
p
for external mass transfer
model of SCWE of carvacrol
Flow rate/
mlmin
1
Mass flow
rate/mgmin
1
Parameter K
D
Parameter
k
e
a
p
/min
1
1 938 70 8.92
2 1876 70 62.013
4 3752 2 20.54
7 COMPARISON OF EXTRACTION MODELS
To quantitatively compare the extraction models,
the mean percentage errors between the experimental
data and the models were considered. Based on the
result in fitting from experimental data, the K
D
model
was generally suitable for the description of extraction
over all the volumetric flow rates tested. On the other
Figure 5 Theoretical curves calculated using Eq. (1) for
thymol extractions from Z. multiflora, controlled by ther-
modynamic partitioning
(flow rate of 2 mlmin
1
, temperature 150C, mean particle
sizes 0.5 mm, pressure 2 MPa)
K
D
= 2; K
D
= 20; K
D
= 50; K
D
= 75;
K
D
= 100; K
D
= 150; + exp, Q= 2 mlmin
1
Chin. J. Chem. Eng., Vol. 17, No. 3, June 2009 365
hand, one-site and two-site kinetic desorption models
describe the extraction data reasonably at lower volu-
metric flow rates. Of all the models considered, however,
the thermodynamic partition with external mass trans-
fer model could best describe the experimental data.
8 CONCLUSIONS
In summary, subcritical water provides a promis-
ing alternative for extraction of the thymol and car-
vacrol from Z. multiflora. Extraction mechanisms
were investigated at 150C, 14 mlmin
1
flow rate
and 0.50 mm mean particle size for 150 min extraction
time. Overall by considering mean average errors of
models, a mathematical model base on the combina-
tion of partition coefficient (K
D
) and external mass
transfer gave a good description of subcritical water
extraction of Z. multiflora, while the kinetic model
reasonably described the extraction behavior at lower
flow rates.
ACKNOWLEDGEMENTS
Financial and technical support is gratefully ac-
knowledged to the Semnan University and the Iranian
Research Organization for Science and Technology
(IROST).
REFERENCES
1 Ayala, R.S., Luquede Castro, M.D., Continuous subcritical water
extraction as a useful tool for isolation of edible essential oils, Food
Chem., 75, 109113 (2001).
2 Ong, E.S., Han Cheong, J.S., Goh, D., Pressurized hot water ex-
traction of bioactive or marker compounds in botanicals and me-
dicinal plant materials, J. Chromatography A, 1112, 92102 (2006).
3 Aynehchi, Y., Pharmcognosy and Medicinal Plants of Iran, Tehran
University Press, 228234, Tehran, Iran (1991).
4 Khajenoori, M., Haghighi Asl, A., Hormozi, F., Eikani, M.H., Noori
Bidgoli, H., Subcritical water extraction of essential oils from
Zataria multiflora Boiss, J. Food Process Eng., 10.111/j. 17454530.
2008. 00245. X (2008).
5 Kubatova, A., Jansen, B., Vaudoisot, J.F., Hawthorne, S.B., Ther-
modynamic and kinetic models for the extraction of essential oil
from savory and polycyclic aromatic hydrocarbons from soil with
hot (subcritical) water and supercritical CO
2
, J. Chromatography A.,
975 (1), 175188 (2002).
6 Shotipruk, A., Kiatsongserm, J., Pavassnt, P., Goto, M., Sasaki, M.,
Pressurized hot water extraction of anthraquinones from the roots
of Morinda citrifolia, Biotechnol. Prog., 20, 18721875 (2004).
7 Schwartzberg, H.G., Chao, R.Y., Solute diffusivities in leaching
process, Food Technol., 36, 7386 (1982).
8 Gertenbach, D.D., Solid-liquid extraction technologies for manu-
facturing nutraceuticals, Shi, J., Mazza, G., Maguer, M.L., eds.,
Functional Foods: Biochemical and Processing Aspects (Volume 2),
CRC Press, Boca Raton, Flordia (2002).
9 Windal, I., Miller, D.J., de Pauw, E., Hawthorne, S.B., Supercritical
fluid extraction and accelerated solvent extraction of dioxins from
high- and low-carbon fly ash, Anal. Chem. 72, 39163921 (2000).
10 Crank, J., The Mathematics of Diffusion, Clarendon, Oxford (1975).
11 Carlslaw, H.S., Jaeger, J.C., Conduction of Heat in Solids, Clarendon,
Oxford (1959).
12 Cacace, J.E., Mazza, G., Pressurized low polarity water extraction
of lignans from whole flaxseed, J. Food. Eng., 77, 10871095 (2006).
13 Anekpankul, Th., Goto, M., Sasaki, M., Pavasant, P., Shotipruk, A.,
Extraction of anti-cancer damnacathal from roots of Morinda citri-
folia by subcritical water, Sep. Purif. Technol., 55, 343349 (2007).