Optimisation of steam extraction of oil from maritime pine needles

Text-only Preview

Author manuscript, published in "Journal of Wood Chemistry and Technology 29, 2 (2009) 87-100"
DOI : 10.1080/0277381092879025
Optimisation of steam extraction of oil from maritime pine needles

University of La Rochelle, LEPTIAB.
EA 4226, UFR Sciences. Avenue M. Crepeau, 17042 La Rochelle, France
tel: 33 (5) 46 45 86 15; fax: 33(5) 46 45 86 16; e-mail : [email protected]
Essential oil from pine maritime needles is generally extracted by steam distillation
process at atmospheric pressure for more than one hour, or by solvent extraction
process. In the last decade, there has been an increasing demand for new extraction
techniques enabling automation, shorter extraction time and reduced consumption of
organic solvent. In this study, Response Surface Methodology (RSM) was used to
evaluate the effects of two processing parameters of an alternative extraction process:
instantaneous controlled pressure drop: "Détente Instantanée Contrôlée" (D.I.C) on the
yield and composition of oil isolated from maritime pine needles (Pinus pinaster). This
process involves subjecting the substrate for a short time to steam varying from 1.5 to
5.5 bar (113 to 155 °C) for 4 to 20 minutes, followed by an instantaneous
decompression to a vacuum (about 50 mbar). We studied the effect of processing
hal-00414223, version 1 - 8 Sep 2009
pressure and processing time on the yield of oil and in three important compounds: ?-
pinene, ?-pinene and germacrene D. Both the processing pressure and time had a
significant effect on all responses studied. For the less volatile compound, ?-pinene, the
maximum quantity was obtained at the lower processing pressure and time, while an
inverse trend was observed for ?-pinene and germacrene D. The models displayed by
the experimental design gave R2 higher than 0.92.
Keywords : Maritime pine (Pinus pinaster), isolation, vacuum, oil, D.I.C process


Essential oil is any class of volatile oil of complex hydrocarbons, mainly terpenes and
some other chemicals which are isolated from plants. One of their characteristics is the
generation of flavour or aroma. Essential oils extracted from plants such as pines are
used as fragrances in cosmetics, flavouring additives of foods and beverages, and
scenting agents in a variety of household products including detergents, soaps or insect
repellent. They are also used as intermediate in the synthesis of perfume chemicals and
for unconventional medicinal purposes as well as in aromatherapy1,2. The conventional
methods for extracting essential oils have some disadvantages. For steam distillation
and hydrodistillation, elevated temperatures can cause chemical modifications of oil
components and a loss of the most volatile compounds3. When using solvent extraction,
it is impossible to obtain a solvent-free products and this process usually also results in
the loss of volatile components. In contrast, extraction by supercritical fluids leads to
high-quality and solvent-free extracts4. However, according to Temelli et al.5 and
Oszagyan et al.6, supercritical fluid extraction is costly. Moreover, several studies7 have
shown that CO2 not only extracts essential oil, but also other compounds such as waxes
hal-00414223, version 1 - 8 Sep 2009
or resins.
The Instantaneous Controlled Pressure Drop process, known as "D.I.C", was developed
and patented in our laboratory some years ago8-9. This process subjects the product to
rapid transition from high steam pressure to vacuum. This transition induces a fast
evaporation of water and volatile compounds. In a previous work10, we showed that
processing by instantaneous controlled pressure drop increases the global diffusivity of
the product and improves the availability of the liquid in the plant.


The essential oil isolation based on this process is an interesting alternative to standard
techniques of essential oil extraction, such as extraction with solvents or steam
distillation. This is because it does not use solvent, and induced cooling when the plant
is rapidly transferred from a high steam pressure to vacuum minimizes thermal
degradation of the essential oil components. Moreover, compared to the steam
distillation, the short time contact (few minutes) between plant and the heated zones of
apparatus avoids the loss and degradation of volatile and thermolabile compounds.
The objectives of this study were to evaluate the effect of two independents process
variables, processing water steam pressure and processing time on : i) the yield of oil
isolated from maritime pine needles; ii) the composition of the isolated oil in three
commercially important compounds, ?-pinene, ?-pinene and germacrene D. In addition,
a mathematical model11-13 predicting the yield allowed the optimisation of the extraction
2.1 Plant material
Maritime pine (Pinus pinaster) was collected from plants growing in west southern
hal-00414223, version 1 - 8 Sep 2009
France. The needles were used at their residual moisture content (63.1 % dry wt basis).
The compounds identified by steam distillation (2.4 section) and their yields are shown
in table 1. The yield of essential oil in fresh raw material was 0.82 % by mass (g of
isolated oil/100 g dm). This value is in agreement with the values cited by Kelkar et al.14
and Dob et al.15.
2.2 Experimental set-up
The experimental set-up was largely described in a recent paper16. It is composed of
three main elements:


The processing vessel where samples were placed and treated.
The vacuum system which consists mainly from vacuum tank with volume (360
l) 130 fold greater than the processing vessel (12 l), and a vacuum pump. The
initial vacuum pressure of vacuum container was maintained at 50 mbar in all
A pneumatic valve that separate the processing vessel from vacuum tank. It can
be opened in less than 0.2 seconds; this ensures a rapid decompression within
the reactor.
2.3 Protocol of extraction by the instantaneous controlled pressure drop process
The needles are placed in the D.I.C vessel which is maintained under vacuum (~ 50
mbar) through its connection to a vacuum container. The vacuum allows a better
diffusion of heating fluid through plant and consequently heat transfer between steam
and wood is improved and time to reach the desired processing pressure (or processing
temperature) is shortened. After closing the electropneumatic valve which connects the
reactor to the vacuum tank, an atmosphere of steam pressure (between 1 and 6 bar in
hal-00414223, version 1 - 8 Sep 2009
this study) is created within the D.I.C. reactor. After a processing time at fixed
processing pressure, the thermal treatment is followed by a rapid decompression
resulting in a rapid drop in pressure. The equilibrium pressure after decompression
depends on operating pressure: the higher the processing pressure, the higher the
equilibrium pressure. Evaporation, which is effected in adiabatic conditions, induces a
rapid cooling of the residual product and the final temperature must be commensurate
with final pressure. Extract and condensed steam are recovered in a specific container.
The volume of obtained mixture was about 400 ml for all experiments.
2.4 Steam distillation


50 g of maritime pine needles chips were placed on a stainless steel grid. This grid was
placed in a glass chamber containing boiling water. The steam crossed the grid during
two hours and was recovered along with volatiles after crossing a refrigerant. The
condensates were separated into aqueous and organic phases by decantation.
2.5 GC/MS conditions
A Varian 3900 gas chromatograph coupled to a Varian Saturn 2100T ion trap mass
spectrometer (Varian, France) was used. The column was a 30m × 0.25 mm, 0.25 µm
CP-Sil 8 CB Low Bleed MS capillary column (Varian, France). Oven temperature 80
°C for 3 minute then programmed from 50 °C to 250 °C at 3 °C/min, then hold at 250
°C for 40 min. Helium as carrier gas at 1 ml/min was used. The extract samples were
injected via a Varian CP-8400 autosampler fitted with a 5µl syringe. Transfer line
temperature was 280°C. Electron impact mass spectra were obtained at 70 eV ionization
potential and peak identity was identified by NIST 2002 Data library.
2.6 Scanning Electron Microscopy
A Philips-FEI Quanta 200 ESEM/FEG Scanning Electron Microscopy operated at 20
kV, with a detector of secondary electrons Everhardt-Thornley, was used to image the
hal-00414223, version 1 - 8 Sep 2009
control sample and some treated maritime pine needles. To improve the quality of the
SEM images, a high vacuum was achieved.
2.7 Experimental design
A central composite rotatable design was developed to evaluate effects of processing
pressure (P) and processing time (t). The design needed 13 experiments with eight (22)
factorial points, four extra points (star points) to form a central composite design and
five replications for the central point. The experiments were run in random order to


minimize effects of unexpected variability due to extraneous factors. In the full factorial
design, processing pressure (P) values varied between 1.5 and 5.5 bar and processing
time (t) between 4 and 20 minutes (table 1). Variables were codified in the way that
their values ranged between –? and +? (?= 1.414), taking as the central point zero.
Thus, P*= (P-3.5)/2 and t*= (t-12)/8; where P, t are the actual values and P*, t* the coded
values of processing pressure and processing time. To avoid thermal reactions we
limited the maximum processing pressure to 6.3 bar, corresponding to a temperature of
160 °C.

Table 2 shows the central composite design matrix, with variables in both coded/non
coded forms. Data were adjusted to a response surfaces which were obtained by using
analysis design procedure of 5.1 version Statgraphics Plus for Windows software17.
?= a0 + a1P + a2 t + a12 Pt + a11P2 + a22 t2
? is the considered response and a0 is the value of the objective function in the central
point conditions. a1 and a2 represent the principal effects associated to the two variables,
hal-00414223, version 1 - 8 Sep 2009
a12 represent the crossed effect among the variables and a11, a22 represent the quadratic
effects of the two studied variables. Thus, the model coefficients reflected the linear,
quadratic and interactive effects.
To evaluate the effect of drying (in an oven at 25 °C) on the composition of the essential
oil extract, extraction by steam distillation were carried out on three samples of
maritime pine needles, the first on fresh moist product, the second on needles dried to
40 % moisture content (d.b), and the third on needles dried to 10 % moisture content.


The results are grouped in table 3. It appears that the extraction yield decreases at lower
moisture content. For the most important compound, ?-pinene, the yield decreased
from 40.5 % for fresh needles to 22.5 % (g of / 100 g isolated oil) for needles dried to
10 % moisture content (d.b). This may be due to some evaporation of the compound
during drying. Thus, for extraction by instantaneous controlled pressure drop process
the needles were treated in fresh state (~63 % d.b).
The results obtained following the experimental design are grouped in table 3. The
validity of the results was confirmed by the low uncertainty limit in the extraction yield
(based on low error estimate) obtained from five replications at processing pressure of
3.5 bar and processing time of 12 minutes.
3.1 Fitting the models
A regression analysis was carried out to fit mathematical models to the experimental
data aiming at an optimal region for the responses studied. The predicted models can be
described by table 4 in term of coded values. The significance of each coefficient was
determined using Fisher test (F-value) and the probability p (p-value) in table 5 which
hal-00414223, version 1 - 8 Sep 2009
displays the variance analysis of the system (ANOVA). Corresponding variables would
be more significant if absolute F-value becomes greater and p-value becomes smaller. It
can be seen for all responses, that processing pressure and processing time have a strong
linear effects. For yield of isolated oil, a significant (p<0.05) quadratic effect of
processing time was also observed, indicating that the yield increases with the
processing time up to a certain value beyond which a diminution is observed due to
thermal degradation. For a processing pressure fixed at its central value, this time
corresponds to 15 min. ?-pinene and ?-pinene exhibited a significant quadratic effect of
processing time. The results suggest that changing in processing pressure or time had a


highly significant effect on the yield of isolated oil and of the three selected compounds.
The coefficients of determinations of models were also given in ANOVA table (table 5).
They were systematically higher than 92 % for the four models suggesting a good fit;
the predicted models seemed to reasonably represent the observed values. Thus the
responses were sufficiently explained by the models.
3.2 Response surfaces for yield of isolated oil
The regression models displayed in table 4 allowed the prediction effect of the two
processing parameters of D.I.C. isolation process. The relationship between independent
and dependant variables can be illustrated in three dimensional representations of the
response surfaces generated by the models (fig 1 and 2).
Fig.1, shows that both processing pressure and time demonstrated a significant linear
increase on yield of isolated oil with the strongest effect for processing pressure. When
processing pressure increases from 1.5 to 5.5 bar, extraction yield increases from 0.2 to
1.8 % according to processing time. Fig. 2 shows that it is possible to obtain a high
yield at low processing times but at high processing pressures (> 4-5 bar). This indicates
hal-00414223, version 1 - 8 Sep 2009
that the mechanical strain induced by the rapid decompression and the brutal
vaporization of water have two main effects: the dehydrating effect due to vaporization
and a subsequent change in the surface tension of the glandular wall, causing it to
crumble or rupture more readily. Similar effects were pointed out by Pare et al.18 for
microwave extraction. The authors reported that an explosion at cell level occurred as a
consequence of the sudden temperature rise generated by microwaves. The same
observation was cited by Spiro and Chen19, who reported that the oil synthesized in the
secretory cells was not released unless an external factor damages the gland. This
observation was also verified by Boutekedjiret et al.20, whom compared the isolation of


rosemary oil by different extraction processes, including the Instantaneous Controlled
Pressure Drop Process. The microstructure of the maritime pine needles is showed in
fig.3 where it can be seen that the size of pores increased with the severity of the
thermomechnaical extraction. The results also show that compared with other "flash"
extraction processes, extraction by D.I.C is more efficient since a high extraction yield
can be obtained at low processing times. It should be noted that steam distillation at
industrial scale is performed in two from three hours. Chen and Spiro21 who worked on
microwave extraction of essential oil from rosemary leaves reported that a long time at
high temperatures could cause rearrangement or polymerization of some of rosemary oil
constituents which are close to the constituents present in the needles of maritime pine
oil. At 4.5 bar processing pressure, 10 minutes are sufficient to extract more than 80 %
of available essential oil. The maximum of extraction yield is almost reached after 15
minutes processing time. Beyond this value, a certain degradation expressed by a more
deepened colouring of oil was observed. By looking figure 1 more closely, we can
observe that the yield evolution versus processing time show a very rapid increase
during the first minutes of isolation process, then gradually levelled to equilibrium value
hal-00414223, version 1 - 8 Sep 2009
at the end of the process. Isolation of oil from maritime pine needles by instantaneous
controlled pressure drop process seems to be regulated by two distinct phenomena
corresponding to two steps. The first one is rapid compared to the second and
corresponds to a free diffusion phenomenon which takes place at the plant surface
The oil recovered in the second step is probably regulated by osmosis phenomena and
slow diffusion through the plant cells towards the surface. However it can be seen that
the oil collected in last step (~5 %) is very low compared to the one collected in first
step (~95 %). It is obvious that the major part of the oil is recovered by a simple process


of free diffusion and evaporation. The proportions of these two parts for steam
distillation extraction are generally lower for the first step and higher for the second
one. This difference may be attributed to the presence of saturated steam under pressure
in the case of instantaneous controlled pressure drop process that allows reaching more
endogenous sites than at atmospheric pressure in the case of steam distillation.
3.3 Response surfaces for the amounts of the three studied compounds.
From an industrial point of view, the qualitative criterion of maritime pine extracts is
based on presence of three compounds namely ?-pinene, ?-pinene and germacrene D,
in defined percentage in the essential oil as commercial standards. The ?-pinene must
represent between 33 and 43 % of isolated oil, ?-pinene between 22 and 32 % and
germacrene D between 0.5 and 4 %. For this reason, the effect of processing pressure
and time on the quantity of these compounds was studied. Fig. 2 (a) represents the three
responses surfaces displayed by generated models for the three compounds. For ?-
pinene, the strongest effect is that of processing time followed by a visible quadratic
effect of processing time whatever the processing pressure suggesting that a large part
hal-00414223, version 1 - 8 Sep 2009
of this compound is located on the surface of the naturally broken glands (exogenous
sites) and are easily extracted according to a free diffusion phenomenon on the plant
surface. Moreover the decreasing of the quantity of ?-pinene clearly indicates a certain
degradation of this compound, which is more volatile. Comelli et al22 reported a same
behaviour for the isomerization reaction of ?-pinene which produces bicyclic and
monocyclic compounds and other products, in presence of catalyst and temperature. For
all selected processing times, the quantity of ?-pinene decreased strongly up to 15 min
and then stabilise. For ?-pinene, which is also an important compound in the maritime