RESULTS AND DISCUSSION
Extraction of FFAs from the microalga N. gaditana FFAs
from the microalga N. gaditana were extracted by the procedure
shown in Fig. 1, which is an adaptation to this microalga of a similar
process applied to the microalgae Phaeodactylum tricornutum (11)
and Isochrysis galbana (13). By this procedure 83.4 3.1 wt% of
fatty acids contained in the microalgal biomass were recovered
and these FFAs were extracted with 73.5 2.6 wt% purity. On the
other hand, saponifiable lipids (SLs) without transforming in FFAs
were also extracted in our laboratory by a similar two-step
process: (i) extraction of lipid from the wet microalgal biomass
with ethanol (96% v/v) and (ii) purification of these extracted
lipids by a second extraction with hexane. This method achieved
a similar SL recovery yield (85%), but the SL purity was only 31%,
the impurities consisting mainly of unsaponifiable lipids
(carotenoids), which obviously can not be transformed into
methyl esters (biodiesel). The transformation of SLs into FFAs by
direct saponification in the biomass (Fig. 1) allows easier
separation of FFAs and unsaponifiable lipids, and, therefore, a
higher purity of the final biodiesel. Besides, the extraction of SLs
as FFAs does not imply an excessive increasing of the economical
cost because only an alkali, as NaOH or KOH, have to be added to
the extraction solvent (ethanol) to transform SLs to FFAs.
Influence of reaction time The experiments of optimization
of the enzymatic reaction conditions were carried out with FFAs
from UVO (used vegetable oil) and the best conditions were applied
to FFAs extracted from the microalgal biomass. Novozym 435 was
chosen to catalyse this reaction because it is one of the most widely
used lipases in industry and for biodiesel production (7,14e17).
These experiments were firstly carried out at small scale (4 g of
FFAs from UVO), using a 1.5:1 methanol/FFA molar ratio and a 0.1
w/w lipase/FFA ratio (40C and 200 rpm). In these conditions it
was observed that for a reaction time of 0.5 h a 92.8% conversion
was attained, and this conversion kept practically constant until
24 h (around 94.6% at 2, 6 and 24 h). This reaction was scaled-up
maintaining the same operational conditions. Fig. 2 shows that at
0.5 h the conversion is again over 90% and the equilibrium yields
attained at both scales are similar (about 95%).
In the following experiment, the conditions previously applied
to FFAs from UVO were applied to FFAs extracted from the microalga
N. gaditana, although in this case 14.3 g of microalgal FFAs were
used (Fig. 2). It was observed that the esterification degree (ED)
exceeded 90% at 0.25 h of reaction, attaining values of over 95% at
only 0.5 h, i.e., the reaction velocity and conversions were equal to
or better than those obtained with FFAs from UVO.
These data (Fig. 2) show the high velocity of transformation of
FFAs to methyl esters, which is much faster than the transesterification
reaction velocity (acylglycerols þ alcohol). This reaction
velocity is higher than that obtained by Da Rós et al. (15) in
the production of biodiesel from cyanobacterial lipids, also using
Novozym 435; these authors attained an ethyl ester yield of 98.1%
after reaction times of 48 h using iso-octane as reaction media. This
result is logical because, in fact, the formation of methyl esters from
SLs by transesterification occurs in a two-step process, that is, the
oil is first hydrolysed to FFAs and partial glycerides and the FFAs
produced are then esterified with methanol (18). By the procedure
assayed in this work the first step of hydrolysis is avoided.
Reuse of lipase As mentioned above, it is known that many
lipases are deactivated in the presence of short chain alcohols, such
as methanol and ethanol. This is particularly noticeable when no
solvent is used, since it is the non-dissolved alcohol which deactivates
the lipase (7). To test the stability of Novozym 435 in the
esterification of FFAs with methanol, experiments were carried out
in which the same batch of Novozym 435 was used to catalyse
successive reactions. In these experiments the operational
conditions shown in Fig. 2 were maintained constant and the
influence of washing the biomass between reactions was also
tested. The washing was carried out with an acetone/ethanol 1:1
(v/v) mixture (19). Table 2 shows the ED attained in six uses of
the same batch of lipase in reactions of 2 and 6 h, both with and
without washing. Table 2 shows that there was no appreciable
diminution of the ED. Neither was the conversion affected by
washing the lipase between the catalysed reactions, and therefore
washing is not considered necessary before reuse. These results
show the stability of lipase Novozym 435 in the operational
conditions required for the esterification of FFAs to produce
methyl esters (biodiesel) in absence of solvent, since the ED
remained constant at approximately 94% over at least six
reactions catalysed by the same batch of lipase. This stability
contrasts, for example, with the loss of activity of this lipase
during repeated experiments carried out by Du et al. (16) in the
transesterification of soybean oil with methanol. According to
these authors this activity loss might be due to the inactivation
effect caused by methanol and the negative effect caused by the
by-product glycerol on the surface of immobilized lipase.
Therefore, this higher stability of the lipase in the esterification
reaction may be due to factors such as the absence of glycerol in
this reaction and the shorter reaction time of esterification
compared with the transesterification times. On the other hand,
Lin et al. (2) stated that a high content of FFAs could protect the
whole cell of lipase-producing Rhizopus oryzae from denaturation.
In the esterification of oleic acid with methanol, these authors
found that the methanol concentration at which this lipase is
inhibited increased with increasing FFA content.
Influence of the methanol/FFA molar ratio Table 3 shows
the ED attained at increasing methanol/FFA molar ratios. It can be
seen that the higher the molar ratio the higher the ED, since the
esterification equilibrium is displaced toward the formation of
products when the molar ratio increases. The highest ED was
obtained with a methanol/FFA molar ratio of 2. However, the ED
only improved by 1% with respect to the 1.5 molar ratio, although
the amount of methanol increased 33%. Therefore a methanol/FFA
molar ratio of 1.5 was chosen for future experiments. It is
important to operate with a low methanol/FFA ratio, since the
higher the methanol concentration is, the higher the possibility
that the lipase is deactivated, especially when no solvent is used
(7). However, some authors use high molar ratios to increase the
conversion. For instance, Da Rós et al. (15) use an ethanol/lipids
molar ratio of 12:1 (equivalent to 4:1 alcohol/FFA), in the
transesterification of the lipids from the cyanobaterium
Microcystis aeruginosa, catalysed by Novozym 435, although these
authors used tert-butanol or iso-octane to preserve the stability
of the lipase. On the other hand, the low methanol/FFA ratio used
in this work contrasts with the high ratios that are required for
the production of biodiesel by esterification of FFAs by acid
catalysis. Su (20) established an optimal methanol/FFA molar
ratio of 10:1 in the esterification of FFAs from soybean oil
catalysed by hydrochloric acid, and Hayyan et al. (21) used a
methanol/oil molar ratio of 8:1 (about 2.7:1 as equivalent FFAs) in
the transesterification-esterification of sludge palm oil with 23 wt
% FFAs.
Influence of temperature and intensity of treatment Table 4
shows the ED attained at three temperatures and several intensity of
treatments (IOTs). The IOT is the Novozym 435 amount reaction
time/FFA amount, and the reaction velocity is proportional to this
variable if no enzyme deactivation occurs (22).
The temperatures at which Novozym 435 is most frequently
used are 30C (7,17) and 40C (14,16), and therefore both temperatures
were tested in the present work. In addition, 25C was
assayed in an attempt to reduce the reaction temperature and increase
the lipase stability. Table 4 shows that the EDs obtained at
25C (between 92% and 93%) were similar to those obtained at 30
and 40C for the esterification of FFAs with methanol in the
experimental conditions shown in this Table 4. Consequently the
lowest temperature, 25C, was chosen.
The IOTwas modified between 0.05 and 0.5 g Novzoym 435 h/
g FFAs and in this range there was no important influence on the
methyl ester yield, which suggests that the equilibrium was
attained at these IOTs. Therefore an IOT ¼ 0.1 g Novozym 435 h/g
FFAs was chosen. This IOT can be attained, for example, with a reaction
time of 4 h and 0.1 g Novozym 435, for 4 g FFAs (i.e., 2.5 wt%
of Novozym 435 with respect to FFAs), although the reaction time
can be reduced using a larger amount of lipase. In the transesterification
of oils to produce biodiesel, the Novozym 435 amount
is often 4 wt% (7,16) or 10 wt% of Novozym 435 (14) with respect to
oil and reaction times in the range of 2472 h. Therefore, IOTs of
around 23 g Novozym 435 h/g oil are used depending on the oil
treated (7,14,16). However, IOTs for the esterification reaction must
be at least 20 times lower than for the transesterification reaction,
which is logically due to the higher reaction velocity of the esterification
reaction. Véras et al. (23) clearly demonstrate this fact in
the production of fatty acid ethyl esters by simultaneous esterification
and transesterification, catalysed by Novozym 435, of highly
acidic feedstock; the first order velocity constants for the esterification
and transesterification reactions were in the ranges
0.921.20 h1 and 0.110.13 h1, respectively. The conversion of
the FFA fraction to fatty acid ethyl esterswas about 90% at a reaction
time of 4 h with 1.5% w/v of Novozym 435, i.e., an IOT of ar