L. Giorno et al. / Journal of Catalysis 247 (2007) 194–200
197
pH, stirring rate) were monitored and controlled during the ex-
periments.
ratio between the standard deviation and the average value over
eight injections of the same sample, was 0.9%.
When an immobilized enzyme was used, a two-separate-
phase membrane reactor was applied (Fig. 1b). The system
comprised an enzyme-loaded membrane module containing
asymmetric membranes made of polyamide. The enzyme was
immobilized by cross-flow ultrafiltration, and the amount of en-
zyme entrapped within the membrane was determined based on
the mass balance between the initial and final solutions. The or-
ganic phase contained the substrate and was recirculated along
the shell side while the aqueous phase extracted the product and
was recirculated along the lumen side. The two phases were
contacted and kept separated at the membrane level. The re-
action was monitored by measuring the product extracted into
the aqueous phase. In this case, observed properties instead of
intrinsic ones were considered, because transport phenomena
through the membrane to the bulk phase also occur. The data
representation neglected transport phenomena, because the re-
action rate and mass transfer rate measurements demonstrated
that the immobilized system still worked in limited reaction
conditions. Therefore, only some delay in the reaction rate was
observed, but the catalytic performance was not affected by the
mass transfer properties.
The total mass of the reaction product at time t was calcu-
lated as follows:
ꢀ
mgtot(t) = Ct Vtot
Here Ct = concentration at time t; Vtot = total aqueous reaction
volume at time t; Cn = concentration of a sample; Vn = volume
of a sample.
+
CnVn.
(1)
The product mass as a function of time allow calculation
of the total conversion, c, the enantiomeric excess, ee, and the
enantioselectivity, E,
mg(S+R
)
)
acid
c =
× 100,
(2)
(3)
mg(S+R
ester
S − R
eep =
× 100.
S + R
Here eep is the enantiomeric excess of the product and S and R
represent the mass of (S)- and (R)-naproxen acids produced.
The enantioselectivity was calculated from the general equa-
tion [10]
ln[1 − c(1 + eep)]
E =
.
(4)
ln[1 − c(1 − eep)]
2.3.1. Lipase activity measurements
For low conversion values, this can be simplified to
The free lipase was tested in the stirred tank reactor. The pH
and temperature were maintained constant at 7.00 and 30 ◦C.
When olive oil was used as a substrate, the reactions were mon-
itored online by titration with NaOH 20 mM using a Mettler
DL25 automatic titrator. The reaction mixture was formed by
19 ml of phosphate buffer 50 mM, 1 ml of olive oil, and 2 ml
of enzyme solution 3 graw powder/L. The mixture was stirred at
500 rpm with a magnetic stirrer to disperse the two immisci-
ble phases and create the oil–water reaction interface. When
synthetic esters were used, the reactions were monitored offline
by taking samples at given intervals of time and then measur-
ing (S)- and (R)-naproxen acid by HPLC.
The reaction mixture comprised 23 ml of phosphate buffer
and 20 ml of substrate solution composed of naproxen ester
5 mM in isooctane and 3 ml of enzyme solution 2 graw powder/L
stirred at 500 rpm.
During the reaction, a 3-ml sample was taken from the aque-
ous phase reaction volume and replaced with 3 ml of a similar
enzyme solution (0.23 mgraw powder/ml) so as to not have a neg-
ative effect on the reaction due to enzyme mass variation. The
samples were filtered with a 50-kDa membrane to stop the re-
action and to purify the sample for HPLC analysis.
The concentration of the chiral product was measured us-
ing a chirobiotic V (250 × 4.6 mm) column protected by the
precolumn chirobiotic V guard (from Astec). The mobile phase
was a mixture of THF and an aqueous solution (pH 7) contain-
ing 0.1% of TEA (10:90 v/v). The analyses were performed at a
flow rate of 1 ml/min and a wavelength of 254 nm. The (S)-na-
proxen acid was eluted faster than the (R)-naproxen acid, with
retention times of 6.5 and 7.5 min, respectively. The concen-
tration of the samples was evaluated using the external standard
method. The error associated with the analyses, estimated as the
S
E =
.
(5)
R
The standard deviation of data and related error were evaluated
over three to five experiments.
3. Results and discussion
This section presents the experimental results obtained with
the free and immobilized lipase used for the kinetic resolu-
tion of the three synthesized esters (methyl, butyl, and octyl
naproxen esters). First, the native catalytic activity of lipase
is characterized with its natural substrate to verify the native
enzyme efficiency. Then the catalytic properties of lipase are
studied using racemic esters with different –OR length groups
synthesized in our laboratories. The different substrates were
used with both the free and immobilized enzymes.
3.1. Catalytic properties of lipase with olive oil
Experiments with olive oil were carried out to evaluate its
native catalytic performance. The experiments were carried out
at 30 ( 1)◦C and pH 7.00. The reaction mixture was composed
of 21 ml of phosphate buffer containing 1.02 mg of proteins
(0.048 mgprotein/ml, with the value taking into account the ratio
of protein with respect to the raw powder) and 1 ml of olive oil
(acidity <1%).
The concentration of acid produced as a function of time
is illustrated in Fig. 2 (with the error calculated over a series
of five experiments). The enzyme had a reaction rate value of
3.2×10−2 ( 8.6×10−3) mmol/ml min and a specific activity
of 2.1 ( 0.55) mmol/minmgprotein
.
Giorno
