Pilot plant study of recovery of lactic acid from ethyl lactate

Article abstract of DOI:10.1021/op800312k

Purified lactic acid is used for production of biodegradable polymer. Esterification with ethanol and subsequent hydrolysis in distillation columns to produce purified lactic acid without catalyst has obvious advantages. In this paper, we present a pilot-

Full text of DOI:10.1021/op800312k

Organic Process Research & Development 2009, 13, 573–575
Pilot Plant Study of Recovery of Lactic Acid from Ethyl Lactate
Prashant P. Barve, Imran Rahman,* and Bhaskar D. Kulkarni
Chemical Engineering and Process DeVelopment DiVision, National Chemical Laboratory (NCL), Pune - 411 008, India
Abstract:
Purified lactic acid is used for production of biodegradable
polymer. Esterification with ethanol and subsequent hydrolysis in
distillation columns to produce purified lactic acid without catalyst
has obvious advantages. In this paper, we present a pilot-plant
study of ethyl lactate hydrolysis to produce 3.86 kg/h lactic acid
(99.85% purity) using three distillation columns. Simulation of
distillation columns at steady state has been carried out, and the
results obtained tally with the experimental results of the pilot
plant.
1. Introduction
Figure 1. Schematic of reactive distillation column.
Polylactic acid, a biodegradable polymer, as a substitute for
petrochemicals-based polymers is beneficial from the environ-
obtained by incorporating the reaction terms in the mass and
mental viewpoint. Production of biodegradable polymers is
heat balance equations of the conventional model for distillation.
becoming increasingly important as a result of diminishing
The kinetic data and vapor-liquid equilibrium for esterification
resources of crude oil.1 Therefore, alternative renewable biomass
of lactic acid described by Delgado et al.8,9 respectively, have
and its transformation through fermentation to L-lactic acid have
been used.
created a lot of interest.
Purification of dilute lactic acid obtained from bacterial
fermentation is difficult due to its low vapor pressure, tendency
to undergo self-esterification, and the presence of troublesome
impurities.1 Esterification of crude lactic acid obtained from the
fermented broth and subsequent distillation of its ester followed
by hydrolysis yield alcohol and pure lactic acid.2
2. Experimental Section
2.1. Materials. Ethyl lactate (>99% purity) was obtained
from Godavari Sugar Mill, India.
2.2. Analysis. Analysis of ethanol and ethyl lactate was
performed on a gas chromatograph (Shimadzu GC-14 B)
equipped with a flame ionization detector. A DB-1, 30 m
Purification of lactic acid through reactive batch distillation
column (J&W Scientific, U.S.A.) was used with nitrogen as a
was investigated by several investigators.1,3,4 Earlier Schopmeyer
carrier gas at a flow rate of 1.5 mL/min. The injector and
et al.5 suggested esterification of lactic acid (40-60 wt %) in
detector temperatures were maintained at 220 and 300 °C,
the presence of a homogeneous catalyst in a jacketed kettle.
respectively. The oven temperature was varied over the range
The homogeneous catalyst used was sulfuric acid which leads to
120-240 °C with the ramp rate of 4 °C/min. The HPLC
corrosion problems. Li et al.6 and Kumar et al.7 investigated
analysis for free lactic acid was performed on a C18 Hypersil
the esterification of lactic acid with methanol in a reactor,
BDS column using a Dionex HPLC system comprising Dionex
followed by hydrolysis of methyl lactate using a cation-
Summit P 680A pump and Dionex ASI 100 autosampler. The
mobile phase consisted of 0.01 M potassium dihydrogen
phosphate (pH ) 2.5) with 7% acetonitrile as an organic
modifier. The chromatographic peaks were monitored using a
exchange resin as catalyst in a continuous column.
In this paper, an experimental study on hydrolysis of ethyl
lactate on a pilot-plant scale (3.8 kg/h) at steady state is reported
and compared with simulation results. The hydrolysis section
UV detector at a detection wavelength of 205 nm.
consists of three columns. The distillation simulation model is
2.3. Apparatus. Figure 1 shows a schematic representation
of the distillation assembly. The distillation assembly consisted
* Author for correspondence. E-mail: r.imran@ncl.res.in.
(1) Seo, Y.; Hong, W. H.; Hong, H. H. Korean J. Chem. Eng. 1999, 16,
556–561.
of a glass column (i.d. ) 25 mm, height ) 1000 mm) packed
with glass packing (10 stages) attached to a jacketed mild steel
glass-lined (MSGL) reboiler (6 L capacity). The distillation
column was equipped with a distillate receiver and packed with
Raschig rings supplied by M/s Silicaware Pvt. Ltd., Mumbai
(2) Vikroy, T. B. Lactic Acid. In ComprehensiVe Biotechnology; Moo-
Young, M., Ed.; Pergamon Press: Oxford, 1985.
(3) Kim, J. Y.; Kim, Y. J.; Hong, W. H.; Wozny, G. Biotechnol.
Bioprocess Eng. 2000, 5, 196–201.
(4) Kim, Y. J.; Hong, W. H.; Wozny, G. Korean J. Chem. Eng. 2002, 19,
808–814.
(5) Schopmeyer, H. H.; Arnold, C. R. U.S. Patent 235,370, 1944.
(6) Li, M. A.; Yang, Z.; Jichu, Y. Chin. J. Chem. Eng. 2005, 13, 24–31.
(7) Kumar, R.; Nanavati, H.; Noronha, S. B.; Mahajani, S. M. J. Chem.
Technol. Biotechnol. 2006, 81, 1767–1777.
(8) Delgado, P.; Sanz, M. T.; Beltran, S. Fluid Phase Equilibr. 2007, 255,
17–23.
(9) Delgado, P.; Sanz, M. T.; Beltran, S. Chem. Eng. J. 2007, 126, 111–
118.
10.1021/op800312k CCC: $40.75  2009 American Chemical Society
Published on Web 03/27/2009
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
573

Table 1. Experimental details for hydrolysis of ethyl lactate
run
F1(kg/h) F2(kg/h) D1(kg/h) B3(kg/h) R1(kg/h) R2(kg/h) TB1(°C) TT1(°C) TB2(°C) TT2(°C) TB3(°C) TT3(°C)
RN1
RN2
RN3
RN4
RN5
3.284
3.528
3.432
3.209
3.240
2.520
2.696
2.572
2.560
2.588
1.896
1.760
1.848
1.828
1.827
3.864
4.164
4.044
3.964
3.976
1.253
1.272
1.276
1.267
1.293
4.770
4.889
4.807
4.81
95.00
95.00
95.00
95.60
95.20
85.50
84.70
85.50
85.90
86.70
98.30
98.60
98.30
98.50
98.40
92.00
92.70
92.10
92.10
91.90
106.50
108.00
106.50
108.40
106.80
96.30
96.60
96.30
96.10
97.10
4.85
(India). To avoid heat losses to the surroundings, the column
was insulated with an asbestos sheet. Sensors (Pt-100) were
provided at the top and bottom of the column to measure
temperatures. Reflux was measured by means of a reflux-splitter
with timer arrangement and attached to SCADA (Supervisory
Control and Data Acquisition). Each column was provided with
feed pumps for feed to the column at a controlled rate and a
reboiler with an overflow outlet to maintain a constant level
and allow product withdrawal. Recycle streams were measured
by weighing on a balance and were fed to the column by a
metering pump. Condensers were equipped with a chiller
capable of achieving a condenser temperature of +6 °C.
2.4. Experiment in Three Hydrolysis Columns. Continu-
ous reactive hydrolysis of ethyl lactate was performed in pilot-
scale columns shown in Figure 1 to produce 3.86 kg/h of pure
lactic acid (65% w/w) as per the following reaction scheme.
3. Results and Discussion
Simulation of reactive distillation processes involves the
simultaneous solution of material and energy balances and
stoichiometric relationships.10 This corresponds to the solution
of a considerably large set of nonlinear equations combined
with the thermodynamic model for predicting the vapor-liquid
equilibrium (VLE). It is recognized that the chemical reaction
may be distributed over several stages. VLE data were correlated
using the UNIQUAC model to describe the chemical and phase
equilibria simultaneously.8,9 Material balance equations were
solved simultaneously using the Newton-Raphson method for
obtaining component mole fraction on each stage.10 The energy
balance equation was solved to obtain the vapor flow rate. The
temperature on each stage was computed by the Newton-
Raphson method.
The lactic acid esterification with ethanol without addition
of a catalyst is assumed to follow the Arrhenius rate law with
ko1 ) 5.298 × 108 mol g- 1 min - 1, EA,1 ) 64.44 kJ/mol.8
The expression for the reaction equilibrium constant was
obtained by Delgado et al.9
The assumptions made for modeling the reactive distillation
column are the following: the vapor and liquid phases are in
equilibrium on each tray; no reaction occurs in the vapor phase;
the liquid phase is always homogeneous; the heat of reaction
is considered negligible.
Ethyl Lactate + Water h Lactic acid + Ethanol (1)
The feed to the first stage of the hydrolysis column was ethyl
lactate and distillate from hydrolysis columns 2 and 3. The
distillate of the first column consisted of ethanol and water. The
overflow of reboiler 1 was fed to reboiler 2, and distillate was
returned to reboiler 1. Again,the overflow of reboiler 2 was fed
to reboiler 3, and distillate was recycled to reboiler 1.
At column temperature, 90% of the ethyl lactate undergoes
hydrolysis. As the equilibrium was disturbed by fractionating
the alcohol, the hydrolysis process progresses in an effort to
restore the alcohol to the equilibrium concentration. Ethanol
and water vapors rising in the column pass through the
condenser and are collected as distillate. The overflow of
reboiler 1 containing lactic acid, ethyl lactate, ethanol, and water
moves to the reboiler 2. In the second column approximately
another 10% of the ethyl lactate was hydrolysed, and the
remaining ethyl lactate along with lactic acid, water, and ethanol
was fed to column 3. The distillate of column 2 consisted of
ethanol, water, and ethyl lactate and was removed by partial
condenser. A water stream was also added in reboiler 3 to
complete the hydrolysis of the remaining ethyl lactate. Pure
lactic acid (65-70%) is drawn off from the reboiler and sent
to an evaporator for concentration to the desired strength. The
distillate which contains water, ethanol, and ethyl lactate was
recycled to reboiler 1.
Convergence of the column model depends on starting
guesses for the composition profile and flow rates in the column.
The use of the Newton-Raphson method requires a proper
starting guess for the composition profile.
The algorithm ends up when global material balance to the
column is verified. A nonplausible column is obtained when
the Newton-Raphson method exceeds a prespecified maximum
number of iterations failing to satisfy the global material balance.
3.1. Comparison of Experimental and Simulation Re-
sults. The experimental data obtained during the pilot plant trials
at steady state for five different runs are given in Table 1.
Typically it takes about 5-6 h to attain the steady state. The
operating parameters in Table 1 were obtained after conducting
a number of pilot-plant trials. The parameters which gave the
best performance are considered optimal and reported in Table
1. F1 is the flow rate of ethyl lactate into reboiler 1 while F2 is
the D.M. water flow rate into reboiler 3. D1 is distillate from
column 1 consisting of ethanol and water. The product
consisting of lactic acid and water is withdrawn from reboiler,
B3. The bottom and top temperatures of columns 1, 2, and 3
are denoted by TT1TT2TT3 and TB1TB2TB3, respectively. Distillate
from columns 2 and 3 are recycled stream (R1 and R2) to reboiler
1. Columns 1 and 2 are operated under conditions of reflux
Initially, ethyl lactate and distilled mixture were fed to
column. Data collection started once all the column attained
steady state. Samples from the top and bottom streams were
collected at regular intervals and analyzed for the product
composition. Both bottom and distillate flow rates were
measured, and the overall material balance was verified.
The objective to obtain the pure lactic acid at the bottom
with 100% conversion of ethyl lactate to lactic acid is achieved
in this three-column configuration.
(10) Suzuki, I.; Yagi, H.; Komatsu, H.; Hirata, M. J. J. Chem. Eng. Jpn.
1971, 4, 26.
574
•
Vol. 13, No. 3, 2009 / Organic Process Research & Development

Table 2. Experimental and simulation details for hydrolysis of ethyl lactate
feed distillate composition bottom composition
temperature (°C)
bottom flow distillate flow
column
1
F1
F2
LA
EtOH EL water LA EtOH EL water bottom
top
rate (kg/h)
rate (kg/h)
sim. 3.284
expt.
-
0.2216 0.0477 0.2040 0.5260
0.2790 0.1070 0.1290 0.4840
0.4400 0.0273 0.0418 0.5170
0.4100 0.0410 0.0610 0.4900
0
0.6460
0.6870
-
0
0
0.3533
0.3420
-
95.00 80.00
93.50 80.00
98.70 80.00
100.00 80.00
106.10 92.72
104.70 96.76
-
-
-
1.896
1.955
-
0
2
3
sim.
expt.
sim.
expt.
-
-
-
-
-
-
-
-
-
-
-
-
-
2.520 0.6530 0.0000 0.0000 0.3470
0.6530 0.0000 0.0000 0.3470
-
-
3.864
3.787
-
-
-
Table 3. Column performance for variation in flow rate of RN1 by simulation
distillate composition bottom composition
temperature (°C)
bottom flow distillate flow
LA
EtOH EL water LA EtOH EL
water
bottom
top
rate (kg/h)
rate (kg/h)
column 1 0.3070 0.0790 0.1050 0.5090 0.6950 0.0050 0.3050 95.00
0
80.00
-
2.029
equal to 1, while column 3 is a stripper. The lactic acid purity
found by HPLC was 99.85%.
by experiments and simulation. The advantage of this process
is that the complete hydrolysis of ethyl lactate is achieved
without using catalyst and thus avoids corrosion problems. The
optimal operating parameters for the process are fixed, and the
process is verified by simulation.
Three hydrolysis columns were individually simulated for
a specified feed, bottom and distillate. Distillate from column
2 and 3 is returned to reboiler of columns 1. These two recycled
streams were determined by material balance calculations. The
experimental conditions and the results of a typical run to
produce 3.86 kg/h of lactic acid (65% w/w) are given in Table
2. The simulation results matches with experimental results.
As the reflux ratio increased from 1 to 1.5 in column
1, a decrease in ethyl lactate hydrolysis from 90% to 86%
was observed. A sensitivity study was also carried out by
changing the flow rate of feed (about (3%) through
simulation for RN1. It was observed that a -3.0% change
in flow rate does not affect the conversion of ethyl lactate.
When there is a +3.0% change in ethyl lactate feed, the
expected column configuration shown for RN1 in Table
1 is simulated. In the distillate the presence of ethyl lactate
is noticed as shown in Table 3, which can be reduced to
zero by increasing the number of stages to 15.
NOMENCLATURE
EA,1
EL
EtOH
F1
apparent activation energy, kJ/mol
ethyl lactate
ethanol
feed to reboiler 1
feed to reboiler 3
F2
k1o
pre-exponential factor for the esterification reaction,
mol/g min
LA
N
lactic acid
number of stages
4. Conclusions
This paper demonstrates a process for obtaining lactic acid
from ethyl lactate. The feasibility of the process is evaluated
Received for review December 16, 2008.
OP800312K
Vol. 13, No. 3, 2009 / Organic Process Research & Development
•
575