4 PB
4 PB
3, 270 – 281
DOI: 10.22146/ajche.72250
D, Hartanto, P.A. Handayani, W. Astuti, R.D. Kusumaningtyas, Y.C. Purwana, Maftukhaturrizqiyah, R.T. Wijayanti, 271
D.Z. Wulansari, R. Wulansarie, I.N. Pradnya, D.S. Hadikawuryan, A.A. Wibowo, R.M. Sholihin, A. Chafidz, I. Khoiroh
obtain the optimum configuration. The NRTL thermodynamic model was used in this study. The
effects of the number of stages (NS), binary feed stage (BFS), entrainer feed stage (EFS), and reflux
ratio (RR) on the ethanol concentration with minimum energy requirements were studied. The
most optimal configurations to produce a high ethanol concentration with less energy are NS 28,
BFS 22, EFS 4, and RR 1.5.
Keywords: Dehydration, Ethanol, Extractive Distillation, [BMIM] [Br] Ionic Liquid, Process
Simulation
(Pan et al., 2019). Ethylene glycol was used as needs to be conducted to understand the
a conventional entrainers in extractive process design and its optimum
distillation (Hartanto et al., 2020; Li et al., configuration. In this work, the simulation of
2021). Unfortunately, ethylene glycol poses the extractive distillation of ethanol/water
toxicity that causes depression in the central with [BMIM] [Br] ionic liquid has been studied
nervous system and failure of the cardio- to understand the effect of the number of
pulmonary and renal systems (Leth and stages, binary feed stages, entrainer feed
Gregersen, 2005). Inorganic salts were also stages, and reflux ratio on the purity of
considered potential entrainers, even though ethanol, total annual cost, and the energy
they cause corrosion and incrustation requirement in the condenser and reboiler.
problems (Hernández-Hernández et al., Moreover, the performance of the ionic liquid
2022). Currently, greener entrainers are has been compared with that of the
greatly interested in being used in extractive conventional entrainer.
distillation technology. Ionic liquids have
become a rising star chemical that can be
used as an entrainer to substitute METHOD
conventional solvents. The advantages of
using ionic liquids are high-temperature Thermodynamic Model
stability, the ability to be used in various The software package Aspen Plus V.11
mixtures, a wide temperature range, and no was used in this study. The rigorous method
vapor pressure (Tsanas et al., 2014). In was applied to simulate extractive distillation.
addition, ionic liquids (ILs) are considered a The nonideality condition of the liquid phase
green solvent and have been applied in was calculated using the Non-Random Two-
extractive distillation technology as one of Liquid (NRTL), assuming that the vapor phase
the promising entrainers (Malik et al., 2023). is in the ideal condition. The NRTL is one of
Some recent studies evaluated the use of ILs the most common thermodynamic models
for the ethanol-water separation (Fadia et al., used in liquid phase calculations because of
2022; Graczová and Steltenpohl, 2022; Ma et its capability to produce good calculation
al., 2019; Momeni and Shekaari, 2022). One of results. The provided optimum binary
the ILs that is considered a green solvent is 1- interaction parameters are listed in Table 1.
butyl-3-methylimidazolium bromide [BMIM] The optimum binary interaction parameters
[Br] (Sudhan et al., 2018; Ghorbani et al., are important as they quantify the molecular
2021). Tsanas et al. (2019) measured the interaction behavior and predict the vapor-
effect of 1-butyl-3-methylimidazolium liquid equilibrium properties. Hence, these
bromide [BMIM] [Br] ionic liquid in properties will affect the whole process
ethanol/water vapor-liquid equilibrium. The simulation's accuracy. Furthermore, the non-
results indicate that the [BMIM] [Br] ionic randomness factor (Cij) of NRTL has an
liquid can break the azeotrope point in the optimum range of 0.2–0.5. However, the
ethanol/water system. However, the effect of the randomness factor of this range
simulation study of this system and the on the curve of excess Gibbs energy is not
performance comparison of ionic liquid with strong. Therefore, the founders of the NRTL
conventional entrainers are not available in model recommend 0.3 as the most optimal
the open literature. Therefore, further study one (Renon and Prausnitz, 1968).
D, Hartanto, P.A. Handayani, W. Astuti, R.D. Kusumaningtyas, Y.C. Purwana, Maftukhaturrizqiyah, R.T. Wijayanti, 273
D.Z. Wulansari, R. Wulansarie, I.N. Pradnya, D.S. Hadikawuryan, A.A. Wibowo, R.M. Sholihin, A. Chafidz, I. Khoiroh
The extended Antoine equation other method may also be used to approach
evaluated the total and partial pressures the ionic liquid with a compound with a
(eq.1). similar chemical structure and properties. In
𝐴2
ln(𝑃 𝑠 ) = 𝐴1 + 𝑇+𝐴 + 𝐴4 𝑇 + 𝐴5 ln 𝑇 + addition, the ionic liquid can be well defined
3
Table 1. Optimum binary interaction parameters for the NRTL model (Tsanas et al., 2014)
Components A1 A2 A3 A4 A5 A6 10-6 A7 A8 A9
Etanol 61.7911 -7122.3 0 0 -7.1412 2.88 2 -114 240.85
Water 62.1361 -7258.2 0 0 -7.3037 4.16 2 0.01 373.95
[BMIM] [Br] -1000 0 0 0 0 0 0 - -
274 Extractive distillation of ethanol/water with 1-butyl-3-methylimidazolium bromide ionic liquid as
a separating agent: process simulation
Table 3. Initial process design parameter ratio (RR). The details of the process flow
diagram are provided in Figure 1. The
Parameters Value
aqueous ethanol solution (FEED stream) and
Ethanol feed mole-fraction 0.7
[BMIM] [Br] ionic liquid (SOLVENT stream)
Theoritical stage numbers 30
enter the EDC. In the first column, ethanol
Entrainer mole fraction 0.5
was separated from the water using the
Feed temperature (oC) 25
entrainer. Higher purity of ethanol (ETHA
Entrainer temperature (oC) 70
stream) was obtained in the top column, and
Binary feed stage 21
a mixture of water/[BMIM] [Br] ionic liquid
Entrainer feed stage 3
was achieved in the bottom column. The
Pressure (kPa) 101.3
mixture of water/[BMIM] [Br] ionic liquid
(RICH-SOL) enters the second column,
followed by the separation process. [BMIM]
RESULTS AND DISCUSSION
[Br] ionic liquid was collected in the bottom
column, and water was obtained in the top
Conceptual Process Design and
column. [BMIM] [Br] ionic liquid (IL-1 stream)
Preliminary Simulation
was cooled to meet the desired temperature.
The conceptual process design for the
The cooler [BMIM] [Br] ionic liquid (IL-2) was
separation of ethanol from water using
recycled to the EDC and mixed with the
[BMIM] [Br] ionic liquid as an entrainer
make-up entrainer to keep a sufficient
through extractive distillation consists of two
amount of entrainer in the EDC. In addition,
main columns, which are the extractive
the preliminary simulation was studied in our
distillation column (EDC) as a first column and
previous work to validate the simulation
the entrainer recovery column (ERC) as a
method (Hartanto et al., 2021). The
second column. The RadFrac block, as a
simulation indicates satisfactory results with a
rigorous calculation model, was used to
small deviation compared to the literature.
simulate EDC and ERC. The main parameters
The simulation results for extractive
in EDC that can be used to optimize the
distillation, recovery column, and stream are
separation performance consist of the
provided in Tables 4 and 5, respectively.
number of stages (NS), binary feed stage
(BFS), entrainer feed stage (EFS), and reflux
MIXER
RICH-SOL
FEED
IL-1
COOLER
IL-2
Fig. 1: Extractive distillation process flow diagram for ethanol and water separation using [BMIM]
[Br] ionic liquid as an entrainer.
D, Hartanto, P.A. Handayani, W. Astuti, R.D. Kusumaningtyas, Y.C. Purwana, Maftukhaturrizqiyah, R.T. Wijayanti, 275
D.Z. Wulansari, R. Wulansarie, I.N. Pradnya, D.S. Hadikawuryan, A.A. Wibowo, R.M. Sholihin, A. Chafidz, I. Khoiroh
1525
with an increased reflux ratio. Thus, the
1500
1475 energy duties of the reboiler and condenser
1450
1425
will also increase due to the product
1400 vaporization and distillate condensation
1375
24 25 26 27 28 29 30 31 requirements. The results align with the
NS previous study, which indicated that an
(a) increasing reflux ratio would significantly
increase reboiler and condenser duties (Zhu
et al., 2016; Stewart, 2014). The highest
2525
2500
RR 1,0
RR 1,1
energy duties for the reboiler and condenser
2475 RR 1,2
RR 1,3
occurred in the highest reflux ratio of 1.5.
2450
2425 RR 1,4 The effect of BFS and EFS on ethanol
2400 RR 1,5
purity was studied. Figure 4 shows that the
QR (kW)
2375
2350 ethanol purity remains constant for the BFS
2325
2300 20 until 29. On the other hand, EFS caused a
2275
2250 significant change in ethanol purity. As the
2225
2200
solvent is supplied to the top stages, it is
24 25 26 27 28 29 30 31 possible to confirm that the entrainer in the
NS
liquid phase is present in all the tower's trays
(b) below the EFS. Therefore, this condition
Fig. 3: Effect of NS and RR on: (a) Condenser results in less contact between vapor and
heat duty (QC) and (b) Reboiler heat duty liquid in a binary mixture, resulting in a less
(QR). effective separation process. The highest
ethanol purity was achieved at EFS 4.
These results also agree with a recent However, the ethanol purity decreased after
study indicating that a higher distillate will be BFS 29 because, as the BFS goes to the lower
achieved at a higher reflux ratio (Zhu et al., part of the column, the bottom part will
2016). The highest ethanol purity of 0.9956 contain more volatile compounds and
mole fraction can be achieved when the reflux therefore less ethanol purity than the top
ratio is 1.5. A higher reflux ratio means the part, the resulting increase in condenser and
energy required will increase because the reboiler duty, as depicted in Figure 5. In this
reboiler and condenser duty will also work, the BFS has the maximum number in
increase. Therefore, maintaining the reflux stage 29. BFS 22 was selected as the optimum
ratio as low as possible should be considered. number because the binary mixture can be
However, a low reflux ratio will only produce fed near the bottom column, which has
ethanol with a purity of less than 99%. Hence, longer contact with the entrainer, while
this work selected the reflux ratio of 1.5 as the maintaining the purity of the ethanol in the
optimum one. top product. Moreover, too low a feed stage
D, Hartanto, P.A. Handayani, W. Astuti, R.D. Kusumaningtyas, Y.C. Purwana, Maftukhaturrizqiyah, R.T. Wijayanti, 277
D.Z. Wulansari, R. Wulansarie, I.N. Pradnya, D.S. Hadikawuryan, A.A. Wibowo, R.M. Sholihin, A. Chafidz, I. Khoiroh
1780
1.00 EFS 2 EFS 4
EFS 3 1760
-Qc (kW)
0.99 EFS 4 1740
0.98 1720
1700
0.97
1680
0.96 1660
0.95 1640
0.94 1620
18 19 20 21 22 23 24 25 26 27 28 29 30 31
0.93
BFS
0.92
18 19 20 21 22 23 24 25 26 27 28 29 30 31 (a)
BFS
Fig. 4: Effect of BFS and EFS on Ethanol 2600
optimum BFS and EFS based on the ethanol (TAC) compared to the other stages, as seen
purity obtained and the energy requirements. in Figure 6. The TAC using NS 28 was 6.02 x
106 USD, lower than the TAC when glycerol is
used as an entrainer, as simulated by Novita
et al. (2018), with a TAC of 6.56 x 106 USD.
CONCLUSIONS