pubs.acs.

org/IECR Article

Production of n‑Butyl Acetate through Continuous Reactive

Distillation Using PTSA−POM as the Catalyst: Lab-Scale
Experiments, Modeling and Validation, Commercial-Scale Design,
and Economic Evaluation
P Nagveni, Reddi Kamesh, and K. Yamuna Rani*
Cite This: Ind. Eng. Chem. Res. 2023, 62, 15070−15083 Read Online
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
Downloaded via SULZER CHEMTECH on May 20, 2025 at 17:44:33 (UTC).

ABSTRACT: Experiments are conducted in a lab-scale continu-

ous reactive distillation column (CRDC) for the esterification of
acetic acid with n-butanol to produce n-butyl acetate using a
poly(o-methylene p-toluene sulfonic acid) (PTSA−POM) catalyst.
A steady-state Aspen model is developed and validated using
experimental data. Optimal design, using this model for data
generation, to maximize conversion and purity yielded 93.4%
conversion and 94.66% purity and 95.75% conversion and 95.01%
purity using response surface methodology and Aspen-based
optimization, respectively. The latter result is experimentally
validated to yield a conversion of 94.2%, with an n-butyl acetate
purity of 96.44%. Further, for the commercial-scale production of
n-butyl acetate, a CRDC is designed in an energy and cost-effective manner through steady-state simulations. The proposed CRDC
with PTSA−POM reduced the reboiler duty by 62.14 and 32.19% and the total annual cost by 30.51 and 36.85% compared to
CRDC using Amberlyst-15 and an ionic liquid as catalysts, respectively.

1. INTRODUCTION catalysts can be used. The conventional method of

n-Butyl acetate, also known as butyl ethanoate, is an ester esterification reaction involves using mineral acids such as
mainly used as an industrial solvent for producing lacquers and H2SO4, HCl, HF, and H3PO4. Usually, homogeneous catalysts
paints and as a flavoring agent in the food industry. Its other such as mineral acids have disadvantages, such as corrosivity,
primary use is in the production of adhesives and hardened toxicity, and difficulty in recoverability. In recent studies, ionic
coatings. It is also used as an extracting agent in liquids have been explored as homogeneous catalysts for this
pharmaceuticals. Commonly used methods for the production reaction due to their high thermal stability, high solubility,
of n-butyl acetate are the esterification of acetic acid with n- non- corrosivity, and nonvolatility. Various ionic liquids such
as pyridinium hydrogensulfate, 2-methylpyridine hydrogensul-
butanol1−4 and trans-esterification of methyl- or ethyl-acetate
fate, 1-methylimidazole hydrogensulfate, N-triethylammonium
with n-butanol.5−9 Researchers have also studied the
sulfate, N-tributyl ammoniumsulfate, butyltriethylammonium
coproduction of ethyl acetate and butyl acetate.10−12 Among
hydrogensulfate, and N-(4-sulfonic acid)butyl triethylammo-
these, the most practiced method is liquid-phase esterification
nium sulfate20,21 are used for this reaction. However, both
of acetic acid with n-butanol in the presence of an acid catalyst.
mineral acids and ionic liquids have difficulties in recover-
This reaction is reversible and exothermic. In various studies,
ability, and the latter are also expensive, making heterogeneous
experiments have been conducted in the absence of a catalyst
catalysts take over. Commonly used heterogeneous catalysts
where the conversion of acetic acid varied from 4.9 to 38.8% at
for liquid-phase esterification are ion exchange resins (such as
different molar ratios of acetic acid to n-butanol and
Amberlyst and Dowex 50), zeolites (such as Al-MCM-41, H-
temperatures ranging from 80 to 115 °C.13−16 It is an
equilibrium-limited reaction having an equilibrium conversion
of 66−69%.17−19 Received: June 13, 2023
Esterification is a relatively slow reaction at room temper- Revised: July 28, 2023
ature and does not proceed to completion. Different catalysts Accepted: August 28, 2023
have been used to enhance the reaction rate and conversion Published: September 11, 2023
because the reaction rate is low without a catalyst. For liquid-
phase esterification, either homogeneous or heterogeneous

© 2023 American Chemical Society https://doi.org/10.1021/acs.iecr.3c01933

15070 Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

M, and H-β), supported heteropoly acids (such as 15 as the catalyst. They concluded that the configuration with
tungstophosphoric acid and molybdophosphoric acid), solid a reactive zone in the middle of the column and rectifying and
acid catalysts (such as sulfated zirconia and WMNLR), and stripping sections as nonreactive zones gave the best results.
biocatalysts (such as Novozym-435 and lipase-CCAB). Singh et al.24 conducted a series of experiments in a lab-scale
Heterogeneous catalysts have various advantages as they are packed RD column. They used Hyflux packing in nonreactive
more efficient, eco-friendly, and easy to recover. Ionic liquids zones and Sulzer Katapak-S packing in the reactive zone. Their
are more advantageous than that of other homogeneous results show that the conversion of acetic acid was >99% for
catalysts (mineral acids and organic acids) due to their high most cases. Luyben and Wu25 performed simulations to study
thermal stability, high solubility, non-corrosivity, and non- different esterification reactions in an RD column. They
volatility. Various ionic liquids such as pyridinium hydro- studied the reaction of acetic acid with methanol, ethanol,
gensulfate, 2-methylpyridine hydrogensulfate, 1-methylimida- isopropanol, n-butanol, and amyl alcohol with Amberlyst-15 as
zole hydrogensulfate, N-triethylammonium sulfate, N-tribu- the catalyst. They developed the possible flowsheets for each
tylammonium sulfate, butyltriethylammonium hydrogensulfate, reaction using the phase equilibria data for these five
and N-(4-sulfonic acid)butyl triethylammonium sulfate20,21 are esterification reactions. They determined the optimum number
used for this reaction. of stages, pressure, number of reactive stages, and feed location
Typical production of n-butyl acetate involves a reactor, by minimizing the total annual cost (TAC). For the
followed by several distillation units to obtain a high-purity esterification of acetic acid with n-butanol, they found that
product. Luyben et al.6 compared the conventional and the optimum design for getting a high-purity product is a single
reactive distillation processes for synthesizing butyl acetate RD column, with 34 stages in which the rectifying section
through the transesterification of methyl acetate with butanol. consisted of 4 stages, the reactive zone consisted of 20 stages,
The conventional process included a CSTR followed by three and the stripping section consisted of 9 stages. Sert and
distillation columns, while the intensified process included a Atalay26 experimentally determined the optimum operating
reactive distillation column (RDC) followed by two distillation conditions for this reaction in an RD column. They conducted
columns. Recently, Shen et al.9 optimized the butyl acetate experiments in a packed RD column with Amberlyst-15 in the
production processes through simulations. They compared the reactive zone. The butyl acetate purity was 82% under their
transesterification and esterification reactions. They used a optimum conditions. Tian et al.27 used Amberlyst-36 wet for
reactor, followed by three distillation columns. They used 1,4 experiments in an RD column. The RD column was also
butanediol as an extracting solvent for production through attached to a fixed bed reactor and a preheater. The maximum
esterification. These production processes are tedious; there- conversion and purity reported by them were more than 96%.
fore, there is a need for a more energy-intensified approach to Peng and Wang28 performed a simulation study using an
produce n-butyl acetate to meet growing global demands. Over [Hpy][HSO4] ionic liquid as a catalyst and compared the
the last two decades, researchers have been exploring energy- results with a column having a heterogeneous catalyst. In
efficient operations to make n-butyl acetate, such as RD, recent studies, external solvents (entrainers) like cyclohexane,
because of its mass usage and industrial importance. First, n-hexane, and heptane are also used to increase water
Venimadhavan et al.4 conducted this reaction in a batch RD separability from the distillate.29,30 Al-Rabiah et al.31 proposed
column with sulfuric acid as a catalyst. They introduced a novel a novel membrane reactor (MR) with Amberlyst- 15 as the
distillate policy to enhance the conversion by removing a catalyst, followed by a distillation column. The butanol
byproduct (mainly water) from the column. In this reaction, conversion was 92%. However, they concluded that the
there is a low boiling heterogeneous azeotrope consisting of performance of RD is better than that of the proposed MR.
butanol, butyl acetate, and water, which, on condensation, In terms of economics, the RD is better than MR.
forms two heterogeneous layers (organic layer and aqueous Recently, Nagveni et al.19 reported the use of a polymeric
layer). Taking advantage of this, they removed the aqueous catalyst, i.e., poly(o-methylene p-toluene sulfonic acid)
layer as a distillate, refluxing the organic layer back into the abbreviated as PTSA−POM, for this system and showed that
column. Hanika et al.3 performed several pilot-scale experi- this catalyst resulted in reaction rates much faster than
ments in a continuous RD column (packed column) using an Amberlyst-15. For this reaction, Amberlyst-15 is the most
acidic ion exchanger. They compared the results with the widely used heterogeneous catalyst in RD, which is thermally
ASPEN plus simulation results. Two configurations were stable until 120 °C,32 while the polymeric catalyst proposed in
considered for experiments, i.e., only an RD column and a our recent study19 is stable until 220 °C. The high thermal
primary reactor followed by an RD column. They observed stability and high catalytic activity of the polymeric catalyst
that the reactor followed by an RD column was slightly better make it a promising heterogeneous catalyst. Therefore, this
in terms of butyl acetate purity than that with a single RD reaction is explored in the RD column by using this polymeric
column. In most of the studies in the open literature, an ion- catalyst, and its performance is determined.
exchange resin, i.e., Amberlyst-15 (A-15), is used as the catalyst In this study, experiments are conducted in a packed RDC
for the reaction in a packed column. to produce n-butyl acetate from the esterification of acetic acid
Steinigeweg and Gmehling22 performed a detailed study of with n-butanol. A polymeric catalyst, i.e., poly(o-methylene p-
this reaction in an RD column, where they studied toluene sulfonic acid) (PTSA−POM) prepared in the lab, is
thermodynamics, kinetics, experiments, and simulations. used for the reaction. The column is divided into three zones.
They have also studied the effect of pressure, reboiler duty, The catalyst is kept in the middle of the column, called the
feed location, number of reactive and nonreactive stages, and reactive zone. Above and below are rectifying and stripping
prereactor use. Their best configuration was with 28 stages, zones filled with nonreactive packing. First, fresh feed, i.e., an
and the conversion of butanol was 98.5%. Gangadwala et al.23 equimolar mixture of acetic acid and n-butanol, is introduced
performed simulations and studied different RD configurations as feed. The kinetics for this reaction was studied in our
and their effect on the conversion and product purity using A- previous study.19 A steady-state model of the continuous RD
15071 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 1. (a) Schematic diagram and (b) lab-scale experimental setup of a continuous RDC.

column is developed in a process simulator, i.e., Aspen plus of heterogeneous catalyst enveloped in a metallic wire mesh.
software and validated through experiments. The developed The column is insulated with heat tape to minimize heat loss
model is optimized by the Box−Behnken design (BBD) across the column. The lab-scale experimental setup is shown
method. Further, optimization is carried out in ASPEN plus in Figure 1b. In experiments, the low boiling feed is usually fed
using a sequential quadratic programming method, and the near the bottom of the column, and the high boiling feed is
optimization results are compared to BBD results and provided near the top of the column, but in this case, as the
experimentally validated. With the existing lab-scale RD boiling points of acetic acid (118 °C) and n-butanol (117.75
column, it is illustrated that conversion and a product purity °C) are too close. Therefore, a mixture containing equal moles
of around 95% could be achieved. of acetic acid and butanol is given as feed. RDC experiments
In the second part of this study, an attempt is made to are conducted at different feed flow rates and reflux ratios. The
design a commercial-scale energy-efficient and cost-effective reboiler is filled with a measured volume of reactants, and the
continuous RDC (CRDC) for this reaction. The TAC is column is operated at total reflux until the top temperature
determined for each configuration, and the objective is to reaches a steady state. Then, the feed pump is started and
minimize it by varying different parameters. The total number continuously fed on top of the reactive zone. Also, the reflux is
of theoretical stages, position of the reactive zone, and feed turned on. As the reaction and separation progress, the vapors,
location are determined to obtain a cost-effective config- upon condensation, get separated into two layers in the reflux
uration. After the design parameters are obtained, sensitivity drum. The top layer consists of butanol, butyl acetate, water,
analysis is carried out to determine the effects of the reflux and a small amount of acetic acid, and the second layer consists
ratio, catalyst mass, and feed stage on the column performance. mainly of water (>99%). The water layer is removed from the
The proposed best configuration is compared with columns system as distillate, and the organic layer is sent back as reflux.
with different catalysts reported in the literature regarding Once the column reaches a steady state, timely samples are
energy and total costs. withdrawn and analyzed using a Shimadzu 2014 gas chromato-
graph equipped with a free fatty acid phase (FFAP) column
2. MATERIALS AND METHODS (length: 30 m and ID: 0.32 mm) and a thermal conductivity
detector. Helium is a carrier gas at a linear velocity of 30 cm/s.
2.1. Materials and Catalyst. Acetic acid (purity >99.7%)
The temperature of both the injector and detector is
and n-butanol (purity >99%) are purchased from FINAR. The
maintained at 250 °C. The column oven temperature is kept
heterogeneous catalyst used is poly(o-methylene p-toluene
at 50 °C and then increased at the rate of 15 °C/min to attain
sulfonic acid) (also known as PTSA−POM). It is a product of
200 °C, and it is maintained there for 2 min.
the copolymerization of para-toluene sulfonic acid and para-
2.3. Continuous Reactive Distillation: Kinetic and
formaldehyde. The catalyst preparation method is reported by
Thermodynamic Models. A process simulator, i.e., Aspen
Liang and Yang.33
Plus software (Version-V11), is used for the simulations. The
2.2. Continuous Reactive Distillation: Experiment.
physicochemical properties of all components are used from
The esterification reaction of acetic acid with n-butanol to
the Aspen databank. The system consists of four components,
produce n-butyl acetate and water is studied. Experiments are
i.e., acetic acid, n-butanol, n-butyl acetate, and water. The
conducted in a lab-scale RDC. The schematic diagram of an
rigorous module, i.e., the RADFRAC module of Aspen plus
RDC is shown in Figure 1a. The total height of the column is 1
software, is used for the simulations. The reaction is
m without a reboiler, and the inner diameter is 0.05 m. The
represented as
bottom of the column has a reboiler with a capacity of 500 mL.
The nonreactive rectifying and stripping sections are packed Acetic acid + n Butanol F n Butyl acetate + Water
with glass beads (OD 0.01 m). The condenser is attached to
the column and connected with a liquid−liquid separator. The A concentration-based pseudohomogeneous kinetic model is
reboiler is connected to a heating/cooling circulator. The used for the rate calculations, and the reaction kinetics
height of the rectifying and stripping sections is 0.4 m each, reported for PTSA−POM19 are incorporated in the simulation.
whereas the height of the reactive zone is 0.2 m, filled with 20 g The rate equation is represented as
15072 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

ri = iMcat(k1CAcetic acidCn Butanol k 1Cn Butyl acetateC Water) sections in the column’s top and bottom portions, respectively.
(1) For all cases, the feed flow rate is 100 kmol/h. All these
configurations and operating conditions are shown in Figure 2.
where Mcat is the mass of the catalyst; ri is the rate of reaction
in terms of component i; ϑi is the stoichiometric coefficient; Ci
is the concentration of component i; and k1, k−1 are the
forward and backward rate constants, respectively. The
Arrhenius equation is used to express the temperature
dependency of the rate constants and is given below
k = k0 e E / RT (2)
where k0 is the pre-exponential factor, E is the activation
energy, T is the temperature, and R is the gas constant. E is
100,820 and 101,420 J/mol, respectively, for forward and
backward reactions and k0 is 5.29 × 109 and 1.71 × 109 (m3)2/
kmol kg s, respectively.19 The UNIQUAC model is used to
determine activity coefficients for the nonideal vapor−liquid
equilibrium (VLE) and vapor−liquid−liquid equilibrium
(VLLE).
The assumptions considered in the present study are listed
below, which are the general assumptions considered in various
studies such as Gangadwala et al.,23 Hanika et al.,3 and
Steinigeweg and Gmehling.22
• An equilibrium stage model is considered.
• The column consists of N number of theoretical stages
and is numbered from top to bottom. The total
condenser is considered stage 1, and the partial reboiler
is the Nth stage. Figure 2. Schematic diagram of an RD column with (a) all stages as
• A decanter is used where the condensate gets collected. reactive stages; (b) the top of the column as a reactive zone; and (c)
As n-butyl acetate forms a heterogeneous azeotrope with rectifying section on the top, reactive zone in the middle, and
n-butanol and water, it gets separated into two layers stripping section on the bottom.
upon condensation. The top layer is an organic layer
consisting of butanol and butyl acetate, refluxed back to
the column. In contrast, the bottom layer is an aqueous The rectifying and stripping sections are nonreactive. The RD
layer continuously removed from the column. column is operated at 2 bar pressure. Aspen Plus distillation
• VLLE is considered throughout the column, and analysis is used to predict azeotropes at 2 bar. These
UNIQUAC is selected as a thermodynamic model components form three minimum boiling heterogeneous
similar to other studies on this system reported by azeotropes and one minimum boiling homogeneous azeotrope.
Venimadhavan et al.,4 Steinigeweg and Gmehling,22 The list of pure components and azeotropes, along with the
Gangadwala et al.,23 and Luyben and Yu.25 molar compositions and their boiling temperature, is given in
• The reaction takes place in the liquid phase only. the Supporting Information as Table S1. First, simulations are
2.4. Development of a Steady-State Model for performed to determine the number of theoretical stages
Validation of Lab-Scale Experiments. A steady-state required to achieve 99.5% pure butyl acetate while minimizing
model is developed using the Aspen Plus software. An the TAC. Then the effect of different parameters is studied.
equilibrium-stage model is developed and validated through 2.6. Process Economics. The TAC is calculated to
experimental results. A heuristic method is used to determine evaluate the column configuration. The equations to determine
the number of theoretical stages. The column is divided into the TAC are reported in Supporting Information in Table
three zones, i.e., rectifying, reactive, and stripping. The feed S2.34 The TAC includes capital and operating costs. The
location is the first stage of the reactive zone. Feed is an objective is to minimize the TAC defined as
equimolar mixture of acetic acid and n-butanol. The feed flow Capital cost
rate is varied in the range 24−36 mL/h. It is assumed that the TAC = Operating cost +
Pay back period
reaction occurs only in the reactive zone. The amount of
catalyst is fixed at 20 g. The column is numbered from top to The capital cost includes the cost of columns, trays, and heat
bottom, i.e., condenser as the first stage and reboiler as the last exchangers (condenser and reboiler). The operating cost
stage. This quaternary system is nonideal and includes different includes the cost of utilities such as steam, water, and catalyst
homogeneous and heterogeneous azeotropes. The first cost. The payback period is considered as 3 years, and the
minimum boiling heterogeneous azeotrope consists of butanol, operational time is 8000 h.
butyl acetate, and water. The column diameter (Dc) is determined using the Aspen
2.5. Design of a Commercial-Scale Continuous Plus tray sizing utility. The column height (Hc) is calculated by
Reactive Distillation Column. Three different RD config- considering 0.6096 m tray spacing and 20% excess space.25
urations are considered and are as follows: (i) all stages as The weir height is considered as 152.4 mm. All of the
reactive stages; (ii) reactive zone at the top of the column; and equations used to determine the sizes and costs are given in
(iii) reactive zone in the middle, rectifying, and stripping Table S2. The Marshall & Swift index (M&S) used for TAC
15073 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

calculation was 1638.2 for the year 2018. The heat transfer Table 1 that the top temperature for all cases is more than 100.
coefficients and temperature difference for the reboiler and Therefore, in simulation, a higher pressure is considered in the
condenser are those reported by Luyben and Yu.25 Catalyst life column to get the top temperature similar to the experiment.
is considered to be 3 months. The amount of catalyst is fixed at For Runs 1, 4, and 5, the column pressure is 1.93 bar. For Runs
50 kg per tray in the reactive zone. 2 and 3, the column pressures are 1.82 and 1.67 bar,
respectively. The accuracy of the developed Aspen model is
3. RESULTS AND DISCUSSION evaluated by comparing the experimental and predicted data.
3.1. Lab-Scale Continuous Reactive Distillation The top and bottom temperatures, conversion, and purity are
Column. The RDC experiments are conducted at different compared and presented in Figure 3 for a 24 mL/h feed flow
flow rates and reflux ratios. The RDC operating conditions,
feed conditions, and results are listed in Table 1. The feed flow
rate varies from 24 to 36 mL/h, and the reflux ratio ranges
from 1.2 to 2.

Table 1. Operating Conditions and Results of RD

Experiments
run 1 2 3 4 5
feed flow rate, F 24 24 36 36 36
(mL/h)
feed composition
acetic acid 0.5 0.5 0.5 0.5 0.5 Figure 3. Comparison of experimental and simulation results: feed
n-butanol 0.5 0.5 0.5 0.5 0.5 flow rate:24 mL/h, reflux ratio:1.2 with mixed equimolar feed having
n-butyl acetate 0 0 0 0 0 acetic acid and n-butanol.
water 0 0 0 0 0
reflux ratio 1.2 1.667 1.2 1.667 2
distillate
rate at a 1.2 reflux ratio. The experimental and simulation
temperature, T1 110 106 104 110 110
(°C) results for each run are presented in Table 2, and the operating
flow rate, D (mL/h) 0.5 2 2.4 4 10 conditions of these runs are listed in Table 1. Top temperature
composition (T1), conversion of acetic acid (Xa), and purity of n-butyl
acetic acid 0.0073 0.0125 0.0067 0.0061 0.0739 acetate (P) are compared, and it can be seen that for each run,
n-butanol 0.0060 0.0032 0.0026 0.0008 0.41 the developed Aspen model gives reasonably good results. As
n-butyl acetate 0.0016 0 0.00158 0 0 can be seen from Table 2, there is a significant difference in the
water 0.9850 0.9844 0.9891 0.9931 0.5161 experiment and simulation values of the bottom temperature
bottom product (TN). In Aspen, there are no heat losses considered for the
temperature, TN 140 139 138 140 139 reboiler, whereas, in the experiment, there are chances of heat
(°C) loss.
flow rate, B (mL/h) 15.5 18.5 37 33 21 3.1.3. Optimization. The objective is to maximize the
composition conversion of acetic acid and the purity of butyl acetate by
acetic acid 0.0829 0.0789 0.1084 0.0966 0.1270 varying the bottom flow rate (B), catalyst mass (Cw), and
n-butanol 0.0917 0.0594 0.1156 0.0754 0.0959 reflux ratio (R). As the steady-state Aspen model is developed
n-butyl acetate 0.8254 0.8617 0.7760 0.8279 0.7771 and validated, response surface methodology (RSM)-based
water 0 0 0 0 0 optimization is compared with the Aspen-based optimization.
conversion, Xa (%) 89.07 81.22 84.07 87.97 87.25 Feed is an equimolar mixture of acetic acid and butanol. The
feed flow rate is 36 mL/h and is fed at the top of the reactive
3.1.1. Experimental Results. For Runs 1 to 5, fresh mixed zone.
feed containing equal moles of acetic acid and n-butanol is First, RSM optimization technique, i.e., a three-level, three-
continuously fed on the top of the reactive zone. Water is factor BBD method, is used for optimization. Minitab software
continuously removed as a distillate and butyl acetate as the is used for the BBD. The coded and uncoded independent
bottom product. Products are periodically analyzed, and the variables and the responses generated using Aspen model are
top and bottom product compositions are presented in Table given in Table 3. The bottom flow rate varies from 18 to 32 g/
1. For all cases, the conversion of acetic acid is more than 80%. h, the reflux ratio is 1.4 to 2.68, and the catalyst weight ranges
The maximum conversion achieved is 89.07% for Run 1. Also, from 4 to 12 g/tray. The conversion of acetic acid and purity of
the butyl acetate purity in this case is 82.54%. n-butyl acetate are the responses. The validated Aspen model is
3.1.2. Steady-State Model and Experimental Validation. considered and simulations are performed for the base case,
The total number of theoretical stages is 15, where the first and for each design, and the results are presented in Table 3.
stage is a condenser, and the 15th stage is a reboiler. Rectifying At lower R, purity decreases with an increase in B, whereas at
section is from 2 to 5, the reactive zone is 6 to 10, and the higher R, it shows the opposite trend. The BBD method takes
stripping section is from 11 to 14. The catalyst is only present the corners and center values to determine the optimal values.
in the reactive zone and is 4 g/stage. It is assumed that the The regression models are obtained from the software and are
reaction occurs only in the presence of a catalyst. At second-order polynomial equations. The models represent the
atmospheric pressure, the minimum boiling ternary heteroge- relationship between the responses and independent variables
neous azeotrope forms at 90.68 °C, whereas it can be seen in and are presented below in terms of coded factors:
15074 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Table 2. Comparison of Experimental and Simulation Results at Different Operating Conditions

experiment simulation
run T1 (°C) TN (°C) Xa (%) P (%) T1 (°C) TN (°C) Xa (%) P (%)
1 110 140 89.07 82.54 108.30 148.22 89.49 82.82
2 106 139 81.22 86.17 106.64 146.88 88.90 84.07
3 104 138 84.07 77.60 104.11 142.25 86.81 77.60
4 110 140 87.97 82.79 108.26 148.84 88.35 82.81
5 110 139 87.25 77.71 108.25 149.00 84.93 77.63

Table 3. BBD Coded and Uncoded Inputs with the values of Cw. A similar trend can be seen in Figures S1c and
Respective Responses S2c, where the combined effect of the reflux ratio and catalyst
weight on responses are demonstrated. At lower R, the purity
input parameters
and conversion are maximum; however, there is no significant
coded uncoded responses change with change in catalyst weight.
B Cw The response optimizer tool of Minitab software is used for
simulation X1 X2 X3 (g/h) R (g/tray) Xa (%) P (%) the optimization. The objective is to maximize both conversion
1 −1 −1 0 18 1.4 8 92.97 94.58 and purity simultaneously. The optimal design values through
2 1 −1 0 32 1.4 8 60.49 31.40 RSM are as follows: B = 18.707 g/h, R = 1.46, and Cw = 11.7
3 −1 1 0 18 2.68 8 37.22 0.48 g/tray. At optimal conditions, the conversion obtained through
4 1 1 0 32 2.68 8 57.47 28.00 Aspen model simulation is 93.4%, and the purity of butyl
5 −1 0 −1 18 2.04 4 81.74 70.37 acetate is 94.6%.
6 1 0 −1 32 2.04 4 58.97 29.68 Second, the optimization is carried out in Aspen plus
7 −1 0 1 18 2.04 12 81.73 70.35 software with the same ranges of independent variables. Aspen
8 1 0 1 32 2.04 12 58.97 29.68 uses the sequential quadratic programming (SQP) method for
9 0 −1 −1 25 1.4 4 74.71 57.61 optimization. For the SQP method, the solver parameters are
10 0 1 −1 25 2.68 4 39.67 12.18 the Kuhn−Tucker error and the Lagrangian function. The
11 0 −1 1 25 1.4 12 74.71 57.61 primary objective is to maximize conversion and purity. In
12 0 1 1 25 2.68 12 39.68 12.18 order to accommodate both of these objectives, the weighted
13 0 0 0 25 2.04 8 46.67 22.64 equations, which have both conversion (C) and purity (P), are
14 0 0 0 25 2.04 8 46.67 22.64
used as objectives. The objective functions, optimal values, and
15 0 0 0 25 2.04 8 46.67 22.64
results are given in Table 4. The maximum conversion is
95.74% for the 0.4C + 0.6P objective function. Also, the purity
P(%) = 658 28.28X1 167.5X 2 11.83X3 + 0.317X12 is more than 95%.
+ X 2 2 + 0.739X32 + 5.06X1X 2 (3)
Table 4. Objective Functions, Optimal Parameters, and the
Simulation Results for ASPEN Optimization
Xa(%) = 505 21.59X1 109.7X 2 9.42X3
objective function B (gm/h) R Cw (gm/tray) Xa (%) P (%)
+ 0.2912X1 + 2.7X 2 + 0.589X32 + 2.943X1X 2
2 2
C 22.28 1.447 4 95.67 94.78
(4) 0.2C + 0.8P 22.26 1.446 4 94.89 95.08
0.4C + 0.6P 22.265 1.446 4 95.75 95.01
where X1 is the bottom flow rate, X2 is the reflux ratio, and X3
0.6C + 0.4P 22.27 1.446 4 95.74 94.92
is the catalyst weight.
0.8C + 0.2P 21.745 2.490 4 36.37 7.51
The 3D surface and contour plots are generated to study the
P 18 1.821 4 94.25 94.70
effect of the interaction of independent parameters on
response variables. The response surface plots are generated
by keeping an independent parameter constant, i.e., at its zero The optimal parameters through Aspen optimization are B =
level, and the other two parameters are varied between their 22.265 g/h, R = 1.446, and Cw = 4 g/tray. For optimal design,
minimum and maximum values. The surface and contour plots the composition, temperature, and flow profiles are shown in
showing the combined effects of operating parameters (B, R Figure 4. It is known that the BBD method uses the corner and
and Cw) on purity (P) and conversion (Xa) are presented in center values for defining the model, whereas, in Aspen
Figures S1 and S2 in Supporting Information, respectively. optimization, the whole range is considered. It can also be seen
Figures S1a and S2a demonstrate the combined effect of the that by only making changes in operating parameters in the
bottom flow rate and reflux ratio on the conversion of acetic existing column, more than 95% purity can be achieved. An
acid and the purity of n-butyl acetate. It can be seen that the experiment is conducted to validate the optimization results at
conversion and purity decrease with increase in bottom flow the optimal parameters obtained from Aspen optimization.
rate and reflux ratio. Figures S1b and S2b show the combined The feed flow rate is 36 mL/h, and the reflux ratio is 1.446.
effect of the bottom flow rate and catalyst weight. It is The total mass of the catalyst in the reactive zone is 20 g.
observed that the purity and conversion are maximum at low Through the simulation, the conversion of acetic acid is
bottom flow rate, whereas the change in catalyst weight has an 95.75%, and the purity of n-butyl acetate is 95.01%, while the
insignificant effect on response variables. As Cw is increased conversion is 94.2% and the purity is 96.44% through the
from 4 to 12 g/tray, conversion is increased, but it is marginal. experiment. The experimental and simulation results are
The response variables are maximum near the lower and upper shown in Figure 5. It can be seen that the predictions match
15075 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 4. Simulation results of the optimal Aspen design: (a) liquid-phase composition profile, (b) temperature profile, and (c) liquid and vapor
flow profiles.

pure butyl acetate while minimizing the TAC. The method-

ology used for designing a CRDC is shown in the Supporting
Information in Figure S3. As shown in the figure, there are four
loops. The first one is to explore different numbers of
theoretical stages (N), and the second one is to consider three
different positions of reactive zones denoted by the index i,
where i = 1 denotes that all stages are reactive stages, i = 2
denotes that reactive zone is at the top of the column, and i = 3
denotes that the reactive zone is in the middle of the column
with rectifying and stripping sections in the column’s top and
bottom portions, respectively.
Also, the first stage in the reactive zone is denoted as RZ1,
and the last stage in the reactive zone is denoted as RZ2; in the
third option of the reactive zone position (i = 3), different
Figure 5. Comparison of experimental and simulation results for the cases with a varying number of stages in the reactive zone are
experiment conducted at optimal parameters. explored, and this is the third loop in the flowchart. In this
loop, the value of RZ1 is fixed, but the value of RZ2 is varied.
the experimental results, which also validate the developed The fourth loop determines the reflux ratio at which the
model. There is a significant difference in experiment and simulation converges, which is executed until convergence is
simulation values of the reboiler temperature; as also achieved by varying the reflux ratio values. For all cases, as per
mentioned in Section 3.1.2, it is due to the heat losses in the the flowchart, simulations have been performed. However, in
experiment. the following subsections, only a few representative cases are
3.2. Design of a Commercial-Scale Continuous discussed, and their results are presented for the sake of
Reactive Distillation Column. First, simulation is carried brevity.
out with the RD column configuration reported by Luyben and 3.2.1.1. Total Number of Theoretical Stages. The objective
Yu25 using the present PTSA−POM catalyst and its kinetics. is to obtain the configuration which has minimum TAC. The
The conversion of acetic acid was 98.3% with Amberlyst-15, total number of theoretical stages is varied in the range of 15−
whereas it is 99.33% with PTSA−POM, implying that the 35. The total annual cost is calculated for different
column performance can be further improved by making configurations and compared to select the cost-effective
specific design changes. Therefore, further simulations are column. For simulations, the mixed feed contains equal
carried out to determine the total number of theoretical stages moles of acetic acid and n-butanol and the flow rate is 100
and the effect of the position of the reactive zone, feed kmol/h. The bottom flow rate is fixed as 49.9 kmol/h for all of
location, and reflux ratio. the cases, and convergence is achieved by varying reflux ratios.
3.2.1. Design and Operating Parameters. The aim of this The feed is introduced in the first stage of the reactive zone.
study is to design an RD column capable of producing 99.5% The RD column design specifications, feed conditions, results,
15076 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Table 5. Design and Operating Parameters, Simulation Results, and the Economics for Different Total Theoretical Stages
N = 15 N = 20 N = 25 N = 30 N = 35
design and operating parameters
total number of stages, N 15 20 25 30 35
reactive zone 5−10 5−15 5−20 5−25 5−30
reboiler duty, Q (MW) 1.287 1.271 1.197 1.157 1.141
diameter of column, Dc (m) 0.976 0.93 0.925 0.92 0.92
simulation results
mole fraction of distillate
xacetic acid 0.00284 0.00140 0.00130 0.00129 0.00129
xn‑butanol 0.00739 0.00611 0.00604 0.00602 0.00601
xn‑butyl acetate 0.00071 0.00071 0.00071 0.00071 0.00071
xwater 0.98906 0.99178 0.99195 0.99198 0.99199
mole fraction of bottom product
xacetic acid 0.00613 0.00484 0.004773 0.00475 0.00474
xn‑butanol 0.00157 0.00012 0.000014 10−6 10−7
xn‑butyl acetate 0.99231 0.99504 0.995213 0.99525 0.99526
xwater 10−11 10−12 10−13 10−14 10−15
conversion of acetic acid (%) 99.104 99.376 99.393 99.397 99.398
economics
total capital cost (103 $) 424.922 451.268 470.278 493.666 512.652
total operating cost (103 $/year) 399.483 404.653 392.009 390.065 395.067
TAC (103 $/year) 541.124 555.076 548.769 554.62 565.951

and economics for different N values are reported in Table 5. 6. The maximum conversion is achieved in case 6, where the
Figure 6a shows that when increasing the total number of reactive zone and stripping section consist of 16 and 4 stages,
stages, conversion and purity increase. For N = 15, the purity respectively. The effect of the position of the reactive zone on
of butyl acetate is only 99.23%, whereas the desired purity is the TAC is shown in Figure 6d. The TAC dropped from case 1
99.5%. Therefore, it is not considered for further comparison. to case 2 and then increased sharply but again dropped until
For N in the range of 20−35, the purity increases from 99.504 case 5. The TAC is minimum for case 2, as can be seen in
to 99.526%, whereas the required energy in the reboiler Figure 6d, but it is not chosen to be the best design because
decreases from 1.287 to 1.141 MW. When N is increased from the conversion of acetic acid is only 96.93%, and the purity of
20 to 30, the TAC drops because the minimum reboiler duty butyl acetate is 97.02%, whereas in cases 3 to 6, the purity is
needed to achieve the desired purity product decreases, more than 99.5%. It is observed that the TAC is maximum for
causing the total energy required to reduce, resulting in a case 3, where the same number of stages is considered in both
decrement in the total cost of heat exchangers and utilities. rectifying and stripping sections. From cases 4 to 6, the
The TAC is minimum for N = 25, as shown in Figure 6b. number of stages in the reactive zone is increased, and it is
Therefore, N is chosen as 25 for further study to determine the observed that the conversion also increases marginally,
effect of the other parameters. although the TAC decreases due to less energy required, and
3.2.1.2. Position of the Reactive Zone. The position of the the column diameter also decreases, resulting in a decrement in
reactive zone plays a crucial role in an RD column. For each N, the capital cost. When case 5 is compared to case 6, the
simulations are performed by considering different positions of conversion and purity are slightly higher for case 6. However,
the reactive zone and studying their influence on the the TAC increases from case 5 to case 6, so the RD column
performance of the column. In the first case, all stages are specifications for case 5 are selected for further study.
considered reactive, and in case 2, the top half of the column is 3.2.1.3. Feed Location. As reported in the literature, the
the reactive zone. In case 3, the reactive zone is in the middle feed location influences the TAC drastically. Generally, the
of the column, with the top section as the rectifying zone and higher boiling feed is fed on top of the reactive zone, and the
the bottom section as the stripping zone. For cases 4−6, 3 lower boiling feed is fed near the bottom. In this reaction, the
stages are fixed in the rectifying section, and the number of boiling points of acetic acid and n-butanol are too close;
stages in the reactive zone is varied from 6 to 16. As mentioned therefore, a mixed feed is considered. The RD column
earlier, the TAC is minimum for N = 25; therefore, the configuration is N = 25, the flow rate of acetic acid and n-
operating conditions, results, and economics for six cases with butanol is 50 kmol/h each, and the reactive zone is from the
different reactive zone locations for N = 25 are reported in fifth to 15th stage. Three different feed locations are
Table 6. It can be seen in Figure 6c that the conversion and considered, i.e., mixed feed on the top of the reactive zone
purity increase from case 1 to case 6. When the number of (denoted as FL 1), acetic acid on the fifth stage and butanol on
stages in the reactive zone is increased, the conversion and the 10th stage (denoted as FL 2), and acetic acid on the fifth
purity also increase, while the energy requirement decreases. stage, i.e., the first stage of the reactive zone, and butanol on
The best results are obtained by keeping the reactive zone the 15th stage, i.e., last stage of the reactive zone (denoted as
between the rectifying and stripping sections. For cases 1 and FL 3). The column performance and economics are listed in
2, the required reflux rate is low and almost the same, resulting Table 7. Purity and conversion are similar for FL1 and FL2, as
in a similar conversion and purity. The required reflux rate is shown in Figure 6e, but the TAC is almost twice that of the
maximum for case 3, and again, it decreases from case 4 to case mixed feed configuration. It can be seen from Figure 6f that the
15077 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 6. Effect of total theoretical stages on (a) conversion of acetic acid and purity of n-butyl acetate and (b) TAC; effect of position of the
reactive zone on (c) conversion of acetic acid and purity of n-butyl acetate and (d) TAC; effect of feed location on (e) conversion of acetic acid and
purity of n-butyl acetate and (f) TAC.

TAC increases very much when acetic acid and butanol are fed the column, which also increases the liquid hold up in the
separately on different stages rather than as a mixed feed on the reboiler, due to which the reboiler duty increases as more
same stage. The required reflux rate and vapor boil-up rate to energy is required to vaporize the liquid resulting in a decrease
achieve high-purity products are much higher for the separate in the conversion.
feed cases than for the mixed feed case. Also, the energy 3.2.2.2. Effect of Catalyst Mass. To study the effect of the
requirement for separate feed cases and required column catalyst weight per stage on the column performance, it is
diameter are too high, resulting in high annual costs. varied from 2 to 50 kg. It is observed from Figure 7c that with
Therefore, the RD column with mixed feed is chosen as the an increase in catalyst weight, the purity increases, but the
best case. desired purity is achieved after 35 kg. Also, with an increase in
3.2.2. Sensitivity Analysis. Sensitivity analysis is performed catalyst mass, the TAC decreases slightly and then increases
to study the effect of various parameters on the column linearly, as shown in Figure 7d.
performance of the proposed configuration. The feed flow rate 3.2.2.3. Effect of Feed Stage. As discussed earlier, the feed
is 100 kmol/h, and the total stages are 25, the reactive zone is stage is an important factor. The mixed feed with equal moles
from the fifth to the 15th stage. The parameters studied are the of acetic acid and n-butanol is fed in the column. In the
reflux ratio, catalyst per tray, and feed stage. proposed column, a mixed feed is fed on the first stage of the
3.2.2.1. Reflux Ratio. To study the effect of the reflux ratio reactive zone. To study the effect of the feed stage, the ratio is
on the column performance, it is varied from 0.65 to 10. It is varied from 2 to 6. When the feed stage is changed from 2 to 5,
observed from Figure 7a that with an increase in reflux ratio, there is a vast increment in conversion, whereas from stage 5 to
the conversion of acetic acid and purity increase and then 6, conversion decreases slightly, as shown in Figure 7e. When
decreases. These values reach a maximum at a reflux ratio of the feed is introduced in stage 2, there is no rectifying section
1.9 and then drop afterward. With the increase in the reflux for separation; therefore, the conversion and purity are around
ratio, the TAC increases, as shown in Figure 7b. Increasing the 93%. When the feed stage is changed from 2 to 3, as a single
reflux ratio increases the amount of liquid entering back into stage is available for separation, there is around a 5% increment
15078 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Table 6. Design and Operating Parameters, Simulation Results, and Economics for Different Positions of the Reactive Zone
case 1 case 2 case 3 case 4 case 5 case 6
design and operating parameters
total number of stages, N 25 25 25 25 25 25
reactive zone 2−24 2−13 10−17 5−10 5−15 5−20
reboiler duty, Q (MW) 1.098 1.099 3.613 1.483 1.175 1.197
diameter of column, Dc (m) 0.92 0.92 1.68 1.057 0.92 0.925
simulation results
mole fraction of distillate
xacetic acid 0.0230 0.0230 10−8 0.0014 0.0013 0.0013
xn‑butanol 0.0085 0.0085 0.008 0.0062 0.0061 0.0060
xn‑butyl acetate 0.0011 0.0011 0.0024 0.0007 0.0007 0.0007
xwater 0.9674 0.9674 0.9895 0.9917 0.9919 0.9919
mole fraction of bottom product
xacetic acid 0.0076 0.0076 0.0085 0.0049 0.0048 0.0048
xn‑butanol 0.0222 0.0222 0.0004 0.00007 0.00005 0.00001
xn‑butyl acetate 0.9702 0.9702 0.9911 0.995 0.9952 0.99521
xwater 10−5 10−18 10−16 10−23 10−19 10−13
conversion of acetic acid (%) 96.93 96.933 99.153 99.37 99.38 99.39
economics
total capital cost (103 $) 445.14 446.30 966.93 546.09 465.45 470.28
total operating cost (103 $/year) 376.07 354.44 1110.22 459.15 375.44 392.01
TAC (103 $/year) 524.45 503.21 1432.52 641.18 530.58 548.77

Table 7. Design and Operating Parameters, Simulation 3.2.3. Optimization. Optimal column configuration is
Results, and Economics for Different Feed Locations determined using the Optimization tool Aspen plus. The
objective is to minimize the TAC with the purity of butyl
FL 1 FL 2 FL 3
acetate equal to or greater than 99.5% as a constraint. The
design and operating parameters reflux ratio and catalyst weight are varied to obtain the
total number of stages, N 25 25 25 objective. The reflux ratio varies in the range of 0.66 to 3, and
acetic acid feed stage, FSacetic acid 5 5 5 the catalyst weight per tray varies from 30 to 50 kg. The
n-butanol feed stage, FSbutanol 5 10 15
optimal configuration has a 0.6763 reflux ratio, and the catalyst
reactive zone 5−15 5−15 5−15
mass is 39.2419 kg. Its TAC is 5,23,378.7 $/year. The
reboiler duty, Q (MW) 1.175 2.225 3.43
corresponding reboiler and condenser duties are 1.1681 and
diameter of column, Dc (m) 0.92 1.3 1.6
−0.916 MW, respectively. Sensitivity analysis results depicted
simulation results
in Figure 7 also lead to similar inferences.
mole fractions of top product
3.2.4. Proposed RD Column Configuration. After evaluat-
xacetic acid 0.0013 0.0052 0.0078
ing the various parameters and minimizing the TAC, the
xn‑butanol 0.0061 0.0021 0.0012
optimal configuration is obtained, which has 25 theoretical
xn‑butyl acetate 0.0007 0.0008 0.0008
xwater 0.9919 0.9919 0.9901
stages with a reactive zone from the fifth to 15th stage and
mole fractions of bottom product
mixed feed on the fifth stage. The rectifying section consists of
xacetic acid 0.0048 0.0009 0.000024
3 stages, and the stripping section consists of 9 stages. The
xn‑butanol 0.00005 0.0041 0.0067
mass of the catalyst on each tray is within the permissible limits
xn‑butyl acetate 0.9952 0.9951 0.9933 according to the column’s hydraulics. The liquid-phase
xwater 10−19 10−17 10−18 composition profiles of all components in the column are
conversion of acetic acid (%) 99.388 99.389 99.213 shown in Figure 8, and it can be seen that the butyl acetate
economics composition increases toward the bottom of the column and
total capital cost (103 $) 465.45 712.025 931.53 reaches almost 0.9952. The temperature and liquid and vapor
total operating cost (103 $/year) 375.44 695.457 1063.05 flow profiles in the column are shown in Figure 9a,b,
TAC (103 $/year) 530.58 932.798 1373.56 respectively.
3.2.5. Comparison with the Literature. The proposed RD
column is compared with configurations reported in the
in the conversion, whereas when it is changed from 3 to 4, the literature. It is compared with the columns where Amberlyst-
increment in conversion is comparatively small. Figure 7e 1525 and ionic liquid28 are used as catalysts in the reactive
shows that a minimum of three stages are needed in the zone. The feed rate is the same in all the cases, i.e., 50 kmol/h
rectifying section to get more than 99.5% conversion and each of acetic acid and butanol. It can be seen from Table 8
purity. TAC is also calculated for each feed stage and is plotted that in the proposed configuration, the conversion of acetic
in Figure 7f. It can be seen that the TAC decreases as the feed acid and the mole fraction of butyl acetate in the bottom
stage shifts toward the bottom of the column. It is a minimum product are higher than those of reported configurations,
for the fifth stage and then increased. Hence, the feed on the whereas the reboiler duty for the proposed configuration is
first stage of the reactive zone is the best possible much less. The reboiler duty, condenser duty, and TAC for
configuration. different configurations are shown in Figure 10. The sign of Qc
15079 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 7. Effect of reflux ratio on (a) conversion and purity and (b) TAC for the proposed column; effect of catalyst mass on (c) conversion and
purity and (d) TAC for the proposed column; effect of feed stage on (e) conversion and purity and (f) TAC for the proposed column.

Figure 8. Liquid-phase composition profile for the proposed RD configuration with N = 25 (inset as a magnified figure).

is negative, but only the magnitude is reported in Figure 10a. than the values considered in the other two cases reported, as
The TAC reported in Table 8 for Amberlyst-15, and ionic shown in Table 8. Despite this difference, the process
liquid, is directly taken from the literature. These simulations economics of the proposed configuration is much better.
indicate that with the proposed configuration, the reboiler duty The TAC of the proposed column is relatively low. The TAC
is reduced by 62.14 and 32.19%, while the condenser duty is of the proposed column is reduced by 30.52 and 36.83% with
reduced by 67.96 and 38.72% with respect to the columns with respect to the columns with Amberlyst-15 and ionic liquid.
Amberlyst-15 and ionic liquid, respectively. It may be noted Therefore, the proposed column performs better due to low
that the cost of steam considered in this study is much higher energy consumption and cost.
15080 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 9. (a) Temperature profile and (b) liquids and vapor flow profile for the proposed RD configuration with N = 25.

Table 8. Configuration, Results, and Economics of the A three-level three factorial BBD-based optimization and
Proposed Column and Reported Literature Aspen optimization are performed and compared. For
optimization, the independent variables are bottom flow rate,
Amberlyst- ionic
1525 liquid28 PTSA−POM catalyst mass, and reflux ratio, whereas conversion and purity
total number of stages 34 28 25
are the responses. The BBD method has resulted in 93.4%
acetic acid feed stage, 5 4 5
conversion, and Aspen optimization yielded 95.75% con-
FSacetic acid version at the respective optimal conditions. The optimal
n-butanol feed stage, FSbutanol 9 4 5 parameters from the BBD design are B-18.707 g/h, R-1.46, and
reactive zone 5−24 2−27 5−15 Cw-11.7 g/tray, while from the SQP method in Aspen, it is B-
product flow rate (kmol/h) 22.265 g/h, R-1.445, and Cw-4 g/tray. Also, an experiment is
top 50.38 49.87 50.1 performed with the obtained optimal parameters from Aspen
bottom 49.62 49.93 49.9 optimization. At optimal conditions, through the simulation,
mole fraction of top products the conversion is 95.75% and the purity is 95.01%, whereas
xacetic acid 0.0167 0.00096 0.00135 through the experiment, the conversion is 94.2% and the purity
xn‑butanol 0.00688 0.00725 0.00611 is 96.44%. Thus, it is concluded that the developed Aspen
xn‑butyl acetate 0.00076 0.00147 0.00071 model can be predicted with reasonable accuracy. Aspen
xwater 0.97566 0.99032 0.99183 optimization provides a more accurate solution compared to
xionicliquid - <10−9 - the conventional optimization technique. Results also show
mole fraction of bottom products that the conversion and purity can be increased significantly by
xacetic acid 0.00004 0.00654 0.00485 changing the operating parameters for the existing column
xn‑butanol 0.01006 0.00109 0.00006 design.
xn‑butyl acetate 0.9899 0.99236 0.99509 Later, a CRDC is designed for the production of n-butyl
xwater <10−8 0.00001 10−19 acetate using a polymeric catalyst. Steady-state simulations are
xionicliquid - <10−9 - carried out to determine the minimum number of theoretical
condenser duty, QC (MW) −2.8579 −1.4942 −0.9156 stages required to achieve high-purity products having the least
reboiler duty, QR (MW) 3.0854 1.7226 1.1681 cost. Also, the operating and design parameters are determined
diameter of column, Dc (m) 1.88 1.23 0.92
at minimum TAC. For the best possible configuration, the
conversion of acetic acid (%) 98.313 99.251 99.381
number of theoretical stages, position of the reactive zone, and
total capital cost (103 $) 1277.26 622.9 464.24
feed location are varied to obtain the maximum conversion and
total operating cost 327.53a 562.3b 368.63c
(103 $/year) purity with a minimum energy requirement. Results revealed
TAC (103 $/year) 753.28 828.6 523.38 that when the middle zone is used as a reactive zone with the
a
Cost of steam is 4.7 $/GJ. Cost of steam is 7.78 $/GJ. cCost of
b top (rectifying) and bottom (stripping) of the column being
steam is 9.88 $/GJ (present study). nonreactive zones, it results in an energy and cost-efficient
process. As the total number of theoretical stages increases, the
conversion of acetic acid and purity of butyl acetate increases.
4. CONCLUSIONS When N is increased from 15 to 25, the conversion and purity
Experiments are conducted in a lab-scale RDC for the increase; however, when it is changed from 25 to 35, the
production of n-butyl acetate by the esterification of acetic increment in conversion and purity is marginal. It is also
acid with n-butanol using poly(o-methylene p-toluene sulfonic concluded that the position of the reactive zone and feed
acid) as the catalyst. Experiments are performed at different location are critical parameters influencing the total costs and
flow rates and reflux ratios. An equimolar mixture of acetic acid column performance. The proposed best column configuration
and butanol is taken as the feed. The conversion is more than is 25 stages, with the 5th to 15th stages as the reactive zone
80% for all runs, and it is highest (89.07%) for the 24 mL/h and mixed feed at the top of the reactive zone.
feed flow rate at a 1.2 reflux ratio. Further, the effect of the reflux ratio, catalyst mass, feed
A steady-state Aspen model is also developed, and the temperature, and feed stage is also studied. Also, the column is
simulation results are compared to the experimental results. optimized for determining the optimal operating parameters.
Results show that the developed model has good accuracy. The proposed cost-effective RD column is compared to the
Further, the validated model is used to carry out optimization. configurations reported in the literature. The conversion of
15081 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

Figure 10. Comparison of the proposed configuration with configurations reported by Luyben and Yu25 (Amberlyst-15) and Peng and Wang28
(ionic liquid) in terms of (a) reboiler and condenser duties and (b) TAC.

acetic acid was 98.31 and 99.25%, respectively, for the column Notes
with Amberlyst-15 and ionic liquid as the catalyst, while the The authors declare no competing financial interest.
conversion of acetic acid is 99.381% for the proposed
configuration. The reboiler duty for the proposed design is
reduced by 62.14 and 32.19%, while the condenser duty is
reduced by 67.96 and 38.72% when compared to the columns
■ ACKNOWLEDGMENTS
The first author acknowledges the Council of Scientific and
with Amberlyst-1525 and ionic liquid28 as a catalyst, while the Industrial Research (CSIR), New Delhi, India, for financial
TAC is reduced by 30.52 and 36.83%, respectively. Hence, it is support in the form of a fellowship. We thank Director CSIR-
concluded that the performance of the proposed configuration IICT for the support (#IICT/Pubs./2023/107).
is better in terms of energy and cost.

■ ASSOCIATED CONTENT
■ REFERENCES
(1) Leyes, C. E.; Othmer, D. F.; Leyes, C. E.; Othmer, D. F.
*
sı Supporting Information Esterification of Butanol and Acetic Acid. Ind. Eng. Chem. 1945, 37
The Supporting Information is available free of charge at (10), 968−977.
https://pubs.acs.org/doi/10.1021/acs.iecr.3c01933. (2) Zhicai, Y.; Xianbao, C.; Jing, G. Esterification-Distillation of
Butanol and Acetic Acid. Chem. Eng. Sci. 1998, 53, 2081−2088.
Azeotropic information on the 4-component system in (3) Hanika, J. H. J.; Kolena, J. J. H.; Smejkal, Q. Butylacetate via
this study; equations used for design calculations and Reactive Distillation - Modelling and Experiment. Chem. Eng. Sci.
economics; and surface and contour plots showing 1999, 54 (21), 5205−5209.
interaction effects of parameters on acetic acid (4) Venimadhavan, G.; Malone, M. F.; Doherty, M. F. A Novel
conversion and butyl acetate purity, along with the Distillate Policy for Batch Reactive Distillation with Application to the
flowchart used for the design approach (PDF) Production of Butyl Acetate †. Ind. Eng. Chem. Res. 1999, 38, 714−
722.

■ AUTHOR INFORMATION
Corresponding Author
(5) Jimenez, L.; Garvin, A.; Costa-Lopez, J. The Production of Butyl
Acetate and Methanol via Reactive and Extractive Distillation. II.
Process Modeling, Dynamic Simulation and Control Strategy. Ind.
K. Yamuna Rani − Process Dynamics and Control group, Eng. Chem. Res. 2002, 41, 6735−6744.
Chemical Engineering & Process Technology Department, (6) Luyben, W. L.; Pszalgowski, K. M.; Schaefer, M. R.; Siddons, C.
CSIR-Indian Institute of Chemical Technology, Hyderabad Design and Control of Conventional and Reactive Distillation
500007 Telangana, India; Academy of Scientific and Processes for the Production of Butyl Acetate. Society 2004, 43,
Innovative Research (AcSIR), Ghaziabad 201002, India; 8014−8025.
orcid.org/0000-0003-4809-3106; Phone: +91-40- (7) Harvianto, G. R.; Ahmad, F.; Lee, M. A Thermally Coupled
Reactive Distillation and Pervaporation Hybrid Process for N-Butyl
27193121; Email: [email protected]; Fax: +91-40- Acetate Production with Enhanced Energy Efficiency. Chem. Eng. Res.
27193626 Des. 2017, 124, 98−113.
Authors (8) Wang, H.; Liu, W.; Gao, L.; Lu, Y.; Chen, E.; Xu, Y.; Liu, H.
Synthesis of N-Butyl Acetate via Reactive Distillation Column Using
P Nagveni − Process Dynamics and Control group, Chemical Candida Antarctica Lipase as Catalyst. Bioprocess Biosyst Eng. 2020, 43
Engineering & Process Technology Department, CSIR-Indian (4), 593−604.
Institute of Chemical Technology, Hyderabad 500007 (9) Shen, Y.; Zhao, F.; Qiu, X.; Zhang, H.; Yao, D.; Wang, S.; Zhu,
Telangana, India; Academy of Scientific and Innovative Z.; Yang, J.; Cui, P.; Wang, Y.; Gao, J. Economic, Thermodynamic,
Research (AcSIR), Ghaziabad 201002, India and Environmental Analysis and Comparison of the Synthesis Process
Reddi Kamesh − Process Dynamics and Control group, of Butyl Acetate. Ind. Eng. Chem. Res. 2020, 59 (50), 21869−21881.
Chemical Engineering & Process Technology Department, (10) Tian, H.; Zhao, S. Y.; Zheng, H. D.; Huang, Z. X. Optimization
CSIR-Indian Institute of Chemical Technology, Hyderabad of Coproduction of Ethyl Acetate and N-Butyl Acetate by Reactive
500007 Telangana, India; Academy of Scientific and Distillation. Chin J. Chem. Eng. 2015, 23 (4), 667−674.
Innovative Research (AcSIR), Ghaziabad 201002, India; (11) Bories, C.; Guzman Barrera, N. I.; Peydecastaing, J.; Etxeberria,
orcid.org/0000-0002-1791-4656 I.; Vedrenne, E.; Garcia, C. V.; Thiebaud-Roux, S.; Sablayrolles, C.
LCA Case Study: Comparison between Independent and Cop-
Complete contact information is available at: roduction Pathways for the Production of Ethyl and n-Butyl Acetates.
https://pubs.acs.org/10.1021/acs.iecr.3c01933 Int. J. Life Cycle Assess 2017, 2019 (Ceresana 2014), 1−16.

15082 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083
Industrial & Engineering Chemistry Research pubs.acs.org/IECR Article

(12) Xie, J. W.; Peng, F.; Li, C. L.; Jiang, T.; Ma, S. M. Coproduction Exchange Resins as Catalysts. Ind. Eng. Chem. Res. 2003, 42 (10),
of Ethyl Acetate and N-Butyl Acetate by Using a Reactive Dividing- 2146−2155.
Wall Column. Chem. Eng. Technol. 2018, 41 (9), 1808−1817. (33) Liang, X.; Yang, Æ. J. Synthesis of a Novel Carbon Based
(13) Tankov, I.; Mitkova, M.; Nikolova, R.; Veli, A.; Stratiev, D. N- Strong Acid Catalyst Through Hydrothermal Carbonization. 2009,
Butyl Acetate Synthesis in the Presence of Pyridinium-Based Acidic 460−463. DOI: DOI: 10.1007/s10562-009-0109-6.
Ionic Liquids: Influence of the Anion Nature. Catal. Lett. 2017, 147 (34) Bîldea, C. S.; Győ rgy, R.; Brunchi, C. C.; Kiss, A. A. Optimal
(9), 2279−2289. Design of Intensified Processes for DME Synthesis. Comput. Chem.
(14) Mitran, G.; Makó, É .; Rédey, Á .; Marcu, I. C.; Mitran, G.; Eng. 2017, 105, 142−151.
Makó, É .; Rédey, Á .; Marcu, I. C. Esterification of Acetic Acid with N-
Butanol Using Vanadium Oxides Supported on ??-Alumina. C. R.
Chim. 2012, 15 (9), 793−798.
(15) Wang, J.; Pan, H.; Wang, A.; Tian, X.; Wu, X.; Wang, Y.
Synthesis and Characterization of S2O82−/ZnFexAl2−xO4 Solid
Acid Catalysts for the Esterification of Acetic Acid with n-Butanol.
Catal. Commun. 2015, 62, 29−33.
(16) Tankov, I.; Yankova, R. DFT Analysis, Reaction Kinetics and
Mechanism of Esterification Using Pyridinium Nitrate as a Green
Catalyst. J. Mol. Liq. 2019, 277, 241−253.
(17) Faith, W. L.; Keyes, D. B.; Clark, R. L.; Lowenheim, F. A.;
Moran, M. K. Industrial Chemicals; Wiley: New York, 1965.
(18) Sepulveda, J. H.; Yori, J. C.; Vera, C. R. Repeated Use of
Supported H3PW12O40 Catalysts in the Liquid Phase Esterification
of Acetic Acid with Butanol. Appl. Catal., A 2005, 288 (1−2), 18−24.
(19) Nagveni, P.; Kamesh, R.; Yamuna Rani, K. Kinetic Modeling of
Liquid-Phase Esterification of Acetic Acid with n-Butanol Using
Heterogeneous Poly(o-Methylene p-Toluene Sulfonic Acid) as
Catalyst. Int. J. Chem. Kinet. 2020, 52, 822−837.
(20) Tao, D. J.; Lu, X. M.; Lu, J. F.; Huang, K.; Zhou, Z.; Wu, Y. T.
Noncorrosive Ionic Liquids Composed of [HSO4] as Esterification
Catalysts. Chem. Eng. J. 2011, 171, 1333.
(21) Tao, D. J.; Wu, Y. T.; Zhou, Z.; Geng, J.; Hu, X. B.; Zhang, Z.
B. 20100_IECR_Kinetics for the Esterification Reaction of N-Butanol
with Acetic Acid Catalyzed by Noncorrosive Brønsted Acidic Ionic
Liquids. Ind. Eng. Chem. Res. 2011, 50, 1989−1996.
(22) Steinigeweg, S.; Gmehling, J. N -Butyl Acetate Synthesis via
Reactive Distillation: Thermodynamic Aspects, Reaction Kinetics,
Pilot-Plant Experiments, and Simulation Studies. Ind. Eng. Chem. Res.
2002, 41 (22), 5483−5490.
(23) Gangadwala, J.; Kienle, A.; Stein, E.; Mahajani, S. Production of
Butyl Acetate by Catalytic Distillation: Process Design Studies. Ind.
Eng. Chem. Res. 2004, 43 (1), 136−143.
(24) Singh, A.; Hiwale, R.; Mahajani, S. M.; Gudi, R. D.;
Gangadwala, J.; Kienle, A. Production of Butyl Acetate by Catalytic
Distillation. Theoretical and Experimental Studies. Ind. Eng. Chem.
Res. 2005, 44 (9), 3042−3052.
(25) Luyben, W. L.; Yu, C.-C. Reactive Distillation Design and
Control; John Wiley & Sons, 2009.
(26) Sert, E.; Atalay, F. S. Esterification of Acetic Acid with Butanol:
Operation in a Packed Bed Reactive Distillation Column. Chem.
Biochem. Eng. Q 2011, 25 (2), 221−227.
(27) Tian, H.; Huang, Z.; Qiu, T.; Wang, X.; Wu, Y. Reactive
Distillation for Producing N-Butyl Acetate: Experiment and
Simulation. Chin. J. Chem. Eng. 2012, 20 (5), 980−987.
(28) Peng, X.; Wang, L. K. Design and Control of Ionic Liquid-
Catalyzed Reactive Distillation for n-Butyl Acetate Production. Chem.
Eng. Technol. 2015, 38 (2), 223−234.
(29) Chen, Y. Le; Hu, X. B.; Wu, Y. T. Binary Bronsted Acidic Ionic
Liquids (BBAILs) as the Reactive Extraction Intensified Catalysts for
the Esterification of Acetic Acid and n-Butanol. J. Chem. Eng. Jpn.
2017, 50 (8), 632−640.
(30) Cho, M.; Han, M. Dynamics and Control of Entrainer
Enhanced Reactive Distillation Using an Extraneous Entrainer for the
Production of Butyl Acetate. J. Process Control 2018, 61, 58−76.
(31) Al-Rabiah, A. A.; Alqahtani, A. E.; Al Darwish, R. K.; Bin
Naqyah, A. S. Novel Process for Butyl Acetate Production via
Membrane Reactor: A Comparative Study with the Conventional and
Reactive Distillation Processes. Processes 2022, 10 (9), 1801.
(32) Gangadwala, J.; Mankar, S.; Mahajani, S.; Kienle, Achim; Stein,
E. Esterification of Acetic Acid with Butanol in the Presence of Ion-

15083 https://doi.org/10.1021/acs.iecr.3c01933
Ind. Eng. Chem. Res. 2023, 62, 15070−15083