Performance Evaluation of Pervaporation ... - ACS Publications

Subscriber access provided by EAST TENNESSEE STATE UNIV

Article

Performance Evaluation of Pervaporation Technology for Process Intensification of Butyl Acrylate Synthesis. Dânia S. M. Constantino, Rui P. V. Faria, Ana Mafalda Ribeiro, José Miguel Loureiro, and Alirio Egidio Rodrigues Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b01328 • Publication Date (Web): 27 Jun 2017 Downloaded from http://pubs.acs.org on June 28, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

1

Performance Evaluation of Pervaporation Technology for Process

2

Intensification of Butyl Acrylate Synthesis

3 4

Dânia S.M. Constantino, Rui P. V. Faria, Ana M. Ribeiro, José M. Loureiro and Alírio. E.

5

Rodrigues*

6

Laboratory of Separation and Reaction Engineering – Laboratory of Catalysis and Materials

7

(LSRE-LCM), Faculdade de Engenharia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-

8

465 Porto, Portugal

9

*Corresponding author. Tel.: +351225081671; fax: +351225081674.

10

E-mail address: [email protected] (A.E. Rodrigues).

11 12

Keywords

13

Butyl acrylate, multicomponent pervaporation data, process intensification, fixed-bed

14

membrane reactor, continuous process simulation

15 16

Abstract

17

Pervaporation-based hybrid processes have been investigated to overcome the drawbacks of

18

equilibrium-limited reactions. Pervaporation processes are strongly recommended for heat-

19

sensitive products and azeotropic mixtures like is the butyl acrylate system case, since it can

20

operate at lower temperatures than distillation. In this work, experimental pervaporation data for

21

multicomponent mixtures in absence of reaction were measured for the compounds involved in

22

the esterification reaction of acrylic acid with n-butanol at different temperatures: 323 K, 353 K

23

and 363 K. A commercial tubular microporous silica membrane from Pervatech was used which

24

is highly selective to water and its performance was evaluated by studying several parameters

25

like the selectivity, permeate fluxes, driving force of species and separation factor. The effect of 1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

temperature and feed composition were assessed for binary, ternary and quaternary mixtures.

2

Increasing the temperature increases significantly the total permeate flux as well as the

3

separation factor, which is higher for quaternary mixtures. The presence of butyl acrylate and

4

acrylic acid reduces the total permeate flux since these molecules hinder the water permeation.

5

The permeance of each species was correlated with temperature according to Arrhenius

6

equation and a mathematical model was proposed to develop an integrated reaction-separation

7

process using the experimental data obtained. The reaction conversion of the fixed-bed

8

membrane reactor at steady-state achieved 98.7 % at isothermal conditions increasing by 66 %

9

the conversion obtained in a fixed-bed reactor (at the same operating conditions).

Page 2 of 36

10 11

1. Introduction

12

Butyl acrylate (BAc) is usually obtained from the esterification between n-butanol and acrylic

13

acid (AAc) having water as side product. Several process intensification based technologies for

14

BAc synthesis have been reported, among which: reactive distillation (RD)1, 2, reactor-

15

separation-recycle (RSR)3, conventional simulated-moving-bed-reactor (SMBR)4 and an

16

enhanced SMBR (FBRplusSMBR)5. Different strategies to overcome the equilibrium

17

limitations and other drawbacks associated to the BAc synthesis process, as the presence of

18

azeotropes and polymerization at high temperatures, have been intensively investigated in the

19

last years.

20

In SMBR based processes4, 5, pervaporation units were proposed for treating the extract stream

21

(composed by n-butanol and water) by dehydration of n-butanol, which was recycled to the

22

process as reactant and/or eluent, depending of its purity. In those works, pervaporation data

23

reported from Sommer’s and Melin6 were used, for a commercial tubular amorphous silica

24

membrane distributed by ECN, Petten (The Netherlands) and by Sulzer Chemtech GmbH

25

Membrane Technology, Neunkirchen (Germany), as Pervap1 SMS. Indeed, membrane

26

pervaporation process is a very interesting technology for organic-water6-13 and organic-organic

27

separations14-16, being more advantageous than distillation which presents considerable 2 ACS Paragon Plus Environment

Page 3 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

thermodynamic limitations17. Moreover, membrane separation processes offer many other

2

advantages such as: high selectivity, low energy consumption, moderate cost to performance

3

ratio, compact and modular design18. The separation by pervaporation involves a transport

4

mechanism based on physical-chemical interactions between the membrane material and the

5

permeating molecules allowing to operate at lower temperatures than distillation which depends

6

of the relative volatility19. Therefore, pervaporation is strongly recommended for separation of

7

heat-sensitive products and azeotropic mixtures19 like is the BAc system.

8

Meanwhile, pervaporation-based hybrid processes have been investigated18, 20-25, mainly for

9

intensification of esterification reactions26, since it enables to reduce the concentrations and the

10

flow rates to be treated and, consequently, the energy requirement and associated costs25. That

11

way, the multicomponent study by pervaporation is really interesting, since this technology can

12

be integrated with chemical reaction helping to shift the reaction equilibrium conversion (which

13

is 63 % at 363 K on batch conditions and using an equimolar mixture of reactants in the BAc

14

system case27, for instance) towards the product (BAc) formation by continuous water

15

permeation. Accordingly, this kind of reaction with very slow kinetic (

16

1 kc,363K = 3.51×10−3 mol.g −A15 .min −1 ,using Amberlyst-15 as catalyst27, for instance) can be

17

favored by the continuous water removal from the reaction medium. Recently, Sert and Atalay

18

published a study28 about the synthesis of BAc combining pervaporation and reaction for the

19

first time. For that, the authors used a Pervap 2201 polymeric membrane in a batch reactor and

20

the effect of the membrane process on the shift of the chemical equilibrium was evaluated by

21

changing the temperature, the initial molar ratio of reactants, the ratio of membrane area to

22

volume of the reaction as well as the catalyst (Amberlyst 131) loading. They concluded that the

23

most important parameter was the temperature, since pervaporation and reaction rates increased

24

with the operating temperature. An increase in conversion of 40 % was observed in that process.

25

However, the effect of driving force was not studied and, according to the literature17, it is a key

26

element to evaluate the pervaporation performance for multicomponent mixtures. Besides that,

27

experimental pervaporation data for multicomponent mixtures in absence of reaction are 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

required to understand the influence of the different compounds in the performance of the

2

membrane for the global process (reaction combined with pervaporation) and these data were

3

missing in the previously cited study28.

4

In this work, a multicomponent pervaporation study for the compounds involved in the BAc

5

production in absence of reaction was performed, for the first time, at different temperatures,

6

323, 353 and 363 K and different parameters were evaluated, including the driving force. A

7

commercial tubular inorganic membrane was used for the experimental measurements since it

8

presents better stability under acidic and high temperature conditions than polymeric

9

membranes, being the best alternative for the dehydration of the reaction medium21. According

Page 4 of 36

10

to the literature29, commercial microporous silica membrane from Pervatech showed better

11

selectivity and water flux than the one from Pervap SMS from Sulzer Chemtech in dehydration

12

of aqueous mixtures, so a Pervatech BV membrane was considered. The activation energies

13

were also estimated by the dependence of permeance on temperature for each compound taking

14

into account the driving force. A mathematical model was developed considering the mass

15

transport under non-isothermal conditions and it was applied for the study of enhanced

16

esterification reaction of AAc with n-butanol by combining a fixed-bed reactor (FBR) with a

17

pervaporation membrane for the BAc synthesis. This configuration was already investigated for

18

other esterification reactions as the esterification of oleic oil with ethanol30, for which an

19

increase of 3 % of the limiting reactant conversion was observed for the pervaporation-assisted

20

reaction, and the esterification of lactic acid with ethanol21, for which the integration of a

21

pervaporation membrane increased more than twice the limiting reactant conversion of a

22

conventional FBR process, considering non-isothermal conditions.

23

Moreover, it is important to refer that all experimental pervaporation data of this work will be

24

useful to design and optimize different configurations of pervaporation-based hybrid processes

25

for BAc synthesis.

26 27

2. Experimental data 4 ACS Paragon Plus Environment

Page 5 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1


2.1. Chemicals and materials

2

The chemicals used in the pervaporation experiments were n-butanol (≥ 99.9 wt.%) from Fisher

3

Scientific, acrylic acid (≥ 99 wt.%) and butyl acrylate (≥ 99.5 wt.%) from Acros Organics.

4

Acrylic acid and butyl acrylate were provided stabilized with inhibitor (about 200 ppm and 20

5

ppm of hydroquinone monomethyl ether, respectively). The additional inhibitor used in this

6

study was phenothiazine (Ptz) (99 wt.%), also from Acros Organics. Isopropanol (≥ 99.9 wt.%)

7

from Fisher Scientific was used as solvent in the chromatographic analysis.

8

A commercial Hybrid Silica AR membrane supplied by Pervatech BV (The Netherlands) was

9

used, which presents hydrophilic characteristics. It consists of a modified silica selective layer

10

coated onto gamma alumina and its separation layer is applied inside of an asymmetric ceramic

11

tube that has an outer diameter of 10 mm, an inner diameter of 7 mm, and a length of 50 cm.

12

This membrane has an effective area per tube of about 110 cm2. It is able to be in contact with

13

any solvent at any concentration, however it is sensitive to extremely acidic and alkaline media

14

being the limit pH range from 0.5 to 8.5.

15 16

2.2. Experimental setup and procedure

17

The experimental data were measured in a pervaporation membrane unit at pilot scale which is

18

represented in Figure 1. It can work either in batch or continuous mode at temperatures up to

19

100 ºC. It is equipped with temperature (TI) (type K thermocouple, accuracy of about ± 2.2 ºC)

20

and pressure sensors (PI), for the absolute and permeate pressures with accuracies of about ± 0.5

21

bar and ± 1 mbar, respectively. The temperature was controlled by a thermostatic bath (Lauda,

22

Germany) with thermal M bath fluid (able to operate from 40 to 170 ºC) that flows through the

23

feed vessel jackets; pressure was set to 1.5 bar by applying an overpressure of helium to the

24

system in order to prevent vaporization of feed mixture over the whole temperature range.

25

Firstly, the feed vessel was charged with approximately 1.5 L of solution and the heating was

26

switched on at the desired temperature. A positive displacement diaphragm pump (Hydra Cell 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 36

1

G-03, Wanner International) was used to recirculate the feed solution over the entire system

2

including the membrane in order to keep it at the same temperature in absence of vacuum on

3

permeate side. When the temperature is constant, the pervaporation experiment was started by

4

applying vacuum to the permeate side with a vacuum pump (Boc Edwards, U.K.). Two parallel

5

glass cold trap partially submerged in liquid nitrogen allowed the vapor permeate condensation.

6

In the beginning, a cleaning procedure was performed by collecting the permeate sample during

7

the first minute for a glass cold trap which was then rejected. After that, the permeate sample

8

was collected in the other cold trap which was defrosted at the end of the experiment to be

9

weighted and analyzed by gas chromatography.

10

Along these measurements it is important to keep the feed composition nearly constant, so the

11

duration of each experiment was conditioned by the trade-off between ensuring a constant feed

12

and a reasonable amount of permeate. Samples were collected before and after each experiment

13

in order to verify the feed composition. The reproducibility was checked by collecting two or

14

three permeate samples under steady-state conditions at each temperature.

15 16

Figure 1. Setup of pervaporation membrane pilot scale unit.

17 18

2.3. Analytical method

6 ACS Paragon Plus Environment

Page 7 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

All samples collected were analyzed (at least two times) in a Shimadzu - GC 2010 Plus gas

2

chromatograph equipped with a flame ionization detector (FID). The compounds were separated

3

using a silica capillary column (CPWax57CB, 25 m x 0.53 mm ID, film thickness of 2.0 µm).

4

The temperature of the injector was set to 523 K while the temperature of the FID was set to be

5

573 K. The initial column temperature was 353 K for 5 min, the temperature was then increased

6

at 353 K.min-1 up to 473.15 K and kept constant for the following 7 min. Helium N50 was used

7

as the carrier gas and the linear velocity was set to 30 cm.s-1. The injection volume used was 1.0

8

µL with a split ratio of 30 for permeate samples and 90 for the retentate samples. Isopropyl

9

alcohol was used as internal standard and acetone as cleaning solvent. The global associated

10

uncertainty of the measured molar fractions was ≤ 0.05.

11

3. Mathematical Model

12

The experimental pervaporation data measured for the silica membrane selected in this work

13

were used for the validation of the mathematical model that was developed to predict the

14

behavior of the separation process, considering the following assumptions: (i) non-isothermal

15

operation due to the heat consumption for species vaporization; (ii) plug flow (retentate stream);

16

(iii) retentate velocity variations inside the membrane due to permeation of components; (iv)

17

concentration polarization due to the global membrane resistance (diffusive transport in the

18

boundary layer combined with the membrane resistance); and (v) continuous process once it is a

19

process subsequent to the FBR unit and both should be operating continuously.

20

Retentate mass balance to component i

21

∂Cret ,i ∂t

+

∂ ( vs Cret ,i ) ∂z

(1)

+ Am J i = 0

22

where 𝑧 is the axial coordinate in the membrane modules, Cret is the liquid phase concentration

23

in the retentate side, 𝑣! is the superficial velocity, 𝐴! is the membrane area per unit membrane

24

modules volume and 𝐽! is the permeate molar flux of specie i, through the membrane, defined

25

as: 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(

)

1

Ji = kov,i xi ai Pi 0 − yi Pperm

2

where 𝑘!",! is the global membrane mass transfer coefficient, that combines the resistance due

3

to the diffusive transport in the boundary layer with the membrane12 resistance:

4

1 kov ,i

=

1 Qmemb,i

+

Page 8 of 36

(2)

ai Pi 0Vmol ,i

(3)

kbl

5

2 For laminar flow and Graetz number, dint vs / ( Dm Lm ) , much greater that one, the mass transfer

6

coefficient for transport in the boundary layer,

0.33

Sc

0.33

⎛ dint ⎞ ⎜ ⎟ ⎝ Lm ⎠

kbl , is determined by the Lévêque correlation31 :

0.33

7

Sh = 1.62Re

8

where

9

Schmidt numbers, respectively,

, (Re < 2300)

Sh = kbl dint / Dm , Re = ρ dint vs / η and Sc = η / ( ρ Dm ) are the Sherwood, Reynolds and

Dm is the solute diffusivity in the boundary layer, 𝑑!"# is the Lm is the membrane length, ρ is the density and η

10

inside diameter of the tubular membrane,

11

is the viscosity.

12

The mole fraction of component i in the vapor phase (permeate side), yi , is defined as:

13

yi =

14

15

(4)

Ji

∑

n

(5)

J

i =1 i

The fluid velocity variation in the membrane feed side is calculated from the total mass balance: n n dV dvs = vs ∑Ci mol ,i − Am ∑J iVmol ,i dz dz i =1 i =1

16

where n is the total number of components and Vmol ,i it is the molar volume of the component i.

17

Retentate heat balance

(6)


Page 9 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

n

n ˆ C ∂T + v Cˆ C ∂T + A h (T − T ) = 0 C ∑ ∑ s p,i ret ,i ∂z m f p ,i ret ,i m ∂t i =1 i =1

2

where Cˆ p ,i is the liquid heat capacity of component i, T is the absolute temperature in the feed

3

side of the membrane, Tm is the membrane absolute temperature, and h f is the heat transfer

4

coefficient in the liquid boundary layer.

5

Membrane heat balance

6

dint 2

( (dint + δ ) / 2)

2 int

− (d / 4)

h f (T − Tm ) =

(7)

n

dint + δ 2

((dint + δ ) / 2 )

− (d

2 int

∑ΔH / 4)

v i

Ji

(8)

i =1

dint is the internal radius of the membrane, δ is the membrane thickness, and ΔH iv is

7

where

8

the heat of vaporization of species i. The heat transport coefficient was estimated by the Sieder-

9

Tate correlation, valid for laminar pipe flow32: 0.33

0.14

10

d ⎞ ⎛ Nu = 1.86 ⎜ RePr int ⎟ L ⎠ ⎝

11

where N u = h f dint / λ and Pr = η Cˆ p ,i / λ are the Nusselt and Prandtl numbers, respectively,

12

η and 𝜂! are the viscosity of the liquid in the feed and in the membrane wall, and 𝜆 is the

13

thermal conductivity.

14

Initial and boundary conditions:

15

t = 0:

18

T = TF

z = 0:

(9)

(10.1)

Cret ,i = C0,i

16

17

⎛η ⎞ ⎜ ⎟ ⎝ ηm ⎠

(10.2)

T = TF

(10.3)

Cret ,i = CF ,i

(10.4)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 36

1

where subscripts 0 and F refer to initial state and membrane feed conditions, respectively.

2

The molar fractions of all the components at the outlet of the membrane and in the permeate

3

stream were monitored:

4

xout ,i =

5

where

6

The FBR was simulated according to the mathematical model described in our previous work33.

7

Numerical solution

8

The numerical solution of this problem was obtained by using the commercial software

9

gPROMS (general PROcess Modeling System) version 4.2.0, using orthogonal collocation in

10

finite elements (OCFEM) with second order polynomials and one internal collocation point in

11

each element; to this end, the axial dimension of the membrane was discretized in 100 finite

12

elements. The DASOLV equation solver was used to solve the resulting system of ordinary

13

differential equations in time. For all simulations a tolerance of 10-5 was used.

Cout ,i

∑

(11)

n

C i =1 out ,i

Cout ,i = Cret ,i ( z =1) .

14 15

4. Results and discussion

16

4.1. Preliminary study

17

A preliminary study was performed to find the minimal feed flow rate required to operate in

18

absence of mass transfer resistance in the boundary layer due to concentration polarization

19

effects. This phenomena is related to the accumulation of the retained species near the

20

membrane surface and, consequently, their concentration will be higher in the boundary layer

21

(adjacent to the membrane surface) than in the bulk34. Generally, concentration polarization

22

results from the depletion of the most permeable component in the vicinity of the

23

feed/membrane interface and it is due to the slow diffusion of the solute from the bulk of the


Page 11 of 36

1

feed to the boundary layer35. The membrane flux and separation efficiency may be severely

2

impaired due to polarization36.

3

In this work, the effect of mass transfer in the boundary layer was investigated with a binary

4

mixture of n-butanol and 31 % of feed water mole fraction at constant temperature, 323.15 K,

5

by varying the feed flow rate. According to Figure 2, this effect is negligible for flow rates

6

higher than approximately 150 L/h, since the total permeate flux remains constant from this

7

value. So, all pervaporation experiments were carried out at 200 L/h in this work ensuring

8

absence of mass transfer resistance.

1.0 0.8

Jtot kg/(h.m2)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


0.6 0.4 0.2

0.0 0

100

150

200

250

300

350

Feed Flowrate (L/h)

9 10 11

50

Figure 2. Total permeate flux as function of feed flow rate (Pperm = 35 mbar, T = 323.15 K, xwater = 0.309).

12 13

4.2. Pervaporation data

14

The performance of the membrane for the n-butanol/AAc/BAc/water system was evaluated by

15

measuring pervaporation data for different mixture compositions in absence of reaction at three

16

different temperatures: 323.15, 353.15 and 363.15 K. A pressure of 45 mbar was set in permeate

17

side for all runs. The esterification reaction between AAc and n-BuOH presents a very slow

18

kinetics, according to the literature27, which allows to measure pervaporation data for

19

quaternary mixtures in absence of reaction even at high temperatures, since no catalyst was 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 36

1

used. This fact was verified by collecting feed samples after each experiment and analyzing it

2

by gas chromatography. All experimental mixture compositions considered in this work are

3

presented in Table 1. The feed compositions reported are the average of the feed composition

4

values at the beginning and at the end of the time interval of the permeate sample collection.

5

However the molar fraction of each component did not change more than 1% .These mixtures

6

were prepared taking into account the miscibility ranges, which were studied by the UNIFAC-

7

DMD model using the database available in the software AspenTech - AspenONE (version 7.1)

8

and can be observed in the ternary diagram presented in Supporting Information (Figure S1).

9

The pair n-butanol/ water (B1 to B4) was initially studied since it represents the extract stream

10

of the SMBR4 and enhanced SMBR5 processes. The AAc/ water binary was not considered due

11

to the high risk of polymerization described in the literature33 apart from membrane safety

12

issues. According to the ternary diagram (Figure S1), there is a large area of immiscibility

13

between BAc and water, so this pair was also not studied. Three different ternary mixtures (T1

14

to T3) were selected taking into account the same feed water molar fraction studied in the binary

15

mixtures for membrane performance comparison besides the miscibility range. One quaternary

16

composition (Q1) was selected to have the same feed water molar composition of binary and

17

ternary mixtures for performance comparison (± 10 % of water). The other mixture (Q2)

18

represents the equilibrium composition of the esterification reaction between n-butanol and AAc

19

with a molar feed ratio (3:1), since n-butanol is the eluent in SMBR based processes being

20

always in excess in relation to the other components.

21

Table 1. Feed molar compositions of the different mixtures studied. Mixture/Component

x,Butanol

x,Water

x,BAc

x,AAc

B1

0.741

0.259

-

-

B2

0.800

0.200

-

-

B3

0.911

0.089

-

-

B4

0.940

0.060

-

-

T1

0.652

0.159

0.189

-

T2

0.795

0.109

0.096

12

ACS Paragon Plus Environment

Page 13 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


T3

0.526

0.061

0.413

-

Q1

0.708

0.095

0.097

0.100

Q2

0.587

0.175

0.176

0.062

1 2

4.2.1. Dehydration of n-butanol: Effect of temperature and feed water composition

3

Temperature is a very important parameter in pervaporation process as reported in the

4

literature19, 28, 37. According to the experimental results presented in Figure 3, it is possible to

5

observe that the total permeate flux increases with the feed water mole fraction and the

6

operating temperature. The use of alumina as intermediate layer increases the hydrophilicity of

7

the membrane, promoting adsorption and diffusion of water38 and leading to higher selectivity

8

for this species. The experimental selectivity values obtained for the studied membrane were

9

between 5 and 160 (𝑆 = 𝑄!"#$% /𝑄!!!"#$ ) which is within of the range reported in the

10

literature for microporous silica membranes7. Thus, the total permeate flux was mainly

11

composed by water, as shown in Figure 4 (yw ≥ 96 %), which substantially increased with the

12

feed water mole fraction due to the higher driving force in mixtures with large feed water

13

content. The effect of the feed water mole fraction and temperature in the water driving force is

14

presented in

15

Figure 5.

16

On the other hand, the separation of liquid mixtures by partial vaporization through a membrane

17

is the principle behind pervaporation processes26, so in this kind of process the mass transfer is

18

accompanied with heat transfer19. In this way, increasing the temperature also increases the

19

driving force for water since the heat required for a phase change from the liquid to the vapor

20

phase, is lower for this compound than for n-BuOH, presenting a higher vapor pressure for the

21

same temperature (see table 2), leading to higher permeate fluxes.



8.0

n-BuOH/water (74/26)

7.0

n-BuOH/water (80/20)

Jtot (kg/(h.m2))

6.0

n-BuOH/water (91/9)

5.0

n-BuOH/water (94/6)

4.0 3.0 2.0 1.0

0.0 320

330

340

350

360

370

T (K)

Figure 3. Total permeate flux as a function of temperature for different n-BuOH/ water binary

1.00

0.05

0.99

0.04

0.98

0.03

0.97

0.02

0.96

0.01

Permeate BuOH mole fraction

mixtures.

Permeate water mole fraction

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 36

0.95 0.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 1

Feed water mole fraction

2

Figure 4. Permeate composition as a function of feed water molar composition for n-

3

BuOH/water binary mixtures (circles: water, triangles: n-BuOH).


Page 15 of 36

50

Driving force (kPa)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


40 30

water, 323 K water, 353 K water, 363 K

20 10 0 0.060

0.089 0.196 Feed water mole fraction

0.259

1 2 3

Figure 5. Influence of temperature and feed water mole fraction on the driving force of water. Table 2. Vapor pressure for water, n-BuOH, BAc and AAc at different temperatures39 Compound/ Temperature

323.15

353.15

363.15

Water

0.124

0.473

0.701

n-BuOH

0.047

0.224

0.349

BAc

0.029

0.110

0.163

AAc

0.024

0.105

0.162

4 5

4.2.2. Multicomponent pervaporation data for BAc system

6

Similarly, a multicomponent pervaporation study was performed for the compounds involved in

7

the BAc production in absence of reaction and different parameters were evaluated. Several

8

ternary and quaternary mixtures were considered according to Table 1. Regarding to the feed

9

water composition and temperature effects, both parameters increase the total permeate flux like

10

happened in the binary mixture as is shown in Figure 6 (ternary mixtures) and Figure 7

11

(quaternary mixtures). The same conclusions reported previously for the binary system are valid

12

for the multicomponent system. However, the total permeate flux decreases about 30 % when

13

the BAc is present in the feed solution (ternary mixture T2) in relation to the binary mixture

14

(B3) with the same feed water mole fraction. Regarding to AAc, its presence together with BAc

15

in the feed solution (Q1) leads to a loss of 56 % of the total permeate flux in relation to the 15 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 36

1

binary with the same feed water mole fraction (B3). In the same way, if we compare the

2

quaternary (Q2) and the binary (B3) mixtures, although the first has almost twice of the feed

3

water content of the second, both show identical total permeate flux. These comparisons are

4

performed in Figure 8. These facts suggests that, the presence of molecules like BAc and AAc

5

induces higher mass transfer resistances in the boundary layer of the membrane significantly

6

hindering the water flux. Nevertheless, their composition on the permeate stream is residual as it

7

is shown in Figures S2 and S3 (see Supporting Information), where are presented the permeate

8

stream compositions for ternary and quaternary mixtures, respectively. The membrane

9

selectivity towards the studied species follows the order: water > n-BuOH > AAc > BAc.

10

According to Table 2, n-BuOH presents the second highest flux trough the membrane.

11

Moreover, BAc has the highest radius of gyration which means that it has the biggest molecular

12

size, hindering its passage through the membrane. This way, AAc is slightly preferably

13

permeate than BAc. The radius of gyration and the dipole moment parameters are shown in

14

Table 3. Furthermore, the high boiling points of the BAc (418.15 K) and AAc (414.15 K) lead

15

to smaller driving force comparing with n-butanol (390.85 K) and water (373.15 K).

16

The influence of temperature and feed water mole fraction on permeate compositions are

17

summarized in Figure 9. Increasing the temperature the water mole composition in permeate

18

side increases whereas the n-butanol, BAc and AAc decreases. Regarding to the feed water

19

mole composition the same behavior is observed, however this parameter seems to have no

20

more influence in the permeate water mole fraction at 363.15 K, since at this temperature the

21

permeate composition is constant for the different molar feed water compositions.


Page 17 of 36

4.0

n-BuOH/BAc/Water (65/19/16)

3.5


Jtot (kg/(h.m2))

3.0


2.5 2.0 1.5

1.0 0.5 0.0 320

330

340

350

360

370

T (K)

1 2

Figure 6. Total permeate flux as a function of temperature for different ternary mixtures: n-

3

BuOH/ Water/BAc (black triangles: T1, grey triangles: T2, white triangles: T3).

2.5

n-BuOH/AAc/BAc/Water (58/6/18/18)

2.0

BuOH/AAc/BAc/Water (70/10/10/10)

Jtot (kg/(h.m2))

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.5 1.0 0.5 0.0 320

4

330

340

350

360

370

T (K)

5

Figure 7. Total permeate flux as a function of temperature for different quaternary mixtures: n-

6

BuOH/ Water/BAc/AAc (black squares: Q1, white squares: Q2).



3.0

n-BuOH/Water (91/9)

Jtot (kg/(h.m2))

2.5

n-BuOH/BAc/Water (79/10/11) n.BuOH/Aac/BAc/water (70/10/10/10)

2.0

n-BuOH/AAc/BAc/Water (58/6/18/18)

1.5 1.0

0.5 0.0 320

330

340

350

360

370

T (K)

1 2

Figure 8. Total permeate flux as a function of temperature for the same feed water molar

3

compositions (± 10 %): circles: binary mixture B3, triangles: ternary mixture T2, black squares:

4

quaternary mixture Q1, white squares: quaternary mixture Q2). Table 3. Radius of gyration and dipole moment for each compound40.

5

Component/ Parameter

Radius of gyration (Ắ)

Dipole moment (debyes)

Water

0.615

1.850

n-BuOH

3.251

2.569

BAc

4.765

1.931

AAc

2.978

0.462

6

Permeate water mole fraction

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 36

1.000 0.995 363

0.990 353

0.985

323

0.980 10

7

18

Feed Water mole fraction

8

Figure 9. Water content on the permeate stream as a function of temperature and feed water

9

mole fraction for quaternary mixtures. 18 ACS Paragon Plus Environment

Page 19 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Membrane productivity is usually characterized by permeate flux, which relates the product

2

separation rate to the membrane area required to achieve the separation19. However, other

3

important factors should be evaluated, particularly for multicomponent mixtures. In this work,

4

the driving force and separation factor were also taken into account. The driving force can be

5

defined as the difference of partial pressures of the respective component at the feed and

6

permeate side (see section 4.2.3). Figure 10 provides information about the driving force for

7

water in the different studied mixtures. In order to have the same water permeate flux, the water

8

requires higher driving force in the quaternary mixtures, like it was expected, due to the

9

presence of the other compounds.

10

The process separation factor, which is described in the following section (see section 4.2.3), is

11

represented as function of temperature for binary, ternary and quaternary mixtures in Figure 11

12

which shows that the separation factor is higher for quaternary mixtures and increases with the

13

temperature. In order words, the membrane presents excellent selectivity for water mainly in

14

ternary and quaternary mixtures at high temperatures (above 353 K), probably due to the lower

15

vapor pressures of AAc and BAc, as mentioned previously (Table 2). Comparing the separation

16

factor determined in this work for a feed water mole fraction of approximately 10 %, at 353 K

17

and 45 mbar (αperv = 665), with the values reported in the literature for the same type of

18

membranes under similar conditions (αperv = 340, at 348 K and 16 mbar)10 it is possible to

19

conclude that the Pervatech membrane used in this work has a separation factor almost two

20

times higher. However, Boutikos et al.7 reported a very similar value to the one found in this

21

work (around 600).

22

It is important to mention that these experimental data can be very useful for future studies

23

about process intensification for BAc synthesis, helping to understand what is the most

24

sustainable way to produce it.



6.0

Jw (kg/(h.m2))

5.0 4.0 3.0 2.0 1.0 0.0 0

1

5

10 15 20 25 30 35 Driving force for water (kPa)

40

45

2

Figure 10. Total permeate flux as a function of the driving force for binary (circles), ternary

3

(triangles) and quaternary mixtures (squares).

10000

Separation Factor

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 36

8000 6000 4000 2000 0 320

4

330

340 350 Temperature (K)

360

370

5

Figure 11. Separation factor as a function of temperature for binary (circles), ternary (triangles)

6

and quaternary mixtures (squares).

7 8

4.2.3. Pervaporation transport and parameters estimation

9

The solution-diffusion model41 was applied to describe the mass transport through the

10

membrane which is given by the following equation (assuming that the mass transfer resistance

11

in the boundary layer can be neglected): 20 ACS Paragon Plus Environment

Page 21 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(

)

1

Ji = Qmemb,i xi ,F ai pi0,F − yi , perm Pperm

2

where Qmemb ,i and 𝑥!,! are the permeance of the membrane in relation to component i and the

3

liquid molar fraction in feed, ai is the activity coefficient (determined using the UNIFAC

4

method), pi0, F , is the saturation pressure in feed (Table 2), yi , perm and Pperm are the vapor molar

5

fraction and the total pressure in permeate side. The last term xi , F ai pi0, F − yi , perm Pperm gives de

6

driving force of the component i.

7

The separation factor, α, is given by:

8

α perv =

9

Where the subscripts F and w corresponding to the feed and water, respectively.

(12)

(

)

yw, perm (1 − xw, F )

(13)

xw, F (1 − yw, perm )

10

In this work, the activation energy required to the phase change of the permeate species were

11

obtained by fitting the experimental results to the Arrhenius equation to describe the permeance

12

dependence on temperature:

13

⎛ −E ⎞ Qmemb,i = Qmemb ,0 exp ⎜ P ⎟ ⎝ RT ⎠

14

where Qmemb ,0 is the pre-exponential factor , T is the absolute temperature, R is the ideal gas

15

constant and EP is the activation energy of permeation which is a combination of the activation

16

energy of diffusion (ED) and the heat of adsorption of the permeate in the membrane (ΔH):

17

EP = ED + ΔH

18

Since temperature influences both membrane permeability and the driving force for mass

19

transport, the activation energy of permeation should be evaluated from the slope ln(Qmemb ,i ) vs

20

1/ T 19. This way, an average of all permeance values for each species and temperature were

21

calculated and the activation energies of permeation as well as the pre-exponential factors were

(14)

(15)



1

determined from the corresponding linear regressions. The Arrhenius adjustments are shown in

2

the Figure 12 for all species. The values of the respective parameters are presented in Table 4.

3

The water presents the lowest activation energy followed by n-BuOH, BAc and the AAc is the

4

compound that requires more activation energy to be permeate through the membrane. The

5

negative value of the activation energy reveals that the permeation of all species investigated is

6

governed by the adsorption, according to the equation (15).

7

The effect of the introduction of additional species to form ternary and quaternary mixtures was

8

also evaluated by comparing the permeation results obtained for water and n-BuOH in binary

9

mixtures The respective activation energies and pre-exponential factors for each compound in

10

the different systems studied are presented in Tables S1 and S2 (in Supporting Information),

11

respectively. From that, it is possible to determine the respective permeance of the each

12

compound for different temperatures, which are displayed for 363.15 K, as an example, in Table

13

S3. According to the results (Table S3), the multicomponent system has more influence in the

14

butanol permeance than in the water permeance when compared with the binary system. The

15

water permeance decreases almost twice while the butanol permeance decreases about 30 times

16

in the quaternary system comparing with the binary one. Even though water permeance

17

decreases it is still species with higher permeation values in the multicomponent system.

0

ln (Qmemb) (mol/(s.m2.Pa)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 36

-5 -10

-15 -20 -25 2.7

18 19

2.8

2.9 1000/T (K-1)

3.0

3.1

Figure 12. Linearized Arrhenius plot for permeance as a function of temperature. 22 ACS Paragon Plus Environment

Page 23 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1 2

Table 4. Overall activation energies and pre-exponential factors for each compound. Eperm, i (kJ/mol)

Q0, i (mol/(s.m2.Pa)

Water

-30.53

3.90E-11

n-BuOH

-26.62

5.66E-13

BAc

-20.85

1.72E-12

AAc

-9.37

5.17E-10

Component

3 4

4.3. Pervaporation membrane-assisted esterification reaction

5

In order to predict the behavior of the continuous reaction-separation process by combining a

6

fixed-bed reactor with a membrane, the experimental data obtained (permeance of each species

7

according to the respective driving force and temperature) were used according to the

8

mathematical model described in section 3 (at the steady-state). The fixed-bed results were

9

obtained from the model reported in our previous work33.

10

The performance of a fixed-bed membrane reactor (FBMR) consisting in a Pervatech tubular

11

membrane packed (inside) with Amberlyst-15 ion exchange resin (A15) was assessed. The

12

reaction kinetic and thermodynamic equilibrium constants considered are reported in the

13

literature27 as well as the adsorption parameters for A15 studied in our previous work33. In Table

14

5 are presented the reactor parameters used in the simulations. The optimal feed molar reactants

15

ratio (n-BuOH/AAc) was investigated for the FBMR towards a more efficient process.

16

According to the results presented in Table S4 (see Supporting Information), the optimal molar

17

ratio can be considered to be 1.3 since the reaction conversion was still very close to 99 % as for

18

the maximum ratio studied (3.0). Furthermore, using this ratio (1.3) the FBMR reaction

19

conversion enhancement in relation to the FBR is much higher (about twice) than the value

20

observed for the maximum studied ratio (3.0) as well as the concentration of BAc attained (4.96

21

mol.L-1) at steady-state and isothermal conditions (around of 80 % higher). In other words,

22

pervaporation has more impact in the global process using smaller ratios (n-BuOH/AAc) since

23

there is less competition with water in the permeation process through the membrane. This way, 23 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 36

1

the water permeance is more efficient leading to high reaction conversions and to higher

2

concentration of the target product (BAc) at the reactor outlet.

3

reactants molar ratio (n-BuOH/AAc) the reaction conversion decreases so, to select the optimal

4

reactants ratio a compromise between the reaction conversion attained and the pervaporation

5

efficiency must be established. Accordingly, the feed reactants molar ratio considered was 1.3

6

(nn-BuOH/nAAc) and the dimensions of the FBR used experimentally in our previous work were

7

kept. Initially, it was considered that the A15 was saturated with n-BuOH at 363 K and the n-

8

BuOH/AAc mixture was fed at 1 ml/min. The performance of a conventional FBR (without

9

membrane), considering isothermal conditions, is presented in Figure 13 (a). After that, a

10

FBMR was simulated at the same operating and design parameters and the respective

11

concentration profiles are given in Figure 13 (b). Like in the previous work33 where

12

experimental FBR data were well predicted with the mathematical model proposed using the

13

experimental adsorption33 and kinetic parameters27, experimental concentration profiles of the

14

FBMR at the steady-state should be reasonably well predicted using the mathematical model

15

presented with the same experimental adsorption and kinetic parameters but now including also

16

the experimental pervaporation data measured in this work. Comparing the concentration

17

profiles in the FBR and the FBMR at isothermal conditions, a considerable improvement is

18

observed in the performance of the FBMR since at steady-state AAc is almost fully converted

19

achieving a reaction conversion of 98.7 % while the FBR reached 59.4 % at the same

20

conditions. This result is a direct consequence of the continuous removal of water from the

21

reaction media by pervaporation (in the FMBR) which displaces the chemical reaction

22

equilibrium, extending the conversion beyond the values predicted by the thermodynamics.

23

According to simulation results, just one membrane (with effective area of 110 cm2) is required

24

to obtain this excellent performance producing BAc with a purity of 97.8 % (in solvent free

25

basis). Figure 13 (c) shows the concentration profiles of the FBMR at steady-state considering

26

non-isothermal operation as well as the temperature profile. The performance is impaired

27

clearly by the temperature drop of about 45 ºC along the membrane due to the heat required for

28

species vaporization. This leads to lower water permeation fluxes and mainly lower reaction

However, decreasing the


Page 25 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

rates, decreasing the conversion about 28 % in relation to the FBR at the same operating

2

conditions. The same behavior was reported by Pereira et al.21 for the esterification reaction of

3

lactic acid with ethanol to produce ethyl lactate. Nevertheless, the good performance attained

4

under isothermal conditions can be ensured by using a heat source to the pervaporation

5

membrane in order to offset the temperature required for species vaporization, depending on the

6

selective layer coating is placed. According to the literature, a heated sweep gas source is the

7

most suitable way to provide heat to the retentate liquid stream when the selective layer coated

8

is inside the membrane tube, like it was assumed in this case. On the other hand, if the selective

9

layer is coated on the external (shell) side of the membrane tube, an appropriate heated solution

10

can be used as heat source by re-circulating it through jacketed modules being an efficient way

11

to do it since the heat can be efficiently transferred to the retentate stream21. This way, the

12

energy required to heat the membrane adsorptive reactor in order to keep the operation closer of

13

the isothermal conditions was determined according to the heat of vaporization of water and

14

taking into account the average molar flux of water (8.16E-04 mol/(dm2.min)) attained at

15

isothermal conditions (363 K) leading to a value of 33.2 J/(dm2.min). Despite this, the energy

16

saving potential of the FBMR comparing with the conventional process is evident since at the

17

end of the FBMR a reaction mixture with approximately 75.7 % of BAc and 22.6 % of n-BuOH

18

(using a ratio of 1.3) is obtained having the remain compounds residual compositions (1.0 % of

19

AAc and 0.7 % of water). This way, just one distillation column shall be required to purify BAc

20

instead of the three distillations columns required by the conventional process. Moreover, the

21

two reactors used in the conventional process can be reduced to one making use of the

22

adsorption and pervaporation technologies as suggested in this work.

23

Table 5. Reactor parameters used in simulation runs. Reactor Parameters Feed concentration (mol/L)

Cn-BuOH = 6.39 ; CAAc = 4.91

Feed temperature (K)

363.15

Feed flowrate (mL/min)

1.0

Permeate pressure (mbar)

45 25

ACS Paragon Plus Environment


Bed porosity

0.415

Bed length (cm)

34.0

Internal diameter (cm)

1.95

12

n-BuOH AAc BAc Water

10

C (mol/L)

8 6 4 2

0 0.0

0.2

0.4

0.6

0.8

1.0

z

1 2

a)

12

n-BuOH AAc BAc Water

10

C (mol/L)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 36

8 6 4

2 0 0.0

3 4

0.2

0.4

0.6

0.8

1.0

z b)

5


12

n-BuOH AAc BAc Water Temp

10

C (mol/L)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


8

370 360 350

6

340

4

330

2

320

0

310 0.0

0.2

0.4

0.6

0.8

T (K)

Page 27 of 36

1.0

z

1 2

c)

3

Figure 13. Concentration profiles at steady-state of the: a) FBR at isothermal conditions; b)

4

FBMR at isothermal conditions; c) FBMR at non-isothermal conditions.

5 6

5. Conclusions

7

Pervaporation data for multicomponent mixtures were measured for the compounds involved in

8

butyl acrylate synthesis in absence of reaction for the first time. The performance of commercial

9

tubular silica membrane supplied by Pervatech BV (The Netherlands) was evaluated by

10

studying different parameters. Selectivity and separation factor were determined and presented

11

similar values to the literature.

12

Besides the selectivity towards water be induced by the alumina used as intermediate layer on

13

the membrane, the driving force of this species is enhanced by increasing the temperature, since

14

water presents the highest vapor pressure. These conditions lead to higher permeate water mole

15

fraction and, consequently, higher total flux since it is composed mainly by water. The feed

16

water mole fraction also increases the driving force of this compound leading to higher

17

permeate fluxes, however from 363.15 K it seems no to have more significant influence.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

The presence of the BAc and AAc harms severely the total permeate flux inducing to higher

2

mass transfer resistances in the boundary layer of the membrane hindering the water flux. The

3

compounds are preferably permeate by the following order: water > n-BuOH > AAc > BAc

4

mainly due to low vapor pressures of BAc and AAc.

5

The performance of a pervaporation-based hybrid process by combining a fixed-bed reactor

6

with a membrane was studied for the first time for the butyl acrylate synthesis. For that, a

7

mathematical model was developed taking into account the experimental data obtained in this

8

work and simulations were carried-out to predict the concentrations profiles of a FBMR at the

9

steady-state. Considering isothermal operation, the FBMR presented a limiting reactant

10

conversion 66 % higher comparing with a FBR at the same conditions due to the continuous

11

water permeate fluxes that allow to remove practically all water produced in the reaction.

Page 28 of 36

12 13

Supporting Information

14

The ternary diagram for n-BuOH/BAc/Water system as well as the permeate stream

15

composition as a function of temperature for the different studied systems (binaries, ternaries

16

and quaternaries) are presented in Supporting Information. Furthermore, the activation energy,

17

the pre-exponential factor and the permeance of each compound in the different systems are

18

summarized. Finally, the reaction conversion obtained in a FBR and FBMR using different

19

reactants molar ratios are shown in that section.

20 21

Acknowledgments

22

This work is a result of project “AIProcMat@N2020 - Advanced Industrial Processes and

23

Materials for a Sustainable Northern Region of Portugal 2020”, with the reference NORTE-01-

24

0145-FEDER-000006, supported by Norte Portugal Regional Operational Programme (NORTE

25

2020), under the Portugal 2020 Partnership Agreement, through the European Regional

26

Development Fund (ERDF) and of Project POCI-01-0145-FEDER-006984 – Associate


Page 29 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Laboratory LSRE-LCM funded by ERDF through COMPETE2020 - Programa Operacional

2

Competitividade e Internacionalização (POCI) – and by national funds through FCT - Fundação

3

para a Ciência e a Tecnologia.

4 5 6

Notation Abbreviations AAc

Acrylic Acid

-

A15

Amberlyst 15 ion exchange resin

-

BAc

Butyl Acrylate

-

FBR

Fixed-Bed Reactor

FBMR

Fixed-Bed Membrane Reactor

gPROMS

General Process Modeling System

-

n-BuOH

n-butanol

-

RD

Reactive Distillation

RSR

Reactor Separation Recycle

-

SMBR

Simulated Moving Bed Reactor

-

Symbols



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Am

membrane area per unit module volume

m2/m3

a

liquid phase activity

-

C Cˆ

liquid phase concentration

mol.L-1

liquid heat capacity

J.mol-1.K-1

internal diameter of membrane

m

solute diffusivity in the boundary layer

m2.s-1

activation energy of diffusion

J.mol-1

activation energy of permeation

J.mol-1

hf ΔH

heat transfer coefficient in the liquid boundary layer

W.K-1

heat of adsorption

J.mol-1

ΔH v

heat of vaporization

J.mol-1

J

total permeate flux

kg.m-2. h-1

kbl

mass transfer coefficient

m.s-1

kov ,i

global membrane mass transfer coefficient

kg.m-2. h-1. bar-1

Lm

membrane length

m

number of compounds

-

Nu

Nusselt number

-

P Pr Qmemb Re

Vapor pressure

bar

Prandtl number

-

permeance

kg.m-2. h-1. bar-1

Reynolds number

-

S

Selectivity of the membrane

Sc

Scherwood number

-

Sh

Schmidt number

-

t

time variable

s

T

temperature

K

νs VM x y z

superficial velocity

m.s-1

molar volume

m3.mol-1

liquid phase molar fraction

-

vapor phase molar fraction

-

axial coordinate in the membrane module

m

η

fluid viscosity in the feed

kg.m-1.s-1

ηm

fluid viscosity in the membrane wall

kg.m-1.s-1

ρ

density

kg.m-3

α

relative volatility

-

α perv

separation factor

-

δ

membrane thickness

m

p

dint Dm ED EP

n

0

Page 30 of 36

1 Greek Letters


Page 31 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


λ

thermal conductivity

(W.m-1.K-1)

Subcripts 0

relative to initial conditions

F

relative to feed

i

Component i (n-BuOH, water, AAc and BAc)

m

membrane

out

at the outlet

ret

retentate

perm

permeate

w

water

1 2

References

3

(1) Niesbach, A.; Fuhrmeister, R.; Keller, T.; Lutze, P.; Górak, A. Esterification of Acrylic Acid

4

and n-Butanol in a Pilot-Scale Reactive Distillation Column—Experimental Investigation,

5

Model Validation, and Process Analysis. Ind. Eng. Chem. Res. 2012, 51, 16444.

6

(2) Niesbach, A.; Kuhlmann, H.; Keller, T.; Lutze, P.; Górak, A. Optimisation of Industrial-

7

Scale n-Butyl Acrylate Production using Reactive Distillation. Chem. Eng. Sci. 2013, 100, 360.

8

(3) Moraru, M. D.; Milea, A.; Bîldea, C. S. Design and Economic Evaluation of a Process for n-

9

Butyl Acrylate Production. U.P.B. Sci. Bull, Series B 2016, 78, 113.

10

(4) Constantino, D. S. M.; Pereira, C. S. M.; Faria, R. P. V.; Loureiro, J. M.; Rodrigues, A. E.

11

Simulated Moving Bed Reactor for Butyl Acrylate Synthesis: From Pilot to Industrial Scale.

12

Chem. Eng. Process. 2015, 97, 153.

13

(5) Constantino, D. S. M.; Faria, R. P. V.; Pereira, C. S. M.; Loureiro, J. M.; Rodrigues, A. E.

14

Enhanced Simulated Moving Bed Reactor Process for Butyl Acrylate Synthesis: Process

15

Analysis and Optimization. Ind. Eng. Chem. Res. 2016, 55, 10735.

16

(6) Sommer, S.; Melin, T. Influence of Operation Parameters on the Separation of Mixtures by

17

Pervaporation and Vapor Permeation with Inorganic Membranes. Part 1: Dehydration of

18

Solvents. Chem. Eng. Sci. 2005, 60, 4509.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 36

1

(7) Boutikos, P.; Pereira, C. S. M.; Silva, V. M. T. M.; Rodrigues, A. E. Performance

2

Evaluation of Silica Membrane for Water–n-Butanol Binary Mixture. Sep.Purif. Technol. 2014,

3

127, 18.

4

(8) Hyder, M. N.; Huang, R. Y. M.; Chen, P. Pervaporation Dehydration of Alcohol–Water

5

Mixtures: Optimization for Permeate Flux and Selectivity by Central Composite Rotatable

6

Design. J. Membr. Sci. 2009, 326, 343.

7

(9) Jiang, L. Y.; Chung, T.-S.; Rajagopalan, R. Dehydration of Alcohols by Pervaporation

8

through Polyimide Asymmetric Hollow Fibers with Various Modifications. Chem. Eng. Sci.

9

2008, 63, 204.

10

(10) Sommer, S.; Melin, T., Performance Evaluation of Microporous Inorganic Membranes in

11

the Dehydration of Industrial Solvents. Chem. Eng. Process. 2005, 44, 1138.

12

(11) Guo, W. F.; Chung, T.-S.; Matsuura, T. Pervaporation Study on the Dehydration of

13

Aqueous Butanol Solutions: a Comparison of Flux vs. Permeance, Separation Factor vs.

14

Selectivity. J. Membr. Sci. 2004, 245, 199.

15

(12) Wijmans, J. G.; Athayde, A. L.; Daniels, R.; Ly, J. H.; Kamaruddin, H. D.; Pinnau, I. The

16

Role of Boundary Layers in the Removal of Volatile Organic Compounds from Water by

17

Pervaporation. J. Membr. Sci. 1996, 109, 135.

18

(13) Zhang, S.; Zou, Y.; Wei, T.; Mu, C.; Liu, X.; Tong, Z. Pervaporation Dehydration of

19

Binary and Ternary Mixtures of n-Butyl Acetate, n-Butanol and Water using PVA-CS Blended

20

Membranes. Sep. Purif. Technol. 2017, 173, 314.

21

(14) Smitha, B.; Suhanya, D.; Sridhar, S.; Ramakrishna, M. Separation of Organic–Organic

22

Mixtures by Pervaporation—A Review. J. Membr. Sci. 2004, 241, 1.

23

(15) Kopeć, R.; Meller, M.; Kujawski, W.; Kujawa, J. Polyamide-6 Based Pervaporation

24

Membranes for Organic–Organic Separation. Sep. Purif. Technol. 2013, 110, 63.

25

(16) Ray, S.; Ray, S. K. Synthesis of Highly Methanol Selective Membranes for Separation of

26

Methyl Tertiary Butyl Ether (MTBE)–Methanol Mixtures by Pervaporation. J. Membr. Sci.

27

2006, 278, 279.


Page 33 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

(17) Luis, P.; Van der Bruggen, B. The Driving Force as Key Element to Evaluate the

2

Pervaporation Performance of Multicomponent Mixtures. Sep. Purif. Technol. 2015, 148, 94.

3

(18) Lipnizki, F.; Field, R. W.; Ten, P.-K. Pervaporation-Based Hybrid Process: A Review of

4

Process Design, Applications and Economics. J. Membr. Sci. 1999, 153, 183.

5

(19) Feng, X.; Huang, R. Y. M. Liquid Separation by Membrane Pervaporation: A Review. Ind.

6

Eng. Chem. Res. 1997, 36, 1048.

7

(20) Sanz, M. T.; Gmehling, J. Study of the Dehydration of Isopropanol by a Pervaporation-

8

Based Hybrid Process. Chem. Eng. Technol. 2006, 29, 473.

9

(21) Pereira, C. S. M.; Silva, V. M. T. M.; Pinho, S. P.; Rodrigues, A. E. Batch and Continuous

10

Studies for Ethyl Lactate Synthesis in a Pervaporation Membrane Reactor. J. Membr. Sci. 2010,

11

361, 43-55.

12

(22) Pereira, C. S. M.; Silva, V. M. T. M.; Rodrigues, A. E. Green Fuel Production Using the

13

PermSMBR Technology. Ind. Eng. Chem. Res. 2012, 51, 8928.

14

(23) Van Hoof, V.; Van den Abeele, L.; Buekenhoudt, A.; Dotremont, C.; Leysen, R. Economic

15

Comparison between Azeotropic Distillation and Different Hybrid Systems combining

16

Distillation with Pervaporation for the Dehydration of Isopropanol. Sep. Purif. Technol. 2004,

17

37, 33.

18

(24) Chandane, V. S.; Rathod, A. P.; Wasewar, K. L. Enhancement of Esterification Conversion

19

using Pervaporation Membrane Reactor. Resource-Efficient Technologies 2016, 2, 47.

20

(25) Ahmad, S. A.; Lone, S. R. Hybrid Process (Pervaporation–Distillation): A Review. Int. J.

21

Sci. Eng. Res. 2012, 3, 1.

22

(26) Jyoti, G.; Keshav, A.; Anandkumar, J. Review on Pervaporation: Theory, Membrane

23

Performance, and Application to Intensification of Esterification Reaction. J. Eng. 2015, 2015,

24

927068.

25

(27) Ostaniewicz-Cydzik, A. M.; Pereira, C. S. M.; Molga, E.; Rodrigues, A. E. Reaction

26

Kinetics and Thermodynamic Equilibrium for Butyl Acrylate Synthesis from n-Butanol and

27

Acrylic Acid. Ind. Eng. Chem. Res. 2014, 53, 6647.

28 33 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 36

1

(28) Sert, E.; Atalay, F. S. n-Butyl Acrylate Production by Esterification of Acrylic Acid with n-

2

Butanol combined with Pervaporation. Chem. Eng. Process. 2014, 81, 41.

3

(29) Casado, C.; Urtiaga, A.; Gorri, D.; Ortiz, I. Pervaporative Dehydration of Organic Mixtures

4

using a Commercial Silica Membrane: Determination of Kinetic Parameters. Sep. Purif.

5

Technol. 2005, 42, 39.

6

(30) Han, Y.; Lv, E.; Ma, L.; Lu, J.; Chen, K.; Ding, J. Coupling Membrane Pervaporation with

7

a Fixed-Bed Reactor for Enhanced Esterification of Oleic Acid with Ethanol. Energy Convers.

8

Manage. 2015, 106, 1379.

9

(31) Lévêque, A. Les Lois de La Transmission de Chaleur par Convection. Dunod, 1928.

10

(32) Welty, J.; Wicks, C.; Wilson, R.; Rorrer, G. Fundamentals of Momentum, Heat, and Mass

11

Transfer; John Wiley & Sons: New York, 2008.

12

(33) Constantino, D. S. M.; Pereira, C. S. M.; Faria, R. P. V.; Ferreira, A. F. P.; Loureiro, J. M.;

13

Rodrigues, A. E. Synthesis of Butyl Acrylate in a Fixed-Bed Adsorptive Reactor over

14

Amberlyst 15. AIChE J. 2015, 61, 1263.

15

(34) Mulder, M. H. V. Chapter 2: Polarization Phenomena and Membrane Fouling. Membrane

16

Science and Technology; Elsevier Science: New York, 1995.

17

(35) Néel, J. Chapter 5: Pervaporation. Membrane Science and Technology; Elsevier Science:

18

New York, 1995.

19

(36) Strathmann, H. Membrane Separation Processes, 4. Concentration Polarization and

20

Membrane Fouling. Ullmann's Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag

21

GmbH & Co. KGaA, 2000.

22

(37) Qunhui, G.; Ohya, H.; Negishi, Y., Investigation of the Permselectivity of Chitosan

23

Membrane used in Pervaporation Separation II. Influences of Temperature and Membrane

24

Thickness. J. Membr. Sci. 1995, 98, 223.

25

(38) Sekulić, J.; Ten Elshof, J.; Blank, D., Separation Mechanism in Dehydration of

26

Water/Organic Binary Liquids by Pervaporation Through Microporous Silica. J. Membr. Sci.

27

2005, 254, 267.


Page 35 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

(39) Yaws, C. L.; Narasimhan, P. K.; Gabbula, C., Yaws' Handbook of Antoine Coefficients for

2

Vapor Pressure (2nd Electronic Edition). In Knovel.

3

(40) Yaws, C. L., Thermophysical Properties of Chemicals and Hydrocarbons. William

4

Andrew: 2008.

5

(41) Wijmans, J. G.; Baker, R. W., The Solution-Diffusion Model: a Review. J. Membr. Sci.

6

1995, 107, 1.

7 8

List of Figure Captions

9

Figure 1. Setup of pervaporation membrane pilot scale unit.

10

Figure 2. Total permeate flux as a function of feed flow rate (Pperm = 35 mbar, T = 323.15 K,

11

xwater = 0.309).

12 13

Figure 3. Total permeate flux as a function of temperature for different n-BuOH/ water binary mixtures.

14

Figure 4. Permeate composition as a function of feed water molar composition for n-

15

BuOH/water binary mixtures (circles: water, triangles: n-BuOH).

16

Figure 5. Influence of temperature and feed water mole fraction on the driving force of water.

17

Figure 6. Total permeate flux as a function of temperature for different ternary mixtures: n-

18

BuOH/ Water/BAc (black triangles: T1, grey triangles: T2, white triangles: T3).

19

Figure 7. Total permeate flux as a function of temperature for different quaternary mixtures: n-

20

BuOH/ Water/BAc/AAc (black squares: Q1, white squares: Q2).

21

Figure 8. Total permeate flux as a function of temperature for the same feed water molar

22

compositions (± 10 %): circles: binary mixture B3, triangles: ternary mixture T2, black squares:

23

quaternary mixture Q1, white squares: quaternary mixture Q2).

24

Figure 9. Water content on the permeate stream as a function of temperature and feed water

25

mole fraction for quaternary mixtures.

26

Figure 10. Total permeate flux as a function of the driving force for binary (circles), ternary

27

(triangles) and quaternary mixtures (squares). 35 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 36

1

Figure 11. Separation factor as a function of temperature for binary (circles), ternary (triangles)

2

and quaternary mixtures (squares).

3

Figure 12. Linearized Arrhenius plot for permeance as a function of temperature.

4

Figure 13. Concentration profiles at steady-state of the: a) FBR at isothermal conditions; b)

5

FBMR at isothermal conditions; c) FBMR at non-isothermal conditions.

6 7 8

For Table of Contents Only

9


Performance Evaluation of Pervaporation ... - ACS Publications

Recommend Documents