- Email: [email protected]
hydrophilic pervaporation process
Chemical Engineering and Processing 77 (2014) 22–29
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Separation and purification of isobutanol from dilute aqueous solutions by a hybrid hydrophobic/hydrophilic pervaporation process Mostafa Omidali a , Ahmadreza Raisi a,b,∗ , Abdolreza Aroujalian a,b a
Department of Chemical Engineering, Amirkabir University of Technology (TehranPolytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran Food Process Engineering and Biotechnology Research Center, Amirkabir University of Technology (Tehran Polytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran b
a r t i c l e
i n f o
Article history: Received 31 July 2013 Received in revised form 8 December 2013 Accepted 11 January 2014 Available online 21 January 2014 Keywords: Pervaporation Isobutanol Hybrid process Composite membranes PDMS PVA
a b s t r a c t In this study, a hybrid hydrophobic/hydrophilic pervaporation process was employed to separate and purify isobutanol from its dilute aqueous solutions. For this purpose, composite polydimethylsiloxane membranes were initially used for the recovery of isobutanol by hydrophobic pervaporation. Then the hydrophilic pervaporation with a composite polyvinyl alcohol membrane was utilized to separate water from the organic phase of the permeate stream of the hydrophobic pervaporation. The effect of feed flow rate on the performance of pervaporation was investigated. The resistance in series model was also applied to calculate the transport resistances through the composite membranes. It was observed that an enhancement in the feed flow rate led to higher permeation flux and selectivity of the more permeable component, while the flux of the less permeable component was almost constant. Also, the ratio of liquid boundary layer resistance to membrane layer resistance decreased by an increase in the feed flow rate. The isobutanol with a purity of higher than 99 wt.% was produced by the hybrid hydrophobic/hydrophilic pervaporation technique from a 2 wt.% aqueous isobutanol solution. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Biofuels are a suitable replacement for fuels gets from sources are not renewable or their renewal period is too long. Depletion of natural or petroleum products and rising prices of raw materials result in a search for renewable energy sources and biofuels [1]. Among biofuels, bioethanol is widely produced and used in different industries. Biobutanol is another biofuel that has properties which are more attractive in comparison with bioethanol. Biobutanol as opposed to bioethanol is non-hygroscopic and corrosion, and has a higher calorific value due to its greater energy content. Also, butanol is combinable with gasoline in any composition and its mixture with gasoline has a lower vapor pressure in comparison with ethanol-gasoline that leads to less fuel vaporization and destruction [1]. Isobutanol is produced by the carbonylation of propylene. It is also produced naturally during the fermentation of carbohydrates [2]. Isobutanol can also be produced by some engineered microorganisms such as corynebacterium [3]. Isobutanol produced from the biological method has a low concentration
∗ Corresponding author at: Department of Chemical Engineering, Amirkabir University of Technology (TehranPolytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran. Tel.: +98 21 64543125; fax: +98 21 66405847. E-mail address: [email protected] (A. Raisi). 0255-2701/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cep.2014.01.002
and needs to be concentrated and purified for use in different applications [1,4]. Separation processes such as distillation, liquidliquid extraction, adsorption, gas stripping and pervaporation have been used to separate alcohols from their aqueous solutions [5–7]. Compared to traditional processes, the pervaporation process has many advantages such as no heat damage to heat-sensitive compounds, low energy consumption, no additional separation treatment for added solvents or absorbents and minimum loss of alcohol. Therefore, pervaporation is an economical and useful technique for extraction of isobutanol from fermentation broth and its aqueous solutions [8,9]. Pervaporation is a membrane process for the separation of liquid mixtures by partial vaporization through a non-porous membrane. There are a large number of studies examining the performance of the pervaporation process for the separation of various alcohols such as methanol [10–12], ethanol [13–18], isopropanol [19–21], n-butanol [8,22–26] and isobutanol [22,27,28] from their mixtures with water. In these studies, a hydrophilic membrane was employed to separate water from the alcohol/water mixtures or a hydrophobic membrane was used to recover alcohol from its aqueous solutions, and further purification was not performed on the products of the pervaporation process. Furthermore, pervaporation-based hybrid processes have employed for the recovery and separation of alcohols from their aqueous solutions [29–31].
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
23
2. Theory List of symbols Nomenclature A constant in Eq. (13) a constant of Sherwood number B constant in Eq. (13) b constant of Sherwood number constant of Sherwood number c d constant of Sherwood number Di diffusion coefficient of component i into the membrane (m2 /s) dh hydraulic diameter of the membrane module (m) Ji flux of component i (g/m2 s) KL mass transfer coefficient of the boundary layer (m/s) L length of the membrane module (m) number of elements N Pi permeability coefficient of component i (cm3 cm/cm2 s cmHg) partial pressure of component i (cmHg) pi equilibrium vapor pressure of component i (cmHg) po,i permeate pressure (cmHg) pp Re Reynolds number RL mass transfer resistance of the liquid boundary layer (s/m) mass transfer resistance of the membrane layer RM (s/m) Rt total mass transfer resistance (s/m) S area of the membrane (m2 ) Sc Schmitt number Sherwood number Sh t time duration of the experiment (s) u feed velocity (m/s) weight of the collected permeate (g) W xi mole fraction of component i Greek letters ˛ selectivity membrane thickness (m) ıM C concentration difference (g/m3 ) kinematic viscosity (m2 /s) Subscripts and superscripts i component index f feed permeate p
The flux of component i varies linearly with the gradient in partial pressure according to [32]: Ji =
Pi (pf,i − pp,i ) ıM
(1)
where pf,i and pp,i are the partial pressure of component i in the feed mixture and vapor permeate, respectively, ıM is the membrane thickness and Pi is the permeability coefficient. The partial pressure of component i in the feed is: pf,i = i xf,i po,i
(2)
where i is the activity coefficient, xf,i is the mole fraction in the feed and po,i is the equilibrium vapor pressure of component i. According to Dalton’s law, the partial pressure of a component in the permeate can be expressed as: pp,i = xp,i pp
(3)
where xp,i is the mole fraction of the component in the permeate and pp is the permeate pressure. Due to a very low total permeate pressure, the partial pressure of the permeate is very small in comparison with the partial pressure of the feed, so it can be neglected. The flux can thus be assumed to vary linearly with the equilibrium vapor pressure of the pure component: Ji =
Pi i xf,i po,i ıM
(4)
Furthermore, according to resistance in series model, the transport of components from the feed solution through the composite membrane occurs by the following steps [33]: (i) diffusion through the liquid boundary layer, (ii) sorption into the membrane active layer, (iii) diffusion of liquid through the membrane active layer, (iv) desorption out of the active layer, and (v) transport of vapors through the porous support. Among these resistances, the liquid boundary layer and membrane active layer resistances control the mass transport in the pervaporation, and other transport resistances are low and negligible [14,16]. Therefore, under the steady state conditions, the flux of component i across the membrane can be expressed as a ratio of the driving force over the total mass transfer resistance (Rt ), as follows: Ji =
Ci Rt
(5)
where Rt = RL + RM In the present study, a hybrid hydrophobic/hydrophilic pervaporation process is used for the recovery and purification of isobutanol from aqueous solution. At first, isobutanol is separated from its dilute aqueous solutions through the hydrophobic pervaporation process with a composite polydimethylsiloxane membrane. Since isobutanol and water dissolve in each other only to a limited extent, the permeate stream of the hydrophobic pervaporation separates into two insoluble liquid phases, an isobutanol-rich phase and a water-rich one. The water-rich liquid phase is returned to the feed solution of the hydrophobic pervaporation process. Then the hydrophilic pervaporation with a composite polyvinyl alcohol membrane is utilized to separate water from the organic phase of the permeate stream of the hydrophobic pervaporation. Furthermore, the resistance in series model is applied to analyze the mass transport resistances and study the transport mechanisms in the composite membrane.
(6)
and RL = RM =
1 KL
(7)
ıM Di
(8)
where KL and Di are the mass transfer coefficient of boundary layer and diffusion coefficient of component i in the membrane, respectively. The Sherwood correlation has been used to determine the mass transfer coefficient (KL ) through the boundary layer [16,34]: Sh =
dh KL L = aReb Sc c D L
d (9)
The constants a, b, c and d for the Sherwood correlation are given in Table 1.
24
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
Table 1 Variables for the Sherwood correlation for plate and frame modules. Reynolds number range
a
b
c
d
Re < 2300 Re > 2300
1.615 0.026
0.33 0.80
0.33 0.30
0.33 –
By substitution of Eqs. (6)–(8) into Eq. (5), the following relation is obtained: Rt =
Ci 1 ıM = + Ji KL Di
(10)
Eq. (9) could be written as:
L 1 dh = Re−b Sc −c KL aD L
Fig. 1. Schematic of the pervaporation instrument used in this work.
−d (11)
or 1 = ARe−b KL
(12)
According to Table 1, coefficient A is independent of Reynolds number. A combination of Eqs. (10) and (12) give: Ci = ARe−b + B Ji
(13)
The coefficients A and B are determined from the slop and intercept of the (Ci /Ji ) versus Re−b plots, respectively. 3. Materials and methods Isobutanol (99.5 wt.%) that was used in the pervaporation experiments was purchased from Merck Co. Ltd. (Darmstadt, Germany) and de-ionized laboratory water was used for making aqueous mixtures. The composite PDMS (polydimethylsiloxane) membranes used in the experiments were kindly supplied by HelmholtzZentrum Geesthacht Zentrum fur Material und Kustenforschung GmbH (Geesthacht, Germany). The composite PVA (polyvinyl alcohol) (PERVAP® 2210) was purchased from Sulzer Chemtech Ltd. (Switzerland). All membranes used in the experiments are listed in Table 2. Separation and purification of isobutanol from its aqueous solutions is carried out in three steps as follows: 1. Pervaporative separation of isobutanol from dilute aqueous solutions by the composite PDMS membranes. 2. Physical phase separation of the permeate of the first step into two insoluble liquid phases (aqueous and organic phases) without any energy consumption. 3. Pervaporative purification of isobutanol from the organic phase of the second step by the composite PVA membrane. The membrane was cut into a 15 cm × 20 cm piece and held in a flat frame membrane module. The pervaporation apparatus has been previously described [35]. A schematic diagram of the experimental apparatus is shown in Fig. 1. The feed was kept in a 1 l reservoir and its temperature was controlled during the experiments with a precision of ±0.5 ◦ C. The feed was pumped from
Table 2 The hydrophobic and hydrophilic membranes used in the experiments. Membrane
Composite layer
Active layer
Active layer thickness (m)
PDMS 08/082-6 PDMS 08/082-11 PERVAP® 2210
PET/PAN/PDMS PET/PAN/PDMS PVA/PAN/PPS
PDMS PDMS PVA
2.3 12.9 1.8–2.8
the feed tank to a stainless steel SEPA CF pervaporation module (Osmonics Inc., Minnetonka, MN, USA). The retentate coming out of the membrane module was recycled to the feed tank. The vacuum was applied to the permeate side of the membrane by a vacuum pump (Busch Inc., Switzerland) in order to provide a pressure gradient across the membrane. The vacuum pressure was controlled by a needle valve and monitored by a vacuum barocel gauge with a precision of ±1 mmHg. The collected frozen permeate was melted and weighed. The isobutanol composition of the feed and permeate streams was determined using Gas Chromatography (Younglin 6000M Series Gas Chromatography System, Anyang Korea). The oven temperature was kept at 100 ◦ C for 7 min. Helium with a column head pressure of 0.7 bar was used as carrier gas. The injector and detector temperatures are 200 and 220 ◦ C, respectively. The injector is operated in the split mode (10:1). The separation performance of the pervaporation process was investigated by flux and selectivity. The total flux (J) was calculated using the following equation: J=
W S·t
(14)
where W is the weight of the collected permeate, S the area of membrane and t is the time duration of the experiments. Isobutanol selectivity (˛) was quantified as the separation factor based on: ˛=
p
p
f
f
xi /xj
(15)
xi /xj
where x is the mole fraction, i and j refer to the isobutanol and water, respectively, as well as f and p referring to the feed and permeate. In this work, initial recovery of isobutanol from the dilute aqueous solutions was carried out by the hydrophobic membrane and the effect of feed flow rate (0.35, 0.70, 2.35, 3.50 and 4.65 l/min) which corresponds to the Reynolds number (150, 300, 1000, 1500 and 2000) on the permeation flux and selectivity was studied for feed concentration of 2 wt.% with two composite PDMS membranes (active layer thicknesses of 2.3 and 12.9 m). Then isobutanol purification was performed with a hydrophilic membrane and the effect of feed flow rate (0.35, 0.70, 2.35, 3.50, 4.65 and 5.60 l/min) which corresponds to the Reynolds number (70, 140, 460, 690, 910 and 1100) at the feed concentration equal to water solubility in the isobutanol was investigated. The feed temperature and permeate side pressure was fixed at 30 ◦ C and 1 mmHg, respectively, for all experiments. Al1 experimental conditions were repeated three times and the average values are reported. The time duration of each experiment was 7 h and a permeate sample was collected every hour. Steady state permeation was reached after an hour at all experimental conditions.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a)
500
hydrodynamic conditions in the feed flow and can be calculated as follows:
140
total flux water flux isobutanol flux
120 200 110
100
(b)
0
1
2
3
4
5
100
Feedflowrate(lit/min) 300
100
total flux water flux isobutanol flux
250
Flux(g/m 2h)
Flux(g/m 2h)
300
0
Re =
130
90
200 80 150 70 100 60
50 0
Flux(g/m 2h)
Flux(g/m 2h)
400
0
1
2
3
4
5
50
Feedflowrate(lit/min)
(c)
30
Selectivity
25 20 15 10 2.3 µm
5
12.9 µm
0
0
1
2
3
4
25
5
Feedflowrate(lit/min) Fig. 2. Effect of feed flow rate on the total and partial fluxes for the PDMS membranes ((a) membrane thickness of 2.3 m and (b) membrane thickness of 12.9 m) and (c) effect of feed flow rate on the selectivity.
4. Results and discussion 4.1. Isobutanol concentration by hydrophobic membrane The effect of feed flow rate can be taken into account to investigate the performance of the pervaporation process. Fig. 2a and b shows the effect of feed flow rate which corresponds to the Reynolds number, on total and partial fluxes for isobutanol separation from 2 wt.% aqueous solution with composite PDMS membranes. The feed flow rate varies in the range of 0.35–4.65 l/min (Reynolds numbers of 150–2000). The Reynolds number (Re) is a well-known dimensionless description of the
udh
(16)
where u and are feed velocity and kinematic viscosity of solution, respectively and dh is the hydraulic diameter of the pervaporation module. Flow regimes into a channel are laminar, transient, and turbulent at Reynolds numbers of Re < 500, 500 < Re < 2000 and Re > 2000, respectively [36]. As shown in Fig. 2a and b, the water flux slightly increases with an increase in the feed flow rate, while the enhancement in the isobutanol flux is more significant. For example, when the feed flow rate changes from 0.35 to 4.65 l/min, the isobutanol flux goes from 56.1 to 63.3 g/m2 .h and the water flux varies from 129.2 to 131.5 g/m2 .h for the thicker PDMS membrane. Thus, for this feed flow rate range, the water flux which is controlled by the rate of diffusion through the membrane is independent of the feed flow rate. The change in the permeation flux of isobutanol by increasing the feed flow rate may be attributed to concentration polarization in the pervaporation process. The concentration polarization leads to a decrease in the permeation rate of the more permeable component, i.e. isobutanol [37]. An increase in the feed flow rate leads to a decline in the concentration polarization effect, and thus the isobutanol flux increases. In other words, when the feed flow rate enhances, the mass transfer coefficient of isobutanol in the solution goes up due to an increase of mixing and relative turbulency in the liquid boundary layer [13,23,35,38,39]. Furthermore, the variations in the isobutanol flux with increasing feed flow rate are more significant for the thinner membrane. These results can be related to the effect of membrane thickness on the membrane resistance and the ratio of liquid boundary to membrane resistance. As the membrane thickness decreases, the membrane resistance reduces and consequently the ratio of liquid boundary to membrane resistance increases. Therefore, the effect of feed flow rate will be more significant for the thinner membrane as can be seen in Fig. 2. To determine the mass transport resistances, the value of (Ci /Ji ) were plotted Re−0.33 and the results are presented in Fig. 3a and b for isobutanol and water, respectively. For different membrane thickness, the liquid boundary layer and membrane layer resistances were calculated from the slope and intercept of each line, respectively. The effect of feed flow rate on the liquid boundary layer and membrane layer resistances of isobutanol and water are indicated in Fig. 4. Also, the ratio of boundary layer resistance and membrane layer resistance to total resistance for both components is presented in Table 3. Fig. 4 shows that an increase in the feed flow rate leads to a decrease in the boundary layer resistance while membrane layer resistance remains constant. Therefore, the ratio of liquid boundary layer resistance to membrane layer resistance declines by an enhancement in the feed flow rate as can be seen from Table 3. The influence of the feed flow rate on the isobutanol selectivity is represented in Fig. 2c. As indicated, the isobutanol selectivity increases as the feed flow rate goes to higher levels. The selectivity can be related to the ratio of isobutanol to water flux. The isobutanol flux enhances with an increase in the feed flow rate while the water flux is almost constant. Therefore, the ratio of the two fluxes which is proportional to the isobutanol selectivity increases when the feed flow rate enhances [37]. This trend is same as the works investigated the effect of feed flow rate on selectivity of other components through the different membranes [13,35,38,39]. The membrane permeability and permeance (the ratio of permeability and membrane thickness) have been defined and used to represent the membrane property. It is a good way to
26
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
Table 3 The ratio of liquid boundary layer and membrane layer resistances to total resistance against isobutanol and water permeation through the PDMS membranes.
RL /Rt
Component
Thickness (m)
0.35
0.70
2.35
3.50
4.65
Isobutanol
2.3 12.9 2.3 12.9
0.2392 0.1952 0.0673 0.0288
0.2001 0.1617 0.0542 0.0230
0.1436 0.1145 0.0370 0.0156
0.1282 0.1019 0.0326 0.0137
0.1181 0.0936 0.0298 0.0125
2.3 12.9 2.3 12.9
0.7608 0.8048 0.9327 0.9712
0.7999 0.8383 0.9458 0.9770
0.8564 0.8855 0.9630 0.9844
0.8718 0.8981 0.9674 0.9863
0.8819 0.9064 0.9702 0.9875
Water
RM /Rt
Isobutanol Water
Feed flow rate (l/min)
Table 4 The permeance coefficient (cm3 /cm2 s cmHg) of isobutanol and water through the PDMS membrane at 30 ◦ C. Component
Membrane thickness (m)
Feed flow rate (l/min) 0.35
0.70
2.35
3.50
4.65
Isobutanol
2.3 12.9
5.81 × 104 3.16 × 104
6.09 × 104 3.34 × 104
6.46 × 104 3.41 × 104
6.64 × 104 3.47 × 104
6.74 × 104 3.57 × 104
Water
2.3 12.9
3.74 × 104 1.82 × 104
3.74 × 104 1.82 × 104
3.79 × 104 1.83 × 104
3.82 × 104 1.83 × 104
3.82 × 104 1.83 × 104
(a) 1.5
values confirmed the results in Fig. 4 and Table 3. It can be seen from Table 4 that the permeance coefficient of the ethanol increases with an enhancement in the feed flow rate. This implies that an increase in the feed flow rate leads to lower liquid boundary layer resistance and consequently the permeance coefficient enhances. Also,
(a) R² = 0.9648
0.9
0.6
RM × 10-5 (s/m)
ΔC/Ji (s/m)
1.2
R² = 0.9913
0.3
2.3 µm
6
3.0
5
2.5
4
2.0
3
1.5
2
1.0
1
0.0
0.00
0.05
0.10
0.15
0.20
0
0.25
0.5
RM
12.9 µm
RL
0
1
Re-0.33
(b) 3.0
RL × 10-5 (s/m)
characterize membrane and can be used as a criterion for membrane selection and material design. The overall permeance coefficients of the studied membranes were calculated according to Eq. (4), the results are presented in Table 4. The permeance
2
3
4
0.0 5
Feed low rate (lit/min)
(b) 1.5
15
1.2
12
0.9
9
0.6
6
2.0
RM × 10-6 (s/m)
ΔC/Ji (s/m)
R² = 0.9679
1.5 R² = 0.9824
1.0 0.5 0.0
3
0.3
2.3 µm
RM
12.9 µm 0.00
0.05
RL× 10-5 (s/m)
2.5
0.10
0.15
0.20
0.25
Re-0.33 Fig. 3. Relationship between (Ci /Ji ) and Re−0.33 for isobutanol (a) and water (b) permeation.
0
RL
0
1
2
3
4
0
5
Feed low rate (lit/min) Fig. 4. Effect of feed flow rate on both liquid boundary layer and membrane layer resistances of isobutanol (a) and water (b) for membrane thickness of 2.3 m.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a)
27
6
150 R² = 0.9800
5
ΔC/Ji (s/m)
Total lux (g/h.m2)
120
90
4 3 2
60 1
Total flux
30
water 0 0.00
water flux isobutanol flux 0
0
1
2
3
4
5
6
(b) 15
Water selectivity
12
9
6
3
0
1
2
3
4
5
0.05
0.10
0.15
0.20
0.25
0.30
Re-0.33
Feed low rate (lit/min)
0
R² = 0.9626
isobutanol
6
Feed low rate (lit/min) Fig. 5. Effect of feed flow rate on the total (a) and partial fluxes (b) as well as the water selectivity (c) for the PVA membrane.
the permeance coefficient for the thicker membranes was lower due to higher membrane layer resistance. 4.2. Isobutanol purification by a hydrophilic membrane The product of the first step of the concentration process separates into an organic phase and an aqueous one. The water content of the organic phase is equal to the maximum solubility of water in isobutanol, i.e. 10 wt.% at 30 ◦ C. The aqueous phase has about 6 wt.% isobutanol which can return to the feed solution of the hydrophobic pervaporation process. A hydrophilic composite PVA membrane was employed to separate water from the isobutanol-rich solution and to purify the isobutanol. Fig. 5a shows the effect of feed flow rate on the permeation flux for separation of water from the isobutanol/water mixtures. It is observed that both total and water flux increase with enhancement of feed flow rate, while the isobutanol partial flux approximately remains constant. The effect of
Fig. 6. Relationship between (Ci /Ji ) and Re−0.33 for component permeation through the PVA membrane.
feed flow rate on the permeation flux of the PVA membrane may be related to the concentration polarization in the pervaporation process, as discussed in details in the previous section. It is quite evident that a higher feed flow rate will promote the flux because it could lessen the concentration polarization on the membrane surface and provide more opportunity for the sorption of water on the hydrophilic membrane. The influence of feed flow rate on the water selectivity is depicted in Fig. 5b. It can be seen that the selectivity slightly enhances from 9.3 to 11.9 when the feed flow rate varies from 0.35 to 5.60 l/min. When the feed flow rate enhances, the water flux increases while the permeation flux of isobutanol remains constant which leads to increase in the water selectivity. Similar observations were reported in the other researches [13,26,35,38,39]. For example, Srinivasan et al. [26] for the pervaporative recovery of 1-butanol from a model pharmaceutical aqueous waste observed that the separation factor of 1-butanol increases from 4.9 to 6.9 while the feed flow rate enhances from 0.4 to 0.8 l/min. Also, Rafia et al. [39] in the pervaporation experiments for aroma recovery from lemon juice reported that the enrichment factor of -pinene and limonene increases from 16.3 to 16.8 and from 16.9 to 17.4, respectively as the Reynolds number varies from 500 to 2300. Fig. 6 represents the linear relationship between (Ci /Ji ) and Re−0.33 and (Eq. (13)) for the water and isobutanol permeation through the PVA membrane. The effect of feed flow rate on the liquid boundary layer and membrane resistances is shown in Fig. 7. Also, the calculated values of the ratio of boundary layer and membrane layer resistances to total resistance for water and isobutanol are given in Table 5. The results indicate that by increasing the feed flow rate, the liquid boundary layer resistance decreases while the membrane layer resistance does not vary. Also, the membrane resistance for water is lower than isobutanol due to the small ˚ in comparison with the isobutanol molecular size of water (2.65 A) ˚ that simplifies diffusion of water molecules molecular size (7.70 A) [40]. As a result, Tables 3 and 5 indicate that the liquid boundary
Table 5 The ratio of liquid boundary layer and membrane layer resistances to total resistance against water and isobutanol permeation through the PVA membrane. Component
Feed flow rate (l/min) 0.35
0.70
2.35
3.50
4.65
5.60
RL /Rt
Water Isobutanol
0.4600 0.1774
0.4039 0.1464
0.3124 0.1031
0.2849 0.0916
0.2662 0.0841
0.2544 0.0795
RM /Rt
Water Isobutanol
0.5400 0.8226
0.5961 0.8536
0.6876 0.8969
0.7151 0.9084
0.7338 0.9159
0.7456 0.9205
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a) 4
with a purity of 99.7 wt.% was obtained from its dilute aqueous solutions by the proposed hybrid method.
Resistance × 10-6 (s/m)
28
5. Conclusions
3
2
1 RM 0
RL 1
0
2
3
4
5
6
15
4
12
3
9
2
6
1
3
RM × 10-7 (s/m)
(b) 5
RL × 10-6 (s/m)
Feed low rate (lit/min)
RM 0
RL 0
1
2
3
4
5
6
0
Separation and purification of isobutanol from its dilute aqueous solutions were performed using a hybrid pervaporation process with hydrophobic and hydrophilic composite membranes. The effect of feed flow rate on the performance of pervaporation was investigated. The proposed pervaporation technique enables to concentrate isobutanol from 2 wt.% to 99.7 wt.% The experimental results showed that for the PDMS and PVA composite membranes, an increase in the feed flow rate led to higher permeation flux and selectivity of the more permeable component, while the flux of the less permeable component was almost constant. The resistance in series model was used to analyze the mass transport resistances and the results indicated that by enhancement in the feed flow rate, the liquid boundary layer resistance of components declined while the membrane layer resistance remained constant. Also, results of modeling indicated that the effect of liquid boundary layer resistance is more significant for the thinner PDMS membrane. It can be concluded that the pervaporation process is an attractive technology for the recovery and purification of isobutanol from its aqueous solution as it yields good separation and operates under mild conditions. Acknowledgement The authors would like to thank Helmholtz-Zentrum Geesthacht Zentrum fur Material und Kustenforschung GmbH (Geesthacht, Germany) for supplying the PDMS membranes.
Feed low rate (lit/min) Fig. 7. Effect of feed flow rate on both liquid boundary layer and membrane layer resistances of water (a) and isobutanol (b).
layer resistance has an obvious role for dilute components permeation through the membrane. The main purpose of this study was the purification of isobutanol from its dilute aqueous solutions. The change in the isobutanol concentration during the hybrid purification process is shown in Fig. 8. It can be seen that the isobutanol was concentrated to a 34.4 wt.% solution from a 2 wt.% isobutanol aqueous solution by the hydrophobic pervaporation process (step 1) at 30 ◦ C. Then the isobutanol concentration increased from 34.4 to 90 wt.% by the physical phase separation (step 2). Finally, the isobutanol was purified to a 99.7 wt.% concentration by the hydrophilic pervaporation process with the PVA membrane (step 3). Therefore, the isobutanol
Fig. 8. The change in the isobutanol concentration during the hybrid purification process.
References [1] W. Kaminski, E. Tomczak, A. Gorak, Biobutanol production and purification methods, Ecol. Chem. Eng. S 18 (2011) 31–37. [2] S. Atsumi, T. Hanai, J.C. Liao, Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels, Nature 451 (2008) 86–89. [3] S.K. Michael, C.K. Myung, C. James, Engineering corynebacteriumglutamicum for isobutanol production, Appl. Microb. Biotechnol. 87 (2010) 1045–1055. [4] N. Qureshi, I.S. Maddox, Continuous production of acetone–butanol–ethanol using immobilized cells of Clostridium acetobutylicum and integration with product removal by liquid–liquid extraction, J. Ferment. Bioeng. 80 (1995) 185–189. [5] W.J. Groot, R. Lans, K. Luyben, Technologies for butanol recovery integrated with fermentations, Proc. Biochem. 27 (1992) 61–75. [6] L.M. Vane, Separation technologies for the recovery and dehydration of alcohols from fermentation broths, Biofuels Bioprod. Bioref. 2 (2008) 553–588. [7] A. Ranjan, V.S. Moholkar, Biobutanol: science, engineering, and economics, Int. J. Energy Res. 36 (2012) 277–323. [8] L.M. Vane, A review of pervaporation for product recovery from biomass fermentation processes, J. Chem. Technol. Biotechnol. 80 (2005) 603–629. [9] K.L. Wasewar, V.G. Pangarkar, Intensification of recovery of ethanol from fermentation broth using pervaporation: economical evaluation, Chem. Biochem. Eng. Q. 20 (2006) 135–145. [10] S.P. Doguparthy, Pervaporation of aqueous alcohol mixtures through a photopolymerised composite membrane, J. Membr. Sci. 185 (2001) 201–205. [11] X. Liu, Y. Sun, X. Deng, Studies on the pervaporation membrane of permeation water from methanol/water mixture, J. Membr. Sci. 325 (2008) 192–198. [12] B. Gonzalez, I. Ortiz, Modeling and simulation of a hybrid process (pervaporation–distillation) for the separation of azeotropic mixtures of alcohol–ether, J. Chem. Technol. Biotechnol. 77 (2011) 29–42. [13] X. Li, S. Wang, Some characteristics of pervaporation for dilute ethanol–water mixtures by alcohol-permselective composite membrane, Sep. Sci. Technol. 31 (1996) 2867–2873. [14] R. Jiraratananon, A. Chanachai, R.Y.M. Huang, D. Uttapap, Pervaporation dehydration of ethanol–water mixtures with chitosan/hydroxyethylcellulose (CS/HEC) composite membranes: I. Effect of operating conditions, J. Membr. Sci. 195 (2002) 143–151. [15] M. Garcia, M.T. Sanz, S. Beltran, Separation by pervaporation of ethanol from aqueous solutions and effect of other components present in fermentation broths, J. Chem. Technol. Biotechnol. 84 (2009) 1873–1882. [16] A. Raisi, A. Aroujalian, T. Kaghazchi, Experimental study and mass transport modeling of ethanol separation from aqueous solutions by pervaporation, Sep. Sci. Technol. 44 (2009) 3538–3570.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29 [17] B. Bolto, M. Hoang, Z. Xie, A review of membrane selection for the dehydration of aqueous ethanol by pervaporation, Chem. Eng. Process. Process Intensif. 50 (2011) 227–235. [18] C.H. Kang, Y.F. Lin, Y.S. Huang, K.L. Tung, K.S. Chang, J.T. Chen, W.S. Hung, K.R. Lee, J.Y. Lai, Synthesis of ZIF-7/chitosan mixed-matrix membranes with improved separation performance of water/ethanol mixtures, J. Membr. Sci. 438 (2013) 105–111. [19] J. Yu, C.H. Lee, W.H. Hong, Performances of crosslinked asymmetric poly(vinyl alcohol) membranes for isopropanol dehydration by pervaporation, Chem. Eng. Process. Process Intensif. 41 (2002) 693–698. [20] G. Dhanuja, B. Smitha, S. Sridhar, Pervaporation of isopropanol–water mixtures through polyion complex membranes, Sep. Purif. Technol. 44 (2005) 130–138. [21] N.R. Singha, S. Kar, S. Ray, S.K. Ray, Separation of isopropyl alcohol-water mixtures by pervaporation using crosslink IPN membranes, Chem. Eng. Process. Process Intensif. 48 (2009) 1020–1029. [22] K.W. Buddeker, G. Bengtson, H. Pingel, Pervaporation of isomeric butanols, J. Membr. Sci. 54 (1990) 1–12. [23] E. Favre, Q.T. Nguyen, Extraction of 1-butanol from aqueous solutions by pervaporation, J. Chem. Technol. Biotechnol. 65 (1996) 221–228. [24] A. Jonquieres, A. Fane, Filled and unfilled composite GFT PDMS membranes for the recovery of butanols from dilute aqueous solutions: influence of alcohol polarity, J. Membr. Sci. 125 (1997) 245–255. [25] N. Qureshi, I.S. Maddox, Reduction in butanol inhibition by perstraction: utilization of concentrated lactose/whey permeate by clostridium acetobutylicum to enhance butanol fermentation economics, Food Bioprod. Process. 83 (2005) 43–52. [26] K. Srinivasan, K. Palanivelu, A.N. Gopalakrishnan, Recovery of 1-butanol from a model pharmaceutical aqueous waste by pervaporation, Chem. Eng. Sci. 62 (2007) 2905–2914. [27] S. Korkmaz, Y. Salt, S. Dincer, Esterification of acetic acid and isobutanol in a pervaporation membrane reactor using different membranes, Ind. Eng. Chem. Res. 50 (2011) 11657–11666.
29
[28] M. Fathizadeh, A. Aroujalian, A. Raisi, M. Fotouhi, Preparation and characterization of thin film nanocomposite membrane for pervaporative dehydration of aqueous alcohol solutions, Desalination 314 (2013) 20–27. [29] F. Lipnizki, R.W. Field, P.K. Ten, Pervaporation-based hybrid process: a review of process design, applications and economics, J. Membr. Sci. 153 (1999) 183–210. [30] Z. Szitkai, Z. Lelkes, E. Rev, Z. Fonyo, Optimization of hybrid ethanol dehydration systems, Chem. Eng. Process. Process Intensif. 41 (2002) 631–646. [31] J.B. Haelssig, J. Thibault, A.Y. Tremblay, Numerical investigation of membrane dephlegmation: a hybrid pervaporation-distillation process for ethanol recovery, Chem. Eng. Process. Process Intensif. 50 (2011) 1226–1236. [32] A. Raisi, A. Aroujalian, T. Kaghazchi, Multicomponent pervaporation process for volatile aroma compounds recovery from pomegranate juice, J. Membr. Sci. 322 (2008) 339–348. [33] M.G. Liu, J.M. Dickson, P. Cote, Simulation of a pervaporation system on the industrial scale for water treatment, Part I: extended resistance-in-series model, J. Membr. Sci. 111 (1996) 227–241. [34] E.L. Cussler, Diffusion, Cambridge University Press, New York, 1989. [35] A. Aroujalian, A. Raisi, Recovery of volatile aroma components from orange juice by pervaporation, J. Membr. Sci. 303 (2007) 154–161. [36] V.L. Streeter, E.B. Wylie, Fluid Mechanics, sixth ed., McGraw-Hill, New York, 1975. [37] G.O. Yahaya, Separation of volatile organic compounds (BTEX) from aqueous solutions by a composite organophilic hollow fiber membrane-based pervaporation process, J. Membr. Sci. 319 (2008) 82–90. [38] F. Peng, Z. Jiang, C. Hu, Y. Wang, H. Xu, J. Liu, Removing benzene from aqueous solution using CMS-filled PDMS pervaporation membranes, Sep. Purif. Technol. 48 (2006) 229–234. [39] N. Rafia, A. Aroujalian, A. Raisi, Pervaporative aroma compounds recovery from lemon juice using poly(octylmethylsiloxane) membrane, J. Chem. Technol. Biotechnol. 86 (2011) 534–540. [40] A. Ravdel, A.M. Ponomarev, Kartkiispravochinkfiziko-khim, Velichin, Leningrad, Himia, 1983.
Contents lists available at ScienceDirect
Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep
Separation and purification of isobutanol from dilute aqueous solutions by a hybrid hydrophobic/hydrophilic pervaporation process Mostafa Omidali a , Ahmadreza Raisi a,b,∗ , Abdolreza Aroujalian a,b a
Department of Chemical Engineering, Amirkabir University of Technology (TehranPolytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran Food Process Engineering and Biotechnology Research Center, Amirkabir University of Technology (Tehran Polytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran b
a r t i c l e
i n f o
Article history: Received 31 July 2013 Received in revised form 8 December 2013 Accepted 11 January 2014 Available online 21 January 2014 Keywords: Pervaporation Isobutanol Hybrid process Composite membranes PDMS PVA
a b s t r a c t In this study, a hybrid hydrophobic/hydrophilic pervaporation process was employed to separate and purify isobutanol from its dilute aqueous solutions. For this purpose, composite polydimethylsiloxane membranes were initially used for the recovery of isobutanol by hydrophobic pervaporation. Then the hydrophilic pervaporation with a composite polyvinyl alcohol membrane was utilized to separate water from the organic phase of the permeate stream of the hydrophobic pervaporation. The effect of feed flow rate on the performance of pervaporation was investigated. The resistance in series model was also applied to calculate the transport resistances through the composite membranes. It was observed that an enhancement in the feed flow rate led to higher permeation flux and selectivity of the more permeable component, while the flux of the less permeable component was almost constant. Also, the ratio of liquid boundary layer resistance to membrane layer resistance decreased by an increase in the feed flow rate. The isobutanol with a purity of higher than 99 wt.% was produced by the hybrid hydrophobic/hydrophilic pervaporation technique from a 2 wt.% aqueous isobutanol solution. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Biofuels are a suitable replacement for fuels gets from sources are not renewable or their renewal period is too long. Depletion of natural or petroleum products and rising prices of raw materials result in a search for renewable energy sources and biofuels [1]. Among biofuels, bioethanol is widely produced and used in different industries. Biobutanol is another biofuel that has properties which are more attractive in comparison with bioethanol. Biobutanol as opposed to bioethanol is non-hygroscopic and corrosion, and has a higher calorific value due to its greater energy content. Also, butanol is combinable with gasoline in any composition and its mixture with gasoline has a lower vapor pressure in comparison with ethanol-gasoline that leads to less fuel vaporization and destruction [1]. Isobutanol is produced by the carbonylation of propylene. It is also produced naturally during the fermentation of carbohydrates [2]. Isobutanol can also be produced by some engineered microorganisms such as corynebacterium [3]. Isobutanol produced from the biological method has a low concentration
∗ Corresponding author at: Department of Chemical Engineering, Amirkabir University of Technology (TehranPolytechnic), Hafez Ave., P.O. Box 15875-4413, Tehran, Iran. Tel.: +98 21 64543125; fax: +98 21 66405847. E-mail address: [email protected] (A. Raisi). 0255-2701/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cep.2014.01.002
and needs to be concentrated and purified for use in different applications [1,4]. Separation processes such as distillation, liquidliquid extraction, adsorption, gas stripping and pervaporation have been used to separate alcohols from their aqueous solutions [5–7]. Compared to traditional processes, the pervaporation process has many advantages such as no heat damage to heat-sensitive compounds, low energy consumption, no additional separation treatment for added solvents or absorbents and minimum loss of alcohol. Therefore, pervaporation is an economical and useful technique for extraction of isobutanol from fermentation broth and its aqueous solutions [8,9]. Pervaporation is a membrane process for the separation of liquid mixtures by partial vaporization through a non-porous membrane. There are a large number of studies examining the performance of the pervaporation process for the separation of various alcohols such as methanol [10–12], ethanol [13–18], isopropanol [19–21], n-butanol [8,22–26] and isobutanol [22,27,28] from their mixtures with water. In these studies, a hydrophilic membrane was employed to separate water from the alcohol/water mixtures or a hydrophobic membrane was used to recover alcohol from its aqueous solutions, and further purification was not performed on the products of the pervaporation process. Furthermore, pervaporation-based hybrid processes have employed for the recovery and separation of alcohols from their aqueous solutions [29–31].
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
23
2. Theory List of symbols Nomenclature A constant in Eq. (13) a constant of Sherwood number B constant in Eq. (13) b constant of Sherwood number constant of Sherwood number c d constant of Sherwood number Di diffusion coefficient of component i into the membrane (m2 /s) dh hydraulic diameter of the membrane module (m) Ji flux of component i (g/m2 s) KL mass transfer coefficient of the boundary layer (m/s) L length of the membrane module (m) number of elements N Pi permeability coefficient of component i (cm3 cm/cm2 s cmHg) partial pressure of component i (cmHg) pi equilibrium vapor pressure of component i (cmHg) po,i permeate pressure (cmHg) pp Re Reynolds number RL mass transfer resistance of the liquid boundary layer (s/m) mass transfer resistance of the membrane layer RM (s/m) Rt total mass transfer resistance (s/m) S area of the membrane (m2 ) Sc Schmitt number Sherwood number Sh t time duration of the experiment (s) u feed velocity (m/s) weight of the collected permeate (g) W xi mole fraction of component i Greek letters ˛ selectivity membrane thickness (m) ıM C concentration difference (g/m3 ) kinematic viscosity (m2 /s) Subscripts and superscripts i component index f feed permeate p
The flux of component i varies linearly with the gradient in partial pressure according to [32]: Ji =
Pi (pf,i − pp,i ) ıM
(1)
where pf,i and pp,i are the partial pressure of component i in the feed mixture and vapor permeate, respectively, ıM is the membrane thickness and Pi is the permeability coefficient. The partial pressure of component i in the feed is: pf,i = i xf,i po,i
(2)
where i is the activity coefficient, xf,i is the mole fraction in the feed and po,i is the equilibrium vapor pressure of component i. According to Dalton’s law, the partial pressure of a component in the permeate can be expressed as: pp,i = xp,i pp
(3)
where xp,i is the mole fraction of the component in the permeate and pp is the permeate pressure. Due to a very low total permeate pressure, the partial pressure of the permeate is very small in comparison with the partial pressure of the feed, so it can be neglected. The flux can thus be assumed to vary linearly with the equilibrium vapor pressure of the pure component: Ji =
Pi i xf,i po,i ıM
(4)
Furthermore, according to resistance in series model, the transport of components from the feed solution through the composite membrane occurs by the following steps [33]: (i) diffusion through the liquid boundary layer, (ii) sorption into the membrane active layer, (iii) diffusion of liquid through the membrane active layer, (iv) desorption out of the active layer, and (v) transport of vapors through the porous support. Among these resistances, the liquid boundary layer and membrane active layer resistances control the mass transport in the pervaporation, and other transport resistances are low and negligible [14,16]. Therefore, under the steady state conditions, the flux of component i across the membrane can be expressed as a ratio of the driving force over the total mass transfer resistance (Rt ), as follows: Ji =
Ci Rt
(5)
where Rt = RL + RM In the present study, a hybrid hydrophobic/hydrophilic pervaporation process is used for the recovery and purification of isobutanol from aqueous solution. At first, isobutanol is separated from its dilute aqueous solutions through the hydrophobic pervaporation process with a composite polydimethylsiloxane membrane. Since isobutanol and water dissolve in each other only to a limited extent, the permeate stream of the hydrophobic pervaporation separates into two insoluble liquid phases, an isobutanol-rich phase and a water-rich one. The water-rich liquid phase is returned to the feed solution of the hydrophobic pervaporation process. Then the hydrophilic pervaporation with a composite polyvinyl alcohol membrane is utilized to separate water from the organic phase of the permeate stream of the hydrophobic pervaporation. Furthermore, the resistance in series model is applied to analyze the mass transport resistances and study the transport mechanisms in the composite membrane.
(6)
and RL = RM =
1 KL
(7)
ıM Di
(8)
where KL and Di are the mass transfer coefficient of boundary layer and diffusion coefficient of component i in the membrane, respectively. The Sherwood correlation has been used to determine the mass transfer coefficient (KL ) through the boundary layer [16,34]: Sh =
dh KL L = aReb Sc c D L
d (9)
The constants a, b, c and d for the Sherwood correlation are given in Table 1.
24
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
Table 1 Variables for the Sherwood correlation for plate and frame modules. Reynolds number range
a
b
c
d
Re < 2300 Re > 2300
1.615 0.026
0.33 0.80
0.33 0.30
0.33 –
By substitution of Eqs. (6)–(8) into Eq. (5), the following relation is obtained: Rt =
Ci 1 ıM = + Ji KL Di
(10)
Eq. (9) could be written as:
L 1 dh = Re−b Sc −c KL aD L
Fig. 1. Schematic of the pervaporation instrument used in this work.
−d (11)
or 1 = ARe−b KL
(12)
According to Table 1, coefficient A is independent of Reynolds number. A combination of Eqs. (10) and (12) give: Ci = ARe−b + B Ji
(13)
The coefficients A and B are determined from the slop and intercept of the (Ci /Ji ) versus Re−b plots, respectively. 3. Materials and methods Isobutanol (99.5 wt.%) that was used in the pervaporation experiments was purchased from Merck Co. Ltd. (Darmstadt, Germany) and de-ionized laboratory water was used for making aqueous mixtures. The composite PDMS (polydimethylsiloxane) membranes used in the experiments were kindly supplied by HelmholtzZentrum Geesthacht Zentrum fur Material und Kustenforschung GmbH (Geesthacht, Germany). The composite PVA (polyvinyl alcohol) (PERVAP® 2210) was purchased from Sulzer Chemtech Ltd. (Switzerland). All membranes used in the experiments are listed in Table 2. Separation and purification of isobutanol from its aqueous solutions is carried out in three steps as follows: 1. Pervaporative separation of isobutanol from dilute aqueous solutions by the composite PDMS membranes. 2. Physical phase separation of the permeate of the first step into two insoluble liquid phases (aqueous and organic phases) without any energy consumption. 3. Pervaporative purification of isobutanol from the organic phase of the second step by the composite PVA membrane. The membrane was cut into a 15 cm × 20 cm piece and held in a flat frame membrane module. The pervaporation apparatus has been previously described [35]. A schematic diagram of the experimental apparatus is shown in Fig. 1. The feed was kept in a 1 l reservoir and its temperature was controlled during the experiments with a precision of ±0.5 ◦ C. The feed was pumped from
Table 2 The hydrophobic and hydrophilic membranes used in the experiments. Membrane
Composite layer
Active layer
Active layer thickness (m)
PDMS 08/082-6 PDMS 08/082-11 PERVAP® 2210
PET/PAN/PDMS PET/PAN/PDMS PVA/PAN/PPS
PDMS PDMS PVA
2.3 12.9 1.8–2.8
the feed tank to a stainless steel SEPA CF pervaporation module (Osmonics Inc., Minnetonka, MN, USA). The retentate coming out of the membrane module was recycled to the feed tank. The vacuum was applied to the permeate side of the membrane by a vacuum pump (Busch Inc., Switzerland) in order to provide a pressure gradient across the membrane. The vacuum pressure was controlled by a needle valve and monitored by a vacuum barocel gauge with a precision of ±1 mmHg. The collected frozen permeate was melted and weighed. The isobutanol composition of the feed and permeate streams was determined using Gas Chromatography (Younglin 6000M Series Gas Chromatography System, Anyang Korea). The oven temperature was kept at 100 ◦ C for 7 min. Helium with a column head pressure of 0.7 bar was used as carrier gas. The injector and detector temperatures are 200 and 220 ◦ C, respectively. The injector is operated in the split mode (10:1). The separation performance of the pervaporation process was investigated by flux and selectivity. The total flux (J) was calculated using the following equation: J=
W S·t
(14)
where W is the weight of the collected permeate, S the area of membrane and t is the time duration of the experiments. Isobutanol selectivity (˛) was quantified as the separation factor based on: ˛=
p
p
f
f
xi /xj
(15)
xi /xj
where x is the mole fraction, i and j refer to the isobutanol and water, respectively, as well as f and p referring to the feed and permeate. In this work, initial recovery of isobutanol from the dilute aqueous solutions was carried out by the hydrophobic membrane and the effect of feed flow rate (0.35, 0.70, 2.35, 3.50 and 4.65 l/min) which corresponds to the Reynolds number (150, 300, 1000, 1500 and 2000) on the permeation flux and selectivity was studied for feed concentration of 2 wt.% with two composite PDMS membranes (active layer thicknesses of 2.3 and 12.9 m). Then isobutanol purification was performed with a hydrophilic membrane and the effect of feed flow rate (0.35, 0.70, 2.35, 3.50, 4.65 and 5.60 l/min) which corresponds to the Reynolds number (70, 140, 460, 690, 910 and 1100) at the feed concentration equal to water solubility in the isobutanol was investigated. The feed temperature and permeate side pressure was fixed at 30 ◦ C and 1 mmHg, respectively, for all experiments. Al1 experimental conditions were repeated three times and the average values are reported. The time duration of each experiment was 7 h and a permeate sample was collected every hour. Steady state permeation was reached after an hour at all experimental conditions.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a)
500
hydrodynamic conditions in the feed flow and can be calculated as follows:
140
total flux water flux isobutanol flux
120 200 110
100
(b)
0
1
2
3
4
5
100
Feedflowrate(lit/min) 300
100
total flux water flux isobutanol flux
250
Flux(g/m 2h)
Flux(g/m 2h)
300
0
Re =
130
90
200 80 150 70 100 60
50 0
Flux(g/m 2h)
Flux(g/m 2h)
400
0
1
2
3
4
5
50
Feedflowrate(lit/min)
(c)
30
Selectivity
25 20 15 10 2.3 µm
5
12.9 µm
0
0
1
2
3
4
25
5
Feedflowrate(lit/min) Fig. 2. Effect of feed flow rate on the total and partial fluxes for the PDMS membranes ((a) membrane thickness of 2.3 m and (b) membrane thickness of 12.9 m) and (c) effect of feed flow rate on the selectivity.
4. Results and discussion 4.1. Isobutanol concentration by hydrophobic membrane The effect of feed flow rate can be taken into account to investigate the performance of the pervaporation process. Fig. 2a and b shows the effect of feed flow rate which corresponds to the Reynolds number, on total and partial fluxes for isobutanol separation from 2 wt.% aqueous solution with composite PDMS membranes. The feed flow rate varies in the range of 0.35–4.65 l/min (Reynolds numbers of 150–2000). The Reynolds number (Re) is a well-known dimensionless description of the
udh
(16)
where u and are feed velocity and kinematic viscosity of solution, respectively and dh is the hydraulic diameter of the pervaporation module. Flow regimes into a channel are laminar, transient, and turbulent at Reynolds numbers of Re < 500, 500 < Re < 2000 and Re > 2000, respectively [36]. As shown in Fig. 2a and b, the water flux slightly increases with an increase in the feed flow rate, while the enhancement in the isobutanol flux is more significant. For example, when the feed flow rate changes from 0.35 to 4.65 l/min, the isobutanol flux goes from 56.1 to 63.3 g/m2 .h and the water flux varies from 129.2 to 131.5 g/m2 .h for the thicker PDMS membrane. Thus, for this feed flow rate range, the water flux which is controlled by the rate of diffusion through the membrane is independent of the feed flow rate. The change in the permeation flux of isobutanol by increasing the feed flow rate may be attributed to concentration polarization in the pervaporation process. The concentration polarization leads to a decrease in the permeation rate of the more permeable component, i.e. isobutanol [37]. An increase in the feed flow rate leads to a decline in the concentration polarization effect, and thus the isobutanol flux increases. In other words, when the feed flow rate enhances, the mass transfer coefficient of isobutanol in the solution goes up due to an increase of mixing and relative turbulency in the liquid boundary layer [13,23,35,38,39]. Furthermore, the variations in the isobutanol flux with increasing feed flow rate are more significant for the thinner membrane. These results can be related to the effect of membrane thickness on the membrane resistance and the ratio of liquid boundary to membrane resistance. As the membrane thickness decreases, the membrane resistance reduces and consequently the ratio of liquid boundary to membrane resistance increases. Therefore, the effect of feed flow rate will be more significant for the thinner membrane as can be seen in Fig. 2. To determine the mass transport resistances, the value of (Ci /Ji ) were plotted Re−0.33 and the results are presented in Fig. 3a and b for isobutanol and water, respectively. For different membrane thickness, the liquid boundary layer and membrane layer resistances were calculated from the slope and intercept of each line, respectively. The effect of feed flow rate on the liquid boundary layer and membrane layer resistances of isobutanol and water are indicated in Fig. 4. Also, the ratio of boundary layer resistance and membrane layer resistance to total resistance for both components is presented in Table 3. Fig. 4 shows that an increase in the feed flow rate leads to a decrease in the boundary layer resistance while membrane layer resistance remains constant. Therefore, the ratio of liquid boundary layer resistance to membrane layer resistance declines by an enhancement in the feed flow rate as can be seen from Table 3. The influence of the feed flow rate on the isobutanol selectivity is represented in Fig. 2c. As indicated, the isobutanol selectivity increases as the feed flow rate goes to higher levels. The selectivity can be related to the ratio of isobutanol to water flux. The isobutanol flux enhances with an increase in the feed flow rate while the water flux is almost constant. Therefore, the ratio of the two fluxes which is proportional to the isobutanol selectivity increases when the feed flow rate enhances [37]. This trend is same as the works investigated the effect of feed flow rate on selectivity of other components through the different membranes [13,35,38,39]. The membrane permeability and permeance (the ratio of permeability and membrane thickness) have been defined and used to represent the membrane property. It is a good way to
26
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
Table 3 The ratio of liquid boundary layer and membrane layer resistances to total resistance against isobutanol and water permeation through the PDMS membranes.
RL /Rt
Component
Thickness (m)
0.35
0.70
2.35
3.50
4.65
Isobutanol
2.3 12.9 2.3 12.9
0.2392 0.1952 0.0673 0.0288
0.2001 0.1617 0.0542 0.0230
0.1436 0.1145 0.0370 0.0156
0.1282 0.1019 0.0326 0.0137
0.1181 0.0936 0.0298 0.0125
2.3 12.9 2.3 12.9
0.7608 0.8048 0.9327 0.9712
0.7999 0.8383 0.9458 0.9770
0.8564 0.8855 0.9630 0.9844
0.8718 0.8981 0.9674 0.9863
0.8819 0.9064 0.9702 0.9875
Water
RM /Rt
Isobutanol Water
Feed flow rate (l/min)
Table 4 The permeance coefficient (cm3 /cm2 s cmHg) of isobutanol and water through the PDMS membrane at 30 ◦ C. Component
Membrane thickness (m)
Feed flow rate (l/min) 0.35
0.70
2.35
3.50
4.65
Isobutanol
2.3 12.9
5.81 × 104 3.16 × 104
6.09 × 104 3.34 × 104
6.46 × 104 3.41 × 104
6.64 × 104 3.47 × 104
6.74 × 104 3.57 × 104
Water
2.3 12.9
3.74 × 104 1.82 × 104
3.74 × 104 1.82 × 104
3.79 × 104 1.83 × 104
3.82 × 104 1.83 × 104
3.82 × 104 1.83 × 104
(a) 1.5
values confirmed the results in Fig. 4 and Table 3. It can be seen from Table 4 that the permeance coefficient of the ethanol increases with an enhancement in the feed flow rate. This implies that an increase in the feed flow rate leads to lower liquid boundary layer resistance and consequently the permeance coefficient enhances. Also,
(a) R² = 0.9648
0.9
0.6
RM × 10-5 (s/m)
ΔC/Ji (s/m)
1.2
R² = 0.9913
0.3
2.3 µm
6
3.0
5
2.5
4
2.0
3
1.5
2
1.0
1
0.0
0.00
0.05
0.10
0.15
0.20
0
0.25
0.5
RM
12.9 µm
RL
0
1
Re-0.33
(b) 3.0
RL × 10-5 (s/m)
characterize membrane and can be used as a criterion for membrane selection and material design. The overall permeance coefficients of the studied membranes were calculated according to Eq. (4), the results are presented in Table 4. The permeance
2
3
4
0.0 5
Feed low rate (lit/min)
(b) 1.5
15
1.2
12
0.9
9
0.6
6
2.0
RM × 10-6 (s/m)
ΔC/Ji (s/m)
R² = 0.9679
1.5 R² = 0.9824
1.0 0.5 0.0
3
0.3
2.3 µm
RM
12.9 µm 0.00
0.05
RL× 10-5 (s/m)
2.5
0.10
0.15
0.20
0.25
Re-0.33 Fig. 3. Relationship between (Ci /Ji ) and Re−0.33 for isobutanol (a) and water (b) permeation.
0
RL
0
1
2
3
4
0
5
Feed low rate (lit/min) Fig. 4. Effect of feed flow rate on both liquid boundary layer and membrane layer resistances of isobutanol (a) and water (b) for membrane thickness of 2.3 m.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a)
27
6
150 R² = 0.9800
5
ΔC/Ji (s/m)
Total lux (g/h.m2)
120
90
4 3 2
60 1
Total flux
30
water 0 0.00
water flux isobutanol flux 0
0
1
2
3
4
5
6
(b) 15
Water selectivity
12
9
6
3
0
1
2
3
4
5
0.05
0.10
0.15
0.20
0.25
0.30
Re-0.33
Feed low rate (lit/min)
0
R² = 0.9626
isobutanol
6
Feed low rate (lit/min) Fig. 5. Effect of feed flow rate on the total (a) and partial fluxes (b) as well as the water selectivity (c) for the PVA membrane.
the permeance coefficient for the thicker membranes was lower due to higher membrane layer resistance. 4.2. Isobutanol purification by a hydrophilic membrane The product of the first step of the concentration process separates into an organic phase and an aqueous one. The water content of the organic phase is equal to the maximum solubility of water in isobutanol, i.e. 10 wt.% at 30 ◦ C. The aqueous phase has about 6 wt.% isobutanol which can return to the feed solution of the hydrophobic pervaporation process. A hydrophilic composite PVA membrane was employed to separate water from the isobutanol-rich solution and to purify the isobutanol. Fig. 5a shows the effect of feed flow rate on the permeation flux for separation of water from the isobutanol/water mixtures. It is observed that both total and water flux increase with enhancement of feed flow rate, while the isobutanol partial flux approximately remains constant. The effect of
Fig. 6. Relationship between (Ci /Ji ) and Re−0.33 for component permeation through the PVA membrane.
feed flow rate on the permeation flux of the PVA membrane may be related to the concentration polarization in the pervaporation process, as discussed in details in the previous section. It is quite evident that a higher feed flow rate will promote the flux because it could lessen the concentration polarization on the membrane surface and provide more opportunity for the sorption of water on the hydrophilic membrane. The influence of feed flow rate on the water selectivity is depicted in Fig. 5b. It can be seen that the selectivity slightly enhances from 9.3 to 11.9 when the feed flow rate varies from 0.35 to 5.60 l/min. When the feed flow rate enhances, the water flux increases while the permeation flux of isobutanol remains constant which leads to increase in the water selectivity. Similar observations were reported in the other researches [13,26,35,38,39]. For example, Srinivasan et al. [26] for the pervaporative recovery of 1-butanol from a model pharmaceutical aqueous waste observed that the separation factor of 1-butanol increases from 4.9 to 6.9 while the feed flow rate enhances from 0.4 to 0.8 l/min. Also, Rafia et al. [39] in the pervaporation experiments for aroma recovery from lemon juice reported that the enrichment factor of -pinene and limonene increases from 16.3 to 16.8 and from 16.9 to 17.4, respectively as the Reynolds number varies from 500 to 2300. Fig. 6 represents the linear relationship between (Ci /Ji ) and Re−0.33 and (Eq. (13)) for the water and isobutanol permeation through the PVA membrane. The effect of feed flow rate on the liquid boundary layer and membrane resistances is shown in Fig. 7. Also, the calculated values of the ratio of boundary layer and membrane layer resistances to total resistance for water and isobutanol are given in Table 5. The results indicate that by increasing the feed flow rate, the liquid boundary layer resistance decreases while the membrane layer resistance does not vary. Also, the membrane resistance for water is lower than isobutanol due to the small ˚ in comparison with the isobutanol molecular size of water (2.65 A) ˚ that simplifies diffusion of water molecules molecular size (7.70 A) [40]. As a result, Tables 3 and 5 indicate that the liquid boundary
Table 5 The ratio of liquid boundary layer and membrane layer resistances to total resistance against water and isobutanol permeation through the PVA membrane. Component
Feed flow rate (l/min) 0.35
0.70
2.35
3.50
4.65
5.60
RL /Rt
Water Isobutanol
0.4600 0.1774
0.4039 0.1464
0.3124 0.1031
0.2849 0.0916
0.2662 0.0841
0.2544 0.0795
RM /Rt
Water Isobutanol
0.5400 0.8226
0.5961 0.8536
0.6876 0.8969
0.7151 0.9084
0.7338 0.9159
0.7456 0.9205
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29
(a) 4
with a purity of 99.7 wt.% was obtained from its dilute aqueous solutions by the proposed hybrid method.
Resistance × 10-6 (s/m)
28
5. Conclusions
3
2
1 RM 0
RL 1
0
2
3
4
5
6
15
4
12
3
9
2
6
1
3
RM × 10-7 (s/m)
(b) 5
RL × 10-6 (s/m)
Feed low rate (lit/min)
RM 0
RL 0
1
2
3
4
5
6
0
Separation and purification of isobutanol from its dilute aqueous solutions were performed using a hybrid pervaporation process with hydrophobic and hydrophilic composite membranes. The effect of feed flow rate on the performance of pervaporation was investigated. The proposed pervaporation technique enables to concentrate isobutanol from 2 wt.% to 99.7 wt.% The experimental results showed that for the PDMS and PVA composite membranes, an increase in the feed flow rate led to higher permeation flux and selectivity of the more permeable component, while the flux of the less permeable component was almost constant. The resistance in series model was used to analyze the mass transport resistances and the results indicated that by enhancement in the feed flow rate, the liquid boundary layer resistance of components declined while the membrane layer resistance remained constant. Also, results of modeling indicated that the effect of liquid boundary layer resistance is more significant for the thinner PDMS membrane. It can be concluded that the pervaporation process is an attractive technology for the recovery and purification of isobutanol from its aqueous solution as it yields good separation and operates under mild conditions. Acknowledgement The authors would like to thank Helmholtz-Zentrum Geesthacht Zentrum fur Material und Kustenforschung GmbH (Geesthacht, Germany) for supplying the PDMS membranes.
Feed low rate (lit/min) Fig. 7. Effect of feed flow rate on both liquid boundary layer and membrane layer resistances of water (a) and isobutanol (b).
layer resistance has an obvious role for dilute components permeation through the membrane. The main purpose of this study was the purification of isobutanol from its dilute aqueous solutions. The change in the isobutanol concentration during the hybrid purification process is shown in Fig. 8. It can be seen that the isobutanol was concentrated to a 34.4 wt.% solution from a 2 wt.% isobutanol aqueous solution by the hydrophobic pervaporation process (step 1) at 30 ◦ C. Then the isobutanol concentration increased from 34.4 to 90 wt.% by the physical phase separation (step 2). Finally, the isobutanol was purified to a 99.7 wt.% concentration by the hydrophilic pervaporation process with the PVA membrane (step 3). Therefore, the isobutanol
Fig. 8. The change in the isobutanol concentration during the hybrid purification process.
References [1] W. Kaminski, E. Tomczak, A. Gorak, Biobutanol production and purification methods, Ecol. Chem. Eng. S 18 (2011) 31–37. [2] S. Atsumi, T. Hanai, J.C. Liao, Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels, Nature 451 (2008) 86–89. [3] S.K. Michael, C.K. Myung, C. James, Engineering corynebacteriumglutamicum for isobutanol production, Appl. Microb. Biotechnol. 87 (2010) 1045–1055. [4] N. Qureshi, I.S. Maddox, Continuous production of acetone–butanol–ethanol using immobilized cells of Clostridium acetobutylicum and integration with product removal by liquid–liquid extraction, J. Ferment. Bioeng. 80 (1995) 185–189. [5] W.J. Groot, R. Lans, K. Luyben, Technologies for butanol recovery integrated with fermentations, Proc. Biochem. 27 (1992) 61–75. [6] L.M. Vane, Separation technologies for the recovery and dehydration of alcohols from fermentation broths, Biofuels Bioprod. Bioref. 2 (2008) 553–588. [7] A. Ranjan, V.S. Moholkar, Biobutanol: science, engineering, and economics, Int. J. Energy Res. 36 (2012) 277–323. [8] L.M. Vane, A review of pervaporation for product recovery from biomass fermentation processes, J. Chem. Technol. Biotechnol. 80 (2005) 603–629. [9] K.L. Wasewar, V.G. Pangarkar, Intensification of recovery of ethanol from fermentation broth using pervaporation: economical evaluation, Chem. Biochem. Eng. Q. 20 (2006) 135–145. [10] S.P. Doguparthy, Pervaporation of aqueous alcohol mixtures through a photopolymerised composite membrane, J. Membr. Sci. 185 (2001) 201–205. [11] X. Liu, Y. Sun, X. Deng, Studies on the pervaporation membrane of permeation water from methanol/water mixture, J. Membr. Sci. 325 (2008) 192–198. [12] B. Gonzalez, I. Ortiz, Modeling and simulation of a hybrid process (pervaporation–distillation) for the separation of azeotropic mixtures of alcohol–ether, J. Chem. Technol. Biotechnol. 77 (2011) 29–42. [13] X. Li, S. Wang, Some characteristics of pervaporation for dilute ethanol–water mixtures by alcohol-permselective composite membrane, Sep. Sci. Technol. 31 (1996) 2867–2873. [14] R. Jiraratananon, A. Chanachai, R.Y.M. Huang, D. Uttapap, Pervaporation dehydration of ethanol–water mixtures with chitosan/hydroxyethylcellulose (CS/HEC) composite membranes: I. Effect of operating conditions, J. Membr. Sci. 195 (2002) 143–151. [15] M. Garcia, M.T. Sanz, S. Beltran, Separation by pervaporation of ethanol from aqueous solutions and effect of other components present in fermentation broths, J. Chem. Technol. Biotechnol. 84 (2009) 1873–1882. [16] A. Raisi, A. Aroujalian, T. Kaghazchi, Experimental study and mass transport modeling of ethanol separation from aqueous solutions by pervaporation, Sep. Sci. Technol. 44 (2009) 3538–3570.
M. Omidali et al. / Chemical Engineering and Processing 77 (2014) 22–29 [17] B. Bolto, M. Hoang, Z. Xie, A review of membrane selection for the dehydration of aqueous ethanol by pervaporation, Chem. Eng. Process. Process Intensif. 50 (2011) 227–235. [18] C.H. Kang, Y.F. Lin, Y.S. Huang, K.L. Tung, K.S. Chang, J.T. Chen, W.S. Hung, K.R. Lee, J.Y. Lai, Synthesis of ZIF-7/chitosan mixed-matrix membranes with improved separation performance of water/ethanol mixtures, J. Membr. Sci. 438 (2013) 105–111. [19] J. Yu, C.H. Lee, W.H. Hong, Performances of crosslinked asymmetric poly(vinyl alcohol) membranes for isopropanol dehydration by pervaporation, Chem. Eng. Process. Process Intensif. 41 (2002) 693–698. [20] G. Dhanuja, B. Smitha, S. Sridhar, Pervaporation of isopropanol–water mixtures through polyion complex membranes, Sep. Purif. Technol. 44 (2005) 130–138. [21] N.R. Singha, S. Kar, S. Ray, S.K. Ray, Separation of isopropyl alcohol-water mixtures by pervaporation using crosslink IPN membranes, Chem. Eng. Process. Process Intensif. 48 (2009) 1020–1029. [22] K.W. Buddeker, G. Bengtson, H. Pingel, Pervaporation of isomeric butanols, J. Membr. Sci. 54 (1990) 1–12. [23] E. Favre, Q.T. Nguyen, Extraction of 1-butanol from aqueous solutions by pervaporation, J. Chem. Technol. Biotechnol. 65 (1996) 221–228. [24] A. Jonquieres, A. Fane, Filled and unfilled composite GFT PDMS membranes for the recovery of butanols from dilute aqueous solutions: influence of alcohol polarity, J. Membr. Sci. 125 (1997) 245–255. [25] N. Qureshi, I.S. Maddox, Reduction in butanol inhibition by perstraction: utilization of concentrated lactose/whey permeate by clostridium acetobutylicum to enhance butanol fermentation economics, Food Bioprod. Process. 83 (2005) 43–52. [26] K. Srinivasan, K. Palanivelu, A.N. Gopalakrishnan, Recovery of 1-butanol from a model pharmaceutical aqueous waste by pervaporation, Chem. Eng. Sci. 62 (2007) 2905–2914. [27] S. Korkmaz, Y. Salt, S. Dincer, Esterification of acetic acid and isobutanol in a pervaporation membrane reactor using different membranes, Ind. Eng. Chem. Res. 50 (2011) 11657–11666.
29
[28] M. Fathizadeh, A. Aroujalian, A. Raisi, M. Fotouhi, Preparation and characterization of thin film nanocomposite membrane for pervaporative dehydration of aqueous alcohol solutions, Desalination 314 (2013) 20–27. [29] F. Lipnizki, R.W. Field, P.K. Ten, Pervaporation-based hybrid process: a review of process design, applications and economics, J. Membr. Sci. 153 (1999) 183–210. [30] Z. Szitkai, Z. Lelkes, E. Rev, Z. Fonyo, Optimization of hybrid ethanol dehydration systems, Chem. Eng. Process. Process Intensif. 41 (2002) 631–646. [31] J.B. Haelssig, J. Thibault, A.Y. Tremblay, Numerical investigation of membrane dephlegmation: a hybrid pervaporation-distillation process for ethanol recovery, Chem. Eng. Process. Process Intensif. 50 (2011) 1226–1236. [32] A. Raisi, A. Aroujalian, T. Kaghazchi, Multicomponent pervaporation process for volatile aroma compounds recovery from pomegranate juice, J. Membr. Sci. 322 (2008) 339–348. [33] M.G. Liu, J.M. Dickson, P. Cote, Simulation of a pervaporation system on the industrial scale for water treatment, Part I: extended resistance-in-series model, J. Membr. Sci. 111 (1996) 227–241. [34] E.L. Cussler, Diffusion, Cambridge University Press, New York, 1989. [35] A. Aroujalian, A. Raisi, Recovery of volatile aroma components from orange juice by pervaporation, J. Membr. Sci. 303 (2007) 154–161. [36] V.L. Streeter, E.B. Wylie, Fluid Mechanics, sixth ed., McGraw-Hill, New York, 1975. [37] G.O. Yahaya, Separation of volatile organic compounds (BTEX) from aqueous solutions by a composite organophilic hollow fiber membrane-based pervaporation process, J. Membr. Sci. 319 (2008) 82–90. [38] F. Peng, Z. Jiang, C. Hu, Y. Wang, H. Xu, J. Liu, Removing benzene from aqueous solution using CMS-filled PDMS pervaporation membranes, Sep. Purif. Technol. 48 (2006) 229–234. [39] N. Rafia, A. Aroujalian, A. Raisi, Pervaporative aroma compounds recovery from lemon juice using poly(octylmethylsiloxane) membrane, J. Chem. Technol. Biotechnol. 86 (2011) 534–540. [40] A. Ravdel, A.M. Ponomarev, Kartkiispravochinkfiziko-khim, Velichin, Leningrad, Himia, 1983.