Performance of commercial composite hydrophobic membranes applied for pervaporative reclamation of acetone, butanol, and ethanol from aqueous solutions: Binary mixtures

Accepted Manuscript Performance of commercial composite hydrophobic membranes applied forpervaporative reclamation of acetone, butanol, and ethanol from aqueous solutions: binary mixtures Katarzyna Knozowska, Anna Kujawska, Joanna Kujawa, Wojciech Kujawski, Marek Bryjak, Ewelina Chrzanowska, Jan K. Kujawski PII: DOI: Reference:

S1383-5866(17)31935-4 http://dx.doi.org/10.1016/j.seppur.2017.07.072 SEPPUR 13929

To appear in:

Separation and Purification Technology

Received Date: Revised Date: Accepted Date:

17 June 2017 26 July 2017 26 July 2017

Please cite this article as: K. Knozowska, A. Kujawska, J. Kujawa, W. Kujawski, M. Bryjak, E. Chrzanowska, J.K. Kujawski, Performance of commercial composite hydrophobic membranes applied forpervaporative reclamation of acetone, butanol, and ethanol from aqueous solutions: binary mixtures, Separation and Purification Technology (2017), doi: http://dx.doi.org/10.1016/j.seppur.2017.07.072

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Performance of commercial composite hydrophobic membranes applied for pervaporative reclamation of acetone, butanol, and ethanol from aqueous solutions: binary mixtures

Katarzyna Knozowska1, Anna Kujawska1, Joanna Kujawa1, Wojciech Kujawski1,*, Marek Bryjak2, Ewelina Chrzanowska1, Jan K. Kujawski2

1

Nicolaus Copernicus University in Toruń, Faculty of Chemistry, 7 Gagarina Street, 87-100 Toruń, Poland

2

Wrocław University of Technology, Chemical Faculty, 27 Wybrzeże Wyspiańskiego, 50-370 Wrocław, Poland

*

Corresponding author: [email protected] (W. Kujawski), phone: +48566114315

Abstract In this study the efficiency of three various commercial membranes based on poly(octylmethyl siloxane), poly(ether-block-amide), and poly(dimethylsiloxane) polymers were investigated in pervaporative separation of acetone, butanol, and ethanol from aqueous binary solutions (0-5 wt. % of organics) at 60oC. The influence of fermentation broth microfiltration on membrane performances in pervaporative removal of ethanol was also investigated. The use of microfiltration improved the effectiveness of the removal of ethanol from the broth comparing with the unfiltered one. Molar ratios of organics to water fluxes and Pervaporative Separation Index (PSI) values were employed to discuss membranes' performance in removal of organic solvents from binary aqueous mixture. It was found that PDMS based (Pervap4060) membrane shows the best separation efficiency in all tested binary aqueous mixtures. Modelling of the process proved the feasibility of pervaporation process for the removal of acetone, butanol and ethanol from binary and quaternary aqueous mixtures.

1

Keywords pervaporation; PDMS, POMS, and PEBA based membranes; acetone, butanol, and ethanol aqueous binary solutions; ethanol fermentation broth microfiltration; modelling of pervaporation process; Abbreviations A – membrane area [m2] ABE – acetone-butanol-ethanol AcO - acetone BuOH – butanol Dr – degree of removal [%] EtOH – ethanol FDn – flux decline [%] H2O – water HSP – Hansen Solubility Parameters Ji – partial fluxes of component i [g m-2 h-1] Jt – total flux [g m-2 h-1] LOD – limit of detection LOQ – limit of quantification MF – microfiltration mi – mass of component [g] MWCO – molecular weight cut off [Da] NTU - Nephelometric Turbidity Unit NTUf – feed turbidity [NTU] NTUp - permeate turbidity [NTU] PDMS – poly(dimethylsiloxane) PEBA – poly(ether-block-amide) POMS – poly(octylmethyl siloxane) PP – polypropylene PSI – Pervaporation Separation Index [g m-2 h-1] PTFE – polytetrafluoroethylene 2

PV – pervaporation PVDF - polyvinylidene fluoride R – rejection coefficient [%] RO - reserve osmosis SW – swelling degree [%] T – temperature [C] t - time of permeate collection [h] TCD – thermal conductivity detector TiO2 - titanium dioxide TPV - thermopervaporation Wd – mass of dry membrane [g] Ws – mass of swollen membrane [g] xi – weight fraction of component i in feed [g/g] xw – weight fraction of water in feed [g/g] Yi – molar fraction of component i [mol/mol] yi – weight fraction of component i in permeate [g/g] yw – weight fraction of water in permeate [g/g] β – separation factor [-] δd - dispersion interactions [MPa1/2] δh - hydrogen bonding interactions [MPa1/2] δp - polar interactions [MPa1/2]

 – distance parameter [MPa1/2]

1. Introduction Pervaporation (PV) is an effective membrane process applied to separate components of liquid mixtures using dense lyophilic membranes [1, 2]. The driving force for PV process is generated by a partial pressure difference between both sides of the membrane [3]. Important benefits of pervaporation, concerning energy requirements and uncomplicated control of the

3

process allowed for a broad utilization of the process, e.g. for dehydration of organic solvents, for separation of organic-organic mixtures and for removal of volatile organic compounds from water as well as fermentation broth [3-6]. The removal of volatile organic compounds from water is viable from an application point of view [3-8]. The rapidly growing demand for energy, observed during last decades contributed to the reduced necessity of fossil fuels by supporting renewable energy resources. Fossil fuels are dominating (∼80%) in energy economy, though biofuels are an interesting alternative [9-11]. Particularly, in the petrochemical industry a lot of attention is paid to bioalcohols, e.g. butanols and ethanol. Biobutanol (BuOH) possesses several advantages comparing with bioethanol, such as lower volatility/flammability, higher energy density, lower hydrophilicity, lower corrosiveness, and compatibility with current gasoline supply infrastructure [12-14]. However, production of biobutanol via fermentation is limited by the low final concentration of BuOH (ca. 13 g L -1 BuOH in the traditional acetone-butanolethanol /ABE/ fermentation process), causing high cost of butanol production if distillation process is applied for the butanol recovery [15-18]. An interesting solution in that case would be utilization of pervaporation or thermopervaporation (TPV) processes, due to their potentials for energy efficiency. Various types of hydrophobic membranes can be used for the pervaporation removal of butanol, ethanol, and acetone (products of ABE fermentation), e.g. poly(dimethylsiloxane) [5, 6, 15, 19-24], poly(ether- block-amide) (PEBA) [25] and poly[1(trimethylsilyl)-1-propyne) (PTMSP) [26-29], polyoctylmethylsiloxane (POMS) [30], polyvinylidene fluoride (PVDF) [31], and hydrohexafluoroisopropylnorbonene block and random copolymers [16]. Borisov et al. [32] proposed novel hybrid process based on thermopervaporation for removal of butanol from model aqueous solution and real fermentation broth. The novel hybrid process uses water rich phase of permeate as a cooling agent. Authors tested

4

commercial membranes like Pervap4060, Pervatech, PolyAn, and MDK-3 membranes. MDK3 membrane showed the highest PSI value in the separation of ABE-water mixtures by the standard thermopervaporation [32]. For that reason this membrane was tested in the novel hybrid process with ABE model solution and real fermentation broth. In the case of model solution MDK-3 membrane demonstrated total flux and separation factor values equal to 0.6 kg/m-2 h-1 and 6, respectively. During 50 h of experiment with fermentation broth around 80% of organic compound were removed from feed solution [32]. Authors performed also IR analysis of MDK-3 membrane before and after TPV experiment with fermentation broth. Analysis showed a high deposition of organics on the membrane surface, however this deposition did not affect the membrane performance [32]. Rom et al. [30] investigated properties of commercial POMS membrane in pervaporation with filtered ABE fermentation broth. POMS membrane was selective towards water, acetone, butanol, and ethanol. Other organic compounds were not found in permeate. However lower organics fluxes were observed during separation of the filtered broth when comparing with model quaternary ABE-water solution. It was suggested that reduction of organics fluxes may result from the presence of lipids in feed solution, which are soluble in alcohols [33]. On the other hands Rom et al. [30] noticed higher separation factor in a contact with fermentation broth. This trend may result from the presence of inorganic salts and the salting-out effect [34-36]. Research proved that POMS membrane can be effectively applied for the separation of ABE-water mixture. Chovau [37] investigated influence of fermentation by-products on the separation efficiency of commercial PDMS based membrane (Pervatech and Pervap4060). To define an influence of fermentation by-products, authors first performed pervaporation with model water-ethanol mixture. Pervaporation experiments showed that Pervap4060 membrane possessed better efficiency in the recovery of ethanol from feed mixture. The highest

5

influence of fermentation by-products was observed during pervaporation with an addition of 0.1 wt. % 2,3-butanediol. Mulder and co-workers [38] tested contribution of pervaporation and ultrafiltration processes in the continuous fermentation and fermentative products removal. It was shown that ethanol removal from the ultrafiltration permeate can be realized via pervaporation using homogeneous, composite or asymmetric membranes. Fermentation broth should not be fed directly to the pervaporation process because feed solution contains particles or suspensions and a pre-treatment step is needed. Mulder et al. [38] proved that ethanol-water mixture can be removed by pervaporation and that the rejection of suspended solids is complete when beer from the fermenter is ultra-filtrated. Ceramic membranes are suitable for ultra- and microfiltration process, especially if the process is realized under harsh pH conditions [38]. Hence, the process can be carried out continuously and the recovered volatile organic compounds such as ethanol, acetone, butanol, and 2-propanol can be utilized within other processes. In the real applications, a microfiltration membrane was utilized prior to PV avoid the fouling of the hydrophobic pervaporation membrane [38]. The separation of products from fermentation broth is one of most important factor affecting production, because downstream processing often contributes in 30 - 40% to the total process cost. Several separation technologies, including extraction, adsorption, pressuredriven

membrane

processes

(e.g.

microfiltration,

nanofiltration,

reverse

osmosis,

pervaporation, and electrodialysis) are available for separation and purification of the target products from the various feedstocks. Among these various separation techniques, membrane separation technology has attracted an increasing attention for several reasons: (i) it is a costeffective and environmentally friendly; (ii) it improves separation efficiency and product quality by minimizing product inhibition because the target product is removed in situ from

6

the fermentation broth; and (iii) it can be flexibly integrated with other separation unit operations [39]. The aim of this research was to investigate and to compare the efficiency of various commercial membranes based on different polymers: poly(dimethylosiloxane) - PDMS, poly(ether-block-amide) - PEBA, and poly(octylmethyl siloxane) - POMS in the removal of acetone, butanol, and ethanol (ABE) from binary aqueous solutions. An important part of this research was an assessment of performance of PDMS based membrane in the separation of ethanol from unfiltered and filtered fermented broths. Moreover, the batch pervaporation process was modelled using an approach based on the mass balance in the system.

2. Material and methods 2.1. Materials Commercially available hydrophobic membranes based on poly(dimethylsiloxane) Pervap4060 (provided by Sulzer Chemtech, Switzerland), poly(octylmethyl siloxane) – POMS, and poly(ether-block-amide) - PEBAX (both provided by Pervatech, the Netherlands) were utilized. Acetone, 1-butanol, ethanol, and 1-propanol were purchased from Avantor Performance Materials Poland S.A. (Poland) and used as received. Aqueous solutions were prepared by using reserve osmosis (RO) water (15 MΩ). Samples of fermentation broth were kindly provided by a local distillery.

2.2. Pervaporation experiments Pervaporation experiments were performed at 60°C using a standard laboratory rig, schematically presented in Fig. 1 and described in detail elsewhere [3]. The membrane samples with an effective area of 170 cm2 were utilized in all experiments. Pervaporation experiments were performed for membranes in contact with pure water, water-acetone, water-

7

butanol, and water-ethanol binary mixtures, using feed mixtures with concentration of organic compounds in the range 0-5 wt. %, and additionally for filtered and unfiltered fermentation broths (within the concentration of 2-7 wt. % of ethanol).

Fig. 1. Pervaporation rig: (1) thermostated and closed stirred feed vessel, (2) recirculation pump, (3) pervaporation module, (4) permeate trap, (5) vacuum pump.

Several factors were employed to assess the separation efficiency and transport properties, i.e. total and partial fluxes (Jt and Ji) – Eqs. (1) and (2), separation coefficient (β) – Eq. (3), and Pervaporation Separation Index (PSI) – Eq. (4) . Total flux was determined according to Eq. (1) [40, 41]:

J

t



Δm A  Δt

(1)

where m is a mass of permeate (g) collected over Δt period of time (h), A is a membrane area (m2). Partial flux of component i was calculated according to Eq. (2):

J i  J t  yi

(2)

where yi is a weight fraction of component i in the permeate. Separation coefficient β is used to describe the membrane separation efficiency, according to Eq. (3) [40]:

8



y y x x i

w

i

w

(3)

where yi and yw are weight molar fractions of component i, and water in permeate and xi, x w are weight molar fractions of component i and water in feed, respectively. Pervaporation Separation Index (PSI) (Eq. (4)) was used to compare efficiency of various membranes during the separation of the same feed mixture [12]:

PSI  Jt (  1)

(4)

Hansen’s Solubility Parameters (HSP) can be also applied for predicting the affinity of a given solvent towards polymer selective layer during the sorption stage of pervaporation. Sorption is one of the steps in the solution-diffusion mechanism used to describe transport and separation in pervaporation process [42]. The similarity of a given solvent to a polymer material can be evaluated by calculating a distance parameter according to Eq. (5) [3, 43]. It was proved that the low values of distance parameters indicate the strong interactions between polymer and solvent [43].



 

 

ΔS,P  δd,sδd,P  δp,Sδp,P  δh,Sδh,P 2

2



2

(5)

The calculated values of the distance parameters are gathered in Tab. 1 and subsequently they will be used in the discussion of the obtained results.

Tab. 1. The Hansen’s solubility parameters for the tested substances and membranes [3, 44] and distance parameter (ΔS,Polymer) calculated according to Eq. (5). δd δp δh Δs-PDMS Δs-POMS Δs-PEBA Solvent (MPa0.5) (MPa0.5) (MPa0.5) (MPa0.5) (MPa0.5) (MPa0.5) PDMS 15.9 0.0 4.1 POMS 14.8 5.6 6.5 PEBA 17.6 7.6 6.8 water 15.5 16.0 42.3 41.4 37.3 36.5 butanol 16.0 5.7 15.8 13.0 9.4 9.3 acetone 15.5 10.4 7.0 10.8 4.9 3.2 ethanol 16.0 5.7 15.8 13.0 13.4 12.7

9

2.3. Microfiltration Microfiltration experiments were realized using an experimental rig provided by Intermasz Membrane Filtration Ltd (Września, Poland) (Fig. 2). The setup was equipped with TiO2 tubular ceramic membrane (MWCO = 300 kDa). The membrane surface active area was equal to 25.2 cm2. The feed was transported to the module from the thermostated tank and retentate was recirculated back to the feed tank. The flux of permeate was determined by a gravimetric method using analytical electronic balance. At the outset of experiments water was used as feed to evaluate the reference flux for the given membrane. In the subsequent step, fermentation broth was used as feed solution. Microfiltration process was realized at temperature of 30C with a transmembrane pressure equal to 2 bar. The flow rate of feed solution over the membrane surface was kept at the constant rate of 75 L h -1. The filtration efficiency was assessed by nephelometric measurement of turbidity. Nephelometric measurements were performed using Turbimeter TN-100 (Eurotech Instrument, Singapore). The rejection coefficient (R) of filtration process was calculated according to the Eq. (6): R  (1 

NTU p ) 100% NTU f

(6)

where NTUp is permeate turbidity and NTUf is feed turbidity.

Fig. 2. Microfiltration experimental rig: 1 – thermostated feed tank, 2 – circulating pump, 3 – manometer, 4 – membrane module, 5 – beaker and balance, 6 – rotameter.

10

2.4. Swelling experiment Swelling experiments were performed using small samples of POMS, PEBAX, and Pervap4060 membranes, following the procedure proposed by Niemistö et al. [45]. Prior to swelling experiments, samples were dried in a vacuum oven at 105°C overnight. Subsequently, membrane samples were immersed into pure solvents: acetone, butanol, and ethanol. After a given period of time, samples were weighted with an accuracy of four decimal places until a constant weight was reached. Degree of swelling (SW) was calculated using Eq. (7) [46].

SW 

Ws  Wd  100% Wd

(7)

where Ws and Wd are masses of swollen (s) and dry (d) samples, respectively.

2.5. Gas chromatography Composition of permeate and feed liquid mixtures were determined using Varian 3300 gas chromatograph with TCD detector. Results of gas chromatography analysis were processed using Borwin software (JMBS, France). The procedure was described in the detail elsewhere [3]. Values of LOQ (limit of quantification) and LOD (limit of detection) for investigated aqueous mixtures were determined in our previous work [3] and are collated in Tab. 2. Table. 2 LOD and LOQ for investigated organic solvents. Solvent LOD 0.05 wt.% Acetone 0.16 wt.% Butanol 0.03 wt.% Ethanol

LOQ 0.15 wt.% 0.50 wt.% 0.10 wt.%

11

3. Results and discussion 3.1. Microfiltration In Fig. 3 the time evolution of water permeate and broth permeate fluxes during microfiltration process are presented. It can be seen that during more than 6h of microfiltration process the flux was stable, however, the system needed around 60 min to reach the steady-state. The average value of water flux was around 118 ± 1.9 L h-1 m-2, whereas in the case of fermentation broth, a significant reduction of flux is observed, as the permeate flux value was equal to 6 ± 0.05 L h-1 m-2. This significant flux reduction is related to the composition of the broth. Unfiltered broth contains a substances which can negatively influence the membrane performance due to the settlement of these particles on the membrane surface as well as inside the membrane pores. Moreover, this phenomenon will promote the fouling. The high effectiveness of the filtration process was proved using the calculated rejection coefficient (R) which was higher than 99.99%. The measured turbidity for fermentation broth was equal 490 NTU (Nephelometric Turbidity Unit), whereas for the filtrated solution it was equal to 0.02 NTU. This is an additional confirmation of the efficiency of filtration process. Such pretreatment step of filtration will influence the transport across the membrane as well as will mitigate the fouling problem due to the smaller amount of particles [38].

12

Fig. 3. Transport properties of ceramic membrane TiO2-300kD contacting water and fermentation broth.

3.2. Swelling properties of POMS, PEBAX, and Pervap4060 membranes Results presented in Tab. 3 show that swelling of all the membranes in butanol was the highest and the lowest one was found for membranes equilibrated with ethanol. It has to be pointed out that SW of PEBAX membrane in all investigated solvents is significantly higher than SW of Pervap4060 and POMS membranes (Tab. 3). Such observation is in a good agreement with calculated values of distance parameter between solvents and polymers according to HSP theory (Tab. 1). The lowest distance parameter values were calculated for interactions between PEBAX and tested solvents. Commercial membranes used in this study are composite ones and due to that their structure consist of polymer selective layer and porous nonselective support layers. Obtained results are in an accordance with sorption results presented by Rozicka et al. [46], however it must be remembered that the support layer of the membrane can also influence the swelling results.

13

Tab. 3. Degree of swelling of PDMS, POMS, and PEBAX based membranes Degree of swelling (SW) [wt. %] Membrane Acetone Butanol Ethanol 45.9 68.8 33.1 POMS 84.0 87.6 74.9 PEBAX 44.5 67.0 56.4 Pervap4060 3.3. Pervaporative removal of ethanol, butanol, and acetone from aqueous binary mixtures The PV performance of Pervap4060, POMS, and PEBAX membranes were tested in pervaporative separation of ABE from binary aqueous solutions at 60°C, within a concentration range of 0-5 wt. % organics. Fig. 4 presents results obtained during separation of water-ethanol mixtures at 60oC. The highest content of ethanol in permeate was observed during experiments with Pervap4060 membrane and the lowest separation efficiency was observed for PEBAX membrane (Fig. 4A). In Fig. 1B it can be seen that the highest ethanol flux was observed during experiments with POMS membrane. Ethanol fluxes and its content in permeate increase with increasing organic solvent content in the feed mixture.

Fig.4. Concentration of ethanol in permeate (A) and permeate fluxes (B) as a function of concentration of ethanol in feed 14

Differences in separation efficiency result from the different sorption rates of organic compounds and different transport rate of water. According to the solution-diffusion mechanism [42], transport in pervaporation depends on the both sorption and diffusion of components across the selective barrier and the resultant separation performance depends on the ratio of organic solvent transport to water transport. It should be highlighted that both the morphology and thickness of membrane selective layer affect the permeate fluxes [46]. For this reason, it is better to discuss the pervaporation results using flux ratios rather than the absolute values of fluxes. Fig. 5 presents ratios of ethanol to water molar fluxes. The higher this ratio is, the more selective membrane is towards the given organic solvent. From Fig. 5 it can be seen that the highest ratio was found for Pervap4060 membrane and the lowest one for PEBAX (Fig. 5).

Fig.5. Comparison of ethanol flux to water flux ratio during recovery of ethanol from ethanolwater mixtures (at 2 wt.% of ethanol in feed).

Lazarova et al. [47] tested POMS membrane in pervaporative separation of diluted water-ethanol mixture. During pervaporation experiments at 25oC in contact with waterethanol mixture containing 5 wt.% of ethanol the total permeation flux equal to 30 g m-2 h-1 and separation factor β=3 were observed. Authors reported that the highest separation factor was achieved for 0.5 wt. % of ethanol in feed. These results showed that POMS membrane

15

can be successfully applied at low ethanol concentration [47]. It should be highlighted that there are many type of POMS membranes, however also in the presented work the POMS membrane was effective and sufficient for VOCs removal (i.e. ethanol). The lower efficiency of the PEBAX membrane is related to the hydrophobic/organophilic nature of the PEBA polymer. Fig. 6 presents the results of separation of ethanol from filtered and unfiltered fermentation broth. Fermentation broth contained the molasses and yeast used to produce ethanol. It can be noticed that the biomass presence negatively influences permeate fluxes and separation efficiency (Fig. 6). Lower separation factor for unfiltered fermentation broth resulted from membrane fouling (Fig 6A). To assess the influence of fouling on the membrane performance the normalized flux decline (FDn) based on the equation (Eq. (8)) was determined. According to the literature data the FDn values can change between 0% and 95%, depending on the various factors like membrane material or separation system [48]. In the investigated system, the flux decline was equal to 1.2% and 13.1% for filtrated and infiltrated broth, respectively.

 J  FDn  1  f  100 [%]  J0  where J0 and Jf correspond to the initial and final permeate fluxes, respectively.

(8)

16

Fig.6. Concentration of ethanol in permeate (A), partial moral fluxes (C) as a function of feed composition and water fluxes as a function of time of experiment (B), Pervap4060 membrane.

As it can be seen from Fig. 6B, during experiments with filtered fermentation broth water flux was higher comparing to the water flux during PV with unfiltered fermentation broth. Moreover, when unfiltered fermentation broth was used as feed, water flux decreased during the experiment, what resulted from membrane fouling, i.e. growing an adsorption layer of fermentation by-products on the membrane surface, which was proved by the calculation of flux decline parameter. Ethanol flux was proportional to its content in the feed mixture, nevertheless lower ethanol flux was observed during pervaporation using an unfiltered fermentation broth (Fig. 6C). Various pervaporation results presenting separation of ethanol from fermentation broth are gathered in Tab. 3. Pervap4060 membrane showed relatively high flux of ethanol for both filtered and unfiltered fermentation broths. However, presence of biomass in feed mixture reduced total flux from 59.8 mol m-2 h-1 for the filtered broth to 45.2 mol m-2 h-1 for the unfiltered one (Tab. 4). Similar relationship was observed by Gaykawad et al. [49]. Authors

17

noticed that membrane fouling causes mitigation of a total flux by ca. 17% compared with the experiment using a model water-ethanol mixture (Tab. 4). Tab. 4. Comparison of various membranes performance in recovery of ethanol from fermentation broth by pervaporation. Feed

Membrane

Cethanol [wt. %]

Jethanol [mol m-2 h-1]

Jwater [mol m-2 h-1]

β [-]

Reference

Unfiltered broth

hollow PDMS

3

0.02

0.3

11

[50]

Unfiltered broth

PDMS

4.9

6.7

64.6

8

[51]

Filtered broth

ZSM5/PDMS

3

0.7

3.7

16

[52]

Unfiltered broth

PDMS

2.2

1.4

25.5

7

[53]

Unfiltered broth

Silicalite membrane

4.6

8.7

5.4

86

[54]

Unfiltered broth

Pervatech

3

1.3

27.7

4

[49]

Model waterethanol mixture Filtered broth

Pervatech

3

1.9

33.0

5

[49]

Pervap4060

4

6.0

53.8

6

This work

Unfiltered broth

Pervap4060

4

3.3

41.9

4

This work

Fig. 7 presents results of separation of butanol from binary aqueous solutions. Similarly, like in the case of water-ethanol mixture (Fig. 4A), the highest butanol content in permeate was obtained for Pervap4060 membrane (Fig. 7A), whereas POMS and PEBAX showed similar membrane separation properties. In should be also mentioned that during ABE fermentation with Clostridium bacteria, acetone, butanol and ethanol are produced in a typical weight ratio of A:B:E=3:6:1 [55, 56], which means that the content of butanol is the highest in the fermentation broth mixture. Content of organic component in permeate during pervaporation experiments with water-butanol mixture was much higher comparing with the results obtained for water-ethanol system. This is related to the higher affinity of butanol to polymers forming the selective layer (Tab. 1). Regarding the fluxes, the highest butanol flux was observed for POMS membrane and the lowest one for PEBAX (Fig. 7B). During the pervaporation experiments water fluxes through PEBAX and Pervap4060 membranes were

18

constant and practically independent on butanol feed concentration. However, in the case of POMS membrane, water flux increased slightly with increasing butanol content at the low content of butanol in the feed mixture (i.e. in the range of 0 to 1 wt. % for butanol). The possible reason for that is a swelling of POMS selective layer in the presence of butanol, which causes the facilitated transport of water.

Fig.7. Concentration of butanol in permeate (A) and permeate fluxes (B) as a function of concentration of butanol in feed.

Comparison of butanol to water molar flux ratios (Fig. 8) confirms that Pervap4060 membrane was the most efficient one among all tested membranes used for butanol recovery from water. Moreover, calculated ratios of butanol to water fluxes were higher comparing with water-ethanol ones. The high separation ability of Pervap4060 membrane results from the limited water flux and the relatively high flux of butanol.

19

Fig.8. Comparison of butanol flux to water flux ratio during recovery of butanol from butanol-water mixtures (at 2 wt.% of butanol in feed).

The presented findings are in a good accordance with the data presented in the literature.

Concerning

POMS

membrane,

Beltran

et

al.

[57]

tested

poly(octylomethylosiloxane) (POMS)/oleyl alcohol (OA) supported liquid membrane loaded with different amount of oleyl alcohol into the membrane (in the range 10 to 50 % of oleyl alcohol) and used for the separation of water-butanol mixture containing up to 1.5 wt. % butanol at 35oC. It was observed that butanol flux increased with increasing feed butanol content, whereas water flux was constant. The best separation efficiency, with PSI equal to 1.52 kg m-2 h-1 and total flux equal to 17.3 g m-2 h-1 for butanol recovery from water-butanol mixture was found for 30 wt. % OA/POMS membrane. Comparing the aforementioned data with those obtained in this work for POMS membrane, it can be stated that doping of the membrane by oleyl alcohol improved the membrane efficiency towards butanol recovery. Li et al. [58] prepared PEBA/ceramic hollow fibre composite membrane which was subsequently tested in pervaporative separation of water-butanol mixtures. The total flux equal to 2011 g m-2 h-1 and separation factor β=20 were found at feed temperature of 40 oC and 1 wt. % of butanol in feed. Authors also observed that permeation flux increased with butanol content in the feed solution. Moreover, the performed research showed that the

20

membrane was stable during 200 hours of the experiment. In our case the PEBAX membrane possess the lowest efficiency in all cases of VOCs recovery from water. This behavior can be explained twofold: firstly because of the hydrophobic/organophilic nature of the PEBA polymer and secondly because of small distance of Hansen Solubility Parameter between this type of the membrane and water (Table 1). It means that PEBAX membrane ones possesses the biggest affinity to water. Results for water-acetone binary mixtures are presented in Figs. 9 and 10. It can be seen that during separation of water-acetone mixture Pervap4060 membrane also showed the highest ability to separate acetone from aqueous solution (Fig. 9A), whereas the lowest separation efficiency toward acetone was again observed for PEBAX membrane. POMS membrane showed similar separation efficiency for acetone comparing with water-butanol mixture. PEBAX membrane demonstrated lower content of organic compound in permeate comparing with water-butanol mixture, whereas, Pervap4060 membrane showed the highest content of acetone in permeate comparing with water-butanol and water-ethanol mixtures. During pervaporation experiments, water fluxes through Pervap4060 and PEBAX membranes were constant (Fig. 9B), however for the POMS membrane water flux increased slightly during the pervaporation experiment what is caused by the swelling effect. Similarly to waterbutanol and water-ethanol mixtures, the highest water flux was observed for the POMS membrane and the lowest one for the PEBAX membrane. Pervap4060 and POMS membranes produced similar acetone permeate fluxes (Fig. 9B).

21

Fig.9. Concentration of acetone in permeat (A) and permeate fluxes (B) as a function of concentration of acetone in feed.

Comparison of acetone to water molar flux ratio (Fig. 10) shows that Pervap4060 membrane was the most efficient one among all membranes used for acetone recovery from water. It should be also highlighted that ratio of acetone to water fluxes obtained with POMS and Pervap4060 membranes were higher comparing with molar flux ratios of butanol to water and ethanol to water. This could result from a higher affinity of acetone to POMS and Pervap4060 membranes. In the case of PEBAX membrane these ratios are similar. It should be also noted that acetone has higher vapour pressure comparing with butanol and ethanol which results in a higher driving force for pervaporative transport [46].

22

Fig. 10. Comparison of acetone flux to water flux ratio during recovery of butanol from butanol-water mixtures (at 2 wt.% of acetone in feed).

Hollein at al. [59] evaluated transport and separation properties of various membranes (based on PDMS, PEBA, and SC in separation of water-acetone mixture. Pervaporation experiments with commercial membranes were carried out at 4.5 wt. % of acetone at 50C. The highest (1.1 kg m-2 h-1) and the lowest (0.2 kg m-2 h-1) acetone fluxes were obtained during experiment with silicone composite (SC) and PEBA membrane, respectively. However, the highest separation coefficient was obtained with PDMS membrane (β=55) at 4.5 wt. % of acetone in feed [59]. Comparing the results discussed by Hollein et. al [59] the results obtained in this work, it can be stated that the similar conclusion can be drawn. The lowest transport and separation properties were observed for the PEBAX membrane (Fig. 9). Kujawski et al. [60] utilized PEBA, Pervap1060 and Pervap1070 membrane for the recovery of acetone from binary aqueous solutions at 40C. Pervap1070 is a commercial PDMS membrane filled with silicalite-1 particles. The best efficiency in separation of acetone from water was obtained for Pervap1070 membrane and the worst one was demonstrated by PEBA membrane. It was also noticed that fluxes of acetone depended strongly on concentration of organic solvent, however water permeate flux was constant. The presence of

23

zeolite filler (Pervap 1070 membrane) caused the diminution of the both water and acetone fluxes. It should be also pointed out that water flux obtained with Pervap1070 membrane was much lower comparing to flux through Pervap1060 one [60]. Similar to our findings PEBA membrane was characterized by the lowest efficiency performance. Moreover, likewise fluxes of separated VOCs depended strongly on concentration of organic solvent whereas water permeate flux was constant.

3.4. Influence of the organic component presence on water flux Tab. 5. Comparison of water fluxes through POMS, PEBAX, and Pervap4060 membranes for pure water and 2 wt. % of organic component at 60 oC. Water flux System Membrane (mol m-2 h-1) 321.7 water 464.6 water-butanol POMS 418.8 water-acetone 341.5 water-ethanol water water-butanol water-acetone water-ethanol water water-butanol water-acetone water-ethanol

PEBAX

155.0 159.8 145.6 144.5

Pervap4060

112.8 111.1 110.9 110.0

Tab. 5 collates water permeate fluxes determined in pervaporation experiments with investigated membranes contacting pure water and binary aqueous solutions containing 2 wt. % of organic solvent. Water fluxes through individual commercial membranes are quite different and the differences result from the different polymeric material, different morphology and the different thicknesses of the membrane separation layers.

24

Interesting information can be drawn out when comparing the influence of the organic component on water permeate flux for each investigated membrane. In the case of POMS membrane it can be noticed that the water flux is much higher in water-butanol and water-acetone mixtures, comparing to the fluxes in contact with pure water or ethanol mixture. These differences are in the agreement with the swelling results of POMS membrane. Structure of POMS membrane possesses long side octyl chains which allow for the easier transport of water molecules together with butanol or acetone molecules. It can be stated that there are coupling between organics and water fluxes, caused by swelling and facilitated transport of water. Reviewing the results for water fluxes through PEBAX membrane (Tab. 5) it can be seen that in the case of water-acetone and water-ethanol mixtures, water flux is slightly suppressed (ca. 145 mol m-2 h-1) comparing to the flux of pure water (155 mol m-2 h-1). Whereas in the case of water-butanol mixtures the flux of water is slightly higher (ca. 160 mol m-2 h-1). It must be remembered that PEBAX membrane possesses hydrophobic/organophilic character. Taking this fact into account it can be concluded that ethanol and acetone molecules are transported mostly through the organophilic parts of polymer reducing slightly water flux. Butanol molecules are transported through hydrophobic parts causing swelling and slightly facilitating water flux. Contrary to the behaviour of POMS and PEBAX membrane, Pervap4060 (PDMS based membrane) is the most stable one regarding the influence of organic component on water fluxes. Data presented in Tab. 5 proved that there is no influence of organic component on water permeate flux. It can be concluded that membrane was crosslinked during the preparation step and the presence of organic solvents in feed mixture does not affect water permeate fluxes. As it was presented in Figs. 4, 7, and 9 water permeate fluxes are

25

independent on the feed mixture composition. The similar trend was observed by Kujawski et al. [41]. 3.5. Membrane performance in removal of organics from water Pervaporative Separation Index (PSI) – Eq.(4) was used to evaluate the potential separation efficiency of different membranes applied for the removal of organic component from binary aqueous mixtures (Fig. 11). According to the definition of PSI, the higher the PSI value, the more effective the membrane is. Inspecting Fig. 11 it can be stated that the lowest separation efficiency was found for PEBAX membrane, whereas both POMS and Pervap4060 membrane showed similar separation ability, as the results for the both membranes are located along the same isopsine line [12]. However it should be underlined that Pervap4060 membrane produced higher separation factors comparing with those for POMS membrane, what means that concentration of organics in permeate using Pervap4060 membrane is higher. Higher values of separation factor with the similar PSI values indicate that Pervap4060 membrane would be better choice for the practical removal of organics from binary and multicomponent aqueous mixtures. Moreover, it should be also taken into account that butanol possesses limited solubility in water and creates biphasic system (i.e. water rich phase and butanol rich phase) when the concentration of butanol is in the range 7.1-79.4% [3]. Pervap 4060 membrane showed much higher value of separation factor (β=30) comparing with POMS membrane (β=9). For that reason Pervap4060 membrane produces higher content of butanol in permeate (Fig.7A) and lower water flux (Fig.7B). Higher content of butanol in permeate means also a larger amount of the organic rich phase and the subsequent dehydration of this phase obtained in pervaporation by hydrophilic pervaporation would be much more efficient. The results on the pervaporative dehydration of organic rich quaternary ABE-water mixture will be presented and discussed in the second part of this research.

26

Fig.11. Comparison of various membranes performance during separation of water-butanol, wateracetone, water-ethanol mixtures; (at 2 wt.% of organic compound in feed).

Taking into account the presented results, the batch pervaporation process applied for the removal of organics from water was modelled assuming the mass balance in the system and following the procedure presented in detail elsewhere [61, 62]. First the degree of removal (Dr) was calculated according to Eq. (9), for feed mixtures containing 2 wt. % of organics (Figs. 12A-C). Calculations were performed assuming the initial feed mass to membrane area ratio equal to 10 kg m-2. Subsequently, removal of ternary ABE mixture from water was simulated assuming that at the outset of the batch process feed mixture contained 2.0 wt. % ABE in the ratio 3:6:1 (i.e. 0.6 wt.% AcO, 1.2 wt.% BuOH, and 0.2 wt.% EtOH) – Fig. 12D. In this latter case, calculation were accomplished using the initial feed mass to membrane area ratio equal to 25 kg m-2.

27

Fig. 12. Modeling of the removal of organics from binary and quaternary aqueous mixtures. Membrane: Pervap4060, T=60oC, A: water-butanol mixture, B: water-acetone mixture, C: waterethanol mixture, D: quaternary ABE-water mixtures (initial content of organics in feed mixture - 2 wt.%). Initial feed mass to membrane area ratio is equal to 10 kg m-2 (A, B, and C) and to 25

kg m-2 (D).

Dr 

m(t ) 100% mtot

(9)

where Dr – degree of removal [%], m(t) – mass of the organic component [kg] removed after time t [min], mtot – initial mass of the organic component [kg] in the feed mixture at the outset of experiments. It can be seen that in the case of water-acetone and water-butanol mixture, the degree of removal of organics reaches a value close to 100% within 40 minutes of the pervaporation process (Figs. 12A and 12B). In the case of ethanol – the removal of this organic solvents can take up to 100 minutes (Fig. 12C). Fig. 12D shows the simulation of the pervaporation

28

process applied for the removal of ABE solvents from water, assuming the initial concentrations equal to 0.6 wt.% acetone, 1.2 wt.% butanol and 0.2 wt.% ethanol, i.e. similar to the concentrations of the fermentation broth. It can be seen, that using the initial feed mass to membrane area ratio equal to 25 kg m-2, the total recovery of butanol and acetone occurs within ca. 100 minutes, whereas after this period of time, only ca. 74% of ethanol is recovered. In the practical application, the remaining ethanol could be recovered using the two stages system. The more detailed discussion of the possible configuration will be presented in the second part of this work.

4. Conclusions Properties of various commercial membranes were evaluated in pervaporative removal of acetone, butanol and ethanol from binary aqueous mixtures. Tested membranes (POMS, PEBA, and PDMS based) were selective towards organics, however PDMS membrane (Pervap4060) showed remarkable good properties and proved the feasibility of pervaporation process in the removal of ABE mixture components from water. The highest efficiency of Pervap4060 membrane resulted from high organics fluxes and relatively low water flux. Microfiltration of fermentation broth should be used prior to pervaporation to suppress the membrane fouling. Modeling proved the effectiveness of pervaporation in contact with both binary and quaternary aqueous mixtures.

Acknowledgements This work were supported by project PVABE TANGO1/266441/NCBR/2015 granted by The National Centre for Research and Development and statutory founds of Nicolaus Copernicus University in Toruń, Poland (Faculty of Chemistry, T-109 "Membranes and membrane separation processes - fundamental and applied research"

29

References [1] V.V. Volkov, Separation of liquids by pervaporation through polymeric membranes, Russ. Chem. Bull., 43 (1994) 187-198. [2] J.G. Crespo, C. Brazinha, 1 - Fundamentals of pervaporation, in: A. Basile, A.F. Khayet (Eds.) Pervaporation, Vapour Permeation and Membrane Distillation, Woodhead Publishing, Oxford, 2015, pp. 3-17. [3] A. Kujawska, K. Knozowska, J. Kujawa, W. Kujawski, Influence of downstream pressure on pervaporation properties of PDMS and POMS based membranes, Sep. Purif. Technol., 159 (2016) 68-80. [4] N. Qureshi, M.M. Meagher, J. Huang, R.W. Hutkins, Acetone butanol ethanol (ABE) recovery by pervaporation using silicalite-silicone composite membrane from fed-batch reactor of Clostridium acetobutylicum, J. Membr. Sci., 187 (2001) 93-102. [5] M. Žák, M. Klepic, L.Č. Štastná, Z. Sedláková, H. Vychodilová, Š. Hovorka, K. Friess, A. Randová, L. Brožová, J.C. Jansen, M.R. Khdhayyer, P.M. Budd, P. Izák, Selective removal of butanol from aqueous solution by pervaporation with a PIM-1 membrane and membrane aging, Sep. Purif. Technol., 151 (2015) 108-114. [6] Y. Zhang, N.E. Benes, R.G.H. Lammertink, Performance study of pervaporation in a microfluidic system for the removal of acetone from water, Chem. Eng. J., 284 (2016) 13421347. [7] S.C. Fraga, A. Kujawska, W. Kujawski, C. Brazinha, J.G. Crespo, Transport of dilute organics through dense membranes: Assessing impact on membrane-solute interactions, J. Membr. Sci, 523 (2017) 346-354. [8] B. Belaissaoui, Y. Le Moullec, E. Favre, Energy efficiency of a hybrid membrane/condensation process for VOC (Volatile Organic Compounds) recovery from air: A generic approach, Energy, 95 (2016) 291-302.

30

[9] V.H.G. Díaz, G.O. Tost, Butanol production from lignocellulose by simultaneous fermentation, saccharification, and pervaporation or vacuum evaporation, Bioresour. Technol., 218 (2016) 174-182. [10] S. Vijay Kumar, C. Arnal-Herault, M. Wang, J. Babin, A. Jonquieres, Multiblock Copolymer Grafting for Butanol Biofuel Recovery by a Sustainable Membrane Process, ACS Appl. Mater. Interfaces, 8 (2016) 16262-16272. [11] V. Garcia, J. Pakkila, H. Ojamo, E. Muurinen, R.L. Keiski, Challenges in biobutanol production: How to improve the efficiency?, Renew. Sust. Energ. Rev., 15 (2011) 964-980. [12] A. Kujawska, J. Kujawski, M. Bryjak, W. Kujawski, ABE fermentation products recovery methods—A review, Renew. Sust. Energ. Rev., 48 (2015) 648-661. [13] M. Lapuerta, R. Ballesteros, J. Barba, Strategies to Introduce n-Butanol in Gasoline Blends, Sustainability, 9 (2017) 589 (581-510). [14] E.E. Johnson, L. Rehmann, The role of 1,3-propanediol production in fermentation of glycerol by Clostridium pasteurianum, Bioresour. Technol., 209 (2016) 1-7. [15] D. Cai, S. Hu, Q. Miao, C. Chen, H. Chen, C. Zhang, P. Li, P. Qin, T. Tan, Two-stage pervaporation process for effective in situ removal acetone-butanol-ethanol from fermentation broth, Bioresour. Technol., 224 (2017) 380-388. [16] D.-G. Kim, T. Takigawa, T. Kashino, O. Burtovyy, A. Bell, R.A. Register, Hydroxyhexafluoroisopropylnorbornene Block and Random Copolymers via Vinyl Addition Polymerization and Their Application as Biobutanol Pervaporation Membranes, Chem. Mater., 27 (2015) 6791-6801. [17] P. Izák, W. Ruth, Z. Fei, P.J. Dyson, U. Kragl, Selective removal of acetone and butan-1ol from water with supported ionic liquid–polydimethylsiloxane membrane by pervaporation, Chem. Eng. J., 139 (2008) 318-321.

31

[18] T. Sarchami, E. Johnson, L. Rehmann, Optimization of fermentation condition favoring butanol production from glycerol by Clostridium pasteurianum DSM 525, Bioresour. Technol., 208 (2016) 73-80. [19] W. Van Hecke, P. Vandezande, M. Dubreuil, M. Uyttebroek, H. Beckers, H. De Wever, Biobutanol production from C5/C6 carbohydrates integrated with pervaporation: experimental results and conceptual plant design, J. Ind. Microbiol. Biotechnol., 43 (2016) 25-36. [20] S. Kießlich, T. Sarchami, G. Munch, K. Gao, L. Rehmann, W. Kujawski, E. Johnson, Pervaporative butanol removal from PBE fermentation broths for the bioconversion of glycerol by Clostridium pasteurianum, J. Membr. Sci, 535 (2017) 79-88. [21] A. Kujawska, J. Kujawski, M. Bryjak, W. Kujawski, Removal of volatile organic compounds from aqueous solutions applying thermally driven membrane processes. 1. Thermopervaporation, Chem. Eng. Process., 94 (2015) 62-71. [22] O. Vopička, V. Hynek, K. Friess, P. Izák, Blended silicone–ionic liquid membranes: Transport properties of butan-1-ol vapor, Eur. Polym. J., 46 (2010) 123-128. [23] W. Van Hecke, P. Vandezande, S. Claes, S. Vangeel, H. Beckers, L. Diels, H. De Wever, Integrated bioprocess for long-term continuous cultivation of Clostridium acetobutylicum coupled to pervaporation with PDMS composite membranes, Bioresour. Technol., 111 (2012) 368-377. [24] A. Figoli, T. Marino, F. Galiano, 2 - Polymeric membranes in biorefinery, in: A. Figoli, A. Cassano, A. Basile (Eds.) Membrane Technologies for Biorefining, Woodhead Publishing, Amsterdam, 2016, pp. 29-59. [25] S. Liu, G. Liu, J. Shen, W. Jin, Fabrication of MOFs/PEBA mixed matrix membranes and their application in bio-butanol production, Sep. Purif. Technol., 133 (2014) 40-47.

32

[26] A.G. Fadeev, Y.A. Selinskaya, S.S. Kelley, M.M. Meagher, E.G. Litvinova, V.S. Khotimsky, V.V. Volkov, Extraction of butanol from aqueous solutions by pervaporation through poly(1-trimethylsilyl-1-propyne), J. Membr. Sci, 186 (2001) 205-217. [27] I.L. Borisov, A.O. Malakhov, V.S. Khotimsky, E.G. Litvinova, E.S. Finkelshtein, N.V. Ushakov, V.V. Volkov, Novel PTMSP-based membranes containing elastomeric fillers: Enhanced 1-butanol/water pervaporation selectivity and permeability, J. Membr. Sci., 466 (2014) 322-330. [28] I.L. Borisov, V.V. Volkov, Thermopervaporation concept for biobutanol recovery: The effect of process parameters, Sep. Purif. Technol., 146 (2015) 33-41. [29] S. Claes, P. Vandezande, S. Mullens, K. De Sitter, R. Peeters, M.K. Van Bael, Preparation and benchmarking of thin film supported PTMSP-silica pervaporation membranes, J. Membr. Sci, 389 (2012) 265-271. [30] A. Rom, A. Friedl, Investigation of pervaporation performance of POMS membrane during separation of butanol from water and the effect of added acetone and ethanol, Sep. Purif. Technol., 170 (2016) 40-48. [31] K. Srinivasan, K. Palanivelu, A. Navaneetha Gopalakrishnan, Recovery of 1-butanol from a model pharmaceutical aqueous waste by pervaporation, Chem. Eng. Sci., 62 (2007) 2905-2914. [32] I.L. Borisov, G.S. Golubev, V.P. Vasilevsky, A.V. Volkov, V.V. Volkov, Novel hybrid process for bio-butanol recovery: Thermopervaporation with porous condenser assisted by phase separation, J. Membr. Sci, 523 (2017) 291-300. [33] A.G. Fadeev, M.M. Meagher, S.S. Kelley, V.V. Volkov, Fouling of poly[-1(trimethylsilyl)-1-propyne] membranes in pervaporative recovery of butanol from aqueous solutions and ABE fermentation broth, J. Membr. Sci, 173 (2000) 133-144.

33

[34] W. Kujawski, S.R. Krajewski, Influence of inorganic salt on the effectiveness of liquid mixtures separation by pervaporation, Sep. Purif. Technol., 57 (2007) 495-501. [35] G. Liu, W. Wei, H. Wu, X. Dong, M. Jiang, W. Jin, Pervaporation performance of PDMS/ceramic composite membrane in acetone butanol ethanol (ABE) fermentation–PV coupled process, J. Membr. Sci, 373 (2011) 121-129. [36] V. Garcia, E. Pongracz, E. Muurinen, R.L. Keiski, Recovery of n-butanol from salt containing solutions by pervaporation, Desalination, 241 (2009) 201-211. [37] S. Chovau, S. Gaykawad, A.J.J. Straathof, B. Van der Bruggen, Influence of fermentation by-products on the purification of ethanol from water using pervaporation, Bioresour. Technol., 102 (2011) 1669-1674. [38] M.H.V. Mulder, J.O. Hendrickman, H. Hegeman, C.A. Smolders, Ethanol—water separation by pervaporation, J. Membr. Sci, 16 (1983) 269-284. [39] P. Boyaval, C. Lavenant, G. Gésan, G. Daufin, Transient and stationary operating conditions on performance of lactic acid bacteria crossflow microfiltration, Biotechnol. Bioeng., 49 (1996) 78-86. [40] R.W. Baker, J.G. Wijmans, Y. Huang, Permeability, permeance and selectivity: A preferred way of reporting pervaporation performance data, J. Membr. Sci., 348 (2010) 346352. [41] J. Kujawski, A. Rozicka, M. Bryjak, W. Kujawski, Pervaporative removal of acetone, butanol and ethanol from binary and multicomponent aqueous mixtures, Sep. Purif. Technol., 132 (2014) 422-429. [42] J.G. Wijmans, R.W. Baker, The solution-diffusion model - a review, J. Membr. Sci., 107 (1995) 1-21. [43] C.M. Hansen, Hansen solubility parameters: a user's handbook, second ed., CRC press, Boca Raton, 2007.

34

[44] S. Heitmann, V. Krüger, D. Welz, P. Lutze, Experimental investigation of pervaporation membranes for biobutanol separation, J. Membr. Sci., 2 (2013) 245-262. [45] J. Niemistö, W. Kujawski, R.L. Keiski, Pervaporation performance of composite poly(dimethyl siloxane) membrane for butanol recovery from model solutions, J. Membr. Sci., 434 (2013) 55-64. [46] A. Rozicka, J. Niemistö, R.L. Keiski, W. Kujawski, Apparent and intrinsic properties of commercial PDMS based membranes in pervaporative removal of acetone, butanol and ethanol from binary aqueous mixtures, J. Membr. Sci, 453 (2014) 108-118. [47] M. Lazarova, P. Bosch, A. Friedl, POMS Membrane for Selective Separation of Ethanol from Dilute Alcohol-Aqueous Solutions by Pervaporation, Sep. Sci. Technol., 47 (2012) 1709-1714. [48] M. Laqbaqbi, J. Sanmartino, M. Khayet, C. García-Payo, M. Chaouch, Fouling in Membrane Distillation, Osmotic Distillation and Osmotic Membrane Distillation, 7 (2017) 334. [49] S.S. Gaykawad, Y. Zha, P.J. Punt, J.W. van Groenestijn, L.A.M. van der Wielen, A.J.J. Straathof, Pervaporation of ethanol from lignocellulosic fermentation broth, Bioresour. Technol., 129 (2013) 469-476. [50] R.H. Bello, P. Linzmeyer, C.M.B. Franco, O. Souza, N. Sellin, S.H.W. Medeiros, C. Marangoni, Pervaporation of ethanol produced from banana waste, Waste Manage, 34 (2014) 1501-1509. [51] Y. Wu, W. Huang, X. Zeyi, Y. Zhong, Ethanol recovery from fermentation broth by pervaporation using a composite polydimethylsiloxane membrane, Chinese J. Chem. Eng., 12 (2004) 586-589. [52] R.D. Offeman, C.N. Ludvik, Poisoning of mixed matrix membranes by fermentation components in pervaporation of ethanol, J. Membr. Sci, 367 (2011) 288-295.

35

[53] M. Esfahanian, A. Ghorbanfarahi, A. Ghoreyshi, G. Najafpour, H. Younesi, A. Ahmad, Enhanced bioethanol production in batch fermentation by pervaporation using a PDMS membrane bioreactor, Int. J. Eng., Trans. B, 25 (2012) 249-258. [54] M. Nomura, T. Bin, S.-i. Nakao, Selective ethanol extraction from fermentation broth using a silicalite membrane, Sep. Purif. Technol., 27 (2002) 59-66. [55] X.Q. Lin, J.L. Wu, J.S. Fan, W.B. Qian, X.Q. Zhou, C. Qian, X.H. Jin, L.L. Wang, J.X. Bai, H.J. Ying, Adsorption of butanol from aqueous solution onto a new type of macroporous adsorption resin: Studies of adsorption isotherms and kinetics simulation, J. Chem. Technol. Biotechnol., 87 (2012) 924-931. [56] G. Merlet, F. Uribe, C. Aravena, M. Rodriguez, R. Cabezas, E. Quijada-Maldonado, J. Romero, Separation of fermentation products from ABE mixtures by perstraction using hydrophobic ionic liquids as extractants, J. Membr. Sci, (2017). [57] A.B. Beltran, G.M. Nisola, E.L. Vivas, W. Cho, W.J. Chung, Poly(octylmethylsiloxane)/oleyl alcohol supported liquid membrane for the pervaporative recovery of 1-butanol from aqueous and ABE model solutions, J. Ind. Eng. Chem., 19 (2013) 182-189. [58] Y. Li, J. Shen, K. Guan, G. Liu, H. Zhou, W. Jin, PEBA/ceramic hollow fiber composite membrane for high-efficiency recovery of bio-butanol via pervaporation, J. Membr. Sci, 510 (2016) 338-347. [59] M.E. Hollein, M. Hammond, C.S. Slater, Concentration of Dilute Acetone-Water Solutions Using Pervaporation, Sep. Sci. Technol., 28 (1993) 1043-1061. [60] W. Kujawski, A. Warszawski, W. Ratajczak, T. Porębski, W. Capała, I. Ostrowska, Application of pervaporation and adsorption to the phenol removal from wastewater, Sep. Purif. Technol., 40 (2004) 123-132.

36

[61] J.K. Kujawski, W.M. Kujawski, H. Sondej, K. Jarzynka, A. Kujawska, M. Bryjak, E. Rynkowska, K. Knozowska, J. Kujawa, Dewatering of 2,2,3,3-tetrafluoropropan-1-ol by hydrophilic pervaporation with poly(vinyl alcohol) based Pervap™ membranes, Sep. Purif. Technol., 174 (2017) 520-528. [62] K. Rychlewska, W. Kujawski, K. Konieczny, Pervaporative performance of PEBA and PDMS based commercial membranes in thiophene removal from its binary mixtures with hydrocarbons, Fuel Process. Technol., 165 (2017) 9-18.

37

Graphical abstract

38

Performance of commercial composite hydrophobic membranes applied for pervaporative reclamation of acetone, butanol, and ethanol from aqueous solutions: binary mixtures

Highlights:



PDMS, POMS, and PEBA based commercial membranes assessed in ABE pervaporation.



PV efficiency in removal of EtOH from filtrated and unfiltrated broth evaluated



PV results discussed using organics to water molar flux ratios and PSI factor.



Pervap4060 PDMS based membrane is the most efficient in removal of ABE from water.



Model calculations proved the feasibility of removal of ABE from water by PV

39