SciELO - Scientific Electronic Library Online

 
vol.16 issue3The influence of specific mechanical energy on cornmeal viscosity measured by an on-line system during twin-screw extrusionEXAFS as a tool for catalyst characterization: a review of the data analysis methods author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632On-line version ISSN 1678-4383

Braz. J. Chem. Eng. vol.16 n.3 São Paulo Sept. 1999

http://dx.doi.org/10.1590/S0104-66321999000300008 

Fractionation of benzene/n-hexane mixtures by pervaporation using polyurethane membranes

 

V. S. CUNHA, R. NOBREGA and A. C. HABERT
Programa de Engenharia Química/COPPE, Universidade Federal do Rio de Janeiro, P. O. Box 68502, CEP 21945-970, Rio de Janeiro-RJ, Brazil.

 

(Received: September 22, 1998; Accepted: July 2, 1999)

 

 

Abstract - In the present work polyurethane membranes obtained from different polyester/MDI-based polymers were used to separate benzene/n-hexane mixtures by pervaporation. In pervaporation experiments, with a 50% wt feed at room temperature, permeate fluxes in the range of 0.3 to 3.2 Kg/m2h (10 mm membrane thickness) and selectivity in the range of 3.8 to 5.6 were obtained. The permeate was always enriched in benzene. Taking into account the compromise between flux and selectivity, the best performance membrane was selected for complementary sorption and pervaporation experiments. Results show that selectivity increases and the permeation flux decreases when the benzene concentration in the feed decreases. In the present application, results also show that sorption is the main factor for selectivity. Using the distillation azeotropic mixture as feed, almost no influence of temperature on selectivity was observed in the range of 25oC to 56oC. The permeate flux increases seven-fold, while selectivity remains constant near 8.0.
Keywords: Pervaporation, polyurethane membranes, organic separation.

 

 

INTRODUCTION

Pervaporation (PV) is a membrane-based process useful for the separation of liquid mixtures. In this process, a liquid mixture is brought into contact with a non-porous membrane on the feed side and the permeate is removed as a vapour on the permeate side. The lower partial pressure on the permeate side is maintained by either a vacuum pump or an inert sweeping gas. Mass transfer through the membrane occurs in basically three steps: selective sorption of the molecules into the polymeric matrix on the feed side, selective diffusion through the membrane and desorption into a vapour phase on the permeate side. Thus, the main factor in the choice of a suitable polymer for a required separation is the difference between solubility and diffusivity of the components of the mixtures. This mechanism of separation is known as the solution-diffusion model (Wijmans and Baker, 1995). The capacity for separating aqueous-organic and organic-organic mixtures was demonstrated three decades ago, but work did not progress because of the lack of market potential due to competition with traditional technologies such as distillation, extraction and adsorption. The energy crisis in the 1970's refocused interest on energy-saving separation techniques. Pervaporation was aggressively investigated because it could be used to dehydrate aqueous alcohol mixtures for fuel usage, and the process was an economic alternative to energy-intensive distillation operations. In addition, new and more efficient membranes are challenging innovations in chemical separations. During the last few years, efforts have been made to develope new membranes for separating organic-organic liquid mixtures in order to extend the application of PV to the petrochemical industry, since the separation of organic-organic mixtures is one of the most important processes in this industry.

The number of potential organic/organic pairs is large and the requirements of membranes may be quite different according to the mixtures to be treated. The investigations mainly concern aromatic/non-aromatic, alcohol/alkane and alcohol/ether separations, an example being the separation of benzene and cyclohexane in benzene-toluene-xylene production plants (Néel, 1995; Luo et al., 1997; Niang et al., 1997). The aromatic/non-aromatic separation was studied by Buijs et al. (1995) using composite membranes with alkylphosphonate PPO as a selective top-layer material. Liquid crystalline polyurethane membranes were used by Wolínska and Jankowski (1995) in the separation of benzene/cyclohexane mixtures. The binary mixtures of benzene/cyclohexane, benzene/n-hexane and acetone/cyclohexane were also separated using poly(ether imide) segmented copolymer membranes by Tanihara et al. (1995). The membranes were preferentially permeable to benzene or acetone over cyclohexane or n-hexane due to preferential sorption and diffusion of benzene or acetone. The authors conclude that the block length of the PEO ( poly(ethylene oxide)) segment affected membrane performance.

In recent work, Yamasaki et al. (1997) used a hydrophilic material to separate the organic-organic mixtures by pervaporation. Membranes based on poly(vinyl alcohol) (PVA) were prepared and used to fractionate benzene/n-hexane and benzene/cyclohexane mixtures. The membranes were benzene selective over all the composition range of the feed. The separation factor of homogeneous PVA membranes was smaller than three.

It may concluded, based on this review, that current membranes still exhibit marginal benefits for most organic separation tasks. In the present work, polyurethane membranes were used in the fractionation of benzene/n-hexane mixtures by pervaporation. These material were chosen because of their greater affinity for aromatic compounds. Polyurethane elastomers are segmented block copolymers composed of alternating soft and hard blocks. Due to their thermodynamic incompatibility, the blocks undergo microphase separation, resulting in a dispersion of soft and hard segmented domains. The degree of phase separation and the domain structure are closely related to the chemical nature of both blocks and their composition and molecular weights (Aithal et al., 1990).

 

EXPERIMENTAL

Membrane

The polyurethanes used in this work were supplied by Cofade-S. Paulo and their basic properties are shown in Table 1. They are synthetized using polyesther as polyol, MDI as diisocyanate and butanediol as the chain extender. Membranes were prepared by casting 15% DMF solutions of polyurethane on a Teflonâ plate and evaporating the solvent at 60° C to a constant weight.

 

Table 1: Characteristics of polyurethanes

       

TPU 1

TPU 2

TPU 3

TPU 4

TPU 5

TPU 6

Ultimate elongation (%)

850

420

800

580

580

520

Density (r )

1.18

1.23

1.18

1.19

1.19

1.20

 

The polyurethanes used in this work are segmented copolymers composed of alternating soft (polyol segments) and hard (diisocyanate segments) blocks. The polyurethanes TPU-1 and TPU-3 have the highest fraction of soft segments (lower density) and TPU-2 has the smallest fraction of soft segments (highest density) (see Table 1).

Pervaporations Experiments

Figure 1 shows the experimental set-up utilized for the pervaporations experiments. Pervaporation experiments were performed at a temperature range of 25° to 56° C, for the whole composition range of the feed solution, using a mixture of benzene/n-hexane and a standard pervaporation cell. The downstream pressure was kept below 5 mmHg in all experiments, and the permeate was collected in a trap cooled by liquid nitrogen. The permeate flux (kg/m2 h) was calculated using Equation 1.

  (1)

 

 

Figure 1: Scheme of pervaporation experimental set-up

 

where J is the permeate flux, m is the weight of the permeate, A is the membrane area and t the time of operation. Membrane thickness is not the same for all membranes; therefore, the flux was normalized to a membrane thickness of 10 m m ( J10 ) using Equation 2.

 a8form2.gif (235 bytes) (2)

where d is membrane thickness. Fickian behavior during mass transport through the membrane was assumed.

The selectivity of the pervaporation process for a binary mixture of compounds A and B is defined as:

(3)

where Y and X are the weight fractions in the permeate and feed, respectively. The composition of the feed and pervaporate were determined by gas chromatography (Chrompack, model CP 9000) using a capillary column.

Sorption Experiments

Strips of dry membranes with a thickness of 200 mm were immersed in benzene/n-hexane mixtures and the pure compounds at room temperature. After the sorption reached equilibrium, the membranes samples were removed, the liquid on their surface wiped off with tissue paper as quickly as possible, and then placed into a dry flask connected to a cold trap and vacuum pump (Figure 2). The collected liquids were weighed and analyzed by gas chromatography.

 

Figure 2: Scheme of sorption experimental set-up

 

Two parameters can be used to characterize the equilibrium sorption properties: the equilibrium swelling ratio, S, and the sorption selectivity factor, aS, defined respectively by:

(4)

(5)

where w and w0 denote the weights of the swollen and dry membranes, respectively, Xm and X represent the weight fraction of the components at equilibrium in the membrane and in the outside mixture, respectively, and A and B are the components of the mixture.

The Flory-Huggins polymer solution model may be used to predict sorption and the sorption selectivity factor. Based on this model, for binary systems, the activity (a) of a penetrant in the polymer is given by Equation 6 as follows:

(6)

where V is the molar volume, f the volume fraction and subscripts i and p indicate the penetrant and polymer, respectively, and ciP the Flory-Huggins interaction parameter. In accord with Flory-Huggins theory, an interaction parameter ciP close to or less than 0.5 indicates a strong polymer-penetrant interaction. On the other hand, a high value of ciP (ciP >0.5) indicates a weak interaction between polymer and penetrant. Using sorption data and Flory-Huggins theory, the interaction parameter can be calculated using Equation 7 (Flory, 1953; Cunha, 1997).

(7)

 

RESULTS

Membrane Selection

The first step of this work was the selection of the best performance membrane. The choice was based on the pervaporation and sorption results.

Sorption Experiments

The six membranes were characterized in sorption experiments with pure compounds and 50% w/w of a mixture of benzene/n-hexane. The sorption behavior of the polyurethane membranes at 25° C (Figure 3) shows that the swelling increases with na increase in benzene concentration, indicating better compatibility of benzene with these polyurethanes. The results are in good agreement with those obtained by Stephan and Heintz, (1992) and Enneking et al. (1996).

Figure 3 shows that the swelling of the polyurethane membranes in pure benzene is greater than 50%, contrasting with the very low swelling in n-hexane, about 6 %. All polyurethane membranes presented a similar sorption behaviour: equilibrium sorption increasing monotonically with increasing benzene concentration. The membrane TPU-2 showed the smallest swelling in both benzene and n-hexane. This behaviour is probably due to the minor fraction of soft segments in this membrane, indicating the great affinity of benzene for the soft segments of polyurethane.

Using the total sorption results of the pure components, the Flory-Huggins interaction parameters (ciP) were calculated through Equation 7. The results obtained are shown in Table 2. All of the polyurethane membranes presented a much greater affinity for benzene (ciP@ 0.6) than for n-hexane (ciP@ 2).

 

Table 2: Flory-Huggins Parameters, ciP, at 250 C

Membrane

Benzene

n-Hexane

TPU-1

0.66

2.11

TPU-2

0.96

2.82

TPU-3

0.62

2.18

TPU-4

0.71

2.75

TPU-5

0.61

2.19

TPU-6

0.72

2.27

 

Pervaporation Experiments

From the above sorption characteristics of the polyurethane membranes, one can conclude that these membranes have a great affinity for benzene than for n-hexane, and should fractionate the benzene/n-hexane mixtures by pervaporation.

Pervaporation experiments were performed at room temperature with a 50% w/w benzene/n-hexane mixture. The permeate flux and selectivities obtained in pervaporation for the different membranes are shown in Table 3. In the same table the results of the sorption selectivity factor (Equation 5) are also shown. It can be observed that all membranes are selective toward benzene. The benzene concentration in the permeate is higher than 80% w/w for the six membranes used in the present work. Table 3 also shows that the values for the sorption selectivity factor and the separation factor in pervaporation (Equation 3) are very close, suggesting that sorption is determining step for selectivity in pervaporation experiments.

 

Figure 3: Total sorption in polyurethane films for pure benzene, n-hexane and mixture with 50% w/w. Temperature 250C

 

It can be observed that the permeate flux varies from 0.39 to 3.26 kg/m2h. Membrane TPU-2 has the largest separation factor (aP=5.6), whereas TPU-3 has the largest permeate flux (J10=3.26 kg/m2h). It is interesting to note that membrane TPU-2, with the smallest fraction of soft segments, showed the smallest permeate flux and the largest separation factor. The most permeable membrane, TPU-3, has the greatest fraction of soft segments. Therefore, the soft segments seem to affect the permeate flux, but not selectivity, which remained in the range of 4 - 5. Similar results were obtained by Ohst et al., (1989) and Wolínska (1995). Taking into account a good compromise between flux and selectivity, membrane TPU-3 was selected to study the effect of the operating temperature and feed concentration on the pervaporation of the benzene/n-hexane system.

 

Table 3: Sorption selectivity and pervaporation performance

Membrane

a S

a P

Flux *
J10(Kg/m2h)

TPU-1

5.5

4.6

2.678

TPU-2

6.6

5.6

0.388

TPU-3

5.5

4.6

3.262

TPU-4

6.0

3.8

1.903

TPU-5

4.7

3.8

2.274

TPU-6

4.5

4.4

1.377

Temperature 250C and concentration feed of 50% w/w
* Flux normalized to a membrane thickness of 10m m.

 

RESULTS FOR THE SELECTED MEMBRANE - MEMBRANE TPU-3

Sorption Results as a Function of Feed Composition

The results of sorption experiments, over all the composition range, using the TPU-3 membrane, are presented in Table 4 and Figure 4 as weight fractions of benzene in the polyurethane membrane as a function of the weight fraction of benzene in the outside solution under thermodynamic equilibrium conditions.

 

Table 4: Sorption results with membrane TPU-3 at T=250C

Benzene in
outside solution
XA(% w/w)

Benzene inside polymeric phase
XAm(%w/w)

Sorption selectivity factor
( a S )

Swelling *
S (g/g)

2.35

21.71

11.5

0.04

11.38

59.26

11.3

0.08

14.95

61.81

9.2

0.08

26.84

71.56

7.1

0.14

46.06

84.45

6.4

0.24

88.00

93.60

2.0

0.63

*Swelling: S ( g of solvent / g of dry polymer )

 

 

Figure 4: Total sorption and sorption selectivity factors for polyurethane membrane (TPU-3) as a function of feed composition. Mixture: Benzene/n-Hexane. Temperature: 250C

 

The total sorption increases when the weight fraction of benzene increases in the liquid mixture outside the membrane, whereas the sorption selectivity factor decreases. The total sorption shows an almost ideal behavior, because the mixture is composed of two non-polar compouds. At low benzene concentrations, the polymer swelling is small and high sorption selectivity factors for benzene are obtained. As the benzene concentration in the outside solution increases, the membrane swelling increases as well. This behaviour is due to the better affinity of benzene for the polymer matrix. As a consequence, the sorption selectivity factor decreases.

Pervaporation Results as a Function of Feed Composition

Table 5 and Figure 5 present the permeate flux and the separation factor for polyurethane membrane TPU-3 for different feed compositions. It can be observed that, as the benzene concentration in the feed increases, the permeate flux increases and the selectivity decreases. This behavior is due to the great affinity of polyurethanes for aromatic coumpounds. The membrane exhibits greater swelling at higher benzene contents in the feed and this decreases the overall selectivity. The higher sorption capacity of this component means a higher fractional free volume for polymer-sorbate system. Greater amounts of the other component of lower affinity for the membrane can penetrate and diffuse into the swollen system.

 

Table 5: Pervaporation results with membrane TPU-3 at 250C

Benzene in the feed
XA(% w/w)

Benzene in the permeate
YA(% w/w)

J10(Kg/m2h)*

Pervaporation selectivity factor
a P

8.26

45.61

0.131

9.4

19.35

67.36

0.334

8.6

68.21

88.56

4.589

3.6

47.09

80.48

3.262

4.6

94.46

96.60

       

1.6

* Flux normalized to a membrane thickness of 10 mm

 

 

Figure 5: Permeate flux and selectivity as a function of feed composition. Mixture: Benzene/n-Hexane. Polyurethane membrane (TPU-3). Temperature: 250C

 

Figure 5 indicates that separation is efficient over all the feed composition range, and operanting pervaporation at a low benzene concentrations is better for the selectivity factors, despite the lower permeate flux. It should be noted that there is a distillation azeotropic system at 4.7% benzene.

Influence of Temperature

Using membrane TPU-3 and a feed mixture near the distillation azeotropic composition, pervaporation experiments were carried out in the temperature range between 25 and 56° C. The results are shown in Figure 6. It can be observed (Figure 6a and 6b) that the permeate flux increases with temperature (seven-fold in this temperature range) and follows an Arrhenius type law:

(8)

 

Figure 6: Permeate flux and selectivity as a function of temperature. Mixture: Benzene (8.0% w/w)/n-Hexane. Polyurethane membrane (TPU-3)

 

where J0 is the pre-exponential factor, EP is the apparent activation energy for permeation, R is the gas constant and T is the absolute temperature. This is a classic behaviour in pervaporation experiments.

On the other hand, the increase in temperature has no significant effect on selectivity. The selectivity factor, aP, remains around 8.5.

In pervaporation, the effect of temperature depends on the major factor that governs the membrane selectivity. If it is mainly due to the sorption step, the variation in selectivity will depend on the heat of sorption of the compounds in the membrane. If it is due to a difference in diffusion rates, then the variation in selectivity will depend on the activation energy for diffusion of each component. A higher activation energy of diffusion of the component that permeates preferentially should lead to an increase in selectivity with the temperature.

In the present work, the pervaporation selectivity for benzene/n-hexane mixtures, using polyurethane membranes (see Table 3), is mainly due to the sorption step. This can be confirmed in Figure 7 where results of both sorption and pervaporation of benzene/n-hexane mixtures at 250C are presented. For a given feed composition, the weight fraction of benzene in the permeate has almost the same value obtained in the sorption experiments. Thus, the observed behaviour for selectivity, shown in Figure 6 is probably related to the different heats of sorption of benzene and n-hexane in the polyurethane membranes.

 

Figure 7: Weight fraction of benzene in the permeate (pervaporation) and in the membrane (sorption) as a function of the weight fraction of benzene in the feed. Polyurethane membrane (TPU-3). Mixture: Benzene/n-Hexane. Temperature: 250C

 

In Figure 7 the pervaporation operating curve is also compared with the liquid-vapor equilibrium curve, showing that the pervaporation process is more efficient than the distillation process in all range of composition, in addition to exhibiting reversed selectivities (in distillation, the vapor phase is enriched in n-hexane, the more volative component).

A comparison of the results for the polyurethane membrane TPU-3 with other works reported in the literature is presented in Table 6. For a 50% w/w of benzene in a benzene/n-hexane system, the polyurethane membranes used in this work show better performance and more promising results.

 

Table 6: Comparison with literature results

A

B

Composition
A (% w/w)

Temp
°C

aA/B

J10
Kg/m2h

Membrane

Benzene

Cyclohexane

50

20

3.0

6.33

Styrene/acrylic acid*

Benzene

n-Hexane

60

50

9.1

0.90

Poly(ether imide)**

Benzene

n-Hexane

60

50

3.0

0.003

PVA***

Toluene

Heptane

60

30

1.4

1.37

Polyethylene****

Benzene

n-Hexane

50

30

1.6

2.40

polyethylene-g-styrene****

Benzene

n-Hexane

50

25

4.6

3.20

This work (TPU-3)

* Ref. Sun and Ruckenstein (1995)
** Ref. Tanihara et al. (1995)
*** Yamasaki et al.(1997)
**** Ref. Aptel et al. (1976)

 

CONCLUSION

The separation of benzene/n-hexane mixtures is suitable using the polyurethane membranes prepared here. Benzene is the component preferentially enriched in the permeate. Sorption is the determining step of the selectivity in the pervaporation process. Membrane TPU-3 was the best membrane tested in this work, with selectivity of 4.6 and permeate flux of 3.2 kg/m2h. When the temperature was varied from 25 to 560 C, using this membrane and a feed mixture with 8.0% w/w of benzene, almost no influence was observed in selectivity, whereas the permeate flux increased seven times.

The pervaporation process showed more efficiency than distillation over all the composition range.

 

NOMENCLATURE

A Membrane area, m2

a Activity

J Permeate flux, Kg/hm2

J10 Normalized permeate flux, Kg/hm2

m Mass, Kg

S Swelling ratio

T Time, h

X Weight fraction in the feed

Y Weight fraction in the permeate

 

Greek letters

a Separation factor

c Flory-Huggins interaction parameter

r density, g/cm3

 

Super/subscripts

m membrane

A, B Components ( Benzene/n-Hexane)

s relative to sorption selectivity factor

p relative to pervaporation

 

REFERENCES

Aithal, U. S., Aminabhavi, T. M. and Balundgi, R. H., Interactions of Organic Solvents with Polyurethane. JMS-Rev. Macromol. Chem. Phys., C30(1), pp. 41-105 (1990).        [ Links ]

Aptel, P., Challard, N., Cuny, J. and Neel, J., Application of the Pervaporation Process to Separate Azeotropic Mixtures. Journal of Membrane Science, 1 271-287 (1976).        [ Links ]

Buijs, H. C. W. M., Wagener, R., Blume, I. and Tinnemans, A. H. A., Composite Membranes with Alkylphosphonated PPO as Selective Top Layer Material in Pervaporation of Aromatics/Non-aromatics. In: Proceedings of Euromembrane`95, Vol.1, pp. I 331 (1995).        [ Links ]

Cunha, V. S.; Membranas de Poliuretano para Uso em Pervaporação de Misturas de Líquidos Orgânicos. Master`s thesis. COPPE/Universidade Federal do Rio de Janeiro (1997).        [ Links ]

Enneking, L., Heintz, A. and Lichtenthaler, R. N., Sorption Equilibria of the Ternary Mixture Benzene/Cyclohexene/Cyclohexane in Polyurethane and PEBA-Membranes Polymers. Journal of Membrane Science, vol 115, pp. 161-170 (1996).        [ Links ]

Flory, P. J., Principles of Polymer Chemistry. Chapter 12, Cornell University Press, Ithaca, NY, (1953).        [ Links ]

Luo, G., Niang, M. and Schaetzel, P., Pervaporation Separation of Ethyl Tert-Butyl Ether and Ethanol Mixtures with a Blended Membrane. Journal of Membrane Science, 125, pp 237-244 (1997).        [ Links ]

Néel, J., Pervaporation. Chapter 5 in Membrane Separation Technology. Edited by Richard D. Noble and S. Alexander Stern. Elsevier, pp 143-204 (1995).        [ Links ]

Niang, M., Luo, G. and Schaetzel, P., Pervaporation Separation of Methyl Tert-Butyl Ether/Methanol Mixtures Using a High-Performance Blended Membrane. J. Appl. Polymer Sci., vol. 64, number 5, pp 875-882 (1997).        [ Links ]

Ohst, H., Hildenbrand, K. and Dhein, R., Polymer Structure/Properties-Correlation of Polyurethane PV-Membranes for Aromatic/Aliphatic Separation. Proceedings of Fifth International Conference on Pervaporation Processes in the Chemical Industry (1989).        [ Links ]

Stephan, W. and Heintz, A., Separation of Aliphatic/Aromatic Mixtures by Pervaporation Using Polyurethane Membranes. Model Calculations and Comparison with Experimental Results. Proceedings of Sixth International Conference on Pervaporation Process in the Chemical Industry, Ottawa, Canada, September (1992).        [ Links ]

Sun, F. and Ruckenstein, E., Sorption and Pervaporation of Benzene/Cyclohexane Mixtures Through Composite Membranes Prepared via Concentrated Emulsion Polymerization. Journal of Membrane Science, 99, pp. 273-284 (1995).        [ Links ]

Tanihara, N., Umeo, N., Kawabata, T., Tanaka, K., Kita, H. and Okamoto, K., Pervaporation of Organic Liquid Mixtures Through Poly(ether imide) Segmented Copolymer Membranes. Journal of Membrane Science, 104, pp. 181-192 (1995).        [ Links ]

Wijmans, J. G. and Baker, R. W., The Solution-Diffusion Model: A Review. Journal of Membrane Science, 107, pp. 1-21 (1995).        [ Links ]

Wolínska-Grabczyk, A. and Jankowski, A., Pervaporation Properties of Liquid Crystalline Polyurethane Membranes Towards Organic-Organic Mixtures. In: Euromembrane`95, Ed. W. R. Bowen, R. W. Field and J. A. Howell, University of Bath, Vol. 2, pp. II-267 (1995).        [ Links ]

Wolínska-Grabczyk, A., Preparation and Characterization of Novel Polyurethane Based Materials for Membrane Organic Liquid Separation. In: Euromembrane`95, Ed. W. R. Bowen, R. W. Field and J. A. Howell, University of Bath, Vol. 1, pp.I-406 (1995).        [ Links ]

Yamasaki, A., Shinbo, T. and Mizoguchi, K., Pervaporation of Benzene/Cyclohexane and Benzene/n-Hexane Mixtures through PVA Membranes. J. Appl. Polyner Sci., vol 64, n° 5, pp. 1061-1065 (1997).        [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License