Preparation and application of a novel ethanol permselective poly(vinyltriethoxysilane) membrane

Abstract

A novel thin poly(vinyltriethoxysilane) membrane with hydrophobic Si–O–Si backbone and vinyl groups is proposed to recover ethanol by pervaporation. It exhibits high flux (>10 000 g m⁻² h⁻¹), which is about ten times higher than that of PDMS, demonstrating that this membrane would facilitate the ethanol industrial production by pervaporation–fermentation process.

---

Due to the increasing demand for bioethanol as a gasoline additive or replacement, pervaporation (PV) as an in situ product recovery technique has attracted more and more attention for overcoming ethanol inhibition and lowering the cost of ethanol production.^1–3 The key element in this integrated process is the PV membrane. Separation factor and flux are two important indexes for PV membranes. According to the solution-diffusion mechanism, the separation factor is affected by the interaction between the components and membrane, especially by the functional groups on the membrane surface, while permeation flux is inversely proportional to the active layer thickness.^4–8 Inspired by the facts above, a thin ethanol permselective membrane with hydrophobic groups on surface will contribute to getting a high flux and high separation factor.

Owing to the lack of proper membranes, especially high flux membranes, the PV–fermentation process cannot achieve a breakthrough in industrial-scale application so far. Currently, polydimethylsiloxane (PDMS)^9–15 and poly(1-trimethylsilyl-1-propyne) (PTMSP)^16–19 are the most common hydrophobic polymeric membranes for ethanol recovery. However, the PV performance of PTMSP membrane decreases with the operating time because of the relaxation processes of the polymeric chains.^16,19,20 Although the PDMS with hydrophobic Si–O–Si bonds is considered as a superior membrane material for organics recovery, PDMS membranes prepared by crosslinking in the polymer with large molecules weight are usually thick, which cause low flux. This low flux of PDMS membrane cannot meet the requirement of industrial application. Based on the economic analysis of PV–fermentation process by O'Brien,²¹ membranes with the high flux and the high separation factor are beneficial to lowering the capital cost and the annual cost. To enhance the overall techno-economic feasibility and industrial applicability of PV–fermentation process, the robust high flux hydrophobic PV membranes are indispensable for ethanol recovery.^22,23 Moreover, the methyl (–CH₃) groups on the surface of PDMS membrane are difficult to modify, which limits the improvement of PV performance.^8,24 Therefore, it is necessary to develop a new high-flux membrane for ethanol recovery.

Herein we propose vinyltriethoxysilane (VTES) to prepare a thin high-flux membrane for ethanol recovery, because of its low molecular weight and the vinyl (–CH=CH₂) group, which can be easily modified for improving the membrane performance.

VTES is one type of organic silicone resins with general structure as R_n–Si–X_(4−n), where X is the –OCH₂CH₃ group for hydrolysis reaction, and R is the –CH=CH₂ group. Its structure is depicted in Fig. 1.

Fig. 1 The molecular structures of VTES, PDMS and PVMTS.

The solubility parameters are used to indicate the compatibility of the two materials. The solubility parameter of VTES is 15.9 J^1/2 cm^−3/2, which is closer to that of ethanol (25.9 J^1/2 cm^−3/2) than water (47.9 J^1/2 cm^−3/2). Thus, the affinity between VTES and ethanol is stronger than that between VTES and water, which indicates the solubility of ethanol in VTES is higher than that of water.

The compatibility between the membrane material and the permeable components should be maintained within a certain range. An excessively high compatibility may result in membrane swelling, and then both the stability and the lifetime of membrane will be negatively affected. To the contrary, the too low compatibility leads to the unsatisfactory adsorption for the prior permeable component.

The silanization of organic silicone resin forms a polymer with Si–O–Si bonds.^25,26 The hydrolysis and condensation processes of VTES are illustrated in Fig. 2, and the formed polymer on the surface is poly(vinyltriethoxysilane) (PVTES). The PVTES membrane provides a hydrophobic surface due to the Si–O–Si bonds and –CH=CH₂ groups. The structure of PVTES can also be verified by FTIR spectrum and ²⁹Si NMR spectrum shown in Fig. 3.

Fig. 2 The mechanism for the PVTES membrane formation.

Fig. 3 (a) FTIR spectrum of the PVETS membrane surface FTIR spectrum and (b) ²⁹Si NMR spectrum for PVETS.

The peaks located at ∼1600 cm⁻¹ and 1410 cm⁻¹ are stretching vibration and characteristic vibration of –CH=CH₂ group, while the other characteristic bands at ∼1010 cm⁻¹ and ∼1160 cm⁻¹ are due to the symmetric and asymmetric Si–O–Si stretches, respectively.²⁷ Both Si–O–Si and –CH=CH₂ are verified by the FTIR spectrum of PVETS. In the ²⁹Si NMR spectrum of PVTES, the peaks of −74.03 ppm and −81.12 ppm are marked as T² and T³, and T represents a trifunctional silicon, while the superscripts 2 and 3 represent the number of siloxane bridges connected to the silicon site (connectivity).²⁷ The present of two peaks, T² and T³ suggest that the majority of –OCH₂CH₃ groups in VTES have been silanized to form Si–O–Si bonds. This hydrophobic structure of PVTES would be better than that of plasma-polymerized poly(trimethoxyvinylsilane) (PTMVS) shown in Fig. 1.²⁸

SEM images of the dense thin PVTES membrane are shown in Fig. 4. The surface of PVTES membrane is smooth and the adhesion between the active layer and the support layer is excellent. The thickness of this thin PVTES membrane is approximate 0.3 μm, which is much smaller than that of most PDMS membranes.

Fig. 4 SEM of PVTES surface and cross-section.

As illustrated in Fig. 5, the XRD spectra of PVTES shows two peaks at 2θ of 9.2° and 21.9°, while that of PDMS exhibits only a characteristic peak at 2θ of 11.5°, which is similar to the XRD results obtained by Lue et al.²⁹ The diffraction angle θ can be used to calculate interplanar spacing by Bragg equation, 2d sin θ = nλ, where λ is 0.154 nm and n equals to one.^30,31 Interplanar spacing d₁ represents the distance of oxygen atoms between adjacent molecules, which would affect the diffusion of the components in the membrane. The d₁ of PVTES is 0.96 nm, which is larger than that of PDMS (0.77 nm), suggesting that the components can diffuse easier in the PVTES membrane than that in the PDMS membrane and this possibly triggers a higher flux of PVTES membrane in comparison of PDMS membrane.

Fig. 5 X-ray diffraction spectra of PVTES and PDMS.

The contact angle between the solution and membrane can be used to judge the interaction between component and membrane. While the contact angle gets bigger, the interaction between component and membrane becomes smaller. As observed from Table 1, the contact angle between water and PDMS is bigger than that between water and PVTES, so the interaction between water and PDMS is smaller, which indicates that the hydrophobicity of the PDMS membranes is stronger, probably leading to a higher separation factor for the organics–water solution compared with the PVTES membrane.

Table 1 Results of contact angle measurement at 25 °C

Membrane	PVTES	PDMS
Pure water contact angle/°	100 ± 1	116 ± 1

---

Mechanical properties are important factors to evaluate the application of the membrane. Two composite membranes, PVTES/PVDF and PDMS/PVDF, are compared with the support layer, PVDF membrane. As shown in Table 2, the tensile strength of PVTES/PVDF membrane is higher than that of PVDF and PDMS/PVDF membranes, while the elongation at break of the PVTES/PVDF and PDMS/PVDF is a little bit lower than that of PVDF membrane, which suggests that the mechanical properties PVTES/PVDF membrane is fine and it can be suitable as PV membrane for ethanol recovery.

Table 2 Mechanical properties of PVDF, PVTES/PVDF and PDMS/PVDF

Membrane	PVDF	PVTES/PVDF	PDMS/PVDF
Tensile strength/MPa	29.5 ± 1.8	40.6 ± 1.1	30.3 ± 2.4
Elongation at break/%	49.3 ± 3.4	37.3 ± 2.3	36.0 ± 3.6

---

The objective of this study is to prepare a thin membrane with hydrophobic group on the surface for ethanol recovery. The results addressed above show that PVTES membrane could be a good choice for this requirement. The thin layer of PVTES would lead to a high flux, which could be higher than the flux of PDMS membranes according to the inversely proportional relationship between the thickness of the active layer and the flux.³² Compared to PDMS with –CH₃ as the hydrophobic group on the surface, –CH=CH₂ on the surface of PVTES offers higher solubility of ethanol than –CH₃, and PDMS and PVTES all possess Si–O–Si as the backbone. Moreover, –CH=CH₂ provides a possibility to graft with other functional groups by free radical polymerization, which can easily improve the performance of membrane, thus extends the scope of PVTES application.

Effect of feed concentration on the PV performance of PVTES membrane is shown in Fig. 6. With the increase of the ethanol concentration from 3 wt% to 13 wt%, the total flux increases from 6000 g m⁻² h⁻¹ to 10 000 g m⁻² h⁻¹ and the separation factor is around 5. The results suggest that PVTES membrane can be used to recover ethanol by PV, which agrees well with the conclusion drawn from solubility parameters. As the ethanol concentration in feed increases, the driving force of ethanol through PVTES membrane enhances. The permeation of ethanol molecules becomes much easier which contributes to the increase of the ethanol flux. However, associating effect of hydrogen bond between water and ethanol makes water flux slightly increase with the increase of ethanol flux, which results in a certain value of the separation factor and the increase of total flux. This result does not agree with the conclusion drawn by D. Van Baelen.³³ The inconsistence may be attributed to the short range of the feed concentration and the difference between the membrane materials used.

Fig. 6 Effect of ethanol concentration on the pervaporation performance of PVTES membrane. The feed temperature is 35.0 °C and the volume of feed solution is 2 L.

According to the data shown in Table 3, the separation factor of PVTES membrane (α = 3.8–6.0) is acceptable, which is close to that of PDMS membrane (α = 1.0–10.8), while the highest flux of PVTES membrane is over 10 000 g m⁻² h⁻¹, which is about one order of magnitude higher than that of PDMS membrane (∼1000 g m⁻² h⁻¹). Although the flux of PTMVS membrane prepared by plasma is as high as that of PVTES, the separation factor is lower (∼2.2) than that of PVTES. These phenomena may be resulted from the molecular structural characteristics and the thickness of membrane. Performance comparison between the PVTES membrane and PDMS membrane on the separation of ethanol–water is shown in Fig. 7. The high flux means a low membrane area requirement of for the ethanol recovery per unit weight, which leads to lower capital investment, lower annual cost and a smaller footprint.^8,21 The separation factor and the high flux of PVTES membrane clearly transcend the upper limit of PDMS membrane reported in most literatures (Fig. 7), which implies that PVTES membrane is more economically attractive than PDMS and PTMVS membranes and offers significant potential for PV integrated with ethanol fermentation process.^1,21

Table 3 Pervaporation data published for separation of ethanol from water with pristine PDMS membranes and poly(trimethoxyvinylsilane) (PTMVS) membrane, in comparison with the results of poly(vinyltriethoxysilane) (PVTES) membranes

Membrane	Cf^a [wt%]	T^a [°C]	l^a [μm]	α^a	J^a [g m^−2 h^−1]	Ref.
a Cf-feed ethanol concentration, T-temperature, l-membrane thickness, α-separation factor and J-total flux.
PDMS	5.0	40	5	8.9	1600	9
	5.0	40	1–2	9.3	1140	10
	4.0	45	5	8.5	1850	11
	5.0	40	8	8.5	1300	12
	8.0	42	1	6.7	1440	13
	8.0	50	—	6.4	265	14
	4.0	45	1	5.0	1600	11
	3.0	50	8	1.0	2800	15
	5.0	40	10	8.8	∼240	34
	2.0	30	50	10.0	∼102	35
	4.3	40	<10	6.3	5150	36
	10.0	40	120	7.4	53.3	37
	6.4	30	100	10.8	25.1	38
	3.0	41	12.5 ± 2	4.6	120	39
PTMVS	10.0	25	—	2.2	7650	28
PVTES	5.0	35.0	∼0.3	5.4	7238	This work
	5.0	45.0	∼0.3	6.0	7740	This work
	9.0	35.0	∼0.3	5.2	9783	This work
	9.0	45.0	∼0.3	3.8	14 059	This work

---

Fig. 7 Pervaporation data published for ethanol recovery from water with pristine PDMS membranes and PTMVS membrane, in comparison with the results of PVTES membranes. The dashed line represents the upper limit of the PDMS membrane. The detail information of data point is shown in Table 3.

Conclusions

In conclusion, we prepare a novel thin PVTES membrane by VTES to recover ethanol by PV process. It consists of –CH=CH₂ groups and the hydrophobic Si–O–Si backbone. The results show that the flux of PVTES membrane is about one order of magnitude higher than that of PDMS membrane, which is caused by the thinner membrane and the polymeric structure of PVTES. This PVTES membrane with a promising performance would be applied in PV integrated with fermentation process for ethanol recovery in the future work, and it will facilitate the industrial application of PV integrated with fermentation process. As demonstrated in this study, and considering the various structures and chemical functionalities of organic silicone resins, like VTES, it is believed that the unique properties of organic silicone resins will offer a new perspective on PV membranes.

Acknowledgements

The authors are grateful to the National Natural Science Foundation of China (21276011 and 21276012) and Program for New Century Excellent Talents in University (NCET-10-0210).

Notes and references

1. L. M. Vane, Biofuels, Bioprod. Biorefin., 2008, 2, 553–588.
2. L. M. Vane, J. Chem. Technol. Biotechnol., 2005, 80, 603–629.
3. S. Nakao, F. Saitoh, T. Asakura, K. Toda and S. Kimura, J. Membr. Sci., 1987, 30(3), 273–287.
4. H. Zhang, H. G. Ni, X. P. Wang, X. B. Wang and W. Zhang, J. Membr. Sci., 2006, 281, 626–635.
5. T. Miyata, H. Yamada and T. Uragami, Macromolecules, 2001, 34, 8026–8033.
6. E. A. Fouad and X. Feng, J. Membr. Sci., 2008, 323, 428–435.
7. A. Raisi and A. Aroujalian, Sep. Purif. Technol., 2011, 82, 53–62.
8. X. L. Liu, Y. S. Li, G. Q. Zhu, Y. J. Ban, L. Y. Xu and W. S. Yang, Angew. Chem., Int. Ed., 2011, 50, 10636–10639.
9. W. Wei, S. Xia, G. Liu, X. Dong, W. Jin and N. Xu, J. Membr. Sci., 2011, 375, 334–344.
10. Y. Luo, S. Tan, H. Wang, F. Wu, X. Liu, L. Li and Z. Zhang, Chem. Eng. J., 2008, 137(3), 496–502.
11. E. Shi, W. Huang, Z. Xiao, D. Li and M. Tang, J. Appl. Polym. Sci., 2007, 104(4), 2468–2477.
12. L. Li, Z. Xiao, S. Tan, L. Pu and Z. Zhang, J. Membr. Sci., 2004, 243(1–2), 177–187.
13. W. Zhang, X. Yu and Q. Yuan, Biotechnol. Tech., 1995, 9(4), 299–304.
14. J. Guo, G. Zhang, W. Wu, S. Ji, Z. Qin and Z. Liu, Chem. Eng. J., 2010, 158(3), 558–565.
15. T. Mohammadi, A. Aroujalian and A. Bakhshi, Chem. Eng. Sci., 2005, 60(7), 1875–1880.
16. C. López-Dehesa, J. A. González-Marcos and J. R. González-Velasco, J. Appl. Polym. Sci., 2007, 103, 2843–2848.
17. V. V. Volkov, A. G. Fadeev, V. S. Khotimsky, E. G. Litvinova, Y. A. Selinskaya, J. D. McMillan and S. S. Kelley, J. Appl. Polym. Sci., 2004, 91, 2271–2277.
18. J. A. González-Marcos, C. López-Dehesa and J. R. González-Velasco, J. Appl. Polym. Sci., 2004, 94, 1395–1403.
19. J. R. González-Velasco, C. López-Dehesa and J. A. González-Marcos, J. Appl. Polym. Sci., 2003, 90, 2255–2259.
20. A. G. Fadeev, S. S. Kelley, J. D. McMillan, Y. A. Selinskaya, V. S. Khotimsky and V. V. Volkov, J. Membr. Sci., 2003, 214, 229–238.
21. D. J. O'Brien, L. H. Roth and A. J. McAloon, J. Membr. Sci., 2000, 166, 105–111.
22. S. Claes, P. Vandezande, S. Mullens, K. D. Sitter, R. Peeters and M. K. V. Bael, J. Membr. Sci., 2012, 389, 265–271.
23. S. Claes, P. Vandezande, S. Mullens, R. Leysen, K. De Sitter, A. Andersson, F. H. J. Maurer, H. V. Rul, R. Peeters and M. K. V. Bael, J. Membr. Sci., 2010, 351, 160–167.
24. D. Beaumelle, M. Marin and H. Gibert, Food Bioprod. Process., 1993, 71(C2), 77–89.
25. W. Yoshida, R. P. Castro, J. D. Jou and Y. Cohen, Langmuir, 2001, 17(19), 5882–5888.
26. N. N. Herrera, J. M. Letoffe, J. P. Reymond and E. B. Lami, J. Mater. Chem., 2005, 15(8), 863–871.
27. H. Dong, M. A. Brook and J. D. Brennan, Chem. Mater., 2005, 17(11), 2807–2816.
28. C. L. Chang and M. S. Chang, J. Membr. Sci., 2004, 238, 117–122.
29. S. J. Lue, J. S. Ou, C. H. Kuo, H. Y. Chen and T. H. Yang, J. Membr. Sci., 2010, 347, 108–115.
30. K. Ishida, D. Sasaki, T. Horiuchi and K. Matsushige, Synth. Met., 1997, 91, 371–373.
31. S. Furukawa and K. Takeuch, Solid State Commun., 1993, 87(10), 931–934.
32. L. Sun, G. L. Baker and M. L. Bruening, Macromolecules, 2005, 38, 2307–2314.
33. D. V. Baelen, A. Reyniers, B. V. Bruggen, C. Vandecasteele and J. Degreve, Sep. Sci. Technol., 2004, 39(3), 559–576.
34. X. Zhan, J. Li, J. Huang and C. Chen, Appl. Biochem. Biotechnol., 2010, 160, 632–642.
35. W. D. Zhang, W. Sun, J. Yang and Z. Q. Ren, Appl. Biochem. Biotechnol., 2010, 160, 156–167.
36. F. Xiangli, Y. Chen, W. Jin and N. Xu, Ind. Eng. Chem. Res., 2007, 46, 2224–2230.
37. T. Uragami and T. Morikawa, J. Appl. Polym. Sci., 1992, 44, 2009–2018.
38. K. Ishihara and K. Matsui, J. Appl. Polym. Sci., 1987, 34, 437–440.
39. A. Dobrak, A. Figoli, S. Chovau, F. Galiano, S. Simone, I. F. J. Vankelecom, E. Drioli and B. V. Bruggen, J. Colloid Interface Sci., 2010, 346, 254–264.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, characterization of PVTES membrane. See DOI: 10.1039/c3ra47623e

---