- Email: [email protected]
New applications of kenaf (Hibiscus cannabinus L.) as microfiltration membranes
Journal of Membrane Science 315 (2008) 141–146
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
New applications of kenaf (Hibiscus cannabinus L.) as microfiltration membranes C.L. Radiman a,∗ , S. Widyaningsih b , S. Sugesty c a b c
Inorganic and Physical Chemistry Division, Faculty of Mathematics and Natural Sciences, Bandung Institute of Technology, Jalan Ganesha 10, Bandung, Indonesia Department of Chemistry, University of Jenderal Sudirman, Jalan H.R. Boenyamin 708, Purwokerto, Indonesia Center for Pulp and Paper, Jalan Dayeuh Kolot 132, Bandung, Indonesia
a r t i c l e
i n f o
Article history: Received 28 September 2007 Received in revised form 11 February 2008 Accepted 13 February 2008 Available online 19 February 2008 Keywords: Kenaf Hibiscus cannabinus L. Cellulose acetate Microfiltration membrane
a b s t r a c t An attempt to explore the possibility of kenaf (Hibiscus cannabinus L.) as microfiltration membrane has been carried out in this work. The pulp was acetylated by anhydride acetic acid to produce cellulose acetate with an acetyl content of 40.40%. Membranes were prepared by phase inversion method using polymer concentrations varied between 14% and 18% (w/w). It was found that membrane composed of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone showed a water flux of 122.29 L/m2 h under an applied pressure of 1 kgf/cm2 , while the rejection towards dextran T-2000 solution was 96.17%. Due to its lower crystallinity index and molecular weight, the acetylated kenaf membrane shows a more porous structure than the one prepared from a commercial cellulose acetate. On the other hand, the Young modulus of acetylated kenaf membrane is higher than the one of commercial cellulose acetate. It is concluded that kenaf as a non-wood plant can be used as alternative raw materials for preparing cellulose acetate microfiltration membranes. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Membrane technology has been developed for several decades after Loeb and Sourirajan discovered the asymmetric membrane preparation for sea water demineralization using cellulose acetate membranes [1]. Until present, cellulose acetate membranes are still made from hardwood. In order to find alternative materials from non-wood plants, this research has been conducted. Compared with hard woods with equivalent cellulose content, non-woods are generally lower in lignin, but higher in silica and ash [2]. So, less chemicals are used in the cooking and bleaching processes. Moreover, in terms of growth rate, non-wood plants take only a few months to reach their full growth while hard wood trees take years. The renewable and sustainable non-wood materials considered in this work is kenaf (Hibiscus cannabinus L.). Kenaf is used for soft fiber, ropes, textiles, paper and automobile industry [3–6]. The American Kenaf Society as well as Japan Kenaf Society reported various new applications of kenaf. This plant is adapted to a wide range of soils and climatic conditions. It grows at temperatures ranging from 15 to 25 ◦ C and can be harvested in 3–4 months from seed. According to Basta et al., membranes prepared from non-wood fibrous materials are still limited [7]. The exploration of non-woods as basic materials should be done in order to find suitable materials for specific membrane applications. So far, no research on the use of kenaf as basic material for synthetic membranes has been
published. Therefore, the objective of this work is to study the possibility of using kenaf (H. cannabinus L.) as the basic material for preparing cellulose acetate membranes used in microfiltration processes. 2. Experimental 2.1. Materials Acetic anhydride 98%, acetone, formamide, glacial acetic acid, hydrochloric acid, sodium hydroxide and concentrated sulfuric acid were obtained from Merck, while dextran T-10, T-40, T-70, T500 and T-2000 were from Sigma. The numbers in the dextran’s code indicate their number–average molecular weight in kg mol−1 . Kenaf was taken from an agricultural experimental station and the pulp was supplied by the Center for Pulp and Paper. Kenaf chips were cooked at 165 ◦ C with a soda-anthraquinone process and bleached with five-stage elemental chlorine free (ECF) bleaching process using oxygen at 95 ◦ C, chlordioxide at 60 ◦ C, extraction and two steps of dioxide at 75 ◦ C. The resulting pulp contained 64.46% alpha-cellulose and 1.22% hemicellulose. The commercial grade cellulose acetate with 40.11% acetyl content was procured from Brataco Ltd., a local trading company. All chemicals were used without further purification. 2.2. Preparation of acetylated kenaf
∗ Corresponding author. Tel.: +62 22 250 2103x2102; fax: +62 22 250 4154. E-mail address: [email protected] (C.L. Radiman). 0376-7388/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2008.02.012
In this experiment, the activation stage was carried out by mixing and stirring 10 g of bleached kenaf pulp and 24 mL of glacial
142
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Table 1 Variation of dope composition Cellulose acetate (wt.%)
Acetone (wt.%)
Formamide (wt.%)
14 16 18
59 57 55
27 27 27
acetic acid at 40 ◦ C for 60 min. Then 40 mL of glacial acetic acid and 0.09 mL of concentrated sulfuric acid as catalyst were added and stirred again for 45 min at the same temperature. The mixture was then cooled until its temperature reached 18 ◦ C and 27 mL of acetic anhydride 98% was added. Another mixture containing 40 mL of glacial acetic acid and 0.6 mL of concentrated sulfuric acid was added into the first mixture and stirred at 40 ◦ C for 20 h. This acetylation stage was then followed by the hydrolysis stage. A solution of 30 mL of acetic acid 67% was added drop by drop within 2 h at 38 ◦ C. The hydrolysis reaction was allowed to continue for 20 h. The product was poured into water with strong agitation and the precipitate was washed with water until the pH became neutral and finally dried at 50 ◦ C. 2.3. Characterization of cellulose acetate The acetyl content of the synthesized cellulose acetate was determined by volumetric method using sodium hydroxide and hydrochloric acid solutions described previously [8]. The viscosity–average molecular weight (Mv ) of cellulose acetate was determined in acetone as solvent by viscometry method using the Mark–Houwink–Sakurada equation: [] = KMva with K = 1.33 × 10−3 and a = 0.616 [9]. The functional groups of the obtained cellulose acetate were analyzed by Fourier Transformed Infra Red (FTIR) spectroscopy using PerkinElmer Fourier Transformed Infra Red spectrophotometer. The crystallinity index and the mean hydrogen bond strength (MHBS) have been calculated according to the method described by O’Connor et al. [10]. So, the crystallinity index was calculated from the ratio of the absorbance of the band maximum at about 1430 cm−1 to the absorbance of the maximum about 900 cm−1 . Meanwhile, the MHBS was calculated from the ratio of the absorbance of O H stretching at about 3390 cm−1 to the absorbance of C H stretching at about 2940 cm−1 . The diffraction pattern was observed by a Jeol diffractometer using a reflection method with a Cu K␣ line. As comparison, similar characterizations were also carried out for purchased cellulose acetate, called afterwards as commercial cellulose acetate.
membranes [12,13]. The thickness of the produced membranes were about 0.22 ± 0.02 mm and kept constant for all formulations. The membranes were subsequently stored in 1 ppm sodium azide solution to prevent microbial growth. 2.5. Characterization of acetylated kenaf membrane The flux and rejection of the produced membranes were measured in a dead-end test cell under a constant applied pressure ranging from 1 to 3 kgf/cm2 . The method of these measurements has been described previously [14]. In order to obtain the molecular weight cut-off (MWCO), the rejection of each dextran was plotted against the logarithm of corresponding dextran’s molecular weight. The MWCO value is determined as the molecular weight at which the rejection is 95%. Membrane cross-sections were observed by scanning electron microscope Jeol JSM-6360LA using gold as the coating agent. The mechanical properties of the obtained membranes were measured by using Autograph Shimadzu AGS-500D. 3. Results and discussion 3.1. Characterization of acetylated kenaf pulp The FTIR spectrum of both commercial cellulose acetate and acetylated product of kenaf pulp in Fig. 1(A) and (B), respectively, show qualitatively the existence of characteristic carbonyl peaks at 1752 and 1237 cm−1 . So, as the spectra of the original kenaf in Fig. 1(C) has no peaks at those regions, it shows that acetyl groups appeared after acetylation. This result is also confirmed by the acetyl content of acetylated kenaf which is 40.40%, while the commercial cellulose acetate has 40.11%. According to Baker, this value refers to cellulose diacetate with acetyl contents ranging from 35% to 43.5% [15]. It can be concluded that the acetylation condition of kenaf pulp is suitable for obtaining cellulose diacetate. It is known that cellulose is a very crystalline substance due to its linear structure and multiple intermolecular hydrogen bonds [16]. During the acetylation process, the swelling agent diffuses
2.4. Preparation of cellulose acetate membranes Based on our previous results, casting solutions composed of cellulose acetate, acetone and formamide were used in this work [11]. The compositions of the mixtures are varied according to Table 1. Each mixture was stirred for 24 h at room temperature until it became a homogeneous dope solution. The dope was allowed to stand for several hours in air tight condition to get rid of air bubbles, then cast on a glass plate and after a partial evaporation for 10 s in the atmospheric condition, the glass plate was gently immersed into cold water at 4 ◦ C. The membrane was gradually formed and, after a complete precipitation, washed with deionized water for several hours until all the solvent and additive have been removed. Casting and gelation conditions were kept constant through all membrane preparations since thermodynamic conditions would largely affect the morphology and performance of the resulting
Fig. 1. FTIR spectrum of: (A) commercial cellulose acetate, (B) acetylated kenaf and (C) kenaf pulp.
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Fig. 2. XRD diffractogram of acetylated kenaf.
inside the interfibrillar spaces of cellulose molecules and the acetylating agent reacts with the hydroxyl groups to form the acetyl groups. The intermolecular hydrogen bonds in the acetylated cellulose decrease and consequently, the product becomes more amorphous than the original cellulose. This is, in fact, supported by the decrease of crystallinity index from 2.31 to 0.38 for the original kenaf and the acetylated one, respectively. Meanwhile, their MHBS value also decreases from 4.87 to 2.98, respectively. Fig. 2 shows the XRD diffractogram of acetylated kenaf. As calculated previously, the crystallinity index of kenaf decreased after acetylation. In fact, the XRD diffractograms of acetylated kenaf and commercial cellulose acetate shown in Figs. 2 and 3, respectively, confirm that broad amorphous peaks are observed in both compounds. Although these diffractograms have similar pattern, there is a slight difference in the region of 2 between 8◦ and 15◦ , which is attributed to cellulose I [7,17,18]. The intensity of those peaks in commercial cellulose acetate is slightly higher than in the acetylated kenaf, showing that the latter is less crystalline than the first one. A quantitative analysis of the IR spectra revealed that the crystallinity index and MHBS value of commercial acetate are 0.41 and 3.02, respectively, while the acetylated kenaf has a crystallinity index of 0.38 and a MHBS of 2.98. According to Ueda et al., some residual hemicelluloses are present in wood pulps as impurities that are chemically or physically resistive to cooking reactions. Therefore, lower crystallinity index of acetylated kenaf might come from the presence of hemicelluloses which also
Fig. 3. XRD diffractogram of commercial cellulose acetate.
143
Fig. 4. Effects of cellulose acetate concentration on water flux under an applied pressure of 1 kgf/cm2 .
result in false viscosity of cellulose acetate in acetone solutions [19]. The determination of viscosity–average molecular weight by Mark–Houwink–Sakurada equation shows that the Mv of acetylated kenaf and commercial cellulose acetate were 1.15 × 104 and 2.4 × 104 , respectively. It proves that the molecular weight of acetylated kenaf is lower than the commercial cellulose acetate due to the presence of hemicelluloses. This data correlates well with the slight difference of their X-ray difractogram patterns and the corresponding crystallinity index obtained previously. In the earlier work, Radiman and Wafiroh showed that the molecular weight of the obtained cellulose acetate depends strongly on the acetylation condition [20].
3.2. Effects of cellulose acetate concentration on permeability and permselectivity of acetylated kenaf membranes Fig. 4 shows that at the same membrane composition, water flux of acetylated kenaf is much higher than that of commercial cellulose acetate. It is well known that high surface porosity of the membrane enhances permeate flux [21]. Regarding the characteristics discussed in the previous section, molecular weight of the polymer can affect the formation of pore structures. So, lower molecular weight results in more porous membranes. Fig. 4 also shows the effects of cellulose acetate concentration on membrane permeability. It is found that higher polymer concentration forms denser pores and results in lower permeability. It can also be seen that higher additive concentration results in higher flux. According to Arthanareeswaran et al. [22], the hydrophylic additive attracts water molecules to enter the membrane matrix. On the other hand, Fig. 5 shows that higher cellulose acetate concentration increases the rejection of dextran T-2000, which was used as one of the standard materials for permselectivity determination. The factors influencing the porosity and pore size of the membranes are complex [23]. Beside polymer concentration, the pore former plays also a key role for the formation of pore sizes in the asymmetric membranes. Its concentration strongly influences the resulting membrane characterizations. Regarding the flux and rejection values in the range of formulations studied in this work, it can be seen that an increase in concentration of cellulose acetate in the casting solution from 14 to 18 wt.% decreases the water flux from 122.3 to 42.8 L/m2 h, but increases the rejection of dextran T-2000 insignificantly from 96.2% to 98.5%. Hence, the dope composition of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone was then chosen for further study.
144
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Fig. 5. Effects of cellulose acetate concentration on rejection of dextran T-2000 under an applied pressure of 1 kgf/cm2 .
3.3. Effects of applied pressure on water flux and rejection of acetylated kenaf membranes Fig. 6 shows that at a given applied pressure the water flux of acetylated kenaf membrane is consistently higher than the one of commercial cellulose acetate. As has been mentioned in some other studies [24–26], the effects of pressure on water flux confirms that higher pressure results in higher flux. According to Mulder [27], higher applied pressure decreases membrane resistance and hence, allows some solutes to permeate through the membranes resulted in lower rejection coefficient. In fact, as seen in Fig. 7, this phenomena is more pronounced in acetylated kenaf membranes. So, both membrane characteristics reveals that acetylated kenaf membranes are more porous than the commercial ones. The presence of hemicelluloses might give looser polymer entanglements resulting in porous membrane. Consequently, the solute molecules are able to permeate more easily through the membrane. It should be noted that acetylated kenaf membrane gives much lower rejection at 3 kgf/cm2 . This finding suggests the presence of loose entanglements among the polymer chains in which the membrane pores becomes more open at high-applied pressure.
Fig. 7. Effects of applied pressure on rejection of dextran T-2000 of membranes with 14% cellulose acetate.
3.4. Molecular weight cut-off of acetylated kenaf membranes The MWCO of membranes corresponds to the value of solute’s molecular weight which rejection is ≥95%. The rejection of dextran tends to increase with its molecular weight due to sieving mechanism in microfiltration which deals not only with pore size but also solute’ size and its molecular weight [23,26,28]. Furthermore, concentration polarization is also a dominant factor which controls the transport of solutes through the membrane [29]. Table 2 shows the molecular weight cut-off of all the prepared membranes. Garcia-Molinaa et al. [25] found that higher concentration of dextran resulted in lower rejection due to the presence of cake layer on the membrane surface which hindered the permeation of water molecules and hence, increased the concentration of solutes in the permeate. So, in order to eliminate the effect of dextran concentration on its rejection values, the concentration of various dextrans was kept constant in all the studies. It can be seen from Table 2 that the MWCO of cellulose acetate from kenaf is higher than the one of commercial membranes. This data suggests that acetylated kenaf membranes have larger pores than the ones prepared from commercial cellulose acetate. The presence of hemicellulose among the cellulose acetate molecules might give looser entanglements and result in large pores. This table also shows that higher applied pressure results in higher MWCO. It indicates that higher pressure enlarges the pores of membranes. According to Whu et al. [24], the effect of applied pressure stays as long as the membrane is under pressure leading to an increased permeate flux. Once the pressure is released, the pore deformation is no longer existed. 3.5. Morphology of acetylated kenaf membranes Figs. 8 and 9 show that both membranes have spongy structure, which is typical for cellulose acetate membranes [25]. However, Table 2 MWCO of acetylated kenaf and commercial cellulose acetate membranes
Fig. 6. Effects of applied pressure on water flux of membranes with 14% cellulose acetate.
Type of membranes
Applied pressure (kgf/cm2 )
MWCO
Acetylated kenaf
1 2 3
7.8 × 105 1.3 × 106 2.4 × 106
Commercial cellulose acetate
1 2 3
2.2 × 105 8.0 × 105 8.9 × 105
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
145
Table 3 Young’s modulus of acetylated kenaf and commercial cellulose acetate membranes Type of cellulose acetate membrane
Cellulose acetate concentration (wt. %)
Young’s modulus (107 N/m2 )
Kenaf
14 16 18
5.38 6.78 6.85
Commercial
14 16 18
5.17 5.40 5.60
4. Conclusions
Fig. 8. SEM photo of acetylated kenaf membrane cross-section.
it can be seen that the cross-section of kenaf membrane is not as dense as the commercial one. This observation is in accordance with the data of membrane characteristics, namely water flux and rejection of dextran T-2000, resulting from the difference of polymer molecular weight and its crystallinity index. This study also confirms the results of Loske et al. [30] who found that the molecular weight of cellulose acetate influenced the spongy structure of membrane morphology. 3.6. Mechanical properties of acetylated kenaf membranes Kenaf fibers have good mechanical properties as they are used for various applications such as ropes, sacks, canvas and carpets [4]. As shown in Table 3, the Young’s modulus of kenaf membranes are actually higher than the commercial ones in all compositions. So, it is a good challenge to develop kenaf applications in membrane technology. Increasing the acetylated kenaf concentration from 14% to 18% (w/w) results in significant increase of Young’s modulus from 5.38 × 107 to 6.85 × 107 N/m2 ; while membranes of commercial cellulose acetate showed only a slight increase. It shows that the structure of kenaf membranes depends strongly on the polymer concentration.
Fig. 9. SEM photo of commercial cellulose acetate membrane cross-section.
This study shows that kenaf as a non-wood plant can be used as alternative raw materials for preparing cellulose acetate membranes applied in pressure-driven processes. Membrane composed of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone shows a water flux of 122.29 L/m2 h under an applied pressure of 1 kgf/cm2 , while the rejection towards dextran T-2000 solution is 96.17%. Due to its lower molecular weight, the acetylated kenaf membrane shows a more porous structure than the one prepared from a commercial cellulose acetate. The Young modulus of acetylated kenaf membranes are higher than the ones of commercial cellulose acetate. Acknowledgements Part of this work has been funded by the Ministry of Research and Technology, Republic of Indonesia under the contract No. 14.09/SK/RUT/2004, for which the authors are very grateful. References [1] S. Loeb, S. Sourirajan, Sea water demineralization by means of an osmotic membrane, in: Presented at the American Chemical Society Meeting, Washington DC, 27 March, 1962. ´ M. Rinaudo, X. Farriol, Synthesis and characterization of [2] C. Barba, D. Montane, carboxymethylcellulose (CMC) from non-wood fibers. I. Accessibility of cellulose fibers and CMC synthesis, Cellulose 9 (2002) 319. [3] T.A. Rymsza, Kenaf and Hemp, appropriateness analysis for paper production, in: Proceedings of the Third Annual American Kenaf Society Conference, Corpus Christi, USA, 23–25 February, 2000. [4] Z. Cheng, B.R. Lu, K. Sameshima, D.X. Fu, J.K. Chen, Identification and genetic relationships of Kenaf (Hibiscus cannabinus L.) germplasm revealed by AFLP analysis, Genetic Resour. Crop Evolut. 51 (2004) 393. [5] S.P. Herath, T. Suzuki, K. Hattori, Multiple shoot regeneration from young shoots of kenaf (Hibiscus cannabinus), Plant Cell Tissue Organ Cult. 77 (2004) 49. [6] C. Zapata, M. Srivatanakul, S.-H. Park, B.-M. Lee, M.G. Salas, R.H. Smith, Improvements in shoot apex regeneration of two fiber crops: cotton and kenaf, Plant Cell Tissue Organ Cult. 56 (1999) 185. [7] A.H. Basta, H. El-Saied, M. Elberry, Cellulose membranes for reverse osmosis. Part II. Improving RO membranes prepared from non-woody cellulose, Desalination 159 (2003) 183. [8] C.L. Radiman, G. Yuliani, Coconut water as a potential resource for cellulose acetate membrane preparation, Polym. Int. 57 (2008) 502. [9] K. Kamide, T. Terakawa, Y. Miyazaki, The viscometric and light-scattering determination of dilute solution properties of cellulose diacetate, Polym. J. 11 (1979) 285. ´ D. Mitcham, Applications of infrared absorption spec[10] R.T. O’Connor, E.F. DuPre, troscopy to investigations of cotton and modified cottons. Part I: Physical and crystalline modifications and oxidation, Text. Res. J. 28 (1958) 382. [11] W. Febrina, C.L. Radiman, Determination of optimum composition of cellulose acetate membrane for microfiltration process, in: Proceedings of the International Conference on Mathematics and Natural Sciences, Bandung, 29–30 November, 2006, pp. 539–542. [12] C. Barth, M.C. Gonc¸alves, A.T.N. Pires, J. Roeder, B.A. Wolf, Asymmetric polysulfone and polyethersulfone membranes: effects of thermodynamics conditions during formation on their performances, J. Membr. Sci. 169 (2000) 287. [13] J.H. Kim, C.K. Kim, Ultrafiltration membranes prepared from blends of polyethersulfone and poly(1-vinylpyrrolidone-co-styrene) copolymers, J. Membr. Sci. 262 (2005) 60. [14] C.L. Radiman, H. Sangkanparan, V.S. Praptowidodo, B.L. Oei, Preparation of asymmetric membranes for desalination, clarification of turbid water and biotechnological down-stream processing, Desalination 93 (1993) 273.
146
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
[15] R.W. Baker, Cellulosic membranes, in: Membrane Technology and Application, Mc.Graw-Hill Co., Inc., New York, 2000, pp. 194–196. [16] C.L. Radiman, A. Putra, Preparation and characterization of Nata-de-coco membranes, in: Presented at The Seventh Pacific Polymer Conference, Oaxaca, Mexico, 3–5 December, 2001. [17] S. Yamanaka, K. Watanabe, N. Kitamura, M. Iguchi, S. Mitsuhashi, Y. Nishi, M. Uryu, The structure and mechanical properties of sheets prepared from bacterial cellulose, J. Mater. Sci. 24 (1989) 3141. [18] M. Iguchi, S. Yamanaka, A. Budhiono, Review bacterial cellulose, a masterpiece of natural’s arts, J. Mater. Sci. 35 (2000) 261. [19] K. Ueda, S. Saka, S. Soejima, Characterizing cellulose acetate prepared by a new acetylation and ripening process, Tappi J. 71 (1988)183. [20] C.L. Radiman, S. Wafiroh, Preparation of cellulose acetate membrane from Abaca (Musa textilis) pulp, in: Proceedings of The Sixth ITB-UKM Joint Seminar on Chemistry, Bali, 17–18 May, 2005, pp. 209–215. ´ ¨ [21] P. Czekaj, F. Lopez, C. Guell, Membrane fouling during microfiltration of fermented beverages, J. Membr. Sci. 166 (2000) 199. [22] G. Arthanareeswaran, P. Thanikaivelan, K. Srinivasn, D. Mohan, M. Rajendran, Synthesis, characterization and thermal studies on cellulose acetate membranes with additive, Eur. Polym. J. 40 (2004)2153.
[23] M. Sivakumar, D.R. Mohan, R. Rangarajan, Studies on cellulose acetatepolysulfone ultrafiltration membranes. II. Effect of additive concentration, J. Membr. Sci. 268 (2006) 208. [24] J.A. Whu, B.C. Baltzis, K.K. Sirkar, Nanofiltration studies of larger organic microsolutes in methanol solutions, J. Membr. Sci. 170 (2000) 159. [25] V. Garc´ıa-Molinaa, S. Esplugas, T. Wintgens, T. Melina, Ultrafiltration of aqueous solutions containing dextran, Desalination 188 (2006) 217. [26] M. Sivakumar, D.R. Mohan, R. Rangarajan, Studies on cellulose acetatepolysulfone ultrafiltration membranes. I. Effect of polymer composition, Polym. Int. 54 (2005) 956. [27] M. Mulder, Separation processes, in: Basic Principles of Membrane Technology, second ed., Kluwer Academic Publisher, Dordrecht, 1996, pp. 214–225. [28] G. Grund, C.W. Robinson, B.R. Glick, Protein type effects on steady state cross flow membrane ultrafiltration fluxes and protein transmission, J. Membr. Sci. 70 (1992) 177. [29] S.K. Zaidi, A. Kumar, Effects of ethanol concentration on flux and gel formation in dead end ultrafiltration of PEG and dextran, J. Membr. Sci. 237 (2004) 189. [30] S. Loske, M. do Carmo Gonc¸alves, B.A. Wolf, Fractionation of cellulose acetate for the investigation of molecular weight influences on the morphology of membranes, J. Membr. Sci. 214 (2003) 223.
Contents lists available at ScienceDirect
Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci
New applications of kenaf (Hibiscus cannabinus L.) as microfiltration membranes C.L. Radiman a,∗ , S. Widyaningsih b , S. Sugesty c a b c
Inorganic and Physical Chemistry Division, Faculty of Mathematics and Natural Sciences, Bandung Institute of Technology, Jalan Ganesha 10, Bandung, Indonesia Department of Chemistry, University of Jenderal Sudirman, Jalan H.R. Boenyamin 708, Purwokerto, Indonesia Center for Pulp and Paper, Jalan Dayeuh Kolot 132, Bandung, Indonesia
a r t i c l e
i n f o
Article history: Received 28 September 2007 Received in revised form 11 February 2008 Accepted 13 February 2008 Available online 19 February 2008 Keywords: Kenaf Hibiscus cannabinus L. Cellulose acetate Microfiltration membrane
a b s t r a c t An attempt to explore the possibility of kenaf (Hibiscus cannabinus L.) as microfiltration membrane has been carried out in this work. The pulp was acetylated by anhydride acetic acid to produce cellulose acetate with an acetyl content of 40.40%. Membranes were prepared by phase inversion method using polymer concentrations varied between 14% and 18% (w/w). It was found that membrane composed of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone showed a water flux of 122.29 L/m2 h under an applied pressure of 1 kgf/cm2 , while the rejection towards dextran T-2000 solution was 96.17%. Due to its lower crystallinity index and molecular weight, the acetylated kenaf membrane shows a more porous structure than the one prepared from a commercial cellulose acetate. On the other hand, the Young modulus of acetylated kenaf membrane is higher than the one of commercial cellulose acetate. It is concluded that kenaf as a non-wood plant can be used as alternative raw materials for preparing cellulose acetate microfiltration membranes. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Membrane technology has been developed for several decades after Loeb and Sourirajan discovered the asymmetric membrane preparation for sea water demineralization using cellulose acetate membranes [1]. Until present, cellulose acetate membranes are still made from hardwood. In order to find alternative materials from non-wood plants, this research has been conducted. Compared with hard woods with equivalent cellulose content, non-woods are generally lower in lignin, but higher in silica and ash [2]. So, less chemicals are used in the cooking and bleaching processes. Moreover, in terms of growth rate, non-wood plants take only a few months to reach their full growth while hard wood trees take years. The renewable and sustainable non-wood materials considered in this work is kenaf (Hibiscus cannabinus L.). Kenaf is used for soft fiber, ropes, textiles, paper and automobile industry [3–6]. The American Kenaf Society as well as Japan Kenaf Society reported various new applications of kenaf. This plant is adapted to a wide range of soils and climatic conditions. It grows at temperatures ranging from 15 to 25 ◦ C and can be harvested in 3–4 months from seed. According to Basta et al., membranes prepared from non-wood fibrous materials are still limited [7]. The exploration of non-woods as basic materials should be done in order to find suitable materials for specific membrane applications. So far, no research on the use of kenaf as basic material for synthetic membranes has been
published. Therefore, the objective of this work is to study the possibility of using kenaf (H. cannabinus L.) as the basic material for preparing cellulose acetate membranes used in microfiltration processes. 2. Experimental 2.1. Materials Acetic anhydride 98%, acetone, formamide, glacial acetic acid, hydrochloric acid, sodium hydroxide and concentrated sulfuric acid were obtained from Merck, while dextran T-10, T-40, T-70, T500 and T-2000 were from Sigma. The numbers in the dextran’s code indicate their number–average molecular weight in kg mol−1 . Kenaf was taken from an agricultural experimental station and the pulp was supplied by the Center for Pulp and Paper. Kenaf chips were cooked at 165 ◦ C with a soda-anthraquinone process and bleached with five-stage elemental chlorine free (ECF) bleaching process using oxygen at 95 ◦ C, chlordioxide at 60 ◦ C, extraction and two steps of dioxide at 75 ◦ C. The resulting pulp contained 64.46% alpha-cellulose and 1.22% hemicellulose. The commercial grade cellulose acetate with 40.11% acetyl content was procured from Brataco Ltd., a local trading company. All chemicals were used without further purification. 2.2. Preparation of acetylated kenaf
∗ Corresponding author. Tel.: +62 22 250 2103x2102; fax: +62 22 250 4154. E-mail address: [email protected] (C.L. Radiman). 0376-7388/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.memsci.2008.02.012
In this experiment, the activation stage was carried out by mixing and stirring 10 g of bleached kenaf pulp and 24 mL of glacial
142
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Table 1 Variation of dope composition Cellulose acetate (wt.%)
Acetone (wt.%)
Formamide (wt.%)
14 16 18
59 57 55
27 27 27
acetic acid at 40 ◦ C for 60 min. Then 40 mL of glacial acetic acid and 0.09 mL of concentrated sulfuric acid as catalyst were added and stirred again for 45 min at the same temperature. The mixture was then cooled until its temperature reached 18 ◦ C and 27 mL of acetic anhydride 98% was added. Another mixture containing 40 mL of glacial acetic acid and 0.6 mL of concentrated sulfuric acid was added into the first mixture and stirred at 40 ◦ C for 20 h. This acetylation stage was then followed by the hydrolysis stage. A solution of 30 mL of acetic acid 67% was added drop by drop within 2 h at 38 ◦ C. The hydrolysis reaction was allowed to continue for 20 h. The product was poured into water with strong agitation and the precipitate was washed with water until the pH became neutral and finally dried at 50 ◦ C. 2.3. Characterization of cellulose acetate The acetyl content of the synthesized cellulose acetate was determined by volumetric method using sodium hydroxide and hydrochloric acid solutions described previously [8]. The viscosity–average molecular weight (Mv ) of cellulose acetate was determined in acetone as solvent by viscometry method using the Mark–Houwink–Sakurada equation: [] = KMva with K = 1.33 × 10−3 and a = 0.616 [9]. The functional groups of the obtained cellulose acetate were analyzed by Fourier Transformed Infra Red (FTIR) spectroscopy using PerkinElmer Fourier Transformed Infra Red spectrophotometer. The crystallinity index and the mean hydrogen bond strength (MHBS) have been calculated according to the method described by O’Connor et al. [10]. So, the crystallinity index was calculated from the ratio of the absorbance of the band maximum at about 1430 cm−1 to the absorbance of the maximum about 900 cm−1 . Meanwhile, the MHBS was calculated from the ratio of the absorbance of O H stretching at about 3390 cm−1 to the absorbance of C H stretching at about 2940 cm−1 . The diffraction pattern was observed by a Jeol diffractometer using a reflection method with a Cu K␣ line. As comparison, similar characterizations were also carried out for purchased cellulose acetate, called afterwards as commercial cellulose acetate.
membranes [12,13]. The thickness of the produced membranes were about 0.22 ± 0.02 mm and kept constant for all formulations. The membranes were subsequently stored in 1 ppm sodium azide solution to prevent microbial growth. 2.5. Characterization of acetylated kenaf membrane The flux and rejection of the produced membranes were measured in a dead-end test cell under a constant applied pressure ranging from 1 to 3 kgf/cm2 . The method of these measurements has been described previously [14]. In order to obtain the molecular weight cut-off (MWCO), the rejection of each dextran was plotted against the logarithm of corresponding dextran’s molecular weight. The MWCO value is determined as the molecular weight at which the rejection is 95%. Membrane cross-sections were observed by scanning electron microscope Jeol JSM-6360LA using gold as the coating agent. The mechanical properties of the obtained membranes were measured by using Autograph Shimadzu AGS-500D. 3. Results and discussion 3.1. Characterization of acetylated kenaf pulp The FTIR spectrum of both commercial cellulose acetate and acetylated product of kenaf pulp in Fig. 1(A) and (B), respectively, show qualitatively the existence of characteristic carbonyl peaks at 1752 and 1237 cm−1 . So, as the spectra of the original kenaf in Fig. 1(C) has no peaks at those regions, it shows that acetyl groups appeared after acetylation. This result is also confirmed by the acetyl content of acetylated kenaf which is 40.40%, while the commercial cellulose acetate has 40.11%. According to Baker, this value refers to cellulose diacetate with acetyl contents ranging from 35% to 43.5% [15]. It can be concluded that the acetylation condition of kenaf pulp is suitable for obtaining cellulose diacetate. It is known that cellulose is a very crystalline substance due to its linear structure and multiple intermolecular hydrogen bonds [16]. During the acetylation process, the swelling agent diffuses
2.4. Preparation of cellulose acetate membranes Based on our previous results, casting solutions composed of cellulose acetate, acetone and formamide were used in this work [11]. The compositions of the mixtures are varied according to Table 1. Each mixture was stirred for 24 h at room temperature until it became a homogeneous dope solution. The dope was allowed to stand for several hours in air tight condition to get rid of air bubbles, then cast on a glass plate and after a partial evaporation for 10 s in the atmospheric condition, the glass plate was gently immersed into cold water at 4 ◦ C. The membrane was gradually formed and, after a complete precipitation, washed with deionized water for several hours until all the solvent and additive have been removed. Casting and gelation conditions were kept constant through all membrane preparations since thermodynamic conditions would largely affect the morphology and performance of the resulting
Fig. 1. FTIR spectrum of: (A) commercial cellulose acetate, (B) acetylated kenaf and (C) kenaf pulp.
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Fig. 2. XRD diffractogram of acetylated kenaf.
inside the interfibrillar spaces of cellulose molecules and the acetylating agent reacts with the hydroxyl groups to form the acetyl groups. The intermolecular hydrogen bonds in the acetylated cellulose decrease and consequently, the product becomes more amorphous than the original cellulose. This is, in fact, supported by the decrease of crystallinity index from 2.31 to 0.38 for the original kenaf and the acetylated one, respectively. Meanwhile, their MHBS value also decreases from 4.87 to 2.98, respectively. Fig. 2 shows the XRD diffractogram of acetylated kenaf. As calculated previously, the crystallinity index of kenaf decreased after acetylation. In fact, the XRD diffractograms of acetylated kenaf and commercial cellulose acetate shown in Figs. 2 and 3, respectively, confirm that broad amorphous peaks are observed in both compounds. Although these diffractograms have similar pattern, there is a slight difference in the region of 2 between 8◦ and 15◦ , which is attributed to cellulose I [7,17,18]. The intensity of those peaks in commercial cellulose acetate is slightly higher than in the acetylated kenaf, showing that the latter is less crystalline than the first one. A quantitative analysis of the IR spectra revealed that the crystallinity index and MHBS value of commercial acetate are 0.41 and 3.02, respectively, while the acetylated kenaf has a crystallinity index of 0.38 and a MHBS of 2.98. According to Ueda et al., some residual hemicelluloses are present in wood pulps as impurities that are chemically or physically resistive to cooking reactions. Therefore, lower crystallinity index of acetylated kenaf might come from the presence of hemicelluloses which also
Fig. 3. XRD diffractogram of commercial cellulose acetate.
143
Fig. 4. Effects of cellulose acetate concentration on water flux under an applied pressure of 1 kgf/cm2 .
result in false viscosity of cellulose acetate in acetone solutions [19]. The determination of viscosity–average molecular weight by Mark–Houwink–Sakurada equation shows that the Mv of acetylated kenaf and commercial cellulose acetate were 1.15 × 104 and 2.4 × 104 , respectively. It proves that the molecular weight of acetylated kenaf is lower than the commercial cellulose acetate due to the presence of hemicelluloses. This data correlates well with the slight difference of their X-ray difractogram patterns and the corresponding crystallinity index obtained previously. In the earlier work, Radiman and Wafiroh showed that the molecular weight of the obtained cellulose acetate depends strongly on the acetylation condition [20].
3.2. Effects of cellulose acetate concentration on permeability and permselectivity of acetylated kenaf membranes Fig. 4 shows that at the same membrane composition, water flux of acetylated kenaf is much higher than that of commercial cellulose acetate. It is well known that high surface porosity of the membrane enhances permeate flux [21]. Regarding the characteristics discussed in the previous section, molecular weight of the polymer can affect the formation of pore structures. So, lower molecular weight results in more porous membranes. Fig. 4 also shows the effects of cellulose acetate concentration on membrane permeability. It is found that higher polymer concentration forms denser pores and results in lower permeability. It can also be seen that higher additive concentration results in higher flux. According to Arthanareeswaran et al. [22], the hydrophylic additive attracts water molecules to enter the membrane matrix. On the other hand, Fig. 5 shows that higher cellulose acetate concentration increases the rejection of dextran T-2000, which was used as one of the standard materials for permselectivity determination. The factors influencing the porosity and pore size of the membranes are complex [23]. Beside polymer concentration, the pore former plays also a key role for the formation of pore sizes in the asymmetric membranes. Its concentration strongly influences the resulting membrane characterizations. Regarding the flux and rejection values in the range of formulations studied in this work, it can be seen that an increase in concentration of cellulose acetate in the casting solution from 14 to 18 wt.% decreases the water flux from 122.3 to 42.8 L/m2 h, but increases the rejection of dextran T-2000 insignificantly from 96.2% to 98.5%. Hence, the dope composition of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone was then chosen for further study.
144
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
Fig. 5. Effects of cellulose acetate concentration on rejection of dextran T-2000 under an applied pressure of 1 kgf/cm2 .
3.3. Effects of applied pressure on water flux and rejection of acetylated kenaf membranes Fig. 6 shows that at a given applied pressure the water flux of acetylated kenaf membrane is consistently higher than the one of commercial cellulose acetate. As has been mentioned in some other studies [24–26], the effects of pressure on water flux confirms that higher pressure results in higher flux. According to Mulder [27], higher applied pressure decreases membrane resistance and hence, allows some solutes to permeate through the membranes resulted in lower rejection coefficient. In fact, as seen in Fig. 7, this phenomena is more pronounced in acetylated kenaf membranes. So, both membrane characteristics reveals that acetylated kenaf membranes are more porous than the commercial ones. The presence of hemicelluloses might give looser polymer entanglements resulting in porous membrane. Consequently, the solute molecules are able to permeate more easily through the membrane. It should be noted that acetylated kenaf membrane gives much lower rejection at 3 kgf/cm2 . This finding suggests the presence of loose entanglements among the polymer chains in which the membrane pores becomes more open at high-applied pressure.
Fig. 7. Effects of applied pressure on rejection of dextran T-2000 of membranes with 14% cellulose acetate.
3.4. Molecular weight cut-off of acetylated kenaf membranes The MWCO of membranes corresponds to the value of solute’s molecular weight which rejection is ≥95%. The rejection of dextran tends to increase with its molecular weight due to sieving mechanism in microfiltration which deals not only with pore size but also solute’ size and its molecular weight [23,26,28]. Furthermore, concentration polarization is also a dominant factor which controls the transport of solutes through the membrane [29]. Table 2 shows the molecular weight cut-off of all the prepared membranes. Garcia-Molinaa et al. [25] found that higher concentration of dextran resulted in lower rejection due to the presence of cake layer on the membrane surface which hindered the permeation of water molecules and hence, increased the concentration of solutes in the permeate. So, in order to eliminate the effect of dextran concentration on its rejection values, the concentration of various dextrans was kept constant in all the studies. It can be seen from Table 2 that the MWCO of cellulose acetate from kenaf is higher than the one of commercial membranes. This data suggests that acetylated kenaf membranes have larger pores than the ones prepared from commercial cellulose acetate. The presence of hemicellulose among the cellulose acetate molecules might give looser entanglements and result in large pores. This table also shows that higher applied pressure results in higher MWCO. It indicates that higher pressure enlarges the pores of membranes. According to Whu et al. [24], the effect of applied pressure stays as long as the membrane is under pressure leading to an increased permeate flux. Once the pressure is released, the pore deformation is no longer existed. 3.5. Morphology of acetylated kenaf membranes Figs. 8 and 9 show that both membranes have spongy structure, which is typical for cellulose acetate membranes [25]. However, Table 2 MWCO of acetylated kenaf and commercial cellulose acetate membranes
Fig. 6. Effects of applied pressure on water flux of membranes with 14% cellulose acetate.
Type of membranes
Applied pressure (kgf/cm2 )
MWCO
Acetylated kenaf
1 2 3
7.8 × 105 1.3 × 106 2.4 × 106
Commercial cellulose acetate
1 2 3
2.2 × 105 8.0 × 105 8.9 × 105
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
145
Table 3 Young’s modulus of acetylated kenaf and commercial cellulose acetate membranes Type of cellulose acetate membrane
Cellulose acetate concentration (wt. %)
Young’s modulus (107 N/m2 )
Kenaf
14 16 18
5.38 6.78 6.85
Commercial
14 16 18
5.17 5.40 5.60
4. Conclusions
Fig. 8. SEM photo of acetylated kenaf membrane cross-section.
it can be seen that the cross-section of kenaf membrane is not as dense as the commercial one. This observation is in accordance with the data of membrane characteristics, namely water flux and rejection of dextran T-2000, resulting from the difference of polymer molecular weight and its crystallinity index. This study also confirms the results of Loske et al. [30] who found that the molecular weight of cellulose acetate influenced the spongy structure of membrane morphology. 3.6. Mechanical properties of acetylated kenaf membranes Kenaf fibers have good mechanical properties as they are used for various applications such as ropes, sacks, canvas and carpets [4]. As shown in Table 3, the Young’s modulus of kenaf membranes are actually higher than the commercial ones in all compositions. So, it is a good challenge to develop kenaf applications in membrane technology. Increasing the acetylated kenaf concentration from 14% to 18% (w/w) results in significant increase of Young’s modulus from 5.38 × 107 to 6.85 × 107 N/m2 ; while membranes of commercial cellulose acetate showed only a slight increase. It shows that the structure of kenaf membranes depends strongly on the polymer concentration.
Fig. 9. SEM photo of commercial cellulose acetate membrane cross-section.
This study shows that kenaf as a non-wood plant can be used as alternative raw materials for preparing cellulose acetate membranes applied in pressure-driven processes. Membrane composed of 14% (w/w) cellulose acetate, 27% (w/w) formamide and 59% (w/w) acetone shows a water flux of 122.29 L/m2 h under an applied pressure of 1 kgf/cm2 , while the rejection towards dextran T-2000 solution is 96.17%. Due to its lower molecular weight, the acetylated kenaf membrane shows a more porous structure than the one prepared from a commercial cellulose acetate. The Young modulus of acetylated kenaf membranes are higher than the ones of commercial cellulose acetate. Acknowledgements Part of this work has been funded by the Ministry of Research and Technology, Republic of Indonesia under the contract No. 14.09/SK/RUT/2004, for which the authors are very grateful. References [1] S. Loeb, S. Sourirajan, Sea water demineralization by means of an osmotic membrane, in: Presented at the American Chemical Society Meeting, Washington DC, 27 March, 1962. ´ M. Rinaudo, X. Farriol, Synthesis and characterization of [2] C. Barba, D. Montane, carboxymethylcellulose (CMC) from non-wood fibers. I. Accessibility of cellulose fibers and CMC synthesis, Cellulose 9 (2002) 319. [3] T.A. Rymsza, Kenaf and Hemp, appropriateness analysis for paper production, in: Proceedings of the Third Annual American Kenaf Society Conference, Corpus Christi, USA, 23–25 February, 2000. [4] Z. Cheng, B.R. Lu, K. Sameshima, D.X. Fu, J.K. Chen, Identification and genetic relationships of Kenaf (Hibiscus cannabinus L.) germplasm revealed by AFLP analysis, Genetic Resour. Crop Evolut. 51 (2004) 393. [5] S.P. Herath, T. Suzuki, K. Hattori, Multiple shoot regeneration from young shoots of kenaf (Hibiscus cannabinus), Plant Cell Tissue Organ Cult. 77 (2004) 49. [6] C. Zapata, M. Srivatanakul, S.-H. Park, B.-M. Lee, M.G. Salas, R.H. Smith, Improvements in shoot apex regeneration of two fiber crops: cotton and kenaf, Plant Cell Tissue Organ Cult. 56 (1999) 185. [7] A.H. Basta, H. El-Saied, M. Elberry, Cellulose membranes for reverse osmosis. Part II. Improving RO membranes prepared from non-woody cellulose, Desalination 159 (2003) 183. [8] C.L. Radiman, G. Yuliani, Coconut water as a potential resource for cellulose acetate membrane preparation, Polym. Int. 57 (2008) 502. [9] K. Kamide, T. Terakawa, Y. Miyazaki, The viscometric and light-scattering determination of dilute solution properties of cellulose diacetate, Polym. J. 11 (1979) 285. ´ D. Mitcham, Applications of infrared absorption spec[10] R.T. O’Connor, E.F. DuPre, troscopy to investigations of cotton and modified cottons. Part I: Physical and crystalline modifications and oxidation, Text. Res. J. 28 (1958) 382. [11] W. Febrina, C.L. Radiman, Determination of optimum composition of cellulose acetate membrane for microfiltration process, in: Proceedings of the International Conference on Mathematics and Natural Sciences, Bandung, 29–30 November, 2006, pp. 539–542. [12] C. Barth, M.C. Gonc¸alves, A.T.N. Pires, J. Roeder, B.A. Wolf, Asymmetric polysulfone and polyethersulfone membranes: effects of thermodynamics conditions during formation on their performances, J. Membr. Sci. 169 (2000) 287. [13] J.H. Kim, C.K. Kim, Ultrafiltration membranes prepared from blends of polyethersulfone and poly(1-vinylpyrrolidone-co-styrene) copolymers, J. Membr. Sci. 262 (2005) 60. [14] C.L. Radiman, H. Sangkanparan, V.S. Praptowidodo, B.L. Oei, Preparation of asymmetric membranes for desalination, clarification of turbid water and biotechnological down-stream processing, Desalination 93 (1993) 273.
146
C.L. Radiman et al. / Journal of Membrane Science 315 (2008) 141–146
[15] R.W. Baker, Cellulosic membranes, in: Membrane Technology and Application, Mc.Graw-Hill Co., Inc., New York, 2000, pp. 194–196. [16] C.L. Radiman, A. Putra, Preparation and characterization of Nata-de-coco membranes, in: Presented at The Seventh Pacific Polymer Conference, Oaxaca, Mexico, 3–5 December, 2001. [17] S. Yamanaka, K. Watanabe, N. Kitamura, M. Iguchi, S. Mitsuhashi, Y. Nishi, M. Uryu, The structure and mechanical properties of sheets prepared from bacterial cellulose, J. Mater. Sci. 24 (1989) 3141. [18] M. Iguchi, S. Yamanaka, A. Budhiono, Review bacterial cellulose, a masterpiece of natural’s arts, J. Mater. Sci. 35 (2000) 261. [19] K. Ueda, S. Saka, S. Soejima, Characterizing cellulose acetate prepared by a new acetylation and ripening process, Tappi J. 71 (1988)183. [20] C.L. Radiman, S. Wafiroh, Preparation of cellulose acetate membrane from Abaca (Musa textilis) pulp, in: Proceedings of The Sixth ITB-UKM Joint Seminar on Chemistry, Bali, 17–18 May, 2005, pp. 209–215. ´ ¨ [21] P. Czekaj, F. Lopez, C. Guell, Membrane fouling during microfiltration of fermented beverages, J. Membr. Sci. 166 (2000) 199. [22] G. Arthanareeswaran, P. Thanikaivelan, K. Srinivasn, D. Mohan, M. Rajendran, Synthesis, characterization and thermal studies on cellulose acetate membranes with additive, Eur. Polym. J. 40 (2004)2153.
[23] M. Sivakumar, D.R. Mohan, R. Rangarajan, Studies on cellulose acetatepolysulfone ultrafiltration membranes. II. Effect of additive concentration, J. Membr. Sci. 268 (2006) 208. [24] J.A. Whu, B.C. Baltzis, K.K. Sirkar, Nanofiltration studies of larger organic microsolutes in methanol solutions, J. Membr. Sci. 170 (2000) 159. [25] V. Garc´ıa-Molinaa, S. Esplugas, T. Wintgens, T. Melina, Ultrafiltration of aqueous solutions containing dextran, Desalination 188 (2006) 217. [26] M. Sivakumar, D.R. Mohan, R. Rangarajan, Studies on cellulose acetatepolysulfone ultrafiltration membranes. I. Effect of polymer composition, Polym. Int. 54 (2005) 956. [27] M. Mulder, Separation processes, in: Basic Principles of Membrane Technology, second ed., Kluwer Academic Publisher, Dordrecht, 1996, pp. 214–225. [28] G. Grund, C.W. Robinson, B.R. Glick, Protein type effects on steady state cross flow membrane ultrafiltration fluxes and protein transmission, J. Membr. Sci. 70 (1992) 177. [29] S.K. Zaidi, A. Kumar, Effects of ethanol concentration on flux and gel formation in dead end ultrafiltration of PEG and dextran, J. Membr. Sci. 237 (2004) 189. [30] S. Loske, M. do Carmo Gonc¸alves, B.A. Wolf, Fractionation of cellulose acetate for the investigation of molecular weight influences on the morphology of membranes, J. Membr. Sci. 214 (2003) 223.