Modified chitosan-based, pH-responsive membrane for protein separation

Abstract

N-Carboxymethyl chitosan (N-CMCS) and O-carboxymethyl chitosan (O-CMCS)-based amphoteric or pH-responsive charged membranes were prepared for protein separation. Both membranes (O-CMCS and N-CMCS) exhibited different charged natures and acidic/alkaline ion-exchange capacities along with 16–19 kDa molecular weight cut-off (MWCO), corresponding to 0.53 nm and 0.48 nm pore radius at pH 7.0 for O-CMCS and N-CMCS, respectively. Water permeability for O-CMCS and N-CMCS membranes varied between 5.0–6.0 × 10⁻⁸ Pa⁻¹ m h⁻¹ (ultra-filter range) and increased with equilibrating medium pH. The dual-charged nature of these membranes (presence of –COOH and –N⁺H₃ groups) was advantageously used to achieve antifouling properties in biomolecule separation. The N-CMCS membrane exhibited a positively charged nature in the 2.0–12.0 pH range. Thus, Donnan exclusion due to mutual electrostatic attraction between the membrane and protein was insignificant in accelerating the mobility of a protein molecule, thus achieving its separation. Meanwhile, the O-CMCS membrane showed pH-tunable charged nature and suitable separation performance for β-casein (β-Cas) and lysozyme (Lys) across alkaline media, as a representative case. Furthermore, no membrane fouling was detected after 50 days of operation in a protein environment.

---

Introduction

Chitosan is a cationic copolymer of glucosamine and N-acetyl glucosamine obtained by deacetylation of chitin. It is widely studied for its pharmaceutical, nonpharmaceutical, hydrogel and membrane applications.^1–3 Chitosan has a unique set of useful characteristics such as bio-renewability, biodegradability, biocompatibility, bioadhesivity and nontoxicity, which make it a polymer of reckon. Natural polymer (chitosan)-based membranes are well accepted for their better biocompatibility and less toxic effect compared to synthetic polymers; they serve as a suitable candidate for biomedical applications and biomolecule separation.^4,5 However, insufficient swelling/wettability of chitosan under neutral physiological conditions and its lack of amphoteric nature have limited its uses.⁶ Numerous chitosan derivatives have been prepared by quaternarization;⁷ by introducing hydrophilic groups like hydroxypropyl, dihydroxyethyl, hydroxyl alkylamino,^8–10 sulphate,¹¹ and phosphate; by adding carboxyalkyl groups such as carboxymethyl, carboxyethyl, and carboxybutyl; or by grafting water-soluble polymers^12,13 to improve its desirable properties and bio-applicability. Among the different derivatives of chitosan, carboxymethyled chitosans (N-carboxymethyl chitosan, (N-CMCS) and O-carboxymethyl chitosan (O-CMCS)) have been widely studied because of their ease of synthesis, amphoteric character, and ample possible applications.¹⁴ Due to the presence of –NH₂ and –COOH groups ionizable to positive and negative charges, both N-CMCS and O-CMCS show pH-dependent charge on the molecule. Thus, it is interesting to investigate the pH-responsive, tunable charged properties of chitosan-derived membranes for the separation of zwitterionic charged proteins. Membrane pore dimension and surface properties can be potentially tuned with new protocols suitable for biomolecule separation.

The pH-responsive, tunable charged nature of the membranes, their stability, rigidity and the fluid mobility of their constituent proteins define the physical environment for the separation of a variety of biomolecules.^15–17 Responsive porous membranes showed tunable permeation/separation properties against external stimuli, such as pH, temperature, pressure, electric field or ionic environment. Ultrafiltration (UF)-based membrane processes with external stimuli, such as pH and pressure, are widely utilized for separation and purification of proteins in the food industry and biotechnology.^18,19 However, membrane fouling (characterized by nonspecific adsorption and deposition) is a serious problem for UF membranes, leading to substantial flux decline.^20,21 Membrane fouling is strongly dependent on protein–membrane and protein–protein interactions, which are affected by a series of factors, such as the surface chemistry of membranes, pH value and ionic strength of the feed solution, the nature of the protein, and the hydrodynamics of the process.^22–26 Charged UF membranes have been reported to avoid the aforementioned problems.^27–30 In these cases, a negative, positive or zwitterionic charged membrane matrix controls the selectivity for charged proteins and reduces the chances of fouling.

Thus, the objective of the present investigation is to modify chitosan into N-CMCS and O-CMCS and develop a pH-responsive, tunably charged, antifouling UF membrane for protein separation. Equimolar mixed solutions of β-Cas and Lys were used as model cases. β-Cas and Lys were selected because of their different natures (pI and molecular weight). β-Cas is a globular protein with 3.9 pI and 20 000 Da molecular weight, while Lys is a linear protein with 9.6 pI and 14 600 Da molecular weight.

Experimental section

Materials and method

Chitosan (100% deacetylation; medium molecular weight), monochloroacetic acid, isopropanol and other chemical reagents were obtained from Sigma-Aldrich. H₂SO₄, NaOH and dimethylformamide (DMF, AR grade) were received from S. D. Fine Chemicals, India. β-Cas (M_w 20 000 Da) and lysozyme (M_w 14 600 Da) were obtained from Hi Media Laboratories, Pvt. Ltd. (India). In all experiments, double distilled water was used.

For the preparations of carboxymethyl chitosan (CMCS), chitosan (1.5 g) was added to a round bottom flask and allowed to swell in 33% NaOH (aqueous) solution at 30 °C for 24 h. Hydrated chitosan was vacuum filtered and transferred to a clean round bottom flask, incubated at 50 °C in water bath for 1 h, and followed by dropwise addition of monochloroacetic acid (20%, w/v) dissolved in 20 mL isopropanol. The solution was stirred for 6 h at 30 °C and then allowed to cool at room temperature. The obtained precipitate was filtered and rinsed with ethanol (100%) and then isopropanol to remove the salts, and then it was vacuum dried at 60 °C. To prepare N-CMCS, CMCS was treated with chloroacetic acid (1 : 10, v/v) for 6 h at 65 °C. The acid residue was removed by washing with ethanol. Further, O-CMCS was prepared by treatment of CMCS with chloroacetic acid (1 : 4, v/v) for 6 h at 30 °C. Preparation procedures for O-CMCS and N-CMCS are described in Scheme 1.

Scheme 1 Synthetic route for carboxymethylation of chitosan and formation of O-CMCS and N-CMCS derivatives at different pH.

For preparing membranes, O-CMCS or N-CMCS was dissolved in DMAC (10%, w/v) under continuous stirring (12 h). The obtained viscous solution was transformed into thin film on a hydrophobized glass plate, dried at 30 °C for 20 min, then dipped in ice-cold water (10 °C) for membrane gelation and pore formation. Average membrane thickness was maintained as 175 μm.

Instrumental analysis and membrane stability

Detailed instrumental characterization has been included in the ESI (Section S1†). The membrane's stability in different solvents was analyzed in terms of weight loss (W_L),where W_d and W_s are the weight of dry and chemically treated dry membrane.

Physicochemical characterization, water flux, protein flux, molecular cut-off and pore radius measurements

Prepared membranes were characterized by estimating their physicochemical properties such as thickness, water content, H⁺/OH⁻ exchange capacities and acidic/alkaline surface charge concentration. The detailed procedure is described in Section S2.† Methods adopted for measuring water flux, protein flux, molecular cut-off and membrane pore radius also are described in Section S3.† Water permeability was estimated from the slope of straight lines obtained from the water flux vs. pressure plot. Water flux was measured at 30 °C between 3.0–8.0 × 10⁵ Pa. For both membranes (N-CMCS and O-CMCS), water permeability was also measured at varied pH (2.0–12.0). Dilute HCl or NaOH (0.10 M) was used to control the water pH.

pH-responsive protein separation

To assess the membranes' performance in discrimination between proteins (Lys and β-Cas as representative cases) and their antifouling nature, protein separation experiments were performed with protein mixture (5.0 mg mL⁻¹, each) in NaAc buffer (0.01 mol L⁻¹) at different pH. O-CMCS and N-CMCS both showed different charged nature and water permeability at varied pH (2.0–14.0). These membranes were also assessed to be stable under the given pH range and showed about >1.0% dimensional charge or weight loss. Further, carboxymethyl chitosan may exhibit good biocompatibility, adhesion towards biomolecules and isoelectric point-based, pH-responsive separation.³¹ The dual-charged nature of these biopolymer-based membranes (presence of –COOH and –N⁺H₃ groups) also may be advantageous for its antifouling properties during biomolecule separation.³² Next, we describe the protein separation performance of O-CMCS and N-CMCS membranes.

Analysis of Lys and β-Cas

Solutions of Lys and β-Cas of desired concentration were freshly prepared in phosphate buffer solution (Na₂HPO₄ and NaH₂PO₄). Further, pH/ionic strength of the solutions were adjusted by the addition of a small quantity of HCl/NaOH, and sodium azide (0.1% w/v) was added to prevent microbial growth. Protein concentrations were determined by UV-Vis spectra (Simazdu 200 spectrophotometer) in 230–380 nm wavelength range. Protein concentration was measured as the average value of three consecutive measurements.

Results and discussion

Synthesis of N-CMCS and O-CMCS

The direct alkylation method utilizing monochloroacetic acid was adopted to prepare N-CMCS and O-CMCS derivatives of chitosan under different reaction conditions (responsible for attaining the N versus O selectivity of carboxyalkylation).^30,33,34 During carboxymethylation of chitosan with monochloroacetic acid (1 : 10, w/v) in the mildly alkaline medium (pH 8.0–8.5) at 65 °C for 6 h, only the amine groups were activated, and N-substitution was expected. Meanwhile, to synthesize O-substituted carboxymethylated chitosan, the reaction was carried out at 0–30 °C with 1 : 4 w/v of chitosan and monochloroacetic acid.

N-CMCS and O-CMCS were derived from a biopolymer (chitosan) containing a long polysaccharide chain. Their structures were elucidated by ¹H NMR and FTIR spectra. ¹H NMR spectrum for O-CMCS (Fig. 1) showed chemical shifts at 4.2 and 4.4 ppm, due to protons on the O-position of the –CH₂–COO– group in the chitosan derivative (C₂ and C₆, respectively). In the modified chitosan, carboxymethyl substituents were observed on a few amino and primary hydroxyl sites. The degree of carboxymethylation on the amino (N-position) and primary hydroxyl (O-position) sites was estimated at approximately 13.5% and 10.0%, respectively, from ¹H NMR spectrum. FT-IR spectra for O-CMCS and N-CMCS and unmodified chitosan (CS) are depicted in Fig. 2. The IR spectrum of CS shows a strong peak at 3428 cm⁻¹, which is assigned to the N–H extension vibration, O–H stretching vibration, and the intermolecular H-bonds of the polysaccharide moieties. The weak peak at 1654 cm⁻¹ is due to the amide C=O stretching. The FTIR spectrum of N-CMCS shows a strong new peak at 1734 cm⁻¹, representing the carboxylate C=O asymmetric stretching. The signal at 1384 cm⁻¹ was assigned to the symmetric stretching vibration of carboxylate C=O. In the H form of O-CMCS, the characteristic peaks are 1741 cm⁻¹ (–COOH), 1154–1029 cm⁻¹ (–C–O–), and 1624 and 1506 cm⁻¹ (–NH₃⁺).

Fig. 1 ¹H NMR spectra of O-CMCS.

Fig. 2 FTIR spectra of O-CMCS, N-CMCS and CS.

Surface morphology of N-CMCS and O-CMCS membranes

Scanning electron microscopy (SEM) images for N-CMCS and O-CMCS membranes (Fig. 3) show gradual reduction in surface segregation due to aqueous coagulation, leading to the arrangement of hydrophilic groups on the surface and pores of the membrane matrix.

Fig. 3 SEM images: (A) N-CMCS, (B) O-CMCS for front surface of membrane, (C) O-CMCS with high magnification, and (D) O-CMCS cross section.

Stability and ion-exchange capacity of O-CMCS and N-CMCS membranes

Thermal gravimetric analysis (TGA) curves for O-CMCS and N-CMCS membranes showed a two-step weight loss (Fig. S1; ESI†). The first-step weight loss at around 120 °C arose due to loss of absorbed water, whereas the second-step weight loss showed degradation of –COOH groups (200–300 °C). Beyond 380 °C, rapid decomposition of the membrane polymer matrix was observed. O-CMCS and N-CMCS membranes exhibited 109.9 °C and 112.4 °C glass transition temperature (T_g), respectively (Fig. S2 ESI†). The crystalline nature of the membrane samples (15 μm thicknesses) was studied by WXRD in reflection mode at 2θ angle (Fig. S3 ESI†). The intensity of the typical peak at 2θ ≈ 20° was high for N-CMCS compared with the O-CMCS membrane. The stability of these membranes was also tested in different solvents (Table S1; ESI†), and membranes were unaffected. This information confirms the thermal and chemical stability of these membranes.

The charged nature of the membranes was investigated in terms of H⁺/OH⁻ exchange capacity and surface charge concentration in the membrane matrix. Acidic and alkaline ion-exchange capacity (IEC) values along with data on the volume fraction of water in the membrane (φ_w) can be used to estimate the fixed charge concentration (X^m) of the membrane in units of (moles of sites)/(unit volume of wet membrane) by the following equation:³⁵

where IEC is expressed in equivalents per gram of dry membrane, ρ_d denotes the density of dry membrane, and τ denotes the membrane void porosity (volume of water within membrane per unit volume of wet membrane) defined by: τ = φ_w/1 + φ_w. Measured H⁺/OH⁻ exchange capacity values and acidic/alkaline surface charge concentration (mmol dm⁻³) for N-CMCS and O-CMCS membranes are included in Table 1. Here, it is interesting to record that the acidic and alkaline group surface charge concentrations on the membrane matrix were estimated at pH 3.0 and 11.0. Due to the presence of –COOH and –NH₂ groups, O-CMCS membrane showed acidic behaviour at low pH due to dissociation of –COOH groups. At pH 3.0, H⁺ exchange capacity and acidic surface charge concentration for O-CMCS membrane were estimated to be 2.07 mequiv. g⁻¹ of dry membrane and 0.76 mmol dm⁻³, respectively. Under the same environment, H⁺ exchange capacity for N-CMCS membrane was relatively low, corresponding to 0.27 mmol dm⁻³ acidic surface charge concentration. Under alkaline medium (pH 11.0), the charged nature of N-CMCS or O-CMCS membranes increased and showed 1.82 mequiv. g⁻¹ OH⁻ exchange capacity and 0.74 mmol dm⁻³ alkaline surface charge concentration. These studies revealed that O-CMCS membrane's strongly charged nature in acidic/alkaline medium may be due to dissociation of –COOH and –NH₂ groups, respectively. The N-CMCS membrane exhibited a weakly charged nature under acidic medium and strongly charged nature under alkaline medium due to the presence of –N⁺CH₃ group. The reaction scheme and schematic structure of O-CMCS and N-CMCS membranes are depicted in Scheme 1 under acidic and alkaline medium.

Table 1 Physicochemical properties of CS, N-CMCS and O-CMCS membrane

Properties	CS	N-CMCS	O-CMCS
a Uncertainty for measurements: 1.0 μm.b Uncertainty for measurements: 0.1%.c Uncertainty for measurements: 0.01 mequiv. g^−1 of dry membrane.d Measured at pH: 3.0.e Measured at pH: 11.0.
Thickness^a (μm)	190	188	181
Water content^b (%)	11.9	16.2	17.9
H^+ exchange capacity^c^,^d (mequiv. g^−1 of dry membrane)	—	0.32	2.07
OH^− exchange capacity^c^,^e (mequiv. g^−1 of dry membrane)	—	1.82	1.83
Acidic surface charge concentration (mmol dm^−3)	—	0.27	0.76
Alkaline surface charge concentration (mmol dm^−3)	—	0.74	0.74

---

MWCO and water permeability

MWCO values for O-CMCS and N-CMCS membranes were estimated by different probe solutes (PEG and dextran) with 10–30 kDa molecular weights. MWCO was defined as the molecular weight of probe solutes that exhibited more than 90% rejection. Both membranes showed 16–19 kDa MWCO values (Fig. 4), which confirmed their ultra-filter nature. Membrane pore diameter was estimated by Ferry equation on the basis of % rejection (R) expected for a membrane with uniform circular pores of diameter (a) as a function of the diameter of spherical solutes (r).^36–38 Molecular/ionic sizes for neutral or ionic probes were obtained from literature.^39–41 Membrane pore diameter was estimated at pH 7.0 and found to be 0.53 nm for O-CMCS, while N-CMCS showed a 0.48 nm diameter because of its inner architecture in the membrane matrix.

Fig. 4 Rejection values vs. molecular weight for different membranes.

Water permeability for O-CMCS and N-CMCS membranes varied between 5.0–6.0 × 10⁻⁸ Pa⁻¹ m h⁻¹ (ultra-filter range) and increased with pH of the equilibrating medium (Fig. 5). Increase in membrane water permeability with pH was attributed to the alteration in the membrane pore size. It seems that the presence of –N⁺H₃ groups in alkaline medium (pH > 7.0) caused a marginal increase in water permeability. It was also observed that pure water permeability was unaltered with time, after prolonged operation.

Fig. 5 Variation in water permeability with pH for N-CMCS and O-CMCS membranes.

pH-responsive protein separation across the N-CMCS membrane

To observe the pH-responsive protein separation properties of N-CMCS membrane, pressure-driven filtration experiments were performed at different pH using a mixture of Lys and β-Cas. For both proteins, individually, protein permeability (J_P) values were initially independent on pH of the medium (pH 2.0–4.0). In the case of β-Cas, J_P increased steeply between pH 4.0–6.0, while for Lys, a rapid increase in J_P was observed between pH 8.0–9.0. Interaction between the charged membrane and protein depended on the nature of the protein and its iso-electric point (IEP). At low pH (<4.0), β-Cas and Lys (IEP: 4.7 and 9.2, respectively) behaved as positively charged proteins (β-Cas⁺ and Lys⁺). In the pH range 2.0–12.0, N-CMCS membrane also showed positive charge due the existence of –NHCH₂COOH group in the equatorial position. Thus, there is less possibility for dissociation of –COOH, perhaps because of steric hindrance. In the pH range 2.0–12.0, only protonation or deprotonation of –NH₂ present in axial position is expected. Thus, a positive charge in the membrane matrix (–N⁺H₃) was expected. Below pH 4.0, J_P values for β-Cas⁺ through the positively charged N-CMCS membrane was relatively low (Fig. 6A). At high pH (>4.0), permeability for β-Cas⁻ through the positively charged O-CMCS membrane (Fig. 6B) increased steeply between pH 4.0–6.0 and attained a limiting value. Similarly, permeability of Lys⁺ (pH < 9.2) through the positively charged N-CMCS membrane was very low. Further, beyond pH 9.2, J_P for Lys⁻ increased quickly. Thus, the N-CMCS membrane behaved as positively charged due to the favoured protonation of the –NH₂ group. Further, an idea about protein separation can be obtained from the flux recovery ratio (F_RR), calculated using the following equation:

Fig. 6 Variation of protein permeability (J_P) with pH for β-Cas and Lys in mixed solution (5 mg mL⁻¹) at 30 °C: A: N-CMCS, B: O-CMCS membrane.

(J_P)_β-Cas and (J_P)_Lys are the permeability of β-Cas and Lys, respectively. F_RR values, which may be defined as the ratio of β-Cas and Lys permeability, showed insignificant variation (Fig. 7). The N-CMCS membrane was assessed to be unsuitable for pH-responsive protein separation. Because of the positively charged nature of N-CMCS membrane in pH range 2.0–12.0, Donnan exclusion due to mutual electrostatic attraction between membrane and protein was insignificant in accelerating the mobility of a protein molecule based on its IEP. Thus, N-CMCS membrane is unsuitable for the separation of zwitterionic protein molecules.

Fig. 7 Separation factor of Lys and β-Cas at different pH across O-CMCS and N-CMCS membranes (30 °C and 5 mg mL⁻¹ concentration).

pH-responsive protein separation across O-CMCS membrane

pH-responsive protein separation properties of O-CMCS membrane were also studied using β-Cas and Lys mixed solutions. J_P values for both proteins are presented at varied pH of the external environment (Fig. 6B). Permeability of β-Cas through the O-CMCS membrane was independent of the pH (2.0–12.0) of the protein solution. Here, it is interesting to record that the charged nature of O-CMCS membrane can be tuned by varying the pH of the external environment. This may be due to the presence of –NH₂ and –COOH groups ionizable to positive and negative charges by varying pH. In the case of low pH (<2), the dominant charge in the membrane matrix were the protonated amino groups, and in the case of high pH (>4), the dominant charges were deprotonated carboxyl groups. It was also reported that at the IEP of O-CMCS (pH 2.0–4.0), the numbers of –N⁺H₃ and –COO⁻ groups in the membrane matrix are equal, and intra-ionic attraction between opposite charges seldom results in residual ionic groups. When pH increased from 4.0 to 12.0, the amino groups were completely deprotonated and contributed to the loss of solubility of the chain segments and also to the formation of new crosslinks by hydrogen bonding.⁴² Thus, the O-CMCS membrane exhibited amphoteric behaviour in pH range 2.0–4.0, and beyond this pH range, the net charge in the membrane matrix was negative. Permeability values for Lys⁺ through amphoteric O-CMCS membrane were relatively low in the 2.0–4.0 pH range, while they quickly increased from pH 4.0–6.0 (Fig. 6B). In this pH range, the membrane carried a net negative charge due to the presence of deprotonated carboxyl groups. This favoured a relatively fast transmission of Lys⁺, probably because of Donnan exclusion due to mutual electrostatic attraction. Beyond pH 9.2 (IEP of Lys), Lys existed as Lys⁻, and its transmission through the negatively charged membrane matrix was extremely low, while in the opposite case (at pH > pI), both proteins carried a positive charge and showed relatively low permeability. This information revealed that the pH-responsive protein separation across the O-CMCS is due to its pH-tunable charged nature. F_RR values for β-Cas and Lys across O-CMCS membrane suggest the possibility of separation of these proteins in alkaline media. This can be explained by considering the pH-responsive, tunable nature of the membrane and the IEP of the proteins. It is also interesting to record that no fouling or deterioration in the membrane flux of O-CMCS membrane was recorded after its application in the protein environment for a prolonged time period (50 days) (Fig. 8). After use, the flux reduced marginally (approx. 5%). Anti-fouling properties of the O-CMCS membrane is attributed to the presence of –N⁺H₃ and –COO⁻ groups in the membrane matrix and thus the amphoteric nature of the membrane in pH range 2.0–4.0. With further increase in the pH of the external environment, dissociation of the –COOH group was suppressed due to deprotonation.

Fig. 8 Membrane-fouling behaviour at pH 7.0: variation of water and NaCl solution flux across O-CMCS with number of days after membrane operation in protein environment.

Effect of protein concentration on permeability

Variation of permeability with protein concentration through N-CMCS (at pH 7.0) and O-CMCS (at pH 10.0) (inset plot) is depicted in Fig. 9A. Across both membranes, a marginal increase in protein permeability with concentration was observed. As expected, F_RR values of β-CAS and Lys through both membranes (N-CMCS and O-CMCS) were also independent of the protein concentration (Fig. 9B).

Fig. 9 (A) Variation of protein permeability through N-CMCS at pH 7.0 and through O-CMCS at pH 10.0 (inset plot), (B) separation factor of β-CAS and Lys through N-CMCS and O-CMCS membranes at different concentrations.

Thus, it is possible to develop pH-responsive amphoteric or charged carboxymethyl grafted chitosan with high water wettability for selective protein separation. The charged nature of modified biomaterial-based membranes rules out any appreciable fouling or deterioration in membrane throughput. In this case, the difference in the IEP values of Lys and β-CAS was quite high, but in general cases where IEP values are close, the prepared membrane also can be efficiently used by varying its pore size according to the molecular size of the protein.

Conclusions

Grafting of carboxymethyl groups on chitosan was achieved at equatorial or axial positions to prepare pH-responsive N-CMCS and O-CMCS membranes. The N-CMCS membrane showed a positive charge on the matrix under pH range 2.0–12.0 due the existence of the –NHCH₂COOH group in the equatorial position and protonation of the –NH₂ group, leading to a positive charge on the membrane matrix in the acidic region. Meanwhile, O-CMCS membrane showed amphoteric nature in the pH range 2.0–4.0 because of protonated amino groups and dissociated –COOH groups. In the pH range 4.0–12.0, the membrane matrix possessed a negative charge due to deprotonated carboxyl groups. The pH-responsive, tunably charged nature of N-CMCS was advantageously used to achieve antifouling properties in the biopolymer-based membrane during biomolecule separation.

Protein permeability studies revealed insignificant variations in F_RR values for β-Cas and Lys through the N-CMCS membrane, and the membrane was assessed to be unsuitable for pH-responsive separation. The O-CMCS membrane showed pH-responsive protein separation properties because of its pH-tunable charge. Protein permeability data suggested efficient separation of β-Cas and Lys in alkaline media using O-CMCS membrane. It is also interesting to record that no fouling or deterioration in the membrane flux of O-CMCS membrane was recorded after its application in the protein environment for a prolonged time period (50 days).

Further studies are aimed to optimize the polymer structure and architecture to develop the best membrane with the same amphoteric properties and pore dimensions for protein separation/purification. Detailed study on structural parameters, long-term durability, degree of cross-linking, and functional group density are necessary to optimize the membrane before practical application.

Acknowledgements

CSIR-CSMCRI registration no.: 082/2014 Instrumental support received from Analytical Science Division, CSIR-CSMCRI is gratefully acknowledged. One of the authors (T. Chakrabarty) would like to thank CSIR-India for providing CSIR-SRF fellowship ASSOCIATED CONTENT.

References

1. M. J. Laudenslager, J. D. Schiffman and C. L. Schauer, Biomacromolecules, 2008, 9, 2682.
2. Z. Xianfang and E. Ruckenstein, J. Mater. Sci., 1998, 148, 195.
3. J. Shang, Z. Shao and C. Xin, Polymer, 2008, 49, 5520.
4. L. Zhang, J. Guo, J. Zhou, G. Yang and Y. Du, J. Appl. Polym. Sci., 2000, 77, 610.
5. X. Peng and L. Zhang, Langmuir, 2007, 23, 10493.
6. L. Zhao, L. Xu, H. Mitomo and F. Yoshii, Carbohydr. Polym., 2006, 64, 473.
7. V. K. Mourya and N. N. Inamdar, J. Mater. Sci.: Mater. Med., 2009, 20, 1057.
8. K. Kurita, Prog. Polym. Sci., 2001, 26, 1921.
9. S. Hitoshi, M. Yutaka, R. Rene and S. Yoshihiro, Chem. Commun., 2000, 909.
10. H. Sashiwa, N. Kawasaki, A. Nakayama, E. Muraki, H. Yajima, N. Yamamori, Y. Ichinose, J. Sunamoto and S. Aiba, Carbohydr. Res., 2003, 338, 557.
11. K. R. Holme and A. S. Perlin, Chitosan N-sulfate, Carbohydr. Res., 1997, 302, 7.
12. T. Ouchi, H. Nishizawa and Y. Ohya, Polymer, 1998, 39, 5171.
13. I. K. Park, T. H. Kim, S. I. Kim, Y. H. Park, W. J. Kim, T. Akaike and C. S. Cho, Int. J. Pharm., 2003, 257, 103.
14. V. K. Mourya, N. Inamdar and N. Tiwari, Adv. Mater. Lett., 2010, 1, 11.
15. Y. Su, C. Mu, C. Li and Z. Jiang, Ind. Eng. Chem. Res., 2009, 48, 3136.
16. Z. P. Zhao, Z. Wang and S. C. Wang, J. Mater. Sci., 2003, 217, 151.
17. V. K. Mourya, N. N. Inamdara and A. Tiwari, Adv. Mater. Lett., 2010, 1, 11.
18. C. M. Valmikinathan, V. J. Mukhatyar, A. Jain, L. Karumbaiah, M. Dasari and R. V. Bellamkond, Soft Matter, 2012, 8, 1964.
19. X. F. Liang, H. J. Wang, H. Luo, H. Tian, B. B. Zhang, L. J. Hao, J. I. Teng and J. Chang, Langmuir, 2008, 24, 7147.
20. T. Y. Liu, S. Y. Chen, Y. L. Lin and D. M. Liu, Langmuir, 2006, 22, 9740.
21. C. A. Tang, E. Ozcam, B. Stout and S. A. Khan, Biomacromolecules, 2012, 13, 1269.
22. M. E. Ibrahim Sherbiny and M. M. Elmahdy, J. Appl. Polym. Sci., 2010, 118, 2134.
23. S. Lee, K. Choo, C. H. Lee, H. Lee, T. Hyeon, W. Choi and H. Kwon, Ind. Eng. Chem. Res., 2001, 40, 1712.
24. T. Tomida, K. Hamaguchi, S. Tunashima, M. Katoh and S. Masuda, Ind. Eng. Chem. Res., 2001, 40, 3557.
25. C. David, F. Pignon, T. Narayanan, M. Sztucki, G. Gesan-Guiziou and A. A. Magnin, Langmuir, 2008, 24, 4523.
26. D. J. Hughes, Z. Cui, R. W. Field and U. K. Tirlapur, Langmuir, 2006, 22, 6266.
27. P. Priyananda and V. Chen, J. Membr. Sci., 2006, 273, 58.
28. S. Sun, Y. Yue, X. Huang and D. Meng, J. Membr. Sci., 2003, 222, 3.
29. K. L. Jone and C. R. O'Melia, J. Membr. Sci., 2001, 193, 163.
30. I. H. Huisman, P. Pradanos and A. Hermandez, J. Membr. Sci., 2000, 179, 79.
31. E. R. Nightingale Jr, J. Phys. Chem., 1959, 63, 1381.
32. M. M. Rohani and A. L. Zydney, J. Membr. Sci., 2012, 397, 1.
33. V. K. Mourya, N. N. Inamdara and A. Tiwari, Adv. Mater. Lett., 2010, 1, 11.
34. N. T. An, P. L. Dung, D. T. Thien, N. T. Dong and T. T. Y. Nhi, Carbohydr. Polym., 2008, 73, 261.
35. A. Zhu, J. Liu and W. Ye, Carbohydr. Polym., 2006, 63, 89.
36. R. K. Nagarale, G. S. Gohil, V. K. Shahi and R. Rangarajan, J. Colloid Interface Sci., 2005, 287, 198.
37. N. O. Necht, D. J. Malik and E. S. Tarleton, Sep. Purif. Technol., 2008, 62, 228.
38. S. C. Chen, Y. C. Wu, F. L. Mi, Y. H. Lin, L. C. Yu and H. W. Sung, J. Controlled Release, 2004, 96, 285.
39. Y. Su and C. Li, J. Membr. Sci., 2007, 305, 271.
40. P. Aimar, M. Meireles and V. Sanchez, J. Membr. Sci., 1990, 54, 321.
41. W. R. Bowen, A. W. Mohammad and N. Hilal, J. Membr. Sci., 1997, 126, 91.
42. L. Vachoud, N. Zydowics and A. Domand, Carbohydr. Res., 2000, 326, 295.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra05314a

---