Preparation of Homogeneous Regenerated Cellulose Nanofiber based Membrane Adsorber for Membrane Chromatography Applications

doi:10.21203/rs.3.rs-4956358/v1

Download PDF

Research Article

Preparation of Homogeneous Regenerated Cellulose Nanofiber based Membrane Adsorber for Membrane Chromatography Applications

https://doi.org/10.21203/rs.3.rs-4956358/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Herein, electrospun regenerated cellulose (RC) nanofibers were fabricated for membrane chromatography applications. A cellulose acetate (CA) nanofiber membrane was prepared using a solvent mixture of N,N-dimethylacetamide and acetone, and glycerol was used to enhance the spin ability of CA, followed by alkaline treatment to obtain the RC nanofiber membrane. Glycerol addition enabled the electrospinning of CA nanofibers at lower concentrations, ensuring uniform nanofiber production and overcoming the limitations of bead formation or electrospraying at low CA concentrations during electrospinning. This study revealed that the compositions of glycerol and CA affect the pore size and fiber diameter of RC nanofiber membranes. To evaluate the practicality of RC nanofiber membranes for membrane chromatography, their binding capacity for bovine serum albumin was examined after grafting anion exchange ligands (3-(methacryloylamino)propyl-trimethylammonium chloride). The nanofiber membrane fabricated using 25 wt.% CA and 10 wt.% glycerol exhibited a superior static binding capacity of 230.9 mg/mL and a dynamic binding capacity of 112.1 mg/mL for bovine serum albumin. These findings indicated that the RC nanofiber–based membrane adsorber has potential applications in various separation and purification processes including adsorption, filtration, and membrane chromatography.

Electrospinning

regenerated cellulose

glycerol

membrane chromatography

Electrospinning stands out from other membrane fabrication techniques because it enables the production of nanofiber mats with high specific surface areas (Xue et al. (2019)). It is widely used for preparing nanofibers, membranes, adsorbers, and functional materials in environmental, chemical, energy, and biomedical applications. In the past decade, researchers have investigated nanofiber–based membrane adsorbers as a promising solution to address the limitations of conventional column chromatography, including its low porosity and slow flow rate due to resin filling. Because nanofiber membranes have much higher porosity than conventional ion exchange resins and phase separation membranes, they exhibit high flow rates and adsorption capacities.

Although various polymeric materials can be used for electrospinning and as membrane adsorbers in downstream bioprocesses, only some polymeric materials, including polyethersulfone, polyvinylidene fluoride, and regenerated cellulose (RC), are mainly used in biopharmaceutical purification processes due to biological safety. Specifically, RC membranes are a functional platform for membrane chromatography because of their chemical and thermal stability and abundant hydroxyl groups that act as reactive sites (Hokkanen, Bhatnagar, and Sillanpää 2016). The hydrogen bonding between the polymer chains of RC membranes is very strong, resulting in excellent chemical stability and organic solvent resistance. Therefore, RC membranes have become important in the chemical and bio industries (Singh, Chen, et al. 2008; Singh, Wang, et al. 2008). Zhou et al. fabricated a composite membrane adsorber via the aqueous phase inversion of RC with 2,2,6,6-tetramethylpiperidine-1-1 oxyl–oxidized cellulose nanofibers (Zhou et al. 2020). Liu et al. prepared a chitosan/cellulose blend hollow fiber membrane using a wet spinning bath and studied its bovine serum albumin (BSA) adsorption performance (Liu and Bai 2005). A cellulose acetate (CA)–based dope solution was prepared and directly coagulated in a NaOH solution to convert CA into RC. Using electrospinning, Zhang et al. successfully fabricated a CA nanofibrous membrane, which was regenerated and functionalized with diethylaminoethyl (DEAE) for BSA separation (Zhang, Menkhaus, and Fong 2008). In general, it is challenging to fabricate cellulose nanofibers directly by electrospinning because selecting appropriate solvents for dissolving cellulose is difficult. To solve this problem, a simple strategy was proposed by Liu et al. (Liu and Hsieh 2002, 2003) and Son et al. (Son et al. 2004). To obtain RC nanofiber, they first produced CA nanofibers and then treated them in an alkaline solution to eliminate acetyl groups.

Among the various factors affecting nanofiber fabrication, the viscosity of the dope solution is crucial (Tabe 2017; Shahriar et al. 2019). The viscosity of the polymer solution can be controlled using an appropriate solvent and cosolvent. For example, N,N-dimethylacetamide (DMAc), N,N-dimethylformamide (DMF), acetic acid, and hexafluoropropylene (HFP), can be used alone or in combinations to prepare CA solutions for electrospinning (Liu and Hsieh 2002; Tungprapa et al. 2007; Liu and Tang 2007; Ye, Watanabe, and Ishihara 2004). However, these solvents reduce the viscosity of the dope solution, resulting in bead formation, and increasing the CA concentration above the critical chain entanglement concentration (C_e) promotes the transition from beads to thick fibers due to intermolecular interactions (Tiwari and Venkatraman 2012). To overcome these challenges and fabricate uniform CA nanofibers, studies have explored different approaches, including single and binary systems and the incorporation of additives (Tiwari and Venkatraman 2012) Majumder et al. studied the solubility and spin ability of CA using different solvent systems (Majumder et al. 2019). They examined the use of pure acetone and a binary solvent mixture of 2:1 acetone/DMAc. Electrospinning CA with pure acetone resulted in cylindrical and ribbon-shaped fibers with a 1 µm diameter, whereas electrospinning CA with the 2:1 acetone/DMAc mixture produced smooth bead-free cylindrical fibers with a diameter ranging from 250 to 350 nm. Additionally, electrospinning CA with a 3:1 acetic acid/water mixture formed beaded fibers. In another study, Naragund et al. successfully fabricated CA nanofiber filtration membranes using electrospinning (Naragund and Panda 2020). They utilized solvent mixtures of methyl ethyl ketone (MEK) and DMAc in different ratios (i.e., 2:1, 1:1, and 1:2) and different CA concentrations (7–19%). MEK was chosen as a substitute for acetone because it has a higher boiling point, which allows longer electrospinning durations by reducing solvent evaporation. However, additives such as plasticizers (Gonçalves et al. 2019) can also be used to fabricate uniform electrospun nanofibers. Plasticizers are low-volatility fluids used to increase the flexibility and extensibility of films by reducing the intermolecular forces between polymer chains. Plasticizers reduce the energy level required to give the chain mobility by reducing the glass transition temperature of the polymer.

Herein, the effect of glycerol on the physicochemical properties of a cellulose nanofiber membrane was investigated to improve its spin ability and adsorption capacity for membrane chromatography applications. Glycerol is one of the most used plasticizers in biopolymers because it is water soluble, polar, and nonvolatile and has a low molecular weight. The addition of glycerol can change the fiber morphology and pore size of electrospun nanofibers (Teixeira et al. 2021). After the electrospinning of CA nanofibers, their deacetylation and grafting of 3-(methacryloylamino)propyl-trimethylammonium chloride (MAPTAC) on an RC nanofiber membrane were performed to prepare a nanofibrous membrane adsorber. The adsorption performance of the prepared RC nanofiber-based membrane adsorbers was evaluated via both static and dynamic binding of BSA.

2.1. Materials

CA (average molecular weight = 30,000 g/mol (gel permeation chromatography) and acetylation degree = 39.8% (w/v)) and MAPTAC were utilized in the study. N,N-Methylenebisacrylamide (crosslinker), 2-hydroxy-2-methylpropiophenone (photoinitiator), and BSA were purchased from Sigma-Aldrich. Acetone, DMAc, hydrochloric acid (HCl), and polyethylene glycol (PEG 400) were purchased from Samchun Pure Chemical Co., Ltd. (South Korea). Analytical-grade glycerol was purchased from Sigma-Aldrich. Sodium chloride (NaCl, 99%) and NaOH were purchased in Samchun Pure Chemical Co., Ltd (South Korea). tris–HCl (99%, Sigma-Aldrich, USA) was used as received. All reagents were utilized as received without any additional purification steps. Experiments and solutions were prepared using deionized water (DW) obtained from a DW purification system (RiOs Essential, Millipore, USA). The DW had a resistivity of 18.2 µΩ/cm, indicating its high purity and low mineral content.

2.2. Electrospinning

The dope solution for electrospinning was prepared using CA, acetone, DMAc, and glycerol without any further purification. Each dope solution was labeled CxGy, where C represents CA, G represents glycerol, and x and y denote the respective weight percentages. For example, in the C20G10 sample, CA accounts for 20 wt.%, glycerol accounts for 10 wt.%, and the remaining 70% is a combination of DMAc and acetone in a 2:1 ratio. To obtain a uniform dope solution, the mixture was stirred for 24 h and then sonicated to prevent aggregation. The viscosity of the dope solution was measured using a viscometer (Brookfield Ametek, Gel Timer DV2T). The temperature of the dope solution was carefully maintained at 20°C throughout the measurement process. The spindle of the viscometer was set to rotate at a speed of 10 RPM. This allowed for the accurate determination of the solution viscosity, providing valuable information about its flow characteristics and consistency. A viscosity measurement method reported in the literature (Genceli et al. 2018) was used.

The electrospinning process (ESR200R2D; ES-ROBOT, South Korea) was performed under specific conditions: 20°C, 25–30% relative humidity, applied voltage = 13 kV, feed rate = 0.3 mL/h, and tip to collector distance = 14 cm. A blunt-tip 23-gauge stainless needle was used, and the collector was drum-type rotating at 300 RPM. Each solution (3 mL) was loaded into the syringe pump and electrospun onto an aluminum foil, which acted as the substrate. Detailed information on the lab-scale electrospinning process is reported in our previous study (Shah et al. 2020). The electrospun nanofiber membrane was separated from the aluminum foil and subjected to heat treatment in a vacuum oven. The nanofiber membrane was kept overnight at 60°C. This heat treatment was performed to enhance the mechanical properties of the nanofiber membrane and ensure complete evaporation of any remaining solvents (Zhang, Menkhaus, and Fong 2008).

2.3. Nanofiber membrane modifications

The overall preparation processes of the nanofiber–based membrane adsorber are illustrated in Fig. 1. Herein, the deacetylation method was used to prepare the RC nanofiber membrane for membrane chromatography applications. This process involved the removal of acetyl groups from the CA nanofiber membrane to obtain the desired RC nanofiber membrane. To initiate the deacetylation reaction, a 0.05 M NaOH solution (9:1 EtOH:water) was used. The deacetylation process was performed by immersing the CA nanofiber membrane in the NaOH solution for 14 h. After the deacetylation reaction, the nanofiber membrane was thoroughly rinsed to remove any remaining NaOH and sodium acetate.

RC membranes were modified with quaternary ammonium anion exchange ligands using MAPTAC, which can be linked to the RC nanofiber membrane via ultraviolet (UV) irradiation, as reported in the literature (Li et al. 2017). C25G10 nanofiber membranes were selected as supports for the grafting of MAPTAC to evaluate the adsorption performance of BSA. To observe the effect of the pore size on the adsorption capacity of BSA, C25G5 was also modified using the same method. The modification solution used for this process was composed of 10.5 g of MAPTAC (monomer), 0.15 g of PEG 400 (disperser), 0.2 g of 2-hydroxy-2-methylpropiophenone photo-initiator), 0.07 g of N,N-methylenebisacrylamide (crosslinker), and 4.08 g of DW (solvent). First, RC nanofiber membranes were immersed in the reaction solution, and any excess solution was removed by rolling. Then, the membrane was dried in a vacuum oven at 60°C for 2 h. Once dry, the RC nanofiber membrane underwent UV irradiation using a 1 kW Hg UV lamp. The lamp emitted light at a wavelength of 321 nm^− 1 with an intensity of 80 W/cm². The top and bottom sides of the membrane were irradiated for 1 min each. After UV irradiation, the grafted RC nanofiber membrane was soaked in DW at 30°C for 30 min and then thoroughly rinsed to remove any residual materials.

2.4. Characterizations of nanofiber membranes

The morphology of the prepared membrane was examined using a field-emission scanning electron microscope (Thermoscientific Quattro S). Before imaging, the membrane was coated with a thin layer of platinum (Pt) to enhance its conductivity. The obtained images were analyzed using ImageJ software, allowing measurement of the fiber diameter based on a sample size of > 50 fibers.

The pore size of the membrane was determined using a POROMETER NV capillary flow porometer (POROLUX 1000). This instrument enabled accurate measurement of the pore size of the membrane, providing insights into its porosity and filtration capabilities. The chemical properties of the membrane were assessed via attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy (Thermo Nicolet 5700) in the range of 700 to 4,000 cm^− 1 with a resolution of 4 and 32 sample scans. This technique allowed the identification of functional groups and chemical bonds present in the membrane. The crystallinity of the membrane was analyzed via X-ray diffraction (XRD; Rigaku Ultima IV X-ray diffractometer) with Cu Kα radiation (λ = 1.5418 Å). The measurements were performed at a speed of 1°/min in the 2θ range of 5–40°, which provided information about the structural arrangement and crystalline phases of the membrane. The relative crystallinity (R_C) of the membrane was calculated using Eq. (1) (Gonçalves et al. 2019).

$$\:{R}_{C}=\frac{(TA-AA)}{TA}\times\:100$$

where R_C is the relative crystallinity of the membrane (%), TA is the total area, and AA is the area of the amorphous region, respectively.

The thermal properties of the membrane were evaluated via differential scanning calorimetry (DSC; TA Instruments Q1000) and thermogravimetric analysis (TGA; Mettler Toledo TGA/DSC 3+). DSC analysis was performed at a rate of 10°C/min from 150°C to 250°C, and TGA analysis was performed at a rate of 10°C/min from 25°C to 600°C. These techniques allowed the investigation of the thermal behavior of the membrane, including phase transitions, thermal stabilities, and decomposition temperatures.

2.5. Static binding capacity (SBC)

First, the MAPTAC-grafted RC nanofiber membrane was trimmed to a diameter of 2.5 cm. Then, it was washed three times with 9.8 mL of tris–HCl buffer for 5 min to ensure thorough cleaning. After cleaning, the MAPTAC-grafted RC nanofiber membrane was sealed and shaken for 14–16 h in 9.8 mL of a 2,000 ppm BSA solution. The absorption of the solution after the SBC experiment was measured at 280 nm using a UV detector with tris–HCl buffer as the reference. Subsequently, the MAPTAC-grafted RC nanofiber membrane was washed twice with 10 mL of tris–HCl buffer for 15 min each. Finally, it was eluted by introducing 10 mL of a 1 M NaCl solution to the MAPTAC-grafted RC nanofiber membrane and allowing it to elute for 1 h. The absorption of the eluted solution was measured at 280 nm using a UV detector with a 1 M NaCl solution as the reference. The quantities of BSA in the feed solution, adsorbed onto the MAPTAC-grafted RC nanofiber membrane, and eluted during the SBC experiment were determined using a UV–visible spectrophotometer (Analytik Jena, SPECORD 210 PLUS) and by examining the samples within the wavelength range of 270 to 290 nm. The protein SBC was determined using Eq. (2) (Singh et al. 2014).

$$\:SBC\:(mg/ml)=\frac{Ew}{Pv}$$

where E_w is the BSA elution amount (mg) and P_v is the membrane volume (MV; mL).

2.6. Dynamic binding capacity

The dynamic binding test was performed using an ÄKTA GO (Cytiva) chromatography system with UNICORN 7.6 software. The experiment involved the use of various solutions, and the system was allowed to run until reaching a UV absorbance equilibrium or a stable state. To initiate the test, first, the system was stabilized using a buffer solution. Then, a BSA solution was introduced to measure the absorbance at 100% breakthrough. To ensure the removal of any remaining BSA from the system, the buffer solution was run through. After the cleaning step, the MAPTAC-grafted RC nanofiber membrane was loaded into the system. Once the buffer solution was stabilized, the elution solution (i.e., 1 M NaCl solution) was passed through the system to achieve equilibrium during the pretreatment phase. Following pretreatment, the BSA solution (2,000 ppm) was loaded onto the MAPTAC-grafted RC nanofiber membrane. The flow rate was controlled from 1 to 20 MV/min, and adsorption was performed until the permeate solution showed the same concentration as the feed solution. DBC_10% was determined when the permeate concentration reached 10% of the feed concentration. The MAPTAC-grafted RC nanofiber membrane was washed with the buffer solution until the absorbance reached a stable baseline; then, BSA was eluted using the elution solution.

3.1. Characterization of the CA nanofiber membrane

Figure 2 shows the scanning electron microscopy (SEM) images of CA nanofiber membranes with and without glycerol. In the absence of glycerol, beads are noticeable in pure CA samples (i.e., C20G0 and C25G0), as shown in Fig. 2a and 2d. This is caused by the low viscosities of the dope solution, as shown in Table 1. C20G0, having a low CA concentration, exhibits a bead-on-string morphology due to the very low viscosity (2,934 cP) of the dope solution. Compared with C20G0, fewer beads are observed in C25G0, which has a higher polymer concentration. C25G0 exhibits a high viscosity of 12,480 cPs, as shown in Table 1, resulting in the average fiber diameter of 261 nm along with the presence of beads and thick fibers (> 1 $\:{\mu\:}\text{m}$). The distributed fiber diameter indicates that the electrospinning conditions are not appropriate for obtaining homogeneous nanofibers. In contrast, the addition of glycerol in samples such as C20G5 (Fig. 2b) and C20G10 (Fig. 2c) dramatically reduces bead formation, resulting in more homogeneous fiber diameters. C20G10 exhibits a low viscosity of 7,680 cPs, leading to a uniform nanofiber diameter of ~ 384 nm without any beads. After the addition of glycerol, the fiber diameter considerably increases at high CA concentrations while maintaining fiber uniformity, as shown in Fig. 2e and 2f. Finally, C25G10 shows the highest viscosity, the largest fiber diameter, and the largest pore size among the prepared samples. Because the pore size of nanofiber membranes is proportional to the fiber thickness, the pore size of prepared CA nanofiber membranes can be effectively tuned from 1 to 6$\:\:{\mu\:}\text{m}$ without bead or nonuniform fiber formation by controlling the viscosity of the dope solution with the glycerol concentration.

Table 1

Dope solution viscosity, electrospun nanofiber diameter, and pore size of CA nanofiber membranes.
Sample	CA (wt.%)	Glycerol (wt.%)	Viscosity (cP)	Fiber diameter (nm)	Pore size (µm)
C20G0	20	0	2,934	Beads on string	-
C20G5	20	5	4,508	293 ± 85	1.178 ± 0.026
C20G10	20	10	7,680	384 ± 72	1.889 ± 0.133
C25G0	25	0	12,480	261 ± 184	2.065 ± 0.057
C25G5	25	5	19,720	564 ± 78	2.164 ± 0.078
C25G10	25	10	31,740	1,310 ± 250	5.903 ± 0.311

These results indicate that the addition of glycerol to the dope solution increases intermolecular interactions via hydrogen bonding and enhances the entanglement between CA chains, eventually reaching the concentration required for forming homogeneous nanofibers (Nie et al. 2008). The schematic for establishing hydrogen bonding between chains via the addition of glycerol is illustrated in Fig. 3. Notably, the increase in the viscosity of the dope solution caused by glycerol addition is considerably smaller than the increase in the viscosity of the dope solution caused by an increase in the CA polymer concentration. Therefore, controlling the glycerol concentration is more advantageous for achieving homogeneous nanofibers than controlling the polymer concentration alone (Nusrat Sharmin 2021).

The ATR-FTIR spectra of C25G0, C25G5, and C25G10 are shown in Fig. 4a. In these spectra, the intensity of the O–H stretching peak in the range of 3,200–3,500 cm^− 1 increases and that of the C–H stretching peak in the range of 2,875–2,940 cm^− 1 also increases due to the remaining glycerol. Glycerol is not removed by thermal drying and remains strongly bonded with CA chains via hydrogen bonds (Huihua Liu 2013).

In the XRD patterns of CA membranes shown in Fig. 4b, two peaks are observed: the first at 8° due to the crystalline structure of CA and the second at 19° C due to the amorphous structure of CA. The peak at 2θ = 8° shows that the crystallinity of CA decreases due to the addition of glycerol (Chen et al. 2016). With an increase in the glycerol concentration, the intensity of the amorphous peak at 19° increases. The presence of glycerol in the polymer matrix causes a rearrangement of the chains via hydrogen bonding, resulting in decreased crystallinity (Yan et al. 2023). Consequently, the fibril structure in C25G10 has 87.8% amorphous and 12.2% crystalline segments whereas that in C25G0 has 60.2% amorphous and 39.8% crystalline segments, as calculated using Eq. 1 (Wan Daud and Djuned 2015). These observations imply that glycerol acts as a plasticizer in the CA membrane, contributing to increased flexibility and amorphousness.

Figure 5a shows the DSC curves of C25G0, C25G5, and C25G10. The DSC curves of nanofiber membranes show two heating cycles. In the first heating cycle, the peak temperatures for endothermic reactions in C25G0, C25G5, and C25G10 are 125°C, 101°C, and 103°C, respectively. The glass transition temperatures (T_g) of membranes were observed in the first and second heating cycles, as shown in Fig. 5a. The T_g values of C20G0, C20G5, and C20G10 in the second heating cycle are 216°C, 201°C, and 166°C, respectively. The T_g of CA nanofiber membranes with glycerol decreases due to a reduction in intermolecular interactions between polymer chains, which is attributed to the presence of different concentrations of glycerol (Alessandro Bonifacio 2023; Rafael Erdmann 2021). This interaction between CA and glycerol reduces the heat flow, as indicated by the endothermic tendency. The other reason for the decrease in T_g is possibly the change in the nature of the nanofiber caused by the addition of glycerol. In detail, nanofibers become more amorphous and their degree of amorphousness increases as the concentration of glycerol increases. This phenomenon is seen in the XRD patterns of membranes in Fig. 4b (Nie et al. 2008; Lucena et al. 2003). Moreover, mixing a plasticizer in the polymer increases the free volume in the material structure; thus, polymer chains can move around at lower temperatures, resulting in a lower T_g of the polymer (Saxena, Shukla, and Gaur 2021).

The TGA curves of C25G0, C25G5, and C25G10 nanofiber membranes are shown in Fig. 5b. In the TGA curve of C25G0, only one thermal event is observed between 300°C and 380°C, exhibiting a considerable mass loss due to the decomposition of CA. Compared to that of the C25G0 nanofiber membrane, the TGA curves of C25G5 and C25G10 nanofiber membranes exhibit two additional thermal events. The first thermal event is observed in the temperature range of 120–200°C, and the second event is observed in the temperature range from 250°C to 300°C. The first event is initiated at 120°C due to the evaporation of absorbed water. Because glycerol is a well-known moisturizer, the glycerol present nanofibers can easily absorb water. The second thermal event starts at 250°C and continues to 300°C before CA degradation and is attributed to the volatilization of glycerol linked between CA chains (Teixeira et al. 2021; Hong, Cho, and Kang 2020). The second thermal event is more clearly seen in the TGA curve of C25G10, indicating the retention of glycerol in prepared CA nanofiber membranes.

3.2. Characterization of MAPTAC-grafted RC nanofiber membranes

Table 2

Fiber diameters and pore sizes of C25G10-CA, C25G10-RC, and C25G10-RC-MAPTAC nanofiber membranes.
Sample	Fiber diameter (nm)	Pore size (µm)
C25G10 (CA)	1,310 ± 250	5.90 ± 0.31
C25G10 (RC)	1,030 ± 227	2.96 ± 0.30
C25G10 (RC-MAPTAC)	1,014 ± 181	1.73 ± 0.08

In the deacetylation process, membrane contraction is observed due to the loss of acetyl groups and formation of additional interchain hydrogen bonding networks of cellulose. As shown in Table 2, the pore size decreases from 5.90 µm to 2.96 µm due to the membrane contraction. Furthermore, high alkalinity induces several degradation reactions along the cellulose backbone, which may have partly occurred during the alkaline hydrolysis of these nanofiber membranes. The partial degradation of cellulose chains likely induces the rearrangement of the membrane structure immersed in an aqueous alkaline solution, which in turn decreases the pore size of the membrane (Tekin and Çulfaz-Emecen 2023; Zaborniak and Chmielarz 2023; Lindman et al. 2017).

The deacetylation and modification of the C25G10 membrane was confirmed using FTIR spectroscopy, as shown in Fig. 6a. In the FTIR spectrum of the C25G10-RC nanofiber membrane, the intensity of the –OH peak in the range of 3,300–3,500 cm^− 1 increases, whereas that of the peaks of the acetyl group in the CA membrane (C = O at 1,734 cm^− 1, C–H stretching at 2,933 and 2,875 cm^− 1, C–O–C at 1,220 cm^− 1) decreases after deacetylation. This indicates that the deacetylation process successfully replaces the acetyl group with a hydroxyl group (Serbanescu et al. 2020). The grafting of MAPTAC on the C25G10-RC nanofiber membrane was also confirmed via FTIR spectroscopy. A peak corresponding to quaternary ammonium is observed at 1,481 cm^− 1 in the FTIR spectrum of C25G10-MAPTAC, which is not present in the FTIR spectra of C25G10 and C25G10-RC. The specific peak values of the MAPTAC functional groups are 3,356 cm^− 1, 1,650 cm^− 1, 1,479 cm^− 1, 1,378 cm^− 1, and 1,314 cm^− 1. Furthermore, nitrogen was detected in C25G10-MAPTAC using SEM energy-dispersive X-ray spectroscopy (EDS), as shown in Fig. 6d. These results confirm the successful attachment of the quaternary ammonium ligand to the C25G10-RC nanofiber membrane (Li et al. 2021). The morphological changes in the C25G10-RC nanofiber after alkaline hydrolysis cannot be visualized in Fig. 6b and 6c, but the fiber diameter and membrane thickness considerably decrease. Although the fiber diameter does not considerably change after ligand modification, the membrane thickness and pore size further decrease after MATPAC modification, as shown in Fig. 6d and Table 2, due to the entanglement of nanofibers by chemical cross linking and physical squeezing during modification processes. Finally, the prepared nanofiber membrane adsorber exhibits a thickness of 200 $\:{\mu\:}\text{m}$ and an average pore size of 1.73 $\:{\mu\:}\text{m}$.

3.3. Evaluation of membrane chromatography

Highly permeable membranes with a uniform pore size are preferred in membrane chromatography because the convection-based well-distributed adsorption of large molecules in 10–20-layer stacked membranes with low permeation resistance is needed. Furthermore, a large surface area modified with adsorption functional groups is needed to obtain a high adsorption capacity. Because there is a trade-off relationship between the pore size and surface area at the same porosity, the pore size and membrane permeance should be considered for membrane chromatography applications. Furthermore, the ligand modification process can lead to a narrow internal structure and reduced membrane permeance and adsorption capacity when the pore size is too small (Jan Schwellenbach 2016b; Jan Schwellenbach 2016a). Therefore, it is important to tailor the membrane and ligand to create a favorable structure, considering factors such as binding capacity and permeability.

Table 3

Pore size, porosity, SBC, DBC_10% of RC nanofiber–based membrane adsorbers at 1, 10, and 20 MV/min.
Sample	Pore size (µm)	Porosity (%)	SBC (mg/mL)	DBC_10% (mg/mL)
Sample	Pore size (µm)	Porosity (%)	SBC (mg/mL)	1 MV/min	10 MV/min	20 MV/min
C25G5-RC-MAPTAC	0.783 ± 0.03	58.0	273.40	101.25	97.69	88.88
C25G10-RC-MAPTAC	1.725 ± 0.08	58.9	239.92	122.53	113.49	112.09

C25G5-RC-MAPTAC and C25G10-RC-MAPTAC with similar porosity were tested, and their binding capacities are shown in Table 3. The SBC of C25G5-RC-MAPTAC is 273.4 mg/mL whereas that of C25G10-RC-MAPTAC is 239.9 mg/mL. Because the SBC indicates the maximum binding capacity of membranes at equilibrium, the SBC increases with a decrease in the membrane pore size (i.e., increase in the surface area). However, the DBC_10% of C25G10-RC-MAPTAC is higher than that of C25G5-RC-MAPTAC, implying a membrane with a larger pore size can more effectively adsorb protein under permeation conditions. Even though C25G5-RC-MAPTAC has a high ligand density, some small pores are not active or blocked during BSA filtration, resulting in low DBC_10% values. The DBC_10% of the membranes under different flow rate conditions was also measured. An increase in the flow rate shortens the contact time between the adsorbent and target material (Conidi, Parker, and Smith 2019). The C25G5-RC-MAPTAC nanofiber membrane exhibits the adsorption performance of 101.25 mg/mL at 1 MV/min, 97.69 mg/mL at 10 MV/min, and 88.88 mg/mL at 20 MV/min. In this case, a relatively large decrease in DBC_10% is observed at high flow rates. C25G10-RC-MAPTAC exhibits the highest DBC_10% of 122.53 mg/mL at 1 MV/min, but the DBC_10% decreases to 113.49 mg/mL at > 10 MV/min. These results show that an increase in the flow rate decreases the contact time between the BSA and membrane, and the adsorption capacity eventually decreases because the residence time decreases due to the high flow rate (Conidi, Parker, and Smith 2019; Liang et al. 2022). Consequently, the pore size and flow rate affect the adsorption efficiency of the membrane adsorbers.

Table 4

Comparison of DBC_10% of C25G10-RC-MAPTAC, reported membranes in literatures, commercial membrane adsorbers, and chromatography resin.
Name	Properties	Flow rate	SBC	DBC_10% of BSA	Reference
C25G10-MAPTAC	Cellulose nanofiber	20 MV/min	220 mg/ml	112.09 mg/ml	This work
PSF-GMA-DEA	Nanofibers	2.85 MV/min	260 mg/ml	201.3 mg/ml	(Chen, Wickramasinghe, and Qian 2020)
PAN-GMA-DMA			40 mg/ml	87.2 mg/ml
S 0% CL	Polypropylene membrane	-	160 mg/ml	-	(He and Ulbricht 2008)
A 0mM	Polypropylene membrane	-	120 mg/ml	-	(He and Ulbricht 2008)
PBT-GMA-DEA	poly(butylenetherephthalate) nonwoven	0.1 ml/min	1.3 g/g	40.62 mg/ml	(Lemma, Boi, and Carbonell 2021)
EA-C	Polyethylene hollow fiber	9 L/min	26 mg/ml	17 mg/ml	(Kubota et al. 1996)
Poly(DMAEMA)	RC membrane	87 MV/min	130 mg/ml	97.5 mg/ml	(Bhut and Husson 2009)
Nano-DEAE	Cellulose nanofiber	4 cm/min	10.7 mg/ml	5.0 mg/ml	(Hardick et al. 2013)
Sartobind Q	Reinforced cellulose membrane	20 MV/min	-	29 mg/ml
Mustang Q	Polyethersulfone membrane	20 MV/min	-	50 mg/ml
Natrix HD-Q	Supported hydrogel	5–25 MV/min	-	200 mg/ml
Eshmuno Q	Q resin	100 CV/h	-	96 mg/ml
Capto Q	Q resin	35 CV/h	-	100 mg/ml
Q Sepharose	Q resin	47 CV/h	-	42–70 mg/ml
*The properties of commercial products are from the manufacture datasheets. CV : column volume

The BSA adsorption capability of the MC membrane, specifically C25G10-RC-MAPTAC nanofiber membrane fabricated in this study, was compared to that of other commercial MC membranes. The results in Table 4 show that C25G10-MAPTAC exhibits considerably superior static and dynamic BSA adsorption capacities. Although the direct performance comparison with the reported data from the literatures due to different operating conditions and binding capacity evaluation methods, the prepared nanofiber based C25G10-MAPTAC exhibited excellent dynamic binding capacity compared to various nonwoven, microfiltration membrane, and nanofiber based MC membranes. In addition, it showed competitive flow rate and dynamic binding capacity compared to chromatography resin products implying a promising applications of MC processes for fast and low cost purification process operation. Notably, when examining the BSA adsorption capacity relative to the pore size, it is evident that C25G10-MAPTAC is superior to other commercial MC membranes. This notable performance can be attributed to the grafted MAPTAC layer formed on highly porous nanofiber RC membranes. Briefly, the combination of a large pore size and elevated protein adsorption capacity signifies the potential of the RC nanofiber-based membrane adsorbers for achieving highly efficient separation and purification with substantial flow rates in biopharmaceutical production, separation, and purification processes (Şimşek, Eroğlu, and Erbil 2019).

The fabricated CA nanofiber membrane shows a beaded or irregular fiber morphology in the absence of glycerol even when the CA concentration is increased due to the low viscosity of the dope solution. When glycerol is added, the regularity of CA fibers improves and the beaded morphology transforms into the fiber morphology. This improvement is attributed to the enhanced chain entanglement achieved via increased intermolecular forces facilitated by hydrogen bonding with glycerol. ATR-FTIR analysis reveals that the intensities of O–H and C–H stretching peaks increase, and XRD analysis shows that the crystallinity of CA decreases from 39.8–12.2% by adding glycerol. The C25G10 membrane exhibits superior properties compared with other fabricated membranes and is used as a membrane adsorber. For this purpose, the C25G10 is deacetylated and further grafted by MAPTAC. ATR-FTIR, SEM, and EDS confirm the deacetylation and grafting, and these modifications decrease the fiber diameter and pore size. Performance evaluation reveals that C25G10-MAPTAC has a high SBC of 218 mg/mL and a DBC_10% of 113 mg/mL, highlighting its potential for high-performance applications in ion exchange chromatography. Overall, incorporating glycerol in CA electrospinning is a promising approach for developing efficient and versatile membranes with enhanced properties.

Acknowledgments

This work was supported by the Ministry of Science and ICT (No. KK2411-10) and the Ministry of Trade, Industry & Energy of the Republic of Korea (20017410).

Funding Declaration

This work was funded by the Ministry of Science and ICT (No. KK2411-10) and the Ministry of Trade, Industry & Energy of the Republic of Korea (20017410).

Conflict of Interest

Authors have no financial or non-financial interests that are related to the work submitted for publication.

Ethics Approval

Not applicable.

Consent for Publication

All authors agree to publish this work.

Alessandro Bonifacio, Lorenzo Bonetti, Etienne Piantanida, Luigi De Nardo 2023. 'Plasticizer design strategies enabling advanced applications of cellulose acetate', European Polymer Journal, 197: 10.
Bhut, Bharat V., and Scott M. Husson. 2009. 'Dramatic performance improvement of weak anion-exchange membranes for chromatographic bioseparations', Journal of Membrane Science, 337: 215-23.
Chen, Jinghuan, Jikun Xu, Kun Wang, Xuefei Cao, and Runcang Sun. 2016. 'Cellulose acetate fibers prepared from different raw materials with rapid synthesis method', Carbohydrate Polymers, 137: 685-92.
Chen, Shu-Ting, S. Ranil Wickramasinghe, and Xianghong Qian. 2020. 'Electrospun Weak Anion-Exchange Fibrous Membranes for Protein Purification', Membranes, 10: 39.
Conidi, D., W. J. Parker, and S. Smith. 2019. 'Effect of solids residence time on dynamic responses in chemical P removal', Water Environ Res, 91: 250-58.
Genceli, Esra, Reyhan Sengur, Gulsum Urper-Bayram, Selin Gümrükçü, Zeliha Güler, Ugur Dagli, Turker Turken, A. Sarac, and Ismail Koyuncu. 2018. 'Effects of carboxylated multi-walled carbon nanotubes having different outer diameters on hollow fiber ultrafiltration membrane fabrication and characterization by electrochemical impedance spectroscopy', Polymer Bulletin, 75.
Gonçalves, Sheyla Moreira, Daiane Cardial dos Santos, Joyce Fagundes Gomes Motta, Regiane Ribeiro dos Santos, Davy William Hidalgo Chávez, and Nathália Ramos de Melo. 2019. 'Structure and functional properties of cellulose acetate films incorporated with glycerol', Carbohydrate Polymers, 209: 190-97.
Hardick, Oliver, Stewart Dods, Bob Stevens, and Daniel G. Bracewell. 2013. 'Nanofiber adsorbents for high productivity downstream processing', Biotechnology and bioengineering, 110: 1119-28.
He, Dongming, and Mathias Ulbricht. 2008. 'Preparation and characterization of porous anion-exchange membrane adsorbers with high protein-binding capacity', Journal of Membrane Science, 315: 155-63.
Hokkanen, Sanna, Amit Bhatnagar, and Mika Sillanpää. 2016. 'A review on modification methods to cellulose-based adsorbents to improve adsorption capacity', Water Research, 91: 156-73.
Hong, Seong Ho, Younghyun Cho, and Sang Wook Kang. 2020. 'Highly porous and thermally stable cellulose acetate to utilize hydrated glycerin', Journal of Industrial and Engineering Chemistry, 91: 79-84.
Huihua Liu, Raju Adhikari , Qipeng Guo , Benu Adhikari. 2013. 'Preparation and characterization of glycerol plasticized (high-amylose) starch–chitosan films', Journal of food engineering, 116: 588-97.
Jan Schwellenbach, Florian Taft, Louis Villain, Jochen Strube. 2016a. 'Preparation and characterization of high capacity, strong cation-exchange fiber based adsorbents', Journal of Chromatography A, 1447: 92-106.
Jan Schwellenbach, Steffen Zobel, Florian Taft, Louis Villain and Jochen Strube. 2016b. 'Purification of Monoclonal Antibodies Using a Fiber Based Cation-Exchange Stationary Phase: Parameter Determination and Modeling', bioengineering, 3.
Kubota, Noboru, Suguru Miura, Kyoichi Saito, Kazuyuki Sugita, Kohei Watanabe, and Takanobu Sugo. 1996. 'Comparison of protein adsorption by anion-exchange interaction onto porous hollow-fiber membrane and gel bead-packed bed', Journal of Membrane Science, 117: 135-42.
Lemma, Solomon Mengistu, Cristiana Boi, and Ruben G. Carbonell. 2021. 'Nonwoven Ion-Exchange Membranes with High Protein Binding Capacity for Bioseparations', Membranes, 11: 181.
Li, Xiang, Huaili Zheng, Baoyu Gao, Yongjun Sun, Bingzhi Liu, and Chuanliang Zhao. 2017. 'UV-initiated template copolymerization of AM and MAPTAC: Microblock structure, copolymerization mechanism, and flocculation performance', Chemosphere, 167: 71-81.
Li, Xiaowei, Haiyang Zhang, Jing Dong, Shujuan Ma, and Junjie Ou. 2021. 'One-pot synthesis of glucose-derived carbonaceous material with high hydrophilicity and adsorption capacity as bilirubin adsorbent', Journal of Materials Science, 56: 18006-18.
Liang, Qihui, Xinxi Fu, Ping Wang, Xinxian Li, and Peiyuan Zheng. 2022. 'Dynamic Adsorption Characteristics of Phosphorus Using MBCQ', Water, 14: 508.
Lindman, Björn, Bruno Medronho, Luis Alves, Carolina Costa, Håkan Edlund, and Magnus Norgren. 2017. 'The relevance of structural features of cellulose and its interactions to dissolution, regeneration, gelation and plasticization phenomena', Physical Chemistry Chemical Physics, 19: 23704-18.
Liu, Chunxiu, and Renbi Bai. 2005. 'Preparation of chitosan/cellulose acetate blend hollow fibers for adsorptive performance', Journal of Membrane Science, 267: 68-77.
Liu, Haiqing, and You-Lo Hsieh. 2002. 'Ultrafine fibrous cellulose membranes from electrospinning of cellulose acetate', Journal of Polymer Science Part B: Polymer Physics, 40: 2119-29.
Liu, Haiqing, and You-Lo Hsieh. 2003. 'Surface methacrylation and graft copolymerization of ultrafine cellulose fibers', Journal of Polymer Science Part B: Polymer Physics, 41: 953-64.
Liu, Haiqing, and Chunyi Tang. 2007. 'Electrospinning of Cellulose Acetate in Solvent Mixture N,N-Dimethylacetamide (DMAc)/Acetone', Polymer Journal, 39: 65-72.
Lucena, Maria da Conceição C., Ana Ellen V. de Alencar, Selma Elaine Mazzeto, and Sandra de A. Soares. 2003. 'The effect of additives on the thermal degradation of cellulose acetate', Polymer Degradation and Stability, 80: 149-55.
Majumder, Sushmita, Md Abdul Matin, Ahmed Sharif, and M. Tarik Arafat. 2019. 'Understanding solubility, spinnability and electrospinning behaviour of cellulose acetate using different solvent systems', Bulletin of Materials Science, 42: 171.
Naragund, Veereshgouda S., and P. K. Panda. 2020. 'Electrospinning of cellulose acetate nanofiber membrane using methyl ethyl ketone and N, N-Dimethylacetamide as solvents', Materials Chemistry and Physics, 240: 122147.
Nie, Huarong, Aihua He, Jianfen Zheng, Shanshan Xu, Junxing Li, and Charles C. Han. 2008. 'Effects of Chain Conformation and Entanglement on the Electrospinning of Pure Alginate', Biomacromolecules, 9: 1362-65.
Nusrat Sharmin, Izumi Sone, James Leon Walsh, Morten Sivertsvik, Estefanía Noriega Fernandez 2021. 'Effect of citric acid and plasma activated water on the functional properties of sodium alginate for potential food packaging applications', Food Packaging and Shelf Life, 29: 10.
Rafael Erdmann , Stephan Kabasci and Hans-Peter Heim. 2021. 'Thermal Properties of Plasticized Cellulose Acetate and Its β-Relaxation Phenomenon', Polymers, 13.
Saxena, Pooja, Prashant Shukla, and MS Gaur. 2021. 'Thermal analysis of polymer blends and double layer by DSC', Polymers and Polymer Composites, 29: S11-S18.
Serbanescu, Oana Steluta, Andreea Madalina Pandele, Florin Miculescu, and Stefan Ioan Voicu. 2020. 'Synthesis and Characterization of Cellulose Acetate Membranes with Self-Indicating Properties by Changing the Membrane Surface Color for Separation of Gd(III)', Coatings, 10: 468.
Shah, Aatif Ali, Young Hoon Cho, Seung-Eun Nam, Ahrumi Park, You-In Park, and Hosik Park. 2020. 'High performance thin-film nanocomposite forward osmosis membrane based on PVDF/bentonite nanofiber support', Journal of Industrial and Engineering Chemistry, 86: 90-99.
Shahriar, S. M. Shatil, Jagannath Mondal, Mohammad Nazmul Hasan, Vishnu Revuri, Dong Yun Lee, and Yong-Kyu Lee. 2019. 'Electrospinning Nanofibers for Therapeutics Delivery', Nanomaterials, 9: 532.
Şimşek, Ceyda, Zeynep Ezgi Eroğlu, and Candan Erbil. 2019. 'Effect of ionomer/multiplet formation on mechanical properties and ascorbic acid release behavior of PNIPAAm hydrogels copolymerized by DMAEMA, DMAPMAAm and MAPTAC', Iranian Polymer Journal, 28: 977-90.
Singh, Naveen Kumar, Roy N. Dsouza, Mariano Grasselli, and Marcelo Fernández-Lahore. 2014. 'High capacity cryogel-type adsorbents for protein purification', Journal of Chromatography A, 1355: 143-48.
Singh, Nripen, Zhen Chen, Namrata Tomer, S. Ranil Wickramasinghe, Neil Soice, and Scott M. Husson. 2008. 'Modification of regenerated cellulose ultrafiltration membranes by surface-initiated atom transfer radical polymerization', Journal of Membrane Science, 311: 225-34.
Singh, Nripen, Jun Wang, Mathias Ulbricht, S. Ranil Wickramasinghe, and Scott M. Husson. 2008. 'Surface-initiated atom transfer radical polymerization: A new method for preparation of polymeric membrane adsorbers', Journal of Membrane Science, 309: 64-72.
Son, Won Keun, Ji Ho Youk, Taek Seung Lee, and Won Ho Park. 2004. 'Electrospinning of ultrafine cellulose acetate fibers: Studies of a new solvent system and deacetylation of ultrafine cellulose acetate fibers', Journal of Polymer Science Part B: Polymer Physics, 42: 5-11.
Tabe, Shahram. 2017. 'A Review of Electrospun Nanofber Membranes', Journal of Membrane Science and Research, 3: 228-39.
Teixeira, Samiris Côcco, Rafael Resende Assis Silva, Taíla Veloso de Oliveira, Paulo César Stringheta, Marcos Roberto Moacir Ribeiro Pinto, and Nilda de Fátima Ferreira Soares. 2021. 'Glycerol and triethyl citrate plasticizer effects on molecular, thermal, mechanical, and barrier properties of cellulose acetate films', Food Bioscience, 42: 101202.
Tekin, Fatma Seden, and P. Zeynep Çulfaz-Emecen. 2023. 'Controlling Cellulose Membrane Performance via Solvent Choice during Precursor Membrane Formation', ACS Applied Polymer Materials, 5: 2185-94.
Tiwari, Sandeep Kumar, and Subbu S. Venkatraman. 2012. 'Importance of viscosity parameters in electrospinning: Of monolithic and core–shell fibers', Materials Science and Engineering: C, 32: 1037-42.
Tungprapa, Santi, Tanarinthorn Puangparn, Monchawan Weerasombut, Ittipol Jangchud, Porntiva Fakum, Somsak Semongkhol, Chidchanok Meechaisue, and Pitt Supaphol. 2007. 'Electrospun cellulose acetate fibers: effect of solvent system on morphology and fiber diameter', Cellulose, 14: 563-75.
Wan Daud, Wan Rosli, and Fauzi Muhammad Djuned. 2015. 'Cellulose acetate from oil palm empty fruit bunch via a one step heterogeneous acetylation', Carbohydrate Polymers, 132: 252-60.
Xue, Jiajia, Tong Wu, Yunqian Dai, and Younan Xia. 2019. 'Electrospinning and Electrospun Nanofibers: Methods, Materials, and Applications', Chemical Reviews, 119: 5298-415.
Yan, Ming, Bihui An, Zhenqiang Zai, Lili Zhang, Xin Li, Shufang Wu, and Jinxia Ma. 2023. 'The mechanism of acetyl groups regulating the morphology and photocatalytic properties of ZnO and its cellulose composite', Cellulose, 30: 6869-85.
Ye, Sang Ho, Junji Watanabe, and Kazuhiko Ishihara. 2004. 'Cellulose acetate hollow fiber membranes blended with phospholipid polymer and their performance for hemopurification', Journal of Biomaterials Science, Polymer Edition, 15: 981-1001.
Zaborniak, Izabela, and Paweł Chmielarz. 2023. 'Polymer-modified regenerated cellulose membranes: following the atom transfer radical polymerization concepts consistent with the principles of green chemistry', Cellulose, 30: 1-38.
Zhang, Lifeng, Todd J. Menkhaus, and Hao Fong. 2008. 'Fabrication and bioseparation studies of adsorptive membranes/felts made from electrospun cellulose acetate nanofibers', Journal of Membrane Science, 319: 176-84.
Zhou, Qi, Yuping Bao, Hao Zhang, Qian Luan, Hu Tang, and Xiuting Li. 2020. 'Regenerated cellulose-based composite membranes as adsorbent for protein adsorption', Cellulose, 27: 335-45.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
01 Sep, 2024
Editor assigned by journal
31 Aug, 2024
Submission checks completed at journal
31 Aug, 2024
First submitted to journal
22 Aug, 2024

You are reading this latest preprint version

Preparation of Homogeneous Regenerated Cellulose Nanofiber based Membrane Adsorber for Membrane Chromatography Applications

Status:

Version 1

Abstract

Figures

1. Introduction

2. Experimental Section

2.1. Materials

2.2. Electrospinning

2.3. Nanofiber membrane modifications

2.4. Characterizations of nanofiber membranes

2.5. Static binding capacity (SBC)

2.6. Dynamic binding capacity

3. Results and Discussion

3.1. Characterization of the CA nanofiber membrane

3.2. Characterization of MAPTAC-grafted RC nanofiber membranes

3.3. Evaluation of membrane chromatography

4. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1