The structure and properties of cellulose acetate materials: A comparative study on electrospun membranes and casted films

Materials

CA (white powder; degree of substitution, 2.43; degree of polymerization, 800–1000; molecular weight of single unit approximately, 303) was obtained from Nantong Acetate Fiber Company in Jiangsu Province, China.

The solvents used in this study were acetone (analytical grade; CH₃COCH₃; molecular weight, 58.08; boiling point approximately 56℃; volatile) and N,N-dimethylacetamide (DMAc; analytical grade; CH₃CON(CH₃)₂; molecular weight, 87.12; boiling point approximately 160℃), all from Shanghai Lingfeng Chemical Reagent Co Ltd China. These chemicals were used as received.

Preparation of polymer solutions and CA films

The 8–13% (w/w) CA solutions were prepared using 2:1 (v/v) acetone/DMAc mixture solvent and gently stirred at room temperature until complete dissolution of polymer occurred.

The CA films were obtained from a 11% solution (w/w) of CA in 2:1 (v/v) acetone/DMAc mixture solvent. The films were cast by drawing on glass plates with a glass rod, followed by air drying for 48 h at room temperature.

Electrospinning setup

The experimental setup used for the electrospinning process is shown in Figure 1. The CA solution was poured into a syringe attached with a capillary tip of 1 mm diameter. A variable high power supply was used to produce a high DC potential across the capillary tip and an aluminum sheet used as the stationary screen collector. The electrode of positive polarity that was emitted was attached to the needle. The voltage used in the experiment was 15 kV and the current was adjusted to be constant. The flow rate was 0.8–1.0 mL/h and the distance between the tip and the collector was fixed at 20 cm. The electrospun nanofibers membranes were air-dried at room temperature for 48 h to remove residual solvent. OPEN IN VIEWER

Measurements of CA materials

The morphology and pore distribution of electrospun membranes and casted films

When only acetone was selected as the solvent for electrospinning, the tip of the needle was easily blocked, which is attributed to the low boiling point of acetone. However, when pure DMAc was selected as the solvent, droplets of large size collected on the aluminum sheet. These phenomena were consistent with the results of Liu and Hsieh [11], who showed that both acetone and DMAc were not a desirable solvent for electrospinning CA. Therefore a mixture of acetone/DMAc solvent system at the ratio of 2:1(v/v) was used to dissolve CA and fabricate electrospun membranes.

Figure 2(a) to (f) shows the SEM images of CA-electrospun membranes fabricated from CA solutions of differing concentrations. When low CA concentrations (8–10%) were used for electrospinning, nanofibers with beads were generated. The numbers of beads decreased, the beads were larger in size and the distances between beads were increased, as the CA concentration increased. Meanwhile, the beads gradually changed in shape from an almost spherical to a spindle-like shape. It is known that the viscosity of the solution is one of the most important factors leading to the formation of beaded nanofibers [15,16]. Higher viscosity favored formation of fibers without beads. Improving CA concentration could lead to higher viscosity, so the beads of CA nanofibers would be reduced. In addition, Figure 2(c) shows that the nanofibers produced from 10% CA solution in the had an increased crosslinking phenomenon. OPEN IN VIEWER

As shown in Figure 2(d) to (f), nanofibers without beads were generated. With the increasing of CA concentration, the diameter of the nanofibers showed a growth trend; however, the uniformity of the nanofibers reduced. Figure 3 shows that the average diameters of nanofibers increased from 333 to 373 nm with CA concentrations ranging from 11–13% (w/w). When a higher CA concentration was used, the error bar of the average diameter was larger, indicating that the CA nanofibers obtained from higher concentration were more uneven. So the electrospun membrane fabricated from 11% CA solution was selected to conduct a comparative analysis with casted film. OPEN IN VIEWER

The casted film was also produced from 11% CA solution. The morphology of casted film is shown in Figure 2(g). A certain number of pores were found to exist on the surface of casted film; however, the size of pores was uneven.

Figure 4 shows the pore size distribution of electrospun membrane and casted film, both produced from 11% CA solution. It was found that the casted film possessed a larger pore size distribution interval that ranged from 0.127 to 13.35 µm, which revealed that the pore size was extraordinarily uneven. Moreover, the main pore size of casted film ranged from 0.127 to 0.30 µm (as the inset shows in Figure 4(a)) and pore size more than 1µm accounted for only 0.1%. In contrast, the pore size distribution of electrospun membrane was more uniform, showing pore size distribution interval ranging from 1.884 to 5.185 µm (as shown in Figure 4(b)). As shown in Table 1, the mean pore size of electrospun membrane was larger than that of casted film. This shows that the pore size distribution of electrospun membrane was more uniform, although the mean pore size was larger. OPEN IN VIEWER

OPEN IN VIEWER

Table 1.

The pore distribution parameters of electrospun membrane and casted film, both produced from 11% CA solution.

Parameter sample	Minimum pore size (µm)	Mean pore size (µm)	Maximum pore size (µm)
Casted film	0.127	0.165	13.35
Electrospun membrane	1.884	3.552	5.185

XRD characterization of CA materials

Figure 5 shows the X-ray diffraction patterns of three different CA materials. In the whole ranges of 2θ angle, it was found that the amorphous peak area of original CA material was significantly larger than that of casted film and electrospun membrane. To determine the degree of crystallization of three different CA materials, the conventional X-ray analysis was adopted and the XRD pattern was divided into two different contributions, amorphous and crystalline phases [17]. Van der Waals halo or amorphous halo, corresponding to the amorphous region of the material, was used to deconvolute the XRD pattern. It is known that the broad halo was usually located around 20° and presented in all organic polymers [18]. OPEN IN VIEWER

For original CA material (Figure 5(a)), when the 2θ angle ranged from 5° to 14°, there were three small intensive peaks, respectively, at 9°, 10° and 13°. In addition, one main diffuse peak exhibited at 20°. Therefore, the fibril of original CA material was constituted by crystalline and amorphous blocks and the maximum diffuse peak suggested that this material possessed basically amorphous characteristics. According to the calculation of software MDI jade, the crystallinity of original CA material was 28.24% and the ratio of corresponding amorphous region was 71.76%. Hydrogen bonding between the chains and water uptake seem to be responsible for this noncrystalline structure [19].

As shown in Figure 5(b), when the 2θ angle ranged from 5° to 60°, the casted CA film had a main intensive peak at 9° and an amorphous peak at 18°. According to the calculation of software MDI jade, the crystallinity of casted CA film was 34.41%. Compared with original CA material, the crystallinity increased by 6.71%, mainly due to the effect of the solvent. When CA was dissolved in the acetone/DMAc solvent mixture, hydrogen bonding between the chains was destroyed and recrystallization phenomenon occurred. Thus the crystallinity of casted CA film increased.

Figure 5(c) shows that the electrospun membrane had an obvious intensive peak at 9°, which was similar to that in the casted film. In addition, a diffuse peak exhibited at 24°. Through the calculation of software MDI jade, the crystallinity of electrospun fibrous membrane was shown to be 51.02%, which showed a dramatic improvement compared with the other CA materials. Kim et al. [17] and Zhao et al. [20] also obtained electrospun cellulose and cellulose derivative nanofibers with high crystallinity, i.e. 40–60%. There were two reasons for the increasing crystallinity of electrospun CA membranes. On the one hand, recrystallization occurred when CA was dissolved in the acetone/DMAc solvent mixture. On the other hand, the CA solution jet and the polymer macromolecule suffered more stretching in the electrospinning process; hence, the crystallinity increased. Qin and Wang [21] and Qin [22] also found that the stretching of electrostatic force is propitious to the formation of crystal of nanofibers.

DSC characterization of CA materials

The thermal data varied greatly for different materials depending on the methods of preparation and analysis [23], so investigations were undertaken under the same conditions. When the DSC test was adopted for the three samples, their initial weights, respectively, were 5.16 mg (Figure 6(a)), 5.39 mg (Figure 6(b)) and 5.44 mg (Figure 6(c)). The DSC thermograms of the three CA materials are shown in Figure 6. OPEN IN VIEWER

Figure 6 shows that for the three samples a similar endothermic hump was observed between 30 and 110℃. This hump appeared at the onset temperature (T_o) of the measurement and was difficult to evaluate. It was believed that this may have been caused by the evaporation of water. However, the hump disappeared in secondary heating curves of all the three CA materials (not shown in this study), indicating that this event could indeed be attributed to water evaporation.

Unfortunately, pronounced glass transition was not observed in all three samples, mainly due to the disturbance of crystalline regions. The secondary heating curves (not shown in this study) showed a similar glass transition temperature of 185℃ for the three samples.

The melting peak and fusion enthalpy were observed in all three curves. It revealed that the three samples appeared approximately at the same T_m, around 237℃. The fusion enthalpy of electrospun CA membrane was about 15.2 J/g, indicating that the electrospun CA membrane presented a more crystalline structure in accordance with the XRD data. However, the original CA material was on the higher side of fusion enthalpy, which was in contradiction with the XRD data. It may be attributed to the slight decomposition of the original CA material.

TGA of CA materials

When TGA was conducted for the three samples, their initial weights, respectively, were 8.4 mg (Figure 7(a)), 8.5 mg (Figure 7(b)) and 7.6 mg (Figure 7(c)). As shown in Figure 7, the three samples showed similar changing trend in the whole ranges of temperature. The decomposition appeared between 350℃ and 410℃, mainly due to a cellulose degradation process such as depolymerization, dehydration and the decomposition of glucosyl. Eventually, charred residues were formed [24,25]. OPEN IN VIEWER

It is known that the T_o exhibited the best repeatability in the TGA cures, so this point was chosen to measure the thermal stability of materials. Table 2 shows the characteristic parameters of TGA curves. The T_o of original CA material was the highest, located at about 377℃, indicating that the thermal stability of casted film and electrospun membrane decreased. However, the electrospun membrane showed better thermal stability, compared with casted film.

OPEN IN VIEWER

Table 2.

The characteristic parameters of thermoanalytical curves on TGA.

Sample parameter	Original CA	Casted film	Electrospun membrane
To (℃)	376.9	359.5	366
Te (℃)	408	398	409.7
Te−To (℃)	31.1	38.5	43.7
Weight loss ratio (%)	76.45	81.49	87.57

As for the whole temperature range of decomposition (T_e−T_o), the electrospun membrane had the largest temperature difference, i.e. 43.7℃, compared with the original material and casted film. It was concluded that the inner structure was more stable; therefore, larger temperature difference would be taken to destroy the chemical structure.

The weight loss ratio could, only to some extent, reflect the change in internal structure and property within the material, but it was not entirely absolute. In the actual testing process, the weight loss ratio was affected by many factors, such as the initial weight and the testing temperature rate. Therefore, the weight loss rate was not used to measure the thermal stability in this study.

Surface hydrophilicity of electrospun membranes and casted films

The hydrophilicity of the CA electrospun membrane and casted film, which may affect their overall performance, was examined by contact angle measurement to compare the effect of nanotexture and casted texture on the surface hydrophilicity. Figure 8 shows that the surface of the casted film was hydrophilic, reporting contact angle of approximately 74.4°. However, the electrospun membrane gave a contact angle of 127.2° and was observably more hydrophobic than the casted film. OPEN IN VIEWER

In theory, the wettability of liquid droplets on solid surfaces was governed mainly by both the chemical composition and the geometrical microstructure [26]. Conventionally, it was difficult to fabricate hydrophobic surface from hydrophilic materials. Herminghaus [27] reported that it was theoretically possible to construct a waterrepellent surface from a material using a material with a contact angle less than 90°. It was proved by Feng [28] and Feng’s group fabricated a superhydrophobic surface (Contact angle = 171.2°) from amphiphilic poly(vinyl alcohol).

In this study, the hydrophobic surface was obtained from hydrophilic CA material by electrospinning. This phenomenon may be attributed to the rearrangement of CA molecules during electrospinng process, which may be similar with the PVA molecular rearrangement found by Lin [28]. Hydrophobic groups were reoriented at the air/solid interface and the hydrogen bonds were formed in the interiors. The other reason was that the electrospun fibrous membrane possessed higher crystallinity (verified in Figure 5) compared with casted film, which was difficult for the diffusion and translation of water molecules. Therefore, it exhibited higher surface hydrophobicity.