A novel preparation of anti-layered poly(vinylalcohol)–polyacrylonitrile (PVA/PAN) membrane for air filtration by electrospinning

Abstract

A novel 3D filtration medium for airborne particles was prepared using an electrospinning technique, wherein PVA and PAN were used as a reinforcing agent and a bulk material for the filtration membrane, respectively. In order to improve the filtration efficiency and mechanical property of the membrane, the membrane was modified by a certain concentration mixture of GA, acetone and acetic acid, which could reduce the number of hydroxyl groups, promote its stability and increase the compactness via a bond between fibers. Simultaneously, the mechanical property of the membrane was measured by a tensile tester, and the results demonstrated that the maximum strain and toughness reached 144.58% and 5.33 J g⁻¹, respectively. After that, the membrane was used as an air purification material for testing air particles content and the particles intercept principle has been discussed. The filtration performance of the membrane demonstrated that the membrane possessed excellent removing efficiency for air suspension with lower pressure drop compared with commercial masks under the same testing conditions. The filtration efficiency for the particles, sized 0.3 to 7 μm, was found to be approximately 99%. Furthermore, we also developed an effective avenue to fabricate a filtration medium based on the a steel wire for applications in ultra condition.

---

Introduction

Air purification has attracted more attention following the increasingly serious environment problems.¹ The particulate matter (PM) pollution had been confirmed as the major cause of jeopardizing the human respiratory tract, which is one of the acute challenges faced in the world today. The industrial production and automotive emissions have aggravated the phenomenon of excessive fine particles.^2,3 As we all know that PM are mixtures of small particles absorbed by lots of viruses and bacteria, and therefore, it poses a serious threat to public body health, making it necessary to avoid these polluted materials entering into our body organs. Recently, hazy days have become more frequent than before. Visibility has decreased severely because of the fine particles (the aerodynamic diameter is ≤2.5 μm). Particularly, for northeast of China, burning of fossil fuel, to provide warmth, in winters is essential, which increases the particulate matter content in air to a great extent. Vehicle emissions also cause great impact on the climate and ecosystems of the environment.⁴ In order to protect human health, several products for the purification of air have been produced. However, filters based on conventional fibres (e.g., glass fibres, melt-blown fibres, and spun-bonded fibres) will not satisfy the demand of people for air purification, particularly for removing the fine particles with diameters in the range of 300–500 nm suffused in the atmosphere.^5–8 Moreover, the bulkiness and thickness of the conventional filter medium would pose as a drawback for their applications under specific conditions, such as tiny filter devices. Therefore, thinner fibres may possess good potential for maximizing the filtration efficiency. To date, enormous efforts have been devoted toward fabricating high filtration efficiency medium, and it should be noted that the sub-microfiber filter media have demonstrated to be an excellent separation material for a remarkable specific surface area, high porosity and interconnected pores, particularly for the controllable diameter fibres.⁹ For the preparation of sub-micro fibres, the electrospinning technique would be an outstanding candidate to create fibrous membranes based on several fascinating features such as remarkable specific surface area, high open porosity, low base weight and interconnected porous structures.^10–12 To date, there have been several polymers which were employed as spinning materials for the removal of particles in air purification.^5,13 However, these membranes still have some defects, which make them unfit for application, such as poor mechanical property. Additionally, the membranes tend to be layered when they get to a certain thickness.¹⁴ Herein, we first present a novel and cost-effective strategy to fabricate the composite membranes through double jet-electrospinning, which show high efficiency filtration for air particles. As we all know that polyvinyl alcohol (PVA) and polyacrylonitrile (PAN) are applied widely in generating organic fibres by electrospinning, which have their own advantages in the applications.^15,16 As for the membrane, PAN fibre serves as bulk material for the filter owing to their chemical stability and outstanding weatherability in the applications.¹⁷ In order to improve the mechanical property of the membrane, enormous studies have been made to attempt the blending of polymer fibres during the electrospinning process and thermal treatments for the composite membrane.^18,19 For example, Ding et al. fabricated FPUPAN/PVB membrane cross linked by thermally treatment, which has greatly enhanced the mechanical properties.²⁰ Herein, we prepared the PAN/PVA membrane cross linked by a chemical bond, which could avoid layering and osteoporosis of surface comparing with uncross-linked membrane. During the process of fabrication, PVA was used as reinforced matrix to immobilize the PAN fibres through connection points after crosslinking; schematic of the experimental set-up is shown in Fig. 1.

Fig. 1 Schematic of the electrospinning set-up.

Then, the composite membranes were immersed in a certain concentration of solution mixed with acetic acid, glutaraldehyde (GA) and acetone.²¹ Glutaraldehyde captures the hydrophilic functional groups of PVA and forms bonding structures among the PVA fibers. (Fig. 2) Because of the interpenetrating PVA and PAN fibers, these crosslinking parts assemble into stable filtration medium, which possess the features of high porosity and packing density via optimizing the contents ratio of PVA and PAN fibers. The cross linked composite membrane retained good morphology and crossed each other via connect points of PVA fibres as compared with uncross liked membrane, as shown in ESI (Fig. S1†). This is because the hydroxyl groups of PVA fibre react with the molecules of GA and form the bond to link the two parts. The detailed plots of crosslinking are shown in the Fig. 2; there are two aldehyde groups at end of GA, which react with hydroxyl of PVA and produce bridge pattern after the acetal reaction. The fibers of PVA intertwine each other, and improve the strength and thermal stabilities to a certain extent.²²

Fig. 2 The mechanism of crosslinking and fibres fixed model.

In this paper, we present the fabrication of multilevel structure filtration medium via multi-jet electrospinning process. Moreover, we also resolve the delamination of single fibrous membrane and contact problem of film-upholder interface in a facile way. Compared with the technique of thermal treatment for the membranes, this approach of preparing the composite membranes has great potential in the future because the materials can retain morphology and chemical properties of the composite membranes.

Experimental

Materials

Polyacrylonitrile (PAN, M_w = 90 000) and polyvinyl alcohol 1799 (PVA, alcoholysis: 99.8–100%) were used in this experiment, which were purchased from Aladdin. Glutaraldehyde (GA, 50%) and acetic acid (99.7%) were purchased from Beijing Chemical Works. The acetone and N,N-dimethylformamide (DMF) were obtained from Sinopharm Chemical reagent Co. Ltd. Stainless steel mesh (SSM, 300 mesh) was purchased from the commercial market. Ultrapure water was used in all experiments. All chemicals were used without further purification.

Methods

At first, PAN was dissolved in DMF with mass percentage concentration of 14% and PVA was dissolved in deionized water with a concentration of 12%. Then, the two different solutions were transferred into two different syringes to generate polymer fibres by electrospinning equipment with boost devices. During the electrospinning process, flow rates of PAN and PVA were controlled at 1.0 mL h⁻¹ and 0.2 mL h⁻¹, respectively, by a peristaltic pump. The high voltage applied to the needle of the PAN solution syringe was 15 kV, and distance from the needle tip to rotating roller covered by SSM was 20 cm. Similarly, the high voltage applied on the PVA syringe was 10 kV and the distance was 20 cm. Then, the composite membranes were cross linked by the mixture of GA, acetic acid and acetone for 1 hour. Moreover, the SSM was immersed in absolute ethyl alcohol to wash the contaminants adhering on the surface for 2 h. Then, the SSM would be put into dilute PVA solutions with the concentration of 2% for 2 hours. After that, it was used as a collector for composite fibres and transferred into the gas mixture of GA and acetic acid.

Characterization and measurements

The membranes were examined by a scanning electron microscope (SEM, SHIMADZUX-550) after applying a gold coating. The Fourier transform infrared spectra of the membranes were obtained using a Paragon 1000 Spectrometer (Perkin Elmer) at a signal resolution of 1 cm⁻¹ within the range of 400–4000 cm⁻¹; the thickness of samples was measured by a micro-screwmeter (0–25 mm, Shanghai).We used a tensile tester to measure the mechanical property of the membranes with stretching velocity of 5 mm min⁻¹. The contact angle of the membrane interface with a drop of water was measured by a contact angle analyzer (Kino SL200B, USA). In order to avail the authenticity of the test results, the contact angles were measured thrice and the average of the values was recorded. Moreover, the filtration performance of the membrane was tested in real polluted air environment; the PM_2.5 diffused in air on that day was at PM_2.5 index > 200. The membrane was used as filters to measure the air particle index for one day. Ultimately, to evaluate the filtration performance of the membranes, an air particle counter (MET ONE 237B) was used to detect the number of particles through the samples and the removal efficiency was calculated by a comparison between the number of particles on the membrane and the PM particle number concentration without air filters. The pressure drop can be calculated from the height difference marked in the tube U. The detection process is shown in the following scheme (Fig. 3).

Fig. 3 The schematic of detection device for airborne particulate matter.

Results and discussion

As shown in Fig. 4, the morphology of the membrane was observed from the SEM images. By adjusting the concentrations of polymers and parameters of electrospinning, the average fibres' diameter could be precisely controlled. The diameter of the fibre is a critical factor for filtration efficiency in the application.²³ As for the PVA fibres, they exhibited smooth, uniform morphology and narrow diameter distribution from 0.55–0.65 μm. Compared with PAN fibers, the PVA fibers play a great role in enhanced tensile strength. The diameter of PAN fibres ranged from 0.25 to 0.35 μm, which was used as the primary material for removing particles and about 80% in the mass ratio compared with PVA. Fig. 4B is the SEM image of the membrane after crosslinking. As observed from the morphology of the cross-linked membrane, the packing density of the film increased evidently and the aperture decreased greatly for the formation of connection points between PVA fibres via crosslinking. The high magnification of SEM image for the cross linked membrane is presented in the Fig. S1,† which demonstrates the formation of bonding structures among adjacent fibres of PVA, highlighted by red dotted circles. Compared with uncross linked fibres, the PVA and PAN fibres' morphology remained almost unchanged. However, there were some changes in the structures between the fibres. For example, the ribbon-shaped bonding structures formed between the PVA fibres are shown in yellow dotted rectangle (Fig. S1†). On the basis of the real weather situation, the membrane was tested for filtration performance under a condition of hazardous PM level equivalent to PM_2.5 index > 200 for 24 hours. Fig. 4C is a low magnification SEM image of membrane, used for 24 hours. The particulate matter in the air was captured by the fibres. The high magnification SEM images are shown in Fig. 4D. From the morphology of intercepted particles, the general capture mechanism for airborne particles is that the particles would aggregate together and wrap around the fibres tightly, which indicates that the roughness of fibre might possess excellent filtration efficiency for the airborne particles than that of uniform fibres. Fig. S2† proved the suspicion and the image reveals that the particles captured were centralized on the parts of one fibre. As we all know that the principles and fundamentals of airborne filtration have been validated with respect to micron particles, such as interception effect, inertial deposition and gravity effect. However, the mechanisms associated with the application of filters for airborne nanoparticles are still uncertain, thereby, the experiment may provide a further research model in the future.^24,25

Fig. 4 The morphology of membranes (A) the PAN/PVA membrane; (B) the morphology of cross linked PAN/PVA membrane; (C) the low magnification morphology of composite membranes used for 24 hours' filtration in the air and (D) SEM image for the back side of membrane used for 24 h. The insets are the corresponding images of the membranes respectively.

In order to measure the changes in the membrane after crosslinking, Fourier transform infrared spectroscopy (FT-IR) was used to characterize the presence of specific chemical groups in the membrane. The red curve represents the FT-IR pattern of the membrane before crosslinking, and the black curve represents the FTIR spectra of the cross linked membrane, which shows some changes in the specific peaks compared with original membrane. As shown in the Fig. 5, the band observed from 3600–3200 cm⁻¹ represents the peak of stretching O–H group from the intermolecular and intramolecular hydrogen bonds, which has considerable reduction in the intensity of O–H because of the reaction of PVA and GA. The result indicates a possible formation of acetal bridges, as we can see, from the mechanism of crosslinking (Fig. 2). The region II refers to the stretching C–H from alkyl groups, and there is an inconspicuous peak for C–H of aldehyde group in the cross linked membrane, which is an evidenced of the crosslinking reaction. The range from 1150 cm⁻¹ to 1010 cm⁻¹ are related to the symmetric C–C stretching mode or stretching of the C=O of a portion of the chain according to the literature.^22,26

Fig. 5 The contrast FT-IR patterns of membranes (red curve for uncrosslinked membrane and the black for crosslinked membrane).

Subsequently, the wettability of the membranes surface was measured for the importance to the filtration materials.²⁷ In order to investigate the wetting behavior of the electrospun membranes and crosslinked membranes, the contact angle hysteresis values were measured. As shown in the Fig. 6, the wettability of the membranes surface should be hydrophobic except for PVA film. Fig. 6 shows the wetting property of the discussed membrane flat surfaces. As observed from the Fig. 6A, the contact angle for a water droplet on the PAN film surface was 122.43 ± 1.3°, indicating that the membrane surface was hydrophobic compared to the PVA film surface²⁸ (contact angle: 42.32 ± 2.2 in Fig. 6D). Fig. 6B and C reveal that the contact angles of composite films were 117.84 ± 3.2° and 110 ± 4.6°, respectively. The results are in good agreement with FT-IR because the hydrophilic groups (–OH) of PVA decreased after crosslinking by the mixture of acetic acid and GA. The hydrophobic property could keep the filtration materials from attaching to the liquid drop, which plays a great role in the applications.²⁹

Fig. 6 Contact angles of electrospun membranes with water drop (the inset labels correspond to the samples and the photos of water drops on flat parts of the films).

Ordinarily, the mechanical property of cross-linked membrane is another important factor to evaluate the property of materials in their applications, therefore, tensile testing was conducted on the as-prepared samples.³⁰ The results are shown in the Fig. 7 and Table 1. Compared with the tensile stress–strain curves of various membranes, we can clearly observe that cross linked membrane corresponds to the highest tensile strength (8.19 MPa) among the three samples. With regard to the phenomenon, the PVA blended in the membrane played a significant role in the enhanced mechanical properties. Based on our previous research, the polymer network of PVA could be tailored into three dimensional network structure in the molecular level, which would enhance the strength for the whole membrane to a great extent.¹⁸ Because of the interpenetrations of PAN and PVA fibers, the PVA immobilizes the PAN in some degree by forming the connection points. Furthermore, owning to the adding of PVA, the toughness of PAN membrane also improve prominently from 3.61 J g⁻¹ to 6.23 J g⁻¹, the detailed data are listed in the Table 1 (Fig. 8 and 9).

Fig. 7 The curves of tensile tests for different membranes.

Table 1 The tensile testing of electrospinning membranes

Samples code	Thickness (mm)	Density (g cm^−3)	Stress (MPa)	Strain break (%)	Area (M Pa)	Toughness (J g^−1)
①	0.038 ± 0.010	1.18 ± 0.07	5.01 ± 0.14	134.76 ± 7.2	4.26 ± 0.91	3.61 ± 0.27
②	0.047 ± 0.003	1.21 ± 0.01	6.93 ± 0.57	182.51 ± 11.3	7.55 ± 0.37	6.23 ± 0.52
③	0.043 ± 0.001	1.23 ± 0.04	8.19 ± 0.31	144.58 ± 4.3	6.56 ± 0.16	5.33 ± 0.14

---

Fig. 8 The comparison of filtration efficiency of three samples.

Fig. 9 The contrast of avulsion interfaces for membranes and SSM (A and B are cross-linked membrane and uncross-linked based on the SSM, respectively).

As the concentrations of airborne particles increased with industrialization, concerns were found to arise with respect to the negative impact on human health, because these particles could cause adverse health problems directly owing to the particles acting as carriers of toxic elements.^31,32 It is obvious that we should emphasize the importance of controlling the particle concentration out of our body. In order to investigate the filtration efficiency of the samples, the air particle counter (MET ONE 237B) was used for detecting the number of particles through the samples. Comparing with the filtration efficiency of two commercial masks, the electrospinning membrane (ESM) exhibited high efficiency for removing airborne particles. The detailed parameters about the test are listed in ESI (Table S1†). The performance efficiency for removing particles, shown in Fig. 6, denotes that the removal efficiency of ESM for particles (PM_≤1.0) had reached to 99.99%, which was higher than that of the other two commercial masks. Furthermore, the average pressure drop of ESM was just about 418 Pa with the air flow velocity of 0.42 m s⁻¹. For comparison, the filtration efficiency of DMM and 3MM were just about 0.8389% and 61.3765% respectively for the size of particles ≤0.3 μm, and the detailed results are listed in the Table 2. Hence, the composite membranes possess great potential as filter media.

Table 2 The detailed plots of membrane for filtration of airborne particles

Particle size (μm)	0.3	0.5	0.7	1	2	5	Thickness (mm)	P (Pa)
Commercial-1 (DMM)	0.8389	6.4943	14.7786	24.9603	43.1035	70.8861	0.336 ± 0.002	536 ± 7
Commercial-2 (3MM)	61.3765	88.7457	94.7796	95.8468	98.7862	99.9945	1.267 ± 0.005	878 ± 3
ESM	99.9961	99.9971	99.9985	99.999	99.9994	100	0.020 ± 0.001	418 ± 23

---

In general, the mechanical properties of filtration have some weakness for the demand of filtration applications under extreme conditions such as high air flow and working pressure. As for the previous researches, the enhanced mechanical properties were modified by introducing blend polymers and thermal treatment. However, to the best of our knowledge, few efforts have been devoted to fabricating 3D electrospinning membranes based on a stainless steel mesh for the hardship of close-knit between the interfaces of membrane and SSM. As shown in Fig. S3,† the modified SSM is the collector to receive the blending fibres of PVA and PAN with the thin layer of membrane. The light part marked by a red dot circle is the joint of stainless steel wires. Therefore, to realize the close integration of interfaces of membrane with SSM, the SSM was rinsed with absolute ethyl alcohol and then modified by dilute PVA solution. After that, the treated SSM was used as a collector for the fibres of PVA and PAN. Afterwards, the filtration medium was cross linked by the atmosphere of GA and acetic acid. As we can see from the images of interfaces of membrane with SSM compared with uncross linked and cross-linked, the cross linked composite membrane adhered on the surface of SSM tightly, which would stop the membrane from falling off from the SSM. This is because the PVA fibre reacts with the PVA adhering on the surface of SSM immersed in the dilute PVA solution, which leads to the cohesion of membrane with SSM.

Conclusions

To summarize, a novel filtration medium has been fabricated via electrospinning technique, which exhibited significant filtration efficiency compared with commercial products. The mechanical property and wettability of membrane have been improved physically by structures of blend polymer fibres and chemically by cross-linked networks between the intertwined PVA fibres. Additionally, we used a stainless steel mesh modified by diluted PVA solution as the substrate for supporting the membrane, which is beneficial for the applications for the filtration material in extreme conditions. PVA fibres have great influence on the qualities of membrane, especially on the mechanical property, which serve as a reinforcing agent after crosslinking by the liquid and atmosphere of GA and acetic acid for the membrane. Overall, the method of preparation for multi-layers of membranes provides a versatile strategy for further design and development of functional composite fibrous membranes for more mainstream applications in filtration technology, which could take advantages of all the features of polymer fibres to ensure the integrated properties of materials.

Acknowledgements

This work was financially supported by Changchun Science and Technology plan projects (14KG078), the National NatureScience Foundation (Grant 51403076).

References

1. V. Thavasi, G. Singh and S. Ramakrishna, Energy Environ. Sci., 2008, 1, 205.
2. J. O. Anderson, J. G. Thundiyil and A. Stolbach, J. Med. Toxicol., 2012, 8, 166.
3. S. Dai, X. Bi, L. Y. Chan, J. He, B. Wang, X. Wang and J. Fu, Atmospheric Chemistry and Physics Discussions, 2014, 14, 28885.
4. E. S. Lee, C. C. Fung and Y. Zhu, Environ. Sci. Technol., 2015, 49, 3358.
5. C. Liu, P. C. Hsu, H. W. Lee, M. Ye, G. Zheng, N. Liu and Y. Cui, Nat. Commun., 2015, 6, 6205.
6. H. Wan, N. Wang, J. Yang, Y. Si, K. Chen, B. Ding and J. Yu, J. Colloid Interface Sci., 2014, 417, 18.
7. N. Wang, Z. Zhu, J. Sheng, S. S. Al-Deyab, J. Yu and B. Ding, J. Colloid Interface Sci., 2014, 428, 41.
8. R. Uppal, G. Bhat, C. Eash and K. Akato, Fibers Polym., 2013, 14, 660.
9. R. Givehchi and Z. Tan, Aerosol Air Qual. Res., 2014, 14, 1.
10. P. Ke, X. Jiao, X. Ge, W. Xiao and B. Yu, RSC Adv., 2014, 4, 39704.
11. J. Jung, C. Lee, S. Yu and I. Kim, J. Mater. Chem. A, 2016, 4, 703.
12. J. Lin, F. Tian, Y. Shang, F. Wang, B. Ding and J. Yu, Nanoscale, 2012, 4, 5316.
13. D. Cho, A. Naydich, M. W. Frey and Y. L. Joo, Polymer, 2013, 54, 2364.
14. A. I. Sidorina and T. V. Druzhinina, Fibre Chem., 2016, 47, 362.
15. J. Matulevicius, L. Kliucininkas, D. Martuzevicius, E. Krugly, M. Tichonovas and J. Baltrusaitis, J. Nanomater., 2014, 2014, 14.
16. Y. Liu, M. Park, B. Ding, J. Kim, M. El-Newehy, S. S. Al-Deyab and H. Y. Kim, Fibers Polym., 2015, 16, 629.
17. Z. Wang xi, L. Jie and W. Gang, Carbon, 2003, 41, 2805.
18. Y. Liang, S. Cheng, J. Zhao, C. Zhang, S. Sun, N. Zhou, Y. Qiuand and X. Zhang, J. Power Sources, 2013, 240, 204.
19. J. Y. Lin, X. F. Wang, B. Ding, J. Y. Yu, G. Sun and M. R. Wang, Crit. Rev. Solid State Mater. Sci., 2012, 37, 94.
20. J. Sheng, M. Zhang, W. Luo, J. Yu and B. Ding, RSC Adv., 2016, 6, 29629.
21. Z. Cheng, F. Zhang, W. Liu, L. Cui and L. Kang, RSC Adv., 2015, 5, 54182.
22. H. S. Mansur, C. M. Sadahira, A. N. Souza and A. A. Mansur, Mater. Sci. Eng., C, 2008, 28, 539.
23. M. Cai, O. Painter and K. J. Vahala, Phys. Rev. Lett., 2000, 85, 74.
24. R. Givehchi and Z. Tan, Aerosol Air Qual. Res., 2014, 14, 1.
25. P. Li, C. Wang, Z. Li, Y. Zong, Y. Zhang, X. Yang, S. Li and F. Wei, RSC Adv., 2014, 4, 54115.
26. C. K. Yeom and K. H. Lee, J. Membr. Sci., 1996, 109, 257.
27. E. Terzaghi, E. Wild, G. Zacchello, B. E. Cerabolini, K. C. Jones and A. Di Guardo, Atmos. Environ., 2013, 74, 378.
28. Y. Liu, N. Wu, Q. Wei, Y. Cai and A. Wei, J. Appl. Polym. Sci., 2008, 110, 3172.
29. M. Tao, L. Xue, F. Liu and L. Jiang, Adv. Mater., 2014, 26, 2943.
30. W. J. Li, C. T. Laurencin, E. J. Caterson, R. S. Tuan and F. K. Ko, J. Biomed. Mater. Res., 2002, 60, 613.
31. G. Oberdörster, A. Maynard, K. Donaldson, V. Castranova, J. Fitzpatrick, K. Ausman and S. Olin, Part. Fibre Toxicol., 2005, 2, 8.
32. A. D. Maynard and E. D. Kuempel, J. Nanopart. Res., 2005, 7, 587.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra13971j

‡ Yingying Zhang and Zhiqiang Cheng contributed to the work equally.

---