Carbohydrate Polymers 198 (2018) 241–248

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

Antibacterial and environmentally friendly chitosan/polyvinyl alcohol T

blend membranes for air filtration
⁎ ⁎
Zhen Wanga,b, Feng Yana,c, , Hongchang Peia,b, Jianxin Lia,b,d, , Zhenyu Cuia,b, Benqiao Hea,b
a
State Key Laboratory of Separation Membranes and Membrane Processes/National Center for International Joint Research on Membrane Science and Technology, Tianjin
Polytechnic University, Tianjin 300387, PR China
b
School of Materials Science and Engineering, Tianjin Polytechnic University, Tianjin 300387, PR China
c
School of Environmental and Chemical Engineering, Tianjin Polytechnic University, Tianjin 300387, PR China
d
Collaborative Innovation Center for Eco-Textiles of Shandong Province, 308 Ningxia Road, Qingdao 266071, PR China

A R T I C LE I N FO A B S T R A C T

Keywords: An antibacterial and environmentally friendly chitosan (CS) /polyvinyl alcohol (PVA) blend membrane for air
Chitosan (CS) filtration was prepared via nonsolvent induced phase separation (NIPS) method. The chemical structure, thermal
Polyvinyl alcohol (PVA) behavior, morphology, mechanical property and surface charge of the resultant CS/PVA membranes were
Nonsolvent induced phase separation (NIPS) characterized. Results showed that CS and PVA were miscible due to the intermolecular hydrogen bond between
Air filtration
them. The blend membrane obtained from over 20 wt.% CS concentration exhibited a gradient interconnected
Antibacterial
porous structure without skin layer. The air filtration efficiency and pressure drop obtained from CS/PVA
membrane with 30 wt.% CS concentration and the thickness of 37 μm under a face velocity of 5.3 cm s−1 were
95.59% and 633.5 Pa, respectively. The performance of air filtration obtained is mainly attributed to the direct
interception of membrane surface. Further, the antibacterial rate of the blend membrane was up to 94.8% for E.
coli and 91.3% for S. aureus.

1. Introduction distinctive 3D structures (Jing et al., 2016; Wang & Pan, 2015; Zhang,
Tang, et al., 2016). In particular, to further improve the filtration per-
Air pollution, especially particulate matter (PM), has been con- formance or tackle the trade-off between the filtration efficiency and
sidered as one of the most serious health threats to the public health, pressure drop, the electrospun nanofibers incorporated with carbon
atmosphere visibility, production efficiency, global climate and eco- nanotubes (CNT), metal-organic framework (MOF) or polytetra-
systems (Apte, Marshall, Cohen, & Brauer, 2015; Huang et al., 2014; fluoroethylene nanoparticles (PTFE NPs) have been fabricated and
Zhang, He, & Huo, 2012). Depending on the particulate size, PM can be show enhanced filtration performance due to the “slip effect” and
classified into PM10 (aerodynamic diameter ≤ 10 μm) and PM2.5 electrostatic adsorption (Li, Wang, Zhang, & Wei, 2014; Wang, Zhao,
(aerodynamic diameter ≤ 2.5 μm). Herein, PM2.5 has a long lifetime Yin, Yu, & Ding, 2016; Zhang, Yuan, et al., 2016). However, unclear
for spreading over thousands of kilometers. It can adsorb toxic pollu- biological toxicity of nanoparticles and the dissipation of charges are
tants and bacteria in the air due to larger surface areas, and penetrate unavoidable issues (Xiong, Yang, Zhu, Chen, & Dong, 2017). More
into the respiratory system and even the blood (Shen et al., 2011; importantly, the electrospun nanofibers usually suffer from low pro-
Turner et al., 2011). Therefore, the development of versatile air filters duction rate and instability owning to their weak mechanical property
to prevent PM2.5 from invading environment and human body is of (Singh, Ravi, Sun, & Tan, 2017). Consequently, other innovative ap-
great importance (Barhate & Ramakrishna, 2007). proaches have been proposed to prepare the filter media.
Nowadays, the nanofibers based filter has been verified to be an For instance, Liu et al. (2015) prepared poly(vinyl alcohol-co-ethy-
effective strategy for the removal of sub-micron aerosol (Leung & Hung, lene) /TiO2 (PVA-co-PE/TiO2) hybrid nanofibrous filters via melt phase
2012). Consequently, many technologies, especially electrospinning separation and suspension coating technique. The prepared nanofibers
have been developed to fabricate nanofibers in recent years (Zhu et al., with diameters in a range of 50 nm–300 nm displayed a filtration effi-
2017). This is because the electrospun nanofibers based filter can be ciency of 95.3% owning to the decrease of pore size and the electro-
prepared by various materials and exhibit controllable morphology and static absorption of TiO2 nanoparticles. Further, a different approach

⁎
Corresponding authors at: State Key Laboratory of Separation Membranes and Membrane Processes/National Center for International Joint Research on Membrane Science and
Technology, Tianjin Polytechnic University, Tianjin 300387, PR China.
E-mail addresses: yanfeng@tjpu.edu.cn (F. Yan), jxli@tjpu.edu.cn (J. Li).

https://doi.org/10.1016/j.carbpol.2018.06.090
Received 8 February 2018; Received in revised form 26 May 2018; Accepted 20 June 2018
Available online 21 June 2018
0144-8617/ © 2018 Elsevier Ltd. All rights reserved.
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

was described by using organic molecules self-assembly to design na- dissolving CS powder in 0.2 mol L−1 adipic acid aqueous solution at
nofibers. The self-assembled nanofibers based air filter reached a fil- 30 °C under magnetic stirring for 6 h. And similarly, PVA solution was
tration efficiency of 89.69% for PM2.5 and 93.08% for PM10 particles prepared after dissolving PVA pellet in 0.2 mol L−1 adipic acid aqueous
(Singh et al., 2017). Very recently, an asymmetric polyvinylidene solution at 90 °C under mechanical stirring for 3 h and cooled to room
fluoride-polyethylene glycol (PVDF-PEG) hollow fibers were developed temperature. Then, the casting solution was prepared by mixing the two
by a dry-jet wet spinning technique and firstly applied to air filtration, solutions in various mass ratio and stirred until to a homogeneous so-
which exhibited excellent filtration efficiency of 99.999% against poly- lution. After degassed, the casting solution was cast on a clean glass
dispersed NaCl particles in both dead-end filtration and cross-flow fil- plate using an automated membrane applicator (Elcometer 4340,
tration (Wang, Yong, Yu, & Chung, 2017). However, the pressure drop England) with a desired thickness, and then immediately immersed into
of the hollow fibers was too high and up to 5000 Pa. the coagulation bath of saturated Na2SO4 aqueous solution at 20 °C for
In addition to the preparation methods mentioned above, recent 30 min. Finally, the prepared membrane was neutralized with 0.1 mol
studies have found that nanofibers made from higher polar polymer L−1 NaOH aqueous solution and rinsed with deionized water to remove
materials can be used as high-efficient filters (Xu et al., 2018). For in- the residual alkali and Na2SO4. Before characterization, the CS/PVA
stance, Zhang, Zhang et al. (2017) reported that the direct e-spinning blend membranes were fully lyophilized. The composition and viscosity
chitosan (CS) nanofibers exhibited more timesaving than other poly- of different casting solutions were listed in Table 1. The polymer con-
mers in removing PM2.5, which was attributed to the strong polarity, centration in all casting solutions was 7 wt.%. The corresponding
positively charged NH3+ group and smaller diameter. Besides, CS is the membranes with different CS concentrations (10–50 wt.%) were named
second most abundant natural cationic polysaccharide with good as M10, M20, M30, M40, and M50, respectively.
membrane forming, biocompatibility and antibacterial ability. How-
ever, the mechanical brittleness of CS materials limits the practical
2.3. Membrane characterization
applications. It is well-known that polymer blend with flexible-chain
polymers is the most common method to improve the brittleness
The chemical structure of CS/PVA blend membrane surface was
(Sionkowska, 2011). Polyvinyl alcohol (PVA) as a common synthetic
characterized using ATR-FTIR spectrophotometer (Nicollet NEXUS-670,
polymer possesses good mechanical property and membrane forming,
USA) at 4000–600 cm−1 with a resolution of 4 cm−1. The thermal be-
which is able to interact with CS through aggregation, ether-bridge
havior of CS/PVA blend membrane was evaluated by a differential
connection and intermolecular hydrogen bond. Moreover, both CS and
scanning calorimeter (DSC) (DSC 204F1, Germany) operated from 50 °C
PVA have biodegradable ability so as to avoid the secondary environ-
to 250 °C at a heating rate of 10 °C min−1 under nitrogen atmosphere
ment pollution. Up to now, CS/PVA blend membranes have been ex-
protection. The crystallinity of PVA component (χ) was calculated as
tensively applied to reverse osmosis desalination, heavy metal adsorp-
Eq. (1) (Bonilla, Fortunati, Atarés, Chiralt, & Kenny, 2014).
tion, and wound dressing (Morgado, Miguel, Correia, & Aguiar, 2017;
Salehi & Madaeni, 2014; Shafiq et al., 2017). χ (%) = ΔH/(ΔH0 × w) × 100% (1)
The aim of this study is to explore a new avenue to tailor an anti-
where ΔH is the enthalpy of melting, ΔH0 is the enthalpy of melting for
bacterial and environmentally friendly CS/PVA blend membrane for air
100 wt.% crystalline PVA, which is 156 J g−1, and w is PVA con-
filtration. A series of CS/PVA blend membranes with various mass ra-
centration in the membrane samples. The surface and cross-sectional
tios and thicknesses were prepared via NIPS. Fourier transform infrared
morphologies of CS/PVA blend membrane was observed by scanning
spectroscopy (FTIR) and differential scanning calorimetry (DSC) were
electron microscopy (SEM) (TM3030, Japan). The membrane was at-
adopted in order to explore the miscibility between CS and PVA.
tached on an aluminum stage, and the cross section of blend membrane
Subsequently, the morphology, structure, mechanical property and
was fractured after frozen in liquid nitrogen. All samples were dried in a
surface charge were characterized. The filtration performance for NaCl
freeze dryer (FD-1A-50, Boyikang, China), and sputter-coated with gold
aerosol particles removal was explored. Additionally, the antibacterial
before observation. The pore size distribution and the average pore size
ability against Escherichia coli (E. coli) and Staphylococcus aureus (S.
of CS/PVA blend membrane were obtained by a capillary flow po-
aureus) was also evaluated.
rometer (3H-2000PB, Beishide, China). The samples were thoroughly
infiltrated with the wetting liquid (Porewick) prior to measurement.
2. Experimental
Moreover, Image J analysis software was used to examine the surface
porosity by analyzing the surface SEM images. Five SEM images of each
2.1. Materials
sample were selected randomly to calculate the surface porosity. The
membrane porosity was determined according to dry-wet weight
CS powder (Mw = 540 kDa, degree of deacetylation = 87.3%) was
method (Li, Cui, Zhang, He, & Li, 2016). The mechanical property of
purchased from National Pharmaceutical Co., Ltd. (China) and PVA
CS/PVA blend membrane was evaluated using an electronic stretching
pellet (Mw = 77 kDa, degree of hydrolysis = 88%) was purchased from
machine (INSTRON-5969, USA) with a crosshead speed of 20 mm
Tianjin Guangfu Technology Development Co., Ltd. (China). Adipic
min−1 and an applied force of 0.25 cN. To minimize the experimental
acid and other reagents were obtained from Tianjin Kermel Chemical
error, each sample with a valid size of 5 mm × 25 mm was tested five
Reagent Co. Ltd. (China). Polyethylene terephthalate (PET) nonwoven
times to obtain the tensile stress and elongation at break. The zeta
fabrics with negligible filtration efficiency (3.0%) and pressure drop
potential of CS/PVA blend membrane surface was analyzed by an
(3.3 Pa) as supporting in the filtration test was kindly provided by
electrokinetic analyzer (Anton Paar, Austria). The samples were tested
Seattle New Material Technology Co., Ltd. (China). E. coli (ATCC8739)
in 1 mmol L−1 KCl aqueous solution at 25 °C under an operating pres-
and S. aureus (ATCC 6538) were obtained from Tianjin Medical
sure of 300 mbar.
University (China). Phosphate buffered saline solution (PBS) (7.16 g
L−1 Na2HPO4·12H2O, 1.36 g L−1 KH2PO4) was prepared in house. Beef
extract, peptone, and agar powder were supplied by Tianjin Sanjiang 2.4. Filtration performance
Chemical Technology Co., Ltd. (China).
The filtration performance of CS/PVA blend membrane was mea-
2.2. Membrane preparation sured by an automatic filter tester (Model 8130, TSI Group, USA) (Gao
et al., 2017). The charge neutralized mono-dispersed NaCl aerosols
The CS/PVA membranes with various mass ratios were prepared via with mass median diameter of 260 nm and count median diameter of
NIPS as demonstrated in Fig. 1(a). Firstly, CS solution was prepared by 75 nm were generated by an aerosol particle generator. Under the help

242
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

Fig. 1. (a) Schematic of preparation process of CS/PVA blend membranes. (b) Proposed interactions among CS, PVA and adipic acid in the casting solution.

Table 1 liquid medium followed by 100-fold diluted with PBS. After that, the
Composition and viscosity of the casting solutions. CS/PVA blend membrane (0.75 g, 5 mm × 5 mm) sterilized by ultra-
Membrane Casting solution composition Viscosity (mPa s) violet radiation were immersed into a flask with 5 mL diluted bacteria
suspension and 70 mL PBS, and the mixture was incubated in an in-
CS (g) PVA (g) Adipic acid solutiona (g) cubator shaker at 25 °C and 150 rpm for 18 h. After the incubation
process, serially 10-fold diluted bacteria suspension (100 μL) was
M10 0.7 6.3 93 588
M20 1.4 5.6 93 685
spread on the prepared agar plates in the Petri dishes and cultured in a
M30 2.1 4.9 93 860 biochemical incubator at 37 °C for 24 h. Finally, the number of viable
M40 2.8 4.2 93 1105 bacteria colonies was counted from the agar plate, and the antibacterial
M50 3.5 3.5 93 1650 rate (Y) was calculated according to Eq. (4).
a
The concentration of adipic acid aqueous solution is 0.2 mol L−1. Y (%) = (Cb − Cs)/Cb × 100% (4)

of an air compressor and a flowmeter, the generated aerosols passed where Cb and Cs are the bacterial concentration of the diluted bacteria
through the membrane with an effective area of 100 cm2 at a constant suspension after the incubation process with the black sample and the
face velocity. The filtration efficiency (η) of the membrane was calcu- CS/PVA membranes, respectively.
lated by Eq. (2). Furthermore, the quality factor (QF) as a trade-off
parameter between the filtration efficiency and the pressure drop was
calculated by Eq. (3). 3. Results and discussion

η (%) = 1 − Ce/Ci (2) 3.1. Miscibility of CS and PVA

−1
QF (KPa ) = −ln(1 − η)/ΔP (3)
The proposed interactions among CS, PVA and adipic acid in the
where Ce and Ci represented the aerosol particle number concentrations casting solution were illustrated in Fig. 1(b). The adipic acid with two
at the inlet and the outlet of the filter holder, which were detected by carboxyl groups provided protons for CS dissolution as well as inter-
laser photometer. ΔP was the pressure drop, which was measured by a acted with the positively charged amino groups or acetyl group via the
pressure transmitter. The filtration experiment was conducted at room ionic interaction (the red dotted circle). Herein, the adipic acid acted as
temperature with a face velocity of 5.3 cm s−1. The filtration efficiency, an ionic cross-linking reagent (Hamdine, Heuzey, & Bégin, 2005). Be-
pressure drop, and QF of the membranes were calculated after ex- sides, these positively charged groups together with hydroxyl groups on
cluding the contribution of PET nonwoven fabrics. The numerical re- CS would interact with PVA through ample hydrogen bonds (the blue
sults were reported as the mean value and standard deviation of five dotted line) after blending.
times measurements. FTIR analysis of PVA pellet, CS powder, and CS/PVA blend mem-
brane was carried out to confirm the proposed interaction mechanism
and the miscibility of CS and PVA as shown in Fig. 2(a). The broad
2.5. Antibacterial experiment absorption band at 3262 cm−1 in the spectrum of PVA pellet corre-
sponded to the stretching υ (OeH) mode of the free hydroxyl groups,
The antibacterial ability of CS/PVA blend membrane was evaluated which involved in the intra- and inter-molecular hydrogen bonds. A
via colony counting method using E. coli and S. aureus as the model similar broad contour was observed in the spectra of chitosan powder
bacteria (Meng et al., 2015). In brief, the bacteria were inoculated in (3429 cm−1) and CS/PVA blend membrane (3000–3400 cm−1), which
250 mL Lysogeny Broth (LB) liquid medium (15 g L−1 peptone, 5 g L−1 was ascribed to the stretching υ (NeH) and υ (OeH) vibrations. It
beef extract and 5 g L−1 sodium chloride, pH = 7.5) and cultivated in meant that hydrogen bonds between −OH groups or between −OH
an incubator shaker at 37 °C and 120 rpm for 24 h. Subsequently, 1 mL and −NH2 groups occurred in CS/PVA blend polymer. Moreover, the
of the bacterial suspension was pipetted into a flask containing 100 mL peak position of bands corresponding to the amide group of CS/PVA
of LB liquid medium and incubated for 18 h to enhance the activity of blend membrane (1630 cm−1 and 1530 cm−1) was clearly lower than
the bacteria. The active suspension was then 100-fold diluted with LB that of CS powder (1650 cm−1 and 1597 cm−1). This proved that the

243
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

(Table 2) with the increase of CS concentration from 10 wt.% (M10) to

30 wt.% (M30) and then decreased to 355 nm and 7.0% with the further
increasing of CS concentration to 50 wt.% (M50). M30 with 30 wt.% CS
concentration had the largest average pore size and surface porosity.
Generally, the pure PVA membrane obtained by NIPS displays a dense
skin layer (Chuang, Young, Yao, & Chiu, 1999). However, the in-
troduction of CS in PVA solution would disrupt the intra- and inter-
molecular interactions among PVA chains and so far as to form new
intermolecular interactions between CS and PVA (Zhang, Zou, et al.,
2017). The new interactions would retrain the aggregation of PVA rich
phase during the phase separation, leading to the obtained blend
membrane with large pore size and high surface porosity. Noteworthily,
the viscosity of casting solution rapidly increased from 588 mPa s to
1650 mPa s with an increase of CS concentration from 10 wt.% to 50 wt.
% in the casting solution (Table 1). This is ascribed to the fact that the
molecular weight of CS (Mw = 540 kDa) is much larger than that of
PVA (Mw = 77 kDa) and both of them are miscible. Consequently, the
casting solution with a high viscosity and more than 30 wt.% CS con-
centration would markedly restrict the movement of polymer chains
and slow down the mutual diffusion between solvent and non-solvent
during phase inversion, finally leading to the obtained blend membrane
with a small pore size and low surface porosity (M40-M50) (Holda,
Aernouts, Saeys, & Vankelecom, 2013; Hung, Wang, Lai, & Chou,
2016).
Similarly, as shown in Fig. 4, the pore size distribution curves of CS/
PVA blend membranes shifted to the positive direction with a large pore
size as the increase of CS concentration from 10 wt.% (M10) to 30 wt.%
(M30) and then back in the opposite direction with a small pore size. In
addition, it could be also found from Table 2 that the membrane por-
osity of CS/PVA blend membrane increased from 64.9% to 74.4% with
the increase of CS concentration from 10 wt.% (M10) to 50 wt.% (M50).
This indicated that the introduction of CS in PVA solution brought
about the evolution of membrane structure from a relatively compact
Fig. 2. FTIR spectra (a) and DSC curves (b) of PVA pellet, CS powder and CS/
structure to a loose structure (Chuang et al., 1999). In sum, the final
PVA blend membrane. morphologies of CS/PVA blend membranes were strongly affected by
the concentration of CS in casting solution.

bonds of the amide group were weaker, because the strong inter-
3.3. Mechanical properties
molecular interactions between CS and PVA occurred after blending.
This proves that the CS and PVA are miscible (Dai & Fan, 2011;
The mechanical properties of filtration media are one of the major
Socrates, 2001; Zając, Hanuza, Wandas, & Dymińska, 2015).
concerns in practical applications. The tensile stress and elongation at
To further confirm the above observations, DSC analysis of PVA
break of CS/PVA blend membrane were illustrated in Table 2 and Fig.
pellet, CS powder, and CS/PVA blend membrane were carried out as
S1. As a comparison, the tensile strength of M10 was up to 6.06 MPa
shown in Fig. 2(b). It was found that the melting endothermic peak
due to the intermolecular interactions between PVA and CS, and the
decreased from 232.0 °C to 209.9 °C with the increase of CS con-
ionic cross-linking between polymers and adipic acid (Chen et al.,
centration from 0 to 50 wt.%. The corresponding melting enthalpies
2008). Interestingly, the largest tensile strength of 6.36 MPa was ob-
(ΔH) were 68.6, 40.1, 27.4, 19.7, 15.5, and 11.6 J g−1 respectively.
tained from M20. The reason was that M20 contained more CS mole-
Hereon, the crystallinity of PVA component (χ) calculated by Eq. (1)
cules, which had a large molecular weight and hard backbones com-
decreased from 44.0% to 14.9%. The reason was that the incorporation
pared with M10 (Zhuang, Li, Fan, Lin, & Hu, 2012). However, up
of CS in PVA solution increased the steric hindrance and changed the
further increasing the CS concentration from 30 wt.% (M30) to 50 wt.%
ordered structure of PVA chains, leading to rearrangement in crystalline
(M50), the tensile strength gradually decreased from 2.82 MPa to
zone as a result of the interaction of intermolecular hydrogen bond
1.00 MPa. This was because of the membrane structure evolution from
between CS and PVA (Çay, Miraftab, & Perrin, 2014; Hang, Tae, & Park,
a relatively compact structure to a loose structure. On the other hand, it
2010). In sum, the above observations are in good agreement with FTIR
also can be found from Table 2 that the elongation at break decreased
analysis and the proposed interaction mechanism.
from 31.0% to 4.20% with the increase of CS concentration from 10 wt.
% (M10) to 50 wt.% (M50) (Bonilla et al., 2014).
3.2. Membrane morphologies
3.4. Filtration performances
The SEM photographs of CS/PVA blend membrane were presented
in Fig. 3. As shown in Fig. 3(a), M10 exhibited a typical asymmetric An excellent filter usually exhibits the hierarchical or gradient
structure with an obvious skin layer and honeycomb porous structure. structure for trapping particles with different diameters during air fil-
Interestingly, the skin layer disappeared at M20 and the honeycomb tration (Zhu et al., 2017). The filtration performance of CS/PVA blend
structure gradually changed into the sponge-like structure with the membrane with a gradient interconnected structure was conducted at a
increase of CS concentration from 20 wt.% (M20) to 50 wt.% (M50). face velocity of 5.3 cm s−1 as shown in Fig. 5.
Specifically, the average pore size and surface porosity of CS/PVA blend It can be seen in Fig. 5 that M10 displayed a high filtration effi-
membrane increased from 59.6 nm and 3.0% to 495 nm and 21.5% ciency of 99.82% for NaCl aerosols removal due to the small pore size of

244
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

Fig. 3. SEM images of CS/PVA blend membrane: (a) M10, (b) M20, (c) M30, (d) M40, (e) M50.

59.6 nm and low surface porosity of 3.00%. However, a large pressure filtration efficiency and pressure drop distinctly decreased from 98.41%
drop of 1500 Pa had reached the threshold value of the filter tester. A and 928.5 Pa to 18.47% and 178.6 Pa with the increase of CS con-
similar phenomenon can be observed in M20. Interestingly, the centration from 30 wt.% (M30) to 40 wt.% (M40). This was because the

245
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

Table 2
The structure and mechanical properties of CS/PVA blend membranes with
various mass ratio.
No. Average Surface Membrane Tensile Elongation at
pore size porosity (%) porosity (%) strength break (%)
(nm) (MPa)

M10 59.6 3.00 ± 0.1 64.9 ± 1.6 6.06 ± 0.1 31.0 ± 2.5
M20 248 14.5 ± 0.3 66.9 ± 2.4 6.36 ± 0.3 18.3 ± 2.1
M30 495 21.5 ± 0.7 71.2 ± 2.2 2.82 ± 0.3 14.1 ± 1.9
M40 401 11.4 ± 0.3 72.9 ± 1.4 1.20 ± 0.1 7.13 ± 0.9
M50 355 7.00 ± 0.2 74.4 ± 2.2 1.00 ± 0.1 4.20 ± 0.1

Fig. 4. Pore size distribution of CS/PVA blend membrane. Fig. 6. Filtration efficiency and pressure drop (a), and quality factor (b) of M30-
1, M30-2, M30-3, and M30-4.

decrease of filtration efficiency and pressure drop was related to the

thinning membrane with short filtration path (or pore channel) for air
flow (Leung, Hung, & Yuen, 2010). This resulted in the weakening of
inertia impact and diffusion effect for capturing aerosol particles.
However, when the membrane thickness further decreased to 16 μm,
M30-4 cannot remain intact during filtration test, which led to a failure
of surface direct interception and then a sharp decrease to 70.4% in
filtration efficiency. The corresponding pressure drop obtained was
355.5 Pa. There are several mechanisms, including direct interception,
inertia impact, Brownian diffusion, and electrostatic adsorption, to re-
move aerosol particles from the air stream (Li et al., 2014). In this
Fig. 5. Filtration efficiency and pressure drop of CS/PVA blend membrane.
study, most of NaCl aerosol particles were captured on the intact
membrane surface by direct interception. Meanwhile, a few of NaCl
aerosol particles were captured in the membrane pore channel through
mechanical property of M40 (1.20 MPa) was so weak that it was hard to inertia impact and diffusion effect.
carry out the filtration test. Similarly, M50 showed a lower filtration In order to further evaluate the filtration performance of CS/PVA
efficiency of 10.85% and pressured drop of 154.6 Pa. Finally, M30 was blend membrane, QF (Eq. (3)) was introduced as shown in Fig. 6(b). QF
selected as a preferable candidate in the following control of membrane values of M30-1, M30-2, M30-3, and M30-4 were 4.23, 4.68, 4.93, and
structure. 4.58 kPa−1, respectively. Obviously, M30-3 exhibited the highest QF
Obviously, the filtration performance is relative to the thickness of value. This implied that the thickness of the membrane is closely related
filtration media. In order to investigate the relationship between fil- to the filtration performance.
tration performance and membrane thickness, CS/PVA blend mem-
branes with thickness of 101, 68, 37, and 16 μm were prepared from the
3.5. Antibacterial properties
casting solution of M30 and designated as M30-1, M30-2, M30-3, and
M30-4, respectively. Herein, CS/PVA blend membranes (M30-1, M30-2,
The antibacterial property of M30 was tested following the colony
M30-3, and M30-4) with different thickness exhibited the similar
counting method. Digital photographs of bacteria colony on the agar
morphology including pore size distribution, surface porosity and
plate related to the blank, PVA membrane, CS membrane and M30 were
average pore size as shown in Fig. S2-3 and Table S1. In other word,
shown in Fig. 7. The bacterial suspension concentration and anti-
M30-1, M30-2, M30-3, and M30-4 have the similar surface direct in-
bacterial rate by counting the bacteria colony were summarized in
terception for the removal of aerosol particles. The filtration perfor-
Fig. 8. It is obvious that the number of E. coli and S. aureus colonies
mance of CS/PVA blend membranes with different thickness was de-
related to the blank and PVA membrane was more than that related to
monstrated in Fig. 6(a). When the blend membrane decreased in the
M30. The suspension concentrations of E. coli and S. aureus were
thickness from 101 μm (M30-1) to 37 μm (M30-3), the filtration effi-
13.5 × 106 CFU mL−1 and 2.53 × 103 CFU mL−1, respectively. Inter-
ciency slowly decreased from 98.41% to 95.59%. Meanwhile, the cor-
estingly, no colony was detected on the agar plate for bacteria cultured
responding pressure drop declined from 928.5 Pa to 633.5 Pa. The
with CS membrane, which indicated that the antibacterial rate was up

246
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

Fig. 7. Photographs of (a) E. coli and (b) S. aureus colonies on the agar plate related to the blank, PVA membrane, CS membrane and M30 (PVA membrane and CS
membrane were prepared via NIPS as the control sample).

to 100% for both E. coli and S. aureus. The antibacterial rate obtained concentration in the casting solution. The morphology of the resultant
from PVA membrane was only 2.2% for E. coli and 1.6% for S. aureus, membrane evolved from honeycomb structure to sponge-like structure
which implied that there was almost no any antibacterial activity. with the increase of CS concentration. M30 with 30 wt.% CS con-
Consequently, it confirmed that the antibacterial activity of M30 with centration exhibited a largest pore size of 467 nm and surface porosity
94.8% for E. coli and 91.3% for S. aureus came from CS component. of 21.5%, a moderate tensile strength of 2.82 MPa, and a positive
Further, Verlee, Mincke, and Stevens (2017) described that the charge of 18.9 mV. The thickness of the membrane is closely related to
antibacterial ability of CS based material was related to the surface the filtration performance and the direct interception on membrane
charge. In this study, the surface charge of CS/PVA blend membrane surface was the dominant mechanism for removing NaCl aerosol par-
was decided by CS concentration as shown in Fig. S3. M30 with 30 wt. ticles. M30-3 with the thickness of 37 μm exhibited the highest quality
% CS concentration exhibited positive charge of 18.9 mV as a result of factor for air filtration and the filtration efficiency of 95.59% and
the high molecular weight chitosan (Mw = 540 kDa) and degree of pressure drop of 633.5 Pa were achieved. Additionally, the resultant
deacetylation (87.3%). The electrostatic interaction between M30 and membrane displayed high antibacterial ability against E. coli and S.
bacteria disrupted the permeability of the cytoplasmic membrane, re- aureus. In brief, our research opens a new avenue to tailor an anti-
sulting in the leakage of intracellular constituents and the death of bacterial and environmentally friendly blend membrane for air filtra-
bacteria (Raafat, von Bargen, Haas, & Sahl, 2008). Besides, M30 could tion.
chelate the trace metal and then inhibit bacteria reproduction (Wang,
Yang, & Zhang, 2010). In addition, it also can be found that the anti-
bacterial rate of M30 for E. coli was higher than that for S. aureus. The Acknowledgments
reason was that E. coli cell wall was thinner and consequently more
susceptible than S. aureus (Goy, Morais, & Assis, 2016). This work was financially supported by Innovative Team in uni-
versity of Ministry of Education of China (Grant No. IRT-17R80), the
National Natural Science Foundation of China (Grant No. 21776218)
4. Conclusions
and the Cultivating Program for Innovative Research Team of Tianjin
High College, China (Grant No. TD13-5044). We also would like to
CS/PVA blend membrane was successfully prepared via NIPS. The
thank Dr Nana Li, School of Textile; Mr. Tianzhi Jia, Ms. Qianqian
CS/PVA blend system was miscible owning to the interaction of the
Zhang and Mr. Saood Qaseem, School of Materials Science and
intermolecular hydrogen bond between them. The structure and prop-
Engineering, Tianjin Polytechnic University, China for their kind help.
erties of the resultant membrane were closely related to the CS

Fig. 8. The bacterial suspension concentration and antibacterial rate by counting the bacteria colony.

247
Z. Wang et al. Carbohydrate Polymers 198 (2018) 241–248

Appendix A. Supplementary data Raafat, D., von Bargen, K., Haas, A., & Sahl, H. G. (2008). Insights into the mode of action
of chitosan as an antibacterial compound. Applied and Environmental Microbiology,
74(12), 3764–3773.
Supplementary material related to this article can be found, in the Salehi, E., & Madaeni, S. S. (2014). Influence of poly(ethylene glycol) as pore-generator
online version, at doi:https://doi.org/10.1016/j.carbpol.2018.06.090. on morphology and performance of chitosan/poly(vinyl alcohol) membrane ad-
sorbents. Applied Surface Science, 288, 537–541.
Shafiq, M., Sabir, A., Islam, A., Khan, S. M., Hussain, S. N., Butt, Z., et al. (2017).
References Development and performance characteristics of silane crosslinked poly(vinyl al-
cohol)/chitosan membranes for reverse osmosis. Journal of Industrial and Engineering
Apte, J. S., Marshall, J. D., Cohen, A. J., & Brauer, M. (2015). Addressing global mortality Chemistry, 48, 99–107.
from ambient PM2.5. Environmental Science & Technology, 49(13), 8057–8066. Shen, G. F., Wang, W., Yang, Y. F., Ding, J. N., Xue, M. A., Min, Y. J., et al. (2011).
Barhate, R. S., & Ramakrishna, S. (2007). Nanofibrous filtering media: Filtration problems Emissions of PAHs from indoor crop residue burning in a typical rural stove: Emission
and solutions from tiny materials. Journal of Membrane Science, 296(1–2), 1–8. factors, size distributions, and gas-particle partitioning. Environmental Science &
Bonilla, J., Fortunati, E., Atarés, L., Chiralt, A., & Kenny, J. M. (2014). Physical, structural Technology, 45(4), 1206–1212.
and antimicrobial properties of poly vinyl alcohol–chitosan biodegradable films. Food Singh, V. K., Ravi, S. K., Sun, W., & Tan, S. C. (2017). Transparent nanofibrous mesh self-
Hydrocolloids, 35(1), 463–470. assembled from molecular LEGOs for high efficiency air filtration with new func-
Çay, A., Miraftab, M., & Perrin, A. K. E. (2014). Characterization and swelling perfor- tionalities. Small, 13(6), 1601924.
mance of physically stabilized electrospun poly(vinyl alcohol)/chitosan nanofibres. Sionkowska, A. (2011). Current research on the blends of natural and synthetic polymers
European Polymer Journal, 61, 253–262. as new biomaterials: Review. Progress in Polymer Science, 36(9), 1254–1276.
Chen, P.-H., Kuo, T.-Y., Liu, F.-H., Hwang, Y.-H., Ho, M.-H., Wang, D.-M., et al. (2008). Socrates, G. (2001). Infrared and Raman characteristic group frequencies (3rd ed.).
Use of dicarboxylic acids to improve and diversify the material properties of porous Chichester, New York, Weinheim, Toronto, Brisbane, Singapore: John Wiley & Sons
chitosan membranes. Journal of Agricultural and Food Chemistry, 56(19), 9015–9021. (Chapter 2).
Chuang, W.-Y., Young, T.-H., Yao, C.-H., & Chiu, W.-Y. (1999). Properties of the poly Turner, M. C., Krewski, D., Pope, C. A., Chen, Y., Gapstur, S. M., & Thun, M. J. (2011).
(vinyl alcohol)/chitosan blend and its effect on the culture of fibroblast in vitro. Long-term ambient fine particulate matter air pollution and lung cancer in a large
Biomaterials, 20(16), 1479–1487. cohort of never-smokers. American Journal of Respiratory and Critical Care Medicine,
Dai, D., & Fan, M. (2011). Investigation of the dislocation of natural fibres by Fourier- 184(12), 1374–1381.
transform infrared spectroscopy. Vibrational Spectroscopy, 55(2), 300–306. Verlee, A., Mincke, S., & Stevens, C. V. (2017). Recent developments in antibacterial and
Gao, H., Yang, Y., Akampumuza, O., Hou, J., Zhang, H., & Qin, X. (2017). A low filtration antifungal chitosan and its derivatives. Carbohydrate Polymers, 164, 268–283.
resistance three-dimensional composite membrane fabricated via free surface elec- Wang, Z., & Pan, Z. (2015). Preparation of hierarchical structured nano-sized/porous poly
trospinning for effective PM2.5 capture. Environmental Science: Nano, 4(4), 864–875. (lactic acid) composite fibrous membranes for air filtration. Applied Surface Science,
Goy, R. C., Morais, S. T. B., & Assis, O. B. G. (2016). Evaluation of the antimicrobial 356, 1168–1179.
activity of chitosan and its quaternized derivative on E-coli and S. aureus growth. Wang, C. C., Yang, F. L., & Zhang, H. M. (2010). Fabrication of non-woven composite
Brazilian Journal of Pharmacognosy, 26(1), 122–127. membrane by chitosan coating for resisting the adsorption of proteins and the ad-
Hamdine, M., Heuzey, M.-C., & Bégin, A. (2005). Effect of organic and inorganic acids on hesion of bacteria. Separation and Purification Technology, 75(3), 358–365.
concentrated chitosan solutions and gels. International Journal of Biological Wang, L. Y., Yong, W. F., Yu, L. E., & Chung, T. S. (2017). Design of high efficiency PVDF-
Macromolecules, 37(3), 134–142. PEG hollow fibers for air filtration of ultrafine particles. Journal of Membrane Science,
Hang, A. T., Tae, B., & Park, J. S. (2010). Non-woven mats of poly(vinyl alcohol)/chitosan 535, 342–349.
blends containing silver nanoparticles: Fabrication and characterization. Wang, S., Zhao, X., Yin, X., Yu, J., & Ding, B. (2016). Electret polyvinylidene fluoride
Carbohydrate Polymers, 82(2), 472–479. nanofibers hybridized by polytetrafluoroethylene nanoparticles for High-efficiency
Holda, A. K., Aernouts, B., Saeys, W., & Vankelecom, I. F. J. (2013). Study of polymer air filtration. ACS Applied Materials & Interfaces, 8(36), 23985–23994.
concentration and evaporation time as phase inversion parameters for polysulfone- Xiong, Z.-C., Yang, R.-L., Zhu, Y.-J., Chen, F.-F., & Dong, L.-Y. (2017). Flexible hydro-
based SRNF membranes. Journal of Membrane Science, 442, 196–205. xyapatite ultralong nanowire-based paper for highly efficient and multifunctional air
Huang, R.-J., Zhang, Y., Bozzetti, C., Ho, K.-F., Cao, J.-J., Han, Y., et al. (2014). High filtration. Journal of Materials Chemistry A, 5(33), 17482–17491.
secondary aerosol contribution to particulate pollution during haze events in China. Xu, W. S., Guo, M. Y., Liu, J., Xiao, Y. T., Zhou, G. Q., Liu, Y., et al. (2018). Poly(lactic-co-
Nature, 514(7521), 218–222. glycolic acid)/polycaprolactone nanofibrous membranes for high-efficient capture of
Hung, W.-L., Wang, D.-M., Lai, J.-Y., & Chou, S.-C. (2016). On the initiation of macrovoids nano- and microsized particulate matter. Journal of Biomedical Nanotechnology, 14(1),
in polymeric membranes – Effect of polymer chain entanglement. Journal of 179–189.
Membrane Science, 505, 70–81. Zając, A., Hanuza, J., Wandas, M., & Dymińska, L. (2015). Determination of N-acetylation
Jing, L., Shim, K., Toe, C. Y., Fang, T., Zhao, C., Amal, R., et al. (2016). Electrospun degree in chitosan using Raman spectroscopy. Spectrochimica Acta Part A: Molecular
polyacrylonitrile-ionic liquid nanofibers for superior PM2.5 capture capacity. ACS and Biomolecular Spectroscopy, 134, 114–120.
Applied Materials & Interfaces, 8(11), 7030–7036. Zhang, Q., He, K., & Huo, H. (2012). Policy: Cleaning China’s air. Nature, 484(7393),
Leung, W. W. F., & Hung, C. H. (2012). Skin effect in nanofiber filtration of submicron 161–162.
aerosols. Separation and Purification Technology, 92, 174–180. Zhang, S., Tang, N., Cao, L., Yin, X., Yu, J., & Ding, B. (2016). Highly integrated poly-
Leung, W. W. F., Hung, C. H., & Yuen, P. T. (2010). Effect of face velocity, nanofiber sulfone/polyacrylonitrile/polyamide-6 air filter for multilevel physical sieving air-
packing density and thickness on filtration performance of filters with nanofibers borne particles. ACS Applied Materials & Interfaces, 8(42), 29062–29072.
coated on a substrate. Separation and Purification Technology, 71(1), 30–37. Zhang, Y., Yuan, S., Feng, X., Li, H., Zhou, J., & Wang, B. (2016). Preparation of nano-
Li, P., Wang, C., Zhang, Y., & Wei, F. (2014). Air filtration in the free molecular flow fibrous metal-organic framework filters for efficient air pollution control. Journal of
regime: A review of high-efficiency particulate air filters based on carbon nanotubes. the American Chemical Society, 138(18), 5785–5788.
Small, 10(22), 4543–4561. Zhang, B., Zhang, Z.-G., Yan, X., Wang, X.-X., Zhao, H., Guo, J., et al. (2017). Chitosan
Li, S., Cui, Z., Zhang, L., He, B., & Li, J. (2016). The effect of sulfonated polysulfone on the nanostructures by in situ electrospinning for high-efficiency PM2.5 capture.
compatibility and structure of polyethersulfone-based blend membranes. Journal of Nanoscale, 9(12), 4154–4161.
Membrane Science, 513, 1–11. Zhang, S., Zou, Y., Wei, T., Mu, C., Liu, X., & Tong, Z. (2017). Pervaporation dehydration
Liu, K., Xiao, Z., Ma, P., Chen, J., Li, M., Liu, Q., et al. (2015). Large scale poly(vinyl of binary and ternary mixtures of n-butyl acetate, n-butanol and water using PVA-CS
alcohol-co-ethylene)/TiO2 hybrid nanofibrous filters with efficient fine particle fil- blended membranes. Separation and Purification Technology, 173, 314–322.
tration and repetitive-use performance. RSC Advances, 5(107), 87924–87931. Zhu, M. M., Han, J. Q., Wang, F., Shao, W., Xiong, R. H., Zhang, Q. L., et al. (2017).
Meng, J., Zhang, X., Ni, L., Tang, Z., Zhang, Y., Zhang, Y., et al. (2015). Antibacterial Electrospun nanofibers membranes for effective air filtration. Macromolecular
cellulose membrane via one-step covalent immobilization of ammonium/amine Materials and Engineering, 302(1), 1600353.
groups. Desalination, 359, 156–166. Zhuang, P.-Y., Li, Y.-L., Fan, L., Lin, J., & Hu, Q.-L. (2012). Modification of chitosan
Morgado, P. I., Miguel, S. P., Correia, I. J., & Aguiar, R. A. (2017). Ibuprofen loaded PVA/ membrane with poly(vinyl alcohol) and biocompatibility evaluation. International
chitosan membranes: A highly efficient strategy towards an improved skin wound Journal of Biological Macromolecules, 50(3), 658–663.
healing. Carbohydrate Polymers, 159, 136–145.

248