- Email: [email protected]
Fabrication of flexible self-standing all-cellulose nanofibrous composite membranes for virus removal
Accepted Manuscript Title: Fabrication of flexible self-standing all-cellulose nanofibrous composite membranes for virus removal Author: Weijuan Huang Yixiang Wang Chao Chen John Lok Man Law Michael Houghton Lingyun Chen PII: DOI: Reference:
S0144-8617(16)30051-0 http://dx.doi.org/doi:10.1016/j.carbpol.2016.02.011 CARP 10766
To appear in: Received date: Revised date: Accepted date:
24-10-2015 18-1-2016 2-2-2016
Please cite this article as: , Nanocomposite membranes showed rejection ratio of ¨ ¨/>pm
Highlights 1. All-cellulose nanocomposite membranes were fabricated as novel filtration system.
ip t
2. Hot pressed electrospun cellulose nanofabric provided mechanical support. 3. Regenerated cellulose gel coating with tiny inter-connected pores acted as barrier.
cr
4. Nanocomposite membranes showed rejection ratio of (98.68 ± 0.71)% against
Ac ce pt e
d
M
an
us
Hepatitis C Virus.
Page 1 of 33
4
Weijuan Huang a, Yixiang Wang a, Chao Chen b, John Lok Man Law b, Michael Houghton b,
5
a, Lingyun Chen *
6
a
7
AB, Canada T6G 2P5
8
b
9
University of Alberta, Edmonton, AB, Canada T6G 2E1
cr us
an
Department of Agricultural, Food & Nutritional Science, University of Alberta, Edmonton,
Li Ka Shing Institute of Virology, Department of Medical Microbiology and Immunology,
M
2
ip t
3
Fabrication of flexible self-standing all-cellulose nanofibrous composite membranes for virus removal
1
Corresponding Author
26
Corresponding author. Tel.: +1-780-492-0038; Fax: +1-780-492-8914.
27
Email address: [email protected] (L. Chen).
Ac ce pt e
d
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1
Page 2 of 33
28
ABSTRACT All-cellulose nanocomposite membranes with excellent performance were successfully
30
fabricated as novel filtration system to remove nanoparticles and virus from aqueous medium.
31
These membranes were composed of two combined layers: an electrospun cellulose
32
nanofabric layer treated by hot-pressing to provide mechanical support and a coating of
33
regenerated cellulose gel with tiny inter-connected pores as barrier. Hot-pressing didn’t affect
34
the fiber shape of electrospun nanofabrics, but significantly improved their mechanical
35
properties due to increased hydrogen bonds. The regenerated cellulose gel formed a porous
36
coating that tightly attached to electrospun nanofabrics, and its pore size varied depending on
37
cellulose source, solution concentration, and drying process. By assembling these two layers
38
together, the nanocomposite membranes showed the notable retention of negatively charged
39
100 nm latex beads (99.30%). Moreover, the electronegative nature of cellulose membranes
40
imparted the rejection ratio of 100% and (98.68±0.71)% against positively charged 50 nm
41
latex beads and Hepatitis C Virus, respectively.
cr
us
an
M
d
Ac ce pt e
42
ip t
29
43
KEYWORDS
44
All-cellulose ultrafiltration membranes, electrospun nanofiber, regenerated gel, flexible self-
45
standing, virus removal
46 47 48 49 50
2
Page 3 of 33
51
1. INTRODUCTION Virus contamination outbreak can often occur in medical or biotechnology products (e.g.,
53
vaccine, monoclonal/polyclonal antibody, plasma, immunoglobulin et al) (Levings &
54
Wessman, 1990). It presents a serious health hazard and results in large economic losses as
55
well as erosion of public trust. In order to reduce the risk of virus contamination, it is
56
necessary to purify biotechnology products. On the other hand, some researchers need to
57
enrich virus for specific experiments. However, the size of virus is extremely small; for
58
example, the swine influenza virus has a typical particle size of 80-120 nm in diameter
59
(Elford, Andrewes & Tang, 1936). It is difficult to separate virus from liquid media by simple
60
filtration. Nano-filtration is a convenient method to isolate small particles like virus (Asper,
61
Hanrieder, Quellmalz & Mihranyan, 2015; Quellmalz & Mihranyan, 2015; Rautenbach &
62
Gröschl, 1990), and has been used to convert sea water into drinking water by filtering salt
63
(Han, Xu & Gao, 2013). Nano-filtration removal or enrichment of virus is a promising
64
technology because it is non-destructive and non-interfering (Dishari, Micklin, Sung, Zydney,
65
Venkiteshwaran & Earley, 2015).
Ac ce pt e
d
M
an
us
cr
ip t
52
66
Recently, the thin film nanofibrous composite (TFNC) membrane consisting of an ultrathin
67
selective barrier layer (top layer), an electrospun nanofibrous scaffold (middle layer), and a
68
non-woven fabric support (bottom layer) has become popular in ultrafiltration systems (Kaur,
69
Barhate, Sundarrajan, Matsuura & Ramakrishna, 2011), because it not only can block
70
nanoscale substances but also has strong mechanical properties and high water flux (Ma et al.,
71
2010a). For example, poly(ethylene terephthalate) (PET) non-woven mats are usually
72
employed as the bottom layer to provide mechanical support, and electrospun
73
polyacrylonitrile (PAN) or polyvinyl alcohol (PVA) nanofibrous membranes constitute the
74
middle layer (Ma et al., 2010b; Yoon, Kim, Wang, Fang, Hsiao & Chu, 2006). Many studies
75
have focused on the fabrication of top barrier layer from a series of polymers, such as
76
cellulose regenerated from ionic liquids (Ma et al., 2010a), cellulose nanocrystals (Ma, 3
Page 4 of 33
Burger, Hsiao & Chu, 2014), chitin membrane (Ma, Hsiao & Chu, 2011), chemically cross-
78
linked PVA (Ma, Burger, Hsiao & Chu, 2012; Ma et al., 2010b), polyamide (Yoon, Hsiao &
79
Chu, 2009), and graphene oxide (Yeh, Wang, Mahajan, Hsiao & Chu, 2013). To the best of
80
our knowledge, TFNC membranes with only two layers that are completely developed from
81
natural polymers have never been reported.
ip t
77
Electrospining is a versatile and easy technology to fabricate nonwoven and continuous
83
nanofibers with diameters ranging from 50 nm to 500 nm by applying a high voltage (Kaur,
84
Sundarrajan, Rana, Matsuura & Ramakrishna, 2012; Liao et al., 2015; Unnithan,
85
Gnanasekaran, Sathishkumar, Lee & Kim, 2014). Electrospun nanofibrous membranes have
86
been widely used in filtration due to their highly porous network structure (Filatov, Budyka &
87
Kirichenko, 2007; Greiner & Wendorff, 2007). Among them, cellulose acetate has been
88
successfully electrospun and then deacetylated to prepare cellulose nanofabrics (Liu & Hsieh,
89
2002; Rodríguez, Renneckar & Gatenholm, 2011; Rodríguez, Sundberg, Gatenholm &
90
Renneckar, 2014; Son, Youk, Lee & Park, 2004; Zhou, Peng, Zhong, Wu, Cao, &Sun, 2016).
91
The obtained cellulose membranes exhibited superior filtration performance in aqueous
92
system due to their high porosity and hydrophilicity. Moreover, cellulose is biodegradable
93
and has good chemical resistance. It won’t react with majority of components in feed solution.
94
Until now, the most common filter membrane for aqueous system is made from cellulose.
95
However, the mechanical properties of electrospun nanofibrous cellulose membrane were
96
poor that limited their applications in filtration. It is reported that the hot pressed nanofibrous
97
membranes revealed better pressure tolerance and mechanical performance when compared to
98
untreated membranes (Kaur, Barhate, Sundarrajan, Matsuura & Ramakrishna, 2011; Wang et
99
al., 2013). Generally, electrospun nanofibers are randomly oriented and not interconnected at
100
joints. Hot press technique compresses the nanofibers together and generates new joints or
101
interactions under high temperature and pressure, thus enhances the structural integrity and
102
mechanical properties of electrospun nanofibrous membranes (Asper, Hanrieder, Quellmalz
Ac ce pt e
d
M
an
us
cr
82
4
Page 5 of 33
103
& Mihranyan, 2015; Lalia, Guillen-Burrieza, Arafat & Hashaikeh, 2013). Nevertheless, the
104
hot pressed electrospun cellulose nanofabrics have never been reported. Cotton and wood are good sources of cellulose because of the high yield and affordability
106
(Klemm, Heublein, Fink & Bohn, 2005). Canada is the largest world producer of newsprint
107
and northern bleached softwood kraft pulp, and the main component of paper and pulp is
108
wood-based cellulose (Kuhlberg, 2005). As a good cellulose solvent, the NaOH/urea aqueous
109
solution developed by Zhang’s group provides a ‘green’ and economical way to rapidly
110
dissolve cellulose at low temperature (Cai & Zhang, 2005). Recently, this technology has
111
been successfully adapted for dissolving wood cellulose after partial acidic hydrolysis (Gong,
112
Wang, Tian, Zheng & Chen, 2014). A series of functional cellulose materials have been
113
fabricated based on this solvent system, such as hydrogels (Zhou, Chang, Zhang & Zhang,
114
2007), microporous membranes (Zhou, Zhang, Cai & Shu, 2002), multifilament fibers (Cai et
115
al., 2007), and microspheres (Luo & Zhang, 2013). Particularly, cellulose solution can form
116
ultrathin gel membrane when being cast on a substrate (Zhou, Zhang, Cai & Shu, 2002). The
117
regenerated cellulose gel membrane exhibits a dense surface with interconnected nanoscale
118
pores (Wang & Chen, 2011). This unique structure provides the possibility to coat electrospun
119
cellulose nanofabrics with regenerated cellulose gel membrane to act as a selective barrier
120
layer to block virus and other microorganisms. Furthermore, the expected excellent
121
compatibility between two cellulose layers would be beneficial to increase water permeation
122
and separation efficiency (Lau, Ismail, Misdan & Kassim, 2012). The two-layer all-cellulose
123
nanofibrous composite membranes were fabricated in current study, where the hot pressing
124
treated electrospun cellulose nanofiber provided the mechanical support and the regenerated
125
cellulose gel coating worked as the separation layer. The structure, mechanical properties and
126
filtration performance of composite membranes were investigated, and the retention rate of
127
Hepatitis C Virus (HCV) was tested to evaluate the practical and functional performance.
Ac ce pt e
d
M
an
us
cr
ip t
105
128 5
Page 6 of 33
129
2. EXPERIMENTAL METHODS
130
2.1 Materials Spruce cellulose (bleached kraft pulp) with -cellulose content of 87.3% was provided by
132
Alberta-Pacific Forest Industries Inc. (AB, Canada). It was hydrolyzed by 20 wt% sulfuric
133
o 4 acid at 30 C for 24 h and the viscosity-average molecular weight (M ) was 5.8 × 10 (Gong,
134
Wang, Tian, Zheng & Chen, 2014). Cotton cellulose (cotton linter pulp) with M of 1.0 × 105
135
was supplied by Hubei Chemical Fiber Group, Ltd. (Xiangfan, China). Cellulose acetate (CA,
136
average MN Ca. 30 000, 39.8 wt% acetyl content) and fluorophore tagged polystyrene latex
137
beads (L9902, 100nm in diameter, sulfonate-modified; fluorescent red; ex~575 nm; em~610
138
nm and L0780, 50nm mean particle size, amine-modified, fluorescent blue, ex~360 nm;
139
em~420) were purchased from Sigma-Aldrich Canada Ltd. (Oakville, ON, Canada).
140
Commercial cellulose filter paper (CM) was purchased from WhatmanTM (GE Healthcare,
141
Buckinghamshire, UK). Dulbecco’s Modified Eagle Medium (DMEM), Non-Essential Amino
142
Acids (NEAA), penicillin and streptomycin were obtained from Life Technologies
143
(Burlington, ON, Canada). Acetic acid and all other chemical reagents were purchased from
144
Fisher Scientific (Markham, ON, Canada) and were used as received unless otherwise
145
described.
cr
us
an
M
d
Ac ce pt e
146
ip t
131
147
2.2 Preparation of all-cellulose nanofibrous membranes
148
2.2.1 Hot pressed electrospun cellulose nanofabrics
149
Nonwoven cellulose acetate nanofabrics were fabricated by a customized digital
150
electrospinning apparatus EC-DIG (IME Technologies, Eindhoven, Netherlands) at room
151
temperature (22 oC). Briefly, 8 g cellulose acetate was dissolved in 42 mL acetic acid/water
152
(75/25, v/v) solution, and then it was forced through a blunt needle with a diameter of 0.8 mm
153
at the rate of 1 mL h-1. The applied voltage was fixed at 23 kV. A rotating drum with a
154
diameter of 10 cm was chosen as the collector, and the distance between the tip and collector 6
Page 7 of 33
was set as 15 cm. The obtained cellulose acetate mats were subsequently immersed in 0.5 M
156
KOH ethanol solution at room temperature for 1 h to generate cellulose nanofabrics. The
157
cellulose nanofabrics were washed with excess deionized water and cut into square pieces (3
158
inch×3 inch). Several square pieces (5, 10 and 15 pieces) were piled up and sandwiched
159
between two plane white PTFE plates. They were pre-heated at 110 oC for 10 min and then
160
hot pressed by a Carver benchtop laboratory press (model 3851, Carver Inc., Wabash, IN)
161
under the pressure of about 7.66 MPa (10 000 pounds/3 inch×3 inch) at 110 oC for 50 min.
162
The resultant cellulose nanofabrics were coded as L5, L10 and L15, corresponding to the
163
different number of pieces used to pile up the nanofabrics. The cellulose sample without hot
164
pressing was coded as RC.
165
2.2.2 Regenerated cellulose gel coating
an
us
cr
ip t
155
Cotton cellulose (CC) and hydrolyzed spruce cellulose (SC) solutions with cellulose
167
concentration of 1, 2 and 3 wt% were prepared as described by Cai and Zhang (Cai & Zhang,
168
2005). Briefly, 7wt% NaOH/12wt% urea aqueous solution was precooled to -12.6 oC, and
169
desired amount of cellulose was added in the solution with mechanical stirring at 2000 rpm
170
for 3 min. The cellulose solutions were degassed by centrifugation at 805 g and 4 oC for 5 min.
171
To form the regenerated cellulose gel coating, the hot pressed electrospun cellulose
172
nanofabric L5 was firstly soaked in 75% acetic acid aqueous solution for 1 min, and then its
173
one face was quickly dipped in the above prepared cellulose solutions to obtain the one-side
174
coated all-cellulose nanofibrous composite membrane. The resultant samples were either
175
directly washed using deionized water, or immersed in pure ethanol for 1 h (solvent exchange
Ac ce pt e
d
M
166
7
Page 8 of 33
cr us
and then thoroughly washed. All the membranes were dried at room temperature and pre-immersed in water before the filtration test. The
177
detailed preparation processing parameters of each sample is shown in Table 1.
178
Table 1. Components, processing methods, and ultrafiltration performance of all-cellulose nanofibrous composite membranes.
M
an
176
Sample
Supporting layer
1CC-L5 2CC-L5
Hot pressed five-layer electrospun cellulose nanofabrics
Ac
2SC-SE-L5
ce
2SC-L5
3SC-SE-L5
180
Coating layer Cellulose Cellulose source concentrate Hydrolyzed spruce 1wt % cellulose Hydrolyzed spruce 2wt % cellulose Cotton cellulose 1wt % Cotton cellulose 2wt % Hydrolyzed spruce 2wt % cellulose Hydrolyzed spruce 3wt % cellulose Cotton cellulose 2wt % Cotton cellulose 3wt %
pt
1SC-L5
ed
179
2CC-SE-L5 3CC-SE-L5 n.d. means not detect.
Rejection ratio / %
Pure water flux / L m-2 h-1
100 nm
50 nm
No
189.87±55.35
77.14±11.51
n.d.
No
93.99±22.51
84.59±4.97
96.84±0.52
No No
130.08±44.75 48.73±1.93
54.86±15.12 82.35±1.62
n.d. 96.14±0.16
Yes
131.20±11.00
92.41±2.15
99.02±2.02
Yes
97.75±5.84
97.79±1.00
99.88±0.11
Yes Yes
113.42±1.76 89.47±2.96
95.56±2.35 99.30±0.66
99.65±0.20 100±0.08
Solvent exchange
181 182
8
Page 9 of 33
183
2.3 Structure and morphology Fourier transform infrared (FT-IR) spectra of electrospun cellulose acetate fibers and the
185
deacetylated cellulose fibers with/without hot pressing treatment were recorded on a Nicolet
186
6700 spectrophotometer (Thermo Fisher Scientific Inc., MA, USA) with KBr pellets. The
187
samples were vacuum-dried for 24 h prior to test. Spectra were recorded as the average of 32
188
scans at 4 cm-1 resolution at room temperature. During measurements the accessory
189
compartment was flushed with dry air.
cr
ip t
184
Morphology observation of electrospun cellulose fibers and regenerated gel coating were
191
carried out with a Philips XL-30 scanning electron microscope (SEM) at an acceleration
192
voltage of 5-10 kV. The all-cellulose membranes were frozen in liquid nitrogen, snapped
193
immediately and then freeze-dried to obtain the cross-sectional fracture surface. The samples
194
were sputtered with gold for 2 min prior to observation and photographing. In SEM photos,
195
fiber diameters were determined with the ImageJ image-visualization software developed by
196
the National Institute of Health (He et al., 2014). Two hundred random positions were
197
selected and measured for each sample.
199
an
M
d
Ac ce pt e
198
us
190
2.4 Mechanical properties
200
Tensile testing of the hot pressed electrospun cellulose nanofabrics was done using an
201
Instron 5967 universal testing machine (Instron Corp., MA, USA) at a crosshead speed of 4
202
-1 mm min and a gauge length of 20 mm according to the ASTM D-638-V standard (He et al.,
203
2014; Wang & Chen, 2014). Five bars with a dimension of 5 cm × 1 cm (length × width) were
204
cut from each fabric membrane, and their thickness was measured by a digital micrometer
205
(Mitutoyo, Japan) with a precision of 1 μm. Before testing, the samples were either vacuum-
206
dried or immersed in water for 24 h.
207 208
2.5 Ultrafiltration performance 9
Page 10 of 33
Pure water flux was tested in a Büchner funnel. All-cellulose nanofibrous composite
210
membranes with an effective filtration area of 5.3 cm2 (26 mm in diameter) were placed in the
211
funnel and 50 mL ultrapure water were used to flow through the membranes. The
212
performance pressure was 10 kPa. The pure water flux was determined by the following
213
equation:
214
Jw=Q/(A* t)
(1)
ip t
209
-2 -1 Where Jw is the pure water flux (L m h ), Q is the quantity of permeation (L), A is the
216
effective membrane area (m2), and t is the permeation time (h).
cr
215
For the particle retention test, 5 μL polystyrene latex bead suspensions were diluted to 10
218
mL with ultrapure water. Afterwards, the diluent was filtered through the all-cellulose
219
nanofibrous membranes under the vacuum pressure of 10 kPa. The fluorescence intensity of
220
the diluent, filtrate and ultrapure water was measured by a SpectraMax M3 microplate reader
221
with SpectraMax Pro Software (Molecular Devices, Inc., USA) at the specified excitation and
222
emission wavelengths. The zeta-potential of 50 nm and 100 nm latex beads dispersed in pure
223
water was measured by a Malvern zeta-sizer (Malvern instruments Inc., UK).
Ac ce pt e
d
M
an
us
217
224
A CLSM 710 Meta confocal laser scanning microscope (Carl Zeiss, Jena, Germany) was
225
used for cross-sectional view of all-cellulose nanofibrous composite membranes after
226
filtration of 100 nm polystyrene latex beads, and the images were recorded using ZEN 2009
227
LE software (Carl Zeiss MicroImaging GmbH, Germany).
228 229
2.6 Retention of Hepatitis C Virus
230
Huh7.5 cells were cultured in DMEM supplemented with 10% FBS, 0.1 mM NEAA and
231
100 μg each of penicillin and streptomycin. Cell culture derived Hepatitis C Virus (HCVcc)
232
was produced using previously described protocol (Lindenbach et al., 2005). Cells were
233
7 washed twice with ice cold PBS and subsequently resuspended to 1.5×10 cells/mL. Then,
234
400 μL of the cell suspension were mixed with 5 μg in vitro transcribed RNA encoding HCV 10
Page 11 of 33
genome in 2 mm gap electroporation curvettes (Bio-Rad, Mississauga, ON, Canada), and 5
236
pulses of 860 V (99 μs, 1.1 s interval) were delivered using the ElectroSquare Porator ECM
237
830 (BTX, Holliston, MA). Post-electroporation, cells were incubated at room temperature
238
for 10 min before plating. Pre-cleared media was collected as virus stocks either 3 or 4 d post-
239
electroporation. The virus titer (50% tissue culture infectious dose (TCID50)) was determined
240
by limited dilution as described previously (Lindenbach et al., 2005).
ip t
235
HCV-JC1 viral stock (5.13×104 TCID50 HCVcc/mL,) was filtered through a sterile
242
membrane in a Bu•chner funnel (Fisher Scientific, Ottawa, ON, Canada) under the vacuum
243
pressure of 10 kPa. Pre- and post-filtration viruses were assessed by NS5A staining. For the
244
NS5A (Nonstructural Protein 5A) staining, 200 μL pre- or post-filtration viruses were diluted
245
to 686 μL and 100 μL dilutions were inoculated to Huh7.5 cells pre-seeded in a 96 well plate
246
(BD, Mississauga, ON, Canada) in quadruplicate. After the 12 h incubation, the viral
247
inoculum was replaced with fresh culture media. Then, 48 h after inoculation, cells were fixed
248
and stained as described previously (Law et al., 2013). The foci per well were detected and
249
counted using a CTL S6 immunospot analyzer (CTL, Cleveland OH) as described previously
250
(Gottwein et al., 2010).
252
us
an
M
d
Ac ce pt e
251
cr
241
2.7 Statistical analysis
253
Experimental results were represented as the mean ± SD. Statistical evaluation was carried
254
out by analysis of variance (ANOVA) followed by multiple-comparison tests using Duncan’s
255
multiple-range test at the 95% confidence level. All of the analyses were conducted using
256
SAS statistical software (SAS Institute, Inc., Cary, NC) with a probability of p < 0.05
257
considered to be significant.
258
3. RESULTS AND DISCUSSION
259
3.1 Structure and mechanical properties of hot pressed electrospun fabrics
11
Page 12 of 33
260
The molecular structure changes of electrospun cellulose acetate (CA) fibers after
261
deacetylation and hot pressing were recorded by FT-IR. As shown in Fig. 1, the electrospun
262
-1 CA fibers exhibited several characteristic absorption peaks at 3496 cm (
263
1375 cm-1 (
264
the characteristic absorption peaks of acetyl group disappeared, indicating CA had been
265
completely converted to cellulose (Ma, Kotaki & Ramakrishna, 2005). At the same time, the
266
absorption peak of
267
3445cm-1, suggesting the improvement of hydrogen bonds due to the recovered hydroxyl
268
groups in cellulose (Liu & Hsieh, 2002),. The
269
broadened, indicating that the stronger hydrogen bonding interactions were developed during
270
hot pressing treatment (Figueiredo, Evtuguin & Saraiva, 2010; Luo, Zhu, Gleisner & Zhan,
271
2011; ŠUTÝ et al., 2012). This enhanced hydrogen bonding could induce the hornification of
272
cellulose and result in the improved dimensional stability (Luo, Zhu, Gleisner & Zhan, 2011;
273
Weise & Paulapuro, 1999).
) (Son, Youk, Lee & Park, 2004). After deacetylation,
cr
became broader and stronger and shifted to lower wavenumber of
peak of hot pressed cellulose fabric further
us
OH
Ac ce pt e
d
M
an
OH
C-O-C
),
C=O
ip t
), and 1237 cm-1 (
C-CH3
), 1754 cm-1 (
OH
274
12
Page 13 of 33
Fig. 1. FT-IR spectra of (a) electrospun cellulose acetate nanofabrics (CA), (b) deacetylated
276
electrospun cellulose nanofabrics (RC), and (c) hot pressed five-layer electrospun
277
cellulose nanofabrics (L5).
Ac ce pt e
d
M
an
us
cr
ip t
275
278 279
Fig. 2. SEM images and diameter distribution of (a) electrospun cellulose acetate nanofabrics
280
(CA), (b) deacetylated electrospun cellulose nanofabrics (RC), and (c) hot pressed 13
Page 14 of 33
281
five-layer electrospun cellulose nanofabrics (L5).
The morphology and width distribution of electrospun CA fibers and cellulose fibers with or
284
without hot pressing are shown in Fig. 2. Uniform CA nanofibers were obtained with flat
285
ribbon-like shape. Similar fibers were reported by Han and co-workers (Han, Youk, Min,
286
Kang & Park, 2008). This was because a thin skin formed on the solution jet and acted as the
287
tube wall. Due to the continuous evaporation of inside solvent, the atmospheric pressure made
288
the tubes collapse and resulted in the elliptical ribbons, which finally became the flat ribbons
289
(Koombhongse, Liu & Reneker, 2001; Ramakrishna, Fujihara, Teo, Lim & Ma, 2005). The
290
hydrolyzed cellulose nanofibers exhibited the same shape and similar width compared to
291
electrospun CA fibers, indicating that the hydrolysis didn’t affect the fiber morphology. It was
292
worth noting the nanoscale fiber shape and porous network structure still well maintained
293
even after hot pressing. This feature could enable the high water flux during filtration. No
294
fusion appeared in the node or intersection of overlapped fibers because cellulose doesn’t
295
melt or carbonize at 110 oC. The width of cellulose nanofibers slightly increased from
296
213±116 nm to 260±130 nm, since the fibers were wet and swelled to some extent prior to hot
297
pressing.
Ac ce pt e
d
M
an
us
cr
ip t
282 283
298 299
Fig. 3. Typical stress-strain curves of membranes at (a) dry status (inset, photo of hot-pressed 14
Page 15 of 33
300
electorspun cellulose nanofabric (L5)), and (b) wet status (inset, photo of hot-pressed
301
electorspun cellulose nanofabric (L5) immersed in water for 24h). Normally, filter membranes are used in liquid media and should be strong enough to stand
303
the pressure. The mechanical strength of electrospun nanofabrics was thus tested to evaluate
304
their handling property. Typical stress-strain curves of cellulose nanofabrics with or without
305
hot pressing (L5, L10, L15, and RC) are shown in Fig. 3. Their tensile strength, elongation at
306
break and Young’s modulus are summarized in Table S1. The mechanical properties of
307
commercial filter paper (CM, Whatman™ 1002-125 Grade 2 Qualitative Filter Paper) were
308
also measured for comparison. The tensile strength of RC was 6.9±1.6 MPa. After hot
309
pressing, the L5 fabric exhibited a greatly improved strength of 11.7±2.0 MPa due to the
310
enhanced hydrogen bonding interactions. However, the tensile strength of L10 and L15
311
samples decreased to 10.4±1.8 and 6.8±3.1 MPa, respectively. It was noticed that the
312
thickness of the fabrics increased when more electrospun cellulose pieces were hot pressed
313
together. The compression force was thus diffused and led to the formation of relatively
314
looser interlayers which affected the efficient stress transfer (Wu, Shuai, Cheng & Jiang, 2014;
315
Zhang, Zhang & Gao, 2011). The mechanical properties of L5 at dry status were only close to
316
those of commercial filter paper, but L5 was three times stronger than commercial product
317
when they were immersed in water. The tensile strength of wet L5 and CM was 1.5±0.1MPa
318
and 0.5±0.1 MPa, respectively, and their elongation at break was 12.6±1.1% and 3.4±0.2%,
319
respectively. It was the result of the developed hydrogen bonding interactions in electrospun
320
cellulose nanofibers treated by hot pressing, which restricted the swelling of cellulose (Weise
321
& Paulapuro, 1999). The photographs of L5 at dry and wet status are inserted in Fig. 3a and
322
Fig. 3b, respectively. It exhibited good flexibility and remained integrated after immersing in
323
water for 24 h. These results indicated that the L5 nanofabric was suitable as the support layer
324
of composite ultrafiltration system and provided sufficient handling properties.
Ac ce pt e
d
M
an
us
cr
ip t
302
325 15
Page 16 of 33
326 3.2 Morphology of regenerated cellulose gel coating
328
Ac ce pt e
d
M
an
us
cr
ip t
327
329
Fig. 4．SEM images of cross-section (left) and surface (right) of all-cellulose nanofibrous
330
composite membranes: (a, b) 1SC-L5, (c, d) 2SC-L5, (e, f) 1CC-L5, and (g, h) 2CC-
331
L5.
332 16
Page 17 of 33
ip t cr us an M d Ac ce pt e
333 334
Fig. 5．SEM images of cross-section (left) and surface (right) of all-cellulose nanofibrous
335
composite membranes: (a, b) 2SC-SE-L5, (c, d) 3SC-SE-L5, (e, f) 2CC-SE-L5, and (g,
336
h) 3CC-SE-L5.
337 338
17
Page 18 of 33
Both cotton cellulose (CC) and hydrolyzed spruce cellulose (SC) were chosen as the
340
coating materials to investigate the effects of cellulose source and concentration as well as
341
drying procedure on the morphology of gel coatings. Fig. 4 shows SEM images of cross-
342
section and surface of composite cellulose nanofibrous membranes prepared by directly
343
drying at room temperature. A thin cellulose top layer with porous structure existed in all the
344
samples. No obvious gap was observed between electrospun cellulose nanofabric and
345
cellulose gel coating. Such good compatibility could be beneficial to increase pure water flux
346
and filtration efficiency. The gel coatings formed by 1% cellulose solutions were too thin to
347
be distinguished (ca. 40 μm) and exhibited relatively large pores on the surface. Dense
348
coating layers with the thickness of about 130 μm were formed in both 2SC-L5 and 2CC-L5
349
samples when cellulose concentration was increased to 2%. The morphology of cross-section
350
and surface of cellulose coating layers made by solvent exchange is shown in Fig. 5.
351
Compared to cellulose coating layers made by directly drying, the coatings generated by
352
solvent exchange and higher cellulose concentration showed relatively larger thickness (150-
353
300 μm) but uniform structures with more and smaller pores. During the directly drying
354
process, free water evaporated to induce a microporous structure that collapsed to certain
355
extent, causing membrane shrinking and cracking. Whereas during solvent exchange process,
356
the water within cellulose gel was replaced by volatile fluid with lower surface tension.
357
According to previous report (Jie, Cao, Qin, Liu & Yuan, 2005), the change of cellulose
358
coating morphology during various drying processes was primarily ascribed to the molecular
359
affinities of cellulose-ethanol, cellulose-water, and water-ethanol. And the molecular affinity
360
of cellulose-ethanol is weaker than that of cellulose-water. So the evaporation of ethanol
361
could maintain the uniform porous structure of cellulose gel (Jawad, Ahmad, Low, Chew &
362
Zein, 2015). The pore size further decreased with raising cellulose solution concentration
363
from 2% to 3%. It was reported that cellulose molecular chain entanglement took place during
364
the regeneration which resulted in the formation of porous gel network with dense surface and
Ac ce pt e
d
M
an
us
cr
ip t
339
18
Page 19 of 33
large inner holes (Wang & Chen, 2011). Thus, smaller pores were generated when the
366
cellulose concentration was higher. These nanoscale pores on the surface provided the
367
possibility to block tiny material during the filtration while the inside large pores ensured the
368
good permeability of water (as shown in Fig. 5).
369
3.3 Ultrafiltration performance of all-cellulose membranes
370
3.3.1 Permeability
ip t
365
The permeability of TFNC membrane is mainly determined by the morphology and
372
thickness of the top barrier layer (Ma et al., 2010a). The pure water flux of all-cellulose
373
membranes was measured and the values were listed in Table 1. The 1SC-L5 and 1CC-L5
374
membranes exhibited the highest water permeability of 189.87±55.35 L m-2 h-1 and
375
130.08±44.75 L m-2 h-1, respectively, due to their very thin coatings. The water flux of 2CC-
376
L5 (48.73±1.93 L m-2 h-1) was obviously lower than that of 2SC-L5 (93.99±22.51 L m-2 h-1),
377
since more pores were generated in the matrix-filler structure of 2SC-L5. The pure water flux
378
of 2SC-SE-L5 and 2CC-SE-L5 were 131.20±11.00 L m-2 h-1 and 113.42±1.76 L m-2 h-1,
379
respectively, which was higher than that of 2SC-L5 and 2CC-L5. It indicated that the solvent
380
exchange process produced more through-pores in cellulose coating layer compared to direct
381
drying at room temperature. Usually, a good filter membrane should have high filtration
382
efficiency and high pure water flux to save time and energy. The all-cellulose nanofibrous
383
composite membranes showed much higher permeability than TFNC membranes composed
384
of a 0.3 μm cotton cellulose layer regenerated from ionic liquid, a polyacrylonitrile
385
nanofibrous scaffold and a melt-blown poly (ethylene terephthalate) (PET) non-woven
386
substrate (28.0 L m-2 h-1 under 10 kPa and 73.7 L m-2 h-1 under 10 psi) (Ma et al., 2010a).
387
Moreover, the dimension of the all-cellulose membranes didn’t change after filtration. The
388
result indicated that cellulose gel membranes prepared from NaOH/urea aqueous solution had
389
promising filtration efficiency.
390
3.3.2 Nanoparticle retention test
Ac ce pt e
d
M
an
us
cr
371
19
Page 20 of 33
The rejection ratio of all-cellulose nanofibrous composite membranes was evaluated by the
392
retention test of fluorescence-labeled polystyrene latex beads. Two kinds of latex beads were
393
employed where the 100 nm beads were modified by electronegative sulfate group and the 50
394
nm ones were modified by positively charged amino group. As shown in Table 1 and Fig. S1,
395
for 100 nm beads, 1SC-L5 and 1CC-L5 exhibited the low rejection ratio of 77.14±11.51%
396
and 54.86±15.12%, respectively, due to their thin cellulose gel coatings and large pores on the
397
coating surface. With the increase of cellulose concentration, more beads were blocked by the
398
regenerated cellulose coatings, and the rejection ratio of 2SC-L5 and 2CC-L5 was
399
84.59±4.97% and 82.35±1.62%. The retention of 50 nm latex beads was even greater and ca.
400
97% beads were blocked by 2SC-L5. The rejection ratio of cellulose coating layer prepared
401
by solvent exchange was much higher than that of directly dried coating. For 100 nm beads,
402
the rejection ratio of 2SC-SE-L5 and 2CC-SE-L5 was 92.41±2.15% and 95.56±2.35%,
403
respectively, and it was even higher with increasing the cellulose concentration. The 3SC-SE-
404
L5 and 3CC-SE-L5 samples could block most beads because of their small surface pore sizes.
405
For 50 nm latex beads, the rejection ratio of 2SC-SE-L5, 2CC-SE-L5, 3SC-SE-L5 and 3CC-
406
SE-L5 was all above 99%. Figs. 6a and b show the morphology of 2SC-L5 after the filtration.
407
Both the 100 nm and 50 nm latex beads were successfully trapped on the surface of cellulose
408
coating. The improved retention ratio to 50 nm beads should be related to the surface charge.
409
As tested, the surface charge of 100 nm and 50 nm latex beads were -54.8±0.96 mV and
410
+25.2±0.62 mV, respectively. At the same time, the cellulose membrane presented the
411
negative zeta potential because of the abundant hydroxyl groups (Ma, Burger, Hsiao & Chu,
412
2011). The attractive interactions existed between 50 nm latex beads and filtration membrane
413
and resulted in the better capture capacity. In order to investigate the distribution of latex
414
beads in all-cellulose membrane after filtration, the confocal scanning microscopy image is
415
shown in Fig. 6c. Most latex beads located on the top surface of cellulose coating layer and a
416
few beads accumulated at the interface of two layers. However, no latex beads were observed
Ac ce pt e
d
M
an
us
cr
ip t
391
20
Page 21 of 33
within the gel network of cellulose coating. It could be supposed that a dense gel structure
418
also formed at the interface of electrospun nanofabric and cellulose coating due to the rapid
419
regeneration of cellulose solution caused by the direct contact of acetic acid. Therefore, a
420
large rejection ratio was achieved by the dense surface/interface of regenerated cellulose gel
421
coating while its porous inside structure enabled the large pure water flux.
us
cr
ip t
417
an
422
Fig. 6. SEM images of all-cellulose nanofibrous composite membrane (2SC-L5) after the
424
retention test: (a) 100 nm and (b) 50 nm latex beads. (c) Confocal microscopic image
425
of cross-sectional view of cellulose coating layer after filtration of 100nm
426
fluorochrome tagged polystyrene latex beads.
428 429 430 431 432 433 434
d
Ac ce pt e
427
M
423
3.3.3 Virus retention test
21
Page 22 of 33
ip t cr us 438
M
an 437
Fig. 7. Focus forming units of Hepatitis C Virus (HCV) solution before and after filtration
d
436
through 2SC-L5 and 2CC-SE-L5 all-cellulose nanofibrous composite membrane.
Ac ce pt e
435
439
3SC-SE-L5 and 3CC-SE-L5 samples exhibited an unstable dimension during drying
440
because of the excessive shrinkage of thick cellulose coatings. Thus, taking both rejection
441
ratio and pure water flux into consideration, 2SC-L5 and 2CC-SE-L5 membranes were
442
selected for the virus retention test. The Hepatitis C Virus (HCV) was chosen as the model
443
that has a spherical shape with positive surface charge and the diameter of 55-65 nm (Kaito et
444
al., 1994; Shimizu, Feinstone, Kohara, Purcell & Yoshikura, 1996). Fig. 7 shows the focus
445
forming units (FFU) in the virus solution before and after filtration. The FFU value treated by
446
2SC-L5 decreased from 183.75±10.80 to 14.75±3.03, indicating that about 92% HCV was
447
removed from the solution. For 2CC-SE-L5 membranes, the rejection ratio increased to
448
(98.68±0.71)%. However, the commercial microfiltration membrane GS0.22 (a mixed 22
Page 23 of 33
449
cellulose esters membrane with 0.22 μm pore size) showed only 90% retention of MS2
450
bacteriophage (Ma, Hsiao & Chu, 2014). We also notice that the filtration performance of
451
2CC-SE-L5 membranes was not sufficient for the requirement from water purification
452
industry, and further modifications were needed to improve this all-cellulose filtration system.
454
ip t
453
4. CONCLUSION
This is the first documentation of all-cellulose nanofibrous composite membrane made by
456
assembling an electrospun cellulose nanofabric layer as the mechanical framework and a
457
coating of regenerated cellulose gel membrane as the barrier. The hot pressing treatment led
458
to enhanced hydrogen bonding interactions, which significantly improved the mechanical
459
properties of supporting layer. The resultant nanofabric possessed a three-fold stronger wet
460
strength compared to commercial filter paper, while maintaining its porous structure. The
461
cellulose gel membrane regenerated from NaOH/urea aqueous solution was well attached on
462
the electrospun cellulose nanofabric. This barrier layer exhibited a unique structure of dense
463
surface/interface and large interior pores, which not only blocked the virus and nanoparticles
464
in water, but also ensured a large water flux. The retention tests indicated that these
465
membranes could remove beads as small as 50 nm. Finally, the rejection ratio of 2CC-SE-L5
466
against HCV was (98.68±0.71)% suggesting its potential applications in virus removal.
468 469
us
an
M
d
Ac ce pt e
467
cr
455
470 471 472
ACKNOWLEDGMENT
473
The authors are grateful to the Natural Sciences and Engineering Research Council of Canada
474
(NSERC), Alberta Crop Industry Development Fund Ltd. (ACIDF), Alberta Innovates Bio 23
Page 24 of 33
Solutions (AI Bio) and Alberta Barley Commission for financial support as well as Canada
476
Foundation for Innovation (CFI) for equipment support. Lingyun Chen would like to thank
477
the Natural Sciences and Engineering Research Council of Canada (NSERC)-Canada
478
Research Chairs Program for its financial support. Weijuan Huang thanks the support from
479
China Scholarship Council (CSC).
ip t
475
480
cr
481 482
us
483
an
484 485
M
486 487
d
488
490 491 492 493 494 495
Ac ce pt e
489
496 497
REFERENCES
498
Asper, M., Hanrieder, T., Quellmalz, A., & Mihranyan, A. (2015). Removal of xenotropic
499
murine leukemia virus by nanocellulose based filter paper. Biologicals, 43(6), 452-456. 24
Page 25 of 33
500 501
Cai, J., & Zhang, L. (2005). Rapid dissolution of cellulose in LiOH/urea and NaOH/urea aqueous solutions. Macromolecular Bioscience, 5, 539-548. Cai, J., Zhang, L., Zhou, J., Qi, H., Chen, H., Kondo, T., Chen, X., & Chu, B. (2007).
503
Multifilament fibers based on dissolution of cellulose in NaOH/urea aqueous solution:
504
structure and properties. Advanced Materials, 19, 821-825.
ip t
502
Dishari, S. K., Micklin, M. R., Sung, K. J., Zydney, A. L., Venkiteshwaran, A., & Earley, J. N.
506
(2015). Effects of solution conditions on virus retention by the Viresolve® NFP filter.
507
Biotechnology Progress, 31, 1280-1286.
an
us
cr
505
Elford, W., Andrewes, C., & Tang, F. (1936). The Sizes of the Viruses of Human and Swine
509
Influenza, as Determined by Ultra-Filtration. British Journal of Experimental Pathology,
510
17, 51-53.
512
d
Figueiredo, A., Evtuguin, D., & Saraiva, J. (2010). Effect of high pressure treatment on
Ac ce pt e
511
M
508
structure and properties of cellulose in eucalypt pulps. Cellulose, 17, 1193-1202.
513
Filatov, Y., Budyka, A., & Kirichenko, V. (2007). Electrospinning of micro-and nanofibers:
514
fundamentals in separation and filtration processes. Journal of Engineered Fibers and
515
Fabrics, 3, 488.
516 517
Gong, X., Wang, Y., Tian, Z., Zheng, X., & Chen, L. (2014). Controlled production of spruce cellulose gels using an environmentally “green” system. Cellulose, 21, 1667-1678.
518
Gottwein, J. M., Scheel, T. K., Callendret, B., Li, Y.-P., Eccleston, H. B., Engle, R. E.,
519
Govindarajan, S., Satterfield, W., Purcell, R. H., & Walker, C. M. (2010). Novel
520
infectious cDNA clones of hepatitis C virus genotype 3a (strain S52) and 4a (strain 25
Page 26 of 33
521
ED43): genetic analyses and in vivo pathogenesis studies. Journal of Virology, 84, 5277-
522
5293. Greiner, A., & Wendorff, J. H. (2007). Electrospinning: a fascinating method for the
524
preparation of ultrathin fibers. Angewandte Chemie International Edition, 46, 5670-5703.
525
Han, S. O., Youk, J. H., Min, K. D., Kang, Y. O., & Park, W. H. (2008). Electrospinning of
526
cellulose acetate nanofibers using a mixed solvent of acetic acid/water: Effects of solvent
527
composition on the fiber diameter. Materials Letters, 62, 759-762.
529
cr
us
Han, Y., Xu, Z., & Gao, C. (2013). Ultrathin graphene nanofiltration membrane for water
an
528
ip t
523
purification. Advanced Functional Materials, 23, 3693-3700. He, X., Xiao, Q., Lu, C., Wang, Y., Zhang, X., Zhao, J., Zhang, W., Zhang, X., & Deng, Y.
531
(2014). Uniaxially Aligned Electrospun All-Cellulose Nanocomposite Nanofibers
532
Reinforced
533
Biomacromolecules, 15, 618-627.
Cellulose
Nanocrystals:
Scaffold
for
Tissue
Engineering.
Ac ce pt e
with
d
M
530
534
Jawad, Z., Ahmad, A., Low, S., Chew, T., & Zein, S. (2015). Influence of solvent exchange
535
time on mixed matrix membrane separation performance for CO2/N2 and a kinetic
536
sorption study. Journal of Membrane Science, 476, 590-601.
537
Jie, X., Cao, Y., Qin, J.-J., Liu, J., & Yuan, Q. (2005). Influence of drying method on
538
morphology and properties of asymmetric cellulose hollow fiber membrane. Journal of
539
Membrane Science, 246, 157-165.
540
Kaito, M., Watanabe, S., Tsukiyama-Kohara, K., Yamaguchi, K., Kobayashi, Y., Konishi, M.,
541
Yokoi, M., Ishida, S., Suzuki, S., & Kohara, M. (1994). Hepatitis C virus particle 26
Page 27 of 33
542
detected by immunoelectron microscopic study. The Journal of general virology, 75,
543
1755-1760. Kaur, S., Barhate, R., Sundarrajan, S., Matsuura, T., & Ramakrishna, S. (2011). Hot pressing
545
of electrospun membrane composite and its influence on separation performance on thin
546
film composite nanofiltration membrane. Desalination, 279, 201-209.
ip t
544
Kaur, S., Sundarrajan, S., Rana, D., Matsuura, T., & Ramakrishna, S. (2012). Influence of
548
electrospun fiber size on the separation efficiency of thin film nanofiltration composite
549
membrane. Journal of Membrane Science, 392, 101-111.
an
us
cr
547
Klemm, D., Heublein, B., Fink, H. P., & Bohn, A. (2005). Cellulose: fascinating biopolymer
551
and sustainable raw material. Angewandte Chemie International Edition, 44, 3358-3393.
552
Koombhongse, S., Liu, W., & Reneker, D. H. (2001). Flat polymer ribbons and other shapes
553
by electrospinning. Journal of Polymer Science Part B: Polymer Physics, 39(21), 2598-
554
2606.
556 557
d
Ac ce pt e
555
M
550
Kuhlberg, M. (2005). (http://www.thecanadianencyclopedia.ca/en/article/pulp-and-paperindustry/).
Lalia, B. S., Guillen-Burrieza, E., Arafat, H. A., & Hashaikeh, R. (2013). Fabrication and
558
characterization
559
electrospun membranes for direct contact membrane distillation. Journal of Membrane
560
Science, 428, 104-115.
561 562
of
polyvinylidenefluoride-co-hexafluoropropylene
(PVDF-HFP)
Lau, W. J., Ismail, A. F., Misdan, N., & Kassim, M. A. (2012). A recent progress in thin film composite membrane: a review. Desalination, 287, 190-199. 27
Page 28 of 33
Law, J. L. M., Chen, C., Wong, J., Hockman, D., Santer, D. M., Frey, S. E., Belshe, R. B.,
564
Wakita, T., Bukh, J., & Jones, C. T. (2013). A hepatitis C virus (HCV) vaccine
565
comprising envelope glycoproteins gpE1/gpE2 derived from a single isolate elicits broad
566
cross-genotype neutralizing antibodies in humans. PLoS ONE, 8, e59776.
ip t
563
Levings, R., & Wessman, S. (1990). Bovine viral diarrhea virus contamination of nutrient
568
serum, cell cultures and viral vaccines. Developments in Biological Standardization, 75,
569
177-181.
us
Liao, N., Unnithan, A. R., Joshi, M. K., Tiwari, A. P., Hong, S. T., Park, C.-H., & Kim, C. S.
571
(2015).
bioactive
poly
(ɛ-caprolactone)–cellulose
acetate–dextran
572
antibacterial composite mats for wound dressing applications. Colloids and Surfaces A:
573
Physicochemical and Engineering Aspects, 469, 194-201.
d
M
Electrospun
an
570
cr
567
Lindenbach, B. D., Evans, M. J., Syder, A. J., Wölk, B., Tellinghuisen, T. L., Liu, C. C.,
575
Maruyama, T., Hynes, R. O., Burton, D. R., & McKeating, J. A. (2005). Complete
576
replication of hepatitis C virus in cell culture. Science, 309, 623-626.
Ac ce pt e
574
577
Liu, H., & Hsieh, Y. L. (2002). Ultrafine fibrous cellulose membranes from electrospinning
578
of cellulose acetate. Journal of Polymer Science Part B: Polymer Physics, 40, 2119-2129.
579
Luo, X., & Zhang, L. (2013). New solvents and functional materials prepared from cellulose
580
solutions in alkali/urea aqueous system. Food Research International, 52, 387-400.
581
Luo, X., Zhu, J., Gleisner, R., & Zhan, H. (2011). Effects of wet-pressing-induced fiber
582
hornification on enzymatic saccharification of lignocelluloses. Cellulose, 18, 1055-1062.
28
Page 29 of 33
583 584
Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2011). Nanofibrous microfiltration membrane based on cellulose nanowhiskers. Biomacromolecules, 13, 180-186. Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2012). Highly permeable polymer membranes
586
containing directed channels for water purification. ACS Macro Letters, 1, 723-726.
587
Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2014). Fabrication and characterization of
588
cellulose nanofiber based thin-film nanofibrous composite membranes. Journal of
589
Membrane Science, 454, 272-282.
us
cr
ip t
585
Ma, H., Hsiao, B. S., & Chu, B. (2011). Thin-film nanofibrous composite membranes
591
containing cellulose or chitin barrier layers fabricated by ionic liquids. Polymer, 52,
592
2594-2599.
M
an
590
Ma, H., Hsiao, B. S., & Chu, B. (2014). Functionalized electrospun nanofibrous
594
microfiltration membranes for removal of bacteria and viruses. Journal of Membrane
595
Science, 452, 446-452.
Ac ce pt e
d
593
596
Ma, H., Yoon, K., Rong, L., Mao, Y., Mo, Z., Fang, D., Hollander, Z., Gaiteri, J., Hsiao, B. S.,
597
& Chu, B. (2010a). High-flux thin-film nanofibrous composite ultrafiltration membranes
598
containing cellulose barrier layer. Journal of Materials Chemistry, 20, 4692-4704.
599
Ma, H., Yoon, K., Rong, L., Shokralla, M., Kopot, A., Wang, X., Fang, D., Hsiao, B. S., &
600
Chu, B. (2010b). Thin-film nanofibrous composite ultrafiltration membranes based on
601
polyvinyl alcohol barrier layer containing directional water channels. Industrial &
602
Engineering Chemistry Research, 49, 11978-11984.
29
Page 30 of 33
603 604
Ma, Z., Kotaki, M., & Ramakrishna, S. (2005). Electrospun cellulose nanofiber as affinity membrane. Journal of Membrane Science, 265, 115-123. Quellmalz, A., & Mihranyan, A. (2015). Citric Acid Cross-Linked Nanocellulose-Based
606
Paper for Size-Exclusion Nanofiltration. ACS Biomaterials Science & Engineering, 1,
607
271-276.
610 611
cr
Electrospinning and Nanofibers. Singapore: World Scientific Publishing.
us
609
Ramakrishna, S., Fujihara, K., Teo, W.-E., Lim, T.-C., & Ma, Z. (2005). An Introduction to
Rautenbach, R., & Gröschl, A. (1990). Separation potential of nanofiltration membranes.
an
608
ip t
605
Desalination, 77, 73-84.
Rodríguez, K., Renneckar, S., & Gatenholm, P. (2011). Biomimetic calcium phosphate
613
crystal mineralization on electrospun cellulose-based scaffolds. ACS applied materials &
614
interfaces, 3, 681-689.
Ac ce pt e
d
M
612
615
Rodríguez, K., Sundberg, J., Gatenholm, P., & Renneckar, S. (2014). Electrospun nanofibrous
616
cellulose scaffolds with controlled microarchitecture. Carbohydrate Polymers, 100, 143-
617
149.
618
Shimizu, Y. K., Feinstone, S. M., Kohara, M., Purcell, R. H., & Yoshikura, H. (1996).
619
Hepatitis C virus: Detection of intracellular virus particles by electron microscop.
620
Hepatology, 23, 205-209.
621
Son, W. K., Youk, J. H., Lee, T. S., & Park, W. H. (2004). Electrospinning of ultrafine
622
cellulose acetate fibers: studies of a new solvent system and deacetylation of ultrafine
623
cellulose acetate fibers. Journal of Polymer Science Part B: Polymer Physics, 42, 5-11. 30
Page 31 of 33
ŠUTÝ, Š., PETRILÁKOVÁ, K., KATUŠČÁK, S., KIRSCHNEROVÁ, S., JABLONSKÝ, M.,
625
VIZÁROVÁ, K., & VRŠKA, M. (2012). Change in the capability of cellulose fibers to
626
retain water during thermally accelerated ageing of paper. Cellulose Chemistry and
627
Technology, 46, 631-635
ip t
624
Unnithan, A. R., Gnanasekaran, G., Sathishkumar, Y., Lee, Y. S., & Kim, C. S. (2014).
629
Electrospun antibacterial polyurethane–cellulose acetate–zein composite mats for wound
630
dressing. Carbohydrate Polymers, 102, 884-892.
us
cr
628
Wang, Q., Cai, J., Zhang, L., Xu, M., Cheng, H., Han, C. C., Kuga, S., Xiao, J., & Xiao, R.
632
(2013). A bioplastic with high strength constructed from a cellulose hydrogel by
633
changing the aggregated structure. Journal of Materials Chemistry A, 1, 6678-6686.
634
Wang, Y., & Chen, L. (2011). Impacts of nanowhisker on formation kinetics and properties of
M
d
all-cellulose composite gels. Carbohydrate Polymers, 83, 1937-1946.
Ac ce pt e
635
an
631
636
Wang, Y., & Chen, L. (2014). Cellulose nanowhiskers and fiber alignment greatly improve
637
mechanical properties of electrospun prolamin protein fibers. ACS Appl Mater Interfaces,
638
6, 1709-1718.
639 640 641 642
Weise, U., & Paulapuro, H. (1999). Effect of drying and rewetting cycles on fibre swelling. Journal of Pulp and Paper Science, 25, 163-166. Wu, M., Shuai, H., Cheng, Q., & Jiang, L. (2014). Bioinspired Green Composite Lotus Fibers. Angewandte Chemie International Edition, 53, 3358-3361.
31
Page 32 of 33
643
Yeh, T.-M., Wang, Z., Mahajan, D., Hsiao, B. S., & Chu, B. (2013). High flux ethanol
644
dehydration using nanofibrous membranes containing graphene oxide barrier layers.
645
Journal of Materials Chemistry A, 1(41), 12998-13003. Yoon, K., Hsiao, B. S., & Chu, B. (2009). High flux nanofiltration membranes based on
647
interfacially polymerized polyamide barrier layer on polyacrylonitrile nanofibrous
648
scaffolds. Journal of Membrane Science, 326, 484-492.
cr
ip t
646
Yoon, K., Kim, K., Wang, X., Fang, D., Hsiao, B. S., & Chu, B. (2006). High flux
650
ultrafiltration membranes based on electrospun nanofibrous PAN scaffolds and chitosan
651
coating. Polymer, 47, 2434-2441.
an
M
653
Zhang, Z., Zhang, Y.-W., & Gao, H. (2011). On optimal hierarchy of load-bearing biological materials. Proceedings of the Royal Society B: Biological Sciences, 278, 519–525.
d
652
us
649
Zhou, J., Chang, C., Zhang, R., & Zhang, L. (2007). Hydrogels prepared from unsubstituted
655
cellulose in NaOH/urea aqueous solution. Macromolecular Bioscience, 7, 804-809.
656
Zhou, Z., Peng, X., Zhong, L., Wu, L., Cao X., Sun R. (2016). Electrospun cellulose acetate
657
supported Ag@AgCl composites with facet-dependent photocatalytic properties on
658
degradation of organic dyes under visible-light irradiation. Carbohydrate Polymers, 136,
659
322–328.
660 661
Ac ce pt e
654
Zhou, J., Zhang, L., Cai, J., & Shu, H. (2002). Cellulose microporous membranes prepared from NaOH/urea aqueous solution. Journal of Membrane Science, 210, 77-90.
662 663 664 32
Page 33 of 33
S0144-8617(16)30051-0 http://dx.doi.org/doi:10.1016/j.carbpol.2016.02.011 CARP 10766
To appear in: Received date: Revised date: Accepted date:
24-10-2015 18-1-2016 2-2-2016
Please cite this article as: , Nanocomposite membranes showed rejection ratio of ¨ ¨/>pm
Highlights 1. All-cellulose nanocomposite membranes were fabricated as novel filtration system.
ip t
2. Hot pressed electrospun cellulose nanofabric provided mechanical support. 3. Regenerated cellulose gel coating with tiny inter-connected pores acted as barrier.
cr
4. Nanocomposite membranes showed rejection ratio of (98.68 ± 0.71)% against
Ac ce pt e
d
M
an
us
Hepatitis C Virus.
Page 1 of 33
4
Weijuan Huang a, Yixiang Wang a, Chao Chen b, John Lok Man Law b, Michael Houghton b,
5
a, Lingyun Chen *
6
a
7
AB, Canada T6G 2P5
8
b
9
University of Alberta, Edmonton, AB, Canada T6G 2E1
cr us
an
Department of Agricultural, Food & Nutritional Science, University of Alberta, Edmonton,
Li Ka Shing Institute of Virology, Department of Medical Microbiology and Immunology,
M
2
ip t
3
Fabrication of flexible self-standing all-cellulose nanofibrous composite membranes for virus removal
1
Corresponding Author
26
Corresponding author. Tel.: +1-780-492-0038; Fax: +1-780-492-8914.
27
Email address: [email protected] (L. Chen).
Ac ce pt e
d
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1
Page 2 of 33
28
ABSTRACT All-cellulose nanocomposite membranes with excellent performance were successfully
30
fabricated as novel filtration system to remove nanoparticles and virus from aqueous medium.
31
These membranes were composed of two combined layers: an electrospun cellulose
32
nanofabric layer treated by hot-pressing to provide mechanical support and a coating of
33
regenerated cellulose gel with tiny inter-connected pores as barrier. Hot-pressing didn’t affect
34
the fiber shape of electrospun nanofabrics, but significantly improved their mechanical
35
properties due to increased hydrogen bonds. The regenerated cellulose gel formed a porous
36
coating that tightly attached to electrospun nanofabrics, and its pore size varied depending on
37
cellulose source, solution concentration, and drying process. By assembling these two layers
38
together, the nanocomposite membranes showed the notable retention of negatively charged
39
100 nm latex beads (99.30%). Moreover, the electronegative nature of cellulose membranes
40
imparted the rejection ratio of 100% and (98.68±0.71)% against positively charged 50 nm
41
latex beads and Hepatitis C Virus, respectively.
cr
us
an
M
d
Ac ce pt e
42
ip t
29
43
KEYWORDS
44
All-cellulose ultrafiltration membranes, electrospun nanofiber, regenerated gel, flexible self-
45
standing, virus removal
46 47 48 49 50
2
Page 3 of 33
51
1. INTRODUCTION Virus contamination outbreak can often occur in medical or biotechnology products (e.g.,
53
vaccine, monoclonal/polyclonal antibody, plasma, immunoglobulin et al) (Levings &
54
Wessman, 1990). It presents a serious health hazard and results in large economic losses as
55
well as erosion of public trust. In order to reduce the risk of virus contamination, it is
56
necessary to purify biotechnology products. On the other hand, some researchers need to
57
enrich virus for specific experiments. However, the size of virus is extremely small; for
58
example, the swine influenza virus has a typical particle size of 80-120 nm in diameter
59
(Elford, Andrewes & Tang, 1936). It is difficult to separate virus from liquid media by simple
60
filtration. Nano-filtration is a convenient method to isolate small particles like virus (Asper,
61
Hanrieder, Quellmalz & Mihranyan, 2015; Quellmalz & Mihranyan, 2015; Rautenbach &
62
Gröschl, 1990), and has been used to convert sea water into drinking water by filtering salt
63
(Han, Xu & Gao, 2013). Nano-filtration removal or enrichment of virus is a promising
64
technology because it is non-destructive and non-interfering (Dishari, Micklin, Sung, Zydney,
65
Venkiteshwaran & Earley, 2015).
Ac ce pt e
d
M
an
us
cr
ip t
52
66
Recently, the thin film nanofibrous composite (TFNC) membrane consisting of an ultrathin
67
selective barrier layer (top layer), an electrospun nanofibrous scaffold (middle layer), and a
68
non-woven fabric support (bottom layer) has become popular in ultrafiltration systems (Kaur,
69
Barhate, Sundarrajan, Matsuura & Ramakrishna, 2011), because it not only can block
70
nanoscale substances but also has strong mechanical properties and high water flux (Ma et al.,
71
2010a). For example, poly(ethylene terephthalate) (PET) non-woven mats are usually
72
employed as the bottom layer to provide mechanical support, and electrospun
73
polyacrylonitrile (PAN) or polyvinyl alcohol (PVA) nanofibrous membranes constitute the
74
middle layer (Ma et al., 2010b; Yoon, Kim, Wang, Fang, Hsiao & Chu, 2006). Many studies
75
have focused on the fabrication of top barrier layer from a series of polymers, such as
76
cellulose regenerated from ionic liquids (Ma et al., 2010a), cellulose nanocrystals (Ma, 3
Page 4 of 33
Burger, Hsiao & Chu, 2014), chitin membrane (Ma, Hsiao & Chu, 2011), chemically cross-
78
linked PVA (Ma, Burger, Hsiao & Chu, 2012; Ma et al., 2010b), polyamide (Yoon, Hsiao &
79
Chu, 2009), and graphene oxide (Yeh, Wang, Mahajan, Hsiao & Chu, 2013). To the best of
80
our knowledge, TFNC membranes with only two layers that are completely developed from
81
natural polymers have never been reported.
ip t
77
Electrospining is a versatile and easy technology to fabricate nonwoven and continuous
83
nanofibers with diameters ranging from 50 nm to 500 nm by applying a high voltage (Kaur,
84
Sundarrajan, Rana, Matsuura & Ramakrishna, 2012; Liao et al., 2015; Unnithan,
85
Gnanasekaran, Sathishkumar, Lee & Kim, 2014). Electrospun nanofibrous membranes have
86
been widely used in filtration due to their highly porous network structure (Filatov, Budyka &
87
Kirichenko, 2007; Greiner & Wendorff, 2007). Among them, cellulose acetate has been
88
successfully electrospun and then deacetylated to prepare cellulose nanofabrics (Liu & Hsieh,
89
2002; Rodríguez, Renneckar & Gatenholm, 2011; Rodríguez, Sundberg, Gatenholm &
90
Renneckar, 2014; Son, Youk, Lee & Park, 2004; Zhou, Peng, Zhong, Wu, Cao, &Sun, 2016).
91
The obtained cellulose membranes exhibited superior filtration performance in aqueous
92
system due to their high porosity and hydrophilicity. Moreover, cellulose is biodegradable
93
and has good chemical resistance. It won’t react with majority of components in feed solution.
94
Until now, the most common filter membrane for aqueous system is made from cellulose.
95
However, the mechanical properties of electrospun nanofibrous cellulose membrane were
96
poor that limited their applications in filtration. It is reported that the hot pressed nanofibrous
97
membranes revealed better pressure tolerance and mechanical performance when compared to
98
untreated membranes (Kaur, Barhate, Sundarrajan, Matsuura & Ramakrishna, 2011; Wang et
99
al., 2013). Generally, electrospun nanofibers are randomly oriented and not interconnected at
100
joints. Hot press technique compresses the nanofibers together and generates new joints or
101
interactions under high temperature and pressure, thus enhances the structural integrity and
102
mechanical properties of electrospun nanofibrous membranes (Asper, Hanrieder, Quellmalz
Ac ce pt e
d
M
an
us
cr
82
4
Page 5 of 33
103
& Mihranyan, 2015; Lalia, Guillen-Burrieza, Arafat & Hashaikeh, 2013). Nevertheless, the
104
hot pressed electrospun cellulose nanofabrics have never been reported. Cotton and wood are good sources of cellulose because of the high yield and affordability
106
(Klemm, Heublein, Fink & Bohn, 2005). Canada is the largest world producer of newsprint
107
and northern bleached softwood kraft pulp, and the main component of paper and pulp is
108
wood-based cellulose (Kuhlberg, 2005). As a good cellulose solvent, the NaOH/urea aqueous
109
solution developed by Zhang’s group provides a ‘green’ and economical way to rapidly
110
dissolve cellulose at low temperature (Cai & Zhang, 2005). Recently, this technology has
111
been successfully adapted for dissolving wood cellulose after partial acidic hydrolysis (Gong,
112
Wang, Tian, Zheng & Chen, 2014). A series of functional cellulose materials have been
113
fabricated based on this solvent system, such as hydrogels (Zhou, Chang, Zhang & Zhang,
114
2007), microporous membranes (Zhou, Zhang, Cai & Shu, 2002), multifilament fibers (Cai et
115
al., 2007), and microspheres (Luo & Zhang, 2013). Particularly, cellulose solution can form
116
ultrathin gel membrane when being cast on a substrate (Zhou, Zhang, Cai & Shu, 2002). The
117
regenerated cellulose gel membrane exhibits a dense surface with interconnected nanoscale
118
pores (Wang & Chen, 2011). This unique structure provides the possibility to coat electrospun
119
cellulose nanofabrics with regenerated cellulose gel membrane to act as a selective barrier
120
layer to block virus and other microorganisms. Furthermore, the expected excellent
121
compatibility between two cellulose layers would be beneficial to increase water permeation
122
and separation efficiency (Lau, Ismail, Misdan & Kassim, 2012). The two-layer all-cellulose
123
nanofibrous composite membranes were fabricated in current study, where the hot pressing
124
treated electrospun cellulose nanofiber provided the mechanical support and the regenerated
125
cellulose gel coating worked as the separation layer. The structure, mechanical properties and
126
filtration performance of composite membranes were investigated, and the retention rate of
127
Hepatitis C Virus (HCV) was tested to evaluate the practical and functional performance.
Ac ce pt e
d
M
an
us
cr
ip t
105
128 5
Page 6 of 33
129
2. EXPERIMENTAL METHODS
130
2.1 Materials Spruce cellulose (bleached kraft pulp) with -cellulose content of 87.3% was provided by
132
Alberta-Pacific Forest Industries Inc. (AB, Canada). It was hydrolyzed by 20 wt% sulfuric
133
o 4 acid at 30 C for 24 h and the viscosity-average molecular weight (M ) was 5.8 × 10 (Gong,
134
Wang, Tian, Zheng & Chen, 2014). Cotton cellulose (cotton linter pulp) with M of 1.0 × 105
135
was supplied by Hubei Chemical Fiber Group, Ltd. (Xiangfan, China). Cellulose acetate (CA,
136
average MN Ca. 30 000, 39.8 wt% acetyl content) and fluorophore tagged polystyrene latex
137
beads (L9902, 100nm in diameter, sulfonate-modified; fluorescent red; ex~575 nm; em~610
138
nm and L0780, 50nm mean particle size, amine-modified, fluorescent blue, ex~360 nm;
139
em~420) were purchased from Sigma-Aldrich Canada Ltd. (Oakville, ON, Canada).
140
Commercial cellulose filter paper (CM) was purchased from WhatmanTM (GE Healthcare,
141
Buckinghamshire, UK). Dulbecco’s Modified Eagle Medium (DMEM), Non-Essential Amino
142
Acids (NEAA), penicillin and streptomycin were obtained from Life Technologies
143
(Burlington, ON, Canada). Acetic acid and all other chemical reagents were purchased from
144
Fisher Scientific (Markham, ON, Canada) and were used as received unless otherwise
145
described.
cr
us
an
M
d
Ac ce pt e
146
ip t
131
147
2.2 Preparation of all-cellulose nanofibrous membranes
148
2.2.1 Hot pressed electrospun cellulose nanofabrics
149
Nonwoven cellulose acetate nanofabrics were fabricated by a customized digital
150
electrospinning apparatus EC-DIG (IME Technologies, Eindhoven, Netherlands) at room
151
temperature (22 oC). Briefly, 8 g cellulose acetate was dissolved in 42 mL acetic acid/water
152
(75/25, v/v) solution, and then it was forced through a blunt needle with a diameter of 0.8 mm
153
at the rate of 1 mL h-1. The applied voltage was fixed at 23 kV. A rotating drum with a
154
diameter of 10 cm was chosen as the collector, and the distance between the tip and collector 6
Page 7 of 33
was set as 15 cm. The obtained cellulose acetate mats were subsequently immersed in 0.5 M
156
KOH ethanol solution at room temperature for 1 h to generate cellulose nanofabrics. The
157
cellulose nanofabrics were washed with excess deionized water and cut into square pieces (3
158
inch×3 inch). Several square pieces (5, 10 and 15 pieces) were piled up and sandwiched
159
between two plane white PTFE plates. They were pre-heated at 110 oC for 10 min and then
160
hot pressed by a Carver benchtop laboratory press (model 3851, Carver Inc., Wabash, IN)
161
under the pressure of about 7.66 MPa (10 000 pounds/3 inch×3 inch) at 110 oC for 50 min.
162
The resultant cellulose nanofabrics were coded as L5, L10 and L15, corresponding to the
163
different number of pieces used to pile up the nanofabrics. The cellulose sample without hot
164
pressing was coded as RC.
165
2.2.2 Regenerated cellulose gel coating
an
us
cr
ip t
155
Cotton cellulose (CC) and hydrolyzed spruce cellulose (SC) solutions with cellulose
167
concentration of 1, 2 and 3 wt% were prepared as described by Cai and Zhang (Cai & Zhang,
168
2005). Briefly, 7wt% NaOH/12wt% urea aqueous solution was precooled to -12.6 oC, and
169
desired amount of cellulose was added in the solution with mechanical stirring at 2000 rpm
170
for 3 min. The cellulose solutions were degassed by centrifugation at 805 g and 4 oC for 5 min.
171
To form the regenerated cellulose gel coating, the hot pressed electrospun cellulose
172
nanofabric L5 was firstly soaked in 75% acetic acid aqueous solution for 1 min, and then its
173
one face was quickly dipped in the above prepared cellulose solutions to obtain the one-side
174
coated all-cellulose nanofibrous composite membrane. The resultant samples were either
175
directly washed using deionized water, or immersed in pure ethanol for 1 h (solvent exchange
Ac ce pt e
d
M
166
7
Page 8 of 33
cr us
and then thoroughly washed. All the membranes were dried at room temperature and pre-immersed in water before the filtration test. The
177
detailed preparation processing parameters of each sample is shown in Table 1.
178
Table 1. Components, processing methods, and ultrafiltration performance of all-cellulose nanofibrous composite membranes.
M
an
176
Sample
Supporting layer
1CC-L5 2CC-L5
Hot pressed five-layer electrospun cellulose nanofabrics
Ac
2SC-SE-L5
ce
2SC-L5
3SC-SE-L5
180
Coating layer Cellulose Cellulose source concentrate Hydrolyzed spruce 1wt % cellulose Hydrolyzed spruce 2wt % cellulose Cotton cellulose 1wt % Cotton cellulose 2wt % Hydrolyzed spruce 2wt % cellulose Hydrolyzed spruce 3wt % cellulose Cotton cellulose 2wt % Cotton cellulose 3wt %
pt
1SC-L5
ed
179
2CC-SE-L5 3CC-SE-L5 n.d. means not detect.
Rejection ratio / %
Pure water flux / L m-2 h-1
100 nm
50 nm
No
189.87±55.35
77.14±11.51
n.d.
No
93.99±22.51
84.59±4.97
96.84±0.52
No No
130.08±44.75 48.73±1.93
54.86±15.12 82.35±1.62
n.d. 96.14±0.16
Yes
131.20±11.00
92.41±2.15
99.02±2.02
Yes
97.75±5.84
97.79±1.00
99.88±0.11
Yes Yes
113.42±1.76 89.47±2.96
95.56±2.35 99.30±0.66
99.65±0.20 100±0.08
Solvent exchange
181 182
8
Page 9 of 33
183
2.3 Structure and morphology Fourier transform infrared (FT-IR) spectra of electrospun cellulose acetate fibers and the
185
deacetylated cellulose fibers with/without hot pressing treatment were recorded on a Nicolet
186
6700 spectrophotometer (Thermo Fisher Scientific Inc., MA, USA) with KBr pellets. The
187
samples were vacuum-dried for 24 h prior to test. Spectra were recorded as the average of 32
188
scans at 4 cm-1 resolution at room temperature. During measurements the accessory
189
compartment was flushed with dry air.
cr
ip t
184
Morphology observation of electrospun cellulose fibers and regenerated gel coating were
191
carried out with a Philips XL-30 scanning electron microscope (SEM) at an acceleration
192
voltage of 5-10 kV. The all-cellulose membranes were frozen in liquid nitrogen, snapped
193
immediately and then freeze-dried to obtain the cross-sectional fracture surface. The samples
194
were sputtered with gold for 2 min prior to observation and photographing. In SEM photos,
195
fiber diameters were determined with the ImageJ image-visualization software developed by
196
the National Institute of Health (He et al., 2014). Two hundred random positions were
197
selected and measured for each sample.
199
an
M
d
Ac ce pt e
198
us
190
2.4 Mechanical properties
200
Tensile testing of the hot pressed electrospun cellulose nanofabrics was done using an
201
Instron 5967 universal testing machine (Instron Corp., MA, USA) at a crosshead speed of 4
202
-1 mm min and a gauge length of 20 mm according to the ASTM D-638-V standard (He et al.,
203
2014; Wang & Chen, 2014). Five bars with a dimension of 5 cm × 1 cm (length × width) were
204
cut from each fabric membrane, and their thickness was measured by a digital micrometer
205
(Mitutoyo, Japan) with a precision of 1 μm. Before testing, the samples were either vacuum-
206
dried or immersed in water for 24 h.
207 208
2.5 Ultrafiltration performance 9
Page 10 of 33
Pure water flux was tested in a Büchner funnel. All-cellulose nanofibrous composite
210
membranes with an effective filtration area of 5.3 cm2 (26 mm in diameter) were placed in the
211
funnel and 50 mL ultrapure water were used to flow through the membranes. The
212
performance pressure was 10 kPa. The pure water flux was determined by the following
213
equation:
214
Jw=Q/(A* t)
(1)
ip t
209
-2 -1 Where Jw is the pure water flux (L m h ), Q is the quantity of permeation (L), A is the
216
effective membrane area (m2), and t is the permeation time (h).
cr
215
For the particle retention test, 5 μL polystyrene latex bead suspensions were diluted to 10
218
mL with ultrapure water. Afterwards, the diluent was filtered through the all-cellulose
219
nanofibrous membranes under the vacuum pressure of 10 kPa. The fluorescence intensity of
220
the diluent, filtrate and ultrapure water was measured by a SpectraMax M3 microplate reader
221
with SpectraMax Pro Software (Molecular Devices, Inc., USA) at the specified excitation and
222
emission wavelengths. The zeta-potential of 50 nm and 100 nm latex beads dispersed in pure
223
water was measured by a Malvern zeta-sizer (Malvern instruments Inc., UK).
Ac ce pt e
d
M
an
us
217
224
A CLSM 710 Meta confocal laser scanning microscope (Carl Zeiss, Jena, Germany) was
225
used for cross-sectional view of all-cellulose nanofibrous composite membranes after
226
filtration of 100 nm polystyrene latex beads, and the images were recorded using ZEN 2009
227
LE software (Carl Zeiss MicroImaging GmbH, Germany).
228 229
2.6 Retention of Hepatitis C Virus
230
Huh7.5 cells were cultured in DMEM supplemented with 10% FBS, 0.1 mM NEAA and
231
100 μg each of penicillin and streptomycin. Cell culture derived Hepatitis C Virus (HCVcc)
232
was produced using previously described protocol (Lindenbach et al., 2005). Cells were
233
7 washed twice with ice cold PBS and subsequently resuspended to 1.5×10 cells/mL. Then,
234
400 μL of the cell suspension were mixed with 5 μg in vitro transcribed RNA encoding HCV 10
Page 11 of 33
genome in 2 mm gap electroporation curvettes (Bio-Rad, Mississauga, ON, Canada), and 5
236
pulses of 860 V (99 μs, 1.1 s interval) were delivered using the ElectroSquare Porator ECM
237
830 (BTX, Holliston, MA). Post-electroporation, cells were incubated at room temperature
238
for 10 min before plating. Pre-cleared media was collected as virus stocks either 3 or 4 d post-
239
electroporation. The virus titer (50% tissue culture infectious dose (TCID50)) was determined
240
by limited dilution as described previously (Lindenbach et al., 2005).
ip t
235
HCV-JC1 viral stock (5.13×104 TCID50 HCVcc/mL,) was filtered through a sterile
242
membrane in a Bu•chner funnel (Fisher Scientific, Ottawa, ON, Canada) under the vacuum
243
pressure of 10 kPa. Pre- and post-filtration viruses were assessed by NS5A staining. For the
244
NS5A (Nonstructural Protein 5A) staining, 200 μL pre- or post-filtration viruses were diluted
245
to 686 μL and 100 μL dilutions were inoculated to Huh7.5 cells pre-seeded in a 96 well plate
246
(BD, Mississauga, ON, Canada) in quadruplicate. After the 12 h incubation, the viral
247
inoculum was replaced with fresh culture media. Then, 48 h after inoculation, cells were fixed
248
and stained as described previously (Law et al., 2013). The foci per well were detected and
249
counted using a CTL S6 immunospot analyzer (CTL, Cleveland OH) as described previously
250
(Gottwein et al., 2010).
252
us
an
M
d
Ac ce pt e
251
cr
241
2.7 Statistical analysis
253
Experimental results were represented as the mean ± SD. Statistical evaluation was carried
254
out by analysis of variance (ANOVA) followed by multiple-comparison tests using Duncan’s
255
multiple-range test at the 95% confidence level. All of the analyses were conducted using
256
SAS statistical software (SAS Institute, Inc., Cary, NC) with a probability of p < 0.05
257
considered to be significant.
258
3. RESULTS AND DISCUSSION
259
3.1 Structure and mechanical properties of hot pressed electrospun fabrics
11
Page 12 of 33
260
The molecular structure changes of electrospun cellulose acetate (CA) fibers after
261
deacetylation and hot pressing were recorded by FT-IR. As shown in Fig. 1, the electrospun
262
-1 CA fibers exhibited several characteristic absorption peaks at 3496 cm (
263
1375 cm-1 (
264
the characteristic absorption peaks of acetyl group disappeared, indicating CA had been
265
completely converted to cellulose (Ma, Kotaki & Ramakrishna, 2005). At the same time, the
266
absorption peak of
267
3445cm-1, suggesting the improvement of hydrogen bonds due to the recovered hydroxyl
268
groups in cellulose (Liu & Hsieh, 2002),. The
269
broadened, indicating that the stronger hydrogen bonding interactions were developed during
270
hot pressing treatment (Figueiredo, Evtuguin & Saraiva, 2010; Luo, Zhu, Gleisner & Zhan,
271
2011; ŠUTÝ et al., 2012). This enhanced hydrogen bonding could induce the hornification of
272
cellulose and result in the improved dimensional stability (Luo, Zhu, Gleisner & Zhan, 2011;
273
Weise & Paulapuro, 1999).
) (Son, Youk, Lee & Park, 2004). After deacetylation,
cr
became broader and stronger and shifted to lower wavenumber of
peak of hot pressed cellulose fabric further
us
OH
Ac ce pt e
d
M
an
OH
C-O-C
),
C=O
ip t
), and 1237 cm-1 (
C-CH3
), 1754 cm-1 (
OH
274
12
Page 13 of 33
Fig. 1. FT-IR spectra of (a) electrospun cellulose acetate nanofabrics (CA), (b) deacetylated
276
electrospun cellulose nanofabrics (RC), and (c) hot pressed five-layer electrospun
277
cellulose nanofabrics (L5).
Ac ce pt e
d
M
an
us
cr
ip t
275
278 279
Fig. 2. SEM images and diameter distribution of (a) electrospun cellulose acetate nanofabrics
280
(CA), (b) deacetylated electrospun cellulose nanofabrics (RC), and (c) hot pressed 13
Page 14 of 33
281
five-layer electrospun cellulose nanofabrics (L5).
The morphology and width distribution of electrospun CA fibers and cellulose fibers with or
284
without hot pressing are shown in Fig. 2. Uniform CA nanofibers were obtained with flat
285
ribbon-like shape. Similar fibers were reported by Han and co-workers (Han, Youk, Min,
286
Kang & Park, 2008). This was because a thin skin formed on the solution jet and acted as the
287
tube wall. Due to the continuous evaporation of inside solvent, the atmospheric pressure made
288
the tubes collapse and resulted in the elliptical ribbons, which finally became the flat ribbons
289
(Koombhongse, Liu & Reneker, 2001; Ramakrishna, Fujihara, Teo, Lim & Ma, 2005). The
290
hydrolyzed cellulose nanofibers exhibited the same shape and similar width compared to
291
electrospun CA fibers, indicating that the hydrolysis didn’t affect the fiber morphology. It was
292
worth noting the nanoscale fiber shape and porous network structure still well maintained
293
even after hot pressing. This feature could enable the high water flux during filtration. No
294
fusion appeared in the node or intersection of overlapped fibers because cellulose doesn’t
295
melt or carbonize at 110 oC. The width of cellulose nanofibers slightly increased from
296
213±116 nm to 260±130 nm, since the fibers were wet and swelled to some extent prior to hot
297
pressing.
Ac ce pt e
d
M
an
us
cr
ip t
282 283
298 299
Fig. 3. Typical stress-strain curves of membranes at (a) dry status (inset, photo of hot-pressed 14
Page 15 of 33
300
electorspun cellulose nanofabric (L5)), and (b) wet status (inset, photo of hot-pressed
301
electorspun cellulose nanofabric (L5) immersed in water for 24h). Normally, filter membranes are used in liquid media and should be strong enough to stand
303
the pressure. The mechanical strength of electrospun nanofabrics was thus tested to evaluate
304
their handling property. Typical stress-strain curves of cellulose nanofabrics with or without
305
hot pressing (L5, L10, L15, and RC) are shown in Fig. 3. Their tensile strength, elongation at
306
break and Young’s modulus are summarized in Table S1. The mechanical properties of
307
commercial filter paper (CM, Whatman™ 1002-125 Grade 2 Qualitative Filter Paper) were
308
also measured for comparison. The tensile strength of RC was 6.9±1.6 MPa. After hot
309
pressing, the L5 fabric exhibited a greatly improved strength of 11.7±2.0 MPa due to the
310
enhanced hydrogen bonding interactions. However, the tensile strength of L10 and L15
311
samples decreased to 10.4±1.8 and 6.8±3.1 MPa, respectively. It was noticed that the
312
thickness of the fabrics increased when more electrospun cellulose pieces were hot pressed
313
together. The compression force was thus diffused and led to the formation of relatively
314
looser interlayers which affected the efficient stress transfer (Wu, Shuai, Cheng & Jiang, 2014;
315
Zhang, Zhang & Gao, 2011). The mechanical properties of L5 at dry status were only close to
316
those of commercial filter paper, but L5 was three times stronger than commercial product
317
when they were immersed in water. The tensile strength of wet L5 and CM was 1.5±0.1MPa
318
and 0.5±0.1 MPa, respectively, and their elongation at break was 12.6±1.1% and 3.4±0.2%,
319
respectively. It was the result of the developed hydrogen bonding interactions in electrospun
320
cellulose nanofibers treated by hot pressing, which restricted the swelling of cellulose (Weise
321
& Paulapuro, 1999). The photographs of L5 at dry and wet status are inserted in Fig. 3a and
322
Fig. 3b, respectively. It exhibited good flexibility and remained integrated after immersing in
323
water for 24 h. These results indicated that the L5 nanofabric was suitable as the support layer
324
of composite ultrafiltration system and provided sufficient handling properties.
Ac ce pt e
d
M
an
us
cr
ip t
302
325 15
Page 16 of 33
326 3.2 Morphology of regenerated cellulose gel coating
328
Ac ce pt e
d
M
an
us
cr
ip t
327
329
Fig. 4．SEM images of cross-section (left) and surface (right) of all-cellulose nanofibrous
330
composite membranes: (a, b) 1SC-L5, (c, d) 2SC-L5, (e, f) 1CC-L5, and (g, h) 2CC-
331
L5.
332 16
Page 17 of 33
ip t cr us an M d Ac ce pt e
333 334
Fig. 5．SEM images of cross-section (left) and surface (right) of all-cellulose nanofibrous
335
composite membranes: (a, b) 2SC-SE-L5, (c, d) 3SC-SE-L5, (e, f) 2CC-SE-L5, and (g,
336
h) 3CC-SE-L5.
337 338
17
Page 18 of 33
Both cotton cellulose (CC) and hydrolyzed spruce cellulose (SC) were chosen as the
340
coating materials to investigate the effects of cellulose source and concentration as well as
341
drying procedure on the morphology of gel coatings. Fig. 4 shows SEM images of cross-
342
section and surface of composite cellulose nanofibrous membranes prepared by directly
343
drying at room temperature. A thin cellulose top layer with porous structure existed in all the
344
samples. No obvious gap was observed between electrospun cellulose nanofabric and
345
cellulose gel coating. Such good compatibility could be beneficial to increase pure water flux
346
and filtration efficiency. The gel coatings formed by 1% cellulose solutions were too thin to
347
be distinguished (ca. 40 μm) and exhibited relatively large pores on the surface. Dense
348
coating layers with the thickness of about 130 μm were formed in both 2SC-L5 and 2CC-L5
349
samples when cellulose concentration was increased to 2%. The morphology of cross-section
350
and surface of cellulose coating layers made by solvent exchange is shown in Fig. 5.
351
Compared to cellulose coating layers made by directly drying, the coatings generated by
352
solvent exchange and higher cellulose concentration showed relatively larger thickness (150-
353
300 μm) but uniform structures with more and smaller pores. During the directly drying
354
process, free water evaporated to induce a microporous structure that collapsed to certain
355
extent, causing membrane shrinking and cracking. Whereas during solvent exchange process,
356
the water within cellulose gel was replaced by volatile fluid with lower surface tension.
357
According to previous report (Jie, Cao, Qin, Liu & Yuan, 2005), the change of cellulose
358
coating morphology during various drying processes was primarily ascribed to the molecular
359
affinities of cellulose-ethanol, cellulose-water, and water-ethanol. And the molecular affinity
360
of cellulose-ethanol is weaker than that of cellulose-water. So the evaporation of ethanol
361
could maintain the uniform porous structure of cellulose gel (Jawad, Ahmad, Low, Chew &
362
Zein, 2015). The pore size further decreased with raising cellulose solution concentration
363
from 2% to 3%. It was reported that cellulose molecular chain entanglement took place during
364
the regeneration which resulted in the formation of porous gel network with dense surface and
Ac ce pt e
d
M
an
us
cr
ip t
339
18
Page 19 of 33
large inner holes (Wang & Chen, 2011). Thus, smaller pores were generated when the
366
cellulose concentration was higher. These nanoscale pores on the surface provided the
367
possibility to block tiny material during the filtration while the inside large pores ensured the
368
good permeability of water (as shown in Fig. 5).
369
3.3 Ultrafiltration performance of all-cellulose membranes
370
3.3.1 Permeability
ip t
365
The permeability of TFNC membrane is mainly determined by the morphology and
372
thickness of the top barrier layer (Ma et al., 2010a). The pure water flux of all-cellulose
373
membranes was measured and the values were listed in Table 1. The 1SC-L5 and 1CC-L5
374
membranes exhibited the highest water permeability of 189.87±55.35 L m-2 h-1 and
375
130.08±44.75 L m-2 h-1, respectively, due to their very thin coatings. The water flux of 2CC-
376
L5 (48.73±1.93 L m-2 h-1) was obviously lower than that of 2SC-L5 (93.99±22.51 L m-2 h-1),
377
since more pores were generated in the matrix-filler structure of 2SC-L5. The pure water flux
378
of 2SC-SE-L5 and 2CC-SE-L5 were 131.20±11.00 L m-2 h-1 and 113.42±1.76 L m-2 h-1,
379
respectively, which was higher than that of 2SC-L5 and 2CC-L5. It indicated that the solvent
380
exchange process produced more through-pores in cellulose coating layer compared to direct
381
drying at room temperature. Usually, a good filter membrane should have high filtration
382
efficiency and high pure water flux to save time and energy. The all-cellulose nanofibrous
383
composite membranes showed much higher permeability than TFNC membranes composed
384
of a 0.3 μm cotton cellulose layer regenerated from ionic liquid, a polyacrylonitrile
385
nanofibrous scaffold and a melt-blown poly (ethylene terephthalate) (PET) non-woven
386
substrate (28.0 L m-2 h-1 under 10 kPa and 73.7 L m-2 h-1 under 10 psi) (Ma et al., 2010a).
387
Moreover, the dimension of the all-cellulose membranes didn’t change after filtration. The
388
result indicated that cellulose gel membranes prepared from NaOH/urea aqueous solution had
389
promising filtration efficiency.
390
3.3.2 Nanoparticle retention test
Ac ce pt e
d
M
an
us
cr
371
19
Page 20 of 33
The rejection ratio of all-cellulose nanofibrous composite membranes was evaluated by the
392
retention test of fluorescence-labeled polystyrene latex beads. Two kinds of latex beads were
393
employed where the 100 nm beads were modified by electronegative sulfate group and the 50
394
nm ones were modified by positively charged amino group. As shown in Table 1 and Fig. S1,
395
for 100 nm beads, 1SC-L5 and 1CC-L5 exhibited the low rejection ratio of 77.14±11.51%
396
and 54.86±15.12%, respectively, due to their thin cellulose gel coatings and large pores on the
397
coating surface. With the increase of cellulose concentration, more beads were blocked by the
398
regenerated cellulose coatings, and the rejection ratio of 2SC-L5 and 2CC-L5 was
399
84.59±4.97% and 82.35±1.62%. The retention of 50 nm latex beads was even greater and ca.
400
97% beads were blocked by 2SC-L5. The rejection ratio of cellulose coating layer prepared
401
by solvent exchange was much higher than that of directly dried coating. For 100 nm beads,
402
the rejection ratio of 2SC-SE-L5 and 2CC-SE-L5 was 92.41±2.15% and 95.56±2.35%,
403
respectively, and it was even higher with increasing the cellulose concentration. The 3SC-SE-
404
L5 and 3CC-SE-L5 samples could block most beads because of their small surface pore sizes.
405
For 50 nm latex beads, the rejection ratio of 2SC-SE-L5, 2CC-SE-L5, 3SC-SE-L5 and 3CC-
406
SE-L5 was all above 99%. Figs. 6a and b show the morphology of 2SC-L5 after the filtration.
407
Both the 100 nm and 50 nm latex beads were successfully trapped on the surface of cellulose
408
coating. The improved retention ratio to 50 nm beads should be related to the surface charge.
409
As tested, the surface charge of 100 nm and 50 nm latex beads were -54.8±0.96 mV and
410
+25.2±0.62 mV, respectively. At the same time, the cellulose membrane presented the
411
negative zeta potential because of the abundant hydroxyl groups (Ma, Burger, Hsiao & Chu,
412
2011). The attractive interactions existed between 50 nm latex beads and filtration membrane
413
and resulted in the better capture capacity. In order to investigate the distribution of latex
414
beads in all-cellulose membrane after filtration, the confocal scanning microscopy image is
415
shown in Fig. 6c. Most latex beads located on the top surface of cellulose coating layer and a
416
few beads accumulated at the interface of two layers. However, no latex beads were observed
Ac ce pt e
d
M
an
us
cr
ip t
391
20
Page 21 of 33
within the gel network of cellulose coating. It could be supposed that a dense gel structure
418
also formed at the interface of electrospun nanofabric and cellulose coating due to the rapid
419
regeneration of cellulose solution caused by the direct contact of acetic acid. Therefore, a
420
large rejection ratio was achieved by the dense surface/interface of regenerated cellulose gel
421
coating while its porous inside structure enabled the large pure water flux.
us
cr
ip t
417
an
422
Fig. 6. SEM images of all-cellulose nanofibrous composite membrane (2SC-L5) after the
424
retention test: (a) 100 nm and (b) 50 nm latex beads. (c) Confocal microscopic image
425
of cross-sectional view of cellulose coating layer after filtration of 100nm
426
fluorochrome tagged polystyrene latex beads.
428 429 430 431 432 433 434
d
Ac ce pt e
427
M
423
3.3.3 Virus retention test
21
Page 22 of 33
ip t cr us 438
M
an 437
Fig. 7. Focus forming units of Hepatitis C Virus (HCV) solution before and after filtration
d
436
through 2SC-L5 and 2CC-SE-L5 all-cellulose nanofibrous composite membrane.
Ac ce pt e
435
439
3SC-SE-L5 and 3CC-SE-L5 samples exhibited an unstable dimension during drying
440
because of the excessive shrinkage of thick cellulose coatings. Thus, taking both rejection
441
ratio and pure water flux into consideration, 2SC-L5 and 2CC-SE-L5 membranes were
442
selected for the virus retention test. The Hepatitis C Virus (HCV) was chosen as the model
443
that has a spherical shape with positive surface charge and the diameter of 55-65 nm (Kaito et
444
al., 1994; Shimizu, Feinstone, Kohara, Purcell & Yoshikura, 1996). Fig. 7 shows the focus
445
forming units (FFU) in the virus solution before and after filtration. The FFU value treated by
446
2SC-L5 decreased from 183.75±10.80 to 14.75±3.03, indicating that about 92% HCV was
447
removed from the solution. For 2CC-SE-L5 membranes, the rejection ratio increased to
448
(98.68±0.71)%. However, the commercial microfiltration membrane GS0.22 (a mixed 22
Page 23 of 33
449
cellulose esters membrane with 0.22 μm pore size) showed only 90% retention of MS2
450
bacteriophage (Ma, Hsiao & Chu, 2014). We also notice that the filtration performance of
451
2CC-SE-L5 membranes was not sufficient for the requirement from water purification
452
industry, and further modifications were needed to improve this all-cellulose filtration system.
454
ip t
453
4. CONCLUSION
This is the first documentation of all-cellulose nanofibrous composite membrane made by
456
assembling an electrospun cellulose nanofabric layer as the mechanical framework and a
457
coating of regenerated cellulose gel membrane as the barrier. The hot pressing treatment led
458
to enhanced hydrogen bonding interactions, which significantly improved the mechanical
459
properties of supporting layer. The resultant nanofabric possessed a three-fold stronger wet
460
strength compared to commercial filter paper, while maintaining its porous structure. The
461
cellulose gel membrane regenerated from NaOH/urea aqueous solution was well attached on
462
the electrospun cellulose nanofabric. This barrier layer exhibited a unique structure of dense
463
surface/interface and large interior pores, which not only blocked the virus and nanoparticles
464
in water, but also ensured a large water flux. The retention tests indicated that these
465
membranes could remove beads as small as 50 nm. Finally, the rejection ratio of 2CC-SE-L5
466
against HCV was (98.68±0.71)% suggesting its potential applications in virus removal.
468 469
us
an
M
d
Ac ce pt e
467
cr
455
470 471 472
ACKNOWLEDGMENT
473
The authors are grateful to the Natural Sciences and Engineering Research Council of Canada
474
(NSERC), Alberta Crop Industry Development Fund Ltd. (ACIDF), Alberta Innovates Bio 23
Page 24 of 33
Solutions (AI Bio) and Alberta Barley Commission for financial support as well as Canada
476
Foundation for Innovation (CFI) for equipment support. Lingyun Chen would like to thank
477
the Natural Sciences and Engineering Research Council of Canada (NSERC)-Canada
478
Research Chairs Program for its financial support. Weijuan Huang thanks the support from
479
China Scholarship Council (CSC).
ip t
475
480
cr
481 482
us
483
an
484 485
M
486 487
d
488
490 491 492 493 494 495
Ac ce pt e
489
496 497
REFERENCES
498
Asper, M., Hanrieder, T., Quellmalz, A., & Mihranyan, A. (2015). Removal of xenotropic
499
murine leukemia virus by nanocellulose based filter paper. Biologicals, 43(6), 452-456. 24
Page 25 of 33
500 501
Cai, J., & Zhang, L. (2005). Rapid dissolution of cellulose in LiOH/urea and NaOH/urea aqueous solutions. Macromolecular Bioscience, 5, 539-548. Cai, J., Zhang, L., Zhou, J., Qi, H., Chen, H., Kondo, T., Chen, X., & Chu, B. (2007).
503
Multifilament fibers based on dissolution of cellulose in NaOH/urea aqueous solution:
504
structure and properties. Advanced Materials, 19, 821-825.
ip t
502
Dishari, S. K., Micklin, M. R., Sung, K. J., Zydney, A. L., Venkiteshwaran, A., & Earley, J. N.
506
(2015). Effects of solution conditions on virus retention by the Viresolve® NFP filter.
507
Biotechnology Progress, 31, 1280-1286.
an
us
cr
505
Elford, W., Andrewes, C., & Tang, F. (1936). The Sizes of the Viruses of Human and Swine
509
Influenza, as Determined by Ultra-Filtration. British Journal of Experimental Pathology,
510
17, 51-53.
512
d
Figueiredo, A., Evtuguin, D., & Saraiva, J. (2010). Effect of high pressure treatment on
Ac ce pt e
511
M
508
structure and properties of cellulose in eucalypt pulps. Cellulose, 17, 1193-1202.
513
Filatov, Y., Budyka, A., & Kirichenko, V. (2007). Electrospinning of micro-and nanofibers:
514
fundamentals in separation and filtration processes. Journal of Engineered Fibers and
515
Fabrics, 3, 488.
516 517
Gong, X., Wang, Y., Tian, Z., Zheng, X., & Chen, L. (2014). Controlled production of spruce cellulose gels using an environmentally “green” system. Cellulose, 21, 1667-1678.
518
Gottwein, J. M., Scheel, T. K., Callendret, B., Li, Y.-P., Eccleston, H. B., Engle, R. E.,
519
Govindarajan, S., Satterfield, W., Purcell, R. H., & Walker, C. M. (2010). Novel
520
infectious cDNA clones of hepatitis C virus genotype 3a (strain S52) and 4a (strain 25
Page 26 of 33
521
ED43): genetic analyses and in vivo pathogenesis studies. Journal of Virology, 84, 5277-
522
5293. Greiner, A., & Wendorff, J. H. (2007). Electrospinning: a fascinating method for the
524
preparation of ultrathin fibers. Angewandte Chemie International Edition, 46, 5670-5703.
525
Han, S. O., Youk, J. H., Min, K. D., Kang, Y. O., & Park, W. H. (2008). Electrospinning of
526
cellulose acetate nanofibers using a mixed solvent of acetic acid/water: Effects of solvent
527
composition on the fiber diameter. Materials Letters, 62, 759-762.
529
cr
us
Han, Y., Xu, Z., & Gao, C. (2013). Ultrathin graphene nanofiltration membrane for water
an
528
ip t
523
purification. Advanced Functional Materials, 23, 3693-3700. He, X., Xiao, Q., Lu, C., Wang, Y., Zhang, X., Zhao, J., Zhang, W., Zhang, X., & Deng, Y.
531
(2014). Uniaxially Aligned Electrospun All-Cellulose Nanocomposite Nanofibers
532
Reinforced
533
Biomacromolecules, 15, 618-627.
Cellulose
Nanocrystals:
Scaffold
for
Tissue
Engineering.
Ac ce pt e
with
d
M
530
534
Jawad, Z., Ahmad, A., Low, S., Chew, T., & Zein, S. (2015). Influence of solvent exchange
535
time on mixed matrix membrane separation performance for CO2/N2 and a kinetic
536
sorption study. Journal of Membrane Science, 476, 590-601.
537
Jie, X., Cao, Y., Qin, J.-J., Liu, J., & Yuan, Q. (2005). Influence of drying method on
538
morphology and properties of asymmetric cellulose hollow fiber membrane. Journal of
539
Membrane Science, 246, 157-165.
540
Kaito, M., Watanabe, S., Tsukiyama-Kohara, K., Yamaguchi, K., Kobayashi, Y., Konishi, M.,
541
Yokoi, M., Ishida, S., Suzuki, S., & Kohara, M. (1994). Hepatitis C virus particle 26
Page 27 of 33
542
detected by immunoelectron microscopic study. The Journal of general virology, 75,
543
1755-1760. Kaur, S., Barhate, R., Sundarrajan, S., Matsuura, T., & Ramakrishna, S. (2011). Hot pressing
545
of electrospun membrane composite and its influence on separation performance on thin
546
film composite nanofiltration membrane. Desalination, 279, 201-209.
ip t
544
Kaur, S., Sundarrajan, S., Rana, D., Matsuura, T., & Ramakrishna, S. (2012). Influence of
548
electrospun fiber size on the separation efficiency of thin film nanofiltration composite
549
membrane. Journal of Membrane Science, 392, 101-111.
an
us
cr
547
Klemm, D., Heublein, B., Fink, H. P., & Bohn, A. (2005). Cellulose: fascinating biopolymer
551
and sustainable raw material. Angewandte Chemie International Edition, 44, 3358-3393.
552
Koombhongse, S., Liu, W., & Reneker, D. H. (2001). Flat polymer ribbons and other shapes
553
by electrospinning. Journal of Polymer Science Part B: Polymer Physics, 39(21), 2598-
554
2606.
556 557
d
Ac ce pt e
555
M
550
Kuhlberg, M. (2005). (http://www.thecanadianencyclopedia.ca/en/article/pulp-and-paperindustry/).
Lalia, B. S., Guillen-Burrieza, E., Arafat, H. A., & Hashaikeh, R. (2013). Fabrication and
558
characterization
559
electrospun membranes for direct contact membrane distillation. Journal of Membrane
560
Science, 428, 104-115.
561 562
of
polyvinylidenefluoride-co-hexafluoropropylene
(PVDF-HFP)
Lau, W. J., Ismail, A. F., Misdan, N., & Kassim, M. A. (2012). A recent progress in thin film composite membrane: a review. Desalination, 287, 190-199. 27
Page 28 of 33
Law, J. L. M., Chen, C., Wong, J., Hockman, D., Santer, D. M., Frey, S. E., Belshe, R. B.,
564
Wakita, T., Bukh, J., & Jones, C. T. (2013). A hepatitis C virus (HCV) vaccine
565
comprising envelope glycoproteins gpE1/gpE2 derived from a single isolate elicits broad
566
cross-genotype neutralizing antibodies in humans. PLoS ONE, 8, e59776.
ip t
563
Levings, R., & Wessman, S. (1990). Bovine viral diarrhea virus contamination of nutrient
568
serum, cell cultures and viral vaccines. Developments in Biological Standardization, 75,
569
177-181.
us
Liao, N., Unnithan, A. R., Joshi, M. K., Tiwari, A. P., Hong, S. T., Park, C.-H., & Kim, C. S.
571
(2015).
bioactive
poly
(ɛ-caprolactone)–cellulose
acetate–dextran
572
antibacterial composite mats for wound dressing applications. Colloids and Surfaces A:
573
Physicochemical and Engineering Aspects, 469, 194-201.
d
M
Electrospun
an
570
cr
567
Lindenbach, B. D., Evans, M. J., Syder, A. J., Wölk, B., Tellinghuisen, T. L., Liu, C. C.,
575
Maruyama, T., Hynes, R. O., Burton, D. R., & McKeating, J. A. (2005). Complete
576
replication of hepatitis C virus in cell culture. Science, 309, 623-626.
Ac ce pt e
574
577
Liu, H., & Hsieh, Y. L. (2002). Ultrafine fibrous cellulose membranes from electrospinning
578
of cellulose acetate. Journal of Polymer Science Part B: Polymer Physics, 40, 2119-2129.
579
Luo, X., & Zhang, L. (2013). New solvents and functional materials prepared from cellulose
580
solutions in alkali/urea aqueous system. Food Research International, 52, 387-400.
581
Luo, X., Zhu, J., Gleisner, R., & Zhan, H. (2011). Effects of wet-pressing-induced fiber
582
hornification on enzymatic saccharification of lignocelluloses. Cellulose, 18, 1055-1062.
28
Page 29 of 33
583 584
Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2011). Nanofibrous microfiltration membrane based on cellulose nanowhiskers. Biomacromolecules, 13, 180-186. Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2012). Highly permeable polymer membranes
586
containing directed channels for water purification. ACS Macro Letters, 1, 723-726.
587
Ma, H., Burger, C., Hsiao, B. S., & Chu, B. (2014). Fabrication and characterization of
588
cellulose nanofiber based thin-film nanofibrous composite membranes. Journal of
589
Membrane Science, 454, 272-282.
us
cr
ip t
585
Ma, H., Hsiao, B. S., & Chu, B. (2011). Thin-film nanofibrous composite membranes
591
containing cellulose or chitin barrier layers fabricated by ionic liquids. Polymer, 52,
592
2594-2599.
M
an
590
Ma, H., Hsiao, B. S., & Chu, B. (2014). Functionalized electrospun nanofibrous
594
microfiltration membranes for removal of bacteria and viruses. Journal of Membrane
595
Science, 452, 446-452.
Ac ce pt e
d
593
596
Ma, H., Yoon, K., Rong, L., Mao, Y., Mo, Z., Fang, D., Hollander, Z., Gaiteri, J., Hsiao, B. S.,
597
& Chu, B. (2010a). High-flux thin-film nanofibrous composite ultrafiltration membranes
598
containing cellulose barrier layer. Journal of Materials Chemistry, 20, 4692-4704.
599
Ma, H., Yoon, K., Rong, L., Shokralla, M., Kopot, A., Wang, X., Fang, D., Hsiao, B. S., &
600
Chu, B. (2010b). Thin-film nanofibrous composite ultrafiltration membranes based on
601
polyvinyl alcohol barrier layer containing directional water channels. Industrial &
602
Engineering Chemistry Research, 49, 11978-11984.
29
Page 30 of 33
603 604
Ma, Z., Kotaki, M., & Ramakrishna, S. (2005). Electrospun cellulose nanofiber as affinity membrane. Journal of Membrane Science, 265, 115-123. Quellmalz, A., & Mihranyan, A. (2015). Citric Acid Cross-Linked Nanocellulose-Based
606
Paper for Size-Exclusion Nanofiltration. ACS Biomaterials Science & Engineering, 1,
607
271-276.
610 611
cr
Electrospinning and Nanofibers. Singapore: World Scientific Publishing.
us
609
Ramakrishna, S., Fujihara, K., Teo, W.-E., Lim, T.-C., & Ma, Z. (2005). An Introduction to
Rautenbach, R., & Gröschl, A. (1990). Separation potential of nanofiltration membranes.
an
608
ip t
605
Desalination, 77, 73-84.
Rodríguez, K., Renneckar, S., & Gatenholm, P. (2011). Biomimetic calcium phosphate
613
crystal mineralization on electrospun cellulose-based scaffolds. ACS applied materials &
614
interfaces, 3, 681-689.
Ac ce pt e
d
M
612
615
Rodríguez, K., Sundberg, J., Gatenholm, P., & Renneckar, S. (2014). Electrospun nanofibrous
616
cellulose scaffolds with controlled microarchitecture. Carbohydrate Polymers, 100, 143-
617
149.
618
Shimizu, Y. K., Feinstone, S. M., Kohara, M., Purcell, R. H., & Yoshikura, H. (1996).
619
Hepatitis C virus: Detection of intracellular virus particles by electron microscop.
620
Hepatology, 23, 205-209.
621
Son, W. K., Youk, J. H., Lee, T. S., & Park, W. H. (2004). Electrospinning of ultrafine
622
cellulose acetate fibers: studies of a new solvent system and deacetylation of ultrafine
623
cellulose acetate fibers. Journal of Polymer Science Part B: Polymer Physics, 42, 5-11. 30
Page 31 of 33
ŠUTÝ, Š., PETRILÁKOVÁ, K., KATUŠČÁK, S., KIRSCHNEROVÁ, S., JABLONSKÝ, M.,
625
VIZÁROVÁ, K., & VRŠKA, M. (2012). Change in the capability of cellulose fibers to
626
retain water during thermally accelerated ageing of paper. Cellulose Chemistry and
627
Technology, 46, 631-635
ip t
624
Unnithan, A. R., Gnanasekaran, G., Sathishkumar, Y., Lee, Y. S., & Kim, C. S. (2014).
629
Electrospun antibacterial polyurethane–cellulose acetate–zein composite mats for wound
630
dressing. Carbohydrate Polymers, 102, 884-892.
us
cr
628
Wang, Q., Cai, J., Zhang, L., Xu, M., Cheng, H., Han, C. C., Kuga, S., Xiao, J., & Xiao, R.
632
(2013). A bioplastic with high strength constructed from a cellulose hydrogel by
633
changing the aggregated structure. Journal of Materials Chemistry A, 1, 6678-6686.
634
Wang, Y., & Chen, L. (2011). Impacts of nanowhisker on formation kinetics and properties of
M
d
all-cellulose composite gels. Carbohydrate Polymers, 83, 1937-1946.
Ac ce pt e
635
an
631
636
Wang, Y., & Chen, L. (2014). Cellulose nanowhiskers and fiber alignment greatly improve
637
mechanical properties of electrospun prolamin protein fibers. ACS Appl Mater Interfaces,
638
6, 1709-1718.
639 640 641 642
Weise, U., & Paulapuro, H. (1999). Effect of drying and rewetting cycles on fibre swelling. Journal of Pulp and Paper Science, 25, 163-166. Wu, M., Shuai, H., Cheng, Q., & Jiang, L. (2014). Bioinspired Green Composite Lotus Fibers. Angewandte Chemie International Edition, 53, 3358-3361.
31
Page 32 of 33
643
Yeh, T.-M., Wang, Z., Mahajan, D., Hsiao, B. S., & Chu, B. (2013). High flux ethanol
644
dehydration using nanofibrous membranes containing graphene oxide barrier layers.
645
Journal of Materials Chemistry A, 1(41), 12998-13003. Yoon, K., Hsiao, B. S., & Chu, B. (2009). High flux nanofiltration membranes based on
647
interfacially polymerized polyamide barrier layer on polyacrylonitrile nanofibrous
648
scaffolds. Journal of Membrane Science, 326, 484-492.
cr
ip t
646
Yoon, K., Kim, K., Wang, X., Fang, D., Hsiao, B. S., & Chu, B. (2006). High flux
650
ultrafiltration membranes based on electrospun nanofibrous PAN scaffolds and chitosan
651
coating. Polymer, 47, 2434-2441.
an
M
653
Zhang, Z., Zhang, Y.-W., & Gao, H. (2011). On optimal hierarchy of load-bearing biological materials. Proceedings of the Royal Society B: Biological Sciences, 278, 519–525.
d
652
us
649
Zhou, J., Chang, C., Zhang, R., & Zhang, L. (2007). Hydrogels prepared from unsubstituted
655
cellulose in NaOH/urea aqueous solution. Macromolecular Bioscience, 7, 804-809.
656
Zhou, Z., Peng, X., Zhong, L., Wu, L., Cao X., Sun R. (2016). Electrospun cellulose acetate
657
supported Ag@AgCl composites with facet-dependent photocatalytic properties on
658
degradation of organic dyes under visible-light irradiation. Carbohydrate Polymers, 136,
659
322–328.
660 661
Ac ce pt e
654
Zhou, J., Zhang, L., Cai, J., & Shu, H. (2002). Cellulose microporous membranes prepared from NaOH/urea aqueous solution. Journal of Membrane Science, 210, 77-90.
662 663 664 32
Page 33 of 33