See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/328579185

Electrospun cellulose acetate nanofibers for airborne nanoparticle filtration

Article  in  Textile Research Journal · October 2018

DOI: 10.1177/0040517518807440

CITATION READS

1 78

4 authors, including:

Qinghai Li Zhongchao Tan

Tsinghua University University of Waterloo,
74 PUBLICATIONS   933 CITATIONS    157 PUBLICATIONS   1,055 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Air Pollution and GHGs book View project

Bioresources View project

All content following this page was uploaded by Zhongchao Tan on 14 May 2019.

The user has requested enhancement of the downloaded file.

Original article
Textile Research Journal
0(00) 1–13

Electrospun cellulose acetate nanofibers ! The Author(s) 2018

Article reuse guidelines:
sagepub.com/journals-permissions
for airborne nanoparticle filtration DOI: 10.1177/0040517518807440
journals.sagepub.com/home/trj

Joerg Ahne1, Qinghai Li2, Eric Croiset1 and Zhongchao Tan1

Abstract
Reported in this paper are the effects of tip-to-collector distance, voltage, deposition time and solution concentration on
the fiber size distribution and filter quality factor of electrospun cellulose acetate (CA)-based nanofibers. Nanofibrous
filter samples were produced by electrospinning in a laboratory setting. The CA solutions were prepared by diluting
various concentrations of CA in a 2:1 (w:w) ratio of N,N-dimethylacetamide (concentration 10–20 wt.%). The electro-
spinning voltages ranged from 8–12 kV, with distances from 10–15 cm and deposition times of up to 30 minutes. The
produced nanofibrous filter samples were then analyzed in terms of fiber size distribution and filter quality factor using
nanosized sodium chloride particles ranging from 4–240 nm in diameter. The maximum filtration efficiency measured was
99.8% for filter samples obtained with an overall deposition time of 30 minutes. The maximum filter quality factor was
0.14 Pa1 for a CA concentration of 20 wt.% and a tip-to-collector distance of 15 cm. The average fiber diameters of the
fibers were between 175 and 890 nm, and CA concentrations below 15% led to the formation of beads.

Keywords
nanofiber, airborne nanoparticle filtration, filtration efficiency, electrospun cellulose acetate, fiber size distribution, filter
quality factor

Nanofibers have been used for air filtration for a few
decades. Nanofibers find applications in many indus-
tries including textiles, clean energy and environment,
medicine and defense.1 Interest in the research and
development of nanofibers by electrospinning has
been growing in recent years due to their potential in
producing sub-100 nm nanofibers. Among several
methods that are available for nanofiber production,
electrospinning is considered superior over others due
to its ability to produce fine fibers from a variety of
materials,2 for example, poly(vinyl alcohol) (PVA),1
polyurethane,3 polypyrrole-poly(acrylonitrile-co-vinyl
acetate),4 polyacrylonitrile (PAN) and dimethylforma-
mide,5 polypropylene,6 alumina,7 activated carbon8 and
polyamide-6.9 A review of multifunctional polymer
nanocomposites for multifunctional finishing of textiles
is also available.10
Electrospinning has found its way into industrial
applications over the last 8 years. A typical electrospin-
ning process is used to fabricate nanofibers with diam-
eters in the order of 100 nm. The electrospinning
process depends on multiple parameters, including
polymer properties (concentration, viscosity, surface
tension, etc.), applied voltage, tip-to-collector distance,
nozzle diameter, and ambient conditions such as tem-
perature and relative humidity. Beaded fibers may be
produced when the tip-to-collector distance, solvent
flow rate or polymer concentration is high.11
Furthermore, a high-viscosity polymer leads to larger
fibers. Thinner fibers are produced at a higher tempera-
tures, high electrospinning voltages and small flow rates
for single jet electrospinning processes.
Nanofibers can be produced for the filtration of air-
borne nanoparticles (1–100 nm in diameter) for the pro-
tection of human health and the environment.12–16
Researchers believe that the features of nanofibers
include high filtration efficiency and low pressure
1
University of Waterloo, Canada
2
Tsinghua University-University of Waterloo Joint Research Centre for
Micro/Nano Energy and Environmental Technology, China
Corresponding author:
Zhongchao Tan, University of Waterloo, 200 University Avenue, West
Waterloo, Ontario N2L 3G1, Canada.
Email: tanz@uwaterloo.ca
2 Textile Research Journal 0(00)
drop. PVA-, nylon 6-, cellulose acetate (CA)- and
PAN-based nanofibers have been successfully fabri-
cated in laboratories for air filtration purposes.
Within the area of air filtration, there is a growing
interest in biodegradable nanofibers, which is mainly
driven by the global promotion of sustainability and
related policies.17,18 Technical challenges arise in the
quest to produce biodegradable polymers in a cost-
effective manner, with performance comparable to syn-
thetic polymers.19 Biodegradable nanofibers can be
made of CA, which is a natural polymer that can poten-
tially replace synthetic polymers. Cellulose is one of the
most widely available organic compounds on Earth.
CA nanofibers offer relatively high modulus, flexible
and tensile strength. The properties of CA nanofibers
can also be tuned for certain contaminants by functio-
nalizing the nanofibers with chemical groups. CA nano-
fibers have been shown to be biocompatible, non-toxic
and biodegradable, thus making them ideal candidates
for filtration applications.20
Researchers have started to investigate the perform-
ance of CA nanofibers on the filtration of airborne
nanoparticles.19 Chattopadhyay et al.21 investigated
the effects of the fiber diameter, solidity and membrane
thickness of electrospun CA fibers on nanoparticle fil-
tration. Their CA fibers were produced by electrospin-
ning. In their study, the tested aerosols were NaCl
particles in the size range of 40–270 nm and diethyl
hexyl sebactate at a face velocity of 45 cm/s, which is
considered to be high. They found that the filtration
performance of electrospun nanofibers was higher
than that of thin commercial filtration materials.
Matulevicius et al.22 tested the effects of operating par-
ameters on the fiber size and filtration performance of
electrospun polymers, including CA. They used mono-
dispersed particles with sizes of 100 and 300 nm in
diameter, which are not nanoparticles. However, the
available data for CA-based materials is still insuffi-
cient, especially for their application in the filtration
of airborne nanoparticles. In addition, there is
limited information available on the air filtration
performance of biodegradable filters at lower face velo-
cities.22 Most of the works published to date only
focuses on morphology and other physical properties
of the filters.
The primary objective of this work is to better
understand the effects of key parameters that affect
the filtration quality factor of CA nanofibrous filters
for nanoparticles. The key parameters include solvent
concentration, voltage, tip-to-collector distance and
deposition time in electrospinning. The electrospun
CA nanofibers were characterized, and tested for filtra-
tion efficiency and quality factors to determine the opti-
mal conditions for the fabrication of CA nanofibers
that aim to achieve nanoparticle filtration. This paper
contributes to the parameter studies of CA and poly-
mer nanofibers in general. Additionally, it provides
practical knowledge of air filtration, which determines
the potential usability of these nanofibers. The main
contribution of this paper lies in the information pro-
vided regarding the production of CA with various par-
ameters. It also enriches the available data concerning
the filtration mechanics of CA nanofibrous materials.
Experimental details
Cellulose acetate was purchased from Sigma Aldrich,
USA (average Mn & 30,000). CA solutions were pre-
pared by diluting various concentrations of CA at a 2:1
(w:w) ratio of N,N-dimethylacetamide (Sigma Aldrich,
99.8%) and acetone (Sigma Aldrich, 99.7%). The
resultant mixtures were stirred at 35 C on a magnetic
stirring heating plate for approximately 2 hours to com-
pletely dissolve CA into the solvent mixture. The poly-
mer molecules were assumed to be adequately dissolved
in the solvent for the electrospinning process. As is typ-
ical practice in this field of research, complete dissol-
ution was confirmed through visual inspection when
the solvent turned to a clear liquid. Similar formula-
tions of the solvent have been reported to work well
for producing CA nanofibers in the literature.23,24
Three different concentrations of CA were tested for
this study: 10, 15 and 20 wt.%.
The CA solution was then placed in a 5 mL syringe
and connected to a capillary tip (gauge 22); this solu-
tion was pressed through the tube and capillary with a
syringe pump at a constant feed rate of 0.6 mL/h. The
capillary needle was connected to a high-voltage supply
(maximum voltage  30 kV). The voltage varied from
8–12  0.1 kV and the distance between the needle
and collector from 10–15 cm  0.1 cm. The nanofiber
samples were deposited onto a stainless steel screen
with a wire diameter of 140 mm and opening size of
368 mm. The screens were placed on aluminum foil
that covered the grounded collector. A schematic of
the setup can be found in Figure 1. The test conditions
for the filter samples are listed in Table 1.
As described in our earlier work,15 the nanofibrous
filter samples were first analyzed by scanning electron
microscopy (SEM) (Zeiss Gemini Model Leo 1550) to
observe their morphology. Then, fiber size distribution
was analyzed using a MATLAB code. More details can
be found the supplemental materials of our earlier
paper.15 The same samples were tested in terms of filtra-
tion efficiency. The analysis steps are introduced below.
In preparation for the SEM analysis, a conductive
gold coating was sputtered onto the nanofiber surfaces
for 120 s to better visualize the nanofibers.
Representative SEM images were obtained at magnifi-
cations from 5000–21,000. MATLAB code developed
Ahne et al. 3

Figure 1. Schematic of electrospinning setup.

Table 1. The electrospinning parameters for the nanofiber samples

Cellulose acetate Deposition Applied Tip-to-collector

Sample name concentration (wt.%) time (minutes) voltage (kV) distance (cm)

CA10_5_8_10 10 5 8 10
CA10_5_8_12.5 10 5 8 12.5
CA10_5_8_15 10 5 8 15
CA10_15_8_10 10 15 8 10
CA15_5_8_10a 15 5 8 10
CA15_5_8_12.5a 15 5 8 12.5
CA15_5_8_15a 15 5 8 15
CA15_5_10_10a 15 5 10 10
CA15_5_12_10a 15 5 12 10
CA15_15_8_10 15 15 8 10
CA15_15_10_10 15 15 10 10
CA15_15_12_10 15 15 12 10
CA15_30_8_10 15 30 8 10
CA15_30_10_10 15 30 10 10
CA20_5_8_10a 20 5 8 10
CA20_5_8_12.5a 20 5 8 12.5
CA20_5_8_15 20 5 8 15
CA20_5_10_10a 20 5 10 10
CA20_5_12_10 20 5 12 10
CA20_15_8_10 20 15 8 10
CA20_30_8_10 20 30 8 10
a
The indicated samples are also displayed in Table 2.

in-house was used to quantify the diameter of the fibers.
The algorithm used for this process included image pre-
treatment, local threshold, smoothness, edge detection
and skeletonization.15 A Euclidean distance transform
matrix was used to obtain fiber diameters by
determining the separation between fiber edges and cen-
terlines. Multiple images were obtained at each magni-
fication and from different locations to get more
representative samples. An equal number of low
(factor 5000) magnification and high (factor 20,000)
4 Textile Research Journal 0(00)

magnification images were averaged out for each compare the filtration performances of the filter sam-
sample in order to get a more representative value for ples. It is a standard parameter in filtration research.
average fiber diameter. The filter quality factor (QF) can be calculated using
The filter testing apparatus is the same as the one that equation 2:
was reported in our previous studies.15,16 The same setup
is shown in Figure 2. The testing setup is comparable to lnð1  Þ
QF ¼  ð2Þ
those used by Chattopadhyay et al.21 and Matulevicius
P
et al.22 A constant output atomizer (TSI model 3076)
was used to generate aerosol particles from a solution where QF is the qualify factor and
P the pressure drop
composed of 0.1 g/L sodium chloride (NaCl) in distilled across the filter sample. Since the particle filtration
water. In this case, compressed air was used to form a efficiency depends on the particle size, QF is size-
high-velocity jet that atomizes the solution into polydis- dependent. To simplify the calculation, in this paper,
persed NaCl aerosol particles. The aerosol particles were total efficiency is calculated as the summation of all
then passed through a diffusion drier (TSI Model 362) to particle sizes tested.
remove water vapor from the generated aerosol.
The aerosol particles were then passed through the
filter medium at an air flow rate of 1.5 L/minute. The Results and discussion
filter medium was placed on a filter holder (BGI Inc.,
Waltham, MA). The filter sample and the filter holder
Fiber size distributions
have a diameter of 25 mm. With a flow rate of Table 2 shows the SEM images and fiber size distribu-
1.5 L/minute, the face velocity on the filter is 5.1 cm/s. tions of the samples indicated in Table 1. It is
The pressure drop across the filter medium (
P) was
monitored and recorded using a differential pressure
gauge (Omegadyne, model DPG409). Each test was 1.20E+07
repeated five times. All tests were conducted in room
conditions, at normal temperature and with a relative 1.00E+07
humidity of about 50%.
dN/dlog(dp) (1/cm³)

8.00E+06
The fractional filtration efficiency was calculated
using the concentrations of particles measured with 6.00E+06
and without the presence of the nanofibrous filter (see 4.00E+06
equation 1):15
2.00E+06
Cw=o
¼ 1 ð1Þ
Cw 0.00E+00
1 10 100 1000
where Cw=o and Cw are the concentrations of nanopar- Parcle Size (nm)
ticles before and after the filter sample, respectively.
Since they have the same unit, equation 1 is dimension-
Figure 3. Particle size distribution of the NaCl particles.
less. In this paper, filter quality factor is used to

Clean, Dry, and Oil- Hygrometer

free Compressed Air HEPA Filter
Filter Holder
Fume
Hood
Diffusion Neutralizer Mass Flow
Dryer Meter

NaCl
Particles

SMPS+E (GRIMM)

Figure 2. Schematic diagram of the filtration setup.

Ahne et al. 5

Table 2. Sample scanning electron microscopy images and determined fiber size distributions

CA15_5_8_10

0.16
Mean df: 145 nm
0.14
Std: 82.8 nm
0.12

Frequency (-)
0.1
0.08
0.06
0.04
0.02
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA15_5_8_12.5

0.2

Mean df: 254.6 nm

0.15

Frequency (-)
Std: 238.6 nm

0.1

0.05

0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA15_5_8_15

0.4
Mean df: 72.7 nm
0.35
Std: 176.6 nm
0.3
Frequency (-)

0.25
0.2
0.15
0.1
0.05
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA15_5_10_10

0.08
Mean df: 283.7 nm
0.07 Std: 254.7 nm
0.06
Frequency (-)

0.05
0.04
0.03
0.02
0.01
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

(continued)
6 Textile Research Journal 0(00)

Table 2. Continued

CA15_5_12_10

0.16
Mean df: 252.5 nm
0.14 Std: 252.2 nm
0.12

Frequency (-)
0.1
0.08
0.06
0.04
0.02
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA20_5_8_10

0.08
Mean df: 252.5 nm
0.07
Std: 245.4 nm
0.06

Frequency (-)
0.05
0.04
0.03
0.02
0.01
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA20_5_8_12.5

0.08
Meand df: 319.4 nm
0.07 Std: 491.4 nm
0.06
Frequency (-)

0.05
0.04
0.03
0.02
0.01
0
0 125 250 375 500 625 750 875 1000
Fiber Diameter (nm)

CA20_5_10_10

0.09
0.08 Mean df: 723.3 nm
Std: 751.2 nm
0.07
Frequency (-)

0.06
0.05
0.04
0.03
0.02
0.01
0

Fiber Diameter (nm)

Ahne et al.
noticeable that all of the distributions have a skewed
log-normal shape and, depending on the production
parameters, the SDs and means of the samples differs
strongly. The impacts of the single parameters voltage,
tip-to-collector distance and concentration on the fiber
distributions are discussed in the following sections.
Effects of CA concentration on the sizes and
morphologies of the nanofibers
The SEM images in Figure 4 are for samples with
10 wt.% CA. They clearly show the beading of poly-
mers. Droplet formation was observed when the poly-
mer concentration was below a certain point or a
critical chain entanglement concentration, ce.25 Such
defects were not observed above this critical concentra-
tion. Similar findings have been reported by others.21 It
has been reported that the magnitude of ce is dependent
on solvent–polymer interactions.24 Bead-free nanofi-
bers are expected at high CA concentrations, that is,
over 15 wt.%. Higher DMAc content in the solvent
leads to strong hydrogen-bonding interactions between
the solvent and the CA molecules. For the 2:1 DMAc/
acetone formulation, the evaporation of the solvent is
expected to be slower, resulting in a wet jet landing on
the collector. In order to minimize surface tension, the
solution on the collector forms spherical beads.
Note that, in an earlier study24 using a 2:1 DMAc/
acetone solvent solution, the beading phenomenon was
observed at concentrations as high as 20 wt.%. Bead-
free nanofibers were only obtained at a CA concentra-
tion of 25 wt.%.24 In the current work, the absence of
droplets in the 15 wt.% micrographs indicates that the
critical chain entanglement concentration for the 2:1
DMAc/acetone solvent mixture is somewhere between
10 and 15 wt.% of CA. This also indicates that other
factors may have played an important role in bead for-
mation. It is likely that the relative humidity or
Figure 4. Scanning electron microscopy images of the nanofibers produced with different cellulose acetate concentrations.
(a) 10 wt.%, (b) 15 wt.% and (c) 20 wt.%.
7
temperature of the air have an impact, since both par-
ameters can influence the conductivity or surface
tension of the solution. Based on the qualitative obser-
vations above, only 15 and 20 wt.% samples were ana-
lyzed from this point on for fiber diameter, filtration
efficiency and filtration quality.
A higher CA solution ultimately leads to higher vis-
cosity of the solution, and consequent higher resistance
of the liquid being precipitated by the electric field. This
resistance leads to lower bending instabilities and ultim-
ately thicker nanofibers. Similar trends have been
observed for CA nanofibers by other research-
ers.20,23,24,26 The increased range of fiber sizes is also
common for increased polymer concentrations.21 In the
study by Chattopadhyay et al., the concentration of
15% CA resulted in a fiber diameter of 302  127 nm,
whereas the fiber diameters corresponding to 20% CA
were 669  341 nm.
Effects of tip-to-collector distance on fiber diameter
Tip-to-collector distance can impact the quality of the
fibers; beads can be observed if the tip-to-collector dis-
tance is either too short or too long.27 Tip-to-collector
distances of 10, 12.5 and 15 cm were considered for the
following study. No beading was observed for these
three tip-to-collector distances.
As seen in Table 3, for the concentration of 15 wt.%
CA, the general trend indicates an increase in fiber
diameter with increasing the tip-to-collector distance
up to 12.5 cm, after which it reaches a plateau. For
the concentration of 20 wt.% CA, the average fiber
diameter increased from 327  245 to 466  262 nm
when the tip-to-collector distance changed from 10 to
12 cm. A further increase of 2.5 cm in the tip-to-collec-
tor distance resulted in an increase in fiber diameter to
836  690 nm. The wide uncertainty associated with
the diameter indicates a poor correlation between
8 Textile Research Journal 0(00)
Table 3. Fiber size distributions at different concentrations and
tip-to-collector distances
Cellulose acetate Tip-to-collector Average nanofiber
concentration distance (cm) diameter (nm)
15 wt% 10 175  82
12.5 343  175
15 344  174
20 wt% 10 327  245
12.5 466  262
15 836  690
tip-to-collector distance and average fiber diameter.
Other researchers have observed that a larger
tip-to-collector distance results in thinner nanofibers.
A larger distance provides a longer time period for
the solvent to evaporate. Therefore, the jet undergoes
increased elongation due to greater bending instabilities
in the system. Hence, thinner fibers can be obtained.
The increased range of fiber diameters could also be
attributed to the increase of bending instability. The
whipping continuously alters the traveling distance of
the fibers, which has a direct impact on the final
fiber size. However, a similar study concluded that
the diameter size and morphology of the fibers were
not significantly influenced by the deposition range
between 5 and 15 cm.20 Similarly, another study saw
a minimal impact of the tip-to-collector distance of
15–30 cm on the corresponding SEM images.20 It
would be desirable to have some data at a shorter dis-
tance, but no fibers could be formed and collected at
distances of 10 cm. An advanced experimental setup
shall be developed for nanofiber fabrication at short
distances. It is worth noting that it is very normal to
have a wide range of fiber sizes. The fiber sizes depend
not only on the operation of the setup (voltage, distance
and concentration), but also on the experimental setup
itself. Chattopadhyay et al.21 used a drum collector,
and Christoforou et al.20 used a horizontal setup with
a plate sized 24  24 cm. They all demonstrated that
collector size directly impact the strength of the electric
field and the final fiber morphology.
Effects of applied voltage on fiber size
Samples were produced at different applied voltages
with two CA concentrations of 15 and 20 wt.% CA,
but with the same deposition time of 5 minutes and
same tip-to-collector distance of 10 cm. These two
groups of samples were used to study the effects of
applied voltage, which were 8, 10 and 12 kV, on the
fiber diameter.
Table 4. Fiber diameters at different concentrations and
applied voltages
Cellulose acetate Voltage Average nanofiber
concentration (kV) diameter (nm)
15 wt.% 8 175  82
10 330  254
12 318  259.
20 wt.% 8 466  262
10 836  690
12 879  783
The experimental results are summarized in Table 4.
For both CA solutions, an increasing trend is observed
for the average fiber diameter with applied voltage. The
sharp increase in the average fiber diameter can be
seen when the voltage was increased from 8 to 10 kV
for both CA concentrations. For the concentration of
20 wt.%, the average fiber diameter increases by 369 nm
when the voltage increased from 8 to 10 kV. When
the applied voltage increased from 10 to 12 kV, the
increase in average fiber diameter was only 43 nm.
A similar trend was also observed when the concentra-
tion was 15 wt.%. However, a marginal decrease
in fiber diameter was observed when the voltage was
changed from 10 to 12 kV for the concentration of
15 wt.% CA.
The effects of the applied voltage on fiber diameter
were not conclusive based on earlier studies.20,21 In a
separate study on CA nanofibers, other researchers
found an inverse effect of applied voltage on the
mean fiber diameter.27 This was attributed to stronger
electrostatic forces at higher voltages, which caused
the fibers to stretch more and become thinner.
Another study showed that voltage had a negligible
effect on the diameter of CA nanofibers.28 Increasing
the applied electric field resulted in a slight decrease in
mean fiber diameter for ethylcellulose nanofibers.
However, an increase in fiber diameter may also be
attributed to a stronger electric field, which can result
in the solution being ejected more quickly from the
capillary tip.
Table 5 shows the results of similar studies in the
literature. It can be seen that the variance of the pro-
duced fibers strongly depends on the operating param-
eters. The parameters also varied between studies for
different setups, especially the collector shape and
dimension. Furthermore, it can be seen that a higher
polymer concentration leads to a greater mean fiber
diameter. The fiber sizes produced in our study herein
for concentrations of 15 wt.% are comparable with
those produced by others.
Ahne et al. 9

Table 5. Fiber diameters and production parameters of other research

Cellulose acetate
concentration Mean fiber diameter Voltage (kV) Distance (cm) Reference
21
11% 0.12  0.07 mm (beading) 75–80 20
21
15% 0.3  0.01 mm 75–80 20
21
17% 0.5  0.019 mm 75–80 20
24
15% 275  195 mm (different molecular weight) 8 15
24
15% 241  111 nm 8 15
24
15% 287  165 nm 8 15
22
11% 350  129 nm (different solvent) 18 16
22
11% 363  131 nm 18 16

(a) (b)
1 0.1
CA15_5_8_10 CA15_5_8_10
0.9 CA20_5_8_10 0.09
CA20_5_8_10
0.8 CA15_15_8_10 0.08
CA15_15_8_10
Quality Factor [Pa-1]
0.7 CA20_15_8_10 0.07 CA20_15_8_10
Efficiency [-]

0.6 CA15_30_8_10 0.06 CA15_30_8_10

CA20_30_8_10 CA20_30_8_10
0.5 0.05
0.4 0.04
0.3 0.03
0.2 0.02
0.1 0.01
0 0
10 100 1000 10 100 1000
Parcel size [nm] Parcel size [nm]

Figure 5. (a) Filtration efficiencies and (b) quality factors at different cellulose acetate concentrations.

Air filtration performance polymer concentration of the solution. A higher con-

Effect of CA concentration on filtration performance
Figure 5 shows the effects of CA concentration on the
filtration efficiencies and quality factors corresponding
to deposition times of 5–30 minutes. The tip-to-collec-
tor distance and applied voltage were kept constant at
10 cm and 8 kV, respectively. Since obvious beading of
the polymer was observed for 10 wt.% CA, only sam-
ples for the concentrations of 15 and 20 wt.% are con-
sidered in this paper. At all deposition times, except for
15 minutes, filtration efficiency and quality factor
increased when the CA concentration increased from
15 to 20 wt.%. For example, in the case of 30 minutes
of deposition time, increasing the CA concentration
from 15 to 20 wt.% led to an increase in average filtra-
tion efficiency from 14.8 to 60.4%. Under the same
conditions, the quality factor increased from 0.0214 to
0.027 Pa1.
The higher filtration efficiency can be explained by
the increased mass deposition of fibers due to the higher
centration of polymer in the solution leads to relatively
less volatile solvent in the jet, which results in a higher
polymer mass deposition rate and solidity of the pro-
duced filters. The higher filtration quality is connected
to the higher filtration efficiency with an equal pressure
drop or less. The slipping effect enabled by the very fine
fibers enables a better filtration performance with a
lower pressure drop. Slip flow occurs when the size of
the fiber approaches the mean free path of the air. As a
result, the air molecules ‘‘slip’’ over the fiber surface,
which leads to a lower pressure drop. As a result, the
deposited mass does not impact the pressure drop
significantly.
Effects of tip-to-collector distance on filtration
performance
Figure 6 shows the effects of tip-to-collector distance on
the filtration efficiency and quality factor of filter sam-
ples obtained with 15 and 20 wt.% CA. The deposition
10 Textile Research Journal 0(00)

(a) 0.5 (b) 0.2

CA15_5_8_10 CA15_5_8_10
0.45 CA15_5_8_12.5 0.18 CA15_5_8_12.5
0.4 CA15_5_8_15 0.16 CA15_5_8_15
CA20_5_8_10 CA20_5_8_10

Quality Factor [Pa-1]

0.35 0.14
CA20_5_8_12.5 CA20_5_8_12.5
Efficiency [-]

0.3 CA20_5_8_15 0.12 CA20_5_8_15

0.25 0.1
0.2 0.08
0.15 0.06
0.1 0.04
0.05 0.02
0 0
10 100 1000 10 100 1000
Parcel size [nm] Parcel size [nm]

Figure 6. (a) Filtration efficiencies and (b) quality factors at different tip-to-collector distances.

(a) (b)
1 0.1
CA15_5_8_10 CA15_5_8_10
0.9 CA15_15_8_10 0.09
CA15_15_8_10
0.8 CA15_30_8_10 0.08 CA15_30_8_10
Quality Factor [Pa-1]

0.7 CA20_5_8_10 0.07 CA20_5_8_10

Efficiency [-]

CA20_15_8_10 CA20_15_8_10
0.6 CA20_30_8_10 0.06
CA20_30_8_10
0.5 0.05
0.4 0.04
0.3 0.03
0.2 0.02
0.1 0.01
0 0
10 100 1000 10 100 1000
Parcel size [nm] Parcel size [nm]

Figure 7. (a) Filtration efficiencies and (b) quality factors at different deposition times.

time and applied voltage were kept constant at 5 min-
utes and 8 kV, respectively. For both CA concentra-
tions, positive trends are observed for filtration
efficiency and quality factor with increasing tip-to-col-
lector distance. There was generally an increase in aver-
age fiber diameter with increasing tip-to-collector
distance. Typically, thinner nanofibers correspond to
higher filtration efficiency. However, this trend was
not observed in the study herein. As seen in Table 2,
although the mean sizes of certain samples were rela-
tively large, these samples have many fibers that are
smaller than the measured fiber diameter. Therefore,
the effects of nanosized fibers can be observed for sam-
ples with a high mean fiber size. In a separate study
using PVA nanofibers,15 a similar correlation was
obtained between tip-to-collector distance and both fil-
tration efficiency and filter quality factor.
Effects of deposition time on filtration performance
The effects of deposition time on the total filtration
efficiency and quality factor are shown in Figure 7
below. All data points were for 15 and 20 wt.% CA
solutions with a tip-to-collector distance of 10 cm
when the applied voltage was 8 kV. It was observed
for both concentrations that both the filtration effi-
ciency and quality factor increased with deposition
time. The filtration performance was relatively poor
when the deposition time was 5 minutes for all tested
samples. As shown in Figure 6(a), the filtration
Ahne et al.
efficiency hovered around 10% for all samples.
However, an increase in deposition time obviously pro-
moted the filtration performance. Increasing the depos-
ition time from 5 to 30 minutes led to a jump of average
filtration efficiency from 1.8 to 60.4%. As discussed
above, slip flow regime was observed when nanofiber
diameter was less than 1 mm. Thicker nanofibers may be
one factor that led to low filtration efficiency.
It can be seen from Figure 7 that the filtration qua-
lities corresponding to the 15% CA sample and 20%
CA concentrations are close to each other. Table 3 and
Table 4 show that a lower concentration resulted in
finer fibers. Thinner fibers can be deposited for a
longer time than thick ones without restricting the air
flow to a point that the pressure drop overcomes the
gain in filtration efficiency.
Effects of applied voltage on filtration performance
Figure 8 shows the effects of applied voltage on the
total filtration efficiency and quality factor. For a
given deposition time, the filtration efficiency increases
with increasing applied voltage. The corresponding
concentration was 15 wt.% with a tip-to-collector dis-
tance of 10 cm. A similar trend was also observed in
terms of the quality factor of the media. This trend in
filtration efficiency is clearly observed for all deposition
times. As the voltage increased from 8 to 12 kV, the
filtration efficiency approached 100%. For example,
for the deposition time of 15 min, the filtration effi-
ciency for the 10 kV sample was 36.7% and was
99.3% for the 12 kV sample. As seen in Figure 8,
higher applied voltages require shorter deposition
times to make filters with filtration efficiencies > 99%.
The local maximum filtration quality factor was
0.06 Pa1, while the maximum quality factor for all
was 0.027 Pa1. Commercial filters have been found
(a)
1 0.1
CA15_5_8_1
0.8
0 CA15_15_8_1
Efficiency [-]
CA15_15_8_
0.6
10
CA15_30_8_
0.4 10
CA15_5_10_
0.2 10
0 0
10 100 1000
Parcel size [nm]
Figure 8. (a) Filtration efficiency and (b) quality factor depending on applied voltage.
11
to have a quality factor as high as 0.007 Pa1.30 Thus,
the nanofibers in this study outperformed these com-
mercial products.
Figure 8 illustrates the relationship between the fil-
tration efficiency and the pressure drop. The pressure
drop increased so much for 10 kV that the rise of the
filtration efficiency was negligible. From the results in
Table 4, we knew that a higher voltage led to a greater
average fiber diameter. It also led to the increase in
pressure drop. The same was true for 12 kV, which
showed a sharp increase in filtration efficiency but a
relatively small increase in filter quality factor. On the
other hand, the increase in the filtration efficiency of the
thinner fibers for 8 kV led to an equivalent increase of
the filtration quality factor. All these results suggest
that the filter with thicker fibers may be more efficient
in capturing nanoparticles, but they also have
high solidity, which led to a high pressure drop.
Furthermore, the slipping effect reduces the pressure
drop for finer fibers. This finding matches the theory
described by the models summarized by Hinds.29,31
Notable differences are samples CA15_5_12_10 and
CA15_15_12_10, which show consistently high filtra-
tion efficiencies with only a minor dip at 100 nm.
Figures 6 to 8 also show the effects of deposition
time on filtration performance. The circles, squares
and rectangles stand for short, medium and long depos-
ition times, respectively. All three graphs show that a
short deposition time corresponds to low filtration effi-
ciencies, while a long deposition time corresponds to
high filtration efficiencies. The higher filtration effi-
ciency is attributed to the thicker filter resulting from
a longer electrospinning time. However, as expected,
the increased amounts of deposited fibers lead to a
higher pressure drop. Nonetheless, the overall filtration
quality is still the highest for all samples made with a
long deposition time.
(b)
CA15_5_8_10
Quality Factor [Pa-1]
0.08
0.06 0
CA15_30_8_1
0.04 0
0.02
10 100 1000
Parcel size [nm]
12
Conclusions
The results of this paper show that applied voltage, tip-
to-collector distance, CA concentration and deposition
time in electrospinning affects the size of the resulting
nanofibers and the consequent filtration quality factors.
With CA concentrations of 15% or less, beads were
formed during the electrospinning process, regardless
of other electrospinning parameters. A higher CA con-
centration resulted in larger fiber diameters, likely due
to the increased viscosity.
Furthermore, a longer tip-to-collector distance and
higher applied voltage also led to larger fiber diameters.
The impacts of the tip-to-collector distance and voltage
are comparable with respect to the increase in fiber
diameter.
Finally, the larger the fiber diameters the higher the
filtration efficiency. The increased filtration efficiency
did not necessarily mean a higher filtration quality
factor, especially for filters obtained via a longer depos-
ition time. The maximum filtration efficiency was
99.8% for filters made with an overall deposition time
of 30 minutes. The maximum filter quality factor was
0.05 Pa1 for a filter corresponding to a CA concentra-
tion of 20 wt.%, a tip-to-collector distance of 12.5 cm, a
voltage of 8 kV and a deposition time of 5 minutes.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with
respect to the research, authorship and/or publication of this
article.
Funding
The authors disclosed receipt of the following financial
support for the research, authorship, and/or publication of
this article: This work was supported by the Natural
Sciences and Engineering Research Council of Canada,
the Canadian Foundation for Innovation, the International
S&T Cooperation Program of China (Grant Number
2015DFG61910) and the OntarioChina Research and
Innovation Fund (OCRIF-2014).
ORCID iD
Zhongchao Tan http://orcid.org/0000-0001-7922-8291
References
1. Fatema UK, Uddin AJ, Uemura K, et al. Fabrication of
carbon fibers from electrospun poly (vinyl alcohol) nano-
fibers. Text Res J 2011; 81: 659–672.
2. Nayak R, Padhye R, Kyratzis IL, et al. Recent advances in
nanofibre fabrication techniques. Text Res J 2012; 82:
129–147.
3. Lee S and Obendorf SK. Use of electrospun nanofiber web
for protective textile materials as barriers to liquid pene-
tration. Text Res J 2007; 77: 696–702.
Textile Research Journal 0(00)
4. Cetiner S, Kalaoglu F, Karakas H, et al. Electrospun
nanofibers of polypyrrole-poly (acrylonitrile-co-vinyl
acetate). Text Res J 2010; 80: 1784–1792.
5. Bagherzadeh R, Latifi M, Najar SS, et al. Transport
properties of multi-layer fabric based on electrospun
nanofiber mats as a breathable barrier textile material.
Text Res J 2012; 82: 70–76.
6. Nayak R, Padhye R, Kyratzis IL, et al. Effect of viscosity
and electrical conductivity on the morphology and fiber
diameter in melt electrospinning of polypropylene. Text
Res J 2013; 83: 606–617.
7. Kang W, Cheng B, Li Q, et al. A new method for pre-
paring alumina nanofibers by electrospinning technology.
Text Res J 2011; 81: 148–155.
8. Liu W and Adanur S. Properties of electrospun polyacry-
lonitrile membranes and chemically-activated carbon
nanofibers. Text Res J 2010; 80: 124–134.
9. Hekmati AH, Rashidi A, Ghazisaeidi R, et al. Effect of
needle length, electrospinning distance, and solution
concentration on morphological properties of polya-
mide-6 electrospun nanowebs. Text Res J 2013; 83:
1452–1466.
10. Gowri S, Almeida L, Amorim T, et al. Polymer nano-
composites for multifunctional finishing of textiles -
A review. Text Res J 2010; 80: 1290–1306.
11. Angammana CJ. A study of the effects of solution and
process parameters on the electrospinning process and
nanofibre morphology. PhD Thesis, University of
Waterloo, Canada, 2011.
12. Liang H-W, Wang L, Chen P-Y, et al. Carbonaceous
nanofiber membranes for selective filtration and separ-
ation of nanoparticles. Adv Mater 2010; 22: 4691–4695.
13. Wang C-S and Otani Y. Removal of nanoparticles from
gas streams by fibrous filters: A review. Ind Eng Chem
Res 2012; 52: 5–17.
14. Givehchi R. Filtration of NaCl and WOx nanoparticles
using wire screens and nanofibrous filters. PhD Thesis,
University of Waterloo, Canada, 2016.
15. Givehchi R, Li Q and Tan Z. Quality factors of PVA
nanofibrous filters for airborne particles in the size
range of 10–125 nm. Fuel 2016; 181: 1273–1280.
16. Givehchi R and Tan Z. The effect of capillary force on
airborne nanoparticle filtration. J Aerosol Sci 2015; 83:
12–24.
17. Ding B, Kimura E, Sato T, et al. Fabrication of blend
biodegradable nanofibrous nonwoven mats via multi-jet
electrospinning. Polymer 2004; 45: 1895–1902.
18. Satyanarayana KG, Arizaga GG and Wypych F.
Biodegradable composites based on lignocellulosic
fibers—An overview. Prog Polym Sci 2009; 34: 982–1021.
19. Buchanan CM, Gardner RM and Komarek RJ. Aerobic
biodegradation of cellulose acetate. J Appl Polym Sci
1993; 47: 1709–1719.
20. Christoforou T and Doumanidis C. Biodegradable cellu-
lose acetate nanofiber fabrication via electrospinning.
J Nanosci Nanotechnol 2010; 10: 6226–6233.
21. Chattopadhyay S, Hatton TA and Rutledge GC. Aerosol
filtration using electrospun cellulose acetate fibers.
J Mater Sci 2016; 51: 204–217.
 Ahne et al.
22. Matulevicius J, Kliucininkas L, Prasauskas T, et al. The
comparative study of aerosol filtration by electrospun
polyamide, polyvinyl acetate, polyacrylonitrile and cellu-
lose acetate nanofiber media. J Aerosol Sci 2016; 92:
27–37.
23. Liu H and Hsieh YL. Ultrafine fibrous cellulose mem-
branes from electrospinning of cellulose acetate. J Polym
Sci B Polym Phys 2002; 40: 2119–2129.
24. Liu H and Tang C. Electrospinning of cellulose acetate in
solvent mixture N, N-dimethylacetamide (DMAc)/acet-
one. Polym J 2007; 39: 65–72.
25. Shenoy SL, Bates WB, Frisch HL, et al. Role of chain
entanglements on fiber formation during electrospinning
of polymer solutions: Good solvent, non-specific poly-
mer–polymer interaction limit. Polymer 2005; 46:
3372–3384.
26. Picq D. Development of electrospun cellulose acetate
nanofiber-based membranes for filtration application.
View publication stats
13
Master’s Thesis, Luleà University of Technology,
Sweden, 2015.
27. Bhardwaj N and Kundu SC. Electrospinning: A fascinat-
ing fiber fabrication technique. Biotechnol Adv 2010; 28:
325–347.
28. Son WK, Youk JH, Lee TS, et al. Electrospinning of
ultrafine cellulose acetate fibers: Studies of a new solvent
system and deacetylation of ultrafine cellulose acetate
fibers. J Polym Sci B Polym Phys 2004; 42: 5–11.
29. Hinds WC. Aerosol technology: Properties, behavior, and
measurement of airborne particles. Hoboken, NJ: John
Wiley & Sons, 2012.
30. Liu C, Hsu PC, Lee HW, et al. Transparent air filter for
high-efficiency PM 2.5 capture. Nat Commun 2015; 6:
6205.
31. Tan Z. Air pollution and greenhouse gases: From basic
concepts to engineering applications for air emission con-
trol. Singapore: Springer, 2014.