Separation and Purification Technology 175 (2017) 79–86

Contents lists available at ScienceDirect

Separation and Purification Technology

journal homepage: www.elsevier.com/locate/seppur

Characterization and evaluate the efficiency of different filter media in

removing nanoparticles
A.C.C. Bortolassi ⇑, V.G. Guerra, M.L. Aguiar
Federal University of São Carlos - UFSCar, Department of Chemical Engineering, Rodovia Washington Luiz, km 235, 13565-905 São Carlos, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Nowadays, removing nanoparticles in a filtration using fibrous filter media become a technique very use-
Received 30 June 2016 ful. There are a lot of applications of these filters which can be found in air conditioner and clean rooms.
Received in revised form 26 October 2016 Filter media used in this work have very similar characteristics and among them can be cited heteroge-
Accepted 9 November 2016
neous distribution and small diameter fibers. There are little studies about the performance of fibrous fil-
Available online 10 November 2016
ter media in removing nanoparticles. So, the aim of this work was evaluate the fiber diameter,
permeability and efficiency of HEPA (High Efficiency Particulate Air Filter) filters with glass and quartz
Keywords:
fibers in 5 g/L sodium chloride solution by the technique of electric mobility. It was possible to simulate
Efficiency
Filter
a contamination of nanoparticles through the experimental system in a stream of ultrapure air using sur-
Air filtration face speed 0.05 m/s and flow rate 1.59 L/min, which was done the particle count before and after passing
Nanoparticle through the filter media. HEPA_2 filter showed the lower permeability and the HEPA_1 filter the higher.
Permeability In general, both filter media were very efficient in removing nanoparticles during the filtration because of
the fibers distribution is heterogeneous. But HEPA_2 filter showed lower penetration of particles in the
range of 7–289 nm comparing to the other filters.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction For this purpose, studies of processes capable of removing

One side there is the concern with the control particulates with
small dimensions and in other cases there is the advancement of
technological applications involving nanoparticles. According to
Aitken et al. [1], ‘‘nanoparticles” are particles with 0.1 lm
(100 nm) dimension or less. For instance, combustion particles
are usually in the 10–50 nm size range, but when they combine
with other particles and agglomerate form larger particulates.
The agglomerate particles may be broken down into smaller parti-
cles and release into air [2].
New researches aim to avoid pollutants affect the environment
conciliating modernization and the industry grow without seri-
ously harming the ecosystem. One way to avoid this is to find a
way to hinder the arrival of nanoparticles to the ecosystem.
Several studies in epidemiology and toxicology say that as smal-
ler as the size of the particles exposed in air more will be the occur-
rence of cardio-respiratory diseases [3]. According to EPA [4], this
is explained by the fact that particles smaller than 10 lm are gen-
erally not removed by the upper respiratory system and, therefore,
eventually affect the pulmonary alveoli and blood flow.
impurities contained in the air stream are made. Filtration is one
of the most commonly processes used to capture particles from a
gas stream. This process occurs by passing the aerosol through
the filter medium in which the particles are being deposited on
the filter surface. This is a process with high efficiency, easy oper-
ation, flexibility and economy [5].
The fibers of the filter media can be classified as microfibers and
nanofibers. However, some authors define ‘‘nanofibers” as fibers
with diameter less than 1 lm [6,7]. But in this work we will con-
sider the fibers used in the filter media as microfibers because they
are measured approximately 0.5 lm.
HEPA filters with glass and quartz fibers are between the
fibrous filter media that have been much used due to high effi-
ciency of removing particles and can be used in air conditioner.
These filters have similar characteristics but it isn’t known the
most worth among them. Thus, this study aims to evaluate the
fiber diameter, permeability and efficiency of different fibrous filter
media in removing nanoparticles using sodium chloride solution.
1.1. Fiber diameter

⇑ Corresponding author. Ascertain fiber diameter in order to analyse the filter medium is
E-mail addresses: anabortolassi@hotmail.com (A.C.C. Bortolassi), vadila@ufscar. very important due to the possibility of particles’ retention on the
br (V.G. Guerra), mlaguiar@ufscar.br (M.L. Aguiar). filter mainly when the fiber diameter is small.

http://dx.doi.org/10.1016/j.seppur.2016.11.010
1383-5866/Ó 2016 Elsevier B.V. All rights reserved.
80 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86

Fig. 1. SEM image of HEPA_1 filter.

Fig. 2. Determination of fiber diameter of HEPA_2 filter SEM - FEG image using Image Pro Plus Program.

Samples of the filter media were prepared in the Structural
Characterization Laboratory DEMa at Federal University of São Car-
los and analyzed by scanning electron microscope (SEM). It was
possible to analyze them on a personal computer using the Image
Pro Plus version 7.0 program after the images are generated by
SEM-FEG and thus obtain the mean diameter of fibers of the filter
media.
Images were divided into 12 equal vertical portions in order to
obtain a large amount of fiber diameters to calculate the average
diameter of each filter medium, as seen in Fig. 1.
Method of analysis consists in observing the width of the fiber
at each intersection with the vertical lines in blue,1 as can be seen
1
For interpretation of color in Fig. 2, the reader is referred to the web version of
this article.
in Fig. 2. Thus, the program operator traced straight perpendicular to
the fibers that cross the vertical lines being represented by yellow
lines and the program will automatically calculate this values, which
are the fibers diameters represented by written in green. Analyzes
were performed on all blue lines from left to right, thus formed a
table of the entire calculated fibers diameter.
Six images were analyzed at different positions from each filter
medium. It was measured approximately 50 fibers diameter in
each image. It was used Eq. (1) to determine the average fibers
diameter of HEPA filters. As we know these filters do not have
homogeneous fibers.
P
ni dfi
Df ¼ ð1Þ
N
 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86 81

ni is the number of fibers of determinate diameter, dfi is the fiber DP l

¼ :v s ð3Þ
diameter in determinated band and N is the total number of fibers. L k1
Van Osdell et al. [8] and Podgórski et al. [9], who used this
Thus, the permeability of the porous filter media can be
method to determine the HEPA filter diameters, highlighted this
obtained by Darcy’s equation. This equation evaluates the flow of
as a simple method but hard-working and impossible to represent
fluids in porous filter media relating pressure drop values with
the real fiber diameter.
superficial velocity.

1.2. Permeability
1.3. Collection efficiency
According to Hung and Leung [10] there are two ways to
The collection efficiency is experimentally obtained through the
improve the filter’s quality. One is to make it more efficient in fil-
technique of electric mobility in which the amount of particles is
tering out aerosol to increase the filtration efficiency and another
calculated before and after passing through the filter medium
one is to make it more permeable to reduce the pressure drop.
determined by the equation:
Forchheimer’s equation is used to evaluate the fluid flow in por-
ous media. The first term of this equation refers to purely viscous Co  Ce
effects and the second term to the inertial effects, according to E¼ ð4Þ
C0
equation [11]:
which Co and Ce represent the concentration of particles before and
DP l qg
¼ :v s þ v 2s ð2Þ after passing through the filter media, respectively.
L k1 k2
which L represents the thickness of the filter media, l is the viscos- 2. Materials and methods
ity of the fluid, k1 e k2 are constants of permeability of filter media,
qg represents the density of the gas and v s is the superficial velocity. 2.1. Particulate material
DP is the pressure drop that is the difference between the inlet and
outlet pressure of the filter during the passage of the air stream. It was used solution of 5 g/L sodium chloride (NaCl) to generate
In this work it was used low filtration velocity, so the second nanoparticles used in the filtration tests. The sodium chloride
term of Eq. (2) can be neglected and the Eq. (3) may be used [11]: (CHEMIS) was pure whose density was 2.17 g/cm3. The commercial

Fig. 3. SEM images of the filter media (a) HEPA_1, (b) HEPA_2, and (c) HEPA_3.
82 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86
inhaler was used to generate particles operated at a pressure of
17 kPa. It was possible to measure particles in the range between
7.4 and 289 nm using a particle analyzer by electric mobility
(SMPS) linked to the filtration line at Environment Control Lab I
of Chemical Engineering Department of Federal University of São
Carlos (UFSCar).
2.2. Filter media
HEPA (High Efficiency Particulate Air Filter) filters with glass
and micro quartz fibers were fibrous filters used in filtration pro-
cess of nanoparticles. HEPA_1 filter classified as H14 (99,995%
MPPS Efficiency) was made of glass fiber and obtained by FILTRA-
COM LTDA company. HEPA_2 filter (EPM2000) was made of glass
fibers and HEPA_3 filter (QM-A) was made of micro quartz fibers
and both filters were obtained by enterprise ENERGY Industry
and Trade LTDA. All of the filter media had 5.3 cm2 filtration areas.
Fig. 3 shows HEPA_1, HEPA_2 and HEPA_3 filters images
obtained by scanning electron microscope (SEM). It can be seen
that filter media have microscale diameters and heterogeneous
fibers which don’t have the same diameters distributions. This is
a feature that makes them very effective in removing particles.
Generally, these filters are low cost and their structure is very
similar.
Because of the little difference between filter media it was nec-
essary to analyze SEM images using Image Pro Plus 7 program. It
was possible to evaluate the mean fiber diameter and filter thick-
ness. Beyond these specifications it’s possible to see in Table 1
the pressure drop across clean filter and filter material.
Table 1
Filter specifications.
Filters HEPA_1 HEPA_2
Mean fiber diameter (lm) 0.59 0.37
Filter thickness (lm) 383 ± 28 249 ± 21
Pressure drop DP (Pa) across 269 ± 16 397 ± 12
clean filter at 5 cm/s
Filter material Glass Glass borosilicate
fibers without aglutinantes
Mean fiber diameter of HEPA_2 filter (Table 1) is almost close
comparing to filters produced by Yun et al. [12] that it was
0.39 lm for a beaded-nanofiber filter prepared by electrospinning.
2.3. Experimental unit
Fig. 4 shows the experimental unit which consists of an air com-
pressor (Shulz), air purification filters (Model A917A-8104N-000
and 0A0-000), particles generator (NS commercial inhaler, model
I- 205), diffusion dryer (Norgren), Kriptônio and Americium neu-
tralizing source (TSI Model 3054), filter apparatus, flowmeter size
3 (Gilmont) and SMPS device formed by electrostatic classifier
(TSI 3080), differential mobility analyzer and ultrafine particles
counter (TSI 3776) at Environment Control Lab I of Chemical Engi-
neering Department of Federal University of São Carlos (UFSCar), as
used in Barros et al. [13].
The experimental unit was used to execute the permeability
and efficiency experiments but in permeability experiments it’s
not necessary to use particles generator, diffusion dryer and SMPS
device because only clean gas stream without particles go through
filtration line.
2.4. Experimental procedure
First, filter media permeability experiments were performed
and the pressure drop was measured using a digital manometer
(TSI 8385A). The filtration tests were performed maintaining
constant the surface speed 5 cm/s, flow rate 1.59 L/min and fil-
tration area of 5.3 cm2. It was possible to obtain the distribution
of particle diameters before and after the filter medium at
the beginning of filtration, making it possible to calculate the
efficiency of the filter media using particle analyzer by electric
HEPA_3
mobility (SMPS) and Eq. (3). The experiments were done in
triplicate.
0.62
Particle count time before and after the filter was equal (315 s).
521 ± 33
418 ± 11 So, each experiment lasted 10 min approximately and the total
time was 30 min. After this, the pump was turned off and started
Micro to count one hour of filtration and the efficiency was measured
quartz
again after one hour of the beginning of the experiment.
Fig. 4. Experimental unit.
 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86
3. Results and discussion
3.1. Fiber diameter
It was possible to obtain the weighted average value of fiber
diameters for each filter media performing the procedure in Sec-
tion 1.1. It’s possible to verify that the higher fibers number is in
the range of smallest fibers diameter (0.075–0.30 lm) and there
are few fibers in the range of 1.94–2.35 lm, which indicates that
the HEPA_1 filter has small fibers diameters and this makes it
highly efficient. It’s possible to verify that major of HEPA_2 filter
is formed by small fibers because the highest amount of fibers is
in range of 0.075–0.30 lm. HEPA_3 filter shows the highest fibers
number is in the range of the smallest fibers diameter (0.075–
0.30 lm) and there are few fibers in the range of 1.94–2.35 lm.
Fig. 5 shows the plot of distribution of fibers using the values of
HEPA_1 filter and it’s possible to note that the distribution of fibers
realized in this work was almost similar to Steffens and Coury [14].
Fig. 6 shows a comparison of distribution of fibers among
HEPA_1, HEPA_2 and HEPA_3 filters. It’s possible to note that the
fibers distribution curve of these filters have a similar behave. Only
one difference is that HEPA_2 filter has a larger amount of small
fibers.
It was possible to evaluate the average fibers diameter of filter
media using Eq. (1) and is presented in Table 2.
As we can see in Table 2, HEPA_2 filter showed the smallest
average fibers diameter compared to the other filter media. Higher
is the capture of particles that pass through the filter medium
when the fiber diameter is smaller because the geometry of the
Fig. 5. Fibers distribution of HEPA_1 filter.
Fig. 6. Fibers distribution HEPA_1, HEPA_2 and HEPA_3 filters.
83
Table 2
Average fibers diameter of filter media.
Filter media HEPA_1 HEPA_2 HEPA_3
Fibers diameter (lm) 0.59 0.37 0.62
fibers promotes the deposition of particles. The difference between
HEPA_1 and HEPA_2 filter is 37% and between HEPA_2 and HEPA_3
is 40%.
3.2. Permeability
Permeability constant of cleaned filter was obtained from
experimental data using Darcy’s equation (Eq. (3)). The superficial
velocity was varied from 1 to 15 cm/s and for each 1 cm/s the pres-
sure drop was measured using a digital manometer. It was deter-
mined the Darcian permeability (k1) for all filter media through
the curves of pressure drop/thickness (DP/L) as function of the
superficial gas velocity (v s ). It was made the average of five perme-
ability experiments of filter media showed in Fig. 7. The deviations
are very small that’s why can’t be noticeable on the graph.
It’s possible to see that the pressure drop was higher for HEPA_2
filter (approximately 38 kPa) and the curves behavior was similar.
So, it’s possible to know that pressure drop is inversely propor-
tional of permeability constant using Darcy’s equation. The perme-
ability constants values for the filter media evaluated are shown in
Table 3 which were determined from a linear fit (Y = ax) of Fig. 4.
It can be seen that the permeability constant value for HEPA_2
filter was the smaller when compared to HEPA_1 and HEPA_3 filter
(Table 3). So, the particles have a greater difficulty in going through
the HEPA_2 filter and are retained easier. But the permeability con-
stants are close and the variation was approximately 15% between
them.
3.3. Nanoparticles distribution
Nanoparticles distributions of HEPA_2 filter generated from
sodium chloride solution can be seen in Fig. 8. All filter media
achieved the same standard particle size distribution curve and
Fig. 7. DP/L as a function of superficial velocity.
Table 3
Permeability.
Filter Permeability (cm2)
HEPA_1 9.62E09 ± 1.78E12
HEPA_2 7.28E09 ± 2.73E13
HEPA_3 8.52E09 ± 3.87E15
84 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86

Fig. 8. Nanoparticles distribution of HEPA_2 filter.

Fig. 11. Media collection efficiency of particles deposition in the beginning of

have been detected particles in the range of 7.4–289 nm using par- filtration using HEPA_2 filter.
ticle analyzer of electric mobility (SMPS) coupled to the filtration
line.

3.4. Efficiency

Fig. 9 shows the experimental collection efficiency over the

deposition of nanoparticles in HEPA_1 filter using 5 g/L of NaCl
solution and 5 cm/s of superficial velocity. Experiments counted
particles before and after the filter media at the beginning of the
filtration and it was possible to obtain the initial media of the effi-
ciency (Fig. 9).

Fig. 12. Media collection efficiency of particles deposition after one hour of
filtration using HEPA_2 filter.

Fig. 9. Media collection efficiency of particles deposition in the beginning of

filtration using HEPA_1 filter.

Fig. 13. Media collection efficiency of particles deposition in the beginning of

filtration using HEPA_3 filter.

After one and a half hour approximately (filtration and particle

Fig. 10. Media collection efficiency of particles deposition after one hour of
filtration using HEPA_1 filter.
count) were performed three more experiments counting the par-
ticles before and after the filter medium (Fig. 10). The whole exper-
iment (three experiments in the beginning of filtration and three
experiments in the end of filtration) was approximately two hours.
Efficiency higher than 99.9% can be seen in Fig. 10. Particles
don’t pass through the filter media and were deposited in the filter.
 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86 85

Fig. 14. Media collection efficiency of particles deposition after one hour of
filtration using HEPA_3 filter.

This probably can be explained by the fact that particles deposited

in the first minutes of filtration reduce the empty space between
the fibers and help retain new particles. Barros et al. [13] also
observed the filter efficiency higher than 99%.
The experiment was conducted with HEPA_2 filter after per-
forming with HEPA_1 filter. Initial media efficiency (Fig. 11) and
final media efficiency (Fig. 12) was also evaluated.
Graphics behaved as HEPA_1 filter in the beginning and in the
end of filtration for the HEPA_2 filter. But there was a small
amount of particles that were able to cross the filter media with
efficiency greater than 99%.
Using the same experimental parameters of these experiments
tests were performed with HEPA_3 filter and are presented in
Figs. 13 and 14. Fig. 15. Media collection efficiency of particles deposition of filtration using
HEPA_1, HEPA_2 and HEPA_3 filter (a) Initial time of filtration and (b) After one
It is possible to analyze that a slight penetration of particles in hour of filtration.
HEPA_3 filter in the beginning of the experiment (Fig. 13) but par-
ticle penetration decreased during the filtration.
The same behavior in the graphs was obtained by the three fil- a smallest amount of particles that were able to cross the filter
ters analyzed (HEPA_1, HEPA_2, HEPA_3). It can be justified media (Fig. 15). Probably it occurred by the fact that this filter is
because these filters have the same characteristics. They are related to the lowest permeability. It’s also possible to notice that
fibrous filters composed of small thickness and diameter of fibers. the deviation in the beginning of the filtration is higher than after
But HEPA_2 filter highlighted among the others because there was one hour of filtration.

Fig. 16. SEM image of HEPA_1 filter with nanoparticles.

86 A.C.C. Bortolassi et al. / Separation and Purification Technology 175 (2017) 79–86
It was also possible to see in Fig. 15 that after one hour of filtra-
tion the efficiency was better than in the beginning of the filtration.
This probably occurred because of the deposition of particles on
the filter media. The deposition of particles decreases the empty
space between the fibers and helps to increase the efficiency
(Fig. 16).
Steffens and Coury [14] evaluated that the efficiency of HEPA
filter were above of 99.999% for the whole size range (8.5–
94.8 nm) for velocities smaller than 9.0 cm/s. So, it’s noticeable
the high efficiencies attainable by the HEPA filter for the smaller
particles. Their HEPA filter is closed to those used in this work.
4. Conclusions
Evaluating the fiber diameter of the filter media was possible to
observe that HEPA_2 filter used in this work has the smaller fiber
diameter than HEPA_1 and HEPA_3 filter. And this is a good quality
in a filter media to retain particles from an air stream.
According to the results obtained, the lowest permeability con-
stant was obtained by the HEPA_2 filter but it can be considered
that the values were very close to the HEPA_1 and HEPA_3 filters;
difference between them is 15%.
It was noted that the entire filter media reached above 99% effi-
ciency for particles in the 7–289 nm range when it was evaluated
the performance of the filter media. It was also possible to note
that the efficiency was very high even the filter media were com-
posed by microfibers. Then, does not necessary imply that HEPA_2
filter is ‘‘absolutely better” than the other filters but it highlighted
between the others filters when it was analyzed the cost benefit.
Acknowledgements
The authors would like to thank the National Counsel of Tech-
nological and Scientific Development (CNPq) for its financial
support.
References
[1] R.J. Aitken, K.S. Creely, C.L. Tran, Nanoparticles: An Occupational Hygiene
Review, HSE Books RR274, Sudbury, UK, 2004.
[2] R.S. Barhate, S. Ramakrishna, Nanofibrous filtering media: filtration problems
and solutions from tiny materials, J. Membr. Sci. 296 (2007) 1–8, http://dx.doi.
org/10.1155/2015/247471.
[3] H.L. Huang, Y.C. Huang, D.M. Wang, Aerosol filtration efficiency of Teflon
fibrous filters, in: Abstracts of the European Aerosol Conference, 2004, pp.
973–974.
[4] EPA (United State Environmental Protection Agency), 2015. <www.epa.gov>.
(Accessed on September).
[5] R.P. Donovan, Fabric Filtration for Combustion Sources, Marcel dekker inc.,
New York, 1985, p. 426.
[6] T. Grafe, M. Gogins, M. Barris, J. Schaefer, R. Canepa, Nanofibers in filtration
applications in transportation, in: Filtration 2001 International Conference and
Exposition of the INDA, Chicago, Illinois, 3–5 December, 2001.
[7] K. Graham, M. Ouyang, T. Raether, T. Grafe, B. McDonald, P. Knauf, Polymeric
nanofibers in air filtration applications, in: Fifteenth Annual Technical
Conference & Expo of the American Filtration & Separations Society,
Galveston, Texas, 9–12 April, 2002.
[8] D.W. Van Osdell, B.Y.H. Liu, K.L. Rubow, D.Y.H. Pui, Experimental study of
submicrometer and ultrafine particle penetration and pressure drop for high
efficiency filters, Aerosol Sci. Technol. 12 (1990) 911–925, http://dx.doi.org/
10.1080/02786829008959403.
[9] A. Podgórski, A. Balazy, L. Gradón, Application of nanofibers to improve the
filtration efficiency of the most penetrating aerosol particles in fibrous filters,
Chem. Eng. Sci. 61 (2006) 6804–6815, http://dx.doi.org/10.1016/j.
ces.2006.07.022.
[10] C. Hung, W.W. Leung, Filtration of nano-aerosol using nanofiber filter under
low Peclet number and transitional flow regime, Sep. Purif. Technol. 79 (2011)
34–42, http://dx.doi.org/10.1016/j.seppur.2011.03.008.
[11] M.D.M. Innocentini, F. Sepulveda, F. Ortega, Permeability, in: M. Sheffler, P.
Colombo (Eds.), Cellular Ceramics: Estructure, Manufacturing, Properties and
Applications, 2005, pp. 313–340, http://dx.doi.org/10.1002/3527606696.ch4b.
[12] K.M. Yun, A.B. Suryamas, F. Iskandar, L. Bao, H. Niinuma, K. Okuyama,
Morphology optimization of polymer nanofiber for applications in aerosol
particle filtration, Sep. Purif. Technol. 75 (2010) 340–345, http://dx.doi.org/
10.1016/j.seppur.2010.09.002.
[13] P.M. Barros, S.S.R. Cirqueira, M.L. Aguiar, Evaluation of the deposition of
nanoparticles in fibrous, Mater. Sci. Forum 802 (2014) 174–179, http://dx.doi.
org/10.4028/www.scientific.net/MSF.802.174.
[14] J. Steffens, J.R. Coury, Collection efficiency of fiber filters operating on the
removal of nano-sized aerosol particles II. Heterogeneous fibers, Sep. Purif.
Technol. 58 (2007) 106–112, http://dx.doi.org/10.1016/j.seppur.2007.07.011.