You are on page 1of 10

Aerosol Science and Technology

ISSN: 0278-6826 (Print) 1521-7388 (Online) Journal homepage: http://www.tandfonline.com/loi/uast20

Novel nonbouncing PM2.5 impactor modified from


well impactor ninety-six

Thi-Cuc Le & Chuen-Jinn Tsai

To cite this article: Thi-Cuc Le & Chuen-Jinn Tsai (2017): Novel nonbouncing PM2.5
impactor modified from well impactor ninety-six, Aerosol Science and Technology, DOI:
10.1080/02786826.2017.1341621

To link to this article: http://dx.doi.org/10.1080/02786826.2017.1341621

Accepted author version posted online: 12


Jun 2017.
Published online: 12 Jun 2017.

Submit your article to this journal

Article views: 13

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=uast20

Download by: [National Chiao Tung University 國國國國國國] Date: 29 June 2017, At: 18:01
AEROSOL SCIENCE AND TECHNOLOGY
https://doi.org/10.1080/02786826.2017.1341621

Novel nonbouncing PM2.5 impactor modified from well impactor ninety-six


Thi-Cuc Le and Chuen-Jinn Tsai
Institute of Environmental Engineering, National Chiao Tung University, Hsinchu, Taiwan

ABSTRACT ARTICLE HISTORY


The impaction well containing an oil-soaked glass fiber (GF) filter of the well impactor ninety-six Received 7 March 2017
(WINS) needs to be replaced every five sampling days to avoid particle overloading which affects Accepted 6 June 2017
the cutoff diameter (dpa50) and the sampling accuracy. This study developed a novel modified WINS EDITOR
(M-WINS) which uses water to wash the wetted GF filter substrate clean continuously to eliminate Kihong Park
the particle overloading effect and extend the service life of the impactor without the need of
impaction well cleaning and replacement of the oil-soaked GF filter. The laboratory test showed
that dpa50 of the M-WINS only varied slightly from 2.44 to 2.49 mm while dpa50 of the WINS
decreased from 2.44 to 2.05 mm when the loaded particle mass increased from 0–6 mg. The field
test showed that dpa50 and the sampling bias of the novel M-WINS met the USEPA requirements
(dpa50 D 2.5 § 0.2 mm and sampling bias <§5%) while the WINS without regular replacement of
the well showed the decrease in dpa50 down to 2.12 mm and the increase in the negative sampling
bias to ¡20% after 17 continuous sampling days. Therefore, the current M-WINS can be used as the
PM2.5 inlet for unattended, continuous sampling for a long time with a very good sampling accuracy.

1. Introduction have sharp and nearly the same collection efficiency


curves if Re ranges between 500 and 3000 (Marple and
Inertia impactors are often used as size-selective inlets in
Rubow 1986). Gravitational force of particles increases
ambient PM10 (particulate matter <10 mm in aerody-
the particle collection efficiency when Re is below about
namic diameter) and PM2.5 (particulate matter <2.5 mm
1500 (Huang and Tsai 2001). Other parameters influenc-
in aerodynamic diameter) sampling systems (Watson
ing the collection efficiency curves are the impaction
and Chow 2011). The performance of inertial impactors
plate diameter, the particle density (Huang and Tsai
is characterized by the cutoff aerodynamic diameter
2002) and the design of impaction surfaces (Tsai and
(dpa50), which is the aerodynamic diameter correspond-
Cheng 1995; Tsai and Lin 2000).
ing to 50% collection efficiency, and the sharpness or
The inertial impactor is known to have problems
geometric standard deviation (GSD) of the collection
associated with particle overloading and solid particle
efficiency curve. The GSD is calculated as (Kenny et al.
bounce, which cause the shift of the cutoff diameter and
2004):
the collection efficiency curve leading to incorrect PM
sffiffiffiffiffiffiffiffiffiffi measurement (Pak et al. 1992; Tsai and Cheng 1995;
dpa84 Kenny et al. 2000; Vanderpool et al. 2001b). To minimize
GSD D [1]
dpa16 solid particle bounce, the impaction plate is usually
coated with vacuum grease, vacuum or silicone oil
where dpa84 and dpa16 are the aerodynamic diameters cor- (Turner and Hering 1987; Pak et al. 1992; Tsai et al.
responding to the collection efficiencies of 84% and 16%, 2012) or covered by a glass-fiber (GF) filter saturated in
respectively. The most important parameter that governs water (Dunbar et al. 2005). Use of rough impaction
the collection efficiency is the Stokes number (Stk), plates (Marjam€aki and Keskinen 2004), porous sub-
which is the ratio of the particle’s stopping distance to strates (Huang et al. 2001; Marjam€aki and Keskinen
the characteristic dimension of the nozzle. Jet Reynolds 2004; Huang et al. 2005), or specially designed substrates
number (Re) is another parameter governing the flow (Tsai and Cheng 1995; Chang et al. 1999; Kim et al.
field and the collection efficiency. Inertial impactors will 2013) also help reduce particle bounce. In addition, the

CONTACT Chuen-Jinn Tsai cjtsai@mail.nctu.edu.tw Institute of Environmental Engineering, National Chiao Tung University (NCTU), No. 1001 University
Road, Hsinchu 30010, Taiwan.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uast.
© 2017 American Association for Aerosol Research
2 T.-C. LE AND C.-J. TSAI

relative humidity of incoming aerosols can be condi- well every 5 days of 24-h sampling (USEPA 1998;
tioned to a higher value to minimize particle bounce Vanderpool et al. 2001b). In comparison, dpa50 of another
(Dunbar et al. 2005; Chen et al. 2011). PM2.5 inlet called Very Sharp Cut Cyclone (VSCC) was
Solid particle bounce from the grease-coated impac- not changed and GSD increased only slightly from 1.16 to
tion plate still occurs when the plate is loaded heavily 1.2 after continuous sampling 90 days at an equivalent
with particles. The incident particles strike on previously concentration of 150 mg/m3 (Kenny et al. 2004). This is
deposited solid particles rather than the coated plate due to a much more diffused particle deposition region in
resulting in possible particle bounce (Tsai and Cheng the VSCC than the single jet deposition area in the WINS.
1995; Vanderpool et al. 2001b). To increase the particle It is important to note that the “equivalent” 90 sampling
loading capacity on the impaction substrate, the special days at the concentration of 150 mg/m3 was determined
designs of impaction plates (Tsai and Cheng 1995), oil- in the laboratory for 30 h with the ATD particle concen-
soaked filters (Peters et al. 2001a), oil-coated impaction tration of 10 mg/m3.
substrate (Turner and Hering 1987; Pak et al. 1992; Tsai This study aims to develop a novel modified WINS
et al. 2012), or rotating substrates (Marple et al. 1991; (M-WINS), which uses water to wash the impaction
Tsai et al. 2012) were used. substrate clean continuously, to extend the 5-day
The Well Impactor Ninety-Six (WINS), which is a requirement to replace the impaction well of the
single-jet, round-nozzle inertial impactor with dpa50 of WINS to a much longer period. The M-WINS uses a
2.48 mm and GSD of 1.18, consists of an upper housing wetted GF filter as the substrate through which a
containing a nozzle and a lower housing supporting an 0.8 mm hole is drilled at the center as shown in
impaction well (Peters et al. 2001a; Vanderpool et al. Figure 1. Deionized (DI) water flow is injected con-
2001a). The well, which contains a 37 mm GF soaked tinuously upward at a flow rate of 0.3 mL/min by
with 1 mL mineral oil as the substrate, is used to mini- using a syringe pump to wash off deposited particles
mize particle overloading and subsequent particle from the substrate. Washed water is drained off by
bounce-off similar to a previous design (Tsai and Cheng using another syringe pump through a hole drilled at
1995). Additionally, the well is also helpful for retaining the edge of the well. The impaction surface is slightly
any blow-off from the tip of the conical build-up that is inclined downward at 5 degrees from the center to
typical after the collection of a large amount of particles the hole to ease the water flow. The wetted GF filter
(Peters et al. 2001a). The collection efficiency curve of is used because it can combine the actions of liquid
the WINS is not affected by particle gravity as Re equals surface tension (capillary forces) and soluble particle
6000 (Huang and Tsai 2001) and there is no particle loss dissolution to prevent particle bounce and overload-
at the sidewall of the well since Stk50 and the well-to-jet ing (Dunbar et al. 2005). The GF filter is relatively
radius ratio equals 0.238 and 9.45, respectively (Hu et al. cheap, has good chemical resistance and dense fiber
2007). The WINS is designated as the inlet for an USEPA weave that retains solution effectively (Peters et al.
PM2.5 Federal Reference Method (FRM) sampler or a 2001a).
PM2.5 Federal Equivalent Method (FEM) monitor. In this article, the M-WINS was tested for the collec-
However, previous laboratory studies showed that tion efficiency and the particle overloading effect in the
dpa50 of the WINS shifted downward to 2.15 mm after laboratory. The original WINS was also tested for com-
loading aloxite powder to 4.5 mg (Kenny et al. 2000) and parison. After that, two field comparison tests were con-
to 2.21 mm after loading Arizona test dust (ATD) to ducted for 12 and 17 days, respectively, for the
16 mg, corresponding to 5 consecutive sampling days performance and particle loading effect of both the
with a negative bias of ¡5% (Vanderpool et al. 2001b).
In the filed sampling studies, dpa50 of the WINS was also
shown to reduce by 12% (2.44 mm to 2.15 mm) after 4
weeks of sampling at a parking lot, corresponding to the
estimated bias in PM2.5 concentrations of ¡8.8% (Kenny
et al. 2000), and decrease to 2.32 mm with a maximum
bias of ¡2.1% in PM2.5 concentrations after 120-h of
sampling (Vanderpool et al. 2001b). This is because that
oil wicks up the previously deposited particle layer and
accumulates particles to form a mound which reduces
the cutoff diameter.
The above studies supported USA Federal regulation Figure 1. Schematic diagram of the impaction well of the
that requires the replacement of the WINS impaction M-WINS.
AEROSOL SCIENCE AND TECHNOLOGY 3

M-WINS and the WINS using the VSCC as the main efficiency data were fitted by following equation to
reference. obtain dpa50 and GSD (Peters et al. 2001a):

" ¡"a ¡ b
hfit ð%Þ D 100%   1  !# ¡ e #
2. Experimental method dpa C dln 2e ¡ 1 ¡ c
£ 1¡ 1 C exp
2.1. Laboratory tests d

Figure 2 shows the experimental setup for determin- [3]


ing the collection efficiency curves of the M-WINS or
the WINS. The experimental method is similar to where a, b, c, d, and e are parameters. This equation
that in detail in the previous study (Peters et al. should be fitted with R2 > 0.99.
2001b). Solid ammonium fluorescein particles (AF A spreadsheet was used to perform the numerical
particles; density is 1.35 g/cm3) with the aerodynamic integration of the collection efficiency curve of the
diameter ranging from 1.5 to 5.0 mm were generated M-WINS or the WINS with the idealized ambient distri-
by a Vibrating-Orifice Aerosol Generator (VOAG, butions to estimate PM2.5 mass concentrations (Csamp),
TSI model 3450, TSI Inc., Shoreview, MN, USA) which is similar to that in the previous study (Vander-
(Vanderpool and Rubow 1998). The electrostatic pool et al. 2001b). The sampling bias in estimated PM2.5
charge of particles was neutralized by a neutralizer concentrations of the M-WINS or the WINS and the
(Kr-85, TSI model 3054) before particles enter the ideal WINS is calculated by
impactors and/or were measured. The upstream and
downstream particle concentrations of the impactors Csamp ¡ Cideal
Bias ð%Þ D £ 100% [4]
were measured by an Aerodynamic Particle Sizer Cideal
(APS, TSI model 3321, USA) to calculate the collec-
tion efficiency, hexp (%), as: where Cideal (mg/m3) is the estimated PM2.5 mass concen-
tration of the ideal WINS. Cideal equals 13.81, 34.28 and
Cin ¡ Cout 78.53 mg/m3 corresponding to idealized fine, typical and
hexp ð%Þ D £ 100% [2]
Cin coarse ambient particle size distributions, respectively
(USEPA 1998; Vanderpool et al. 2001b).
where Cin and Cout are the upstream and downstream In order to study the particle overloading of the
aerosol number concentrations (#/cm3). The collec- M-WINS or the WINS, Al2O3 powder (QF-Al-500,
tion efficiency for each particle diameter was calcu- Sipernat, Japan) was dispersed by using a Small-Scale
lated from at least six upstream and downstream Powder Disperser (SSPD, TSI model 3433) at the aerosol
measurements. The APS was calibrated first with four flow rate of 18.5 L/min. Generated particles were then
standard PSL particles (polystyrene latex, 1, 2, 3, and introduced into the 5 L stainless steel chamber and
4 mm) before the tests and the accuracy was found to diluted by 3.2 L/min clean airflow. The upstream particle
be approximately §3%. The discrete collection mass concentration was determined by filter cassette 01

Figure 2. Experimental setup of the particle collection efficiency test.


4 T.-C. LE AND C.-J. TSAI

at 5.0 L/min and remaining 16.7 L/min passed through was used to replace the original virtual impactor in the
the M-WINS or the WINS. The downstream mass con- Dichot so that only PM2.5 was sampled. A 47 mm Teflon
centration was determined by filter cassette 02 at 5.0 L/ filter (Teflo R2PL047, Pall Corp.) was used in the Parti-
min. 47 mm Teflon filters (Teflo R2PL037, Pall Corp., sol-VSCC and two 37 mm Teflon filters (Teflo R2PL037,
NY, USA) installed in cassette 01 and 02 were weighed Pall Corp.) were used in the Dichot-M-WINS and the
before and after sampling after being conditioned for at Dichot-WINS, respectively. The filter conditioning, pre-
least 24-h in an environment conditioning room where weighting, and post-weighting procedures were similar
the relative humidity and temperature were kept at to those described in the previous section.
40 § 2% and 21 § 1 C, respectively. The electrostatic In period #1, the WINS and the VSCC were cleaned
charge of the filters was eliminated by an ionizing air manually every day (i.e., clean inlets) to evaluate the per-
blower (Model CSD-0911, MEL-SEI, Japan) before the formance of the M-WINS. In period #2, the VSCC was
filters were weighted by a microbalance (Model CP2P-F, cleaned while the WINS was not cleaned during the sam-
Startorius, Germany). The precision of weighting was pling period to examine the particle loading effect of
determined to be 2.0 mg by repeated weighing for at least both the WINS and M-WINS (i.e., uncleaned inlets). In
five times. The total loaded particle mass, ML (mg), in both periods, the substrate of the M-WINS was cleaned
the WINS or the M-WINS was calculated as automatically by injecting water at the flow rate of
0.3 mL/min. After the test of period #2 was over, the
16:7 ðL=minÞ impaction wells of the M-WINS and the WINS were
ML ðmg Þ D ðMin ¡ Mout Þ £ [5]
5 ðL=minÞ tested for the cutoff diameter in the laboratory following
the method described in the previous section.
where Min and Mout are the collected particles mass of fil- In order to evaluate the sampling precision of
ter cassette 01 and 02, respectively (mg). M-WINS or the WINS, paired sample t tests were
Two syringe pumps were used to keep the impaction conducted using SPSS (Statistical Package for the Social
substrate of the M-WINS clean during the particle load- Sciences, version 22.0, SPSS Inc., Chicago, IL, USA) to
ing test at the water flow rate of 0.3 mL/min. After pre- determine whether two PM2.5 data sets measured by
liminary tests, the water flow rate was set at 0.3 mL/min two collocated pairs of samplers are statically the same
to not only wash off the deposited particles effectively (p > 0.05).
but also minimize the amount of water used. After each
loaded particle mass of 0.5, 4.0 and 6.0 mg, the M-WINS
or the WINS was tested for the particle collection effi- 3. Results and discussion
ciency curve immediately by using the APS.
3.1. Laboratory results—particle collection
efficiency
2.2. Field comparison tests
Figure 3 shows the particle collection efficiency curve of
The field comparison tests were conducted at National the WINS. The cutoff diameter equals 2.44 § 0.05 mm,
Chiao Tung University (NCTU) from Oct. 29 to Nov. which is very close to the cutoff diameters of 2.48 §
10, 2016 (period #1, 12 days) and from Nov. 15 to Dec. 0.05 mm by Peters et al. (2001a) and 2.44 § 0.03 mm by
3, 2016 (period #2, 17 days). The NCTU sampling site Kenny et al. (2000). The GSD of the collection efficiency
was located at the rooftop of the five-floor building of curve of the WINS is 1.2, indicating that the particle col-
the Institute of Environmental Engineering of NCTU, lection efficiency curve is less sharp than that in Peters
which is about 1 km away from heavy-traffic roads. Sam- et al. (2001a) (GSD D 1.18) but sharper than that in
plers were set up at a sampling platform with 2-meter Kenny et al. (2000) (GSD D 1.23). The particle collection
height inlet tubes. efficiencies of the coarse particles (2.5 to 4 mm) are lower
One Dichot (Dichotomous PM2.5 sampler, Model 241, than those obtained by Peters et al. (2001a) while the
Andersen Inc., Georgia, USA) equipped with the M- particle collection efficiencies of the fine particles are
WINS and the other Dichot equipped with the WINS slightly lower than those obtained by Kenny et al. (2000).
were collocated with a PartisolÒ -FRM 2000 sampler con- The difference in GSDs between the present study and
taining the VSCC and a TEOM-FDMS (tapered element that of Peters et al. (2001a) is mainly due to more data
oscillating microbalance with a filter dynamic measure- points acquired for coarse particles and hence better data
ment systems, Model 1405-DF) as the reference sam- fitting in the current study, although the same kind of
plers. Only the base-mode PM2.5 data of the TEOM- test particles (monodisperse AF particles) were used. The
FDMS were used for comparison without correcting for test particles used in Kenny et al. (2000) were solid, poly-
the reference-mode PM2.5. The M-WINS or the WINS disperse spherical glass microspheres with the number
AEROSOL SCIENCE AND TECHNOLOGY 5

the USEPA FRM regulation (2.5 § 0.2 mm) (USEPA


2001). The dimensions of the nozzle, which are not
changed in the M-WINS, were designed to achieve dpa50
of »2.5 mm (Peters et al. 2001a). The GSD of the collec-
tion efficiency curve of the M-WINS is 1.24, indicating
that the collection efficiency curve of the M-WINS is
slightly less sharp than that of the WINS when using
water-saturated GF filter instead of oil-soaked GF filter,
especially in the particle diameter range of 2.5 to 4 mm.
This is because the particle collection characteristics of
the impactor is notably influenced by the nature of
impaction surface (Marjam€aki and Keskinen 2004). In
this study, the silicon oil (KF96SP, Shin-Etsu Chemical
Co. Ltd., Taiwan) with the kinetic viscosity of 100 mm2/s
used in the WINS could retain particles on the impaction
substrate more effectively as compared to water with
a lower kinetic viscosity of 0.89 mm2/s used in the
M-WINS.
Figure 3. Particle collection efficiency curve of the WINS. In Figure 4, it is apparent that the M-WINS with a dry
impaction plate without adhesive coating has low particle
median diameter around 1.5 mm. These different types of collection efficiency for coarse particles due to particle
solid particles may create the differences in the GSDs. bounce. It is observed that particles deposit in the inner
Despite the small differences in the cutoff diameters and wall and some are trapped in the antispill ring of the
GSDs, the estimated biases of the WINS in the present impaction well after bouncing from the plate. This parti-
study are less than §1.1%, which still meet cle loss is considered a part of the collection efficiency
the criteria specified in the USEPA FRM regulation determined by the APS, which explains the steady-state
(<§5%) (USEPA 2001). collection efficiency of 80% for particles larger than
Figure 4 shows the collection efficiency curve of the 4.0 mm. When the plate is wetted by water, solid particle
M-WINS with the wetted GF filter substrate. The cutoff bounce is suppressed, dpa50 reduces to 2.5 mm which still
diameter of the M-WINS equals 2.46 § 0.05 mm, which satisfies the acceptance criteria of the USEPA. However,
agrees with the theoretical cutoff diameter of 2.4 mm GSD increases to 1.37 indicating a lower coarse particle
(Marple and Willeke 1976; Peters et al. 2001a), is similar collection efficiency and less sharp efficiency curve com-
to that of the WINS and meets the criteria specified in pared to that of the WINS with the wetted GF filter sub-
strate (Figure 4). Since Re of the M-WINS is 6000, a
large fraction of the particles will impact on the substrate
within a spot whose radius equals to the nozzle radius
(Sethi and John 1993; Huang and Tsai 2002). With the
GF filter substrate, injected water could trap and wash
off deposited particles effectively. In comparison, part of
the injected water can be blown away without trapping
particles. This explains a lower collection efficiency of
the wetted plate than that of the wetted GF filter sub-
strate. Therefore, the wetted GF filter in the M-WINS
improves the collection efficiency of coarse particles over
the wet impaction plate since it makes the impaction sur-
face rough and the filter media retains water keeping it
from being blown off by the aerosol jet. This tells why
the wetted GF filter has to be used in the M-WINS.
The sampling biases of the M-WINS with the wetted
GF filter substrate are shown in Table 1. There is a small
positive bias of C2.25% for the idealized coarse distribu-
tion, and small negative biases of ¡0.58% and ¡0.15% for
Figure 4. Particle collection efficiency curve of the M-WINS. the idealized fine and idealized typical distributions,
6 T.-C. LE AND C.-J. TSAI

Table 1. Estimated bias of the M-WINS in PM2.5 concentrations study. This is because the loaded particles (Al2O3 powder:
for three ambient aerosol size distributions. 0–4 mm) with the Mass Median Aerodynamic Diameter
M-WINS Fine aerosol Typical aerosol Coarse aerosol (MMAD) of »2 mm and GSD of »1.93 are smaller than
Estimated mass (mg/m ) 3
78.08 34.23 14.12
those in Kenny et al. (2000) (Aloxite, MMAD D 6 mm
Bias (%) ¡0.58% ¡0.15% C2.25% and GSD D 1.4) and Vanderpool et al. (2001b) (ARD,
MMAD D 5 mm and GSD D 3). Previous studies also
found that smaller loaded particles cause greater overload-
respectively. That is, the M-WINS can be used as the ing effect and more shift in dpa50 than larger particles at the
PM2.5 inlet to measure the PM2.5 mass concentration as it same loaded mass (Chen and Huang 1998; Kenny et al.
meets the performance FRM requirements (dpa50 D 2.5 § 2000). Smaller loaded particles form a steady mound
0.2 mm and sampling bias <§5%). underneath the nozzle, which can grow gradually while
Besides, it was observed that AF particles, which are bigger loaded particles form an unsteady mound, which
water soluble, were dissolved completely in DI water can be broken off. The type of loaded particles also contrib-
after striking on the substrate, indicating that washed utes to the differences in Figure 5a in different studies of
water could be further analyzed for determining the the WINS (Chen and Huang 1998).
chemical constituents for soluble coarse particles. In Figure 5b, a slight decrease in the sharpness of
collection efficiency curve (i.e., GSD slightly increases
from 1.2 to 1.25) was observed as the loaded particle
3.2. Laboratory results—particle loading effect
mass increases from 0.0 to 6.0 mg since the increase in
The laboratory test results of the cutoff diameters and the the collection of particles smaller than the cutoff diame-
collection efficiency curves of the WINS at different loaded ter occurs. It is due to the formation of the mound
particle masses are shown in Figures 5a and b, respec- which enhances the collection efficiency of fine particles
tively. It can be seen that dpa50 of the WINS decreases (Jurcik and Wang 1995). The consistent relationship of
from 2.44 to 2.05 mm and the collection efficiency curve the decrease in the sharpness and the cutoff diameter is
shifts to the left as the loaded particle mass increases from similar to that in Vanderpool et al. (2001b) and Kenny
0.0 to 6.0 mg. When striking on the oil-soaked GF sub- et al. (2000). The estimated negative biases of the
strate, the deposited particles retain and pile up on the loaded WINS increase (¡1.1% to ¡6.9%) with the
substrate, resulting in the change of the flow field in the increase in loaded mass (0.0–6.0 mg), indicating that
impaction region and the increase in a likelihood of parti- particle overloading causes the underestimation in
cle collection. Thereby the cutoff diameter decreases and PM2.5 mass concentrations. When the loaded mass is
the collection efficiency curve shifts to the left. greater than 4 mg, the estimated bias of the loaded
In Figure 5a, the trend of decreasing cutoff diameter is WINS exceeds the FRM requirement of §5%. For com-
similar to the results in previous studies (Kenny et al. 2000; parison, Kenny et al. (2000) estimated the mass loading
Vanderpool et al. 2001b) but the decrease in dpa50 is more for the WINS performance to deteriorate significantly
significant with increasing loaded particle mass in this is less than 3 mg.

Figure 5. (a) Cutoff aerodynamic diameter and (b) collection efficiency curve of the loaded WINS. (Note: the data of Vanderpool et al.
(2001b) at loaded particle mass of 16 mg are not shown.)
AEROSOL SCIENCE AND TECHNOLOGY 7

Figure 6. Field comparison of clean WINS and M-WINS (a) with clean VSCC and (b) with TEOM-FDMS as the references.

In contrast, dpa50 of the M-WINS only varies slightly between the performance of the M-WINS and that of the
from 2.44 to 2.49 mm with the loaded particle mass rang- VSCC (R2 > 0.99), and the TEOM-FDMS (R2 > 0.98).
ing from 0 to 6 mg. That is, the cutoff diameter of the The data of the WINS correlates well with those of the
M-WINS is not affected by the loaded particle mass. No VSCC (R2 > 0.997) and the TEOM-FDMS (R2 > 0.991).
particle mound was observed at the center of the impac- The WINS data show a better linear fit to the reference
tion substrate in the laboratory study and in the field data with higher R2 values than those of the M-WINS,
study to be discussed in the following section. It was indicating that the WINS outperforms the M-WINS.
observed that the Al2O3 test particles, which are indissol- This is because the sampling effectiveness (dpa50 and
uble, were washed off by continuous injected water pre- GSD) of the WINS is quite similar to the VSCC. Besides,
venting the accumulation of the deposited particles. The the linear regression coefficients of the M-WINS and the
pictures of the substrates with the mound (WINS) and references meet the USEPA performance requirements
without the mound (M-WINS) taken after 17 sampling for comparability of PM2.5 candidate equivalent method
days can be seen in Figure 8 for comparison. That is, it is samplers (i.e., slope D 1.0 § 0.05, intercept D 0.0 §
demonstrated that the M-WINS is able to eliminate the 1.0 mg/m3, and R2 > 0.97) (Kenny et al. 2004).
particle loading effect on the cutoff characteristics under Figure 7 shows that the uncleaned WINS under-
a high loading condition. samples PM2.5 concentrations compared to the VSCC

3.3. Field comparison results


Figure 6 shows field comparison results for the M-WINS
and the clean WINS with the VSCC (Figure 6a) and with
the TEOM-FDMS (Figure 6b) as the references at the
NCTU sampling site. The results of paired sample t test
show that the PM2.5 concentrations of the M-WINS
ranging from 3.7 to 34.1 mg/m3 agree well with those of
the VSCC (period #1: p D 0.142 > 0.05; period #2: p D
0.886 > 0.05) and the TEOM-FDMS (period #1: p D
0.220 > 0.05; period #2: p D 0.413 > 0.05). Similarly, the
PM2.5 concentrations of the clean WINS ranging from
7.9 to 34.8 mg/m3 have no significant difference with
those of the clean VSCC (period #1: p D 0.058 > 0.05)
and the TEOM-FDMS (period #1: p D 0.059 > 0.05).
This indicates that the Dichot-M-WINS is not statisti-
cally different from the other samplers for PM2.5 sam-
pling over a wide concentration range.
The slope, intercept, and correlation coefficient (R2)
of linear regression lines of pairwise comparisons are Figure 7. Sampling biases of the uncleaned WINS and the
shown in Figure 6. It shows a good comparability M-WINS using the VSCC as the reference.
8 T.-C. LE AND C.-J. TSAI

Figure 8. Impaction substrates of (a) the uncleaned WINS and (b) the M-WINS after 17 sampling days.

with the sampling biases increasing from ¡3.0 to ¡5.7% et al. 2001c) or »10% (Chow et al. 2008) for
(day 1 to day 5), ¡9.0 to ¡17.1% (day 6 to day 11, if the determining the mass. The cutoff diameter of the
data at day 9 and day 11 with the sampling biases of M-WINS only shifts from 2.46 to 2.51 mm, which not
»30% were disregarded), and ¡15.7 to ¡21.3% (day 12 only meets the accuracy of dpa50 § 0.07 (Kenny et al.
to day 17). Because of low PM2.5 concentrations (<11.0 2000) but also falls well in the range of the USEPA
mg/m3) at day 9 and day 11, biases become high. During requirement of 2.5 § 0.2 mm. The water injected
period #2, the PM2.5/PM10 ratio is 0.62 § 0.19 and the through the wetted GF filter of the M-WINS not only
measured PM2.5 concentration range of the VSCC is 3.7– washes off the deposited particles but can also dis-
33.5 mg/m3. The deposited particle mound was observed solves most of the inorganic species dominant at the
on the WINS substrate as shown in Figure 8 at the end NCTU sampling site. Black color was seen at the cen-
of sampling. The cutoff diameter of the uncleaned WINS tral spot on the substrate due to the deposition of
shifts downward from 2.48 to 2.12 mm as the estimated un-dissolvable particles such as soot, which were not
loaded mass is increased to 4.1 mg and the total sampled washed off completely. However, no particles piled up
PM2.5 is calculated to be 5.8 mg after 17 sampling days. on the substrate were observed after 17 continuous
The trend is similar to the results in previous studies sampling days as shown in Figure 8.
as described in the introduction (Kenny et al. 2000; Van-
derpool et al. 2001b). However, the present sampling
bias of the uncleaned WINS (<¡20%) is higher than the 4. Conclusion
estimated sampling bias by Kenny et al. (2000) The novel M-WINS using the wetted GF filter as
(<¡8.8%) in which the similar total sampled PM2.5 the substrate was developed in this study. In the
(»6 mg) and PM2.5/PM10 ratios (0.5–0.75) were M-WINS, the injected water captures and washes off
reported. The present sampling bias of the uncleaned the deposited particles on the wetted GF filter contin-
WINS is ¡19.6% as the calculated loaded mass is 2.3 mg uously, preventing the particle overloading effect. The
after 12 continuous sampling days but a small estimated laboratory tests showed no shift in the cutoff diame-
sampling bias of < ¡1.8% was reported by Vanderpool ter with the loaded particle mass ranging from 0.0 to
et al. (2001b) at the same estimated loaded mass 6.0 mg and the field comparison study showed accu-
(»2.3 mg) and PM2.5/PM10 ratio (0.62). In addition, after rate PM2.5 concentration measurement during 17 con-
5 days of sampling, the sampled PM2.5 concentration of tinuous sampling days. The study demonstrates that
the uncleaned WINS is »6% lower than that of the clean the M-WINS has a great potential to be used as the
VSCC, which is beyond the acceptance criteria of the PM2.5 inlet for continuous PM2.5 sampling and moni-
estimated PM2.5 concentration (sampling bias <§5%). toring for a long period with the sampling bias less
This agrees well with USA Federal regulation that than the USEPA requirement of §5%. In the future,
requires replacing the well of the WINS every 5 days of the collected liquid sample can also be coupled with
24-hr sampling. an Ion Chromatograph for the real time measurement
The sampling biases of the novel M-WINS are of water-soluble ions in coarse particles.
<§3.6% when PM2.5 concentration is greater than
7.1 mg/m3 (day 1 to 8 and 11 to 17). The biases are
increased slightly to ¡6.4% when PM2.5 concentration Funding
is low: 3.7mg/m3 at day 9 and 4.6 mg/m3 at day 10.
The final support of this work by the Taiwan Ministry of
These results are still in good agreement with the Science and Technology via the contracts MOST 105-2622-8-
results of the previous studies which found that the 009-007-TE4 and MOST 105-2221-E-009-005-MY3 is grate-
sampling biases of PM2.5 FRM could be 2–6% (Peters fully acknowledged.
AEROSOL SCIENCE AND TECHNOLOGY 9

References Marple, V. A., Rubow, K. L., Behm, S. M. (1991). A Microori-


fice Uniform Deposit Impactor (MOUDI): Description,
Chang, M., Kim, S., Sioutas, C. (1999). Experimental Studies on Calibration and Use. Aerosol Sci. Technol., 14:434–446.
Particle Impaction and Bounce: Effects of Substrate Design Pak, S. S., Liu, B. Y. H., Rubow, K. L. (1992). Effect of Coating
and Material. Atmos. Environ., 33:2313–2322. Thickness on Particle Bounce in Inertial Impactors. Aerosol
Chen, C. C., Huang, S. H. (1998). Shift of Aerosol Penetration Sci. Technol., 16:141–150.
in Respirable Cyclone Samplers. Am. Ind. Hyg. Assoc. J., Peters, T. M., Vanderpool, R. W., Wiener, R. W. (2001a).
60:720–730. Design and Calibration of the EPA PM2.5 Well Impactor
Chen, S. C., Tsai, C. J., Chen, H. D., Huang, C. Y., Roam, G. D. Ninety-Six (WINS). Aerosol Sci. Technol., 34:389–397.
(2011). The Influence of Relative Humidity on Nanoparticle Peters, T. M., Vanderpool, R. W., Wiener, R. W. (2001b).
Concentration and Particle Mass Distribution Measure- Methodology for Measuring PM2.5 Separator Character-
ments by the MOUDI. Aerosol Sci. Technol., 45:596–603. istics Using an Aerosizer. Aerosol Sci. Technol., 34:398–
Chow, J., Doraiswamy, P., Watson, J., Chen, L., Ho, S., Sode- 406.
man, D. (2008). Advances in Integrated and Continuous Peters, T., Norris, G., Vanderpool, R., Gemmill, D., Wiener, R.,
Measurements for Particle Mass and Chemical Composi- Murdoch, R., McElroy, F., Pitchford, M. (2001c). Field Per-
tion. J Air Waste Manag. Assoc., 58:141–163. formance of PM2.5 Federal Reference Method Samplers.
Dunbar, C., Kataya, A., Tiangbe, T. (2005). Reducing Bounce Aerosol Sci. Technol., 34:433–447.
Effects in the Andersen Cascade Impactor. Inter. J. Pharma- Sethi, V., John, W. (1993). Particle Impaction Patterns from A
ceutics., 301:25–32. Circular Jet. Aerosol Sci. Technol., 18:1–10.
Hu, S., Seshadri, S., McFarland, A. R. (2007). CFD Study on Tsai, C. J., Cheng, Y. H. (1995). Solid Particle Collection
Compound Impaction in a Jet-in-Well Impactor. Aerosol Characteristics on Impaction Surfaces of Different Designs.
Sci. Technol., 41:1102–1109. Aerosol Sci. Technol., 23:96–106.
Huang, C. H., Tsai, C. J. (2001). Effect of Gravity on Particle Tsai, C. J., Lin, T. I. (2000). Particle Collection Efficiency of
Collection Efficiency of Inertial Impactors. J. Aerosol Sci., Different Impactor Designs. Sep. Sci. Technol., 35(16):2639–
35:375–387. 2650.
Huang, C. H., Tsai, C. J. (2002). Influence of Impaction Plate Tsai, C. J., Liu, C. N., Hung, S. M., Chen, S. C., Uang, S.
Diameter and Particle Density on Collection Efficiency of N., Cheng, Y. S., Zhou, Y. (2012). Novel Active Personal
Round-Nozzle Inertial Impactors. Aerosol Sci. Technol., Nanoparticle Sampler for the Exposure Assessment of
36:714–720. Nanoparticles in Workplaces. Environ. Sci. Technol.,
Huang, C. H., Tsai, C. J., Shih, T. S. (2001). Particle Collection 46:4546–4552.
Efficiency of an Impactor with Porous Metal Substrates. J. Turner, J. R., Hering, S. V. (1987). Greased and Oiled Sub-
Aerosol Sci., 32:1035–1044. strates as Bounce-free Impaction Surfaces. J. Aerosol Sci.,
Huang, C. H., Chang, C. S., Chang, S. H., Tsai, C. J., Shih, T. S., 18:215–224.
Tang, D. T. (2005). Use of Porous Foam as the Substrate of USEPA. (1998). Monitoring PM2.5 in Ambient Air Using Des-
an Impactor for Respirable Aerosol Sampling. J. Aerosol ignated Reference or Class I Equivalent Methods, in Quality
Sci., 36:1373–1386. Assurance Handbook for Air Pollution Measurement Sys-
Jurcik, B., Wang, H. (1995). On the Shape of Impactor Effi- tems, Volume II, Part II, Section 2.12. Office of Air Quality
ciency Curve. J Aerosol Sci., 26:1139–1147. Planning and Standards, Air Quality Assessment Division,
Kenny, L. C., Gussman, R., Meyer, M. (2000). Development of RTP, NC27711.
a Sharp-cut Cyclone for Ambient Aerosol Monitoring USEPA. (2001). Ambient Air Monitoring Reference and
Applications. Aerosol Sci. Technol., 32:338–358. Equivalent Methods. Federal Register 40CFR Parts 53,
Kenny, L. C., Merrifield, T., Mark, D., Gussman, R., Thorpe, A. Subpt. F, Table F-1.
(2004). The Development and Designation Testing of a Vanderpool, R. W., Rubow, K. L. (1998). Generation of Large,
New USEPA-Approved Fine Particle Inlet: A Study of the Solid, Monodisperse Calibration Aerosols. Aerosol Sci.
USEPA Designation Process. Aerosol Sci. Technol., 38 Technol., 9:65–69.
(S2):15–22. Vanderpool, R. W., Peters, T. M., Natarajan, S., Tolocka, D. B.,
Kim, W. G., Yook, S. J., Ahn, K. H. (2013). Collection Effi- Gemmill, D. B., Wiener, R. W. (2001a). Sensitivity Analysis
ciency of Rectangular Slit-Nozzle Inertial Impactors with of the USEPA WINS PM2.5 Separator. Aerosol Sci. Technol.,
Impaction Plates of Elliptical Concave Curvature. Aerosol 34:465–476.
Sci. Technol., 47:99–105. Vanderpool, R. W, Peters, T. M, Natarajan, S., Gemmill, D. B.,
Marjam€aki, M., Keskinen, J. (2004). Effect of Impaction Plate Wiener, R. W. (2001b). Evaluation of the Loading Charac-
Roughness and Porosity on Collection Efficiency. J. Aerosol teristics of the EPA WINS PM2.5 Separator. Aerosol Sci.
Sci., 35:301–308. Technol., 34:444–456.
Marple, V. A., Willeke, K. (1976). Impactor Design. Atmos. Watson, J. G., Chow, J. C. (2011). Ambient aerosol
Environ., 10:891–896. sampling, in Aerosol Measurement: Principles, Techni-
Marple, V. A., and Rubow, K. L. (1986). Theory and Design ques and Applications, P. Kulkarni, P. A. Baron, and K.
Guidelines, in Cascade Impactor, P. J. Lodge, T. L. Chan, and Willeke, eds., John Wiley & Sons, Inc., New Jersey, pp.
T. L. Chan Jr, eds., Am. Ind. Hyg. Assoc., Ohio, pp. 79–101. 821–844.

You might also like