0048-9697/02/$ - see front matter䊚 2002 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 8 - 9 6 9 7 Ž 0 2 . 0 0 0 1 5 - 3
The effect of environmental conditions and electrical charge on the
weighing accuracy of different filter materials
Chuen-Jinn Tsai *, Chung-Tso Chang , Bow-Huei Shih , Shankar G.Aggarwal ,
a, a a aShou-Nan Li , Hung Min Chein , Tung-Sheng Shih
b b cInstitute of Environmental Engineering, National Chiao Tung University, No. 75, Poai Street, Hsin Chu 300, Taiwan a
Center for Environmental Safety and Health Technology Development, Industrial Technology Research Institute, Hsin Chu, b
Taiwan
Institute of Occupational Safety and Health, Council of Labor Affairs, Taipei, Taiwan c
Abstract
Different filter materials and electrical charge elimination methods were used to investigate the weighing accuracy of filter papers under different environmental conditions.The results show that the standard deviations (S.D.) of weight data for glass fiber and MCE filters were substantial without environmental control, whether or not the electrical charge eliminators were used. Values of 0.157 and 0.349 mg were determined for glass and MCE filters, respectively. The accuracy of weighing was substantially improved and the S.D. was reduced to 0.01 and 0.09 mg for glass fiber and MCE filters, respectively, after applying the environmental control conditions.For PVC and Teflon filters, the accuracy of weighing was good, even in the uncontrolled environmental conditions, whether or not the electrical charge eliminators were used.The S.D.values of weighing data of PVC and Teflon filters were 0.007 and 0.011 mg, respectively.䊚 2002 Elsevier Science B.V. All rights reserved.
Keywords: Filters; Environmental conditions; Electrical charge; Weighing accuracy
1. Introduction
Accurate weighing of filters before and after sampling is important in order to obtain the true concentrations of aerosols existing in the atmos-phere.Weighing results are influenced by the
ambient relative humidity (RH), temperature and
electrostatics on filters.Koistinen et al.(1999)
indicated that a buoyancy effect should be taken into account during weighing, hence, the weight
*Corresponding author.Tel.: 3-5731880; fax: q886-3-5727835.
E-mail address: cjtsai@cc.nctu.edu.tw(C.-J. Tsai).
of a filter sample must be corrected by the follow-ing expression:
9
Ž .
DbsV Ø r yr Ø10f a2 a1 (1)
where Db(mg) represents the buoyancy effect, Vf
(m ) is the volume of the filter, and r and r3
a1 a2
(kgym ) are the air densities at filter weighing3
before and after filter sampling, respectively.The air density is given by:
w3.848ØPy 0.00252ØTy0.02058 ØRHŽ . x
r sa (2)
Ž273.2qT.
Table 1
The weighing results of glass fiber filters with three different electrical charge eliminators
Glass fiber Without controlled Controlled
(T, RH) (T, RH)
Without charge Ave.s87.933 Ave.s87.844 elimination Std.s0.0628 Std.s0.0118
R.S.D.s0.0714 R.S.D.s0.0134 Using Po210electrical Ave.s91.314 Ave.s88.999
charge neutralizer Std.s0.0653 Std.s0.0038 R.S.D.s0.0715 R.S.D.s0.0043 Nilstat air ionizer Ave.s90.433 Ave.s89.497
Std.s0.0713 Std.s0.0057 R.S.D.s0.0788 R.S.D.s0.0064
Domestic electrical Ave.s87.455 Ave.s89.256 charge neutralizer Std.s0.0724 Std.s0.0032
R.S.D.s0.0828 R.S.D.s0.0036
Ave.: average weight, mg; Std.: standard deviation, mg; R.S.D.: relative deviation, %.
is the atmospheric temperature and RH(%) is the
ambient relative humidity.
The electrostatic charge is gained through the actions of friction, contact or stripping between
materials (Council of Labor Affairs, 1996).The
Council of Labor Affairs (1997) declared that the
variation of temperature and RH should be kept within "2.0 8C and "10.0%, respectively, during weighing, before and after sample collection.Elec-trostatic effects appear on the filter by touching tweezers or scales and causes a weight error.In order to prevent such errors, the use of an electrical charge neutralizer to eliminate the electrostatics from the filter is necessary.
The US EPA (1988) specification of standard
operation process for PM10 filters stated that the average temperature of the environment should be kept between 15 and 30 8C, and the temperature variation should be maintained within "3.0 8C over 24 h, whereas, the average RH of the envi-ronment should be kept between 20 and 45%, and the RH variation should be maintained within
"5.0 8C over 24 h, before and after weighing.
Following the standard operation process for
PM2.5 filters, the average environmental
tempera-ture should be kept between 20 and 23 8C and the temperature variation should be maintained within
"2.0 8C over 24 h, whereas the average RH of
environment should be kept between 30 and 40%, and the RH variation should be maintained within
"5.0 8C over 24 h before and after weighing.
Among the several different kinds of filter mate-rials which are used, glass fiber filters can sustain higher temperatures and are less hygroscopic than
a cellulose filter (US EPA, 1994; Hering, 1989).
Li and Lundgren (1999) suggested that the
stan-dard deviation (S.D.) of weighing data of PVC
filters under controlled environmental conditions
(i.e. the variation range of the temperature is 21–
33 8C, and the RH is 50–98%) is less than 0.006
mg in deviations within 14 runs of weighing tests. MCE filters are usually used for collecting the sample for further chemical analysis to determine the species concentration.Charell and Hawley
(1981) tested MCE and PVC filters at different
RHs, and showed that the weight of the filters increases with the increasing RH, and that a MCE filter is much more sensitive than a PVC filter
with respect to this effect.Mark (1974) tested
glass fiber, MCE and PVC filters at different RHs and found that MCE shows the largest hygrosco-picity when compared with glass fiber and PVC filters.
Teflon filters are widely used for sampling
particles and gases (Willeke and Baron, 1993).
Tsai et al.(1997) tested MCE and Teflon filters in a chamber over a temperature range of 15–22 8C and a RH 40"5%.The results show that the S.D. of MCE and Teflon filters was 0.0151 and 0.0039 mg, respectively, whereas, Teflon filters performed better in weighing accuracy than MCE filters.
In this study, four different filter materials and three different electrical charge eliminators were used under controlled and uncontrolled weighing environments to understand the influence of envi-ronmental conditions and electrostatics on the weighing accuracy of filters.
2. Methods and materials
The four types of filter materials used in this
experiment were: glass fiber (Cat 噛1882866,
Whatman Corp., dia. 37 mm); PVC (Cat
噛P-503700, Omega specialty instrument Co., dia. 37 mm); Teflon (Cat 噛LZ0204750, Omega specialty
instrument Co., dia. 47 mm); and MCE (PyN
64678, Gelman Sciences Co., dia. 37 mm).Three
Fig.1.Effect of ambient relative humidity on glass fiber filters under uncontrolled environmental conditions and in the absence of the electrical charge eliminators.
Table 2
The weighing results(mg) of Teflon filters without using electrical charge eliminators
Day 1 2 3 4 5 6 7
Uncontrolled 244.367 244.363 244.370 244.378 244.374 244.375 244.373
(T, RH)
Controlled 244.364 244.361 244.372 244.376 244.373 244.372 244.371
(T, RH)
namely: Po210 electrical charge neutralizer; Nilstat
air ionizer (Nilstat 2020, Ion Systems, Inc.); and
domestic electrical charge neutralizer(Model:
SN-2001, Taiwan).The Nilstat air ionizer has the
dimensions of 10=7=5 cm and has four needles operating at 24 VAC.The power of the internal fan of the ionizer was disconnected and the four needles were centered at the exit of the fan to enhance the effect of electrostatic elimination.The domestic electrical charge neutralizer has dimen-sions of 13=11=9 cm with one discharge needle, operating at 5000 VAC in front of the 25-cm snake tube.
The experiment was conducted in a controlled chamber.Four different filters were weighed after
24 h of environmental conditioning (RHs40–
45%).The weighing of filters with an electronic
scale (Sartorius A200S) was performed at
con-trolled(the temperature and RH ranges were 22–
26 8C and 39.5–41%, respectively) and
uncontrolled(the temperature and RH ranges were
24–28.5 8C and 40–65%, respectively) weighing
environments.
3. Results and discussions
3.1. Glass fiber filter
The effect of RH on the weighing data of glass fiber filters under uncontrolled environmental con-ditions is shown Fig.1, the weights of glass fiber filters are increased with the increasing RH.The high hygroscopic behavior of glass fiber filters may be one of the reasons for this.Table 1 shows the weighing results of the glass fiber filter under controlled and uncontrolled environmental
condi-tions.The S.D.of the weighing data of filters
indicate that the weighing results show better accuracy under controlled environmental condi-tions rather than under uncontrolled condicondi-tions. With the electrical charge eliminators, the experi-mental results show the same consequences of weighing accuracy.The results show that the weighing accuracy of glass fiber filters is signifi-cantly influenced by environmental conditions, such as temperature and ambient RH.The results also show that the weighing accuracy of filters is not improved by using electrical charge eliminators
under uncontrolled environmental conditions
before weighing (S.D.s calculated under this
con-dition are all larger than 0.065 mg).In contrast,
the weighing accuracy of filters is obviously improved under controlled environmental condi-tions and the weighing accuracy can be further improved by using electrical charge eliminators before weighing.The S.D.of weighing data can
Table 3
The weighing results of Teflon filters with three different elec-trical charge eliminators
Teflon Uncontrolled Controlled
(T, RH) (T, RH)
Without charge Ave.s244.371 Ave.s244.370 elimination Std.s0.0047 Std.s0.0049
R.S.D.s0.0019 R.S.D.s0.0020 Using Po210 Ave.s272.801 Ave.s272.800
electrical charge Std.s0.0040 Std.s0.0087 neutralizer R.S.D.s0.0015 R.S.D.s0.0032 Nilstat air ionizer Ave.s277.004 Ave.s277.003
Std.s0.0073 Std.s0.0096 R.S.D.s0.0026 R.S.D.s0.0035
Domestic electrical Ave.s254.636 Ave.s254.637 charge neutralizer Std.s0.0029 Std.s0.0082
R.S.D.s0.0011 R.S.D.s0.0032
Ave.: average weight, mg; Std.: standard deviation, mg; R.S.D.: relative deviation, %.
Table 5
The weighing results of PVC filters with three different elec-trical charge eliminators
PVC Uncontrolled Controlled
(T, RH) (T, RH)
Without charge Ave.s13.259 Ave.s13.257 elimination Std.s0.0021 Std.s0.0023
R.S.D.s0.0160 R.S.D.s0.0174 Using Po210 Ave.s13.056 Ave.s13.052
electrical charge Std.s0.0029 Std.s0.0023 neutralizer R.S.D.s0.0220 R.S.D.s0.0176 Nilstat air ionizer Ave.s12.190 Ave.s12.187
Std.s0.0030 Std.s0.0033 R.S.D.s0.0244 R.S.D.s0.0273
Domestic electrical Ave.s10.758 Ave.s10.752 charge neutralizer Std.s0.0019 Std.s0.0013
R.S.D.s0.0179 R.S.D.s0.0122
Ave.: average weight, mg; Std.: standard deviation, mg; R.S.D.: relative deviation, %.
Table 4
The weighing results(mg) of PVC filters without using electrical charge eliminators
Day 1 2 3 4 5 6 7
Uncontrolled 13.260 13.258 13.262 13.260 13.259 13.261 13.255
(T, RH)
Controlled 13.255 13.254 13.260 13.257 13.257 13.257 13.261
(T, RH)
be decreased from 0.01 mg (without electrostatic
elimination) to 0.003 mg (with electrostatic
elim-ination), whereas the relative standard deviation
(R.S.D.) can be decreased from 0.013% to
0.004%.
3.2. Teflon filter
The weighing results of Teflon filters without using electrical charge eliminators are shown in Table 2.The weighing data are very close for both controlled and uncontrolled environmental
condi-tions (the largest S.D. of weighing data is less
than 0.015 mg).The results given in Table 3 show
that the weighing accuracy is good enough and not dependent upon the use of electrostatic elimi-nators before weighing. The S.D.s indicate that no effect of environmental conditions was observed
on weighing accuracy (S.D. for both, controlled
and uncontrolled conditions was 0.005 mg).
3.3. PVC filter
The PVC filter showed similar characteristics to those shown by the Teflon filter.Table 4 shows that the weighing results were very close to each other and did not depend on the environmental conditions.The largest variation in weighing data, when no electrical charge eliminator was applied was 0.007 mg with a R.S.D. of 0.017%.
Table 5 shows that the weighing accuracy of PVC filters is good even at uncontrolled
environ-mental conditions (S.D.s0.0021 mg, R.S.D.s
0.016%).The data show that the weighing
accuracy of PVC filters could not be improved with electrical charge eliminators.
Fig.2.Effect of ambient relative humidity on MCE filters under uncontrolled environmental conditions and in the absence of the electrical charge eliminators.
Table 6
The weighing results of MCE filters with three different elec-trical charge eliminators
MCE Uncontrolled Controlled
(T, RH) (T, RH)
Without charge Ave.s41.582 Ave.s41.435 elimination Std.s0.1132 Std.s0.0431
R.S.D.s0.2722 R.S.D.s0.104 Using Po210 Ave.s42.803 Ave.s42.606
electrical charge Std.s0.1173 Std.s0.0354 neutralizer R.S.D.s0.2740 R.S.D.s0.0831 Nilstat air ionizer Ave.s42.968 Ave.s42.791
Std.s0.1101 Std.s0.0354 R.S.D.s0.2563 R.S.D.s0.0828
Domestic electrical Ave.s42.862 Ave.s42.664 charge neutralizer Std.s0.1257 Std.s0.0377
R.S.D.s0.2933 R.S.D.s0.0883
Ave.: average weight, mg; Std.: standard deviation, mg; R.S.D.: relative deviation, %.
Fig.3.The weighing results of glass fiber filters before and after buoyancy modification under uncontrolled environmental conditions and in the absence of the electrical charge eliminators.
3.4. MCE filter
MCE (mixed cellullose ester) filters are eco-nomical and suitable for air monitoring applica-tions.Fig.2 shows that the weight of the MCE filters was increased with the increasing RH under uncontrolled environmental conditions and in the absence of electrostatic eliminators.This proves the strongly hygroscopic character of MCE filter materials.Due to this character, the weight of the MCE filters was increased by 0.349 mg, which was approximately three times higher than the increase in weight of 0.157 for glass fiber filters, when the ambient relative humidity was 40–60%. The results given in Table 6 show that the weighing accuracy cannot be improved by using electrical charge eliminators instead of controlling the environmental conditions. The S.D.s were not reduced very much after using electrical charge eliminators under controlled environmental
condi-tions (S.D.s before and after use were 0.043 and
0.038 mg, respectively).
3.5. Correction of buoyancy effect
The correction of the buoyancy effect is shown
in Eqs. (1) and (2).Fig.3 shows the weighing
results of glass fiber filters under uncontrolled conditions without using the electrostatic elimina-tors.Comparing the original weight of glass fiber filters(87.933"0.0582 mg) with the weight after
(87.970"0.0581 mg), no improvement in results
was observed for accuracy, and the buoyancy effect was not obvious.Therefore, the weighing of filters should be conducted in well-controlled environ-mental conditions, and it is hard to obtain accurate results just by considering the buoyancy effect. The buoyancy effect is also shown not to improve the weighing accuracy of Teflon filters, MCE filters or PVC filters.
4. Conclusions
The present study documents the accuracy of weighing different filter materials.The weighing accuracy of glass fibers under controlled environ-mental conditions is good enough, and could be improved if the electrical charge eliminators are
used simultaneously (R.S.D. is reduced to
0.006%).MCE filters show similar weighing
char-acteristics with respect to environmental conditions and the use of electrical charge eliminators.The hygroscopicity of MCE filters is more serious than glass fiber filters, and the R.S.D. was found to be 0.088% under the controlled environmental con-ditions and after using the electrical charge elimi-nators.PVC filters and Teflon filters showed good weighing results, and weighing data were not influenced by the environmental conditions and electrical charge eliminators. The R.S.D.s calculat-ed for PVC and Teflon filters were 0.016% and 0.002%, respectively, under uncontrolled environ-mental conditions and in the absence of the elec-trical charge eliminators.The weighing accuracy of PVC and Teflon filters could not be improved under controlled environmental conditions and after using the electrical charge eliminators.
Acknowledgments
The authors would like to thank the financial support of the Institute of Occupational Safety and
Health(IOSH), Taiwan.
References
Charell RP, Hawley RE.Characteristics of water adsorption on air sampling filters.Am Ind Hyg Assoc J 1981;42:353 – 358.
Council of Labor Affairs.Reference method for the analysis of nuisance dust.Institute of occupational safety and health
(IOSH) 1997.(RM017A).
Council of Labor Affairs.Hazard of electrostatics, investiga-tion and preveninvestiga-tion.Institute of Occupainvestiga-tional Safety and Health(IOSH) 1996.(85–S342; 4–6).
Hering SV.Air sampling instruments for evaluation of atmos-pheric contaminants.7th ed.USA: American Conference of Governmental Industrial Hygienists, Inc, 1989.p.311 –322. Koistinen KJ, Kousa A, Tenhola V, et al.Fine particle(PM )2.5
measurement methodology, quality assurance procedures, and pilot results of the EXPOLIS study.J Air Waste Manage Assoc 1999;49:1212 –1220.
Li SN, Lundgren DA.Weighing accuracy of samples collected by IOM and CIS inhalable samplers.Am Ind Hyg Assoc J 1999;60:235 –236.
Mark D.Problems associated with the use of membrane filters for dust sampling when compositional analysis is required. Ann Occup Hyg 1974;17:35.
Tsai CJ, Shih TS, Sheu RN.Characteristic of lead aerosols in different work environments.Am Ind Hyg Assoc J 1997;58:650 –656.
US EPA.PM2.5mass weighing laboratory standard operating procedure for the performance evaluation program.1988. US EPA.Diesel particulate filter handling and weighing 1994.
(TP714C).
Willeke K, Baron PA.Aerosol measurement principles, tech-niques, and applications.USA: Van Nostrand Reinhold, Inc, 1993.p.628 –630.