vvEPA
United States
Environmental Protection
Agency
Environmental Monitoring and
Support Laboratory
Research Triangle Park NC 27711
EPA-600/4-80-024
April 1980
Research and Development
Improvement and
Evaluation of
Methods for Sulfate
Analysis
600480024
Part II.
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
IMPROVEMENT AND EVALUATION OF METHODS FOR SULFATE ANALYSIS
PART II
Final Report
by
B. R. Appel, E. M. Hoffer, W. Wehrmeister
M. Haik and J. J. Wesolowski
Air and Industrial Hygiene Laboratory Section
California Department of Health Services
2151 Berkeley Way
Berkeley, California 94704
EPA Grant No. 805-447-1
Project Officer
John C. Puzak
Quality Assurance Division
Environmental Monitoring Systems Laboratory
Research Triangle Park, Ncrth Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views
and policies of the U.S. Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
Measurement and monitoring research efforts are designed to anticipate
potential environmental problems, to support regulatory actions by developing
an in-depth understanding of the nature and processes that impact health and
the ecology, to provide innovative means of monitoring compliance with regula-
tions, and to evaluate the effectiveness of health and environmental protection
efforts through the monitoring of long-term trends. The Environmental
Monitoring Systems Laboratory, Research Triangle Park, North Carolina, has
responsibility for: assessment of environmental monitoring technology and
systems; implementation of agency-wide quality assurance programs for air
pollution measurement systems; and supplying technical support to other groups
in the Agency including the Office of Air, Noise, and Radiation, the Office of
Toxic Substances, and the Office of Enforcement.
The work covered in this report details efforts performed for the
Environmental Monitoring Systems Laboratory to improve methodology used to
monitor air pollution concentrations. Several procedures for analyzing the
sulfate content of ambient aerosols collected on various filter types were
evaluated for precision, accuracy, working range, and intermethod comparability.
The work reported here and in phase I of this project (EPA-600/4-79-028,
April, 1979) should provide air pollution agencies with information about the
reliability of several different sulfate analytical procedures and help each
agency choose the analytical procedure which best fulfills its needs.
<£--'
Thomas R/ Hauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina
iii
-------
ABSTRACT
Methods for extraction of sulfate from glass fiber hi-vol and Teflon lo-vol
samples were evaluated. Efficiencies were found to vary with sampling
location up to 20%. Mechanical shaking in water at room temperature was
significantly more efficient than ultrasonic or reflux techniques with
hi-vol samples. While Teflon filters are not wet by water, pre-wetting
of filters with methanol did not significantly enhance sulfate extraction.
A turbidimetric sulfate method using SulfaVer IV was evaluated for
ruggedness, precision and intermethod agreement. Its precision was at
least equal to that of a conventional turbidimetric method but its accuracy
was somewhat less, especially at lower sulfate levels. The Dionex Model 10
ion chromatograph was evaluated for low level sulfate analysis using both
a sample pre-concentrator and large (0.5 ml) sample loop. The latter was
the preferred technique for samples <_ 20 yg/ml. Accuracy was within
15% in the range 2 to 20 yg/ml with a median C.V. of 6.5% for 2k atmos-
pheric samples. This range will permit sulfate analysis of 2U hour fine
particulate samples collected with dichotomous samplers. Use of a sample
pre-concentrator permitted analysis of samples containing < 1 pg/ml sulfate.
This work is submitted in fulfillment of Grant No. 805-^7-1 by the
California Department of Health Services under the sponsorship of the
U.S. Environmental Protection Agency. This report covers the period
October 1, 1978 to Sept. 30, 1979, and work was completed as of May 27, 1979.
iv
-------
CONTENTS
Abstract iii
Figures vi
Tables - vii
Acknowledgements viii
I. Introduction _ 1
II. Summary and Conclusions 3
III. Sulfate Extraction Efficiency Studies with Glass Fiber
Hi-vol Filter Samples. 7
IV. Sulfate Extraction Efficiency Studies with Teflon
Lo-vol Filter Samples. 19
V. Shelf Life of Pre-mixed Reagent for Turbidimetric
Sulfate Analysis. 27
VI. Evaluation and Improvement of a Turbidimetric Method
for Sulfate Using SulfaVer IVR. 32
VII. Sulfate Analysis with the Dionex Model 10 Ion Chromatograph 38
VIII. Intermethod Comparison 63
References 71
Appendices
A. Ultrasonic Extraction Procedure 73
B. Reflux Procedure from AIHL Method 6l 7^
C. Mechanical Shaking Procedure from BAAPCD Method S-U-2 75
D. Sulfate Extraction from Teflon Filters by Mechanical
Shaking. 76
E. Sulfate Extraction from Teflon Filters by Ultrasonic
Extraction with Pre-wetting with Methanol. 77
F. Sulfate Extraction from Teflon Filters by Heating in
Water at 80°C. 78
G. AIHL Method 79. Determination of Sulfate in High Volume
Particulate Samples Using SulfaVer IVR. 79
H. Ion Chromatographic Analysis of Sulfate in the Range 0 to
20 yg/ml. 89
v
-------
FIGURES
Number Page
1 Comparison of Sulfate Recovered by Successive Extractions of 23
Fluoropore Filter Samples
2 Effect of Pre-mixed Reagent Age in Turbidimetric Sulfate
Analysis Working Curve Slope vs. Time 28
3 Effect of Pre-mixed Reagent Age in Turbidimetric Sulfate
Analysis Working Curve Intercept vs. Time 29
k Effect of Pre-mixed Reagent Age in Turbidimetric Sulfate
Analysis Working Curve Sy.x vs. T:'.me 30
5 Sulfate Data Reduction Procedures for 1C 39
6 Typical Working Curve for Sulfate Analysis by Dionex 1C With
Pre-concentrator 53
7 Working Curve for Sulfate Analysis by Dionex 1C With 0.5 ml
Sample Loop 57
8 Accuracy as a Function of Sulfate Concentration by Dionex 1C
With 0.5 ml Sample Loop 58
9 Scatter Diagram of Results With Hi-vol Filter Samples Comparing
SulfaVer IV and Colovos MTB Sulfate Results 67
10 Scatter Diagrams of Results With Lo-vol Filter Samples Using
Three Sulfate Methods 70
VI
-------
TABLE
Number Page
1 Accuracy and Precision of the MTB Method Using EPA Audit Strips 10
2 Determination of Variability Between Quarters Cut from 8 x 10"
Hi-vol Filters 12
3 Recovery of Sulfate from Extraction of Quarters from 8 x 10"
Glass Fiber Filter (yg SO^") I1*
h Average Efficiencies for Extraction of Water Soluble Sulfate
from 2^-hour Hi-vol Glass Fiber Filter Samples 1"
5 Mean Recoveries of Sulfate by 60-Minute Mechanical Shaking as
a Function of Location IT
6 Recovery of Sulfate and Efficiency of Extraction With Lo-vol
Teflon Filter Samples from Berkeley 2h
7 Recovery of Sulfate and Efficiency of Extraction With Lo-vol
Teflon Filter Samples from Los Angeles 25
8 Factors for Evaluation in Ruggedness Test of SulfaVer Method 33
9 Results of Ruggedness Test of Sulfate Analysis by Turbidimetry 35
10 Interference Effect of Nitrate on Sulfate Determination hi
11 The Effect of Nitrate on Sulfate Retention Times ^3
12 Analysis of EPA Sulfate Audit Strips by the Dionex 1C hU
13 Change in Peak Height and Area With Time With Sulfate Standards h6
l Instrument Settings, Sulfate Level for 80% Full Scale and
Accuracy With Standards Using the Dionex 1C With Pre-
concentrator 52
15 Retention Times for Sulfate , Nitrate and Related Species
Using 250 mm Anion Separator Column 56
l6 Day-to-Day Change in Working Curve of Dionex 1C Using 250 mm
Column for Sulfate Analysis 60
17 Sulfate Analysis of EPA Audit Strips by 1C Using the 0.5 ml
Sample Loop 62
18 Results of Intermethod Comparison With Hi-vol Filter Samples
(ug sulfate/ml) 6U
19' Average Agreement and Precision of Sulfate Methods With Hi-vol
Filter Samples 65
20 Results of Intermethod Comparison With Teflon Lo-vol Filter
Samples (yg sulfate/ml) 69
vii
-------
ACKNOWLEDGEMENTS
Other participants in this study included Ms. L. Raftery who provided
assistance in the laboratory, in filter sample collection, and with data
reduction. Dr. Evaldo Kothny assisted in development of experimental
procedures, in supervision of some of the experimental work and in review
of reports. The atmospheric samples used in this study were provided,
in part, by Mr. J. Wendt, California Air Resources Board and by
Mr. R. J. Schwall, Rockwell International. The SulfaVer IVR.pillows
were furnished by S. Balestrieri of the Bay Area Air Quality Management
District. The cooperation and assistance of all persons named are
gratefully acknowledged.
Mr. J. C. Puzak served as Project Officer for this program. His help-
fulness throughout this work has been sincerely appreciated.
viii
-------
I. INTRODUCTION
In preceding EPA-sponsored programs, a series of wet chemical sulfate
methods was evaluated and compared to one another and in some cases,
1-3
to total sulfur determinations by x-ray fluorescence analysis.
These methods were:
—Barium sulfate turMdimetric procedures (Public Health Service,
AIHL Method 6l and an improved version, AlHL Method 75),
•p
—A barium sulfate turbidimetric method using SulfaVer IV ,
-Automated methylthymol blue procedures (Midwest Research Institute,
the Colovos and AIHL versions),
7
-The AIHL microchemical method,
—Two modifications of the thorin method as developed by C. Brosset,
9
-A manual barium chloranilate method,
-The Dionex ion chromatograph.
Typically, the methods were evaluated for precision, accuracy, working
range, interference effects and comparability of results with atmospheric
samples. In one case a ruggedness test was performed.
The current program includes work done in the period October 1978-
March 1979 to complete EPA Grant No. 805-UUT-l. It continues sulfate
studies including (l) an evaluation of sulfate extraction procedures
-------
for glass fi"ber hi-volume filter and Teflon low-volume filter samples,
(2) a determination of the shelf-life of the pre-mixed reagent used in
barium sulfate turbidimetric AIHL methods 6l and 75» (3) a ruggedness
test and optimization of a turbidimetric procedure using SulfaVer IV^,
(U) an evaluation of the Dionex ion chromatograph for sulfate analysis
of extracts from low-volume filter samples such as anticipated with a
dichotomous sampler network and (5) an intermethod comparison with
the methods evaluated.
-------
II. SUM4APY MID CONGLUCIONE
Evaluation of procedures for aqueous extraction of sulfate from glass
fiber hi-vol filter samples has demonstrated that 30 minutes ultrasonic
extraction and 60 minute reflux procedures are not significantly
different. However, these techniques give sulfate recoveries 2-3%
lower than mechanical shaking for 60 minutes at room temperature.
Ultrasonic extraction for 5 minutes is substantially poorer in efficiency.
Systematic variation in sulfate recoveries with sampling location was
observed.
Using Berkeley low-volume atmospheric samples on Teflon membrane filters,
four sulfate extraction techniques gave results which were equal within
experimental error. However, with samples collected adjacent to a
Los Angeles freeway, heating in hot water at 80°C in sealed tubes was
notably less efficient than 30 minutes ultrasonic extraction with or
without pre-wetting with methanol, or mechanical shaking for 60 minutes.
The latter averaged about 90%. The use of methanol to pre-wet the
filters did not cause a consistent improvement in extraction efficiency
and is not recommended. The choice between mechanical shaking and
ultrasonic extraction can probably be based on convenience and personnel
costs. It remains unclear what effect, if any, simultaneous ultrasonic
extraction of large numbers (> 8) samples has on extraction efficiency.
Similarly the effect of position within the bath, in relation to
standing waves set up by ultrasonic vibration, was not evaluated.
Finally, the need exists to obtain a quantitative measure of ultrasonic
energy output. Lacking such a measure the generality of the current
study, which used a Bransonic Model 42 150 watt input ultrasonic bath,
remains unclear.
-------
Sulfate was substantially more difficult to extract from the freeway
particulate enriched samples. This may relate to the presence of oily
particulates, or, and less likely, relatively insoluble sulfates
(e.g. lead sulfate). The former hypothesis would be consistent with
results with the hi-vol filter samples.
A pre-mixed reagent for stabilizing suspensions of barium sulfate in
turbidimetric sulfate analysis (AIHL Methods 6l and 75) was shown to
have a shelf life of more than 15 months.
A turbidimetric sulfate method utilizing SulfaVer IVR was subjected to an
11 parameter ruggedness test. Choice of sulfate level at 300 or 1300
yg/20 ml was the dominant source of variance in the method; at the
lower sulfate level results were substantially in error. The optimized
procedure (Appendix G) utilizes reagent-sample mixing as well as
absorbance readings in one inch diameter, sealed test tubes to eliminate
sample transfers.
The Dionex Model 10 ion chromatograph was evaluated for use in analyzing
low level (< 20 yg/ml) sulfate samples such as obtained from dichotomous
samplers. Both a large sample loop and a sample pre-concentrator were
evaluated for this application. The latter was found especially useful
for samples < 1 yg/ml and showed excellent linearity and accuracy with
standards. However, for routine analysis, an 0.5 ml sample loop provided
a simpler procedure and an adequate analytical range. Accuracy with the
0.5 ml sample loop was hampered, however, by non-linearity of the
working curve below about 5 yg/ml. Nevertheless, the method remained
accurate within 15% in the range 2 to 20 yg/ml, as measured with EPA
U
-------
sulfate audit strips. Precision, as expressed by C.V. values, was <_ 5$.
Nitrate was shown to provide insignificant interference, even with the
use of a 250 mm anion separator column, if the trailing peak height
method was used.
An intermethod comparison was done for the SulfaVer IV method using
2U hi-vol extracts. It was compared to an automated methylthymol blue
(MTB) procedure with results for the latter being calculated with and
•p
without correction for initial sample absorbance. The SulfaVer IV
method was, on average, 10$ higher than the corrected MTB results
and U$ higher than the uncorrected MTB results. The median C.V. for
2k samples, analyzed with three determinations on separate days, was
3.2$. This is somewhat better precision than previously found with
this method or with Method 6l and 75 probably as a result of the use
of a better quality spectrophotometer and the elimination of sample
transfers. The accuracy of the method is somewhat poorer compared to
conventional turbidimetric sulfate nethods using barium chloride.
An intermethod comparison was done between the Dionex 1C, MTB (0-10
yg/ml range) and AIHL microchemical methods. On average, 1C results
were lower than those by the MTB procedure by 7$, but agreed within 2%
with those by the AIHL micro method. The median C.V. for the 1C method
was 6.5$.
Based on evaluations of accuracy, precision and intermethod comparison
R
the SulfaVer IV procedure, AIHL Method 79, can be employed as an alter-
native to conventional turbidimetric sulfate analysis in 2k hour
hi-volume filter samples. Similarly, the Dionex ion chromatograph with
0.5 ml sample loop can be used for analysis of sulfate in the range
5
-------
<_ 20 yg/ml such as can be obtained with extracts from 2U hour dichotomous
filter samples. However, the non-linearity of the working curve observed
below 5 Pg/nl decreases accuracy unless additional standards are employed.
For samples < 1 pg/ml, a sample pre-concentrator is necessary.
-------
III. SULFATE EXTRACTION EFFICIENCY STUDIES WITH GLASS FIBER HI-VOL FILTER SAMPLES
A. Introduction
At least three procedures are in use by monitoring organizations
for the extraction of water soluble ions from atmospheric samples.
These include 30-minute ultrasonic extraction at room temperature,'
12
60-minute heating under reflux, and 60-minute mechanical shaking
13
in water at room temperature. The aims of the present study were
l) to compare these procedures, and 2) to determine their absolute
efficiencies for removal of water soluble sulfate. The specific
procedures evaluated were as follows:
1. Ultrasonic extraction in 50 ml H^O as in Reference 12 but for
5 minutes and using 60 ml Erlenmeyer flasks with ground glass
stoppers.
2. As in (l) but for 30 minutes (the time specified in Reference
12).
I
3- 60 minutes boiling under reflux as in Reference 13 which includes
filter rinsing. Final volume 100 ml.
H. Mechanical shaking with aBurrell wrist action shaker in 50 ml.
H20 at room temperature contained in 250 ml flasks sealed with
Parafilm.13
Detailed protocols for the procedures followed are included as
appendices A-C.
-------
In all cases extraction was followed by filtration through an 0.7 ym
pore size cellulose ester Millipore filter using a Millipore
filtration apparatus. Only with the reflux procedure were the glass
fiber and Millipore filters rinsed following filtration. As a result
this study represents a comparison of the extraction process, itself,
minimizing variations introduced by subsequent sample handling.
After comparing sulfate recoveries by the four methods, the efficiency
of each method for extraction cf sulfate was established by deter-
mining the amount of sulfate remaining unextracted following initial
extraction with one of the four methods. The approach used was to
repeatedly re-extract the filter residue and analyze these extracts
for sulfate. By choosing the reflux procedure for this purpose the
problem of sulfate in the extract remaining wetting the filter was
minimized since in this procedure, the filter is rinsed with water.
Following the initial extraction by refluxing, the residual filter
was re-extracted by the reflux method using 35 ml 1^0 and, with
washings, brought to 50 ml for analysis. The residue from this
extraction was extracted a thirl time in 15 ml water by 30 minutes
ultrasonic extraction. The sum of the sulfate recovered on the
quarters by successive extractions was taken as the total water
soluble sulfate. Efficiencies for each extraction method were
calculated relative to these totals. Thus, only the efficiency for
extracting water soluble sulfate is being measured. Insoluble
sulfates, if present, are not considered. (Past studies comparing
x-ray fluorescence analyses for sulfur to water extractable sulfate
have failed to establish a significant difference. Accordingly, the
distinction between "total sulfate" and "total water soluble sulfate"
might be academic).
8
-------
B. Preliminary Evaluation to Establish Variability Between Filter Quarters
The experimental procedure requires sectioning a set of 8 x 10" hi-vol
filter samples into quarters and extracting each of the four quarters
from a given filter by one of the four methods. The ability to discern
differences "between extraction methods is limited "by the inherent
variability between the quarters of a given filter and the precision
of the analytical method. To measure this variability, four hi-vol
filters were quartered and the l6 quarters extracted by the 30-minute
ultrasonic extraction procedure. For this trial the filter quarters
included the usual two borders without particulate. Since filters
are rarely mounted in such a way as to yield equal borders on all
sides, effort was made to quarter the filters to provide equal loaded
areas. However, because the sealing gasket on the sampler is not a
perfect rectangle (it is typically curved slightly into arcs) the
quarters could not be conveniently cut into identically loaded area.
'This would contribute to any variability observed.
The extracts were analyzed by an automated MTB method as solutions
*
in the range 16 to 6l yg/ml sulfate. The analytical protocol has
2
been described and evaluated previously. A check of accuracy and
precision of the method using EPA audit strips to provide solutions
in this range is given in Table 1, and was made as part of the current
study. Results indicate that the differences between the theoretical
and recovered sulf ate for the N-I'B method are 3% or less and the
coefficient of variation for four strips at each level, is in the
range 0.5 to k.9% increasing with decreasing sulfate concentrations.
*
Analyses were done in the range 0-100 ug/ml using MTB levels uncorrected for
impurities and a third order regression fit to the resulting non-linear
working curve.
-------
Table 1
Accuracy and Precision of the
MTB Method Using EPA Audit Strips
Theoretical Value
Sample (yg/strip) (yg/ml) 0"bs .a/Theore^. C.V.
9000 Series
712-5000 Series
712-6000 Series
7^5.6
2250
2700
lU.9
U5.0
5^.0
1.00
0.98
0.97
U.9
0.5
1.8
a. Mean results for four strips extracted by 30-minute ultrasonic
extraction, calculated using third order regression data analysis.
10
-------
The results of the quarter filter variability study are given in
Table 2 expressed as jag sulfate per quarter. The results indicate
a coefficient of variation ranging from 1 to 10$, again, increasing
with decreasing sulfate loading. If variation in loaded filter
area were the dominant source of variability it would be expected
that the C.V. would be invariant with loading in contrast to the
results obtained. Since quarters were cut to equalize loaded areas
rather than filter surface, some variation in the sulfate contri-
buted by the blank filter would be expected. However, the mean
*
sulfate blanks for all batches of EPA Grade glass fiber filters
are < 0.7 yg/cm2 suggesting negligible contribution to the vari-
ability observed. We conclude therefore, that, except for the most
lightly loaded sample, the observed variability reflects principally
the variability of the analytical method. For the exception,
variability in sulfate loading and/or loaded filter area contributes
roughly equally.
To minimize observed variability between quarters for extraction
method comparisons, the 2h filter samples used were restricted to
those from sites within California's South Coast Air Basin likely
to exhibit relatively high sulfate levels. As a result, the minimum
sulfate per quarter proved to be about 1700 jag. To further reduce
variability the borders from all filters were removed "before
quartering leaving a rectangle containing only loaded filter area.
This was then quartered to provide four quarters equivalent in an
area (about 98 cm2) within an estimated \%. Based on the above
and results for sample k, Table 2, the variability between quarters
*
This is the manufacturer's designation and does not imply approval by the
U.S. Environmental Protection Agency.
11
-------
Table 3
= i d
2ry of Sulfate from Extraction of Quarters from 8 x 10" Glass Fiber Filter (yg SOi^ ) '
5 Min
Filter Ultrasonic
1 2600
2 231*0
3 2760
1* 3l*l*0
3350
5 3UlO
1 J*bOO
2 10500
3 5890
1 1*320
2 2960
3 2880
1* 2110
5 7120
6 1*11*0
1 6500
2 ll*00
3 1*070
1*030
i iyiu
2 2560
3 2760
1* 5280
5 2290
6 3670
7 2970
30 Min
Ultrasonic
3030
2690
3190
3550
3650
5000
111*00
651*0
6520
6690
1*910
3280
3200
3130
2370
8070
1*780
80YO
1680
1+720
2080
2800
3^50
5700
3200
1*170
3330
31*00
60 Min
Reflux
2950
2970
3260
371*0
3860
5990
12200
661*0
5070
301*0
3190
21*80
8230
1*670
7280
1780
1*1*10
21bO
2820
3190
5790
2720
1*120
3200
60 Min
Shaking
3020
2920
2920
3l*UO
3780
3830
5750
5600
12300
6300
61*30
5170
32l*0
3320
2700
8670
1*1*90
7&90
171*0
5350
1*810
22l*0
2980
31*1*0
5810
3100
1*320
35^0
Re-extraction After
60-Minute Reflux
First13 Second0
73.7
39-2
72.0
60.6
118
106
302
122
69.7
70.1
55-8
1*9.1
158
73.6
121
71.8
77-1
3l*. 3
72.3
66.3
ll*2
69. ^
11*3
251*
< 7.5
< 7.5
< 7.5
< 7-5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
< 7.5
3859
1*368
1*1*07
1*533
101
are 2l|-hour hi-vol samples obtained with EPA Grade filters from J. ¥endt, GARB.
aute reflux method but with final volume 50 ml.
lute ultrasonic extraction method but in 15 ml H^O.
values for a given sample indicate replicate analyses.
-------
Based on mean recovered sulfate levels, the average efficiency of
each procedure for extraction of water soluble sulfate from 2H-hour
hi-vol glass fiber filter samples is given in Table U.
The high sulfate recovery by the 60-minute mechanical shaking method
prompted further data evaluation to determine if this efficiency
might "be subject to variation with sample type as implied by
differences in sampling location. Mean recoveries of sulfate,
expressed as percents of the total water-soluble sulfate, are given
in Table 5• Data have been arranged to list sites by increasing
distance from Long Beach. Since in some cases only 3 or 5 samples
were obtained at a given site, no firm conclusions may be made.
However, the data suggest that with samples obtained at increasing
distance from Long Beach, mechanical shaking in cold water becomes
relatively more efficient. Since the Long Beach area is one es-
pecially rich in hydrocarbons because of oil fields and refineries,
aerosols may be especially oily and difficult to wet by aqueous
extraction. This offers at least a simplistic rational for these
observations. Further studies would be needed to confirm the validity
of this site-specificity.
We conclude from these studies that the 30-minute ultrasonic and
60-minute reflux procedures are not significantly different in
efficiency. Except for relatively unusual sampling locations,
the mechanical shaking procedure usually provides the highest
sulfate recovery. Since this method offers decided advantages in
simplicity and reduced equipment cost compared, at least, to the
reflux method, consideration of its wider use should be given.
15
-------
Table
Average Efficiencies for Extraction of Water Soluble Sulfate
from 2U-hour Hi-vol Glass Fiber Filter Samples
Method Efficiency
5 min. ultrasonic 85.6
30 min ultrasonic 96.9
60 min reflux 97.8
60 min shaking 100
a. Calculated relative to sulfate recovered by successive extractions
by the reflux method.
16
-------
Table 5
Mean Recoveries of Sulfate by 60-Minute
Mechanical Shaking as a Function of Location
Distance Inland from Pacific
Site
Long Beach
Anaheim
Downtown
Los Angeles
Santa Ana
Pasadena
Coast at .Long Beach, KM
0
22
32
32
U2
N
3
5
6
7
3
Mean % Recovered
96.0
98.8
101.6
102.6
105.6
a. Relative to the total water soluble sulfate determined by successive
extractions by the reflux method.
IT
-------
Further work is needed to determine the significance of sulfate
recoveries above 100% observed by this method with nearly 70%
of the samples. One possible cause would be greater extraction
of a negative interferent (e.g. Ba 2) from the glass fiber filter
at reflux compared to room temperature.
18
-------
IV. SULFATE EXTRACTION EFFICIENCY STUDIES WITH TEFLON LO-VOL FILTER SAMPLES
A. Introduction
Previous AIHL studies have evaluated procedures for the extraction
of sulfate and nitrate from filter samples collected on kj mm
cellulose acetate membrane and glass fiber hi-vol filters. With
the low-volume samples a micropercolation technique was shown to
be about 99% efficient for sulfate extraction. With glass fiber
filters ,sulfate extraction efficiency for micropercolation was 92%,
equivalent to that by a simpler procedure, immersion of the sample
in water at 80°C in sealed test tubes.
Aqueous extractions of sulfate from Teflon filter samples is more
difficult, in relation to cellulose ester and glass fiber filter
samples, because of its non-wettability in water and low density
causing it to float. Stevens et al reported use of ultrasonic
extraction for 20 minutes with water at room temperature, the
filter being held submerged and unfolded by a fluted Teflon pipe,
the end resting on the unloaded edge of the filter. The samples
were continuously moved within the bath because of concern about
variability in agitation with bath location. By comparison with
x-ray fluorescence analysis for sulfur, the efficiency for sulfate
extraction was inferred to be 95-98%.
B. Experimental Procedure
The present study has emphasized aqueous extraction procedures
potentially useful for processing large numbers of samples such
as anticipated from the dichotomous sampler network. The procedures
19
-------
evaluated were:
1. Mechanical shaking 60 minutes with an Eberbach platform shaker
with samples in test tubes and filters cut into quarters
(Appendix D).
2. Ultrasonic extraction for 30 minutes, with sample pre-wet with
methanol and weighted down with a glass rod (Appendix E).
3. Same as 2 omitting methanol.
4. Heating at 80°C for two hours in sealed test tubes with filters
cut into quarters (Appendix F).
The study employed two groups of 2k filters each. Group A included
. *
24-hour samples collected without size segregation in Berkeley
approximately 27 m above street level using 47 mm Fluoropore filters
mounted in an open face filter holder. Group B were 24-hour
fine particulate samples collected on 37 mm Fluoropore filters
using two dichotomous samplers. Samplers were located 8 meters
east of the eastern edge of the San Diego Freeway, in West Los
Angeles, about 2 meters above the roadway. Average traffic volume
on this freeway is 250,000 cars/24 hours. By employing samples
of diverse origin it was intended to provide differing matrices
for sulfate extraction. Since higher sulfate levels were anticipated
for the Los Angeles samples, anc. since sufficient extract for inter-
method comparison was needed, the Los Angeles samples were extracted
in 20 ml H20. Berkeley samples used 10 ml H20.
*
Samplers at 2151 Berkeley Way, in downtown Berkeley.
20
-------
Six samples from each of the two locations were extracted "by each
of the four extraction procedures. Following vacuum filtration the
quantity of sulfate in the solution remaining wetting the particulate
matter on the filter was determined by weighing the Teflon filters
wet and after drying to constant weight at 105°C. With the weight
difference and the sulfate concentration measured in the extract,
the sulfate remaining in the aqueous phase on the filter was calculated.
Sulfate in the extract was determined "by an automated methylthymol
blue (MTB) procedure and by the AIHL microsulfate method. The MTB
procedure followed was that described in Reference 2 in which no
effort is made to linearize the working curve by altering reagents
but the working curve is fit by non-linear regression. By eliminating
sample dilution the procedure was applied in the 0-10 yg/ml range.
No correction for sample color was made. However, the resulting
error is expected to be small relative to that with hi-vol samples
(e.g. 6% error with samples described in Table 19, page 65).
Calculation of extraction efficiency employed the mean results
from the two procedures.
To determine the sulfate remaining unextracted after the first
extraction, two filters were extracted together by the mechanical
shaking procedure (l), but for 2k hours, in 5 ml H20. Filters
were combined since it was considered likely that the sulfate in
extracts from single filters would be too dilute for reliable
analysis. Extracts from this second extraction were analyzed with
the Dionex 1C using a sample pre-concentrator.
21
-------
The efficiency of the initial extraction "by each method was
calculated as follows:
% Extraction Efficiency = -——A—;rr- x 100
DI + (b2-ti)
where Si = total micrograms of sulfate recovered by the initial
extraction from extracting two filters separately.
82 = total micrograms of sulfate recovered in the second
extraction of the two filters combined.
R = total calculated sulfate remaining in aqueous phase
clinging to the two filters after initial extraction.
C. Results
The levels of sulfate recovered by the second extraction were
relatively low, and, in principal, might reflect the influence of
contamination or other artifacts. Accordingly, a relationship
was sought between the sulfate recovered by the first and second
extractions for filter pairs (Figure l). The figure suggests a
significant positive relationship between levels of sulfate re-
covered by successive extraction. Thus sample contamination or
other artifacts do not appear to be influencing the level of
sulfate recovered by the second extraction. The Los Angeles
samples yielded substantially higher second extraction sulfate
levels.
The extraction results are tabulated in Tables 6 and 7 for Berkeley
and Los Angeles samples, respectively. In contrast to expectations
22
-------
Fi gure 1
COMPARISON OF SULFATE RECOVERED BY SUCCESSIVE EXTRACTIONS
OF FLUOROPORE FILTER SAMPLES
40
• Los Angeles collection site
0 Berkeley collection site
oo
30
60
C
•
« 20
2
•u
X
W
o
u
10
©
•
^
•
©
©
••—©T
©
0
50
100
150 200 250
First Extraction (jig Sulfate)
300
350
400
-------
Table 6
Recovery of Sulfate and Efficiency of Extraction with Lo-Vol Teflon Filter Samples from BerkeleyJ
ro
from
yg S0i+ in
yg
in Second
Average
Extraction
Mean (C .V., %}
Extraction
Method
Mechanical
Shaking
Ultrasonic
with MeOH
Ultrasonic
Heating at
80°C
a. Samples
Filter
Coded
B1A
B2A
B3A
B5A
B6A
BIB
B2B
B3B
BUB
B5B
B6B
B1C
B2C
B3C
BUG
B5C
B6C
BID
B2D
B3D
B5D
B6D
collected
First Extraction Residual Extract Extract of
AIHL MTB Wetting Filter Filter Pair by 1C
76.2
290
250
82. U
101
101
1U5
"^
25.3
UU.o
96.5
192
79.9
131
28.9
36.0
119
3U.8
llU
6l.O
52. U
62.2
199
121
during the
b. Calculated using mean of MTB
c. Excluded
from mean
*
76.8
266
96.5
108
9U.U
1U5
U7.2
52.5
96.8
161
199
81.8
1^5
*^U 9
U3.6
138
66.8
135
62.7
67^6
210
lUO
period 10/23/78
and AIHL method
1.03
0.7U
0.9U
0.22
0 .5U
l.OU
0.82
0.09
0.11
0.33
0.50
0.58
0.71
0.73
0.08
0.32
1.78
0.29
1.38
0.62
1.2U
1.30
1.99
1.78
to 11/28/78.
results from the
_L4 . ~>
U.25
.
.15
2. 35
0.75
12. U
2.10
29.0
2.50
U.10
2.UO
3.65
first extraction,,
Efficiency for Efficiency
Filter Pairb for Method
96.7
99.1
r\Q <-,
90.7
99.2
99.8
96.8
100.0
( N c
( (J-->/
99.8
99. U
100.1
100.0
expressed as
98.2 (1.3)
0
98.6 (1.6)
99-9 (0.1,
99-8 (O.U)
a percent.
d. Mean of two values.
-------
Table 7
Recovery of Sulfate and Efficiency of Extraction with Lo-Vol Teflon Filter Samples from Los Angeles
Method
ro
yg 301+ from
Filter First Extraction
Codeb AIHL MTB
yg bU^ in
Residual Extract
Wetting Filter
Mg SO^ in Second
Extract of
Filter Pair by 1C
Average
Extraction
Efficiency for
Filter Pairb
Mean (C.V.f.)
Extraction
Efficiency
for Method
Mechanical
Shaking
Ultrasonic
with MeOH
Ultrasonic
Heating in
80°C
L1A
L2A
L3A
L4A
L5A
L6A
LIB
L2B
L3B
L4B
L5B
L6B
L1C
L2C
L3C
L4C
L5C
L6C
LID
L2D
L3D
L4D
L5D
L6D
a. Samples collected
b. Calculated
64.0
74.1
ill
6l.O
64.0
144
92.0
138
126
86.9
152
139
123
115
33.0
178
109
28.2
17.7
42.1
39.7
21.8
54.4
43.0
during the
using mean of MTB
68.2
77.4
124
64.0
72.0
157
95.8
146
138
94.0
159
149
136
124
32.2
190
116
27.0
24.0
48.4
43.6
22.4
47.4
44.4
period 11/1/78
and AIHL method
0.06
0.21
0.25
0.21
0.30
0.28
0.01
0.04
0.01
0.19
0.00
0.05
0.08
0.15
0.03
0.42
0.37
0.08
0.17
0.09
0.17
0.14
0.24
0.42
to 11/13/78.
results from the
19-1
4.60
28.2
3.90
31.2
34.0
31.0
31.4
24.7
12.6
16.8
33.3
first extraction,
88.3
97-7 92
88.8
98.4
87.8 92
89.8
89.0
87-5 87
85.2
83.9
79.5 80
75.8
expressed as a percent.
(5-8)
(6.1)
(2.2)
(5.1)
-------
the Berkeley samples proved to be more heavily loaded with sulfate.
With the Berkeley samples, efficiencies were > 98% "by all methods.
However, the Los Angeles samples appeared to be more difficult to
extract for sulfate, with method (h), heating at 80°C, significantly
less efficient than the other procedures. The latter averaged about
90%. The use of methanol to pre-wet the filters did not produce a
consistent improvement in efficiency. The choice between mechanical
shaking and ultrasonic extraction can probably be based on con-
venience and personnel costs.
The lower efficiency for extraction of Los Angeles vehicular effluent-
enriched samples compared to those from Berkeley parallels the results
for hi-vol samples (Section III); the reduced efficiency may reflect
the influence of oily particulate matter (e.g. aerosolized lubri-
cating oils) in encapsulating other particulate constituents.-
Alternatively, elevated levels of relatively insoluble lead sulfate
might contribute to reduced recoveries of sulfate in initial ex-
tractions. However, assuming the solubility product for pure
_6
PbSOij in water, 1.8 x 10 , to be applicable to the atmospheric
sample, then for reasonable levels of FbSO^ (e.g. < 50% of the
total Pb) lead sulfate formation would not be a significant source
of reduced sulfate recovery.
26
-------
V. STABILITY OF PRE-MIXED REAGENT FOR TURBIDIMETRIC SULFATE ANALYSIS
A. Introduction
The "barium sulfate turbidimetric methods (AIHL Methods 6l and 75)
employ a pre-mixed reagent composed of glycerol, HC1 and water for
stabilizing the colloidal suspension. A ruggedness test performed
3
during the previous phase of this grant compared freshly prepared
reagent with a batch prepared two years earlier. The results
indicated that the choice of the old or new reagent was the source
of 78% of the total variance observed, a result which exceeded the
variance of the dummy variable at the 95% confidence level.
Based on these results, the stability of this reagent on storage
was evaluated. For this purpose a batch of reagent prepared in
October 1977 was used periodically to prepare working curves for
turbidimetric analysis. Changes were sought in slopes, intercept
and standard error of the estimate, S , from linear regression
for standards in the range 300-1600 yg/20 ml samples.
B. Results
Results with the three parameters are plotted against reagent age
in Figures 2-U. The working curve slope, sometimes defined as the
sensitivity of the method, displayed no significant trend over
about 15 months. Over this period, the intercept increased
slightly. The most interesting results are those for S . Except
for the cluster of data at about 200 days, S remained approxi-
mately constant. The exception was associated with the use of a
different B & L Model 20 spectrometer, the usual instrument being
temporarily unavailable. Thus the precision of results may be
influenced by the instrument used.
27
-------
no
CO
c*l
o
fl
X
W
on
.5750
.5700
.5650
.5600
.5550
.5500
.54:0
.5400
.5350
.5300
.5250
.5200
,
EFFECT OF PRE-MIXED REAGENT AGE IN TURBIDIMETRIC SULFATE ANALYSIS
WORKING CURVE SLOPE VS. TIME
J_
±
±
JL
J_
1
I
J
0 35
70
105 140 175
210
245 280 315 350 385 420 455 490
TIME (DAYS)
Figure 2
-------
INTERCEPT
o
o
u>
Ul
o
Ul
tjl
H
a i
tj W
U)
to
Ul
to
oo
o
(Jl
u>
Ul
o
UJ
OO
Ul
to
o
Ul
Ul
O
ON
o
O
^J
o
O
OO
o
O
sO
o
O
o
to
o
Ul
o
w
Tl
Tl
W
O W
js a
si
G
r
Tl
>
w
*:
1/5
-------
EFFECT OF PRE-MIXED REAGENT AGE IN TURBIDIMETRIC SULFATE ANALYSIS
WORKING CURVE Sy.x VS. TIME
LO
O
Sy.x
.0275
.0250
.0225
.0200
.0175
.0150
.0125
.0100
.0075
.0050
.0025
* /
I
I
I
I
I
I
I
I
I
I
J
0 35 70 105 140 175
210 245 280
TIME (DAYS)
Figure h
315 350 385 420 455 490
-------
G. Conclusions
The shelf-life of the pre-mixed reagent is at least 15 months.
The sensitivity of the method to'choice of 2-year old or a new
reagent in the previously performed ruggedness test may have been
due to factors other than aging (e.g. contamination).
31
-------
VI. EVALUATION AND IMPROVEMENT OF A TURBIDIMETRIC METHOD FOR SULFATE
USING SULFAVER IV1*
A. Introduction
3
Preceding studies under this Grant included development of a
protocol utilizing SulfaVer IV . Accuracy and precision with
EPA sulfate audit strips,working range,and agreement with other
methods were also determined. The current program has provided
additional evaluation of this technique employing a ruggedness
IT
test ' to optimize the method and an intermethod comparison with
the optimized procedure.
B. Ruggedness Test
The ruggedness test protocol followed is similar to that given in
Appendix C of Reference 3. The eleven factors evaluated are given
in Table 8. Factors B and F, organics and colloidal clay, were
included because prior studies of other turbidimetric sulfate
methods indicated these to be interferents. The levels of organics
used, absorbance 0.025 and 0.1 per cm at HOO nm, compares to a
maximum value of 0.07 per cm observed for extracts from St. Louis
samples. The levels of colloidal clay, 200 and 1000 yg kaolinite/
20 ml, compare to a maximum 90° light scattering (at 600 nm) for
St. Louis extracts, expressed in the corresponding clay concen-
tration, of 1300 pg/20 ml, following filtration through a fine
glass frit. All samples were mixed with SulfaVer Iv and their
absorbances determined in the same container ,thereby avoiding
potential errors introduced vith multiple transfers (c.f. AIHL
Methods 6l and 75, Reference 3).
32
-------
Table 8
FACTORS FOR EVALUATION IN RUGGEDNESS TEST OF SULFAVER METHOD?
Factor
A = Spectrophotometer
B = Organics concentration
C = HC1 concentration
D = Reaction time
E = Sulfate level
Q
F = Colloidal clay level
G = Loss of SulfaVer
H = Shaking speed
I = Shaking time
J = Reaction vessel and
Spectrophotometer cell
K = Dummy
Low (-)
B & L Model 20
Absorbance 0.025 /cm
at hOO nm
Zero
5 minutes
300 yg/20 ml
200 yg/20 ml kaolinite
Discard 50% of SulfaVer
from each pillow
270 oscillations/minute
3 minutes
Screw cap test tube
(25 x 150 mm)
High (+)
B & L Model 21
Absorbance O.I/cm at ^00 nm
0.3 N
20 minutes
1300 yg/20 ml
1000 yg/20 ml kaolinite
Discard no SulfaVer
90 oscillations/minute
1 minute
Cuvet (25 x 150 mm)
a. Concentrations and absorbances shown are for 20 ml samples prepared to
simulate hi-vol filter extracts.
b. Yellow organics isolated from hi-vol filter aqueous extractions as
described in Appendix G, Reference 3.
c. Used to simulate the source of turbidity seen in some filter extracts.
d. Using an Eberbach platform shaker.
33
-------
The mean results of each of the twelve experiments, each run three
times, calculated by third order regression for the working curve,
were expressed as the ratio of the observed to the theoretical
sulfate level. The effect of each factor was evaluated as the
difference between mean results for the runs with high (or plus)
and low (or minus) levels. Table 9 ranks the observed effects,
squares the effects to estimate the variance of the method due to
that effect and determines the proportion of the total due to each
factor.
The results show the most significant sources of variability in
result to be the sulfate level and the choice of spectrophotometer.
At the low sulfate level (300 yg/20 ml) results averaged about 30%
high causing the measured effect for Factor E to be substantially
negative. The substantial variance observed for Factor A (choice
of spectrophotometer) followed from average results by the B & L 21
which were 26% too high compared to about 3% too low with the
B & L 20. ¥e believe this reflects primarily an interaction with
the effect of sulfate and interferent levels; the results for
Factor A are strongly influenced by the results for three runs at
300 yg/20 ml sulfate (Runs k, 5 and 10). It is more reasonable
that the high results in these three runs (average U8% positive
error) resulted from the relatively high interferents and low
sulfate levels rather than selection of spectrophotometer. The
value for S , the standard error of the estimate , for the
yx'
working curve for six trials was (7-9 to 22) x 10~3 with the B & L 21
and (13 to 21) x 10~3 with the B & L 20. Thus the degrees of scatter
were about equal with the two instruments. Results for Factor A
are, therefore, considered to be insignificant. Results for other
3U
-------
Table 9
RESULTS OF RUGGEDNESS TEST OF SULFATE ANALYSIS BY TURBIDIMETRY
Factor Identification
E Sulfate level
A Spectrophotometer
I Shaking time
G Loss of SulfaVer
K Dummy
B Organics concentration
D Reaction time
C HC1 addition
F Colloidal clay
H Shaking speed
J Reaction vessel and cell
,E2 as
E
-0.326
0.275
0.153
0.129
-0 . 115
0.10U
-O.OT58
-0.06*15
-0.031*!
-0.0285
0.0133
E2
0.106
0.0753
0.0233
0.0166
0.0131
0.0107
0.0057
O.OOUl
0.0011
0.0008
0.0001
% of Total
Ul.3
29.3
9-1
6.5
5.1
U.2
2.2
1.6
O.U
0.3
o.oU
E = Effect of variable = difference between mean results for runs
with high (or plus) and low (or minus) levels.
35
-------
factors differ from those for the dummy factor "by less than a factor of
two or show variance less than that of the dummy. Therefore, only
sulfate level (Factor E) is considered to be a significant source of
variance.
•p
It may be noted that discarding half of the contents of the SulfaVer IV
pillows had no significant effejt on results. In trials with 10 pillows,
the variability (C.V.) in contents transferred to the samples was 10$.
Thus loss of 50% would be greater than would ever be expected. Clearly,
the quantity of reagent is in large excess compared to that required
at up to 1300 yg/20 ml.
C. Comparisons with Prior Study
The preceding study of the SulfaVer Iv method employed an analytical
procedure analogous to that in AIHL Method 6l, (i.e. reagent and sample
were mixed in graduated cylinders and transferred to 2 cm cylindrical
cuvets for turbidity measurement with a B & L Model TO spectrophotometer)
This procedure yielded recoveries within 10$ of the theoretical sulfate
values using EPA audit strips in the range 300 to 1700 yg/20 ml, with a
C.V. < 6%. Furthermore, the working range, based on precision and
relative accuracy of a single atmospheric extract diluted to various
concentrations, was determined to be from 180 to at least 1^00 yg/20 ml
(accuracy within \%, C.V. <_ 6%}. Finally, in analysis of 2\ atmos-
pheric hi-vol filter samples, a median C.V. of 5-3$ was found.
The present ruggedness test shoved positive error of about 30% at
300 yg/20 ml with Factor J (reaction vessel and cell) without signi-
ficance. However, in contrast to the evaluation of accuracy,
precision and working range described above, all solutions in the
36
-------
ruggedness test contained added colloidal clay (200 or 1000 yg/20 ml)
and yellow organics (absorbance 0.025 or O.I/cm at UOO nm). While
the atmospheric extract previously used to determine working range
also showed absorbance at UOO nm, when diluted to provide <_ 300 yg
sulfate/20 ml the absorbance was below 0.025 cm 1 at ^00 nm, the
lower level in the ruggedness test due to organics. Thus the results
from the ruggedness test are probably not in conflict with the prior
work.
Aside from the problem of accuracy, the lack of significant sensitivity
of the method to the level of colloidal clay appears surprising.
However, in interference studies employing barium chloride-glycerol-
HC1-H20 (e.g. AIHL Method 75) for sulfate analysis by turbidity, with
750 to 1200 yg/20 ml sulfate, a change from 200 to 1000 yg/20 ml
colloidal clay (kaolinite) caused only a 9 to 12% decrease in observed
sulfate. The effect of this change in clay concentration at 200
yg/20 ml sulfate was, however, large (-82%). If the change at
300 yg/20 ml sulfate, as used ir. the ruggedness test, were similar
to that at 200 yg/20 ml, then a significant variance due to Factor F
would be expected. Further work is needed to explore interference
effects with the SulfaVer IV method.
D. Procedure for Sulfate Analysis vith SulfaVer IV
Based on prior studies and the ruggedness test,a procedure has been
prepared (AIHL Method 79) suitable for sulfate analysis of 2U-hour
high volume filter samples. It is included as Appendix G. The
extraction procedure specified, mechanical shaking at room
temperature, is based on results obtained in Section III.
37
-------
VII. STUDIES WITH THE DIONEX MODEL 10 IOE CHROMATOGRAPH
A. Introduction
During the preceding phase of this grant the Dionex Model 10 ion
chromatograph (1C) was evaluated for use in analyzing hi-vol filter
samples. The principal focus of 1C studies in the current grant
period vas on evaluating its use with smaller samples such as
obtained with low-volume, dichotomous samplers. Two approaches
were evaluated, a larger sample loop and a sample pre-concentrator.
In addition, the system was modified by replacing the 500 mm anion
separator column with one of 250 mm length to reduce analysis time.
Before beginning work at lower sulfate ranges, the problem of
3
interference effects by nitrate and drifts in calibration curves
was re-examined.
B. Data Reduction Techniques and Interference Effects of Nitrate on
Sulfate Determination
Previous studies directed toward analysis of hi-vol filter samples
established a 2-3% positive interference in sulfate measurement
when nitrate was present at equal concentration (by weight).
This interference was observed using peak heights measured from an
extrapolated base line (Figure 5A). Before modifying the 1C for
lo-vol sample analyses, interference effects were evaluated at
higher NO3 /SO^ ratios with data reduction by the previously used
peak height method as well as the trailing peak and integration
techniques (Figure 5A-C).
38
-------
U)
VO
Peak Height
SOL
NO-
Trailing Peak
Peak Area
SO
Figure 5 SULFATE IATA REDUCTION PROCEDURES FOR 1C.
-------
For thir; r;tudy the 1C war; calibrated twice each day using standards
without nitrate and the data reduced usinp; the corresponding
calibration. As discussed below, the calibration shift within one
day was usually significant. Experimental conditions were as
follows:
Range: 10 ymho (linear scale)
Column: 3 x 100 mm precolumn and 3 x 500 mm anion separator
Eluent: 0.002U M Na2C03 + 0.0030 M NaHC03
Eluent Flow Rate: 2.5 ml/min
Sample Loop: 30 yl
Temperature: 35°C
Recorder: 1.0 V full scale (equivalent to Dionex meter)
Integrator: Autolab Minigrator set for
peak width = 99
sensitivity = 99999
baseline = 1.0
trailing peak =0.0
The results summarized in Table 10 confirm the significance of nitrate
interference using the peak height method with extrapolated baseline.
Use of the trailing peak and peak area methods yield reduced error.
However the precision of the trailing peak height method appears
somewhat better compared to peak areas. We conclude that for
atmospheric samples containing nitrate at concentrations ^_ that of
sulfate the trailing peak method should be used. Alternatively, the
1C should be operated to achieve base line separation of the peaks.
-------
Table 20
Interference Effect of Nitrate on Sulfate Determination
(% Error )a
SO^
(yg/ml)
20
20
20
20
20
20
20
20
20
20
ho
hO
hO
hO
hO
hO
80
80
80
N03
(yg/ml)
ho
ho
ho
ho
Mean:
60
60
60
Mean:
100
100
100
Mean:
80
80
80
Mean:
120
120
120
Mean:
160
160
160
Mean:
Peak Trailing
Height13 Peakc
+ 11.1 +
+ 6.1
+ 8.8
+ 6.7
+ 8.2 +_ 2.3
+ 10.6
+ 10.9
+ 8.2
+ 9-9 ±1.5
+16.5 +
+ 15.1
+ I.h.6 +
+ 15. U ± 1.0 +
+ 7.3
+ 6.7
+ U.7
+ 6.2 +_ i.U
+10.6 +
+ 8.8
+ 7.5
+ 9-0 +_ 1.6
+ 7.5 +
+ h.h
+ h.2.
+ 5.h +_ 1.9
1.3
1.3
0.2
0.7
0.2 +_ 1.1
0.9
1.3
0.7
1.0 +_ .3
0.3
2.3
2.3
0.1 ± 2.3
0.0
O.U
2.0
0.8 ± 1.1
0.3
0.7
1.0
0.5 +_ .7
l.U
0.8
0.6
0.0 +_ 1.2
PeaK
Area
+ 9.2
d
- 1.0
+ 7-0
+ 5.1 ±
+ 5.6
- 3.U
- 0.7
+ 0.5 +_
+ U.2
- 6.0
- 1.6
- i-1 ±
+ 3.2
- 0.5
+ 2.U
+ 1.7 ±
+ 2.5
- 0.3
+ O.U
+ 0.9 ±
d
d
d
5.U
U.6
5.1
2.0
1.5
a. 100 x (Observed-True) /(true)
b. From extrapolated base line (Figure 5A)
c . See Figure 5B .
d. Integrator did not function correctly.
hi
-------
Aside from partial overlap of the sulfate and nitrate peaks it was
considered possible that the presence of nitrate might increase the
sulfate peak by reducing its retention time and, therefore, sharpening
the peak. The influence of nitrate on sulfate retention time is
shown in Table 11. Results are from data obtained on a single day
and, except as noted are means +_ 1 a for two trials. The data
indicate no significant effect of nitrate on sulfate retention time.
Furthermore, increasing sulfate concentration, alone, did not
influence retention time.
C. Accuracy of Sulfate Determination by 1C and the Effect of Shifting
Calibration Curves
Preceding studies demonstrated a persistent positive bias in sulfate
determinations using EPA sulfate audit strips. The positive error
was especially pronounced for a sample with N03 /SO^. weight ratio
of ca. 2. Since the reported accuracy might have been influenced
by the method of data reduction (the peak height method as shown
in Figure 1A), a set of audit strips was extracted (30 minutes
ultrasonic) and analyzed by 1C using the trailing peak method.
Instrument conditions were as given in section (A) above. Two
calibration .curves were obtained daily from three sulfate standards
(10, 20 and ho yg/ml). Data from samples was reduced using the most
recent calibration. Results frcm analysis of the audit strips are
given in Table 12. Consistent with prior observations the 9000
series samples with high nitrate levels showed the largest positive:
error. However the mean ratio, observed/theoretical of 1.03 for
this series by peak heights compares to 1.15 previously reported.
1*2
-------
Table 11
The Effect of Nitrate on Sulfate Retention Times
N03~ S04~ Retention Time
(yg/ml) (yg/ml) _ (seconds) _
10 0 5HOa
20 0 538 +_ 5
20 Ho 539 +. 0
20 60 539 ± 0.7
20 100 5H5 1 9
ho 0 537 ± 2.8
HO 80 539 + 5
Ho 120 5Hl H^ H
80 0 53H +_ 11
80 160 529 + 0
a. Single trial
H3
-------
Table 12
ANALYSIS OF EPA SULFATE AUDIT STRIPS BY THE DIONEX 1C
a
Sample
712-7000 Series
9000 Series
712-5000 Series
Theoretical _ _ Mean
Value S0tf~/N03~ . Observed Value (yg/ml)
SOU /ml) Wt. Ratio Pk Ht Tr Pk Pk Area
C.V. (%}
Pk Ht Tr Pk Pk Area
10.0
Ik. 9
^5.0
k.2
0.60
2.8
10.1
15.3
U1+.5
10.1
lit. 6
1+3.9
10.0
1U.3
kk.2
3.1
3.9
1.6
3.1
3.0
1.6
5-3
3.9
1.2
Observed/Theoretical
Pk Ht Tr Pk Pk Area
1.01 1.01 1.00
1.03 0.98 0.96
0.99 0.98 0.98
a. Results are means for four strips from each series extracted by 30 minutes ultrasonic extraction in 50 ml water.
-------
The peak height technique gave slightly higher results than with the
other techniques. None, however differed, on average by more than
U% from the true values. Precision, as measured "by coefficients of
variation for four strips, was approximately equal by all techniques.
The above results suggest that something other than nitrate inter-
ference was the dominant cause of the positive errors previously
reported. In seeking this cause we have evaluated changes in the
instrument response within one day's operation which could lead to
shifts in calibration curves. Table 13 .indicates the shifts observed.
In all cases the change is in the direction of increased instrument
response. The magnitude of the shift is from U to 11% over 2-3 hour
periods and affects both peak heights and areas. The variability
of the shift appears to depend upon operating parameters such as
eluent flow rate. For example, increasing eluent flow rate increases
the rate of depletion of the suppressor column, resulting in a faster
calibration shift per unit time. The rate of change is also high
just after and prior to suppressor column regeneration, and after
the pump is shut off for any reason. Thus, depending on the time
between calibration and sample analyst's a given sample would show
varying error but usually in the positive direction. We consider
this change in calibration to be the principal source of positive
errors in sulfate analyses with 1C.
Discussions with E. Johnson, Dionex Corp. indicate that this
phenomenon may relate to changes in retention of protonated sulfate
on the suppressor column. As the protons of the suppressor column
are replaced by sodium from the eluent, the resulting neutralized
-------
Table 13
CHANGE IN PEAK HEIGHT AND AREA WITH TIME WITH SULFATE STANDARDS*
First Calibration Second Calibration Change in Calibration (%}
(yg/ml) Pk. Ht. Pk. Area Pk. Ht. Pk. Area Pk. Ht. Pk. Area
10
20
1*0
8.0
7.1
16.2
15.2
3^.8
32.2
1978527
1881778
3992628
3787132
81*36753
77^8065
8.1*
8.3
17-7
16.8
36.7
35-7
2178198
2108238
1*289327
1*1711*39
877H02
8502336
+ 5.0
+16.9
+ 9.3
+10.5
+ 5.5
+10.9
+10.1
+12.0
+ 7.1*
+10.2
+ i*.o
+ 9-7
a. Results shown for two trials made on successive days.
b. Interval between calibrations 2-3 hours.
-------
resin exhibits an affinity for the HSO^" ion. The result, he suggests,
is to sharpen somewhat the sulfate peak leading to increased peak
height. However, peak areas should not, in principle, be affected
in contrast to our observations. Peak areas, Johnson notes, are
inherently less precise to use with a pulsing chromatographic system
compared to peak heights. He recommends installing a long section
of Teflon tubing to attenuate pump pulses and then to use peak areas.
Since in our work peak, areas have no clear advantage regarding
accuracy and in our previous studies have been found somewhat less
precise, the peak area technique will not be used.
There appears to be sources of negative drift in the instrument
response as well (e.g. temperature decreases in the eluent supply
which was not thermostated). These may, occasionally, partially
or completely offset the positive drift associated with consumption
of the suppressor column.
D. Sulfate Analysis Using a Pre-concentrator Column
1. Introduction
The sample pre-concentrator column (SPC) consists of a length
of glass tubing, approximately 50 mm long, of which 35 mm is
packed with anion resin of the type used in the separator columns.
Like the latter columns, the SPC is 3 mm in I.D. and has Altex-
Durrim plastic fittings, with about 100 mm lengths of Teflon
tubing on each end, for connection to the slide valve in place
of the usual sample loop. The volume of the column and tubing
is about 0.25 ml.
-------
In use, one injects the sample through the sample port, passing
it through the resin, which retains and. separates to some degree
• the anions from the sample. Then the slide valve is switched
so that the eluent passes through the SPC, eluting out the
collected ions. The analysis then proceeds as usual.
2. Advantages
The advantages of the pre-concentrator compared to the 30 yl
sample loop previously used are:
a. The ability to analyze very low sulfate concentrations since
the total in a large sample (e.g. 15 ml) can be analyzed.
For example, samples of distilled water were determined to
have 0.00^ to 0.01 yg sulfate/ml by this technique.
b. Unlike the loop, sample volume is easily varied. Thus if
a sample exceeds the working range a smaller aliquot is
injected rather than having to dilute the sample.
3. Disadvantages and Problems Encountered
a. The pre-concentrator had to be regenerated after about
125 samples (using a solution containing 0.5N each of
NaHC03 and Na2C03).
b. Injection of sample from the pre-concentrator causes a
mementary drop in line pressure and, therefore, eluent flow
rate. As a result, one or more spurious positive peaks occur
initially, followed by a negative peak about a minute later.
-------
c. Sample injection is more complex; a syringe is first loaded
with a known volume of sample and injected followed "by an
injection of distilled water to flush the syringe, tubing
and SPC to insure all sample is on the resin.
d. Manual injections of relatively large volumes are made
against a large back pressure leading to operator fatigue.
Dionex recommends use of a pump.
e. After about 100 injections the resin in the pre-concentrator
had been compressed to about one-half its initial volume.
This increased the back pressure and created more free void
in the glass tubing.
Procedure for Sample Injection Using the Sample Pre-concentrator
Column (SPC)
The procedure adopted for sample injection with method evaluation
studies and sample analyses is as follows:
a. Equipment
(l) Disposable syringe, 12 cc volume, "Monoject", Sherwood
Medical Industries or equivalent, used without a needle.
(2) Adjustable dispensing pipet, Pipetman, or equivalent,
providing ca. 1% precision over desired range.
b. Set the pump so that the minimum pressure during injection
is 50 to 100 psi, consistent with an acceptable eluent flow
rate.
-------
c. Set the injector switch GO "Load" position. Inject 2.0 ml
of distilled water through the SPC with the syringe.
d. Draw in 3 ml minus sample aliquot of distilled water into
the syringe.
e. Using the pipet, dispense the sample aliquot (0.10 ml to
3.0 ml) into the tip of the syringe. For this, the pipet
tip must fit inside and seal the syringe. -Inject the sample
into the SPC.
f. Carefully withdraw the syringe from the port and draw in
another 3.0 ml of distilled water* and inject this into the
port.
g. After the baseline is stable, set the injection switch to
"Inject". After 2.0 min, flip the switch to "Load".
h. If a spurious negative or positive peak (due to injection)
interferes with the elution of the sample peaks, change the
time in g. above to eliminate this problem.
*The total amount of distilled water is not critical but should be minimized
to decrease operator fatigue and compression of the SPC resin.
-------
5. Test Operation Parameters
Range: 30, 100 and 300 ymho (linear scale)
Columns: 3 x 100 mm precolumn + 3 x 250 mm anion column
Eluent: 0.0030 M NaHC03 + 0.002U M N02C03
Elution rate: 2.5 ml/min
Recorder: 0.20 and 1.0 volt full scale
Bath: 35°C and ambient
Integrator: As given in Section A
6. Range and Precision
The ranges used, the resulting approximately 80% of full scale
sulfate level, and our typically obtained error are given in
Table lU. Figure 6 illustrates a working curve for analysis of
low concentration sulfate samples.
1. Nitrate Interference in Sulfate Analysis
Using sulfate samples in the 0-2 yg range with nitrate/sulfate
ratios of 2 w/w, the error in sulfate by the trailing peak
height method averaged 1.8$, and by peak areas, 5-8$.
8. Atmospheric Sample Analysis
Because of its potential for analysis of very low concentration
samples, the Dionex with pre-concentrator was used for analyses
of extracts obtained in studies comparing extraction procedures
for low-volume Teflon filter samples. Results from the analyses
are included in Section IV.
51
-------
Table ih
Instrument Settings, Sulfate Level for 80% Full Scale and Accuracy awith Standards
Using the Dionex 1C with Pre-concentrator
Approx.
Instrument Range Recorder Full Scale, Effective Range, Full Scale Sulfate, Mean Error
ymho
100
300
30
volts
1.0
0.2
0.2
ymho^
100
60
6
MS
ho
20
2
Peak Ht.
3.9
3.U
1.9-7.7°
Peak Area
U. 3
U.8
2.0-7.^°
a. Measured by the mean of the absolute values for percent differences between the true sulfate
ro concentrations used for calibration and the values obtained by linear regression of the
working curve constructed between 10 and 100% of the Q0% full scale sulfate value shown.
b. Effective range, in pmho, equals (instrument Range) x (Recorder Full Scale Voltage).
c. Range of mean values from three trials.
-------
TYPICAL WORKING CURVE FOR SULFATE ANALYSIS BY DIONEX 1C WITH PRE-CONCENTRATOR
80
70
60
50
D
4->
rt
6
K
^
rt
40
30
20
10
Conditions
30 ymho scale
0.2 volt recorder f.s.
35'C
i i i
0.50
1.0
Total Sulfate.^g
Figure 6
53
1.5
2.0
-------
9. Conclusions
The use of a pre-concentrator permits analyses of sulfate samples
containing very lov (e.g. < 0.1 yg/ml) sulfate levels. However,
for analysis of samples to be expected from 2k hour collection
with dichotomous samplers (e.g. > 5 yg/ml) this method is
unnecessarily complex compared to use of a sample loop.
E. Sulfate Analysis of Lo-Yol Filter Samples Using an 0.5 ml Sample Loop
1. Introduction
For sulfate analysis of lo-vol filter samples such as provided
by dichotomous samplers a method providing accurate and precise
results in the range 0-20 pg/ml is needed. Increasing the
sample loop size from 0.03 to 0.5 ml provided a sufficient
increase in instrument sensitivity to accomplish this by 1C.
As in the studies with the pre-concentrator, the column size
was reduced from 500 to 250 mm to decrease the analysis time
required.
The present study has evaluated l) the resolution of anions
under these condition, 2) the interference by S03 and N03 ,
3) the linearity and day-to-day reproducibility of the working
curves, U) precision and accuracy using EPA sulfate audit strips,
and 5) intermethod comparison with two other procedures. As in
previous work the trailing peak height method proved to be the
most accurate and precise; the results reported here employed
only this procedure.
-------
2. Resolution of Sulfate and Nitrate Related Species
Experimental conditions for this work were as follows:
Range: 100 ymho (linear)
Recorder: 0.5 volts full scale
Columns: 3 x 100 mm precolumn + 3 x 250 mm anion column
+ 6 x 250 mm anion suppressor column.
Eluent: 0.0030M NaHC03 + 0.002^M Na2C03
Elution Rate: 2.5 ml/min
Column and Detector Temperature: 35°C
Sample Loop: 0.5 ml
(A detailed procedure for 1C of samples is included in Appendix H)
Under these conditions, retention times for sulfate, nitrate and
nitrite were as given in Table 15. With the shorter separator
column, nitrite, nitrate and sulfate are still separated. The
retention time difference for sulfate and nitrate, 88 seconds,
compares to 160-179 seconds with the 500 mm separator column.
Sulfite and sulfate have identical retention times using the
standard eluent mixture.
3. The Working Curve for Sulfate Analysis
The working curve for sulfate analysis in the range 0-20 yg/ml is
shown in Figure 7- In contrast to results obtained with the
30 yl sample loop, the working curve appears to be distinctly
non-linear below 5 yg/ml. The errors in sulfate determination
resulting from a regression equation for the best single line
are compared to those using a two straight line fit (0-2 yg/ml,
5-20 yg/ml) in Figure 8. For this purpose "observed" values
55
-------
Table 15
Retention Times for Sulfate, Nitrate and
Related Species Using 250 mm Anion Separator Column
Species Retention Time (sec) _A-c.(sec_)a
nitrite (N02~) 157 + 5 125
nitrate (N03~) 19U +2 88
sulfite (S03~) 282 +^5 0
sulfate (S04=) 282 + 5
a. Relative to sulfate.
56
-------
90
WORKING CURVE FOR SULFATE ANALYSIS
BY DIONEX 1C WITH 0.5ml SAMPLE LOOP
80
70
60
50
U
o
40
30
20
10
Conditions:
lOO^imho scale
0.5 volt recorder f.s.
35°C
Single line fit
—— — Two straight line fit
I
5 10
Sulfate Concentration,
Figure 7
57
15
20
-------
ACCURACY AS A FUNCTION OF SULFATE CONCENTRATION
BY DIONEX 1C WITH 0.5ml SAMPLE LOOP
w
TJ
o
D
a>
1.80
1.70
1.60
1.50
1.40
1.30
1.20
1.10
1 on
f
|
-
t
~\
\
I
Single line regression (0-20.ug/ml)
Two line regression (0-2, 5-i
0.90 -
0.80 -
0.70 -
0.60 -
sZ*
10
15
True Sulfate Concentration,jug/ml
20
Figure 8
58
-------
are sulfate concentrations for the standards used in constructing
the working curve obtained from the regression equation for the
working curve. The ratio of observed to true sulfate levels may
be used as a measure of accuracy at varying concentrations. By
both procedures, accuracy remains within 1.0% in the range 1 to
20 yg/ml with errors tending to be negative in the range 1-10
yg/ml. As expected from Figure 7, the difference in accuracy
by the two techniques is maximized at 5 yg/ml. Based on these
results and the need to minimize numbers of standards run to
achieve reasonable sample output, the balance of this study
employed two working curves. One was obtained from a straight
line regression fit for standards 5 to 20 yg/ml and was used
for analysis of samples >_ 5 yg/ml. The second was the straight
line based on two points, 0* and 5 yg/ml; it was used for analysis
of samples < 5 yg/ml. Where substantial numbers of samples below
5 yg/ml are expected, additional standards in this range should
be used.
The day-to-day change in the working curve for the range 5 to 20
yg/ml is given in Table l6. At the time of this work the anion
separator column had been used for about ^00 samples and thus
was nearing the end of its useful life. The relatively large
change observed in slope is considered symptomatic of the column
age. Within one day's operation, the changes observed were
insignificant, however.
*The twice distilled water contained < 0.01 yg/ml sulfate as measured using
the sample pre-concentrator.
59
-------
Table 16
Day-to-Day Change in Working Curve of Dionex 1C
Using 250 mm Column for Sulfate Analysis'1
Date
1-3-79 3.906 -It.07
Slope Intercept r S
~ y.x
a. For Standards in the range 5 to 20 ug/ml.
b. Mean of four trials.
c. Mean of two trials.
60
0,9991 0.9718
l-,fc-79 4.124b _3.lK>*
1-8-79 i
0.9996b o.76T5b
0.9997 0.6383
-------
k. Accuracy and Precision Using EPA Sulfate Audit Strips
Aqueous extracts of EPA sulfate audit strips were prepared by
30-minute ultrasonic extraction and diluted, if necessary, to
obtain samples covering the range 2 to 15 yg/ml. Four extracts
were prepared for each level. An additional set of four extracts
was prepared for 15 yg/ml to provide samples high in nitrate.
The results are summarized in Table IT and indicate a coefficient
of variation of < 5% and accuracy within 15% in all cases. No
interference from the high nitrate concentration was seen in
the 9000 series samples.
61
-------
Table IT
Sulfate Analysis of EPA Audit Strips by 1C
Using the 0.5 ml Sample Loopa
Sample
712-7000 series
712-7000 series
9000 series
712-5000 series
Theoretical
Value (yg/ml)
2.0b
10.0
14.9
15.0C
Wt. Ratio
SOu'VNOa"
U.2
U.2
0.6
2.8
c.v.U)
i.od
0.8
2.8
5.1
Obs . /Theoret .
0.86d
0.95
0.92
0.92
a. Results are mean values for four extracts at each level using the
trailing peak height method. All strips extracted in 50 ml H20
vith further dilution as noted. Working curve based on standards
5, 10, 15 and 20 ml, except as noted.
b. Diluted fivefold.
c. Diluted threefold.
d. Working curve based on standards 0 and 5 yg/ml.
62
-------
VIII. INTERMETHOD COMPARISONS WITH ATMOSPHERIC SAMPLES
A. Hi-vol Filter Sample Methods
•p
To compare results by the SulfaVer IV method described in Appendix G
with those by a previously evaluated procedure, extracts from samples
collected on EPA Grade glass fiber filters were analyzed by this
technique as well as the Colovos-MTB procedure , operated in the
0-80 yg/ml range. With both methods, analyses were done with three
determinations obtained on separate days. The 2U filter samples
employed were described in Section III. Extracts for intermethod
comparison were prepared by pooling solutions remaining from the
evaluation of extraction procedures.
The MTB method, is often employed without correction for the
-| Q
initial color of the sample. Accordingly, samples were run without
reagent to determine absorbance. Sulfate results were calculated
both with and without correction by subtraction of initial sample
absorbance.
The results, expressed as yg/ml of aqueous extract, are given in
Table 18. The undiluted extraci s covered a concentration range
from ca. 30 to 230 yg/ml. With the SulfaVer IV method, samples
exceeding 55 yg/ml were diluted prior to analysis. For the MTB
method, samples exceeding about 80 yg/ml were diluted prior to
analysis.
The results are compared as ratios of means relative to corrected
results by the Colovos-MTB method in Table 19. On average, the
SulfaVer IV method yielded results which were 10$ higher.
63
-------
Table 18
Results of Intermethod Comparison with Hi-Vol Filter Samples
(yg sulfate/ml)
Colovos-MTBa
Uncorrected Corrected SulfaVer IV
32.it +_ 0.1;
U2.1 +_ O.U
50.0 +_ 0.4
55-7 ± 0.8
.5U.9 ± 0.5
55.0 +_ 1
58.9 ±0.5
61.6 +_ O.U
62.lt +_ 0.3
62.9 ± 0.2
65.0 + 1
66.5 ± 0.1
71.9 ± O.U
73.5 ± 0.3
81.8 + 0.9
91.8 +_ 0.7
93.6 +_ 0.3
99.2 +_ O.U
105 + 1
116 +_ 1
128 +_ 1
lU2 +_ 2
156 i 2
230 + 3
29.7 ± o.H
39-5 ± O.lt
45.9 +_ o.u
51.1 + 0.8
51.5 ± 0.5
52.0 ±1
55.0 +_ 0.6
57-6 +_ 0.5
58.2 + o.U
58.8 + 0.1
60.0 +_ 1
62.2 +_ 0.6
66.6 ± O.U
68.7 1 0.6
76.2 +_ 0.9
85. U ± 0.7
87.6 +_ 0.3
93.0 i 0.5
98.8 +_ 0.9
110 +_ 1
122 +_ 1
13U +_ 2
150 ±2
22U + 3
32 + 1
U5 + 1
5U +_ 1
55 + 1
59 + 1
59 ± 3
60 + 1
80 +_ 20
66 + 2
6U + 3
65 + 6
69 ± U
75 ± 3
78 ± 2
90 +_ 2
9U +_ 2
95 + 5
102 +_ 6
105 ±. i
123 + 3
127 ± U
lU6 +_ 2
162 + 9
230 + 20
a. Samples diluted to < 80 pg/ml before analysis where necessary.
b. Samples above 55 yg/ml diluted to provide solutions below this
concentration.
6U
-------
Table 19
Average Agreement arid Precision of
Sulfate Methods with Hi-Vol Filter Samples
Colovos-MTB (corrected)
Colovos-MTB (uncorrected)
SulfaVer IV
Median
Range
Ratio of Means C.V. (%) C.V. (%}
1.00 1.0 0.1 to 2.1
1.06 0.9 0.3 to 2.0
1.10 3.2 1 to 20
a. Results expressed relative to those for the Colovos MTB method.
65
-------
Correction of the MTB values for initial absorbance caused a decrease
P
averaging 6%. Thus the uncorrected MTB and SulfaVer IV method
differ, on average by k%. These results may be compared to those
previously found with slightly different versions of the SulfaVer IV
and MTB methods ; in that case the SulfaVer Iv results averaged 6%
higher than those by Midwest Research Institute version of the MTB
method.
The precisions of the methods are expressed by the median and range
of the coefficients of variation. The 3.2% C.V. for the SulfaVer
Iv method compares to 5-3$ found with the earlier version of this
method. A more detailed comparison of the SulfaVer IV and Colovos-
MTB (corrected) data is given in Figure 9• The results by the two
. methods are highly correlated. The substantial positive intercept
leads to larger percentage differences for lower concentration
samples.
B. Low-Volume Filter Sample Method
To compare the 1C procedure for lo-vol filter samples to other
procedures, extracts from the 2k samples collected in Los Angeles,
described in Section IV, were analyzed by an MTB (0-10 yg/ml)
*
method, the AIHL microchemical method with ion exchange pretreat-
•7
ment and by the 1C. Analyses by the 1C procedure were performed
with three determinations on separate days. Because of the limited
sample available, analyses by the other procedures were done with
a single trial.
See page 19
66
-------
SCATTER DIAGRAM OF RESULTS WITH HI VOL FILTER SAMPLES
COMPARING SULFAVER IV AND COLOVOS MTB SULFATE RESULTS
200 r
150
.£
J£
*Z 100
o
J
3
C/3
50
Sulfaver = 1.03 (Colovos-MTB) + 5.91
r = 0.997
Sy.x=3.50
50 100
Colovos MTB (^g/ml)
150
200
Figure 9
6.7
-------
The results, expressed as yg/ml of aqueous extract, are given in
Table 20 together with the ratio of means relative to the MTB
procedure. The extracts covered a concentration range from about
1 to 9 yg/ml. On average, the three methods agreed within 6% with
both the AIHL micro and Dionex 1C results somewhat lower than by
the MTB method. This trend for lo-vol samples by 1C differs
somewhat from that with hi-vol samples previously observed. The
latter study found, on average, 1C results using a 30 yl sample
loop to be about h% higher than by two automated MTB procedures.
Similarly, the earlier study found results by the AIHL micro method
to be on average, about 2% higher than by the two MTB procedures.
The precision of the 1C method with lo-vol samples, as expressed by
the median coefficient of variation, was 6.5$ (range 1.1 to 33.7$).
A more detailed comparison of the methods is given in Figure 10.
The results by the three methods are highly correlated with the
lower results by the 1C and AIHL methods persisting throughout the
concentration range studied.
68
-------
Table 20
Results of Intermethod Comparison with Teflon Lo-Vol Filter Samples
(yg sulfate/ml)a
Sample ID
LID
LHD
L6C
L3C
L3D
L2D'
L6D
L5D
LHA
L1A
L5A
L2A
LHB
LIB
L5C
L3A
L2C
L1C
L3B
L2B
L6B
L6A
L5B
MTB
1.20
1.12
1.35
2.18
2.22
2.
3.
3.
3.
3.
37
20
Hi
60
87
H.70
H,
5,
6.
.79
.82
,22
6.20
6.81
6.78
7-30
7.HH
7.85
7-93
9-51
AIHL Micro
0.88
1.09
l.Hi
1.65
1.98
2.10
2.15
72
05
20
29
3.71
H.35
H.60
• H3
,53
,77
6.1H
6.32
6.91
6.9H
7.18
7.6l
8.92
Dionex 1C
0.92
0.91
1.38
1.72
2.02
1.99
1.93
2.12
2.92
3.23
3.36
3.62
H.3H
H.35
5.56
5.85
5.87
6.68
6.31
6.72
6.59
7.H6
7.20
8.88
+ 0.15
+ 0.19
+ 0.18
+ 0.58
+ 0.19
+ 0.18
+ 0.21
+ 0.11
+ 0.32
+ 0.39
+ 0.36
+ 0.3H
+ 0.08
+ 0.06
+ 0.11
+ O.H5
+ 0.11
+ 0.21
+ 0.18
+ 0.16
+ O.lH
+ 0.37
+ 0.18
+ 0.10
Ratio of Means:
1.00
0.9HH
0.928
a. All samples analyzed without dilution.
mean + 1 a for three determinations.
Results for Dionex 1C
69
-------
Figure 10
SCATTER DIAGRAMS OF RESULTS WITH LO VOL FILTER SAMPLES
USING THREE SULFATE METHODS
10
9
8
7
1 6
y 5
X
w
Z 4
Q
3
2
1
0
DIONEX 1C =0.939 (MTB) - 0.053
r =0.998
Sy.x =0.157
I
I
I
I
456
MTB (pg/ml)
10
10
9
y
•5
AIHL MICRO =0.911 (MTB)+0.1 54
r =0.995
Sy.x =0.236
456
MTB (jig/ml)
10
-------
References
1. B. R. Appel, E. L. Kothny, E. M. Hoffer and J. J. Wesolowski, Comparison
of Wet Chemical and Instrumental Methods for Measuring Airborne Sulfate,
Interim Report. EPA-600/2-76-059 (1976).
2. B. R. Appel, E. L. Kothny, E. M. Hoffer and J. J. Wesolowski, Comparison
of Wet Chemical and Instrumental Methods for Measuring Airborne Sulfate,
Final Report. EPA-600/7-77-128 (1977).
3. B. R. Appel, E. M. Hoffer, M. Haik, W. Wehrmeister, E. L. Kothny and
J. J. Wesolovski, Improvement and Evaluation of Methods for Sulfate
Analysis, Final Report (1978).
k. Selected Methods for the Measurement of Air Pollutants, Public Health
Service Publication No. 999-AP-ll (196*0.
5. Tentative Method for the Determination of Sulfates in the Atmosphere
(Automated Technicon II Methylthymol Blue Procedure) Environmental
Protection Agency, Quality Assurance Branch, July 15, 1977-
6. G. Colovos, et al, Anal. Chem. hQ_ 1693 (1976).
7. E. M. Hoffer, E. L. Kothny and B. R. Appel, Simple Method for Microgram
Amounts of Sulfate in Atmospheric Particulates, Atmos. Environ. 13
303 (1979).
8. C. Brosset and M. Ferm, "An Improved Spectrophotometric Method for the
Determination of Low Sulfate Concentration in Aqueous Solutions",
Swedish Water and Air Pollution Research Laboratory, U02, 2k Gothenburg,
Sweden (19 7*0.
9. Barium Chloranilate Method for Determination of Sulfates in the Atmosphere,
March 1976. Prepared for U.S. EPA Environmental Monitoring and Support
Laboratory, Research Triangle Park, North Carolina.
10. H. Small, et al, Anal. Chem. Hj_ 1801 (1975).
11. Hach Chemical Company. Ames, Iowa.
12. Determination of Sulfate in High Volume Particulate Samples: Turbidi-
metric Barium Sulfate Method, AIHL Method 6l, revised July 1976.
13. Determination of Sulfate in Glass Fiber High Volume Filters, Bay Area
Air Pollution Control District Method S-h-2 (June 30, 1976).
1*». E. M. Hoffer and B. R. Appel, AIHL Report No. l8l "A Comparative Study
of Extraction Methods for Sulfate and Nitrate from Atmospheric Parti-
culate Matter", November 1975•
71
-------
15. H. W. Hermance et al, Environ. Sci. and Technol. 5_ ?8l (1971).
16. R. K. Stevens and T. G. Dzubay, Atmos. Environ. 12_ 55 (1978).
17. W. J. Youden, Statistical Techniques for Collaborative Tests, Association
of Official Anal. Chem. (1973).
18. Technicon Industrial Method 118-71W, Technicon Industrial Systems,
Tarrytown, NY.
72
-------
APPENDIX A
Ultrasonic Extraction Procedure
The procedure used was taken from EPA-EMSL Method "Tentative Method for
the Determination of Sulfates in the Atmosphere (Automated Technicon II
Methylthymol Blue Procedure)":
The filters are removed from the folder, opened flat, and cut into
1.9 "by 20.3 (3A x 8 in.) strips using a pizza cutter. The filter
should be cut with the particulates face up. One or more filter
strips are placed in a 60-ml (2-oz) glass bottle. A random 5-10%
of the filters should be extracted in duplicate for use as quality
control samples. Fifty milliliters of distilled water are pipetted
into each bottle. The bottles are then closed with polyseal caps.
The samples are placed in the sonic bath, which should be refilled
before each set of extractions with fresh cold tap water to the level
of the liquid in the bottles. The sonic bath is operated for 30 min.
The extracts are immediately vacuum filtered using the Buchner funnels
and the vacuum filtering apparatus. The samples are filtered directly
into polyethylene bottles. The filters should not be washed or
squeezed, and the filtrates are not diluted. After filtering is
complete, the polyethylene bottles are capped with polyseal caps and
stored upright until analyzed. The samples are stable at room
temperature for at least two weeks.
For the present study the above procedure was modified by employing 60 ml
Erlenmeyer flasks with ground glass stoppers in place of 60 ml glass bottles,
Filtration was done with a Millipore filtration apparatus using 0.7 ym
cellulose ester filters, discarding the filter after each sample.
During ultrasonic extraction 8 flasks were extracted simultaneously,
distributed uniformly around the bath.
73
-------
APPENDIX B
, *
Reflux Procedure from AIHL Method 6l
One-fourth of the filter is cut into about 5-cm lengths for ease in
handling and placed into the 125 ml boiling flask containing 50 ml of
distilled water. The sample is refluxed for 60 minutes. The hot
solution is filtered through a Whatman No. k2 filter paper which has
been previously rinsed free of sulfate with at least 50 ml of boiling
distilled water. The filtrate is collected in a 100 ml glass stoppered
graduated cylinder. Both the boiling flask and sample filter are rinsed
3 times with about 10 ml each of boiling distilled water. After cooling,
the final filtrate volume is brought up to 100 ml with distilled water.
For the present study this procedure was modified by filtration as
described in Appendix A.
-------
APPENDIX C
. *
Mechanical Shaking Procedure from BAAPCD Method S-4-2
Cut up one quarter of the exposed glass filter into strips of about
3/V by 1 1/2", place in 250 ml Erlenmayer flasks and add 50 ml
distilled water. Seal the tops of the flasks with parafilm and shake
the contents of the flask for one hour on the Burrel Shaker. Filter
the samples thru dry filter paper into any suitable container for
storage. Do not wash the residue or filter paper.
For the present study this procedure was modified by filtration as
described in Appendix A.
75
-------
APPENDIX D
Sulfate Extraction from Teflon Filters by Mechanical Shaking
The filters were cut into quarters in a laminar flow clean bench and
inserted into test tubes sealed with Teflon lined screw caps. Berkeley
samples used l6 x 120 mm plastic tubes and Los Angeles samples, l6 x 150 mm
glass tubes. To the Los Angeles samples was added 20 ml twice distilled
*
H20 and to the Berkeley samples, 10 ml. The tubes were shaken in
batches mounted horizontally on an Eberbach platform shaker at 90
oscillations/min for one hour. Sample filtration was performed using a
Millipore vacuum filtration apparatus.
*
The dead volume in the tubes was approximately equal for the Berkeley
and Los Angeles samples.
76
-------
APPENDIX E
Sulfate Extraction from Teflon Filters by
Ultrasonic Extraction with Pre-wetting with Methanol
Uncut 37 or ^7 mm filters were placed, -unfolded, loaded side up in 100 ml
plastic wide mouth containers. The Berkeley filters were wet by spotting
with 0.2 ml anhydrous methanol, and the Los Angeles samples, with O.H ml.
The filters were then weighted down with short sections of 3 mm glass
rod bent into a "V", touching the filter at two points. To Los Angeles
samples was added 20 ml twice distilled t^O and to Berkeley samples,
10 ml. The apex of the glass rod extended above the liquid level.
Filters were extracted for 30 minutes in batches of eight distributed
uniformly within an ultrasonic bath. The liquid level in the bath was
adjusted to be equal to that in the samples. Sample filtration was
performed using a Millipore vacuum filtration apparatus.
77
-------
APPENDIX F
Sulfate Extraction from Teflon Filters by Heating in Water at 80°C
Filters were cut in quarters and inserted into 16 x 150 mm Teflon lined,
screw capped test tubes. To Los Angeles samples was added 20 ml twice
distilled H20 and to Berkeley samples, 10 ml. Samples were heated two
hours at 80°C in thermostated heating blocks (Labline Inst. Co., #2090).
Samples were then shaken briefly by hand and allowed to cool overnight.
Sample filtration was performed using a Millipore vacuum filtration
apparatus. No effort to weight down the filters was made. During heating
filters are wet by condensing water vapor.
78
-------
AIHL Method
DRAFT
APPENDIX G
DETERMINATION OF SULFATE IN HIGH VOLUME
PARTICULATE SAMPLES USING SULFAVER
Analyte:
Application:
Matrix:
Procedure:
Date First
Issued:
Sulfate
Air Pollution
Air
Collection on filter by
high-volume sampler,
extraction with water
followed by turbidi-
metric analysis
Method No:
Working Range:
79
180 to >_ ll*00 yg
sulfate/20 ml
Detection Limit: Not determined
Precision:
Accuracy:
<_ 6% coefficient of
variation in working
range
Within 6%, on average,
using EPA Audit Strips
1. Principle of the Method
1.1 Atmospheric suspended particulate matter is collected over a 2^-hour
period on a 20 by 25-cm (8 by 10-inch) filter by using a high-volume
sampler.
P
1.2 A water extract of the filter sample is treated with SulfaVer from
which barium chloride forms a barium sulfate colloidal suspension.
The turbidity of the suspension is measured spectrophotometrically
at 500 nm.
1.3 The extract is filtered through a Millipore filter to eliminate
turbidity due to suspended particles or fibers.
l.U Barium sulfate formation and turbidity measurements are done in
test tubes (25 x 150 mm), thereby eliminating all sample transfers.
The procedure was developed by E. M. Hoffer. Evaluation of the procedure
is given in References 1 and 2.
Prepared by staff of the Air and Industrial Hygiene Laboratory Section,
State Department of Health Services, Berkeley, California.
79
-------
2. Interferences
2.1 Sample coloration and/or turbidity may interfere with the analysis.
These interferences are minimized "by filtration through a Millipore
filter and by measuring the absorbance (A}) of the filtrate before
the addition of the SulfaVer Iv. This value is subtracted from
•p
the absorbance (A£) of the sample after the addition of SulfaVer IV .
2.2 Sulfur-containing anions are generally strong positive interferents
2
probably due to air oxidation to sulfate.
2.3 Glass fiber filters contribute to observed sulfate both from a
2 3
"blank" value and by artifact sulfate formation. ' Artifact
sulfate can be minimized by employing pH neutral filters (e.g.
quartz fiber). With all filter types, the background or blank
sulfate concentration should be measured (Section 7-3-*0 for
every lot and type of sampling filter used and the results corrected.
3. Precision and Accuracy
3.1 The precision of the method was established by three determinations
on each of the extracts from 2k high-volume atmospheric samples
ranging in concentration from 2^0 to 1500 yg sulfate per 20 ml
solution. The median coefficient of variation was 5-3 (range 1.0
to 9.1$).
3.2 Accuracy was established by analyzing EPA audit strips (i.e filter
strips loaded with known quantities of sulfate). For solutions in
the range 300 to 1700 yg/20 ml, the ratio of observed to theoretical
concentration ranged from 1.00 to 1.10 with mean value 1.06.
3.3 The extraction procedure, mechanical shaking in water at room
temperature, was shown to extract, on average, 100% of the total
water soluble sulfate in 2k high-volume filter samples.
80
-------
U. Working Range
U.I Working range is defined as the sulfate concentration range providing
approximately constant coefficient of variation and "relative accuracy"
The latter indicates the accuracy of the method relative to the value
obtained in the optimal concentration range of the method. This is
determined using a pooled, concentrated atmospheric sample extract,
diluted to varying degrees.
k.2 This procedure yielded a relative accuracy within U% in the
concentration range 180 to 1^00 yg sulfate/20 ml solution with a
C.V. of <_ 6%.
5. Equipment
5.1 High-volume Sampler. A motor blower-filtration system with a
sampling head which can accommodate a 20 by 25-cm filter and capable
of sampling at an initial flow rate of about 1.7 m3/min (60 ft3/min).
5.2 Filters. 20 by 25-cm (8 by 10-inch) filters.
5.3 Wrist Action Shaker. Burrell Model CC, Burrell Corp., Pittsburgh, PA
5.^ Filter Assembly
5.^.1 Funnel, 300 ml, Teflon faced, Millipore Catalog No. XK10k72k
5.H.2 Base, Teflon faced, Millipore Catalog No. XX10U722
5.^.3 Spring clamp, anodized aluminum, Millipore Catalog No. XX10U703
5.U.U Stopper, Neoprene, No. h to fit Fisher Filtrator
5-5 Fisher Filtrator. low form (Catalog No. 9-788).
5.6 Millipore Filter. H7 mm plain white cellulose acetate, pore size
in range O.U5 to 1.2 ym.
5-7 Filtrate Receivers. 60 or 100 ml polypropylene bottles with liquid
tight caps.
81
-------
5.8' Screw Capped Test Tubes. 25 x 150 mm, Teflon-lined. The tubes
should be unscratched. Add a vertical fiduciary mark to permit
reproducible positioning in the spectrophotometer.
5-9 Pipets. 5, 10, 20 ml and other sizes as required.
5.10 Spectrophotometer. Bausch and Lomb, Model 21 or equivalent.
5-11 Repipet. 10 ml capacity.
5.12 Platform Shaker. Eberbach Model 6000.
6. Reagents
6.1 SulfaVer IVRPillows. Catalog No. 12.065 obtainable from Hach Co.,
Loveland, Colorado 80537.
6.2' Standard Sulfate Solution (lOOO pg sulfate/ml). Dry anhydrous
sodium sulfate at 105°C for U hours and cool in a desiccator.
Dissolve 1.1*79 8 of the dried sodium sulfate in distilled water
and dilute to 1 liter. This solution contains 1000 yg sulfate
per ml.
7. .Procedure
7.1 Sampling. Using the high-volume sampler, collect the particulate
matter from approximately 2,000 m3 of air. Twenty-four hours is
the usual sampling period. Note and record the air flow rates at
the start and end of the sampling period.
7.2 Sample Preparation. The sample filter should be delivered to the
laboratory unfolded in a glassine envelope. Take an aliquot of the
filter for analysis (Appendix 1 discusses sectioning the filter for
various analyses). Cut one-fourth of the filter (Quadrant A an
shown in Appendix l) into strips about 1 to 1.5 cm wide for ease
in handling. Place the strips into a 250 ml glass Erlenmeyer
flask containing exactly 50 ml of distilled water, cover with
82
-------
parafilm and shake for one hour on the Burrell shaker. Filter
through an unused Millipore filter, (dull side up) using the
Millipore filter assembly with a Fisher filtrator using vacuum
(Figure l). Place a 60 ml polyethylene container into the Fisher
filtrator to receive the filtrate.
7.3 Analytical Procedure
7.3.1 Pipet aliquots of the filtrates, normally 20 ml, into a
series of clean and dry screw-capped test tubes. When a
smaller aliquot is used, dilute to 20 ml with distilled
water.
7.3.2 Using the screw cap test tube as the spectrometer cell,
with the fiduciary mark aiding reproducible positioning,
determine the absorbance at 500 nm against distilled water.
•n
7-3.3 The contents of one SulfaVer IV pillow are to be added to
each of a batch of 12 samples in a test tube rack. To
facilitate transfer, attach pillows to a jig in which the
pillows are clipped at intervals corresponding to the
intervals of the tubes in the test tube rack. Tap the
pillows,to settle contents to the bottom. Cut off pillow
•tops and carefully transfer contents, simultaneously, to
a set of samples or standards contained in the test tubes.
Tighten caps in each tube and mount batch of tubes hori-
zontally in an Fjberbach shaker set at 90 oscillations per
minute. Shake 1 minute. After shaking, place the samples
in a vertical position and wait 20 minutes at room temperature
before reading absorbance (A2). Read batch of 12 tubes
83
-------
within 5 min. By spacing batches at about 5 minute
intervals, determinations per day are maximized.
7.3.^ A correction for the concentration of sulfate in the filters
must be made for each nev lot of filters. This value (B)
must be the average of at least 5 determinations using 5
filters from each lot of 100 filters using the entire
analytical procedure and must be subtracted as filter blank
(Section 9-2).
8. Standards and Calibration
8.1 Using the 1000 vig/ml sulfate standard prepare fresh weekly 100 ml
of the following working standards :
yg SOU~/20 ml ml of 1000 yg/ml standard
200 1.0
^00 2.0
600 3-0
800 U.O
1000 5.0
1200 6.0
7.0
Analyze 20 ml aliquots of these calibrating solutions together
with each day's samples.
8.2 Plot the difference in absorbance readings (A2-A\) on the vertical
axis versus the corresponding yg of sulfate on the horizontal axis
using a rectilinear graph paper. The relation between absorbance
and amount of sulfate should be approximately linear between 500
and litOO yg/20 ml. By restricting samples to this range, linear
Qk
-------
regression can be employed. For analyses in the range 180 to
1^00 yg/20 ml, a third order regression equation is used of the
form y = a + bx + ex2 + dx3 where x = yg/20 ml sulfate and
y = absorbance.
9. Calculations
9-1 Air Volume Calculation
a. For samples collected at altitudes less than 2000 feet above
mean sea level, use the calibrated air flow rate, which is
approximately equal to the flow rate under standard conditions
(760 Torr and 25°C).
b. For samples collected at altitudes of 2000 feet or greater,
i^
calibrate the high-volume sampler using the ARE procedure
which corrects the flow rate to standard sea level conditions.
c. Using the flow rate determined in (a) or (b) above, calculate
the air volume from the sampling time and the average of the
air flow rates at the start and end of the sampling period.
Where: Qj = air flow rate at start of sampling period (m3/min)
cubic feet per minute x 0.02832 = m3/min.
Q2 = air flow rate at end of sampling period (m /min)
cubic feet per minute x 0.02832 = m3/min.
t = sampling period (min).
V = sample volume in cubic meters (m3) at standard
conditions .
85
-------
9.2 Subtract Aj from A2 and, calculating from the regression equation
obtained in Section 8, determine the equivalent yg of sulfate (C)
in the aliquot. Calculate the concentration of sulfate in the 20
by 25 cm filter sample, in yg/m3 as follows:
yg sulfate/m3 =
3 _
! x F2 x C) - B
V
Tn_ „ total ml of filtrate
Where: Fi = •
1 ml of filtrate taken for analysis
_ total sample area of filter sample
2 sample area of filter quadrant analyzed
C = yg sulfate in aliquot of sample taken
B = yg sulfate/20 x 25-cm filter blank
V = air sample volume in m3 (determined as in Sec. 10.l)
10. References
Final Report, EPA Grant No. 805-^7-1 "Improvement and Evaluation
of Methods for Sulfate Analysis, B. R. Appel, E. M. Hoffer, M. Haik,
¥. Wehrmeister, E. L. Kothny, and J. J. Wesolowski, October 1978.
Final Report, EPA Contract No. EPA 68-02-1660, "Comparison of Wet
Chemical and Instrumental Methods for Measuring Airborne Sulfate",
B. R. Appel, E. L. Kothny, E. M. Hoffer and J. J. Wesolowski,
February 1976.
Final Report, Effect of Environmental Variables and Sampling Media
on the Collection of Atmospheric Sulfate and Nitrate, NTIS Reports
PB 286U80/AS and PB 286WJ1/AS.
Standard Procedure for the Calibration of Hi-vol Samplers and
Plotting of Flow Calibration Curves Corrected for Altitude,
California Air Resources Board, September 1975, Sacramento, Calif.
86
-------
To Vacuum
FISHER
FILTRATOR
-Millipore Filtration
Assembly
(47mm Millipore Filter
on Frit)
No. 4 Neoprene Stopper
-60ml Polyethylene Bottle
Support
Millipore Filter Assembly With Fisher Filtrator
Figure 1
87
-------
Cutting of Glass-Fiber Hi^h-Volume Filters
1. Remove the glass-fiber filter from the shipping envelope.
2. Using a clean cutting tool, preferably stainless steel, cut the filter in
half. Then cut one half into two equal quadrants as shown in Figure 1.
•P
cut
B
25 cm
T
20 cm
Figure 1
3. Use quadrant "A" for the determination of sulfate .
88
-------
APPENDIX H
Ion Chromatographic Analysis of Sulfate in the Range 0 to 20 ug/ml
1. Equipment
1.1 lon-Chromatograph. Dionex System 10 Ion Chromatograph, Dionex Corp.,
1228 Titan Way, Sunnyvale, CA 9^087.
1.2 Varian A-25 Recorder
1.3 Containers .• ^oz. polypropylene containers with plastic screw-caps.
l.U Filters (extraction). 0.^5 y disposable filter unit (Swinnex-25
Filter Unit, Millipore , or equivalent).
1.5 Syringes. 12 cc disposable syringes, without needle, graduations
0.2 cc. (Monoject Sterile Disposable Syringe, Cat. Ho. 512S, or
equivalent).
2. Reagents
2.1 Water. The water for all reagents and suppressor column rinse should
be distilled to a resistance of approximately 15 megohms, or conduc-
tivity of 0.1 to 1.0 micromho/cm or better. The water should be
filtered free of particles larger than 0.20 ym unless a pre-column
is used. Fill reservoir labelled "H20" in the chromatograph.*
2.2 Eluent. Prepare 0.003 M NaHC03-0.002U M Na2C03 solution as
follows: In a 2-liter volumetric flask, dissolve 1.008 g NaHC03
(sodium bicarbonate, MCB, Cat. No. SX320 or equivalent) and 1.0175 g
Na2C03 (sodium carbonate, MCB, Cat. No. SX395-CB705 or equivalent)
with distilled filtered water prepared as in 2.1 above. Invert
gently to dissolve, make to the mark with deionized water, mix.
Transfer to the eluent reservior Labelled "E^" or "E2" in the chroma-
tograph.* Add an additional 2 liters water and mix well.
*
When filling reserviors, avoid air bubbles which may cause pumps to lose
their prime—see instruction manual for this procedure.
89
-------
2.3 Regenerant. Prepare 1 N P^SOi, as follows: Into a 2 liter volumetric
flask containing approximately 1 liter of distilled filtered water,
introduce 111 ml of concentrated sulfuric acid, mix, cool. Make
to the mark with' deionized water, mix. Fill reservoir labelled
"Regenerant" in the chromatograph. Add an additional 2 liters
water to make a total of H liters, and mix well.
2.U Stock Standard Sulfate Solution. (1000 yg sulfate/ml). Dry
(NHtt)2S04 (ammonium sulfate, NBS certified) powder at 105°C for
k hours, cool in a desiccator. Dissolve 1.376 g of the dried
ammonium sulfate in distilled water and dilute to 1 liter.
2.5 Stock Standard Nitrate Solution. (1000 yg nitrate/ml). Dry KN03
(potassium nitrate, NBS certified) powder at 105°C for U hours,
cool in a desiccator. Dissolve 1.631 g of the dried potassium
nitrate in distilled filtered water and dilute to 1 liter.
2.6 Sulfate Working Standards. Prepare working standards of 0, 2,
5, 10, 15, 20 yg/ml sulfate concentrations. To obtain the best
accuracy and precision, weigh out the required amounts of stock
standards into small beakers, and then transfer the contents to
Class A volumetric flasks. Samples of higher concentration are
set aside, diluted and analyzed at a later time.
Working Standard ml or gms of Sulfate
yg/ml Sulfate Stock Std. Added
0 0
2 0.20
5 0.50
10 1.00
15 1.50
20 2.00
After addition of stock standards to a 100 ml Class A volumetric
flask, add sufficient double distilled water to the mark.
90
-------
3. Analytical Procedure
3.1 Chromatograph Parameters
Range: 100 yniho
Recorder: 0.5 volts full scale, 10 inches/hour chart speed
Columns: 3 x 100 mm pre-column
3 x 250 mm anion separator column
6 x 250 mm anion suppressor column
Eluent: 0.0030 M NaHC03 + 0.002U M Na2C03
Elution Rate: 2.5 ml/min
Column and Detector Temperature: 35°C
Sample Loop: 0.5 ml
Under these conditions 20 yg/ml sulfate yielded a peak height of
80$ of full scale.
3.2 Put the toggle switch on the front panel of the chromatograph in
the LOAD position; using a syringe, inject 2 ml of sample solution
into the injection port. Leave the syringe in place during
chromatography.
3.3 Using the OFFSET COARSE or FINE knobs, adjust the indicator needle
on the SPECIFIC CONDUCTANCE meter to 0.0.
3.1* Flip the toggle switch to the INJECT position, at the same time
start the integrator or strip chart. After 15-30 seconds, flip
the toggle switch back into the LOAD position.
3-5 Record the sample I.D. on the chart.
3.6 After the run is completed, rinse the sample loop with 3 ml of
deionized filtered water with the toggle switch still in the
LOAD position.
3-T Inject the next sample as described in Sections 3.2 through 3.6.
91
-------
4. Chromatograph Start-Up. (Review lon-Chromatograph Instrument Manual)
4.1 Using the regulator on a dry air or nitrogen compressed gas
cylinder, adjust the pressure to 90-100 psi for the air actuated
valves in the chromatograph. Flip the toggle labelled AIR, on the
front panel of the chromatograph, to the ON position.
4.2 Flip the toggle labelled POWER to the ON position.
4.3 Place the toggle labelled FLUSH down.
4.4 Place the toggle labelled Tr. .• down.
LUAl)
4.5 Place the toggle labelled
WATER
4.6 Place the toggle labelled £2 up if EI is empty.
4.7 Place the toggle labelled E2 down if reservoir E^ is full.
4.8 Place the toggle labelled ANALYT up_.
4.9 Place the toggle labelled SUPPRESS up_.
4.10 Flip the toggle labelled PUMP to the ON position. Adjust the vernier
dial on the front pump so that the flow rate is appropriate, e.g.,
2.5 ml min. (The flow rate may be checked by disconnecting the
output tubing from the suppressor column and placing the tubing in
a graduated cylinder). Allow the system to run for 30 minute, or
until the baseline drift is reasonably stable. Check for leaks
in the tubing connections. Wear safety glasses when opening the
column door. Check reservior levels.
4.11 Turn MODE switch to LIN.
4.12 Turn yMHO FULL SCALE to 100, initially, as in Section 3.1.
4.13 Set SPECIFIC CONDUCTANCE needle to 5-0 with the OFFSET COARSE and
FINE knobs. (Allow sufficient positive baseline to account for
any negative drift).
92
-------
5. Standards and Calibration
5.1 Inject 3 ml of each of the standards described in paragraph 2.6.
Record the reading by measuring the recorder trace (chromatogram)
peak height.
5.2 Calculate linear regression lines for the standards from 5 to
20 yg/ml, based on the trailing peak height method. The response
in this range is linear.
Calculate additional regression lines, for the standards from 0
to 5 yg/ml, as above. The Dionex response in this range is
curvilinear and the results, therefore, less accurate using
linear regression.
Alternately and preferably, set aside the samples below 5 Vg/ml
and rerun, using a more sensitive scale (e.g. 10 ymhos) and
standards of 0, 1, 2, 3, 5 yg/ml.
5-3 To check for calibration drift over a day's run, rerun 5 and 20 yg/ml
standards at 2 hour intervals. If the change in the regression
slope is greater than 3$, calculate additional regression lines as
needed, based on the new standards.
6. Regeneration of Suppressor Column
6.1 At the end of each day's run, the suppressor column may require
regeneration as indicated by a color change in the column resin
bed from tan to whitish tan, or by a swift rise in the conductance.
6.2 On the chromatograph, make the following settings:
6.2.1 Flip toggle switch labelled PUMPS to the OFF position.
Set the switch labelled MODE to ZERO.
6.2.2 Flip toggle switch labelled SUPPRESS to the down position.
Check the liquid levels in the regenerant and "H20" reservoirs.
93
-------
6.2.3 Set TIME MIN indicators to 10 on REG side and 50 on RIN side.
6.2.U Depress the green button labelled START, the rear pump should
begin pumping. Set the vernier on the rear pump to approxi-
mately 90.
6.2.5 The cycle of regeneration is now automatic. At the end of
the cycle, the pump will stop and the suppressor column will
be tan colored. The cycle may be stopped prematurely by
depressing the red colored button labelled RESET.
7- Chromatograph Shut Down
T.I Flip the toggle switch labelled PUMP to the OFF position.
7.2 If the regeneration cycle is in process, and premature termination
is necessary, depress the red button labelled RESET.
7.3 Flip toggle switch labelled POWER to the OFF position.
7.U Flip AIR toggle switch to the OFF position.
7.5 Turn off regulator on compressed air or nitrogen cylinder.
7.6 Protect the integrator from dust using a plastic cover.
7.7 If the chromatograph is to be shut down for a long period of time
(e.g., 2 months) rinse both columns with distilled water beforehand.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
2.
3. RECIPIENT'S ACCESSION NO.
4.
LE AMD SUBTITLE
Improvement and Evaluation of Methods for Sulfate
Analysis - Part II
5. RPPORT DATE
April 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
B.R. Appel, E.M. Hoffer, W. Wehrmeister, M. Haik,
J.J. Wesolowski
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Air and Industrial Hygiene Laboratory Section
California Dept. of Health
2151 Berkeley Way
Berkeley, CA 94704
10. PROGRAM ELEMENT NO.
A09A1D
11. CONTRACT/GRANT NO.
Grant 805-447-1
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Research and Development
Environmental Monitoring Systems Laboratory
U.S. EPA
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOC
Oct. 1978 - Mav 1979
14. SPONSORING AGENCY CODE
JCOVE.RED
:inal
Report -
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Methods for extraction of sulfate from glass-fiber hi-vol and Teflon lo-vol
samples were evaluated. Efficiencies were found to vary with sampling location up
to 20%. Mechanical shaking in water at room temperature was significantly more
While Teflon
did not signifi-
using SulfaVer IV
Its precision
but its accuracy
Model TO ion
efficient than ultrasonic or reflux techniques with hi-vol samples.
filters are not wet by water, pre-wetting of filters with methanol
cantly enhance sulfate extraction. A turbidimetric sulfate method
was evaluated for ruggedness, precision and intermethod agreement.
was at least equal to that of a conventional turbidimetric method,
was somewhat less, especially at lower sulfate levels. The Dionex
chromatograph was evaluated for low level sulfate analysis using both a sample pre-
concentrator and large (0.5 ml) sample loop. The latter was the preferred technique
for samples <_ 20 yg/ml. Accuracy was within 15% in the range 2 to 20 yg/ml with a
median C.V. of 6.5% for 24 atmospheric samples. This range will permit sulfate
analysis of 24 hour fine particulate samples collected with dichotomous samplers.
Use of a sample pre-concentrator permitted analysis of samples containing < 1 yg/ml
sulfate.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution and Control
Environment
Air Monitoring
b.IDENTIFIERS/OPEN ENDED TERMS
Measurement Methods
Sulfates
c. COSATI Field/Group
68A
43F
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS {This Report)
UNCLASSIFIED
21. NO. OF PAGES
95
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
-------
-------
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA-335
Official Business
Penalty for Private Use, $300
4TH CUSS
SPECIAL BOOK RATE
Please make all necessary changes on the above label,
detach or copy, and return to the address in the upper
left-hand corner
If you do not wish to receive these reports CHECK HERE n,
detach, or copy this cover, and return to the address in the
upper left-hand corner
EPA-600/4-80-024
------- |