&EPA
United States
Environmental Protection
Agency
Environmental Sciences Research EP-i
Laboratory
Research Triangle Park NC 2771 1
Research and Development
Analysis of Houston
Aerosol Samples by
GC/MS Methods
Final Report
-------
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 PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-071
April 1980
ANALYSIS OF HOUSTON AEROSOL SAMPLES
BY GC/MS METHODS
Final Report
by
F. W. Karasek and M. L. Parsons
Arizona State University
Department of Chemistry
Tempe, Arizona 85281
Contract No. 68-02-2961
Project Officer
Kenneth J. Krost
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratories
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH 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 Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does men-
tion of trade names or commercial products constitute endorsement or recommen-
dation for use.
ii
-------
ABSTRACT
An analysis procedure developed to give a qualitative and quantitative
analysis for organic compounds adsorbed on aerosols collected by Hi-Vol fil-
ters was adapted and applied to a similar analysis of aerosols collected by
dichotomous filters. The procedure involves a 12 hour Soxhlet extraction with
raethanol, concentration of extract by a factor of 200 to 2000 by rotary eva-
poration under aspirator vacuum and analysis by GC and GC/MS. Analysis
was conducted for five dichotomous samples and two Hi-Vol samples collected
in the Houston study. Estimated concentration levels for dichotomous filters
of 0.1 to 20 ng/m3 were reported for the 27 organic compounds searched.
Compounds detected included carboxylic acid esters, phthalates, n-alkanes and
polycyclic aromatic hydrocarbons. Because of the uncertainties, low level
of compounds and need for multiple analyses involved, concentrates were not
reduced below 50 yl volumes. The quantities observed in the samples were
near the detection limits of the GC/MS-SIM analysis, and blanks of the instru-
mental system, solvent, procedure and filter elements were necessary to iden-
tify artifacts introduced. Filter and solvent blanks contained compounds
in the same or higher quantity ranges as did the filter samples, indicating
the teflon elements of the dichotomous filters were not cleaned sufficiently
prior to use. Results from analysis of one-quarter portions of the Hi-Vol
filter samples showed a very light loading of organic compounds, giving values
of 1 to 17 ng/m^ for the same types of compounds as found in the dichotomous
filters. Duplicate analyses of these samples showed a high degree of qualita-
tive reproducibility for the procedure. While these samples contained a
higher concentration of many organic compounds by a factor of 5 to 10 than the
dichotomous samples, they did not appear to be typical of Hi-Vols from, indus-
trialized areas.
iii
-------
CONTENTS
Abstract i i i
Figures vi
Tab! es vi i
Acknowl edgement vi i i
1. Introduction 1
2. Conclusions 4
3. Experimental Procedures 5
Sample History 5
Glassware Cleaning 7
Dichotomous Filter Extraction 7
Hi-Volume Filter Extraction 9
Analysis by Gas Chromatography 9
Analysis by Gas Chromatography/Mass Spectrometry 13
Solvent Study 16
4. Results and Discussion - Part 1 19
5. Results and Discussion - Part II 38
References 48
Appendix - GC/MS Data , 4g
-------
FIGURES
Number Page
1 GCPLOT of analysis of methanol solvents concentrated as
procedure blank 21
2 SIM plot of PG-2 analysis 24
3 GCPLOT of SIM data for dichotomous filters -.. 29
4 SIM plot of system blank for 3F-016 30
5 SIM plot of 3F-016 analysis 32
6 SIM plot of GC CALIBRATION MIXTURE analysis for dichoto-
mous s amples 34
7 GCPLOT of duplicate Hi-Vol filter samples and blanks 36
8 GCPLOT of dichotomous filter samples and solvent
blank 37
9 SIM plot of Hi-Vol sample 0052-1 39
10 SIM plot of Hi-Vol sample 0052-2 40
11 SIM plot of Hi-Vol sample 0086-1 41
12 SIM plot of Hi-Vol sample 0086-2 42
13 GCPLOT of SIM data of Hi-Vol samples 0052-1 and
0086-1 for components identified in Table 11 45
14 Plots from Dual-Mode analysis of 0052-1 46
15 Plots from Dual-Mode analysis of 0086-1 47
VI
-------
TABLES
Number Page
1 Sample History 6
2 Steps in Glassware Cleaning Procedure 8
3 Steps in Dichotomous Filter Extraction and Analysis
Procedure 10
4 Steps in Hi-Vol Filter Extraction and Analysis
Procedure 11
5 Comparisons of Extraction Procedure - ASU 12
6 GC Conditions - ASU (Dichotomous & Hi-Vol Filter
Analysis) 14
7 Conditions for Analysis (GC/MS) (UW) 15
8 Ions Monitored in SIM and GC/MS Analysis 17
9 SIM Data for Dichotomous Filters using PG solvent 22
10 Concentrations (ng/m3) of Organic Compounds in Dichotomous
Filter Extracts 27
11 SIM Quantitation of Hi-Vol Samples 0052-1 and 0086-1 44
Vll
-------
ACKNOWLEDGEMENTS
We wish to acknowledge the careful and extensive work done on this
very difficult analytical problem by W. D. Bowers of ASU and G. A. Eiceman
and R. E. Clement of DW.
VI11
-------
SECTION 1
INTRODUCTION
Sampling of aerosols by drawing large amounts of air through a fiber
glass filter element has been an established procedure since 1967 (1). The
apparatus used routinely for such sampling (Hi-Vol), however, gives no infor-
mation concerning particle size distribution. Such data are important for
health studies since only the very fine particles can penetrate deep into
the human respiratory system (2). Particle size information can be obtained
by sampling air with^the use of a cascade impaction sampling device (3,4).
This technique can be used to collect particulate matter in several size
ranges. Van Cauwenberghe and co-workers (3,4) have combined the use of a
6-stage cascade impactor for sampling with the technique of gas chromatogra-
phy-mass spectrometry (GC/MS) for the comparison of organic pollutant levels
on various particle sizes of aerosols. Organic compounds of low volatility
were found to be primarily associated with the small particle sizes of less
than 3 um diameters.
For most studies, extensive fractionatlon of particles into many size-
distribution categories is not necessary. Liu has reported that the volume
of aerosol in ambient air has a bimodal distribution with a relative minimum
at particle diameters of about 2 um (2). Also, Harris and Lippmann have
established that particles smaller than 3.5 um diameter can penetrate deeply
into the human respiratory system, while larger particles are trapped in the
upper respiratory system, while larger particles are trapped in the upper
respiratory passages (5). In addition, separation of aerosol into many sizes
may require lengthy sampling periods to obtain the quantity of sample needed
to perform organic and inorganic analyses.
For most studies, therefore, separation of aerosol into two particle
size ranges would seem to be sufficient to develop useful, information.
-------
Forley has described a single stage, variable slit impactor which is capable
of determining entire particle-size distributions, or fractionating particles
into two size classes for mass or chemical analysis (6). Stevens used a
dichotomous sampler to collect fine (<3.5 ym diameter) and coarse (> 3.5 urn)
particles for determination of atmospheric sulphates (7). More than 70 per-
cent of the sulfur was determined to be present in the fine particle frac-
tion for at least 90% of the samples analyzed. Use of the dichotomous sam-
pler is particularly suited for inorganic analysis by techniques such as
X-ray Fluorescence (XRF) (8,9,10). By using suitable filter elements,
aerosols may be analyzed by XRF directly after sampling, with no additional
sample treatment necessary. Use of the dichotomous sampler provides a
reasonable compromise between fractionation of atmospheric aerosols by size
for health studies and quantity of particulate collected for analytical pur-
poses.
Little work has been done to date to provide for the analysis of or-
ganic compounds adsorbed onto particulate matter collected by dichotomous
samplers. Because of the suspected carcinogenic effects of some organic com-
pounds, notably the polynuclear aromatic hydrocarbons (PAHs), it is important
to know the distribution of these compounds between respirable and non-
respirable fractions of airborne particluate matter.
In methods reported for organic analysis, adsorbed organic constituents
are solvent extracted from particulates on the filter surface, and analyzed
by GC and GC/MS. Methods of extraction vary greatly between laboratories,
some of which require a great deal of sample handling. Steps involved in
sample preparation may include extraction, sample clean-up, preseparation of
particular components of interest or separation into basic and acidic frac-
tions.
Recently, a simplified, effective analytical procedure has been deve-
loped to give a qualitative and quantitative analysis of organics compounds
adsorbed onto aerosols collected by Hi-Vol Filtration (11). This procedure
involves a 2 to 12 hour Soxhlet extraction with methanol, concentration by
a factor of 200 by rotary evaporation under aspirator vacuum, and analysis
by GC and GC/MS, with no further sample treatment. Data is processed and
presented via specialized computer programs in a format permitting easy
sample comparisons. Elimination of extra treatment steps greatly speeds
-------
analysis time and reduces chances for sample losses. For samples with very
low organic loadings, GC/MS analysis using the technique of Selected Ion
Monitoring (SIM) may be used to increase analytical sensitivity for compounds
of interest. The study reported here was undertaken to evaluate the appli-
cability of this rapid procedure, or some modification of it, for the analy-
sis of organic compounds associated with aerosols collected by dichotomous
filters.
-------
SECTION 2
CONCLUSIONS
These results pertain specifically to the procedure and instrumental
system used: namely, Soxhlet extraction with methanol, condensation, and GC,
%
GC/MS analysis with instrumentation involving special Aue-type packed columns
and a quadrupole mass spectrometer equipped with a membrane interface.
Although several of the common compounds expected to be associated with
these aerosol samples were detected, application of this analysis procedure
gave inadequate results for routine analysis of dichotomous filter samples
for organic content. This work showed the need for high solvent purity,
careful determination of solvent, column, instrumental system and filter
element blanks to separate artifacts introduced from compounds detected.
The need for a different approach to this analysis which minimizes the
artifact introduction and increases the extraction efficiency and instrumen-
tal sensitivity is seen. An approach utilizing an inert, sealed extraction-
condensation apparatus could permit a condensation increase by a factor of
5 and minimizes artifact introduction. If the extraction efficiency could
be increased by a factor of 2 using ultrasonic techniques and another factor
of 10 increase in sensitivity achieved by modification of the GC/MS operating
conditions, an overall increase in detectability of a factor of 100 would be
achievable. This higher detectability-to-background ratio would be able to
provide useful information on organic content.
-------
SECTION 3
EXPERIMENTAL PROCEDURES
SAMPLE HISTORY
Dichotomous Filters
Ten dichotomous filter samples were taken from sites in Houston, Texas
by the Radian Corporation, the agent responsible for collecting and shipping
filter samples. The ten samples, 3F-017, 3C-017; 3F-031, 3C-031; 3F-028,
3C-028; 3F-016, 3C-016; and 3F-032, 3C-032 and two blanks 3F-033, 3C-Q33
were received on October 18, 1978 and transferred from the shipping contain-
ers, where filters were cooled with dry ice, directly to a freezer (-19°C)
until analyzed. The filters were inspected visually and filters 3F-017 and
3F-031, which appeared most heavily loaded with particulate matter, were
extracted using the Soxhlet apparatus method (1,2) and pesticide grade (PG)
methanol (Fisher Scientific Company, Fairlawn, NJ). They were extracted
with PG methanol since reported levels of organic impurities in this solvent
were low.
It was later found that the "distilled-in-glass" BJ methanol (Burdick
and Jackson, Muskegan, MI) contained a lower level of impurities. The
remaining filters and filter blanks were then extracted by the Soxhlet appa-
ratus method using the BJ methanol solvent. Table 1 lists the history of the
samples analyzed.
High-Volume Filters
High-Volume filter [Hi-Vol] samples 0052 and 0086 were received
February 5, 1979 in sealed glass tubes packaged in dry ice. The tubes were
transferred directly to a freezer (-19°C) until analyzed. A one-quarter
portion of each filter was extracted using a soxhlet apparatus and the BJ
methanol. Following extraction, samples were analyzed by gas chromatography
-------
TABLE 1
HISTORY OF SAMPLES ANALYZED
0\
Sample
3F-017
3F-031
3F-028
3F-016
3F-032
3C-031
3F-033
Blank
3C-033
Blank
0052*
Hi-Vol
0086*
Hi-Vol
0085
Hi-Vol
Blank
Sample
Houston
Site 19
Houston
Site 21
Houston
Site 17
Houston
Site 19
Houston
Site 21
Houston
Site 21
-
-
unknown
Houston
Site 1
Date and Time
Sample Collected
9-28-78:0608 to
9-28-78:1808
10-9-78:0600 to
10-9-78:1800
10-6-78:0600 to
10-6-78:1852
9-27-78:1808 to
9-28-78:0608
10-10-78:0600 to
10-10-78:1800
10-9-78:0600 to
10-9-78:1800
-
-
unknown
10-3-78:0610 to
10-3-78:1800
~
Date Sample
Received ASU
10-18-78
10-18-78
10-18-78
10-18-78
10-18-78
10-18-78
10-18-78
10-18-78
2-5-79
2-5-79
5-8-79
Date
Extracted
ASU
11-23-78
11-21-78
1-28-79
1-28-79
1-29-79
1-29-79
1-31-79
1-31-79
2-20-79
2-20-79
5-13-79
Date of Analysis
GC(ASU) GC/MS(UW) GC/MS(UW)
SIMS
11-23-78
11-22-78
2-2^-79
2-3-79
1-30-79
1-30-79
2-2-79
2-1-79
2-21-79
2-24-79
5-14-79
11-28-78
11-28-78
2-22-79
2-23-79
2-23-79
2-23-79
2-23-79
2-23-79
3-27-79
3-27-79
~*
11-28-78
11-28-78
2-22-79
2-23-79
2-23-79
2-23-79
2-23-79
2-23-79
4-02-79
4-02-79
»
* These filter elements were not cleaned for organic analysis and will contain high and variable
organic background.
-------
at Arizona State University (ASU) and sent to the University of Waterloo (UW)
for analysis by gas chromatography/mass spectrometry and Selected Ion Moni-
toring (SIM). A Hi-Vol filter blank, sample 0085, supplied by the contrac-
tor at a later date was extracted and analyzed by the same procedure.
GLASSWARE CLEANING
All glassware was washed with an aqueous solution of Alconox detergent
(Alconox, Inc., New York, NY), thoroughly rinsed with analytical reagent (AR)
grade methanol (Fisher Scientific Co., Fairlawn, NJ) , rinsed with BJ grade
methanol and dried at 115°C for several hours. Soxhlet glassware was cooled,
assembled and operated with 100 mis of methanol for 12 hours as a final
cleaning procedure. Following this procedure the 100 mis of BJ methanol were
discarded and the glassware was used immediately. After drying, sample vials
were boiled in BJ methanol for 5 hours and used. The cleaning procedure is
outlined in Table 2.
DICHOTOMOUS FILTER EXTRACTION PROCEDURE
Dichotomous filter samples were removed from the freezer using Teflon
forceps and the filters separated from the plastic holders with a razor blade
and Teflon forceps. The Teflon filters were not in contact with any object
that was not cleaned and also methanol rinsed. Each filter was placed in a
80 mm high x 25 mm I.D. coarse fritted glass extraction thimble and the
thimble inserted into a 30 mm I.D. Pyrex Soxhlet tube (Fisher Scientific Co.,
Fairlawn, NJ). The entire extraction apparatus filled with 100 milliliters
of BJ methanol was assembled. Ground glass joint greases were not used in an
attempt to avoid introducing contaminants. Filters were continuously ex-
tracted for 12 hours in the Soxhlet apparatus.
After cooling to room temperature, the 100 milliliter extract was
condensed to approximately 5 milliliters using a rotary flash evaporator
(Buchler, Fort Lee, NJ). After quantitative transfer to a 10 milliliter
round bottom flask with a clean disposable pipet, the extract was further
reduced in volume to 0.5 milliliters. The extract was quantitatively trans-
ferred to a 1 milliliter reacti-vial (Altech Associates, Arlington Heights,
IL) sealed with screw cap and Teflon liners. A stream of ultra pure
(99.9%) nitrogen (Liquid Air, Phoenix, AZ), directed across the mouth of the
-------
TABLE 2
STEPS IN GLASSWARE CLEANING PROCEDURE
I. Detergent wash
a) Aqueous solution of Alconox detergent rinse
b) Distilled water rinse
c) AR grade methanol rinse
d) BJ methanol rinse
II. Dry in oven @ 115°C - two hours
SOXHLET APPARATUS
III. Soxhlet glassware assembled &
operated 12 hours as a pre-
rinse with 100 ml of BJ
methanol
SAMPLES VIALS
III. Boiled in BJ methanol for
5 hours
IV. [After prerinse completed]
Methanol discarded and
glassware used immediately
IV. Ready for use
-------
sample vial was used to reduce the volume of the extract to 50 microliters.
Table 3 outlines steps in the extraction procedure.
HIGH-VOLUME FILTER EXTRACTION PROCEDURE
The sealed glass tube containing the Hi-Vol filter was removed from the
freezer and allowed to reach room temperature. The tube was opened at one
end and the foil-wrapped filter removed. The foil was separated from the
filter using Teflon forceps and the filter cut in half. One half was re-
wrapped in the foil, inserted into the tube and the tube resealed. The tube
was returned to the freezer (-19°C). The remaining half was cut into halves
(a quarter of the original filter). Quarter filters were cut into 1 cm
squares, placed in the glass extraction thimble and continuously extracted
with 100 ml of distilled in glass methanol in the Soxhlet apparatus for 12
hours.
The methanol extract, after cooling, was condensed in volume to 5
milliliters using the rotary flash evaporator. The extract was quantitative-
ly transferred to a round bottom flask and further condensed to approximately
0.3 ml, and transferred to a sample reacti-vial and was adjusted to a final
volume of 0.5 milliliters.
Following gas chromatographic analysis, 0.2 milliliters of the final
concentrate were transferred to 0.1 railliliter reacti-vial, and reduced in
volume to 50 microliters using a stream of high purity nitrogen (Twin Cities
Welding, Kitchener, Ontario) directed over the mouth of the reacti-vial. The
more concentrated extract was used in GC/MS analysis. The remaining 0.3
milliliters was placed under refrigeration (2-5°C). Extraction procedures
for Hi-Vol filters are given in Table 4.
A comparison of volume and concentration factors between dichotomous
filter and Hi-Vol filter extraction procedures is listed in Table 5.
ANALYSIS BY GAS CHROMATOGRAPHY
All extracts were first analyzed on a 5830A Hewlett-Packard digital
gas chromatograph employing a 2 meter x 2mm I.D. glass column containing Aue
packing and equipped with a flame ionization detector. Aue packing consists
of an ultra-thin film of Carbowax 20M which is physically bonded to Chromo-
sorb W that has been exhaustively extracted with an aqueous hydrochloric acid
-------
TABLE 3
STEPS IN THE DICHOTOMOUS FILTER EXTRACTION
AND ANALYSIS PROCEDURE
I. Soxhlet extraction for 12 hours with 100 ml BJ methanol
II. Condensation of extract to 50 yl
a) 100 ml evaporated to * 5 ml
b) Transferred to 10 ml flask and evaporated to = 0.5 ml
c) Transferred to clean sample vial and evaporated to 50 yl with
stream of ultra-pure N2 gas purging across top of sample vial
III. GC analysis of final 50 yl extract
a) Column blank
b) Filter extract - 3 pi injection
c) Hydrocarbon standard
IV. GC/MS analysis (Peakfinder)
a) Column blank
b) Methanol system blank
c) Filter extract of 2-5 yl injections
d) Calibration mixture
V. GC/MS analysis (SIM)
a) Column blank
b) Methanol system blank
c) Filter extract of 2 to 5 yl
d) Calibration mixture
(a) steps I to III conducted at ASU; steps IV and V at UW
10
-------
TABLE 4
STEPS IN THE HIGH-VOLUME FILTER EXTRACTION
AND ANALYSIS PROCEDURE
I. Soxhlet extraction for 12 hours with 100 ml BJ methanol using 1/4
filter
II. Condensation of extract to 0.5 ml
a) 100 ml evaporated to a 5 ml
b) Transferred to 10 ml flask and evaporated to s 0.3 ml
c) Transferred to sample vial and brought to final volume of 0.5 ml
III. GC analysis of 0.5 ml extract
a) Column blank
b) Filter extract 3 yl injection
c) Hydrocarbon standard
IV. Further sample condensation
a) 0.2 ml evaporated to 50 yl
V. GC/MS analysis (peakfinder)
a) Column blank
b) Methanol system blank
c) Filter extract - 3 yl injection
d) Calibration mixture
VI. GC/MS - (SIM)
a) Column blank
b) Methanol system blank
c) Filter extract - 3 yl injection
d) Calibration mixture
(a) steps I to III conducted at ASU; steps IV to VI at UW
11
-------
TABLE 5
COMPARISON OF EXTRACTION PROCEDURES - ASU
Initial Extrac- Final Sample Concentration Volume of Air Volume of Air per
tion Solvent Volume (yL) Factor Sampled (M3) of Concentrated
Volume (mL) Extract
Soxhlet Extraction 100 50 2000 12 0.24
Dichotomous Filters
Hi-Vol Filters
(One-Fourth of Filter 100 500 200 500 1.0
Extracted)
-------
solution (12). The high efficiency of separation and the application of Aue
packing for analysis of complex environmental samples has been demonstrated
(11,13). In addition, Aue packing yields rapid and convenient analyses and
exhibits low bleed properties making it suitable for application in GC/MS
methods. Gas chromatographic conditions are listed in Table 6.
Column blanks were run before each sample to determine levels of arti-
facts introduced by the septum or column. Three microliters of the samples
were then analyzed. A hydrocarbon standard was also analyzed following each
sample and hydrocarbon retention times used to calculate retention indices
under temperature programmed conditions (11). Retention indices were used
in computer generated plots of peak area against retention index. These
plots were developed at the University of Waterloo and make possible rapid
and convenient comparisons of qualitative and quantitative distributions of
organic compounds in complex mixutres (14). The plots, referred to as
RIPLOTS, also can be generated using retention times and are then called
GCPLOTS. A similarly equipped 5830A gas chromatograph was used at the Univer-
sity of Waterloo where the ASU GC operating conditions were duplicated.
ANALYSIS BY GAS CHROMATOGRAPHY/MASS SPECTROMETRY
Sample extracts were analyzed with a Hewlett-Packard 5992A GC/MS which
was calculator controlled and equipped with single floppy disk, X-Y plotter,
membrane separator and 2 or 2.7 meter x 2mm I.D. glass column containing Aue
packing. Operating parameters for GC/MS system were optimized daily using
a perfluorotributylamine standard and software supplied by the manufacturer
(15). Extracts were analyzed using the GC/MS in both the scanning (Peakfind-
er, Dual-Mode) and Selected Ion Monitoring modes of operation. Opera-
ting conditions for the GC/MS analysis are given in Table 7.
Scanning Mode
During temperature programmed gas chromatographic analysis the mass
spectrometer scanned continuously from m/e 500 to m/e 40 at a rate of 330
atomic mass units/second. Software supplied by the manufacturer (Peakfinder)
stores mass spectra taken at the top of eluting peaks and stores the spectra
on disc. An eluting peak is detected when a threshold value for total ion
abundance is exceeded. A low threshold value was used for this study so
13
-------
TABLE 6
GC CONDITIONS - ASU
(DICHOTOMOUS & HI-VOL FILTER ANALYSIS)
Temperature 1
Time 1
Temperature Program Rate
Temperature 2
Time 2
Injection Port Temperature
Flame lonization Detector
Temperature
Attenuation
Helium Carrier Flow Rate
Sample Size
Column
Packing
80°C
4 minutes
4°C/minute
240°C
30 minutes
240°C
240°C
240°C
8
30 ml/minute
3 yl
2m x 2mm I.D.
Glass
Aue Packing, 80/100 mesh
14
-------
TABLE 7
CONDITIONS FOR ANALYSIS (GC/MS) (UW)
Temperature 1
Time 1
Rate
Temperature 2
Time 2
Injection Port Temperature
Helium Carrier Flow Rate
Sample Size
Column
Packing
Separator
lonization Source
Mass Spectrometer Type
Detector
EM Voltage
Solvent Time Out
GC Peak Detect Threshold
Mass Peak Detect Threshold
90°C
0 min
4°C/min
250°C
15 min
250°C
40 ml/min
2 to 5 pi
2 or 2.7m x 2mm I.D. Glass
Aue Packing 80/100 mesh
Silicone Membrane
Electron Impact
Quadrapole
Channeltron Electron
Multiplier (EM)
Autotune Value Used
2.5 to 3 minutes
200 to 1000
2000
15
-------
spectra were saved from even minor components. Mass scans taken at the low-
est abundance values between consecutive peaks were also saved as background.
Each background spectrum was subtracted from its corresponding peak spectrum
when the GC/MS run was terminated and this process provided spectra more
suitable for interpretation.
A modified version of Peakfinder, called Dual-Mode, developed at the
University of Waterloo, became available for use in March 1979 and was used
in the analysis of the Hi-Vol samples [16]. The Dual-Mode software, in addi-
tion to saving mass spectra as in Peakfinder, is also capable of storing time/
abundance data on disc for later generation of reconstructed gas chromato-
grams and as many as six mass chromatograms. Conditions for analysis were
unchanged from those used in Peakfinder.
Selected Ion Monitoring (SIM)
The SIM mode of analysis allows consecutive sampling of as many as six
ions which are pre-selected by the operator. During a GC run the abundance of
a particular ion is measured for a specified dwell time before sampling the
next ion. Dwell times used were 166 msec for all ions monitored by SIM. For
each sample analyzed two separate SIM runs were made giving a total of 12
ions which were monitored. By knowing retention times of compounds associated
with these ions, identifications can be made. The ions which were monitored
are given in Table 8, along with the compounds or compound classes character-
istic of these ions. Quantitation was provided by running a calibration mix-
ture which contained normal hydrocarbons, phthalates, alcohols and selected
polynuclear aromatic hydrocarbons (PAHs). Use of SIM allows detection of
even minor components since sensitivity is greatly increased by dwelling on
a particular ion for a long time period relative to the short dwell time of
Peakfinder or Dual Mode.
Solvent Study
Since the pesticide grade methanol concentrate contained impurities of
about the same quantities as the organic compounds extracted from the dicho-
tomous filters, this seriously interferred with determination of compounds
which were extracted from the dichotomous filter. It was known from work done
at the University of Waterloo that the Burdick and Jackson methanol was quite .
16
-------
TABLE 8
IONS MONITORED IN SIM AND GC/MS ANALYSIS
Ion in m/e Compound or Class
SIM
43.1 n-Hydrocarbon series
57.1 n-Hydrocarbon series
60.1 n-Aliphatic acids, n-Hydrocarbon series
74.1 n-Methyl ester series
85.1 n-Hydrocarbons series
149.1 Phthalate esters
154.1 Biphenyl, Acenaphthene
163.1 Dimethyl Phthalate
166.1 Fluorene
178.1 Anthracene
202.1 Fluoranthene, Pyrene
228.1 Triphenylene, Chrysene, Benzol-
anthracene
252.1 Benzo[a]pyrene, Benzo[k]fluoranthene
Dual Mode GC/MS
74.1 n-Methyl ester series
85.1 n-Hydrocarbon series
149.1 Phthalate esters
178.1 Anthracene
202.1 Fluoranthene, Pyrene
252.1 Benzo[a]pyrene, Benzo[k]fluoranthene
17
-------
low in impurities. Therefore, a study to compare the impurity levels of va-
rious types of methanol was conducted to determine which methanol would be
best to use for the extraction of the dichotompus filters. The three types
of methanol investigated were analytical reagent (AR), pesticide grade (PG),
and the "distilled in glass" (BJ).
The glassware was cleaned as in the general procedure. After the 12
hours of operation as a prerinse in BJ solvent, the used solvent was discard-
ed and 100 mis of fresh BJ solvent was added to the apparatus. After the 12
hour extraction period, the solvent blank, BJ-1, was condensed down to fifty
microliters as in the general procedure. This solvent blank was then analy-
zed by GC and GC/MS. A second solvent blank, BJ-2 was done by the same pro-
cedure with essentially the same results.
The AR and pesticide grade methanol were treated in the same manner and
procedure as the BJ solvent. These solvent blanks were designated AR-1 and
PG-1 representing the analytical reagent grade and pesticide grade methanol
respectively.
18
-------
SECTION 4
RESULTS AND DISCUSSION - PART I
GC/MS ANALYSIS DICHOTOMOUS FILTER ANALYSIS
Methanol extracts of dichotomous filters were found to contain 0.1 to
20 ng/m3 per organic compound, with one compound, dibutyl phthalate, being
present at 150 ng/m3. Near 0.1 ng/m3 the detection limits of SIM analysis
are approached and stringent steps were required to identify sources of con-
tamination and prevent loss of sample integrity during extraction, concentra-
tion, and analysis. Artifacts from the instrumentation are also possible at
these very high levels of sensitivity and were monitored with system blanks.
System blanks indicated that syringe and instrumental contamination was at or
near the noise level of the instrument.
The dichotomous filter analysis can be divided into three phases.
First, filters 3F-017 and 3F-031 were extracted with Pesticide Grade (PG)
methanol and analyzed by GC/MS (Peakfinder Mode) and SIM. The results of
these analyses indicated a solvent with far less impurities than present in
the PG grade would be necessary to obtain useful results. Second, different
grades of methanol, PG, AR and BJ were concentrated and analyzed by GC-FID.
The AB grade methanol concentrate, which contained the most impurities, was
analyzed by GC/MS. Finally, the remaining dichotomous filters were extracted
with the BJ solvent, which had the least impurities, and the extracts analy-
zed using SIM.
Initial results from the GC/MS (Peakfinder Mode) analysis of filter
3C-031, which was extracted with BJ methanol and which contained the highest
concentration of organic compounds, indicated that concentration levels for
individual components were near or below detection levels. Therefore, all
remaining dichotomous filters (extracted with BJ methanol) were analyzed by
SIM which is a more sensitive technique than regular GC/MS analysis.
19
-------
GC/MS (Peakfinder Mode)
Three filters 3F-017, 3F-031, and 3C-031 were analyzed using GC/MS
(Peakfinder Mode). The first two filters were extracted with PG methanol,
and showed 12 to 15 components. However, the PG methanol concentrate itself
contained many of the same components which, if they had originated from a
filter sample, would have shown an estimated equivalent concentration of 0.1
to 10 ng/m3 per component. The GC/MS results from these three filters were
inconclusive.
Three grades of methanol were then analyzed by GC to determine the con-
tribution of organic compounds to filter extracts following volume reduction
to 50 or 100 vl. Figure 1 is the GCPLOT from the chromatographic analysis of
different methanol solvent concentrates. The major components found in PG
solvent concentrate are methyl rayristate, methyl palmitate, methyl stearate,
diethyl phthalate and dioctyl phthalate. Since these components predominate
in 3F-017 and 3F-031, it is difficult to determine the components which are
contributed by the filter samples.
Figure 1 also clearly shows that BJ methanol contains the lowest levels
of impurities and it was chosen for extraction of 3C-031 and all other fil-
ters.
Analysis of 3C-031 showed no detectable components by GC/MS and was ana-
lyzed by SIM, as were all remaining filter extracts.
SIM Analysis
Results from the GC/MS analysis of 3F-017, 3F-031, and PG-2 were con-
firmed by SIM analysis and are given in Table 9. Results from SIM analysis
of PG-2 are shown in Figure 2. Because the SIM analyses were run on sample
concentrates after 4 months storage, and relative concentration factors bet-
ween solvent and sample are different and unknown, only a qualitative compa-
rison is possible.
Table 10 and Figure 3 summarize the results of the SIM analyses of the
dichotomous filter extracts with the BJ solvent. The results are given in
estimated concentration levels for 27 organic compounds found in extracts of
6 filters and a system blank. Identity was made by characteristic ions and
retention times. The procedure used was to run a SIM system blank for the
GS/MS system prior to a series of sample runs by injecting a 2 pi sample of
20
-------
1001
75
50-
25-
UJ
100-
75-
50-
25-
100-
75-
50-
25-
BJ-2
I I I I I I i
flR-1
I T I I I I
F.S. 17000
F.S. 17000
F.S. 17000
30
10 20
RETENTION TIME(MIN)
Figure 1. GCPLOT of analysis of methanol solvents concentrated
as procedure blank
140
21
-------
TABLE 9
SIM DATA* FOR DICHOTOMOUS FILTERS USING PG SOLVENT
Compound
Biphenyl
Acenaphthalene
n-Ci6H3it
n-Ci7H36
Dimethyl Phthalate
n-G18H38
Fluorene
Methyl Myristate
Diethyl Phthalate
Anthracene
n-CigH^o
Methyl Palmitate
n-Cao^a
n-CaiH^
Dibutyl Phthalate
Methyl Oleate
Methyl Stearate
m/e
154.1
154.1
85.1
85.1
163.1
85.1
166.1
74.1
149.1
178.1
85.1
74.1
85.1
85.1
149.1
74.1
74.1
Retention Time
in minutes
3.0
5.0
5.4
7.6
9.0 -
9.8
10.0
10.1
10.5
11.0
11.5
14.0
14.5
16.5
17.0
17.5
18.0
System Blank
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
PG-2
10
ND
26
35
ND
35
ND
270
ND
ND
26
2200
ND
ND
120
110
550
3F-017
1
ND
4
6
ND
6
ND
8
ND
2
6
150
ND
ND
50
8
48
3F-031
2
ND
6
9
ND
12
ND
9
ND
3
6
600
ND
ND
80
30
2
*Data are presented only for qualitative comparisons and are given in SIM peak heights. Concentra-
tion factors and sample age differ considerably between PG-2 and samples.
-------
TABLE 9 (Cont.)
SIM DATA FOR DICHOTOMOUS FILTERS USING PG SOLVENT
tsJ
Compound
n-C22H,6
Fluoranthene
n-CaaH^e
Pyrene
n-C21+H50
n~G25H52
n-C26H5[t
Dioctyl Phthalate
Benzo [k] f luor-
m/e
85.5
202.1
85.1
202.1
85.1
85.1
85.1
149.1
252.1
Retention Time
in minutes
18.5
19.5
20.5
20.6
22.5
24.5
26.0
28.0
35.0
System Blank
ND
ND
ND
ND
ND
ND
ND
12
ND
PG-2
ND
, 6
17
21
9
9
ND
170
ND
3F-017
ND
5
ND
3
ND
ND
ND
140
0.5
3F-031
ND
11
6
4
6
6
6
370
0.1
anthene
Benzo[a]pyrene
252.1
37.0
.3
ND
-------
I UN IBU.l. FULL SCFIIE = 23.9
inn ISM. I FULL 5O1LE = 22.S
N>
P-
I UN IE3.I FULL SCflLE -.53.2
I DM IM9.I FULL SCHLE = 398.S
IDN 7M.I FULL 5CHLE = 2217.9
IDM S7.I FULL SCflLE = 303.S
B.I Hi '12.1 IHB 'IC.I IB.H 2H.B 22.i '2MB '26J '21.1 ».! '32.8 '3HJ
Figure 2. SIM plot of PG-2 analysis
-------
N3
to
tr .
P6-2
JL
ION ZS2.I FULL 5CRLE = 10.0
ION 202.1 FULL SCflLE =38.7
ION 178.1 FULL 5CRLE -.44.3
IDN 60.1 FULL 5CRLE = 21.B
IDN H3.I FULL 5CHLE = 310.9
IDN BS.I FULL SCflLE = 173.3
Ti IT1 ts3 *WM '11.1 'iz.i 'IH.I 'IG.I 'IB.I 'a.i '22.1 'JH.I 'JBB 'a.i 'ai.i '32.1 '34.
I '3B.I ^Tl TO ngjl ->H.|
MINUTES >
Figure 2. (Cont.) SIM plot of PG-2 analysis
-------
pure BJ methanol. This gives the background SIM components contributed by
the syringe needle, septum puncture, GC column and other unknown elements.
The system blanks run prior to each series of analyses were almost identical
to the typical one listed in Table 10. Using the difference between these
values and those found in the filter extract analyses, along with the SIM
peak values from a known calibration mixture run, an estimate of quantities
of components given in Table 10 was made. Figures 4,5,6 containing the data
for sample 3F-016 illustrate this procedure. Other pertinent data are given
in the appendix.
Table 10 shows that most organic compounds detected are at low concen-
trations and that instrumental artifacts are also very low. Filter blanks
3C-033 and 3F-033 appear to contain methyl myristate, methyl palmitate, and
methyl stearate at net higher levels than after similar filters had been used
for sampling. Although no unequivocal explanation is possible for these data,
they suggest that the large volumes of air passing over the filters during
sampling may remove esters which are present as initial impurities. Each fil-
ter contains a range of n-hydrocarbons which are detected at quantities above
that found on the fine filter blank. Only n-C^iH^ and n-C22Hi»6 were signi-
ficantly larger on the blanks than on the filters. Filters 3F-028 and 3F-
032 had increased levels of dibutyl phthalate, while other filters had levels
at or near those of the filter blank. No dimethyl phthalate or diethyl
phthalate were seen in any sample or blank. Levels of dioctyl phthalate were
10 to 20 ng/m3, but similar levels were also present in the filter blanks.
Significantly, four common aromatic and polynuclear aromatic hydrocarbons
were not detected in filter blanks or filter samples. These compounds were
biphenyl, acenaphthene, fluorene, and anthracene. A trace of fluoranthene
might have been in 3F-033 and pyrene was detected in 3F-016, 3F-028, and
3C-031. Benzofajpyrene and benzofk]fluoranthene were also detected at levels
from 0.1 to 1 ng/m3 in 3F-016, 3F-028, and 3C-031.
In summary, a few of the dichotomous filters analyzed in this study con-
tained very low levels of some other common polynuclear aromatic hydrocarbons.
Low levels of n-alkanes were detected in every filter sample and a new loss
in methyl esters was observed between filter blanks and filter samples. The
quantities found, in the 0.1 to 10 ng/m3 range, were at the detection limits
of the procedure used. This indicates very low loading of organic compounds
26
-------
TABLE 10
CONCENTRATIONS (ng/m3)* OF ORGANIC COMPOUNDS
IN DICHOTOMOUS FILTER EXTRACTS
NJ
Compound
Biphenyl
Acenaphthene
n-C16H3lf
n-Ci7H36
Dimethyl
Phthalate
n-C18H38
Fluorene
Methyl
Myristate
Diethyl
Phthalate
Anthracene
n-CigH^Q
Methyl
Palmitate
n-C20H92
, Retention
m/e Times in
minutes
154.1
154.1
85.1
85.1
163.1
85.1
166.1
74.1
149.1
178.1
85.1
74.1
85.1
3.0
5.0
5.4
7.6
9.0
9.8
10.0
10.1
10.5
11.0
11.5
14.0
14.5
System
Blank
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
3C-033
Filter
Blank
ND
ND
ND
ND
ND
ND
ND
0.6
ND
ND
ND
4
<0.1
3F-033
Filter 3F-016
Blank
ND ND
ND ND
ND <0.1
ND <0.1
ND ND
ND <0.1
ND ND
0.5 <0.1
ND ND
ND ND
ND <0.1
5 0.3
ND <0.1
3F-028
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
<0.1
0.8
<0.1
3F-032
ND
ND
ND
ND
ND
ND
ND
<0.1
ND
ND
ND
0.6
0.6
3C-031
ND
ND
<0.1
<0.1
ND
0.1
ND
0.3
ND
ND
0.1
5
0.3
ND is not detected.
<0.1 indicates a range
of 0.01 to 0.09
*estimated precision = ±50%
-------
TABLE 10 (Corit.)
CONCENTRATIONS (ng/m3)* OF ORGANIC COMPOUNDS
IN DICHOTOMOUS FILTER EXTRACTS
N)
00
Compound
n-C21H^
Dibutyl (1)
phthalate
Methyl
oleate
Methyl
stearate
n-C22Hi*6
Fluoranthene
n-C33H^8
Pyrene
n-C21fH50
n-C25H52
n-C26H51+
Dioctyl
phthalate
Benzo[k]
Fluoranthene*
Benzo[a]pyrene
m/e
85.1
149.1
74.1
74.1
85.1
202.1
85.1
202.1
85.1
85.1
85.1
179.1
252.1
252.1
Retention
Times in
Minutes
16.5
17
17.5
18
18.5
19.5
20.5
20.6
22.5
24.5
26.0
28.0
35
37
System
Blank
ND
ND
ND
ND
ND
ND
ND
0.2
ND
ND
ND
10
ND
0.3
3C-033
Filter
Blank
2
2
0.4
2
0.2
ND
ND
0.4
ND
ND
ND
20
ND
<0.1
3F-d33
Filter 3F-016
Blank
0.5 <0.1
1 1
ND ND
2 0.3
<0.1 <0.1
<0.1 ND
ND <0.1
ND 1
ND <0.1
ND <0.1
ND <0.1
10 10
ND 0.2
ND 1
3F-028
0.2
4
ND
0.5
<0.1
ND
<0.1
0.4
<0.1
<0.1
-------
iiw
75-
50-
25-
100-
75-
so-
25-
100-
75-
so-
25'
100-
75-
50-
25-
100-
75-
50-
25-
100-
75-
50-
25-
100-
75-
50-
25-
3C-031 |
[ 3F-032 J
3F-028 I
i i i i 1,11 1 i i | r-
1 3F-016 |
L 3F-033J
1 |
3C-033 J
|STS BLflNKl
1 i -i i i i ii
< , 1, ,.,,, , ,
| .r^,-
1 F.S. 33 ]
1 F.S. 150 1
[ F.S. 23 1
1 F.S. 10 i
, ,,,,.,!.,
1 F.S. 10 I
FILTER BLANK
IF.S. 201
FILTER BLANK
I F.S. 10 |
-i ; r i i 1 1 i i
10 20
RETENTION TIMEQ1IN)
30
HO
Figure 3. GCPLOT of SIM data for dichotomous filters in full scale (F.S.)
of ng/m3
29
-------
3F-HIG/ 3F-H20/ 3F-032/ 3F-033/ 3C-03I/ 3C-H33
SYSTEM BLONK
r
|[JN l'ia.1 FULL 5CF1LE = '111.7
Ul
w<
z
cr
o
z
rj
ui
o:
I UN IG3.I FULL SCRI.E = H.H
i
ION 7M.I FULL SCflt.E =2.5
i
I UN fiH.I FULL SCflLE =1.3
I [IN M3.I FULL SCfllE = 0.3
I UN OS. I FULL 5CI1I.F. = 10.7
i.i B.« O.B
MINtlTFIB >
Figure 4. SIM plot of system blank for 3F-016
-------
3F-0IE/ 3F-B2B/ 3C-03I/ 3C-033
SYSTEM BLRNK
S.H
v
tc
IUN 2H2.2 FULL 5CI1LE = II.S
tu
ir
IHN IBB. I FULL 5G1LE = HI.3
IS'I.I FULL 5CHI.E = 21.0
BjiIB.II
I UN 57.1 FULL SCHLE = C7.I
HIHLITEB »
Figure A. (Cont.) SIM plot of system blank for 3F-016
-------
eo
Jl -
1
hi
III
IUN i'in. i nn i. tiCHt.r - '«).
IMI IL3.I FULL aCllli" = 2A\
IUH 7M.I FULL SOLE = £.
KULL SCIILE = H.B
I UN 43.1 FULL SCHLE = 12.'I
I [IN US. I FULL SOH.F. = 3.2
' ' ' ' '
JM.U ac.a 'JKU '>ui.u ''c.u '>i'iu
MINUTEH -)
iii.u la.u i>i.u IB. a lu u
LU.B ai.u
Figure 5. SIM plot of 3F-016 analysis
-------
1.1
i i
S;
n
at
inn iiiiii.i nil. lioii-r "
I UN IEiE.1 FULL SCfll.E = 23. B
I UN ISM. I FULL 5CIII.E = 23.1
I UK S7. FULL SCflLE = 102.
'.» 'yt.it 'ai.H 'aj.H 'iu.u ''til
HINUTELi ^
Figure 5. (Cont.) SIM plot of 3F-016 analysis
-------
bv CIILIhklll HIM IIIXIIIi:!
njy.3
1-
LI
in
IL
Hill 2ffi.2 FULL SCHl.E - 3UU'I.3
I UN I7B.I FULL SCHLE = Qy,U
OJ
IUN IGG.t FULL 5CHLE = 3l3ll.li
ION ISH.I FULL SCHLE = BIMU.I
IUN S7.I FULL SCHLE = 2722.1]
'liaj 'aa.a '3'i.n r'Mi.» ^aTii^uji ''IH.M
MINUTtS ^
TO MO TuTa MO MuTi '/iia 'ii.u '2'i.a
Figure 6. SIM plot of GC CALIBRATION MIXTURE analysis for Dichotomous Samples
-------
on the dichotomous filters. A direct comparison of the organic loading ex-
tracted from these filter samples with those from lightly loaded Hi-Vol fil-
ters can be seen by comparing the GCPLOTS of these two sample sets shown
in Figures 7 & 8. In general, the intensity of the GC peaks for the Hi-Vol
samples were greater by a factor of 10. The Hi-Vol blank filter element also
showed less organic compounds, indicating it was cleaned more rigorously than
were the dichotomous filter elements.
35
-------
ioo-
75-
50-
25-
fyfc
lOOn
75-
50-
25-
i rtr>
100-
75-
50-
UJ
g 25-
CO
d 100-
x 75-
50-
25-
i on**
100
75-
50-
25-
« nn-i
100
75-
50-
25-
EPfl
u , il.
r'fn
JEFfl
I i r
EPf)
|EPR
,
EPfl
i i I
|EPR
, V1,
I " T
052-1
,!,,' 1,
052-2
limlilUj
i f
085-1
085-2
082-1
. ..il
082-2
a.
.,i !i
1 ' '
i.iii ill
t ii"ip'
i ,, i
li|!nll|!|||i
|(
j.i.ji.i
itfiti Ii ill J
10
l
1,1; 1
n f u|
Ii
l!j
FT
l«J(
ullil
.ill
III!
ill
i
i
ilii
iih
ji
1
TT"
II
|
1 1
Ji
,|
r
1,1
F
II
|l!
2
1,
7
1,
r
|,
'
i
:,
0
1
1
J
ill
1 '
1
1
1
llji
1
,11!
I 1
|
1'.
1 1 -
I,,
' I
| T
in
i
i
,
r ' 1
tl
r* ' i
30
F.S. 150000 |
jni i| i .,, . r, , lil || i . i , 1
F.S. 150000 |
i r i r i i i i r tiii
F.S. 15000 |
FILTER BLANK
F.S. 15000 |
FILTER BLANK
F.S. 150000 |
i, . i, .... .11. i i |
F.S. 150000 |
i -i1 i-l , 1 i, ill
I I i i'i i i r i i " r i i i "
40 SO
RETENTION TIME(MIN)
Figure 7. GCPLOT of duplicate Hi-Vol filter samples and blanks
36
-------
too-
75-
so-
25-
100-
75-
50
25
100-
75
50-
25
SOLVENT BLANK
100
75
50-
25
100
75-
so-
25-
100-
75-
so-
25
100-
75-
so-
25-
BJ-1
3F-016
I 3F-028
3F-033
3C-O33
"i i iii
F.S. ISOOO
...I i
-it-i i i
3C-031 I
F.S. ISOOO
I F.S. ISOODl
I F.S. 15003 |
FILTER BLANK
I F.S. 15000 |
FILTER BLANK
IF.S. 15000
I F.S. 15000 |
iii i i i i i i ri
10 20 30
RETENTION TIME CHIN}
SO
Figure 8. GCPLOT of dichotomous filter samples and solvent
blank
37
-------
RESULTS AND DISCUSSION - PART II
GC/MS ANALYSIS - HIGH VOLUME FILTER ANALYSIS
Visual inspection of the two Hi-Vol filter samples analyzed, 0052 and
0086, indicated that both filters contained a light particulate loading.
One-half of each filter was chosen for analysis. Each half chosen was fur-
ther divided so that four filter portions were available for separate extrac-
tion and analysis, each portion comprising one-quarter of the original filter.
The two quarter-filters from filter 0052 are designated 0052-1 and 0052-2,
and the two quarter-filters from 0086 are called 0086-1 and 0086-2. This
analysis scheme allows evaluation of the reproducibility of the procedures
employed. Regular GC/MS analysis was difficult due to the light particulate
loadings, therefore GC/MS SIM analysis was employed.
GC/MS SIM
SIM analysis for both 0052 and 0086 produced similar results. Compari-
sons between plots of SIM data for duplicate samples (0052-1 with 0052-2 and
0086-1 with 0086-2) show good reproducibility of the analytical method
(Figures 9-12).Retention times and peak shapes for major and minor components
in duplicate filter extracts are almost identical qualitatively.
A comparison of full-scale values between Figures 11 and 12 shows quan-
titative reproducibility within a factor of two for most compounds. This
holds for components present at levels approaching the sensitivity limits for
GC/MS-SIM analysis. The good reproducibility also demonstrates the even
distribution of organic compounds on corresponding quarter-filters sections.
Sample 0052-2 reached dryness after GC analysis and was reconstituted with 50
microliters of methanol before GC/MS analysis. No qualitative changes are
apparent, however some quantitative changes were detected, most notably a
decrease in methyl ester concentration.
Data shown previously in Figure 7 compares results of GC analysis of
duplicate sample sets from filters 0052, 0086 and a Hi-Vol filter blank
38
-------
EPH BBSZ-I
ION 252.1 FULL SOLE = IB2.0
IOM22Q.I FULL 5CHLE = 202.E
IDN202.I FULL SOLE = 272.1
ION I7B.I FULLSCflLE = 33.0
ION ISH.I FULL 5CBLE = H3.I
IDN IMS
r
.1 FULL 5CRLE = B07.7
IOHBS.I FULL 5CBLE = S23.3
IOH7M.I FULL5CRLE = 29IS.2
a 22 « 'zs 'a a '32 '»
Figure 9. SIM plot of Hi-Vol sample 0052-1
-------
EPH 00S2-2
E - 37.1
o
5CRLE = 3HJ
DM IH9.I FULL SOLE = IH3.B
ION 7H.I FULL 5CI1LE = HB3.7
T2 'H 1 ^ Tl ^12 IM "IE ^
FIGURE 10. SIM plot of Hi-Vol sample 0052-2
-------
CPR H0BE-I
ION 2B2.I FULL SOLE = 72.7
^JONITB.1 FULL SCflLE = IS.H
I DM ISH.I FULL5CHLE = 2G.S
I DM TO. I FULL 50U.E = SH3.2
A
ION 74.1 FULL5CRLE = ISS.H
' '' ' ' ' '
I 2 1 G
II 12 II IS IB a 22
2E a'aT2^) x "a
MINUTES ^
Figure 11, SIM plot of Hi-Vol sample, 0086-1
-------
EPH 00BB-2
= 74.E
ON 22Q. FULL 5CRLE = S2.
NJ
5CHLE = 22.9
ISH.I FULL 5CRLE = 39.B
JL
JDN 149.1 FULL SOLE = 49E.7
IDN7H.I FULL 5CHLE = 20H.2
ra
MINUTES >
Figure 12. SIM plot of Hi-Vol sample 0086-2
-------
(sample 0085). Plots are at full scale sensitivity values which are compa-
rable to relative condensation factors. The filter blank shows very low
levels of only a few contaminants. These data further support the qualita-
tive and quantitative reproducibility of the analytical procedure.
Table 11 and Figure 13 show the SIM data from analysis of 0052-1 and
0082-1. Quantitation is based on comparison of sample component peak
heights to peak heights of standards. Calculations include corrections for
injection volumes, sample volume changes, volume of air sampled and relative
response factors. Independent quantitation by GC-FID confirmed the estimated
concentration range for most compounds detected to be from 1 to 17 ng/m3.
Most compounds detected were found in all the filter extracts and at similar
concentrations. Only dioctyl phthalate, methyl palmitate and methyl stearate
were found to differ significantly between the samples. Some methyl esters,
other phthalates and n-hydrocarbons were not detected in the SIM analysis.
GC/MS Analysis - Generation of Mass Chromatograms
Software [Dual-Mode] has been developed at the University of Waterloo
which allows storage of the total ion current and up to 6 mass chromatograms
in a GC/MS run, in addition to the mass spectra taken at the top of eluting
GC peaks (16). Results of the analysis of 0052-1 and 0086-1 using the Dual-
Mode program, are shown in Figures 15 and 16. These results further confirm
the major differences between these two samples to be in the amounts of
methyl palmitate and methyl stearate. Mass spectra contained significant con-
tributions from unresolved and unidentified organic compounds, even after
subtraction of background spectra. Much of this mass of compounds may con-
sist of unsaturated and branched hydrocarbons. Components of low abundance
detected on a high background of unresolved hydrocarbons are not easily iden-
tified. The low abundances obtained by running column blanks confirmed the
presence of high background material in the samples. Mass chromatograms
which were generated support results obtained from the SIM analyses.
43
-------
TABLE 11
SIM QUANTITATION OF HI-VOL
SAMPLES 0052-1 and 0086-1
Compound
Retention Time
in minutes
0052-1
ng/m3
0086-1
PAHs
Benzo [ a ] pyr ene
Benzo [k] f luoranthene (1)
Benzanthracene (1)
Pyr ene (2)
Fluoranthene
Anthracene
Fluorene
Biphenyl
PHTHALATE S
Dioctyl Phthalate
Dibutyl Phthalate
Diethyl Phthalate
HYDROCARBONS
Methyl Palmitate (3)
Methyl Stearate (3)
(1) Based on Benzo [a] pyr ene.
(2) Based on Fluoranthene.
* Retention Times are n-Cj
(3) Based on response of n-C
35
34
28
20.5
19.5
10.5
7.5
3.5
29.3
18.3
10.5
*
14.5
18.5
2
3
3
0.6
0.6
17
0.6
15
8
Response Factor = 1
Response Factor = 1
3^0:10.2 minutes :n-C2oHi| 2 :14.5 minutes
'21^1*2 relative response factors are 15.
1
1
1
0.3
0.3
11
4
0.3
1
for Methyl Palmitate and 1.2
for Methyl Stearate.
-------
75-
50-
UJ
rf 25-
en
1
*» i nrt-«
J 100-
i.
^ 75-
50-
25-
0052-1
i i i I I
0086-1
ill i i I i
.
ill
iii
i i i r i ii
i 1111 ii
i
F.S. 20 |
1
III i i i i
F.S. 20 |
iii i i i i
10 20
RETENTION TIME(MIN)
30
Figure 13. GCPLOT of SIM data of Hi-Vol samples 0052-1 and 0086-1
for components identified in Table 11
45
-------
CE
ca
a:
EPB BK2-I
iii|iiii|iiri|r
0 s
"r~( i i i i | ' ' ' i I ' ' ' IT"1" ' ' i-r~r-r~<
10 IS 20 2S 30 3S H0
1DN S2.2
KT1
mm.\
mm
in m.i
Km
in ra.i
Rl?fi
IWBS.I
R1Z
HKTI.!
fSEIH
ric
fSII
MINUTES
Figure 14. Plots from Dual-Mode analysis of 0052-1
-------
[PR 00BFH
JL
i
UJ
V
z
cr
to
cr
0
1 i r
2S
IIII 52.2
F57S
R7I
IDN II. I
PS'ID
IDN I'B.I
Rail!
IMBS.I
F5I37
IDN?). I
FSB79
10 15 20 25 30 35 H0 MINUTE5
Figure 15. Plots from Dual-Mode analysis of 0082-1
-------
REFERENCES
1. G.A. Jutze and K.E. Foster, J.A.P.A., IT., 17 (1967)
2. K.T. Whitby, R.B. Husar and B.Y.H. Liu, J.Colloid Interface Sci., .39,
177 (1972)
3. L. VanVaeck and K. Van Cauwenberghe, Atmos. Environ., 12, 2229 (1978)
4. L. VanVaeck, G. Broddin, W. Cautreels and K. Van Cauwenberghe, Sci. Total
Environ., 1JL, 41 (1979).
5. M. Lippmann and W.B. Harris, Health Phys., j}, 155 (1962)
6. L.J. Forwy, Rev. Sci. Instrum., 47, 1264 (1976)
7. R.K. Stevens, T.G. Dzubay, G. Russwurm and D. Rickel, Atmos. Environ.,
17., 55 (1978)
8. T.G. Dzubay, R.K. Stevens and C.M. Peterson in "X-ray Fluorescence Analy-
sis of Environmental Samples," p. 95-105, T.A. Dzubay, Ed., Ann Arbor
Science Pub. Inc., Mich., 1977
9. A.R. McFarlard, C.A. Ortiz and R.W. Bertch, Jr., Environ. Sci. Technol.,
^2, 679 (1978)
10. T.B. Dzubay and R.K. Stevens, Environ. Sci. Technol, 9^, 663 (1975)
11. F.W. Karasek, D.W. Denney, K.W. Chan, and R.E. Clement, Anal. Chem., 50,
82 (1978)
12. W.A. Aue and C.R. Hastings, J. of Chromatogr., 89. 369 (1974)
13. H.H. Hill, Jr., K.W. Chan, and F.W. Karasek, J. of Chromatogr., 131, 245
(1977)
14. R.E. Clement, "Application of Computer Techniques to the Collection and
Analysis of Analytical Data," M.Sc. Thesis, University of Waterloo,
Waterloo, Ontario, August 1976
15. F.W. Karasek, Research/Development, 27, 42 (1976)
16. L.C. Dickson, "Software Improvements for GC/MS/Calculator Used In Trace
Organic Analysis of Environmental Samples," Department of Chemistry Re-
port, University of Waterloo, Waterloo, Ontario, April 1979
48
-------
APPENDIX
GC/MS DATA
Number page
A-l SIM plot of GC/MS column blank 50
A-2 SIM plot of system blank for dichotomous filters 51
A-3 SIM plot of system blank for dichotomous filters 52
A-3 (Cont.) SIM plot of system blank for dichotomous filters. 53
A-4 Nine ion SIM plot of GC CALIBRATION MIXTURE for
dichotomous filters 54
A-5 Nine ion SIM plot of instrument blank for
dichotomous filters 55
A-6 Six ion SIM plot HYDROCARBON STANDARD MIXTURE 56
A-6 (Cont.) Six ion SIM plot HYDROCARBON STANDARD MIXTURE 57
A-7 Six ion plot of GC CALIBRATION STANDARD MIXTURE 58
A-8 Six ion plot of METHYL ESTER STANDARD MIXTURE 59
A-9 Total ion chromatogram and mass chromatograms
of column blank GC/MS 60
A-10 Total ion chromatogram and mass chromatograms of Hi-Vol
sample 0052-1 GC/MS analysis. 61
A-ll Total ion chromatogram and mass chromatograms of Hi-Vol
sample 0086-1 GC/MS analysis 62
49
-------
1* -
UI
n:
z - -
tn
0.
GC/HQ - COLUMN BLTINK
IDNHSH.I
inn ?B2.Z FULL SCHLE = B.I
IDN I EG. I FULL 5CF1LE = in 7
ION IS'I.I FULL SCHI.F. = 17.D
HIM S7 I FULL 5CMI.E - 'IU.I
T.I*TeTT
'III.B *TTii fTTe ' in.B 'in.n '?O *?*.ii
'?n.n
HINIJITFi »
A-l. SIM plot of GC/MS column blank
-------
i +
I
3F-HIB/ 3F-02H, 3F-032/ 3F-H33/ 3C-B3I/ 3C-H33
SYSTEM BLHNIC
I
IBM I'13.1 FULL SCflLE = 'HI.7
I UN IB3.I FULL SCflLE = 2.0
I
IDM 7H.I FULL BCPLE = 2.E
I UN CB.I FULL SCflLE =1.9
I UN H3.I FULL SCHLE = 3.3
jv.
11IN OS. I FULL 5CPIE = IB.7
ITS TTi ^Ti TTH8.B II.B u.H I'l.i IB.B ill.
fl.l
MiNiiTf:!; ^
A-2. SIM plot of system blank for dichotomous filters
-------
r- -
ui
t-o
3F-0IG/ 3F-02B/ 3C-B3I/ 3C-033
SYSTEM BLFINK
IDN 2S2.I FUIY S«1LE = S.M
V
IUN ZBH.H FULL 5CDLE = II.S
IUN IGB.I FULL 5CIILE = 21.3
I UN ISM. I FULL SCflLE = 21.0
IDN S7.I FULL SCflLE - G7.I
Fi IT5 ^Ti *e3 *IJi 'n.i 'I?.H 'i'i.« ' IB n 'm.a '?n.n 'lit 'yi.n 'm.ii 'am f.in.
i aa.ii .TI.II
n.ii 'JII.B ''in.ii '-K.II '
MINUTES ->
A-3. SIM plot of system blank for dichotomous filters
-------
3F-B32/ 3F-033
SYSTEM BLriNK
iDM 252. i
LJ
-A-
IDN 2H2.2 FULL 5OTLE =0.3
Ln
Co
IUN IEG.I FULL 5CRLE - 21.S
IUN ISM. I FULL SO1LE = 22.'I
OtTiT?
'la.n 'U.H 'ri.« 'id. a 'in. ii 'JII.B
i»(.ii W.B ain
iUN 57.1 FULL sanx = i;n.3
'M,B 'TI.B 'an.ii '.ni.ii '-11.11 'MZ.II
ri ->
A-3. (Cont.) SIM plot of system blank for dichototnous filters.
-------
EC CRLIHRRTIDN MIXTURE
IUN2S2.I FULL 90ILE = 229.G
t-
LJ
-------
INSTRUMENT BLHNK
J
to
1C
ION 2192.2 FULL SCBLE = B.l
V
ION IEE.I FULL SCHLE = 18.7
ION ISH.I FULL SCflLE = 17.0
ION 57.1 FULL SCflLE = H3.I
1.1 '». '1.1 't.i 'BI 'ti.i 'K.I 'IH.I 'IB.I 'IBI 'a.i 'n.i '?... 'g.i 'an 'M.I 'a.i '?i.i 'as.i 'a.. 'HI.I
n mures «
A-5. Nine ion SIM plot of instrument blank for dichotomous filters
-------
HYDRtKFIRBQN 5THNDRRD
I EG.I FULL 5CBLE = 44.1
8 I
EC
LLLUluju
ION 154.I FULL SCflLE = 614.E
Jl IL
to
ir
ION 149.1 FULL 5CHLE = 81.7
IDN IB3.I FULL 5CRLE =
IDN 74.1 FULL SCflLE =
IDN S7.I FULL 5CHLE = II73S.3
n 11 u
i
B.I J.B H.I E.I B.I IB.B 12.B IH.B IE.I IB.I a.l 22.1 2H.R 26.B 2B.I 3D.I 32.1 34.B 36.fl 3D.B ' I K2.I MM B
HINUTE5 *
A-6. Six ion SIM plot HYDROCARBON STANDARD MIXTURE
-------
VI
z
IE
Ci
Ul
HYDRdCRRBQN 5TRNDRD
JL
. * * » ~.
JiLUlLJL.
ION 252.1 FULL 5CRLE = III.H
IDN202.I FULL SCBLE = 2B.H
ION I7B.I FULL 5CRLE = IS.G
IOME0.I FULL 5CRLE = 7.1
JjJJJAliLLU
IDNH3.I FULL 5CHLE = SI4I.B
u
fl 'll.i '
IDNBS.I FULL 5CHLE = B7H7.7
I.I 2.1 S.I B.I BB ll.l 12.1 14.1 IB.I IB.I a.l 22.1 2H.I 26.1 2B.I 3B I '32.1 3M.I 3G.I *3B.I '.! H2.I
MINUTES ^
A-6. (Cont.) Six ion SIM plot HYDROCARBON STANDARD MIXTURE
-------
G.C. CHLIBRHTIDN 5TRNDHD
ui
z
a:
z -
to
IE
Ln
00
ILL BCHLE = 2S02.7
I ON 2H2.1 FULL 5CHLE = 9261.0
I ON 178.1 FULL BCHLE = 273.H
IDM EH.I FULL SCRLE = 10.0
IDNH3.I FULL 5CHLE = 29S2.7
1
IDNBS.I FULL 5CHLE = 4ISG.S
'H.I Tn rO 'IIB 'izTi 'IH.I IB.I^'IBJ 'ai.i '22.1 ZM JE.I a.n 'aa.i '32.1
A,
E.I 'a.i 'HB.I '12.1
MINUTES »
A-7. Six ion plot of GC CALIBRATION STANDARD MIXTURE
-------
Ul
8 .
z
tc
z
to
IE
11
METHYL-ESTER 5TRNDRRD
IW S7.I FULL ?.-?'. I' ' :r-'.n
L
ill
IDNBS.I FULL 5CRLE = 2BB.H
l- H l
ION 252.1 FULL 5CRLE = IH.B
IOM2B2.I FULL SCflLE = 27.2
IDN IH3.I PULL 5CRLE = 2E0.S
ION 7H. 1 FULL SCflLET « 3887.6
1.1 lii'"N.i"l|iMr 'a.i 'iii*n.i 'IH.I Vi
>!>» '.I..' he. l.o.n'-^
*BT-
M mints >
A-8. Six ion plot of METHYL ESTER STANDARD MIXTURE
-------
DC
CO
en
EPfl COLUMN BLRNK
JvJujA,*^^
^^^^^^^^
10
IS
20
30
3S
H0
RUN
ION 2S2.2
F5 B
RUN I00H
ION 202.1
F5 B
RUN I00H
ION I7B.I
F5 II
RUN I00H
ION 1MB.I
F5 27
RUN I00H
ION BE.I
F5 B4
RUN I00M
ION 7H.I
F5 32
RUN I00H
TIC
F5 3SB3
MINUTES
A-9. Total ion chromatogram and mass chromatograms of column blank GC/MS
-------
flBUNDRNCEi
EPfl 0052-1 j\rt
-_..,.. j , <« j ,, -._ uuj.J ^"-
____JL^
___JV_ _
.. L
V
. . (I . 1
. l r -i i i i^, , , i i i i | i i i i | i i i i | i i i i | i i i i | i i i , | i i i
0 S 10 IS 20 25 30 3S H0
RUN I00S
ION 2S2.2
-F5 79
RUN 1005
ION 202.1
n... ,. f. 1 I'l
RUN 1005
ION I7B.I
** F5.E3
RUN 1005
IDN.IH9.I
F5 I3H5
RUN 1005
. ION.BSU
F5 H02
RUN 1005
ION 7H.I
RUN 1005
TIC
V* ???R9
HINUTE5
A-10. Total ion chromatogram and mass chromatograms of Hi-Vol sample 0052-1
GC/MS analysis
-------
to
cn
EPH 0BBE-I
1
1
Ujw^^
IB
T"1
IS
fT"1
213
' ' i '
30
« i
35
MB
RUN IBBE
IDN 252.2
F5 7S
RUN I00E
ION 202.1
F5 71.
RUN I00E
IDN I7B.I
F5 M0
RUN IBBG
IDN IH9.I
F5 2BIB
RUN IBBG
IDN 05.1
F5 137
RUN IBBG
ION 7H.I
F5 E79
, RUN IBBG
TIC
F5 9GBG
HINUTC5
A-ll. Total ion chromatogram and mass chromatograms of Hi-Vol sample
0086-1 GC/MS analysis
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
FPA-60Q/2-80-Q71
3. RECIPIENT'S ACCESSION-NO.
TITLE AND SUBTITLE
ANALYSIS OF HOUSTON AEROSOL SAMPLES BY GC/MS METHODS
Final Report
5. REPORT DATE
ril 1980
Apr
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
F. W. Karesek and M. L. Parsons
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Arizona State University
Department of Chemistry
Tempe, Arizona 85281
10. PROGRAM ELEMENT NO.
1AA603 AH-16(FY-78)
11. CONTRACT/GRANT NO.
68-02-2961
2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences and Research Laboratory- RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final 9/78 - 6/79
14. SPONSORING AGENCY CODE
EPA/600/09
5. SUPPLEMENTARY NOTES
16. ABSTRACT
An analysis procedure developed to give a qualitative and quantitative analysis
for organic compounds adsorbed on aerosols collected by Hi-Vol filters was adapted
and applied to a similar analysis of aerosols collected by dichotomous filters.
Analysis was conducted for five dichotomous samples and two Hi-Vol samples collected
in the Houston study. Estimated concentration levels for dichotomous filters of 0.1
to 20 ng/m3 were reported for the 27 organic compounds searched. Compounds detected
included carboxylic acid esters, phthalates, n-alkanes and polycyclic aromatic
hydrocarbons. The quantities observed in the samples were near the detection limits
of the GC/MS-SIM analysis, and blanks of the instrumental system, solvent, procedure
and filter elements were necessary to identify artifacts introduced. Filter and
solvent blanks contained compounds in the same or higher quantity ranges as did the
filter samples, indicating the teflon elements of the dichotomous filters were not
cleaned sufficiently prior to use.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
*Air pollution
*Aerosols
*0rganic compounds
*Quantitative analysis
Qualitative analysis
*Gas chromatography
*Mass spectrometry
13B
07D
07C
14B
DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
SECURITY CLASS ri
UNCLASSIFIED
71
20 SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
orm 2220-1 (9-73)
63
------- |