EPA-600/4-77-004
January 1977
Environmental Monitoring Series
ENVIRONMENTAL APPLICATIONS OF ADVANCED
INSTRUMENTAL ANALYSES:
Assistance Projects, FY 75
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia 30601
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
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.
-------
EPA-600/4-77-004
January 1977
ENVIRONMENTAL APPLICATIONS OF ADVANCED INSTRUMENTAL
ANALYSES: ASSISTANCE PROJECTS, FY75
by
Ann L. Alford
Analytical Chemistry Branch
Environmental Research Laboratory
Athens, Georgia 30601
Project No. 16020 GHZ
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30601
-------
DISCLAIMER
This report has been reviewed by the Environmental Research 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.
-------
FOREWORD
Nearly every phase of environmental protection depends on a
capability to identify and measure chemical pollutants in the
environment. As part of this Laboratory's research on the
occurrence, movement, transformation, impact, and control of
specific environmental contaminants, the Analytical Chemistry
Branch develops techniques for identifying and measuring chemi-
cal pollutants in water and soil.
This report is the fifth in a series of annual summaries of
analytical chemistry support requested by other organizations.
In most cases, these analyses contributed to the solution or
better understanding of a pollution incident. This report will
acquaint other researchers and administrators with the type of
information that can be obtained with current analytical
techniques.
David W. Duttweiler
Director
Environmental Research Laboratory
Athens, Georgia
ill
-------
ABSTRACT
The Analytical Chemistry Branch of the Athens Environmental
Research Laboratory identified arid measured aquatic pollutants
under eight projects in response to requests for assistance from
other EPA organizations and other government agencies. In most
cases these analyses helped us to solve, or at least to
understand more clearly, the related pollution incident, and in
some cases the analyses provided evidence for enforcement of
regulatory legislation. Under an additional project, analytical
consultations were held as requested by various organizations
concerned with pollution incidents. The report describes those
projects.
This report was submitted in fulfillment of Project 16020 GHZ by
the Environmental Research Laboratory, Athens, Georgia.
Projects discussed we,re completed during FY 1975.
-------
CONTENTS
Page
Foreword iii
Abstract iv
Tables vi
Abbr evi at ions vi i
Acknowledgments viii
I Introduction 1
II Recommendations 2
III Discussion 3
Chemical Elements in Great Lakes Fish
Samples 3
Multielement Analysis of Poultry Farm Water
Samples 5
Chemical Elements in Drinking Water
Samples 6
Mercury in Seawater 6
Chemical Elements in Plant and Soil Samples
for Biomonitoring of Power Plant Emissions 7
Chemical Elements in Various Environmental
Samples 10
Organic Contaminants in New Orleans Drink-
ing Water 11
Organic Contaminants in Drinking Water
From 10 United States Cities 23
Dissemination of Analytical Information 31
IV References 35
-------
TABLES
Number Page
1 Chemical Elements Detected by SSMS in Great
Lakes Fish Samples 4
2 Mercury in Seawater Samples Analyzed by
Neutron Activation 7
3 Neutron Activation Data for Mercury and
Cadmium in Standard Samples 9
U Organic Compounds Identified in New Orleans
Drinking Water 16-19
5 Concentration Classification of Organic
Components Detected in New Orleans Drinking
Water 15
6 Comparison of Number of Identified Compounds
in Different Sample Types 22
7 Comparison of Recoveries from Filters #1
and #2 of the 70-Year Sample 24
8 Organic Compounds Identified in Drinking
Water of Ten U.S. Cities 28-30
9 Classification of Identified Compounds
According to Type of Source Water 32
-------
AAS
ACB
EPA
GC
GC-FTIR
GC-MS
INAA
LMRF
MS
NFIC
NOAA
PCB
SSMS
LIST OF ABBREVIATIONS
atomic absorption spectrophotometry
Analytical Chemistry Branch
U.S. Environmental Protection Agency
gas chromatography
combined gas chromatography and Fourier
transform infrared spectrophotometry
combined gas chromatography and mass
spectrometry
instrumental neutron activation analysis
Lower Mississippi River Facility
mass spectrometry
National Field Investigations Center of the
Environmental Protection Agency
National Oceanic and Atmospheric Administration
polychlorinated biphenyl
spark source mass spectrometry
VI1
-------
ACKNOWLEDGMENTS
The following Athens Environmental Research Laboratory personnel
were the principal investigators of the projects discussed in
Section III: C. E. Taylor, R. V. Moore, A. W. Garrison, L. H.
Keith, and the staff of the Analytical Chemistry Branch.
The principal investigators gratefully acknowledge the
assistance of A. L. Alford, F. R. Allen, L. V. Azarraga, M. H.
Carter, T. L. Floyd, S. Hensley, J. D. Pope, O. W. Propheter, W.
J. Taylor, A. D. Thruston, Jr., R. G. Webb, and G. D. Yager.
vxii
-------
SECTION I
INTRODUCTION
The Analytical Chemistry Branch (ACB) at the Athens
Environmental Research Laboratory (Athens ERL) develops techni-
ques for identifying and quantifying chemical pollutants and
identifies specific compounds associated with various pollution
sources. The ACB has analyzed many samples related to a variety
of specific pollution problems. Analytical results were
reported only to the persons who requested the analyses and
therefore had limited distribution. Earlier problems studied
by the ACB have been summarized in annual reports 1~1+ to acquaint
other researchers and administrators with the type of informa-
tion that can be obtained and to inform environmental chemists
of technique applications and developments. This report
summarizes fiscal year 1975 projects.
-------
SECTION II
RECOMMENDATIONS
Existing analytical techniques should be continually improved,
and new techniques should be investigated for applicability to
pollutant analysis. Information about specific pollution
incidents should be widely disseminated to help solve and
perhaps prevent future environmental problems.
-------
SECTION III
DISCUSSION
CHEMICAL ELEMENTS IN GREAT LAKES FISH SAMPLES
To provide baseline data for future pollution studies.
Environment Canada, through the U.S. Bureau of Sport Fisheries
and Wildlife in Ann Arbor, Michigan, requested spark source mass
spectrometric (SSMS) survey analyses of chemical elements in
four fish samples. Personnel from Environment Canada collected
samples of chubs and burbot from Lake Huron and lake trout from
Coppermine Point and Apostle Island in Lake Superior. Each
sample was an entire fish carcass, including intestines, that
was ground, blended, frozen, and shipped to the Athens
Environmental Research Laboratory, where it was prepared for
SSMS analysis.
After addition of yttrium as an internal standard, a 5-g aliquot
of each sample was digested in 50 ml of formic acid for 6 hr to
destroy proteinaceous material. Graphite was added, and each
sample was dried under infrared light. To destroy all remaining
organic material, each sample was heated in a muffle furnace at
460°C for 6 hr.
Duplicate analyses of the four fish samples provided
concentration data for 38 elements (Table 1). Four other
elements (Zr, Cu, Ni, and Al) that were detected in the fish
samples were not reported, because comparable amounts were found
in the blank sample. These elements were probably introduced
during formic acid digestion.
These data were reported to the supervisory chemist of the
Analytical Chemistry Group of Environment Canada in Winnipeg,
Manitoba, where they will be compiled with other pollution
parameters and maintained for future use.
-------
Table 1. CHEMICAL ELEMENTS DETECTED BY SSMS IN GREAT LAKES FISH
SAMPLES
Concentration, yg/kga
Element
Ca
P
Na
S
Zn
I
K
Fe
Mg
Sn
Cl
Br
Rb
Mn
Cr
Sr
Co
Ge
Se
Ti
Pb
V
Ba
So
Ag
In
Mo
F
Ga
Cs
As
La
Te
Cd
Nd
Rh
Pr
Ce
a
b
^
Trout from
Coppermine Point
>45,000
>26,000
>24,000
>23,000
21,000
16,000
>16,000
14,000
7,000
7,000
> 5,000
3,100
2,800
2,700
1,600
1,200
360
240
180
160
100
85
55
55
45
30
25
20
15
10
2
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
Average of duplicate
•.
«•
--.-., ::_._., z
_..yl —
Trout from
Apostle
>71,000
>38,000
>32,000
>22,000
27,000
25,000
>16,000
12,000
82,000
430
>53,000
8,200
3,800
3,700
980
2,300
370
270
200
200
110
85
120
120
180
130
220
460
5
20
5
130
10
N.D.
35
20
N.D.
5
analyses.
Not detected; detection level
Island Burbot
>54,000
>26,000
>25,000
>22,000
76,000
18,000
>13,000
22,000
22,000
820
>50,000
8,500
2,400
1,700
680
3,200
240
160
190
130
95
75
250
75
60-
25
30
830
20
N.D.
5
25
50
N.D.
N.D.
N.D.
N.D.
N.D.
is 0.5 u9/kg.
Chubs
>27,000
>15,000
>14,000
>60,000
25,000
12,000
> 6,000
11,000
9,000
350
>60,000
5,300
2,200
2,400
1,800
3,800
1,000
1,300
280
160
75
260
300
440
80
55
N.D.
270
120
N.D.
13
60
N.D.
170
N.D.
N.D.
20
N.D.
-------
MULTIELEMENT ANALYSIS OF POULTRY FARM WATER SAMPLES
Investigation of possible causes of a poultry liver disease
prompted a request for the ACB to analyze chemical elements in
farm water samples. The incidence of this disease. Fatty Liver
Syndrome (FLS), was much greater among laying hens in south
Georgia than in north Georgia. The disease was suspected to be
related to the presence of manganese in the hens' drinking
water, because manganese had been correlated with fat
accumulation in rats.
Scientists in the Department of Poultry Science at the
University of Georgia were studying the problem under the
auspices of the Georgia Egg Commission and requested the ACB's
assistance. Instrumental neutron activation analysis of water
samples was desired to confirm manganese concentration data that
had been obtained by emission spectrography and to provide
information about other chemical elements present.
The ACB was given samples that had been collected from 9 farms
with a history of FLS and from 14 farms without any apparent FLS
problem. Each 1-liter sample had been collected in a plastic
bottle at the point where the chickens received drinking water.
After a preservative (chloroform) was added, the samples were
transported to the University and divided into aliquots for
analysis.
Several elements (Mn, Ca, Mg, Sn, Na, Fe, Cu, Zn, Al, and B)
were detected and measured with a direct reading emission
spectrograph at the Soil Testing Laboratory of the University
Cooperative Extension Service. Instrumental neutron activation
analysis (INAA) by ACB personnel provided manganese
concentration data and detected other chemical elements present
at concentrations <1 pg/1, the approximate detection level for
most elements.
No correlation between manganese content and geographic location
of the sampling sites was revealed by INAA data. Manganese
content varied from a level that could not be detected to 65
yg/1, with mean values of 8.5 ±2.1 pg/1 in water samples from
farms without the disease and 19.6 ±8.0 yg/1 from those with the
disease. These INAA manganese concentration data correlated
with the emission spectrographic analyses, which produced mean
values of 7.0 ±2 ]jg/l and 14.7 +7 pg/1, respectively.
The INAA manganese concentration data and a list of all elements
detected were reported to the scientists at the University's
Department of Poultry Science. Their comparison of INAA data
with emission spectrographic data produced a possible
-------
correlation between the Fatty Liver Syndrome and water hardness.
Most poultry farms with the disease problem had water with
considerably higher concentrations of calcium and magnesium than
were found in water samples from farms without the disease. The
University of Georgia scientists concluded that further
investigation is necessary to determine whether lipid metabolism
in laying hens is directly related to the mineral content of
their drinking water.
CHEMICAL ELEMENTS IN DRINKING WATER SAMPLES
To compare and evaluate several methods for multielement
analysis of chemical elements in water, the EPA's Health Effects
Research Laboratory in Cincinnati, Ohio, requested multielement
analysis of about 100 drinking water samples. The ACS agreed to
analyze all of these samples with neutron activation and to
obtain comparative data for about 10 of these samples by
analyzing them with SSMS.
Drinking water samples collected from all over the country were
used for these comparative analyses. The first 39 samples
requiring neutron activation analysis arrived at the Athens ERL
in April and May 1975 and some analyses were completed in the
latter part of FY 75. Preliminary INAA data indicated that
Public Health Service water quality criteria for copper,
manganese, cadmium, and chromium might have been exceeded at one
or more sampling sites. Three samples requiring SSMS analysis
arrived in May 1975, and quantitative data for 40 elements were
obtained from duplicate analyses.
Work on this project will continue during FY 76. As data are
obtained, they are reported to the Health Effects Research
Laboratory, where they will be correlated with data from other
analytical methods.
MERCURY IN SEAWATER
To confirm unexpectedly high mercury concentrations in seawater
samples analyzed by atomic absorption spectrophotometry (AAS),
the EPA's Region II Quality Control Coordinator requested
neutron activation analysis of four seawater samples. These
samples had been collected about 150 miles offshore of New
Jersey at the 106 mile ocean dump site, the deepest and farthest
offshore site assigned to industries with permits for ocean
dumping. This site is intended for the most toxic and hazardous
wastes and is monitored periodically by the National oceanic and
-------
Atmospheric Administration (NOAA) to guard against environmental
degradation.
In the summer of 1974, samples collected at this site by NOAA
were analyzed by AAS at the EPA Region II laboratory. Levels of
mercury were 20 times higher than previously encountered
concentrations. To confirm the AAS data, four of these samples
were sent to the ACB for analysis by another technique. Mercury
concentration data obtained from duplicate neutron activation
analyses (Table 2) confirmed those obtained with AAS. Because
sample contamination during collection procedures was thought to
be the probable cause of these high mercury concentrations, the*
Quality Control Coordinator recommended that no action be taken
unless Region II AAS analysis of the next set of samples,
collected with careful attention to possible contamination
sources, indicated that they also contained high levels of
mercury.
Table 2. MERCURY IN SEAWATER SAMPLES ANALYZED BY
NEUTRON ACTIVATION
Mercury Concentration, yg/kg
Sample * Run 1 Run 2 Average
1
2
3
4
2.65
3.94
2.84
11.7
2.57
3.41
3.09
13.9
2.61
3.68
2.96
12.8
CHEMICAL ELEMENTS IN PLANT AND SOIL SAMPLES FOR BIOMONITORING OF
POWER PLANT EMISSIONS
To determine if certain desert plants can be used for biological
monitoring of chemical elements emitted from power plant
smokestacks, the EPA's Environmental Monitoring and Support
Laboratory (EMSL) at Las Vegas, Nevada, requested multielement
analyses of plant tissues and soil samples. They hoped to
correlate sampler chemical element content with stack gas
-------
composition and with distance of the sampling site from the
power plant.
Initially, neutron activation analysis of five standard samples
was requested to confirm mercury and cadmium concentration data
obtained by Zeeman spectroscopy, a new instrumental method in
use at the Las Vegas laboratory. Certified values were
available for the mercury and cadmium content of two samples
(NBS Standard Reference Materials for coal and orchard leaves)
but not for the other three samples (soil and fish and plant
tissue). Satisfactory confirmation was obtained for mercury
data, but INAA was not sensitive enough to detect cadmium in
four of the five samples (Table 3) .
Later, the ACB was asked to determine concentrations of 20
specific elements in 30 plant tissue samples and 3 soil samples.
These samples were collected at three sites at varying distances
downwind from a power plant in the Four Corners area of the
western United States. Concentration data were needed for six
elements (Be, Pb, Ni, Se, F, and S) that could not be determined
satisfactorily by INAA, because detection limits for these
elements are higher than the concentrations expected in these
samples. Since these elements can be detected with SSMS at a
lower level ( 'v 1 ng/g) than with INAA (<1 pg/g) , SSMS was
expected to be the more useful analytical technique.
All 33 samples were analyzed by SSMS; 3 soil samples and 20
plant samples were also analyzed by INAA. Samples of soil and
three varieties of vegetation were collected at all three
sampling sites, one of which was a control site. Five plant
varieties were collected at only 2 sites; the remaining 11 plant
varieties were collected at only one site.
8
-------
Table 3. NEUTRON ACTIVATION DATA FOR MERCURY AND CADMIUM
IN STANDARD SAMPLES
Sample
Mercury
Concentration, yg/kg
Cadmium
Concentration,
NBS Orchard
Leaves
NBS Coal
Oregano
Tuna
Mine Mtn.
Soil #3
INAA
Data
151 b
120 a
654 a
366 a
280 a
Certified INAA
Value Data
155 N.D.
130 N.D.
N.D.
N.D.
14,300
Certified
Value
110
__
— —
__
—
a Average of duplicate analyses
b Not detected; detection limit is ^ 1
For SSMS analysis, weighed samples of approximately 10 g were
ashed at 450°C for 4 hr. The residue was mixed and formed into
an electrode. A single analysis by SSMS was performed on each
sample, and no attempt was made to adjust sample size so that
major elements could be determined quantitatively.
The INAA samples were irradiated in duplicate in plastic vials.
Data were obtained after both long and short periods of
irradiation. Long irradiation periods lasted 4 hours with
subsequent counting for 600 sec after 4 days' decay; after 8 to
15 days these samples were counted again for 2 hours. Short
irradiations lasted 3 to 30 sec; counting began as soon after
irradiation as possible and lasted 600 sec. For elements
measured after both long and short irradiations, separate
isotopes were measured to provide independent confirmation;
copper and potassium were exceptions. For these elements, the
same isotopes were measured after both long and short
irradiations.
Because many of the elements determined quantitatively by INAA
exceeded the concentration range for SSMS, comparison of data
from the two different techniques was frequently not possible.
When possible, comparisons showed generally lower values were
obtained by SSMS, especially for chlorine. The lower SSMS
values for chlorine probably resulted from loss of organic
chlorine during sample preparation.
-------
Data interpretation by conventional analysis of variance showed
that location of sampling site was insignificant as a source of
experimental variance. However, factor analysis suggested 8
elements (Ir Ba, Rb, La, Al, Mo, Ce, and Cl) may discriminate
between samples taken at the control site and those taken at the
other two sampling sites. Particular plant species are even
more promising discriminators: Larrea divaricata with respect
to I, Ba, and Cl content, and Ephedra nevadensis with respect to
Al, La, Mg, and Zn content. Of these elements, only zinc is
likely to be relatively more concentrated in fly ash than in the
soil. Consequently, a large question remains as to whether the
indicated correlation between location and the elemental content
of these plants actually reflected contamination by the power
plant plume.
The ACB's report to the EMSL concluded that these results did
not provide conclusive evidence of soil and plant contamination
from the power plant plume. However, these results may be
helpful in planning similar future experiments. In such
studies, samples of the suspected source of pollution should be
taken to obtain plant tissue samples that are free of surface
contamination by soil and that are as homogeneous as possible.
CHEMICAL ELEMENTS IN VARIOUS ENVIRONMENTAL SAMPLES
During the year, the ACB analyzed a variety of samples for the
EPA's Region IV Surveillance and Analysis Division to assist
them in locating pollutant sources. Their monitoring activities
required multielement analyses of industrial wastewaters,
reservoir water, river sediment, air participates, and sewage
sludge.
Survey analyses of the industrial wastewaters were requested to
determine if pollutant discharge limits were exceeded.
Multielement analysis by SSMS identified chemical elements
present in significant concentrations and provided an elemental
profile for each type of industrial effluent.
Most industrial wastewaters conformed to the appropriate
discharge limits. Industries monitored included three can
manufacturing plants, four steel mills, two synthetic fiber
plants, a molecular sieve manufacturing facility, and a wire
manufacturing plant. The wire plant wastewater contained
excessive copper, lead, and zinc; the concentration of each
element exceeded the total amount permitted for all heavy
metals.
10
-------
Other environmental samples were analyzed to help locate sources
of specific pollution problems. Sediment from a stream used for
cattle drinking water was analyzed to determine if heavy metal
contamination could be causing cattle poisoning; no heavy metals
were detected in significant concentrations. A fluffy air
pollutant, which was collected by a citizen from his patio, was
suspected to be insulation material emitted from a nearby
insulation manufacturing plant. Data from 3SMS analysis were
compatible with the composition of the suspected material,
perlite. Each of six sewage sludge samples contained 35-37
elements at concentrations >1 pg/1. A 5-day composite sample
of reservoir intake water contained 25 elements at
concentrations >1 yg/1.
Results of these analyses were reported to Region IV personnel
for use in enforcement proceedings, when appropriate.
ORGANIC CONTAMINANTS IN NEW ORLEANS DRINKING WATER
In response to a request by the EPA»s Region VI Surveillance and
Analysis Division, the ACB identified and measured non-purgeable
volatile organic compounds in a city drinking water. Because of
allegations that the New Orleans, Louisiana, drinking water
contained pollutants hazardous to consumers1 health, state and
city officials requested the EPA to examine the city's drinking
water. This task required the cooperative efforts of several
EPA laboratories. The ACB had demonstrated the use of combined
gas chromatography-mass spectrometry for identification of
specific organic contaminants in water *-* and had experience in
sample extraction and separation techniques. Therefore, ACB
personnel were asked to perform some analyses and to recommend
extraction and separation procedures to be used.
The types and numbers of samples to be collected and analyzed
were determined by EPA personnel in Region VI. Samples were
collected by personnel from the EPA's Lower Mississippi River
Facility (LMRF) at, Slidell, Louisiana, the Region VI laboratory
that was responsible for coordination of this project.
ACB personnel analyzed various extracts from ten samples and
their corresponding blanks. Gas chromatography (GC) and
combined gas chromatography-mass spectrometry (GC-MS) were used
to identify and quantify as many as possible of the volatile
organic components.
11
-------
Sample Collection and Extraction
In July and August 1974, four types of samples were collected
from three different water treatment plants serving the New
Orleans metropolitan area. Region VI personnel collected three
types of samples using carbon sorption techniques and extracted
them with chloroform to remove sorbed organic material. At one
water treatment plant, a large ("mega") sample was collected
during a 7-day period by passing 1.1 million liters of water
through approximately 20 kg of activated carbon. Carbon from
this mega sampler was extracted at the EPA's Robert Taft Center
in Cincinnati, Ohio. The carbon was extracted for 40 hr with
approximate y 190 £ of chloroform in a large-scale reflux
extraction unit. Excess chloroform was distilled off until
about 2 liters remained in the pot.
A sample from each of the three water treatment plants was
collected by passing 25,500 & of water through two Pyrex
cy inders, each containing approximately 340 g of activated
charcoal, connected in series. This sample, which was collected
over a 7 day period, represented the theoretical amount a person
would drink during 70 years at the rate of 1 £/day. The carbon
from each cylinder was extracted at the EPA's Robert S. Kerr
Laboratory at Ada, Oklahoma, in a special room designed to
minimize contamination. Chloroform extraction for 48 hr in a
Soxhlet apparatus was followed by vacuum concentration of the
extract at temperatures <27°C to a volume of 30 to 60 ml. After
chloroform extraction, the first of the two carbon filters in
series was also extracted with ethyl alcohol.
At each of the three sites, the third type of carbon adsorption
sample was collected during a two-day period by passing 60 £ of
water through a small polyvinyl chloride tube containing
approximately 70 g of carbon. Each of these samples was assumed
to be equivalent to a person's water consumption during 2
months. These samples were extracted at the EPA's Region VI
laboratory in Houston, Texas. A final volume of 25 ml was
obtained after extraction in a Soxhlet apparatus.
ACB personnel participated in collection of the other two types
of samples, which were concentrated using resin adsorption and
tetralin extraction techniques. The resin adsorption technique
had been developed7 by Gregor Junk (Ames Laboratory, Iowa State
University, Ames, Iowa) who participated in this project. Small
glass columns containing cleaned Rohm and Haas XAD-2 resin were
connected to finished water lines at all three treatment plants,
and approximately 320 £ of water were passed through each tube.
Ethyl ether extracts of these samples were concentrated to 1 ml.
12
-------
A high-boiling solvent, tetralin (1,2,3,4-
tetrahydronaphthalene), was used to extract triplicate samples
from all three water treatment plants. This technique,
originally suggested by B. F. Dudenbostel of EPA's Region II,
permits GC analysis of the very volatile components that elute
with and are obscured by conventional solvents, such as
chloroform. A 2-ml portion of tetralin was used to extract each
of three 1-& water samples from each plant.
Each sample extract or blank extract was placed in an
appropriate container and shipped to the ERLA for analysis by GC
and GC-MS. Portions of the resin extracts were also analyzed by
GC and GC-MS at the Ames Laboratory.
Extract Concentration and Separation Procedures
Mega Sample—
At the ERLA the chloroform extract of the mega sample was
concentrated in a Kuderna-Danish apparatus to make 1 ml of each
extract equivalent to 1 H of water passed through the filter.
Several fractionation techniques were investigated to obtain the
best feasible GC separation of sample components. Solubility
separation of the total extract into acid, basic, and neutral
fractions showed that most sample components were neutral
compounds, hut did not improve GC separation significantly.
Steam distillation neatly separated the extract into volatile
distillable and non-distillable components but did little to
improve GC separation of chromatographable components. No
polynuclear aromatic hydrocarbons were detected by thin layer
chromatography using techniques designed to isolate this class
of compounds. However, two chlorinated pesticides were detected
by electron capture GC after fractionation by column
chromatography.
Seventy-year Samples—
Extracts of the 70-year sample filters, which had sorbed organic
material from 25,500 H of water were also concentrated to make 1
Vl equivalent to 1 & of filtered water. As with the mega
sample, column chromatography was used to fractionate the
chloroform extracts of the 70-year sample filters from one water
treatment plant.
Two-month Samples--
Chloroform extracts of the 2-month samples were concentrated in
a Kuderna-Danish apparatus to 0.3 yl to make 1 yl of extract
equivalent to 100 ml of filtered water.
13
-------
Resin Samples—
Ethyl ether extracts of the XAD-2 resin filters were
concentrated to 1 ml in a micro Kuderna-Danish apparatus.
Further solvent evaporation with a stream of dry nitrogen
reduced the volume to make 1 yl equivalent to 1 £ of filtered
water.
Tetralin Extracts—
»
Because the tetralin extracts were obtained to identify very
volatile water contaminants, these extracts could not be
concentrated by solvent evaporation.
Analysis
All sample extracts and blanks were analyzed by gas
chromatography (flame ionization detector). After preliminary
examination by GC techniques, low resolution electron impact
mass spectral data were acquired with a computer-controlled GC-
MS system. Some samples were also analyzed with electron
capture GC and with chemical ionization and high resolution
electron impact GC-MS systems.
Preliminary identifications of sample components were obtained
by computer matching of unknown mass spectra with standard mass
spectra in a central data bank and by visual comparison of
unknown spectra to standard spectra in scientific publications
or on file at the ERLA. When possible, standard samples were
obtained to confirm tentative identifications by comparing GC
retention time data and mass spectra of sample components with
those of standards analyzed under the same conditions. To
confirm the presence of two sample components, combined gas
chromatography-Fourier transform infrared spectrophotometry was
used to compare data from sample components and reference
compounds.
Quantitative data were calculated from GC peak area
measurements. Atrazine* a compound present in all the New
Orleans water extracts was chosen as an internal standard.
Solutions of known amounts of pure reference compounds were
prepared and mixed with a stock solution of the internal
standard. The flame response and retention time of each
reference compound relative to the internal standard were used
to calculate sample component concentrations. In some cases,
the flame response calculated for a reference compound was also
used for other compounds of the same chemical class. Blank
samples, which contained no atrazine, were spiked with this
internal standard for component concentration calculations.
14
-------
Results
Analysis by GC and GC-MS permitted tentative identification of
76 organic components of the drinking water extracts (Table 4);
the presence of 49 of these compounds was confirmed. Of the 76
compounds, 9 were not specifically identified as to particular
isomer, and 8 others were identified only by compound class or
by certain structural features. Compounds present in blanks
were not reported as sample components except in cases where at
least five times greater amounts were found in the sample
extract.
Calculated contaminant concentrations ranged from about 0.01
yg/1 (the detection limit for most compounds identified by the
techniques used) to 12 yg/1. Most compounds detected were
present at the lower end of this range; the relationship between
concentration level and the number of detected compounds is
shown for two samples from one water treatment plant (Table 5).
These calculated concentrations are only estimates; unknown
carbon sorption/desorption characteristics of sample components
preclude stating concentrations with a high degree of accuracy.
Table 5. CONCENTRATION CLASSIFICATION OF ORGANIC COMPOUNDS
DETECTED IN NEW ORLEANS DRINKING WATER
Concentration
Range, yg/1 Number of Compounds Detected
70-year Sample Mega Sample
Extract Extract
0.01-0.09
0.10-0.99
1.0 -10
74
19
4
84
17
2
Total 97 103
Fifteen compounds present in low concentrations (0.1 to 0.3
yg/1) produced usable mass spectra but were not identified.
Many other compounds in even lower concentrations were detected
by GC, but did not produce usable mass spectra. For example,
about 100 peaks were counted in the gas chromatogram (from a
packed column) of the chloroform extract of one 70-year sample,
but only 47 compounds were identified.
15
-------
Table 4. ORGANIC COMPOUNDS IDENTIFIED IN NEW ORLEANS DRINKING WATER
Concentration , yg/la
Water Treatment Plant A Hater Treatment Plant B Water Treatment Plant C
Compound Name
RRTb
Mega
Sample
CCE°
70-Year
Sample
2-Month
Sample Resin
CCEC Extract
70 -Year
Sample
CCEC
2-Month
Sample Resin
CCE Extract
70-Year
Sample
CCEC
2-Month
Sample
CCEC
Resin
Extract
Aliphatic Hydrocarbons
1.
2.
3.
4.
5.
6.
7.
8.
9.
n-decane
Secane , branched
n-dodecane
iT-nonane^
n-pentadecaned
n-tetradecane^
n"-tridecaned
n-undecaned
undecane, branched
0
0
0
0
0
0
0
0
0
.48
.44
.67
.38
.86
.80
.72
.58
.52
0.06
0.03
..—
-_
—
__
0.03
0.04
__
0.02
0.01
0.03
0.03
0.02
0.01
«
0.06
2.4
5.8
0.10
2.4
0.10
0.30
2.5
5.3
__
—
—
~
— —
—
—
~~
2.0
5.4
0.40
2.4
0.10 0.01
0.10
0.17
-------
Table 4. ORGANIC COMPOUNDS IDENTIFIED IN NEW ORLEANS DRINKING WATER (Continued).
Concentration , yg/1
Water Treatment Plant A
22.
23.
24.
25.
26.
27.
Compound Name
1,1, 1-trichloropropane
1,2, 3-trichloropropane
chlorinated, fluori-
nated hydrocarbon
chlorinated, fluori-
nated hydrocarbon
dichlorinated hydro-
carbon, MW 200
dichlorinated hydro-
carbon, MW 249
RRTb
M).
0.
0.
0.
0.
•fcl.
07
36
78
82
82
0
Mega
Sample
CCEC
—
<0.2
--
—
0.05
<0.02
70 -Year
Sample
CCEC
__
— -
<0.01
<0.01
--
<0.04
2 -Month
Sample Resin
CCEC Extract
<0.1
«
— <0.1
~ <0.1
-_ __
-- "
Water Treatment Plant B
70 -Year
Sample
CCEC
__
—
—
--
—
*%
De
2 -Month
Sample Resin
CCE° Extract
__
_» -_
— — ^0 . 1
—— ^0 . 1
__ __
-_ -._
Water Treatment
70-Year
Sample
CCEC
—
--
—
—
~
A
D6
2 -Month
Sample
CCEC
—
«
~
— -
~
— —
Plant C
Resin
Extract
—
—
<0.1
<0.1
~
——
Chlorinated Aliphatic
Ethers
28.
29.
bis (2-chloroethyl)
ether"3
bis (2-chloroisopropyl)
etherd
0.
0.
44
51
--
D6
0.04
0.18
_- -—
—
0.16
0.08
— __
~
0.12
0.03
~
—
--
~
Aromatic Hydrocarbons
30.
31.
32.
33.
34.
35.
36.
37.
38.
Cj-alkylbenzene
C,-alkylbenzene
cf-alkylbenzene
ethylbenzene
m-ethyl toluene":
o-ethyl toluene"
toluened
trims thy Ibenzene isomer
o-xylene°
-------
Table 4. ORGANIC COMPOUNDS IDENTIFIED IN NEW ORLEANS DRINKING WATER (Continued).
00
Concentration, ug/la
Water Treatment Plant A Water Treatment Plant B
Compound Name
Chlorinated Aromatics
39. m-dichlorobenzene^
40. 1,2,3,4,5,7,7-hepta-
chl or onorbornene
41. heptachloronorbornene
isomer
42. pentachlorophenyl
methyl ether
Herbicides and Related
Compounds
43. alachlor (Lasso)
44. alachlor homolog
(alachlor + 1 chlorine)
45. atrazine'-'
46 . atrazine homolog
(atrazine - C^H^) .
47. butachlor (Macnetey
48. cyanazine (Bladex) "
49. propazine"
50. simazined
Pesticides
51. a-chlordane
52. chlordened
53. dieldrind
54. endrin
RRTb
0.
0.
0.
g
i.
i.
i.
0.
i.
i.
M..
•v.1.
g
g
i.
i.
46
94
98
09
14
00
96
21
13
0
0
23
27
Mega
Sample
CCEC
0
0
<0
0
0
2
0
0
<0
-------
Table 4. ORGANIC COMPOUNDS IDENTIFIED IN NEW ORLEANS DRINKING WATER (Continued).
Concentration, yg/la
Water Treatment Plant A Water Treatment Plant B
Water Treatment
Mega 70-Year 2-Month 70-Year 2-Month 70-Hear 2-Month
. Sample Sample Sample Resin Sample Sample Resin Sample Sample
Compound Name RRT" CCEC CCE° CCEC Extract CCE= CCEC Extract CCEC CCEC
Phthalates
55. benzyl butyl phthalated 1.29 0.24 0.64 1.4 — 0.83 1.8
56. dibutyl phthalated 1.12 0.09 0.10 — 0.05 0.36 — 0.01
57. diethyl phthalated 0.91 0.02 0.03 0.24 -- 0.03 0.10 0.03
58. di-2-ethylhexyl
phthalated 1.38 11 0.10 0.40 0.05 0.46 0.50
59. dihexyl phthalate 1.24 — 0.03 — 0.05
60. diisobutyl phthalated 1.06 0.59 — <1 <0.05
61. dimethyl phthalated 0.83 0s 0.27 0.60 — 0.13 0.82
62. dipropyl phthalated 1.02 0.01 0.07 -- — 0.13
63. unknown phthalate 1.22 0.01
64. unknown phthalate 1.24 0.01
65. unknown phthalate 1.36 0.12
Miscellaneous
66. benzaldehyded 0.44 0.03
67. 2,6-di-t-butyl-p_-
benzoquinoned 0.84 — 0.22 — — 0.21
68. di-2-ethylhexyl adipated 1.32 0.10
69. dihydrocarvone 0.63 — 0.14 — — 0.06
70. ethyl acetate — — — — De
71. isophoroned 0.60
72. limonened 0.50 — 0.03
73. methyl benzoated 0.58 — — — — — — <0.01
74 1,3,5-trimethyl
isocyanurate^ 0.73 0.01 0.01
75. triphenylphosphatea 1.31 0.03 0.12
76. unknown compound MW 145 ^0.40 <0. 3 <0.7 — — <0.9
a Blank spaces indicate that the compound was not detected in that particular sample.
b GC retention time relative to the internal standard, atrazine.
c CCE = the chloroform extract of the carbon filter.
d The compound identification was confirmed by matching the mass spectrum and GC retention
component with those of a standard analyzed with the same instrumental conditions.
0.75 1.6
0.23
0.01 0.18
0.27 1.2
__ __
__ —
0.18 0.74
0.14
__ __
—
~— —
—
0.25
—
0.07
— —
— __
~~ «.«
—
_« _«
<0.9
time of the sample
Plant C
Resin
Extract
0.08
0.03
--
0.16
0.16
—
—
--
--
—
— —
—
—
--,
A
D6
--
—
—
e Detected but not quantified because of low concentration or interference from other sample components.
f Chloroform was detected only in the tetralin extract of water from treatment plant A.
g Retention time not calculated; detected only in column chromatographic fraction.
h Includes the amount of atrazine found in the ethyl alcohol extract.
i The compound identification was confirmed by comparison of the infrared spectrum of the
with that of a standard; both spectra were obtained with a GC-FTIR system.
sample component
-------
No carboxylic acids or phenols were identified in any of the
sample extracts, including the carbon filter alcohol extract,
which was examined before and after methylation with
diazomethane. No polynuclear aromatic hydrocarbons were
identified even after fractionation by thin layer
chromatography. Some compounds (endrin, a-chlordane, chlordene,
and pentachlorophenyl methyl ether) were not detected in the
total extracts but were identified in fractions separated by
column chromatography.
Identification of some compounds whose mass spectra were not in
the central data bank or in spectra collections required special
techniques. Other compounds were detected only after special
sample treatment.
Dieldrin and Endrin—
These two organochlorine pesticides were detected by electron
capture GC after fractionation of the mega-sample extract by
column chromatography. The presence of both compounds was
confirmed by matching retention times of sample components with
those of standards using three different GC columns of varying
polarity. Dieldrin was later detected in the total extract by
GC-MS analysis but only by using a special technique. This
pesticide was detected by computer searching for characteristic
ions in the stored mass spectra collected with repetitive
computer-controlled data acquisition as GC peaks eluted.
Simazine and Propazine—
Computer searching of stored spectra for characteristic ions
also provided confirming evidence of the presence of simazine
and propazine in water extracts. These two compounds were
obscured in sample chromatograms by larger concentrations of
atrazine, which had approximately the same retention time with
the GC conditions used. Mass spectra acquired continually as
the GC peaks eluted indicated that the atrazine peak was
contaminated with other sample components; molecular ions and
some significant fragments of these other components were
observed. Computer searching for definitive ions indicated that
simazine was present in the leading edge of the atrazine GC
peak, and propazine was present in the trailing edge.
Alachlor—
The identification of a significant component of sample extracts
from all three water treatment plants required a combination of
analytical techniques. Its mass spectrum was not in the
computerized data bank or in published spectra collections. The
compound1s molecular weight (269) was determined from its
20
-------
chemical ionization mass spectrum. • High resolution electron
impact GC-MS data produced possible empirical formulae for the
molecular ion and important fragment ions. Correlation of these
data with reasonable fragmentation modes provided the most
probable molecular formula. Alachlor, a herbicide with this
molecular formula and the appropriate structural
characteristics, was a reasonable contaminant. The sample
component and a standard sample of alachlor produced the same
low resolution mass spectra, had the same GC retention times,
and produced matching infrared spectra (analysis by GC-FTIR).
These data confirmed the identification of alachlor, 2-chloro-
2',6' -diethyl-N-(methoxymethyl)acetanilide.
Alachlor Analog—
Some of the same techniques were used to identify a compound
that produced a low resolution electron impact mass spectrum
similar to that of alachlor. Chemical ionization and high
resolution electron impact mass spectrometry provided additional
structural information. These data indicated that the molecular
structure of the sample component was very similar to that of
alachlor. This unknown was identified as 2,2-dichloro-2',6'-
diethyl-N-(methoxymethyl) acetanilide, which is equivalent to
alachlor with an additional chlorine. A standard was not
available to confirm this tentative identification. Laboratory
experiments indicated that this compound was not formed during
chlorination treatment of water containing alachlor.
Trimethyl Isocyanurate--
By computer matching of mass spectra, one sample component was
tentatively identified as trimethyl isocyanurate (1,3,5-
trimethyl-2,4,6-trioxohexahydrotriazine) . A reference sample of
this compound was not commercially available but was synthesized
at the Athens ERL. The identification was confirmed by matching
the mass spectrum and GC retention time of the synthesized
material with those of the sample component.
This study did not produce any conclusions as to the best
sampling method. Analysis of only one sample type, whether the
mega sample, the 70-year sample, the 2-month sample, or the
resin sample, would not have permitted identification of all
these 76 compounds. Although nine compounds were detected in
the mega sample but not in the other three samples collected
from Plant A, four of these nine compounds were detected in
fractions obtained with column chromatography, which was not
used for the other Plant A samples. More compounds were
identified in the 70-year samples than in the 2-month or resin
samples, but still more were identified when both 70-year and
two-month samples were analyzed (Table 6).
21
-------
Table 6. COMPARISON OF NUMBER OF IDENTIFIED COMPOUNDS
IN DIFFERENT SAMPLE TYPES
Number of Identified
Sample Type Compounds
70-Year Samples 55
2-Month Samples 29
Resin Samples 28
70-Year and 2-Month Samples 60
70-Year and Resin Samples 53
2-Month and Resin Samples 42
The tetralin extracts proved to be too dilute for detection of
any volatile components except chloroform, which was the only
compound identified by this technique. Most GC peaks observed
in tetralin sample chromatograms were also present in blank
chromatograms. In the future, 1-£ samples will be extracted
with 1 ml of tetralin that has been purified to eliminate
interfering peaks. Whether this will increase the sensitivity
of this technique enough to be useful for drinking water samples
has not been determined.
In general, more compounds and higher concentrations were
observed in carbon filter extracts than in resin extracts, but
more data are needed to make valid comparisons of XAD resin and
carbon filter sorption and extraction efficiencies.
The alcohol extract of the 70-year carbon filter produced only a
few, poorly resolved GC peaks when examined without further
treatment. Since most components were expected to be polar
compounds such as carboxylic acids and phenols, the extract was
methylated with diazomethane. Six compounds were identified in
this methylated extract, but all were present and identified in
the chloroform extract also. No carboxylic acids or phenols
were found.
To obtain an indication of how well the organic contaminants
were sorbed by the carbon filters, chloroform extracts of the
two filters connected in series were analyzed separately for the
70-year sample. For each of the 29 compounds detected at a
total concentration >0.04 yg/1, the percentage recovered in each
filter was calculated (Table 7). Comparison of the amounts
found on filters t1 and #2 showed that >96% of 2H compounds was
extracted from filter f1. A general high degree of sorption was
22
-------
indicated by the wide variety of compound classes included in
these 24 compounds, but significant exceptions were noted. For
example, more hexachloroethane was extracted from filter t2 than
from filter #1, and only 64 to 85% of three trihalogenated
methanes or ethanes was adsorbed by filter #1.
Comparison of sample component concentrations showed that up to
10 times higher concentrations of the lower-boiling compounds
were found in the 2-month sample extract than in the 70-year and
mega sample extracts. However, concentrations of the less
volatile components were approximately the same in the three
types of carbon extracts. Because the samples were not
collected at the same time, contaminant concentrations could
have changed, but concentration variations should not have been
a function of volatility. These results suggest that the 70-
year filter was overloaded with organic material, but this
explanation is contradicted by recovery data (Table 7).
Concentrations of mega sample components were low compared to
those of the 70-year sample. Of 30 compounds quantitated by
both methods, 21 were found in the mega sample at concentrations
only 25 to 76% of concentrations in the 70-year sample.
Differences in solvent evaporation techniques may be partially
responsible for the concentration discrepancies.
Results of these analyses of carbon and resin extracts were
reported to the EPA's Region VI Surveillance and Analysis
Division. The Athens ERL data were compiled with data from
other laboratories participating in the survey of New Orleans
drinking water contaminants and were incorporated into an EPA
summary report8. These contaminant identifications and other
data from the survey will permit other researchers to begin
assessing the relationship between contaminants and public
health.
ORGANIC CONTAMINANTS IN DRINKING WATER FROM 10 UNITED STATES
CITIES
As part of a nationwide EPA survey of water supply systems, the
ACB analyzed drinking water samples from 10 United States cities
to identify organic components that were sorbed from water by
carbon and desorbed from carbon by solvent extraction. This
survey was initiated by the EPA Administrator, Russell E. Train,
to identify and estimate the concentration of organic compounds
in drinking water from all over the nation. Part of the survey
required analysis of drinking water from 80 U.S. cities to
determine concentrations of 6 specific halogenated compounds
before and after disinfection by chlorination. These data were
23
-------
Table 7. COMPARISON OF RECOVERIES FROM FILTERS #1 AND #2
OF THE 70-YEAR SAMPLE
Total Concentration I From % From
Compound yg/1 Filter #1 Filter #2
Alachlor 0.82 97 3
Alachlor homolog 0.28 100 0
Atrazine 4.9 98 2
Atrazine, deethyl 0.51 96 4
Benzyl butyl phthalate 0.64 100 0
Bromoform 0.57 100 0
Butaahlor 0.05 100 0
Cyanazine 0.35 100 0
Chlorodibromomethane . 1.1 64 36
bis-2-Chloroethyl ether 0.04 100 0
bis-2-Chloroisopropyl ether 0.18 100 0
Dibromodichloroethane isomer 0.33 100 0
Dibutyl phthalate 0.10 .80 20
2,6-Di-t-butyl-p-benzoquinone 0.22 100 0
Dichloroiodomethane 1.1 85 15
Dieldrin 0.04 100 0
Di-2-ethylhexyl phthalate 0.10 100 0
Dimethyl phthalate 0.27 100 0
Dipropyl phthalate 0.07 100 0
1,2,3,4,5,7,7-Heptachloronorbornene 0.06 100 0
Heptachloronorbornene isomer 0.06 100 0
Hexachloro-1,3-butadiene 0.16 100 0
Hexachloroethane 4.3 36 64
Isophorone 1.6 98 2
1,1,2-Trichloroethane 0.35 69 31
Trimethylbenzene isomer 0.04 100 0
Triphenyl phosphate 0.12 100 0
Undecane, branched 0.06 100 0
o-Xylene 0.33 100 0
aOnly compounds present in total concentration of
0.04 yg/1 or greater are included.
24
-------
needed to determine the extent to which chlorination of drinking
water forms new chlorinated compounds or increases the
concentration of those already present. The six halogenated
compounds sought were chloroform, bromodichloromethane,
dibromochloromethane, bromoform, carbon tetrachloride, and 1,2-
dichloroethane, several of which are so volatile that they
require special analytical techniques.
Although more extensive analysis of the drinking water from all
80 cities was desirable, the more feasible approach was
determined to be extensive analysis of drinking water from
several cities that are scattered throughout the country and use
water from several different types of sources. Therefore, 10 of
the 80 cities were selected for further analysis. In an attempt
to determine if certain types of source water contain fewer
organic contaminants than other types, the 10 cities were
selected to include 2 cities using water from each of 5
different source types, as follows:
Type ofSource Water
Uncontaminated upland water
Ground water
Water Possibly Contaminated
by Agricultural Runoff
Water Possibly Contaminated
by Industrial Wastes
Water Possibly Contaminated
by Municipal Wastes
Sample Site
Seattle, Washington
New York, New York
Miami, Florida
Tucson, Arizona
Otumwa, Iowa
Grant Forks, North Dakota
Cincinnati, Ohio
Lawrence, Massachusetts
Philadelphia, Pennsylvania
Terrebonne Parish, Louisiana
The survey was coordinated by the EPA's Municipal Environmental
Research Laboratory (formerly the Water Supply Research
Laboratory) in Cincinnati, Ohio. They also analyzed samples
from all 80 cities to determine concentrations of the 6
halogenated hydrocarbons and collected samples from 10 cities
for analyses by the ACB.
At each sampling site, drinking water was passed through carbon
filters for 7 days. Two cylindrical filters (3 in. x 18 in.),
each containing approximately 3UO g of granular activated
carbon, were used to sorb organic contaminants from 6000 & of
25
-------
finished drinking water. To obtain reasonable blank samples,
two filters that were to be used as blanks were placed behind
two filters that had been flushed with 20 £ of water; 20 & of
drinking water were passed through all four filters connected in
series. These samples were shipped to the EPAfs Robert S. Kerr
Environmental Research Laboratory at Ada, Oklahoma, where they
were extracted with chloroform. The extracts were shipped to
the Athens ERL for analysis by GC and GC-MS.
During January and February of 1975, samples were collected from
the first five cities. Analysis of the chloroform extracts of
these samples began in February. Comparison of gas
chromatograms of sample extracts with those of corresponding
blank extracts showed that the sample extracts were contaminated
with organic compounds that apparently were on the carbon before
samples were collected. Contamination of carbon used in the
filters made these first samples useless. More than 11 times as
much organic contamination was found in the carbon used to
filter these water samples than was found in the carbon used for
the New Orleans water samples. Since 26 times as much New
Orleans water was passed through the same size filters,
significantly less organic matter was adsorbed by filters used
to collect samples from these five cities. After a cleaner
carbon was located in sufficient amounts to complete sampling
for the survey, water from the first five cities was resampled
in March 1975. These chloroform extracts arrived at the Athens
ERL in April 1975.
For analysis by GC and GC-MS, the chloroform extracts were
concentrated to 6 ml, making 1 pi of sample extract equivalent
to 1 & of water passed through the carbon filter. Each blank
was concentrated to a volume equivalent to its corresponding
sample extract. A 30 m x 0.4 mm I.D. glass capillary GC column
coated with SP-2100 was used to separate sample components. The
improved separation available with this column had been
demonstrated by comparison of chromatograms obtained with a
packed GC column (SP2100 liquid phase; 10 ft x 1/8 in.) with
those obtained with the capillary column. For example, 182
components were detected in a sample chromatogram from the
capillary column, but only about half that many were detected
when the same sample was analyzed with a packed column. The low
flow rate for the capillary column carrier gas (approximately 2
ml/min) permitted direct coupling to the mass spectrometer
without using an enrichment device. Since some sample is always
lost through an enrichment device, its eliminination from the
system increased the effective sensitivity of the analytical
method. Because all GC eluants entered the mass spectrometer,
the amount of injected sample was limited to 0.4 pi to avoid
filament damage or excessive pressure in the ionizing region.
26
-------
Sample components were identified and measured with the same GC
and GC-MS techniques used for the New Orleans drinking water
samples, except that concentration data were obtained from mass
spectral data rather than GC data. With the computer-controlled
system, mass spectra were acquired continuously as sample
components eluted from the GC. After data acquisition, the
spectra were retrieved from the storage device. Peaks similar
to those in gas chromatograms were obtained by plotting total
ion current versus spectrum number. Computer programs providing
summations of ion current for each peak were used to calculate
the approximate quantities of sample components after background
data were subtracted. Calculations were based on data obtained
from reference compounds analyzed under the same conditions as
the sample. When a particular standard was not available, a
reference compound of similar molecular structure was used to
estimate the sample component concentration.
The reported concentrations (Table 8) are only approximations of
amounts of organic compounds present in the water. Several
factors affect analytical accuracy and precision: instrumental
instability between analysis of the samples and reference
compounds, variation of ionization between reference compounds
and sample components of similar structure, and unknown carbon
sorption/desorption characteristics of sample components.
Because of these factors, a more precise calculation would not
improve the accuracy of the reported concentrations.
Fifty different compounds were tentatively identified in the
drinking water samples; 15 compounds were identified in more
than one sample (Table 8). Thirty-four of these tentative
identifications were confirmed by comparison of mass spectra and
GC retention times of the sample component with those of a
reference standard analyzed under the same conditions.
Calculated concentrations of identified components varied from
<0.01 to 30 pig/1. Of the 100 calculated concentrations, 55 were
<0.09 yg/1 and 34 were between 0.1 and 0.99 yg/1.
The data did not permit evaluation of the different types of
source water; no valid conclusions could be reached as to the
type containing fewer organic contaminants (Table 9). Some of
the identified compounds may have resulted from chlorination,
some may come from agricultural runoff, and some may occur
naturally in water. However, most probably come from municipal
or industrial waste discharges; many of the identified compounds
were previously identified in industrial wastewaters*-*/»-t2.
No polynuclear aromatic hydrocarbons or polychlorinated
biphenyls were among the identified compounds.
Analysis of the samples from the first five cities was completed
in time for the data to be included in a report issued by the
27
-------
Table 8. ORGANIC COMPOUNDS IDENTIFIED IN DRINKING OF 10 U. S. CITIES
00
Conroound Concentration . ua/1
Uncontaminated
Upland Water
Source
Compound Name Seattle New York
Aliphatic Hydrocarbons
1. 2-nethyl-5-ethylheptane
2. n-nonane — 0.02
Halogsnated Aliphatic
Hydrocarbons
3. broraodichloromethane 0.1 1.3
4. bromoformk v — —
5. chlorodibromomethane — 0.4
6. hexachloro-1 ,3,-
butadiene" . -— —
7. hexachloroethane — —
8. isoamylchloride , — 0.01
9. tetrachloroethylene — 0.05
Aromatic and Cyclic
Hydrocarbons
10. t-butyltoluene
11. p_-ethyltoluene b 0.05
12. isopropylbenzene
13. isopropylmethylbenzene
(cymene) isomer — —
14. n-propylbenzene ~~
15. n-propylcyclohexane
16. tetramethylbenzene isomer
Ground Water
Source
Miami Tucson
4.5
1.5 3.0
15 0.01
0.5
0.1 <0.01
0.1
0.05
0.2
0.2
Source Contaminated
by Agricultural Source Contaminated
Runoff by Industrial Wastes
Ottumwa Grand Forks Cincinnati Lawrence
0.6 1.0 0.6
0.1 0.5 0.01
<0.01
0.2 0.1 0.07
0.01
Source Contaminated by
Municipal Waste
Philadelphia Terreborne
Parish
0.01
1.0 2.0
0.- 1.0
0.01
0.01
-------
Table 8. ORGANIC COMPOUNDS IDENTIFIED IN DRINKING OF 10 U. S. CITIES (Continued).
to
vo
Compound Name
Chlorinated Aromatics
17. chlorobenzene b
18. p_-chlorotoluene ,
19. o-dichlorobenzeneb
20. m-dichlorobenzene.
21. g-dichlorobenzene
Aldehydes
22. acetaldehyde
23. 2-ethylbutanal
24. 3-methyl-3-pentanal
25. n-pentanal
26. trichloroacetaldehyde
(chloral)0
Ke tones
27. acetone .
28. acetophenone
29. camphor'3 ,
30. cyclohexanone
31. 2,6-di-t-butyl-p_-
ben zoquinone^
32. di-t-butyl ketone
33. 2-methylhexa-3-ene-2-
one b
34. 2 -pen tanone
35. l,l,3,3-tetrachloro-2-
Dronanoneb
Uncontaminated
Upland Water
Source
Seattle New York
0.1
0.05
3.5 0.02
1.0
0.5
0.07
Ground Water
Source
Miami Tucson
1.0
1.5
1.0
0.5
0.5
0.5
0.1
0.2
Compound Concentration, yg/la
Source Contaminated
by Agricultural Source Contaminated
Runoff by Industrial Wastes
O'ttumwa Grand Forks Cincinnati Lawrence
0.02 — 0.04
1.0
0.5
0.01 2.0
0.1 -- 0.1
0.1
0.02
0.1
0.5
Source Contaminated by
Municipal Waste
Philadelphia
0.1
5.0
1.0
1.0
Terreborne
Parish
0.01
1.0
-------
Table 8. ORGANIC COMPOUNDS IDENTIFIED IN DRINKING OF 10 U. S. CITIES (Continued).
Compound Concentration, yg/la
Compound Name
Herbicides and Pesticides
36. atrazine ,
37. lindane (ct-BHC)
Phthalates
38. di-n-butyl phthalateb
39. diethyl phthalateb
40. di-2-ethyhexyl phthalate
41. di-n-propyl phthalateb
Miscellaneous Compounds
42. diethyl malonate
43. l,4-dioxaneb
44. methylethylmaleimide
45. 6-santalene
46. o-terpineolb
47, tetramethyltetrahydro-
furan isomer .
48. tri-n-butylphosphate
49. trichlorinitromethane
(chloropicrin) b
50. trimethylisocyanurate
Un contaminated
Upland Water
Source
Seattle New York
0.01
0.01 0.01
0.02
0.01
Ground Wate-r
Source
Miami Tucson
5.0
1.0
30
0.5
0.5
Source Contaminated
by Agricultural
Runoff
Ottumwa Grand Forks
0.1
0.1
0.1
0.5
0.5
0.05
Source Contaminated
by Industrial Wastes
Cincinnati Lawrence
0.01
0.1
0.01
0.05
0.5
0.01
0.04
0.08
0.01
Source Contaminated by
Municipal Haste
Philadelphia
0.05
0.01
0.5
Terreborne
Parish
0.02
0.04
Blank spaces indicate that the compound was not detected in that particular sample.
The compound identification was confirmed by comparison of the mass spectrum and GC retention time of the sample
component with those a standard analyzed with the same instrumental conditions.
-------
EPA in June 197513. Data for all 10 samples were reported to
the project coordinator at the EPA«s Municipal Environmental
Research Laboratory in Cincinnati, Ohio, where they were
compiled with results of analyses performed by that laboratory.
These data will be used in future research projects designed to
determine the health significance of the compounds at the levels
found in water, to investigate their sources, and to evaluate
techniques for preventing their presence in water.
DISSEMINATION OF ANALYTICAL INFORMATION
The ACB's analytical expertise was utilized by universities,
colleges, local schools, scientific organizations, industrial
laboratories, a congressman, and state and federal agencies. In
addition to the usual requests for publication reprints,
standard samples, and information about analytical methods, the
ACB responded to a variety of requests for assistance.
Consultations
• Assistance with identification of rat metabolites of
Arochlor 1016, a commercial polychlorinated biphenyl
mixture, was given to scientists from a university
medical school.
• The use of capillary columns for improved GC
separation of complex mixtures of organic compounds
was demonstrated at a Cincinnati, Ohio, EPA
laboratory, where an ACB-prepared capillary column
was installed in a GC-MS system used for pollutant
identifications.
• The EPA's laboratory at Gulf Breeze, Florida,
requested assistance with an analytical method for
nitrilotriacetic acid in seawater; the ACB provided
advice and training in the operation of a
differential pulse polarograph.
• Each of two university seniors, both chemistry
majors, received 4 weeks of training in current
methods of water pollution analysis. ACB personnel
directed their work on identification of organic
compounds in raw and treated paper mill wastewaters
and in municipal wastewater before and after
chlorination.
• Throughout the year, all EPA laboratories with
computerized GC-MS systems were given assistance with
the use of minicomputer programs for data
30.
-------
Table 9. CLASSIFICATION OF IDENTIFIED COMPOUNDS ACCORDING TO
TYPE OF SOURCE WATER
Number of Compounds Identified
Source Type
City #1
City #2
Both Cities
Uncontaminated Upland
Water
Ground Water
Contaminated by
Agricultural Runoff
Contaminated by
Industrial Wastes
Contaminated by
Municipal Wastes
9
22
11
12
10
10
3
9
8
3
3
4
5
32
-------
manipulation and with the mass spectra matching
program that compares sample spectra obtained in
individual laboratories with standard spectra stored
in a central data bankllf.
The mass spectra matching program was used at the
Athens EKL by a scientist from the National
Institute of Environmental Health Sciences, where the
necessary computerized equipment was not available.
Mass spectra obtained with a GC-MS system in the
North Carolina laboratory were brought to the Athens
EKL and were compared with standard spectra stored
in the central data bank.
A college professor who was working on the problem of
prevention of accumulation of s-triazine herbicides
in soils was provided a low resolution electron
impact mass spectrum of a reaction product of
atrazine.
The use of a combined gas chromatography and Fourier
transform infrared spectrophotometry (GC-FTIR)
system, which is still in the developmental stage and
is not readily available in analytical laboratories,
was demonstrated to a pesticide manufacturer. The
GC-FTIR system was used to analyze a complex mixture
of metabolites isolated from food crops (corn and
oranges) that had been treated with a newly developed
pesticide.
At the request of the EPA's office of Program
Integration, the ACB participated in planning and
preparing a thorough critical review of all available
data that pertain to setting effluent guidelines and
standards for mercury and its compounds in air,
water, and solid wastes.
Several reports and research proposals were reviewed
by ACB personnel having expertise in the particular
subjects being addressed. These included a
manuscript on characterization of polychlorinated
biphenyls in the marine environment, a comprehensive
report on the environmental impact of asbestos and
mercury, research proposals in response to the EPA's
stated need for identification and quantification of
organic compounds in surface waters from selected
industrialized areas of the United States, a con-
tractor's plans for electrochemical treatment of
textile dye wastes, and a study of organic pollutants
in industrial effluents and their fate in receiving
waters. The latter project, which was sponsored by
3*3
-------
the National Science Foundation, involved assistance
in planning and reviewing the proposal, contacting
prospective industries, planning sampling sites and
schedules, and reviewing research results and
reports.
At the request of the local school system, ACB
personnel made several educational presentations to
acquaint students with water pollution projects
at the Athens EEL.
The office of a Maryland congressman concerned with
water pollution was given a list of all the organic
compounds that the ACB has identified three or more
times in waters of all types.
Symposia
During FY 75, AGB personnel planned and arranged three
scientific symposia. On May 19-21, 1975, the Fifth Annual
Symposium on Recent Advances in the Analytical Chemistry of
Pollutants was held at Jekyll Island, Georgia. It was sponsored
by the U.S. Environmental Protection Agency, the University of
Georgia, and the Division of Environmental Chemistry and
Analytical Chemistry of the American Chemical Society. The
symposium's purpose was to provide a forum for communication
between environmental analytical chemists and experts in
advanced analytical techniques with environmental applications.
The 225 participants, including 27 from 9 foreign countries,
represented diverse analytical interests and were about evenly
distributed among industry or research institutes, academic
institutions, and government facilities.
An ACB scientist organized a special symposium, The
Identification and Analysis of Organic Pollutants in Water,
which was held in conjunction with the First Chemical Congress
of the North American Continent in Mexico City on November 31-
December 5, 1975. This symposium was sponsored by the American
Chemical Society as well as Canadian and Mexican chemical
organizations. The 48 invited speakers presented results of
recent research activities at various types of scientific
organizations concerned with water pollution. This symposium
promoted international cooperative efforts toward preventing and
controlling water pollution on the North American continent.
A symposium on Trace Analysis was organized and conducted for
the American Society-for Mass Spectrometry. It was held during
the twenty-third annual conference of the Society in Houston,
Texas, during the week of May 26, 1975. Approximately 100
conference participants attended the 12 research presentations.
34
-------
SECTION IV
REFERENCES
1. Keith, L. H. and S. H. Hercules. Environmental
Applications of Advanced Instrumental Analyses: Assistance
Projects, FY 69-71. U. S. Environmental Protection Agency,
Athens, Georgia. Publication Number EPA-R2-73-155. May
1973. 83 p.
2. Alford, A. L. Environmental Applications of Advanced
Instrumental Analyses: Assistance Projects, FY 72. U. S.
Environmental Protection Agency, Athens, Georgia.
Publication Number EPA-660/2-73-013. September 1973. 46
P-
3. Alford, A. L. Environmental Applications of Advanced
Instrumental Analyses: Assistance Projects, FY 73. U. S.
Environmental Protection Agency, Athens, Georgia.
Publication Number EPA-660/2-74-078. August 1974. 31 p.
4. Alford, A. L. Environmental Applications of Advanced
Instrumental Analyses: Assistance Projects, FY 74. U. S.
Environmental Protection Agency, Athens, Georgia.
Publication Number EPA-660/4-75-004. June 1975. 30 p.
5. McGuire, J. M., A. L. Alford, and M. H. Carter. Organic
Pollutant Identification Utilizing Mass Spectrometry. U.
S. Environmental Protection Agency, Athens, Georgia.
Publication Number EPA-R2-73-234. July 1973. 48 p.
6. Webb, R. G., A. W. Garrison, L. H. Keith, and J. M.
McGuire. Current Practices in GC-MS Analysis of Organics
in Water. U. S. Environmental Protection Agency, Athens,
Georgia. Publication Number EPA-R2-73-277. August 1973.
p. 5-13.
7. Junk, G. A., J. J. Richard, M. D. Grieser, D. Witiak, J. L.
Witiak, M. D. Arquillo, R. Vick, H. J. Svec, J. S. Fritz,
and G. V. Calder. Use of Macroreticular Resins in the
Analysis of Water for Trace Organic Contaminants. J.
Chromatogr. 9_9, 745-762, 1974.
35
-------
8. Region VI, Surveillance and Analysis Division. New Orleans
Area water Supply Study. U. S. Environmental Protection
Agency. Publication Number 906/9-75-003. September 1975.
9. Keith, L. H. Chemical Characterization of Industrial
Effluents. (Presented at the 163rd National Meeting of the
American Chemical Society, Division of Water, Air, and
Waste Chemistry, Boston, Massachusetts, April 1972.)
10. Keith, L. H. and J. M. McGuire. Computer-Controlled Mass
Spectral characterization of Industrial Organic Pollutants.
(Presented at the 164th National Meeting of the American
Chemical Society, Division of Water, Air, and Waste
Chemistry, New York, NY, August 1972.)
11. Keith, L. H. Chemical Characterization of Industrial
Wastewaters by Gas Chromatography-Mass Spectrometry. Sci.
Total Environ. 3, 87-102, 1974.
12. Keith, L. H. Analysis of Organic Compounds in Two Kraft
Mill Wastewaters, 0. S. Environmental Protection Agency,
Athens, Georgia. Publication No. EPA-660/4-75-005. June
1975. 99 p.
13. Preliminary Assessment of Suspected Carcinogens in Drinking
Water, Interim Report to Congress. U. S. Environmental
Protection Agency, Washington, DC. June 1975. 33 p.
Appendices I-VI, 214 p.
14. Heller, S. R., J. M. McGuire, and W. L. Budde. Trace
Organics by GC/MS. Environ. Sci. Technol. 9, 210-213,
1975.
36
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/4-77-004
2.
3. RECIPIENT'S ACCESSIOWNO.
4. TITLE AND SUBTITLE
ENVIRONMENTAL APPLICATIONS OF ADVANCED INSTRU-
MENTAL ANALYSES? ASSISTANCE PROJECTS, FY75
5. REPORT DATE
January 1977 Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Ann L. Alford
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Environmental Research Laboratory - Athens, GA
Office of Research and Development
U.S. Environmental Protection Agency
Athens, Georgia 30601
11. CONTRACT/GRANT NO.
16020 GHZ
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Same as above
14. SPONSORING AGENCY CODE
EPA/600/01
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The Analytical Chemistry Branch of the Athens Environmental Research
Laboratory identified and measured aquatic pollutants under eight
projects in response to requests for assistance from other EPA organiza-
tions and other government agencies. In most cases these analyses
helped us to solve, or at least to understand more clearly, the. related
pollution incident, and in some cases the analyses provided evidence
for enforcement of regulatory legislation. Under an additional project,
analytical consultations were held as requested by various organiza-
tions concerned with pollution incidents.
This report was submitted in fulfillment of Project 16020 GHZ by the
Environmental Research Laboratory, Athens, Georgia. Projects discussed
were completed during FY 1975.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Gioup
chemical analysis, water analysis,
organic compounds, trace elements,
mass spectroscopy, gas chromato-
graphy, neutron activation analysis,
contaminants
multielement analysi:
spark source mass
spectrometry, GC-MS
5A
13. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
45
20. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (9-73)
37
u.s. GOVERNMENT PRINTING OfFICE: 1977-757-056/5553 Region No. 5- 1 1
------- |