PB-246 213
SAMPLING AND ANALYSIS OF SELECTED TOXIC SUBSTANCES
TASK II - ETHYLENE DIBROMIDE
John Going, et al
Midwest Research Institute
Prepared for:
Environmental Protection Agency
September 1975
DISTRIBUTED BY:
KTLT
National Technical Information Service
U. S. DEPARTMENT OF COMMERCE
-------
-------
EPA 560/6-75-001
SAMPLING AND ANALYSIS OF SELECTED TOXIC SUBSTANCES
TASK II - ETHYLENE DIBROMIDE
September 1975
FINAL REPORT
U;S. Environmental Protection Agency
Office of Toxic Substances
Washington, D.C. 20460
-------
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
EPA 560/6-75-001
3. Recipient's Accession No.
4. Title and Snhmlr
Sampling and Analysis of Selected Toxic Substances, Task II
Ethylene Dibrotni.de
5. Report Date
Sept., 1975
6.
7. Anilmr(s)
Jolin Going and Sue long
8. IVrforrning Orgnnizmion Hrpt
NtMRI Rep.
9. I Vi lornun;' < >rp,;mi/.u ion N;t'iir :uid Address
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
. Projcct/TiisJf/Vl'ork Unit No.
ask II *-
10
Task
11. Contraci/(irnnt No.
68-01-2646
12. Sponsoring Or^ani/mion Nnnic iinil Address
Environmentatl Protection Agency
Office .of Toxic Substances (WH-557)
401 M St., SW
Washington, DC 20460
13. Type of Report A Period
Covered
14.
15. Supplcmentiiry Notes
16. AKilnicts
Ethylene dibromide has recently been reported by the NCI to be a potential carcinogen.
~ts commercial use is predominately as a scavenging agent for lead in gasoline.
. minor usage is a pesticide fumigant. A protcol was developed for the sampling
and analysis of ethylene dibromide in ambient air and surface water. The range
in concentration of ethylene dibromide in the air and the site where it was •
collected are as follows: manufacturing sites, 4.2 to 115y^g/M3; oil refinery,
0.23-1.65/pcii-l;iid<-d Terms
Ethylene Dibromide
18. Avuilnbility Smtcmeni
Release Unlimited
19. Securii'y Class (This
Report)
UNCLASSIFIED
21. No. of PaRes
20. Security Clnss (This
Hage
UNC1.ASSII-IKO
-------
BACKGROUND STATEMENT
Ethylene dibromide (EDB), which is also known as 1,2-dibro-
moethane, is used principally as an additive in leaded gasoline.
(It is not needed in unleaded gasoline.) Relatively small quanti-
ties of EDB are used as pesticides in soil fumigants (as a nema-
ticide) and in grain and commodity fumigants. Over one hundred
EPA-registered pesticides contain EDB. Small quantities are also
used as an industrial solvent, and as a chemical intermediate.
Five U.S. companies - Dow Chemical, Ethyl, Great Lakes Chemical,
Northwest Industries (Michigan Chemical), and PPG Industries
(Houston Chemical) - produce over 300 million pounds of EDB
annually.
The National Cancer Institute issued a "memorandum of alert"
on ethylene dibromide in October, 1974. When administered directly
into the stomach of rats or mice at the maximum tolerable dose and
at half that level, EDB produced stomach cancer at a high rate and
with a very short latent period. Tumors appeared as early as ten
weeks after the first administration of EDB. This rapid effect is
most unusual and may be unprecedented. Details of these experi-
ments can be found in the Journal of the National Cancer Institute,
51, 1993 (1973). The carcinogenicity of EDB with other routes of
exposure has not been determined at the present time, although
additional experiments are being designed. However, it may be
prudent to consider EDB to be a suspect carcinogen by any route of
exposure.
Besides being carcinogenic, EDB is known to cause pneumonia
and other lung disorders in rats at 50 ppm, and adversely affect
fertility in bulls (4 mg EDB/kg body weight) and in hens (10 ppm
EDB in grain). It has also shown mutagenic properties in micro-
organisms.
-------
Ethylene dibromide is used as a scavenger for lead in leaded
gasoline. The chief source of EDB pollution from automotive
sources is from evaporative emissions from the fuel tank and
carburetor of cars operated on leaded fuel. Emissions from these
sources have been estimated to range from 2 to 25 mg/day for 1972
through 1974 model-year cars. Emissions from pre-1972 cars would
be somewhat higher than these estimates. EDB is not expected to
be emitted' in the exhaust because it is not expected to survive
the combustion process.
Ethylene dichloride is also used as a scavenger along with
EDB. Although ethylene dichloride is cheaper than EDB, it alone
cannot be used satisfactorily because the lead chloride formed is
not sufficiently volatile to keep all parts of the combustion
chamber free of lead deposits. Although ethylene dichloride is
probably less of a hazard than EDB, the health implications of
exposure to it are not fully understood.
The increased use of unleaded gasoline in some 1975 and later
model-year cars, in conjunction with possible regulations to lower
the average lead content of gasolines, should reduce any possible
risks associated with the use of EDB as an additive to gasoline.
The very limited and preliminary air monitoring data set
forth in this report show air concentration values of 0.07-0.11 yg/m
(about 0.01 ppb) in the vicinity of gasoline stations along traf-
fic arteries in three cities (Phoenix, Los Angeles, and Seattle),
3 •
0.2-1.7 yg/m (about 0.1 ppb) on the property of an oil refinery
3
in Kansas City, and 90-115 yg/m (10-15 ppb) at EDB manufacturing
sites in Arkansas. This suggests that EDB is present in ambient
air at very low concentrations. Suitable sampling sites associated
with pesticidal uses of EDB were not identified in this study.
Further work to clarify ambient air levels and to assess whether
ib.
-------
they may represent a health risk is anticipated, but the very low
levels detected to date are not thought to present a significant
risk to health.
Concentrations on the order of 1 ppb of EDB were found in two
samples from streams of water on industrial sites. Although EDB
is not thought to be a serious water pollutant, this matter will
be further investigated. Ethylene dibromide is not believed to
accumulate in the environment. Limited information suggests that
it degrades at moderate rates in both water and soil.
Office of Toxic Substances
Environmental Protection Agency
September 29, 1975
-------
EPA 560/6-75-001
SAMPLING AND ANALYSIS OF SELECTED TOXIC SUBSTANCES
TASK II - ETHYLENE DIBROMIDE
by
John Going
Sue Long
EPA Contract No. 68-01-2646
EPA Project Officer: William A. Coniglio
For
Environmental Protection Agency
Office of Toxic Substances
' 4th and M Streets, S.W.
Washington, D.C. 20460
September 1975
-------
TABLE OF CONTENTS
Page
List of Tables '. lit
List of Figures iii
I. Summary and Conclusions 1
II. Introduction 2
III. Screening and Selection of Sampling Sites 4
A. Manufacturing Sites 4
B. Granaries 4
C. Soil Fumigation 5
D. Urban Areas 6
E. Leaded Gasoline Storage and Bulk Transfer 6
IV. Discussion of Sampling Sites 7
A. Manufacturers 7
B. Granaries 7
C. Soil Fumigation 7
D. Urban Areas 11
E. Leaded Gasoline Storage and Bulk Transfer 18
V. Development of Sampling and Analysis Protocol 20
A. Method Development and Evaluation 20
B. Protocol Adopted 24
VI. Results and Discussion p 26
VII. Recommendations 29
References. 30
11
-------
LIST OF TABLES
Np.
I Ethylene Dlbrdmide Producers In the United States 2
2 Results of EDB Analysis 22
3 Gas Chromatographic Columns and Conditions Column 23
4 Retention Times 27
5 Water Analysis and Retention Times 28
LIST OF FIGURES
No. Page
1 Air Sampling Location - Dow Chemical, Magnolia, Arkansas . . 8
2 Water Sampling Locations - Dow Chemical, Magnolia,
Arkansas 9
3 Air Sampling Location - Ethyl Corporation, Magnolia,
Arkansas 10
4 Air Sampling Site - Phoenix, Arizona 12
5 Air Sampling Location - Phoenix, Arizona ... 13
6 Air Sampling Site - Los Angeles, California 14
7 Air Sampling Location - Los Angeles, California 15
8 Air Sampling Site - Seattle, Washington 16
9 Air Sampling Locations - Seattle, Washington 17
10 Air and Water Sampling Locations - Oil Refinery,
Sugar Creek, Missouri 19
iii
-------
I. SUMMARY AND CONCLUSIONS
A limited sampling and analysis program was conducted to dctermin'r
if ethylene dibromide (EDB) is present in ambient air or surface water. The
protocol for sampling air utilized a sampling medium of Tenax-GC® resin at
dry ice temperature to trap EDB. The sampling train was assembled with a
filter and a drying tube ahead of the collecting medium. A critical orifice
was placed after the sampling train to insure an air flow of 1 liter/min. A
single Tenax-GC® trap was demonstrated to be 947» effective in trapping EDB.
EDB was extracted from the Tenax-GC® with hexane for analysis. Water samples
were extracted twice with hexane using Vaq/Vhexane ratio of 20:1. The hexane
extracts were analyzed for EDB by electron capture gas chromatography. Sev-
eral column packings were utilized.
The three general categories of sampling locations selected for
this study were: manufacturing sites, oil refineries, and heavily trafficked
urban areas. Sampling and analysis revealed 4 to 115 ug/M^ of EDB in ambient
air near the manufacturing sites, and 0.23 to 1.65 ug/M^ at the oil refinery,
near a bulk transfer site. Samples collected in Phoenix, Arizona, Los Angeles,
California, and Seattle, Washington, were found to contain 0.069 to 0.11 ug/M^
of EDB. The relative concentrations in the three general areas were in the
order expected. Water samples taken from small streams below the oil refinery
and at the downstream edge of a manufacturing site contained 1.13 and 1.05 ppb
of EDB, respectively. The stream at the oil refinery is directed into the com-
pany's biodegradation ponds. The stream coming off the manufacturing site is
quite small and nearly stagnant.
-------
II. INTRODUCTION
Ethylene dibromide (1,2-dibrbmoethane) is commercially produced
by reaction of ethylene and bromine. Most of the United States production
sites are located near the brine fields around Magnolia and El Dorado,
Arkansas. The brine, rich in bromide salts, provides a ready supply of
bromine. The known major producers-i are listed in Table 1.
Table 1. ETHYLENE DIBROMIDE PRODUCERS IN THE UNITED STATES
Company Location
Dow Chemical, USA Magnolia, Arkansas
Midland, Michigan'
Ethyl Corporation Magnolia, Arkansas
Great Lakes Chemical Corporation El Dorado, Arkansas
Northwest Industry, Inc. El Dorado, Arkansas
PPG Industry, Inc. Beaumont, Texas
Historically, ethylene dibromide has been used commerically as a
gasoline additive and as a fumigant. Its value as a gasoline additive is
to serve as a scavenger agent for lead released in the combustion process.
The concentration of EDB in gasoline is variable but is on the order of
0.025% (wt/vol).-7
The second use of EDB has been as a fumigant. It has been used
in disinfecting fruits, vegetables, foodgrains. seeds, seedbeds, mills, and
warehouses,2/ It has also been shank injected— into soil to combat nema-
todes in tobacco fields. While normally used by commercial fumigators on
large installations in the United States, it is possible that EDB is used
at the individual farm level—' for fumigating grains in small containers.
The level of application is very dependent upon the nature of the
specific fumigation task. Frequently, EDB is added in a mixture with other
fumigants such as ethylene dichloride, carbon tetrachloride, carbon disul-
fide or methyl bromide.—*— The nature of the mixture effects the applica-
tion rate. Typical of the recommended doses are: 1 to 5 gal. of 7% EDB
-------
per 1,000 bushels stored grain;-'' 1.5 to 3.0 Ib of EDB per 1,000 ft3 of
space; 4.5 to 6 gal. of 83% EDB per acre of tobacco fields.—' Commercially
fumigated grain showed residues of EDB from less than 0.01 to 6.1 ppm.—'
The EDB residue level on grains fumigated in laboratory experiments did
not change significantly in 7 weeks after application.—'
Very little is known about the presence of EDB in ambient air or
surface water. It appears that EDB has not been previously reported to be
present in urban atmospheresZ/ or public drinking waters.— A potential
exists, however, for the introduction of EDB into the atmosphere. The most
probable sources are: evaporative loss from leaded gasoline both from re-
tail stations and automobiles; incomplete combustion; evaporative loss from
leaded gas mixing, storage and bulk transfer locations; granaries or fields
fumigated with EDB containing fumigants; fumigant mixing, storage, and bulk
transfer locations; as well as manufacturing, storage and bulk transfer lo-
cations of EDB. Thus, a potential exposure may exist not only for a few
hundred people at isolated manufacturing sites but also for the entire ur-
ban population.
This monitoring program was designed to answer the following
questions:
1. Is EDB present as an environmental pollutant in either ambient
air or surface water?
2. How great is the exposure potential at critical manufacturing
und high use locations?
-------
III. SCREENING AND SELECTION OF SAMPLING SITES
The task called for the selection of sites where critical exposure
to EDB would be anticipated. The following were specified as sampling sites:
1. Manufacturers - 2 Sites
2. Granaries - 2 Sites
3. Soil fumigation - 1 Site
4. Urban areas - 3 Sites
5. Leaded gasoline storage and - 1 Site
bulk transfer
Ambient air and surface runoff water, if available, were to be collected at
all locations.
The normal factors to be considered in site selection would include,
among others, production volume, degree and method of discharge, control, the
nature of the treatment facilities, the competence and cooperativeness of the
facility management, meteorological factors, population densities and nature
of nearby industrial operations. ; For high priority situations, such as in
this case, all adverse factors effecting a quick response time were heavily
weighted. Such factors included: site and equipment modifications required
prior to sampling; administrative problems encountered in gaining access to
specific sites; and the effect of other externalities, e.g., climatic condi-
tions, on the feasibility of immediate sampling. All sites were selected to
represent the maximum release of toxic materials rather than a representative
nationwide exposure of man and/or the environment.
A. Manufacturing Sites
All known United States producers of EDB are listed in Table 1.—
Two industrial facilities owned by Dow Chemical, USA and Ethyl Corporation,
were selected as sampling locations. The two sites are in close proximity
to each other at Magnolia, Arkansas.
B. Granaries
Since EDB has been used for postharvest fumigation of sorghum,
wheat and corn, an intensive attempt was made to locate a granary using
-------
this method of pest control. The facilities contacted and responses elicited
are listed below.
1. Ed Sorenson, Manager, Farmaro Elevator, Hutchinson, K;ms;is - "No
EDB used in Hutchison."
2. Don Anderson, Manager, Farmers Coop, Wellington, Kansas - "No
EDB used, phostoxin is the present substitute."
3. Manager, Wolcott-Lincoln Elevator, Kansas City, Missouri - "Will
not use EDB."
4. Manager, Van Watters and Rogers Elevator, Salt Lake City, Utah -
"A gas, probably methyl bromide will be used." •
5. Dr. Dewitt, Extension Agent, Iowa State University, Ames, Iowa -
"EDB not used on corn in Iowa."
6. Mildred Mayer, Farmers Coop, Clovis, New Mexico - "EDB is not
used; phostoxin is used."
7. Bob Strader, New Mexico Mill and Elevator, Clovis, New Mexico -
"I have used Dowfume EB-5 in the past but do not expect to do so this year."
8. Mr. Sullens, Industrial Fumigants, Kansas City, Missouri - "EDB
is used generally in spot fumigation of milling equipment but phostoxin is
used for granaries."
9. Research Products, Inc. (Commerical Fumigator), Kansas City,
Missouri - "We do not use EDB."
C. Soil Fumigation
It had been reported that EDB containing fumigants are used for
postharvest fumigation of potato fields in Wisconsin and Idaho, and tobacco
fields in Connecticut. Our investigation, summarized below, did not identify
any significant users of EDB-containing fumigants in these areas.
Our contacts and their responses are listed below.
1. L. K. Binning, University of Wisconsin, Horticulture Department,
To his knowledge, EDB fumigants are not used by the tarping technique on the
Wisconsin potato fields. He believed that a small quantity was used on seed
beds.
-------
2. W. Sparks, University of Idaho, School of Agriculture, To his
knowledge, very little EDB containing fumigant is used in Idaho, due mainly
to the expense. However, when EDB is used, it is injected 1 to 2 in. below
the soil surface.
3. Gordon Taylor, State Department of Agriculture, Windsor Experi-
mental Station, Connecticut, He indicated that Dowfume W-85 has been shank In-
jected into shade tobacco fields at the rate of 4 to 6 gal/acre.
4. Robert Borg, Professional Fumigator, West Suffield, Connecticut,
He believes Dowfume W-85 to be inadequate and he no longer uses it.
D. Urban Areas
Ethylene dibromide is an additive to leaded gasoline at a rate ap-
proximating 0.0257» (wt/vol). Consequently, EDB may be introduced into the
urban atmosphere by evaporative losses from automobiles and' retail gasoline
stations and possibly from incomplete combustion. To explore this possibility,
sampling sites were chosen because they are heavily trafficked, are adjacent
to several retail gasoline stations, and are in downtown areas or other areas
having heavy pedestrian use. The sites were also chosen to represent a wide
range of light and temperature conditions. The sites selected for sampling
were Phoenix, Arizona, Los Angeles, California, and Seattle, Washington.
E. Leaded Gasoline Storage and Bulk Transfer
Generally, lead and related additives are mixed with gasoline at a
refinery site. The gasoline is stored in large tanks and ultimately trans-
ferred into trucks for delivery to retail stations. Loading may be through
the top or through the bottom of the trucks. Loss by volatilization is greater
when trucks are top-loaded. The oil refinery at Sugar Creek, Missouri, was
chosen as a representative storage and transfer facility and two locations in
close proximity to the mixing shed, gasoline storage tanks and the bulk trans-
fer dock were established as the sampling points. •
-------
IV. DISCUSSION OF SAMPLING SITES
A. Manu fa c tu re r s
The Dow production facilities were located approximately 10 miles
west of Magnolia, Arkansas. The air sampling site location, with respect- to
the EDB production facilities, is shown in Figure 1. The site was approxi-
mately 250 ft northwest of the production area and situated on the inner edge
of the access road. The sampling probe was 6 ft high and directed into the
wind as well as toward the facility. Throughout the sampling period the wind
direction was constant (from the southeast) but velocity varied from 0 to 5
mph for the period of 0900 to 1600. Rain occurred from 1600 to 1700, however,
sampling had been terminated at 1430. There was no detectable wind from 1700
to 1930. Temperatures at selected sampling times were 74°F (0900), 90°F (1115),
100°F (1320), 89°F (1600) and 84°F (1900).
Water samples were collected at two sites on the Dow property (Figure
2). The streams, normally slow moving, were flowing rapidly due to a 1.7 in.
rainfall the previous night.
A site map of the Ethyl Corporation facility, located approximately
10 miles south of Magnolia, Arkansas, is shown in Figure 3. The sampling
equipment was placed approximately 100 ft north northeast of the EDB facili-
ties, adjacent to the plant administration building. The sampling probe was
again 6 ft high and directed into the wind. No rain was recorded at this
site, although the wind behavior was identical to that at the Dow site. Tem-
peratures at selected times were 94°F (1045), 80°F (1300), 87°F (1530) and
85°F (1735). The relative humidity at both the Dow and Ethyl sites ranged ,„„..
from 65 to 100%. This variable was not determined at any of the other sites'."1"' ,<•''
No water samples were available.
B. Granaries -• -t't/
. If"
No samples were obtained due to the reasons previously discussed,^**
••••-•-.* ^vis^-' ! ^-f"9V -
• • ...-•••.»?^ i .I,
- • S
C. Soil Fumigation
No samples were obtained due to the reasons previously discussed.
-------
Access Road
N
I
1. EDB Production
2. EDB Storage
3. Tank Car Loading
4. EDB Related
5. Admin. Building
6. Sampling Location
(Velocity)
0-5MPH (0900-1600)
~OMPH (1600^1930)
Figure 1. Air sampling location - Dow Chemical, Magnolia, Arkansas
-------
Air Sampling
Location
N
\
D
Note;
Streams "a" and "b" Merge East of Highway 132
Stream "b".
CN
CO
D)
Stream "a".
Figure 2. Water sampling locations - Dow Chemical, Magnolia, Arkansas
-------
N
X X-
Lab
Office
Main Entrance
X-.
x-x-x
X-X-X
•Sampling
Location
(Velocity) X0'
0-5 MPH (0900-1600)
~OMPH (1600-1930)
100 Feet
1. EDB Process Building
2. EDB Storage Area
3. Tank Car Loading Area
Figure 3. Air sampling location - Ethyl Corporation, Magnolia, Arkansas
10
-------
D. Urban Areas
1. Phoenix, Arizona: A map of Phoenix showing the gener.il loca-
tion of the air sampling site is presented in Figure l\. A more detailed de-
scription of the specific site is given in Figure 5. The. sampler located on
the sidewalk at 1804 Camelback Street was set at a height of 6 ft and directed
into the wind. Camelback is a six-lane street carrying an average traffic
load of 36,000 vehicles per day. Six retail gasoline stations located within
three city blocks of the site sell about 475,000 gal/month. The vicinity may
be characterized as a retail shopping area located in a residential district.
A light wind blew from the east during the period of 0700 to 1200 and shifted
180 degrees continuing from the west during the period of 1200 to 2400. Tem-
peratures during the sampling period were 99°F (0700), 100°F (1200), 119°F
(1520), 118°F (1800) and 100°F (2400).
2. Los Angeles, California; The general location of the air sam-
pling site is shown in Figure 6. A detailed drawing of the actual sampling
location is presented in Figure 7. Sampling equipment was located at the pay
booth of the Kirkaby Center parking lot on Lindbrook Avenue in Westwood Village.
The site was approximately 100 ft east of Westwood Street and was 1 block north
of Wilshire Boulevard. The daily vehicular traffic loads were 26,000 and 53,000
for Westwood and Wilshire Boulevard. Four major retail gasoline stations were
located within approximately 500 ft of the sampler. As before, the probe was
at a height of 6 ft and directed into the wind. A light westerly wind was ob-
served from 0800 to 1400. At this time it shifted and continued in an easterly
flow from 1400 to 2000 hr. Temperatures during the sampling periods were 79°F
(0800), 90°F (1200), 91°F (1600) and 78°F (2000).
3. Seattle, Washington; A map showing the general location of the
air sampling site is found in Figure 8. A detailed drawing of the two sampling
locations established in Seattle is shown in Figure 9. One location was approx-
imately 200 ft south of 45th N.E. on Roosevelt Way, and situated by the sidewalk
on the University Datsun car lot. The second location was approximately 100 ft
north of 45th N.E. on Roosevelt Way and placed by the sidewalk on the University
Chevrolet car lot. In both cases, the probe was set at a height of 6 ft and di-
rected into the wind. The vehicular traffic loads were reported as 32,000 and
9,000/day for 45th N.E. and Roosevelt Way, respectively. Three major retail
gasoline stations are located on 45th N.E., all within approximately 300 ft of
the sampling sites. The wind was light and from the southwest during the sam-
pling. Temperatures during the sampling periods were 58°F (0730), 61°F (1110),
67°F (1500) and 65°F (1900).
11
-------
Figure 4. Air sampling site - Phoenix, Arizona
12
-------
GO
L
Sampling
Location
Wind
Light (0700-1200)
Light (1200-2400)
Retail Gasoline Station
_c
.+_
•o
N
\
Comelback
Figure 5. Air sampling location - Phoenix, Arizona
13
-------
•riv-wo;-;-
Figure 6. Air sampling site - Los Angeles, California
14
-------
O
Sampling-'
Location
Wind —* Light (0800 -1400)
•••—Light (1400-2000)
H Retail Gasoline Station
N
\
f ->
Westwood
Lindbrook
Wilshire Blvd
Figure 7. Air sampling location - Los Angeles, California
15
-------
"•••., ."11,1 \ •w.'rtK-««-!'-
Figure 8. Air sampling site - Seattle, Washington
16
-------
Sampling
Location a
Bank
Parking Lot
45th N.E.
v
9)
vt
8
Sampling Location b
N
Light (0710-1930)
Retail Gasoline Station
Figure 9. Air sampling locations - Seattle, Washington
17
-------
E. Leaded Gasoline Storage and Bulk Transfer
A map depicting the location of the two air sampling locations arid
the stream used for water sampling appears in Figure 10. Air sampling loca-
tion "a" was approximately 50 ft south of the tank truck loading facilities
at which loading is done through the top with an open stream. Location "b_"
was at the "process laboratory," 400 ft north of the bulk transfer facilities
and was used only during July 31 to August 1 sampling. During the entire sam-
pling period the wind was from the south at 5 to 10 mph. Temperatures during
the sampling period were, on July 31, 77°F (0800), 90°F (1200), 94°F (1600)-
and 88°F (2000), and on August 1, 72°F (0800). ;'
During repeat sampling on August 15, the wind was from the north
at 5 to 10 mph. Temperatures during the sampling period were 80°F (0800),
82°F (1200) and 85°F (1600). The gasoline transfer facility was open and
used 24 hr/day.
The water sample was taken from a small stream, designated as "e?y
that surfaced south of the bulk loading station, and was then covered and '
channelled into the company's biodegradation ponds.
18
-------
r-'ip,}{^?f
1
>\\" rt / -./
\\ / /-A. > . """^ -•
6:,) / A^XX/"
Wind (Velocity) N
7/31 - 8/1 f (0925-1000)
8/15 | (0945-1545)
S
Figure 10. Air and water sampling locations -
oil refinery, Sugar Creek, Missouri
19
-------
V. DEVELOPMENT OF SAMPLING AND ANALYSIS PROTOCOL
A. Method Development and Evaluation
1. Air: Most of the published literature on the sampling and
analysis of EDB has been related to its use as a fumigant. Interstitial
air concentrations in closed bins or silos have been determined by analyz-
ing gas samples with flame ionization gas chromatography.Z/ The limit of
detection by this technique has been reported as 2 ug/liter (0..24 ppm) .
Based on a reported sensitivity of 50 pgZ/ for electron capture detection
gas chromatography, the calculated limit of detection by direct gas sam-
pling would be 50 ng/liter (0.006 ppm). In this study it was anticipated
that in most instances the EDB concentrations in ambient air samples would
be considerable lower than either of the above detection limits.
Therefore, an analytical procedure was required that concentrated
the EDB present in a relatively large volume of air. Sampling periods of up
to 24 hr were anticipated since neither the actual source of emissions nor
the emission patterns were known or predictable.
The initial approach to sampling air for EDB was to use a Tenax-GC®
trap at ambient temperature. Experiments run under laboratory conditions in-
dicated that good recovery of EDB should be possible. A stream of helium at
200 ml/min for 5 hr was used to evaporate and carry 500 ng of EDB into a
Tenax-GC® trap at ambient temperature. Recovery was 84%. Therefore, Tenax-G(
at ambient temperature was used to sample the oil refinery. Two Tenax-GC®
traps were connected in series and air was drawn through at 1 liter/rain.
EDB was found on both Tenax-GC® traps. Under field conditions EDB either
was not stopped or was migrating through the tubes. Additional laboratory
experiments indicated that EDB migrated through the columns. A glass tube
containing 2,000 ng EDB in hexane was connected to two Tenax-GC® traps (T-l
and T-2) in series. Helium was passed through at 1 liter/min. After 2 hr,
the second trap, T-2, was replaced by T-3. Two hours later, T-3 was replaced
by T-4. T-4 was left in place for 4 hr. All the tubes were extracted and
analyzed for EDB.
Tube EDB
T-l 160 ng
T-2 568 ng
T-3 645 ng
T-4 370 ng
While about 60 to 70% of the EDB was originally stopped by T-l,
under the constant flow of helium, EDB migrates down and out of the tube.
Thus, it was concluded that at ambient temperatures Tenax-GC® was ineffective
for long term air sampling of EDB.
20
-------
To be certain that EDB would be retained by the Tenax-GC® traps,
the tubes were cooled with dry ice. This modified cryogenic technique has
been used to trap vinyl chloride on Tenax-GC®.—' Silica gel drying tubes
were inserted ahead of the Tenax-GC® trap to remove water vapor and avoid
the formation of ice plugs. The efficiency of the air sampling train shown
in the air sampling protocol was evaluated by the following experiment. A
glass tube containing 500 ng EDB was connected to the drying tube and two
Tenax-GC® tubes were cooled by dry ice. Helium was passed through at a rate
of 750 ml/min for 4 hr. The tubes and their contents were then extracted
with hexane and analyzed for EDB by the procedure described in the analytical
protocol. The results of duplicate runs are as follows:
Run 1 Run 2 Average
Percent EDB recovered by first extraction 73 79 76
Percent EDB recovered by second extraction 21 14 18
Percent EDB retained by drying tube 000
Percent EDB retained by first Tenax-GC® tube 94 93 94
Percent EDB retained by second Tenax-GC® tube 0 0 0
The first Tenax-GC® trap was extracted twice in this experiment.
In analyses of field samples, only one extraction was used and therefore a
recovery factor of 1/0.76 or 1.3 has been applied to the results. As seen
in the above results, EDB was not retained by the drying tube and did not
pass through the first Tenax-GC® tube. The procedure works well in moderate
humidity ranges. With high humidity, the drying tubes can be readily satu-
rated and will require periodic changing.
The hexane extracts from the first sampling of the oil refinery
(July 31 to August 1) were reanalyzed 6 days after the initial analysis. A
considerable decrease in the amount of EDB present was observed (see Table
2). These samples had been stored in a refrigerator. Since degradation was
rapid, all subsequent samples were stored in a freezer at -13°C. This pro-
cedure appeared to stabilize the samples. The analysis of hexane extracts
containing EDB was done by gas chromatography using an electron capture de-
tector. The electron capture detector is extremely sensitive to EDB and can
detect as little as 1 to 2 pg/injection. Calibration plots were found to be
linear from 20 to 1,000 pg/injection. All samples were first chromatographed
on Carbowax 20M according to the conditions listed as Column 1 in Table 3.
Additional columns were used for confirmation of identity. Chromatographic
conditions are also listed in Table 3.
21
-------
Table 2. RESULTS OF EDB ANALYSIS
(Air Analysis)
Site location
Oil Refinery
a. Bulk loading
b. Process lab
Oil Refinery
a. Bulk loading
Phoenix
to
to. Los Angeles
Seattle
a. N of 45th
street
b. S of 45th
street
Dow
Ethyl Corporation
Sampling time;
temperature
Date tnin-max
7/31-
3/1 a.
b.
8/15
a.
8/5 a.
8/7 a.
8/20
a.
b.
8/26 a.
b.
8/26 a.
b.
c.
d.
e.
0925-1000
0945-0945
72-94°F
0945-1545
80-85°F
0700-2400
0800-2000
78-91°F
0730-1930
0710-1945
60-67 °F
0900-1330
1330-1430
1045-1530
1530-1730b-/
1730-1845
1630-1730
1730-1830
74-100°F
Volume
sampled Filter
Ct) (ng)
1,364
1,318
174
765
641
715
705
116
43.8
113.8
80.8
46.9
44.4
43.5
207
220
NDS./
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Drying Tenax, ng Tenax, ng
tube ambient dry Ice EDB EDB—
(nR) 1st 2nd 1st 2nd ug/m3
710 780 - - 1.09
"-- -/ (180)S/ (120)S/ (0.22)£/
- ; 725 816 - - 1.27
(21)S/ (10)S./ (0.024)3./
ND - - 31 ND 0.18
ND - - 41 ND 0.053
ND - - 55 ND 0.086
ND - - - 45 ND 0.063
ND - - 45 ND 0.064
ND - - 9,400 ND 81.0
3,750 115 . - 88.2
ND - - 7,900 ND 69.4
ND - T 260 ND 3.2
ND - - 330 ND 7.0
360 ND - - 8.1
5"15 ND - - 11.8
ppb
0.131
(0.026)3.
Ug/m3 -
1.42
/ (0.29)-'
0.152 1.65
(0.0029)£/(0.031)S/
0.022
0.0064
0.010
0.0076
0.0077
9.72
10.6
8.32
0.38
0.84
0.97
1.42
0.23
0.069
0.11
0.082
0.083
105.0
115.0
90.2
4.2
9.1
10.5
15.3
ppb
0.170
(0.034>i/
0.20
(Q.0038)^
0.029
0.0083
0.013
0.0099
0.010
12.6
13.8
10.8
0.49
1.09
1.26
1.85
a/ Second analysis, 6 days later.
b/ Rain In area - 1600-1700
Wind - SE 2-5 nph 0900-1700, than - 0
£/ Corrected by multiplying by 1.3 to account for actual recovery of 76%.
ND = Not detected.
-------
Table 3. GAS CHROMATOGRAPHIC COLUMNS AND CONDITIONS COLUMN
NJ
10
Packing
Column Dimensions
Column Temperature
Injector Temperature
Detector Temperature
Flow Rate, He
Retention Time
I-
57, Carbowax 20 M
on 80/90 Anakron
6 ft x 1/8 in.
stainless steel
115°C
205°C
210°C
32 ml/min
4.1 min
2
57, QF-1 on 100/120
Sulpelcoport
6 ft x 1/4 in.
stainless steel
75°C
205°C
210°C
32 ml/min
4.0 min
3
57, DECS on 80/100
Chromosorb W
6 ft x 1/8 in.
stainless steel
80°C
205°C
210°C
20 ml/min
2.4 min
4
17, OV-101 on 100/120
Chromosorb G-HP
6 ft x 1/8 in.
stainless steel
70°C
205°C
210°C
20 ml/min
1.38 min
-------
2. Water; No reports of EDB being present in surface waters were
found in the literature. Several procedures for multiresidue analyses of EDB
on grains were expected to be satisfactory for analyzing water. In one pro-
cedure^.' grains coated with EDB were added to water, toluene was added, and
the EDB was quantitatively removed by steam distillation. Alternately, EDB
can be removed by a nitrogen sparge of a boiling solution..§/ In this proce-
dure water present in the gas stream was removed by a drying trap of Chromo-
sorb W. The EDB was then trapped in isooctane held at -80°C.
A simpler procedure than the above was desired. It was determined
that water samples containing EDB could be quantitatively extracted by hexane
using two extractions at a Vaq/Vorg ratio of 20:1. Five-hundred milliliters
of water containing 5 ppb EDB was subjected to sequential extractions; re-
coveries of 82 and 20% were obtained, yielding a total recovery of 1027o.
If necessary, the extracts can be reduced in volume using a Kuderna-
Danish evaporator. Two 500 ml water samples containing 0.1 ppb EDB were ex-
tracted with two 25-ml portions of hexane. The volume of the hexane was re-
duced to 5.0 ml by the evaporator. Recovery was 47 and 52%.
Attempts to reduce the volume below 5 ml by a slow N2 stream lead to
excessive loss of EDB.
The hexane extracts were analyzed for EDB by the same procedure used
for the hexane extracts of the Tenax-Gd®.
B. Protocol Adopted
1. Air: The sampling train is shown below:
• Critical Orif ii <•
a
_^ I „
i—i >r~ i—
\ ^i\7 v
\ \/ A
A~{ J * Vacuum i'nmp
c n • T T L Dry Ice Chest
Filter Drying Tenax Tubes
Tube
The filter was a millipore, Type A with a mean pore size of 8 u. The drying
and Tenax-GC® tubes were constructed of glass 15 cm long, 12 cm O.D., 10 cm
I.D., and were packed to within 2 to 3 cm of each end. Glass wool previously
extracted with hexane was inserted at each end. Minimum lengths of rubber
tubing were used to connect the filter and collection tubes. The tubing
..should be replaced after each use.
i- .
24
-------
Flow rate was controlled by a critical orifice needle at 1 liter/
tnln» Flow rates were checked every 1 to 2 hr. The inlet was positioned at
a height of 5 to 6 ft and directed into the wind. All tubes and the filter
were stored over dry ice after use. The filter, Tenax-GC@ and silica gel
were extracted with 2.0, 20, and 10 ml portions of pesticide grade hexane
using an ultrasonic bath. After extraction, the solutions were stored in
a freezer until analyzed.
2. Water; Water samples were first filtered if necessary to re-
move particulate material. The samples were then extracted twice with pest-
icide grade hexane at a Vaq/Vhexane rat*° °f 20. The hexane extracts were
combined and diluted to 50.0 ml. The extracts were stored in a freezer until
analyzed. If necessary the volume can be reduced to 5.0 ml by use of a Kuderna-
Danish evaporator.
25
-------
VI. RESULTS AND DISCUSSION
Table II summarizes the results obtained for the .'ilr samples col-
lected at the sites discussed previously. The oil refinery site was sam-
pled on two separate occasions; the first time using Tenax-GC® at ambient
temperature, the second time using the adopted protocol. The hexane ex-
tracts held above +4°C showed a time related drop in concentration.
Samples at Dow and Ethyl were collected by both Tenax-GC® traps
at ambient and dry ice temperature. The air was sampled for only 1 hr
through the ambient temperature Tenax-GcM tubes in order to minimize loss
through the tubes. Backup tubes were used to evaluate the efficiency of
the first tube. The results obtained at Dow and Ethyl indicated that
Tenax-GC® at ambient temperature may be useful for short sampling periods.
Breakthrough was only 3% at Dow and EDB was not detected in the second
Tenax-GC® tubes at Ethyl. .
The results obtained at ambient and dry ice temperatures at both
sites agree quite well. For example, comparable samples collected at Dow
were found to contain 105 and 115 ug/M->. Similar results at Ethyl showed
9.1 and 15.3 ug/M3 at dry ice and ambient temperatures, respectively.
The results obtained at Ethyl for the period from 1045 to 1530
hr are similar to those obtained at Dow. At 1600 hr, rain occurred at the
Dow site and the wind ceased at both sites. As would be expected, the am-
bient air levels at Ethyl dropped accordingly. The sampling at Dow had
been terminated prior to the rainfall.
The second cold Tenax-GC® trap served as a check of the effi-
ciency of the first trap and as a system blank. EDB was not detected in
any of these samples, thus verifying the purity of the hexane and Tenax-GC®
resin.
Table IV lists the columns and experimental retention times. The
chiromatographic peaks observed in the field samples are reported as relative
retention values based on the ratio of the retention time of each peak to
the retention time of EDB standards. The .standards were chromatographed
before and after the field samples. The consistency of the ratios provides
strong evidence of the presence of EDB in the samples. It was not possible
in all cases to obtain second and third column confirmation because of in-
terference of large unidentified peaks from the sample.
Table V summarizes the results obtained from the water samples
and lists the retention times where appropriate.
26
-------
Table 4. RETENTION TIMES
Sites
Oil Refinery
a. 7/31-8/1
b. 7/31-8/1
a. 8/15
Ehoenix
Los Angeles
Seattle
N
S
Dow - .
a.
b.
Ethyl
a.
b.
c.
d.
e.
tR
?! 60^
4.80
4.40
4.33
5.43
5.51
4.13
4.13
4.13
4.13
4.13
4.13-
4.13
1
Carbowax
CR (STD)
7.57-s/
4.72
4.33
4.33
5.47
5.47
4.13
4.13
4.13
4.13
4.13
4.13
4.13
2 3
20M QF-1 DECS
fcR/cR (STD) tR. CR (STD) CR/tR (STD) CR CR (STD) tR/tR (STD) CR
.1.00 4.18 4.25 0.98 Not used Not used
1.00
1.02
1.02 4.06 4.00 1.01 -. Not used Not used
1.00 4.10 4.00 1.02 Not used Not used
0.99 - - - Not used Not Used
1.01 - -
1.00 Not used 2.44 2.44 1.00 1.38
1.00 2.44 2.44 1.00 1.38
1.00 Not used 2.40 2.44 0.98 1.38
1.00
1.00
1.00
1.00
4
OV-101
CR (STD) tR/tR (STD)
1.38 1.00
1.38 1.00
1.38 1.00
a/ Measured on a 6 ft x 4 mm I.D. glass column packed with 47. Carbowax 20M +.C-870 KOH.on
Carbopak at 115°C, 30 ml/min. All others at the conditions given in Table 3.
R = Retention time in minutes.
-------
Table 5. WATER ANALYSIS^ AND RETENTION TIMES
Carbowax 20 M
Site Location
Oil Refinery^
£.
Dow
a.
b.
Date Volume Analyzed EDB (vig)
7/31 1.0 liter 1.13
8/26 . 1.0 liter 1.05
8/26 1.0 liter 0
EDB (ppb) tR tR(STD)
1.13 7.58 7.57
1.05 4.13 4.13
0 00
tR/tR(STD)
1.00
1.00
0
Note: tR = Retention time in minutes
noff water not available at oth
e sample came from a small cree
the company's biodegradation ponds.
a/ Runoff water not available at other sites.
10 —
00 J)/ The sample came from a small creek flowing past the bulk loading station. This water is directed into.
-------
VII. RECOMMENDATIONS
This study represents only a preliminary invest i).;;ition of the
problem. The imposed time limit did not allow for a comprehensive study.
Naturally, many questions are as yet unanswered and should be considered
in future programs. For example, temperature, humidity, diurnal, seasonal
and geographical effects are not known. The actual source of EDB in the
urban areas is not known. The source may be evaporative losses from retail
gasoline stations or automobiles, incomplete combustion or combinations of
all three factors. The points of emission at the manufacturing sites are,
likewise, not known.
EDB levels in the air or water in communities near manufactur-
ing sites should be determined. Apparently, 127° of the EDB produced in the
United States is used as a fumigant. This study has not uncovered signif-
icant uses of EDB fumigants in the United States or significant export data.
It should also be determined where and how the fumigants are mixed, stored
and transferred for shipment.
29
-------
REFERENCES
1. Directory of Chemical Producers, Chemical Information Service, Stanford
Research Institute, Menlo Park, California 94025 (1975).
2. Private communication, George Weiss, Sugar Creek, Missouri.
3. Girish, G. K., R. K. Goyal, and K. Krishanamurthy, Bull. Grain Tech.,
1£:120 (1973).
4. Private communication, Gordon Taylor, State Department of Agriculture,
Windsor Experimental Station, Connecticut.
5. Malone, B., J. Assoc. Offic. Agr. Chem., 5j2:800 (1969).
6. Malone, B., J. Assoc. Offic. Agr. Chem.. 53j742 (1970).
7. Berck, B., J. Agr. Food Chem.. 2^:977 (1974).
8. Olson, W. A., R. T. Haberman, E. K. Weisburger, J. W. Ward, and J. H.
Weisburger. J. Nat. Cancer Inst.. 5J_(6) : 1993 (1973).
9. Dimitriades, B., C. F. Ellis, and D. E. Seizinger, "Atmospheric Analysis
by Gas Chromatograph," Advances in Chromatography, Vol. 8, Marcel
Dekker, New York, New York (1969).
10. Dowty, B. J., D. R. Carlisle, and J. L. Laseter, Environ. Sci. Technol..
9:762 (1975).
11. Ives, N. F., J. Assoc. Offic. Agr. Chem.. 58:457 (1975).
30
------- |