Battelle
Columbus laboratories
Report
ENVIRONMENTAL MONITORING
NEAR INDUSTRIAL SITES
ACRYLAMIDE, ORGANOTINS, ARYL PHOSPHATES
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF TOXIC SUBSTANCES
December, 1977
-------
RESEARCH REPORT
on
ENVIRONMENTAL MONITORING
NEAR INDUSTRIAL SITES
ACRYLAMIDE, ORGANOTINS, ARYL PHOSPHATES
to
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF TOXIC SUBSTANCES
December, 1977
by
G. W. Kinzer, P. P. Kelly, B. E. Sherwood,
A. Graffeo, C. J. Riggle, D. L. Sgnotz,
and P. A. Clarke
Contract No. 68-01-1983
. BATTELLE
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
-------
TABLE OF CONTENTS
PART I. ACRYLAMIDE 1
Introduction 1
Literature Review 2
Gas Chromatography-Electron Capture 2
Ultraviolet Detection 4
Polarographic 4
Wet Chemical 4
Analytical Methods Development 4
Detection Equipment and Column Detection 5
Equipment Optimization 6
Acrylamide Identity Confirmation 8
Direct Injection Methods Development 8
Concentration Methodology Development 12
Internal Standard Selection 15
Collection and Analysis of Environmental Samples 16
Water 16
Air 20
Soil 21
GC Analysis and MS Confirmation 21
Recommendations 23
PART II. ORGANOTINS 25
Introduction 25
Literature Review 25
Analytical Methods Development 27
Recovery Studies During the Sample Concentration
Procedure 32
Speciation of Organotins by Thin Layer Chromatography . 36
Collection and Analysis of Environmental Samples 38
Reading Site 38
Avondale Site 40
Conclusions and Recommendations 46
PART III. ARYL PHOSPHATES 49
Introduction 49
Literature Review 49
Hydrolysis 49
Paper and Thin Layer Chromatography 50
Gas Chromatography 50
iii
-------
TABLE OF CONTENTS (Continued)
Page
Analytical Methods Development 50
Calibration of Gas Chromatograph 51
Extraction Recoveries 53
Collection and Analysis of Environmental Samples 54
Water 54
Soil 54
Air 55
Results 55
REFERENCES 65
APPENDIX - AMBIENT AIR SAMPLING SYSTEM A-l
FIGURES
Number Page
1 GC response at detection limit for acrylamide in
methanol: 2 y£ of 1 ppm solution 7
2 GC chromatogram of acrylamide in an environmental sample 9
3 Reconstructed gas chromatogram of acrylamide in an
environmental water sample from Finnigan 3200 GC-MS. . 9
4 Mass spectrum of acrylamide from confirmation study of
an environmental water sample 10
5 Standard response curve of Hewlett Packard 7670A GC
with nitrogen detector to acrylamide in methanol ... 11
6 Map of sampling sites for acrylamide at Midland,
Michigan 17
7 Topographic map of sampling sites for acrylamide at
Midland, Michigan 18
8 Concentration and subsequent elution of organotins ... 28
9 Bis(tri-n-butyltin) oxide standard curve 30
10 Methyltin trichloride standard curve 30
11 Atomic absorption response to bis(tri-n-butyltin) oxide. 31
12 Apparatus for adsorbing volatile organotins 35
13 Map of sampling sites for organotins at Reading, Ohio. . 39
14 Topographic map of sampling sites for organotins at
Reading, Ohio 40
iv
-------
FIGURES (Continued)
Number Page
15 Map of sampling sites for organotins at Avondale
Shipyard at New Orleans, Louisiana 43
16 Topographic map of sampling sites for organotins at
Avondale Shipard, New Orleans, Louisiana 44
17 Gas chromatography trace of standard triaryl phosphate
mixture 52
18 Map of sampling sites for aryl phosphates at
Gallipolis Ferry, West Virginia 58
19 Topographic map of sampling sites for aryl phosphates at
Gallipolis Ferry, West Virginia 59
20 Map of sampling sites for aryl phosphates at Lordstown,
Ohio 62
21 Topographic map of sampling sites for aryl phosphates
at Lordstown, Ohio 63
22 Aerial map of sampling sites for aryl phosphates at
Lordstown, Ohio 64
TABLES
1 Acrylamide Analytical Methods 3
2 Acrylamide Sample Pretreatment Methods 19
3 Acrylamide Analytical Data for Water Samples from
Midland, Michigan 20
4 Acrylamide Analytical Data for Air Samples from
Midland, Michigan 22
5 Acrylamide Analytical Data for Soil Samples from
Midland, Michigan 23
6 Separation of Organotin Classes by Thin Layer
Chromatography 29
7 Reproducibility of Atomic Absorption Response for
Methyltin Trichloride at Different Concentrations. . . 31
8 Recovery Study of Methyltin Trichloride from Water ... 33
-------
TABLES (Continued)
Number Page
9 Recovery Study of Bis(tri-n-butyltin) Oxide from Water . 33
10 Recoveries of Bis(tri-n-butyltin) Oxide from Soil. ... 34
11 Rf Values of Organotins on Thin Layer Chromatography
Silica Gel Plates 37
12 Organotin Analytical Data for Water Samples from
Reading, Ohio 38
13 Organotin Analytical Data for Air Samples from
Reading, Ohio 42
14 Organotin Analytical Data for Soil Samples from
Reading, Ohio 45
15 Organotin Analytical Data for Water Samples from
Avondale, Louisiana 46
16 Organotin Analytical Data for Air Samples from
Avondale, Louisiana 47
17 Organotin Analytical Data for Soil Samples from
Avondale, Louisiana 46
18 Retention Time Data for Aryl Phosphates 51
19 Extraction Recoveries for Aryl Phosphates 53
20 Aryl Phosphate Analytical Data for Water Samples from
Gallipolis Ferry, West Virginia 56
21 Aryl Phosphate Analytical Data for Soil Samples from
Gallipolis Ferry, West Virginia 56
22 Aryl Phosphate Analytical Data for Air Samples from
Gallipolis Ferry, West Virginia 57
23 Aryl Phosphate Analytical Data for Water Samples from
Lords town, Ohio 60
24 Aryl Phosphate Analytical Data for Soil Samples from
Lords town, Ohio 60
25 Aryl Phosphate Analytical Data for Air Samples from
Lordstown, Ohio 61
vi
-------
EXECUTIVE SUMMARY
This document describes the development and application of techniques
for the analysis of acrylamide, organotins, and aryl phosphates in environ-
mental samples.
The analytical technique developed for acrylamide involves a sample
pretreatment consisting of concentration of aqueous solutions or methanol
extracts by direct evaporation and gas chromatography with a nitrogen
detector. Acrylamide detection levels were 1 ppb in air, 25 ppb in water,
and 20 ppb in soil. Air (vapor and particulate), water, and soil samples
were collected in the vicinity of an acrylamide plant and analyzed.
Acrylamide was detected in only one sample, a water sample taken at an
outfall from the plant; the concentration of acrylamide in this particular
sample was in the range of 25 to 125 ppb.
The method developed for organotin compounds comprised three steps:
(1) concentration of organotin compounds by dynamic adsorption on silica
gel, (2) speciation by major organotin compounds of commercial interest by
chromatographic methods, and (3) quantification by atomic absorption
spectroscopy. Detection levels of tin compounds in environmental samples
by the method in terms of tin were: air, 0.1 yg/m3; water, 1 ppb; soil,
10 ppb. Environmental samples were collected at two sites: a production
site and a user site (shipyard). No organotin was detected in air from
the two sites at the detection levels of the method. Tin in soil samples
from the production site ranged from 15 to 120 ppb and in water samples
from less than 1 to 4 ppb. The concentration in soil samples from the user
site was less than 10 ppb and in water less than 1 ppb.
The analytical method developed for aryl phosphates is based on
capillary gas chromatography utilizing a flame photometric detector specific
for phosphorus. Reference samples selected for this study were: tricresyl
phosphate, triphenyl phosphate, cresyl diphenyl phosphate, isopropylphenyl,
diphenyl phosphate, and 2-ethylhexyl diphenyl phosphate. The detection
limits of the method based on triphenyl phosphate were 10 ppb in water, 100
ppb in soil, and 0.2 ppb in air. Air, water, and soil samples were collected
in the vicinity of two industrial facilities: a manufacturer of aryl phos-
phates and an industrial plant which makes extensive use of hydraulic
fluids. None of the selected aryl phosphates were found in the vicinity of
these two facilities in quantities exceeding the detection limits of the
analytical method.
vii
-------
PART I. ACRYLAMIDE
INTRODUCTION
Acrylamide has gained wide use in its polymerized form as a water-
proofing agent in construction, an antierosion agent for the farming
industry and, in some areas, a flocculent for water treatment. Typically,
it is the monomer which is applied to building foundations or fields, with
a catalyst to initiate polymerization. It is the residual monomer which
is of concern here because of its reported toxicity. The acute oral 1059
in rodents for acrylamide is approximately 200 mg/kg (Norris, 1967).
However, the compound possesses a high degree of cumulative toxicity. It
is known, for example, that 15 mg/kg/day for 22 days will cause peripheral
neurotoxicity in Beagle dogs (Thomann et al., 1974). Furthermore, acrylo-
nitrile, which is structurally similar and a pyrolysis or catalyzed
dehydration product of acrylamide, is currently being investigated as a
possible carcinogen (Chemical & Engineering News, 1977).
Acrylamide is very water soluble and could be expected to accumulate
in water supplies by runoff or direct treatment. The acceptable limit of
its concentration in drinking water is discussed in the following appraisal
of a British research group (Croll et al., 1974).
"The high chronic toxicity of acrylamide [D. D. McCollister,
F. Oyen, and V. K. Rowe, Toxic. Appl. Pharmac., j6, 172
(1964)] makes it an undesirable contaminant of potable water
supplies. In 1966 the then Ministry of Housing and Local
Government set up a standing committee on New Chemicals for
Water Treatment to advise water undertakings in the U.K. on
the suitability of new water treatment chemicals, including
acrylamide polymers and co-polymers, for use in potable
water treatment. Typically an acrylamide polymer or co-
polymer is considered unobjectionable for this use if no
batch contains more than 0.5 percent acrylamide and the
average polymer dose to the water does not exceed 0.5 mg
I"1. The maximum dose must never exceed 1.0 mg 1 [Housing
and Local Government Ministry, Committee on New Chemicals
for Water Treatment, Wat. Treat. Exam., 18, 90 (1969)].
These figures correspond to a highest acceptable average
level of acrylamide in water of 0.25 ug 1~^ (0.00025 mg 1-1)
and a maximum acceptable short term level of 0.5 ug 1 ."
Considering the cumulative toxicity of acrylamide and dose limits of
acrylamide in water treatment, an acrylamide analysis method should display
-------
a detection limit into the low ppb range to be generally useful for analysis
of environmental levels.
One of the analyses referenced in Table 1 displays such a limitthe
bromination/gas chromatographic analysis of Croll. Because the bromination
step is both time-consuming and difficult, this study proceeded toward de-
velopment of a direct injection method using a gas chromatograph, but with a
nitrogen detector instead of electron capture detector as used by Croll.
Primary considerations for this approach were the availability and simplicity
of the instrumentation, no chemical pretreatment would be necessary, and the
method would allow for direct injections of aqueous solutions. Another
apparent advantage of the nitrogen detector system was the decrease in
potential interference from hydrocarbons since the nitrogen-detector is
relatively insensitive to compounds which do not contain nitrogen. Finally,
it was thought that the sensitivity of the nitrogen detector to untreated
acrylamide samples would, like electron capture detector after bromination
treatment, be much greater than that of a flame-ionization detector. When
the method was found to reveal only ppm levels, a concentration step was
added.
Gas chromatography-mass spectroscopy (GC-MS) was employed for confirma-
tion of the acrylamide in selected samples. Like the GC-nitrogen detector
system no chemical pretreatment of the sample is required and direct injec-
tion of aqueous samples can be performed. No attempt was made to develop a
GC-MS method for acrylamide because the large sample load would have
required a dedicated instrument which was not available for this program.
Moreover, the instrumentation would be a considerably greater financial
investment for a monitoring laboratory.
LITERATURE REVIEW
An extensive literature search, including an automated search via NTIS
(National Technical Information Service), was performed at the initiation of
the program. The results are summarized in Table 1 and each method is
critically evaluated below.
Gas Chromatography-Electron Capture
Of the acrylamide detection procedures reported in the literature, only
one, the electron capture gas chromatographic method, displays the low ppb
sensitivity that would be required in trace environmental analyses. The
final detection methodology is potentially applicable to the trace environ-
mental analyses. However, the extensive bromination-extraction-concentration
pretreatment procedure required is a disadvantage. The work-up involved
would require one-man day or more to process at most 5 or 6 samples. Also,
the author points out that in a previous flame ionization detector study a
preextraction with ether at pH 1 was required in some samples to remove GC
interfering species. Last, care must be taken to regiment the bromination
procedure if accurate and reproducible results are to be obtained. The yield
of dibromopropionamide is highly variant with the concentration level of
-------
TABLE 1. ACRYLAMIDE ANALYTICAL METHODS
Method
Detection Limit
Substrate
Reference
Cation Exchange Chromatography/ 0.5 yg
Ultraviolet Spectrophotometry
Derivatization/Ultraviolet 2 ppm
Spectrophotometry (2-50 ppm)
Polarography 100 ppm
Differential Pulse Polarography 1 ppm
Flame lonization Gas 400 ppm
Chromatography
Bromination/Electron Capture 0.25 ppb
Gas Chromatography
Polyacrylamide
Neat and urine
Polyacrylamide
Polyacrylamide
Acrylamide polymers
and copolymers
Water
Schmoetzer (1971)
Mattocks (1968)
MacWilliams et al. (1965)
Betso and McLean (1976)
Croll (1971)
Croll and Simkins (1972)
-------
acrylamide and amount of exposure to light during the photochemically
initiated bromination.
Ultraviolet Detection
UV detection of underivatized acrylamide is performed at 200 nm; it
is sensitive only to greater than 0.5 ug and it is easily interfered with
by a number of UV adsorbing species. Derivatization of the acrylamide
first can be useful in some cases where only ppm sensitivity is required,
but a few of the derivatizations require anhydrous media. Besides the
derivatives described by Mattocks (Table 1), acrylamide pretreatment with
iodine monochloride is reportedly useful for detecting as low as 0.5 ppm
by UV spectrophotometry (Rapaport and Ledovskikh, 1972).
Polarographic
Direct dc polarographic analysis described by MacWilliams et al.
(Table 1) is useful only for concentrations above 100 ppm. The Differen-
tial Pulse Polarographic (DPP) method of Betso and McLean (Table 1) is
sensitive to something less than 1 ppm and requires no chemical pre-
treatment of the acrylamide. However, in addition to nonsufficient
sensitivity for trace environmental analysis, a few species can interfere
with the acrylamide detection, such as alkyl esters of acrylic acid,
nitrilotripropionamide (reaction product of acrylamide and ammonia) and
large amounts of ammonia or alkali metal cations. Extensive ion-exchange
may be required to remove the latter.
Wet Chemical
A number of wet chemical methods for acrylamide analysis are known
including bromination (Narita et al., 1963), chlorobromination (Belcher
and Fleet, 1965), thiol addition (Beesing et al., 1949), and morpholine
addition (Belcher and Fleet, 1965). None of these methods are sufficiently
sensitive as presently employed for trace environmental analysis of
acrylamide.
ANALYTICAL METHODS DEVELOPMENT
This section describes the equipment selections that were made and
laboratory studies performed to develop and refine an acrylamide analytical
methodology. The work was directed toward a complete analytical protocol
development that would allow direct analysis of environmental samples
containing substantial quantities of acrylamide, ppm levels or higher, yet
with an appropriate sample workup would also allow for the detection of
trace quantities of acrylamide, ppb levels or less. The recommended analy-
tical protocol outlined in the next section was derived from these labora-
tory studies and fundamental analysis considerations.
-------
Detection Equipment and Column Detection
Detection Equipment Selection. The first decision to be made concerned
the selection of final detection equipment for acrylamide. Based upon the
literature survey and assessment and Battelle's analytical experience, GC-
electron capture, GC-nitrogen detector, and GC-MS were considered. Only
these systems could provide the sensitivity required together with certainty
of identification of the species being detected and quantitated.
The major disadvantage of the GC-electron capture system is that it
requires a chemical pretreatment procedure (bromination) regardless of the
level of acrylamide in the sample. The GC-nitrogen detector and GC-MS
systems would not require sample pretreatment for high level aqueous
acrylamide solutions and air samples. All three systems of course require
some form of acrylamide concentration step to allow for detection of trace
quantities, ppb levels or less.
Although GC-MS facilities are available at Battelle the GC-nitrogen
detector equipment would be much less of an investment for other labora-
tories to acquire. Further, the GC-nitrogen detector system is simpler to
operate and maintain for any routine environmental monitoring activities
and yet would be very adequate for the analyses. GC-MS would be an ideal
tool for acrylamide species confirmation on selected samples. Under these
considerations the GC-nitrogen detector system was chosen for development.
The work was first performed on a Varian 1700 Gas Chromatograph equipped
with a Perkin-Elmer 023-0301 nitrogen phosphorus detector. Later a
Hewlett Packard 5730A with a dual nitrogen-phosphorus detector (Model
18789A) was used. This thermionic detector utilizes an electrically heated
rubidium bead and produces and detects cyanamide ions from nitrogen-
containing organic compounds. The bead in the Hewlett-Packard instrument
showed better sensitivity over a longer period of time than that from the
other detector.
Column Selection. Several column packings were investigated for
analysis of acrylamide by gas chromatography. A discussion of the results
obtained with SP-1000, OV-17, Tenax-GC, and Chromosorb-101 is presented
here. Poropak columns have been investigated by other researchers (Goodyear
Tire & Rubber Co., 1977) and found not to be useful for acrylamide analyses.
SP-1000. This packing supplied by Supelco, Inc., is Carbowax 20 M
terminated with 2-nitroterphthalic acid, similar to FFAP packing reported
to be useful for acrylamide analysis by Croll (1971). Battelle examined
10 percent SP-1000 on Supelcoport 80/100 mesh, which is an acid washed,
silanized chromosorb W support, in 3 and 6 feet by 1/4-inch columns. The
retention time for acrylamide at 180 C was 2 min and 6 min, respectively.
A number of temperature manipulations were made including temperature
programming and it was concluded that 170 to 180 C isothermal gave the
best results considering retention time and peak symmetry.
Prolonged detector imbalance was evident when aqueous solutions were
injected on the SP-1000 columns. This phenomenon limits the use of SP-1000
for analysis of aqueous samples since this form of "tailing" extends out to
-------
the retention time for acrylamide. This problem was not encountered with
several organic solutions of acrylamide including acetone, methanol, and
ethyl acetate.
OV-17. Another column packing examined was 3 percent OV-17 (50:50
methyl phenyl silicone polymer) on 100/120 gas chrom Q support. At 180 C
on a 10-ft column of OV-17> acrylamide was unretained. At low temperatures
(50 C and 85 C) it was slightly retained but the detection limit in aqueous
solutions would be about 1 percent due to the fact that the acrylamide
comes off the column during the time that peak tailing occurs because water
temporarily imbalances the detector.
Tenax-GC. Tenax-GC 40-60 mesh, in a 6-ft by 1/4-inch column, retained
acrylamide for 4 min at 170 C isothermal. As with the other column supports,
various temperatures including temperature programming were studied, the
conclusion being that 170 C isothermal gave the best results with regard
to retention time and peak symmetry. Aqueous solutions could be chromato-
graphed on the Tenax-GC without any interference with either the acrylamide
resolution or detection by the water present. This was not the case with
the materials examined previously.
In changing to and from high temperatures needed to drive off various
contaminants in environmental samples, however, the Tenax-GC was found to
shrink. This resulted in gaps in the column and deterioration of peak
shape over the course of several days.
Chromosorb-101. This packing material was selected as having tolerance
for aqueous samples and better resolution of peaks than the Tenax-GC.
Observed shrinkage was much less than that of Tenax-GC. A 3-ft by 1/4-inch
column of 100-120 mesh Chromosorb-101 retains acrylamide for 5.5 min at
210 C isothermal. The detection limit without concentration was 1 ppm, with
column replacement only after several weeks as deposits of contaminants from
environmental samples built up on the column.
Equipment Optimization
Initial limits of detectability were established with the flow rates
for the Hewlett Packard Gas Chromatograph and nitrogen detector set
according to manufacturers recommendations (H2 flow 3.5 ml/min, He flow
20 ml/min and air flow 50 ml/min). The 3-ft Chromosorb-101 column was used
isothermally at 100 C. Under these conditions, 20 ng of acrylamide produced
a 70-line high peak with a retention time of 8 min at an attenuation setting
of 1 x 8. The detector voltage setting was 14.5 for this determination,
which produced a pen deflection of 10 lines at 1 x 32 attenuation with no
electrometer bucking applied. The detector was set at 250 C, the injector
at 200 C.
Temperature adjustments were studied first. Temperature programming
over several ranges altered the retention time but yielded much broader
peaks. Increasing the isothermal temperature from 180 C to 210 C decreased
-------
retention time from 8 to 5.5 min while maintaining a sharper, more symmetrical
peak shape.
Further optimization was accomplished by varying the gas flow rates.
During this procedure it was desirable not only to improve the height and
shape of the acrylamide peak but also to minimize the detector voltage
necessary to obtain a pen deflection of 10 lines at an attenuation setting
of 1 x 32. This sensitivity was chosen on the basis of manufacturer's
recommendations and was later increased to 20 lines at attenuation 1 x 32
in efforts to maintain greater sensitivity.
The optimal H2 carrier gas flow was found to be 3.1 ml/min. Further
adjustments yielded best results with air flow at 50 ml/min and He2 flow at
22 ml/min.
The maximized instrument conditions as used for subsequent experiments
are summarized below. These resulted in a 4-line peak height for 2 ng
acrylamide (Figure 1).
Temperature
Hydrogen flow
Air flow
Helium flow
Injector temperature
Detector temperature
Column
210 C isothermal
3.1 ml/min
50 ml/min
22 ml/min
200 C
250 C
3 ft x 1/4 in. glass Chromosorb-101
30
o
01
g 20
*o
M
.t? 10
*0>
X
o
01
OL
o»
c
CM
TJ
'E
o
>%
L.
O
I I
I
I
2345678
Retention Time (minutes)
10
Figure 1. GC response at detection limit for acrylamide
in methanol: 2 yl of 1 ppm solution.
-------
A typical chromatogram run under these conditions is shown in Figure 2.
After each run the column temperature was raised to 230 C for 16 min after
the acrylamide came off, to elute high boiling contaminants that could
interfere with subsequent analyses.
Acrylamide Identity Confirmation
At elevated temperatures it is known that acrylamide can be pyrolyzed
to acrylonitrile. Hence, the structural integrity of acrylamide under the
GC conditions was established by GC-MS. This GC-MS confirmation run was
done with a 5-ft x 1/4-inch column packed with Chromosorb-101 (Figures 3
and 4). Using an authentic standard solution of acrylamide in water, the
material which passes through the column under the temperature conditions
of the analysis was shown to be intact acrylamide. Further, with all of
the column support materials except Chromosorb-101, acrylonitrile was
injected along with acrylamide, verifying that the two compounds exit the
column with different retention times.
The GC-MS equipment is a very appropriate tool for species confirma-
tion. It may be utilized during any environmental monitoring program to
analyze a few selected samples and confirm that the species being quanti-
tated is acrylamide.
Direct Injection Methods Development
A preliminary standard curve was prepared to evaluate the linearity of
the equipment's response to various concentrations of acrylamide in aqueous
solution. Five solutions of acrylamide in methanol were prepared so as to
provide total injected quantities from 2 to 200 ng (2 u£ injections of 1 to
100 ppm solutions). These experiments were conducted with the GC-nitrogen
detector system equipped with a Hewlett Packard Model 7670A automatic
sampling tray and a Spectro-Physics SP-4000 chromatography data system.
A graph of the integrated area values versus the amount of acrylamide
injected is shown in Figure 5. The curves are nearly linear for the entire
range of acrylamide concentrations. Beyond 200 ng (100 ppm) the amount of
acrylamide approached the detector saturation limit and the relative response
of the detector began to level off.
During this study and a previous one where random injections of two
acrylamide concentrations were injected, no ghost peaks or other evidence of
nonquantitative retention and desorption from the Chromosorb-101 column were
observed.
These results indicate that about 1 to 100 ppm aqueous acrylamide samples
of environmental origin may be analyzed by direct injection of a few micro-
liters of the samples onto the GC column. Further, high levels of acrylamide
in other environmental media that can be leached or extracted into water may
also be analyzed by direct injection of the resultant aqueous solution. The
-------
Figure 2. GC chromatogram of acrylamide in an
environmental sample (attenuation
1x4).
com cr OTT-LE c-an. r»-a
«> _
I! ' I1
10 ZS
W C5 Gfl 7fl
Figure 3. Reconstructed gas chromatogram of
acrylamide in an environmental water
sample (Finnigan 3200 GC-MS).
-------
srerrrui TG, - -M
ci-cc-rc cow or ornz c-an. aa-a
U.O
QW)
Ifl
"I I
Cfl
n/c
i
60
70 00
I '
90
I I ' I ' I '
100 110 1Z0 130
1C0
Figure 4. Mass spectrum of acrylamide from confirmation
study of an environmental water sample
(Finnigan 3200 GC-MS).
10
-------
500,000
400,000
<
g 200,000
100,000
I
J I
10 20 30 40 50 60 70 80 90 100 110 120 130
Concentration Acrylomide, ppm
Figure 5. Standard response curve of Hewlett
Packard 7670A GC with nitrogen
detector to acrylamide in methanol.
11
-------
optimal range for direct analysis on this column would appear to be between
10 and 200 ng contained in 2 to 4 y£ solvent.
The standard curve data were used in a program written for the Spectro-
Physics SP-4000. This calculated the concentrations of acrylamide in
environmental samples directly from the peak area.
The program was also useful in monitoring the stability of the system.
Standard solutions were injected periodically and the calculated concentra-
tions were checked to see that results were consistent.
Concentration Methodology Development
Using the GC-nitrogen detector system as the method for final detection,
a concentration factor of 1000-fold or greater is required to allow for
detection of trace levels, 1 ppb and less, in environmental samples. An
intermediate degree of concentration, approximately 10 to 100-fold, would
be required for samples containing between 10 ppb and 1 ppm of acrylamide.
Several potential concentration methods were investigated using aqueous
samples. It was anticipated that environmental samples of other media
besides water and air could be readily transformed to aqueous samples. Air
is a special media case where, by simply extending the sampling time, samples
can be acquired with high enough levels of acrylamide to allow for direct GC
analysis.
The investigation of each potential procedure to concentrate aqueous
acrylamide samples is discussed below. Direct evaporation under a nitrogen
stream was selected for this study because it was the simplest method with
the least chance for sample loss.
Solvent Extraction. A simple separatory funnel extraction of a 12.5 ppm
aqueous acrylamide solution with an equal volume of ethyl acetate was per-
formed. Analysis by GC of the ethyl acetate layer after 1/2-hour exposure to
the aqueous acrylamide with three intermittent mixings revealed a concentra-
tion of 6.55 ppm, which demonstrates that about 52 percent of the acrylamide
was extracted into the organic phase. Therefore, about 5 consecutive extrac-
tions or a continuous extraction operation would be required to recover more
than 95 percent of acrylamide from aqueous systems using ethyl acetate.
Solvent extraction was not investigated further, since the distribution
coefficients for several other solvents are reported (Croll, 1971) to be
even less than the 52 percent we found for ethyl acetate.
Polyurethane Foam Sorbent. High density polyurethane foam was examined
for its ability to adsorb acrylamide from water solutions. The foam was
precleaned by soxhlet extraction with water, followed by acetone, and dried
at 80 C. Aqueous acrylamide (33 ml, 100 ppm) was exposed to 4.43 g of the
polyurethane foam in a 1-cm-diameter column for 1-1/2 hours and then compared
by GC for acrylamide content with an unexposed 100 ppm acrylamide sample. The
resulting chromatograms revealed very little (less than 5 percent), if any,
adsorption of acrylamide by the foam.
12
-------
Tenax-GC Sorbent. First, a preliminary evaluation of Tenax-GC as a
sorbent for acrylamide from aqueous solutions was made. The Tenax-GC,
40/60 mesh, was precleaned by soxhlet extraction with methanol for 48 hours
and dried at 80 C. Aqueous acrylamide (10 ml, 100 ppm) was stirred with
2 g of Tenax-GC in a 50 ml beaker for 2 hours. The mixture was filtered
and the resultant solution was compared by GC with a 100 ppm control sample.
The results indicate that about 33 percent of the acrylamide was adsorbed
by the Tenax. It was possible that the low efficiency of acrylamide
adsorption by Tenax-GC in this experiment could be due to saturation of
the adsorption sites with the high level of acrylamide employed (10 ml,
100 ppm); so another Tenax-GC evaluation experiment was performed using
1 ppm aqueous solutions.
Four 1-cm-diameter glass columns were charged with 2 g of Tenax-GC
in each. All glassware was silylated for this experiment. Acrylamide
adsorption was then attempted by passing 100 ml of 1 ppm aqueous acrylamide
through one column, 200 ml through a second, 500 ml through a third, and
1000 ml through a fourth at flow rates of 2, 2, 5-1/2, and 10-1/2 ml/min,
respectively. A slight pressure generated with nitrogen gas was necessary
to obtain the two highest flow rntes. Desorption of the acrylamide was
then attempted by eluting the Tenax-GC, after drying for 1/2 hour with a
stream of nitrogen, with 10 ml of methanol at a flow rate of 1/2 ml/min
while collecting 1 ml aliquots. An additional 20 ml of methanol eluent
was used on the Tenax column which had been exposed to 1000 ml of aqueous
acrylamide. The methanol eluate fractions from all of the columns were
examined by GC for acrylamide content and none were found to contain more
than our direct injection detection limit of 1 ppm. This experiment dem-
onstrated that Tenax-GC was totally inefficient as used for concentrating
acrylamide from low level (1 ppm) aqueous solutions.
XAD-2 Sorbent. Two grams of XAD-2 resin which had been precleaned by
soxhlet extraction with methanol and dried at 80 C was placed in a 1-cm-
diameter column. One liter of a 1 ppm aqueous acrylamide solution was
passed through the XAD-2 column at 10 ml/min. The column was dried for
1/2 hour with a stream of nitrogen and eluted with 30 ml of methanol at a
flow rate of 1/2 ml/min while collecting 1 ml aliquots of eluent. The
methanol aliquots were analyzed for acrylamide content by GC using the
nitrogen detector on a 6-ft Tenax-GC column at 170 C isothermal.
No acrylamide was detectable in the methanol eluates and our direct
injection detection limit is about 1 ppm. Thus, XAD-2 resin as employed
was totally inefficient for concentrating acrylamide from low level (1 ppm)
aqueous solutions.
Charcoal Sorbent. The charcoal used for these experiments was
petroleum based, 30/60 mesh, Lot 104 from SKC, Inc., and was precleaned
by soxhlet extraction with methanol for 48 hours, followed by drying at
80 C. Three 1-ctn-diameter colums were charged with 2 g each of charcoal;
1 £ of 1 ppm aqueous acrylamide was passed through one column, 500 ml through
a second, and 100 ml through the third column at gravity-controlled flow
rates of 4, 9, and 6.5 ml/min, respectively. Each of the columns was then
13
-------
dried with a stream of nitrogen for 1/2 hour and eluted with 30 ml of
methanol; 1 ml fractions of eluate were collected for analysis.
The methanol fractions from the first column (1000 ml) were examined
by GC. Acrylamide was found in fractions 2 through 18. The fractions were
combined, diluted to 25 ml, and analyzed by GC using the nitrogen detector
on a 6-ft Tenax column at 170 C isothermal. Comparison of the peak height
response of this solution to that of a 40 ppm authentic acrylamide standard
revealed an approximate concentration of 30 to 31 ppm. These semiquantitative
data indicates an overall recovery efficiency of 76 percent for the acry-
lamide from the 1 & of 1 ppm aqueous solution, llethanol eluates from the
columns loaded with 100 and 500 ml of the 1 ppm solution were not analyzed.
The GC analysis also revealed that severe contamination of the charcoal
which interfered with the subsequent analysis could occur easily during prepa-
ration of the charcoal absorption tubes. For this reason this method of con-
centrating aqueous samples of acrylamide was abandoned in favor of concentration
by the technique of direct evaporation described below.
Direct Evaporation. Direct evaporation of the aqueous acrylamide
by rotary evaporation, vortex evaporation, and evaporation under a nitrogen
stream was examined. Rotary evaporation of 100 ml of a 10 ppm solution of
acrylamide to just dryness followed by reconstituting with water resulted
in approximately 75 percent loss of the acrylamide. The 25 percent recovery
was estimated using the GC-nitrogen detector system by comparing the peak
height response of the evaporated reconstituted solution with that of an
authentic 100 ppm acrylamide standard.
It was suspected that a large amount of the acrylamide loss may be due
to adhesion to the glass vessel upon evaporation to dryness. Thus, the
experiment was repeated, stopping the rotary evaporation at about 5 ml
and reconstituting the sample to 10 ml with water. Comparing again with a
100 ppm acrylamide standard, nearly 100 percent recovery of the acrylamide
was demonstrated.
Thus, using rotary evaporation, beginning with 300 to 500 ml of aqueous
samples and reducing to 10 ml, one could achieve a 30 to 50-fold concen-
tration without significant loss of the acrylamide solute. This operation
would take approximately 1 hour.
Vortex evaporation of aqueous acrylamide samples was investigated using
a Buchler Instruments (Fort Lee, New Jersey) Model 3-2200 vortex evaporator
connected to a water aspirator vacuum system. Ten milliliters of a 1 ppm
aqueous acrylamide solution were vortex-evaporated to dryness in two separate
experiments, one with silylated and one with nonsilylated glass tubes. The
samples were reconstituted to 1 ml with distilled water and semiquantitated
by comparison with a 10 ppm authentic acrylamide standard using GC. The
recovery of acrylamide was only 10 percent with the nonsilylated tube, but
was nearly quantitative (greater than 95 percent) with the silylated tube.
-------
The experiments demonstrate that vortex evaporation can be used without
substantial acrylamide loss, provided silylated glassware is employed. This
method is limited however by the quantity that can be evaporated, 10 ml,
which controls the concentration factor that can be acquired and the time
required, which would be about 3 to 5 hours per 10 ml of sample with water
solutions.
For samples with acrylamide ranging from 50 ppb to 1 ppm, concentration
in silylated tubes can be accomplished by blowing the samples with nitrogen
while heating them slightly (40 C) to prevent freezing. While reducing a 20 ml
sample to 0.5 ml takes 3 to 5 hours, the process does not require constant
monitoring. The simple procedure also diminishes the chances of loss from
transfer of material since only one test tube is employed.
Internal Standard Selection
The use of an internal standard for GC analysis can eliminate the need
for determining the final volume of samples from a concentration process.
The ratio of the acrylamide response to that of a given amount of internal
standard placed in the sample can be used to quantify the acrylamide regard-
less of the sample volume. This technique is particularly useful when
final sample volumes are less than 1 ml and accurate volume determination
is difficult. Using the concentration procedures described above, the
methanol or water solutions must be concentrated to 1 ml and less to quantify
environmental acrylamide levels of 1 ppb and less. Therein lies the need
for a suitable internal standard.
The criteria for an internal standard are substrate and workup system
dependent. For the acrylamide analyses, a search for an internal standard
proceeded with these criteria:
Is soluble in water and methanol at about 10 ppm levels
Contains nitrogen which efficiently activates the
nitrogen detector
Is relatively nonvolatile
Has a GC retention time different from that of
acrylamide (4 min) on a 3-ft Chromosorb-101 column
at 210 C isothermal but not greater than 15 to
20 min
Is structurally similar to the compound to be
analyzed so that its behavior on the column can
be expected to be similar.
15
-------
Dimethylformamide, phenylacetonitrile, benzoamide, and benzonitrile
were examined as potential internal standards. Benzamide retention time
was far too great at 170 C. Phenylacetonitrile had a retention time of
20 rain, usable but somewhat long. Dimethylformamide retention overlapped
with acrylamide and was not useful.
Benzonitrile was found to have retention time of 7 min, gave a
sufficient detector response and exhibited good peak symmetry. It was
used in initial GC response studies. Later, however, it was abandoned
as not being sufficiently like acrylamide to be a reliable standard.
Subsequently, direct volume measurements of the concentrated solution
were made.
COLLECTION AND ANALYSIS OF ENVIRONMENTAL SAMPLES
Ambient air and soil samples were collected from six sites surrounding
the Dow Chemical complex (includes Dow Chemical and Dow Corning plants) at
Midland, Michigan. Ambient air samples were collected by an impinger method.
Water samples were collected upstream and downstream of the Tittabawassee
River and at the inlet of Lingle Drain into this river. The map of sampling
sites is shown in Figure 6; a topographic map of the area is shown in
Figure 7. The sampling system is described in the Appendix.
Water and air impinger samples were concentrated directly and analyzed
as aqueous solutions. Filters, which had been placed upstream of the
impingers for air sampling, were extracted with methanol and analyzed to
detect acrylamide in airborne particulate. Soil samples were dried and
extracted with methanol. Table 2 compiles the pretreatment methods used.
Gas chromatography with nitrogen detection was used for analysis. Limits
for detection of acrylamide varied with degree of concentration but were
as low as 1.0 ppb for air. Confirmation studies of selected samples were
performed using GC-mass spectrometry.
Water
Grab samples of 1000 ml water were collected and preserved by refri-
geration in brown bottles until analysis. Screening by gas chromatography
indicated that concentrations were less than 1.0 ppm; this step was deemed
unnecessary because of the low levels, and subsequent samples were not
screened prior to pretreatment.
Direct concentration in silylated tubes was found to be quite satisfac-
tory for aqueous solutions containing as little as 50 ppb acrylamide. A 20
ml sample is evaporated to approximately 0.5 ml (exact volume noted) using a
Kuderna-Danish or similar tube heater set for 40 C (to prevent freezing) and
a slow nitrogen stream.
Gas chromatography with nitrogen detection was used for analysis.
Operating conditions for the Hewlett-Packard 5730A gas chromatograph are
listed at the end of this section. Injected volume for all samples was
16
-------
# SITE NUMB
HIGHWAY
I SOIL/AIR SAMPLE
A WATER SAMPLE
Figure 6. Map of sampling sites for acrylamide
at Midland, Michigan.
17
-------
ono-bnitv - - tv ' L L - Bttal-
h. , A- ^ "K.7 I n^**^«*. I -»"%3-t**
^v^-^-v-oi^ - r 'v> r
r^^. ^\'-i-V\^ 5 * 1
»£ JSpifasarf '«V -'** f s
. T- i ~*""i~"4^r«or"
jAv'v. V--^#6
20'
u.
:-/"^\
*/Ek^
g
oo
DOW CHEMCAIL
-
DOW CORNING
I ; x-/'
?.,/.... . - ..
A WATER SAMPLE
PLANT SITE
SOIL/AIR SAMPLE
1/2
MILE
Figure 7. Topographic map of sampling sites for acrylamide at Midland, Michigan
-------
TABLE 2. ACRYLAMIDE SAMPLE PRETREATMENT METHODS
Media
Method
Status
WATER GC screening:
AIR
SOIL
,if >10 ppm; direct analysis
if <10 ppm; direct concentration
.particulate: collection on filters,
extract with methanol
-vapor: water impinger collection
Duplicate methanol extraction, proceed as with water
GC screening unnecessary
because of low environmental
levels. Verified, effective.
treat as water Verified, effective.
Verified.
-------
2 u&. Figure 1 (page 7) represents the gas chromatogram obtained from an
effluent water sample from Midland, Michigan. GC-mass spectroraetry confirmed
the presence of acrylamide. The chromatogram and spectrum are shown in
Figures 3 and 4 (pages 9 and 10).
Data from the analysis of water samples are compiled in Table 3. The
only acrylamide detected was from an outfall near the Dow Corning plant,
Sampling Station C.
TABLE 3. ACRYLAMIDE ANALYTICAL DATA FOR WATER
SAMPLES FROM MIDLAND, MICHIGAN
Sampling U.S. Geological
Station Description Survey Coordinates
A
B
C
D
E
Tittabawassee River, upstream
of chemical plant
Tittabawassee River, downstream
of chemical plant
Effluent from outfall at Lingle
Drain near Dow Corning
Tap water samples from Holiday Inn
Snow sample taken at air Sampling
Site 3
722442 m E
4832000 m N
726231 m E
4827500 m N
725308 m E
4829692 m N
723591 m E
4832910 m N
Acrylamide
Analysis,
PPba
<25
<25
25-125
<25
<25
Detection limit 25 ppb; quantitation limit 125 ppb.
Air
Air samples were collected using a midget impinger (EPA Method 6 type)
containing 25 ml distilled water and 10 ml methanol to prevent freezing.
Make-up methanol was added during the sampling period as necessary to
replace that lost by evaporation. Sampling was conducted for a period of
24 hours at a rate of 1 to 1.5 £/min. The impinger sample was transferred
quantitatively into a 25 ml volumetric flask with several rinses of
distilled water until the final volume was 25 ml. Total air volume sampled
was 2 m3. Samples were preserved by refrigeration in brown bottles until
analysis.
Pretreatment for air impinger samples was direct concentration, as
described for water. The 25 ml samples were reduced to approximately 0.5
ml and transferred to vials for the automatic sampler. The volume was
brought up to 1 ml with several rinsings from the concentrator tube. GC
analyses were performed as for water.
20
-------
Airborne Particulate. Filters were placed upstream of the impinger
system to collect airborne participate separately. The filters were then
extracted three times by shaking 10 min with 7 ml of methanol. The combined
extracts were evaporated to approximately 0.5 ml (exact volume noted) and
analyzed by GC in the manner described for water.
Table 4 lists the results of the environmental air analyses. Detection
limits for both acrylamide vapor and particulate were 1 ppb which could be fur-
ther lowered by concentration to 0.1 ml or less. Acrylamide in the vapor samples
was not detected at the 1 ppb level. Some of the particulate samples yielded
small peaks at the proper retention time for acrylamide; however, confirmation
studies by GC-MS indicated that no acrylamide was present at the 1 ppb level.
Soil
Soil samples were preserved by refrigeration prior to analysis. The
soil was dried for 2 hours at 100 C and 50 g samples of the dry soil were
then extracted by shaking 1 hour with 70 ml of methanol. The extract was
decanted and the soil extracted a second time in a similar manner. The
combined extracts were filtered using a suction flask, and partially concen-
trated with a rotary evaporator. The volume was further reduced 0.5 to
1.0 ml (exact volume recorded) under a nitrogen stream. Precipitation was
observed upon concentration to volumes less than 1.0 ml and a microfiltra-
tion step would be recommended if a lower detection limit is desired. GC
analysis was performed under conditions discussed subsequently, with confirma-
tion of selected samples by GC-MS.
As indicated in Table 5, no acrylamide was detected in the soil samples.
All, however, displayed interfering peaks on gas chromatograms similar to
those noted in the airborne particulate data. Results of confirmation
studies were negative. The soil samples with the highest interfering peaks
were from the same sites that yielded the particulate samples which displayed
interferences. This correlation could be expected from airborne soil from
those sites.
GC Analysis and MS Confirmation
Acrylamide analyses were conducted by gas chromatography using a nitrogen
detector. The GC conditions on a Hewlett-Packard 5730A gas chromatograph or
similar instrument are
Column: 3-ft x 1/4-inch glass, packed with
Chromosorb-101, 100-120 mesh
Column Temperature: 210 C isothermal
Injector Temperature: 200 C
Detector Temperature: 250 C
Carrier Gas: Helium at 32 ml/min.
21
-------
TABLE 4. ACRYLAMIDE ANALYTICAL DATA FOR AIR SAMPLES FROM MIDLAND, MICHIGAN
Weather Conditions
Sampling
Station
1
2
3
4
5
6
U.S. Geological
Survey
Coordinates
722448 m E
482819 ra N
722417 m E
4831905 m N
723591 m E
4832910 m N
725810 m E
4831445 m N
725793 m E
4831088 m N
726286 m E
4831833 m N
Wind
Direction, Speed,
degrees m/sec
245-340 1.8-4.8
Ditto 1.8-4.8
1.8-4.0
1.8-3.0
1.2-3.8
1.8-5.0
Temp
Range ,
F
20-45
20-47
20-49
20-46
20-38
20-30
Acrylamide Analysis
Parti- b
culate, Vapor
General Sampling Period VJg/m3 ppb ug/m3
Partly cloudy 4/6/77 1308 to <1.0 <1.0 <1.0
to clear 4/7/77 1323
Ditto 4/6/77 1408 to <1.0 <1.0 <1.0
4/7/77 1417
11 4/6/77 1533 to <1.0 <1.0 <1.0
4/7/77 1583
" 4/6/77 1667 to <1.0 <1.0 <1.0
4/7/77 1683
" 4/6/77 1792 to <1.0 <1.0 <1.0
4/7/77 1820
" 4/6/77 1100 to <1.0 <1.0 <1.0
4/7/77 1122
Detection limit: 1.0 yg/m3.
Detection limit: 1.0 ppb.
-------
TABLE 5. ACRYLAMIDE ANALYTICAL DATA FOR SOIL
SAMPLES FOR MIDLAND, MICHIGAN
Sampling
Station
1
2
3
4
5
6
U.S. Geological
Survey Coordinates
722448 m E
4828198 m N
722417 m E
4831905 m N
723591 m E
4832910 m N
725810 m E
4831445 m N
725793 m E
4831088 m N
726286 m E
4831833 m N
Acrylamide
Analysis,
ppba
<20
<20
<20
<20
<20
<20
Detection limit 20 ppb; quantitation limit 100 ppb.
The nitrogen detector conditions to be used on a Hewlett Packard
nitrogen-phosphorus detector are
Air Flow Rate: 50 ml/min
Hydrogen Flow Rate: 3.1 ml/rain
Bead Voltage/Temperature: Sufficient to give 20
linear deflection at 1 x 32 attenuation.
A Finnigan 3200 GC-MS was used for confirmation studies. The chemical
ionization mode was used. GC conditions were identical to those previously
indicated. A 5-ft column of Chromosorb-101 produced an expected increase in
retention time over the 3-ft column used for the gas chromatograph.
RECOMMENDATIONS
Our present analysis method using gas chromatography with nitrogen
detection is capable of analyzing down to 1 ppm (approximately 2 ng) of
acrylamide in solution. Detection below 1 ppm (approximately 2 ng) is
difficult for two reasons: the response of acrylamide to nitrogen detection
is only moderate compared with that of more favorable nitrogen compounds, and
the gas chromatography of amides gives rise to tailed peaks which limit their
detection. Nevertheless, with a simple concentration step we can detect
down to 50 ppb of acrylamide by this method.
23
-------
In order to analyze this compound at lower levels, a derivatization
procedure is necessary. Bromination followed by GC with electron capture
detection has been used to analyze for acrylamide at the 1 ppb level, but
the derivation step is variable, making quantification difficult. We
believe that the following approach may provide the methodology long needed
for acrylamide analysis. Acrylamide can be derivitized to a pyrazoline (II)
quite easily using diazomethane as a derivitizing reagent (Mattocks, 1968):
CH2N2
CH2=CH-CONH2 * | [TCONH2
II
This pyrazoline can then be converted to a highly UV active Schiff base by
reaction with 4-dimethylaminocinnamaldehyde III:
i?
-CONH2 ^ p |jC>NH2
N
I I
I N
v/
N
H CH
II
Both of these reactions are carried out quantitatively at room temperature
in a single test tube. The attractiveness of this method is that there are
two possible modes of analysis: GC with nitrogen detection of the pyrazoline
and HPLC detection of the Schiff base. The pyrazoline should show both
increased sensitivity due to the increase in nitrogen content and also
better chroma tographic properties (reduced tailing) due to the increase in
molecular weight. HPLC of the Schiff base should be straightforward and
very sensitive. Schiff base can easily be chromatographed with excellent
peak shape and its extinction coefficient is approximately 100,000. This
should allow the analysis of program levels of acrylamide. Since the wave-
length monitored would be 538 nm, little interference would be suspected and
a highly selective and sensitive method should result.
Because of the difficulty involved in analyzing amide in general by GC,
it is recommended that this derivatization approach be tried in future
acrylamide analysis.
24
-------
PART II. ORGANOTINS
INTRODUCTION
The purpose of this research program was to monitor the levels of
organotin compounds in environmental samples. After a literature search
was conducted, it became evident that a sensitive and specific method
capable of analyzing for all important organotin compounds was not avail-
able. We then divided the research effort into two phases: Phase 1 Methods
Development, and Phase 2 Sample Analysis. During Phase 1 an analytical
method was developed in which major organotin compounds of commercial
interest could be speciated as their acetates and quantified in environ-
mental media. During Phase 2, air, water, and soil samples were analyzed
for organotins from two sitesa production site and a users site.
Organotin compounds are widely used in the plastics industry as
stabilizers for poly(vinylchloride), agriculturally as fungicides and
insecticides, and in antifouling paints as biocides. Because of their
widespread and divergent use, there is considerable concern about their
entry into the environment and possible toxicity to man.
LITERATURE REVIEW
Clear evidence of the difficulty in analyzing for organotin compounds
is the number of different analytical methods that have been tried.
Spectrophotometry, electrochemistry, chromatography, and atomic absorption
spectroscopy have all been used with varying degrees of success.
Early analysis for organotins involved extractions followed by colori-
metry with reagents such as toluene-3,4-dithiol (Corbin, 1970), dithizone
(Aldridge and Cremer, 1957; Chromy and Uhacz, 1968), catechol violet (Adcock
and Hope, 1970; Corbin, 1973), 4-(2-pyridylazo) resorcinol (Sawyer, 1967),
diphenylcarbazone (Skeel and Bricker, .1961), quercetin (Filer, 1971), and
3-hydroxylflavone (Vernon, 1974). These methods are fairly nonspecific and
relied on elaborate extraction and distillation schemes to speciate the
organotins prior to analysis. Also, the sensitivity of these methods is not
high enough for environmental monitoring of low concentrations.
Anodic stripping voltametry (ASV) has been used to analyze for organo-
tins before and after oxidation to elemental tin (Booth et al., 1970; Woggon
et al., 1972). The sensitivity of this method is quite good but the method
lacks the capability to speciate the various organotin compounds. Neverthe-
less, ASV could be successfully used as part of an analytical method for the
25
-------
analysis of organotins. Alternating current polarography has also been used
for the direct estimation of organotin compounds (Jehring and Metiner, 1963).
The most widely used method for the speciation of organotins has been
chromatography. Numerous paper and thin layer chromatography systems have
been developed for the separation of certain groups of organotins such as
the octyl- or phenyltins (Figge, 1969; Simpson and Currell, 1971; Cassol et
al., 1965; Williams and Price, 1964). Gas chromatography has also been used
extensively for the analysis of tetraalkyltins (Gauer et al., 1974;
Steinmeyer et al., 1965; Putnam and Pu, 1965; Tonge, 1965; Geissler and
Kriegsmann, 1965). Recent studies have used gas chromatography combined
with atomic absorption spectroscopy for the specific analysis of volatile
organotins (Brinkman et al., 1976; Parris et al., 1977) and GC-MS analysis
(Meinema et al., 1977). The main problem associated with any gas chroma-
tography method is the nonvolatility of many transformation products of
organotins, such as mono- and di-compounds. Volatile derivatives, therefore,
must be made. The most widely used procedure is to methylate prior to gas
chromatography. This precludes the speciation of methyItins and also can
lead to erroneous quantitation due to rearrangements of the organic moieties
on the tin molecule.
Atomic absorption chromatography is an ideal method for the analysis of
tin compounds. It is both highly sensitive and specific. This method has
been used widely (Freeland and Hoskinson, 1970; Meranger, 1975; George et al.,
1973; Jeltes. 1969) and compares favorably with polarographic and spectro-
scopic methods (Williams, 1973; Engberg, 1973). As with other methods,
speciation is not possible. Therefore, atomic absorption spectroscopy must
be combined with chromatographic steps for speciation and quantitation. As
we have mentioned, GC-AA has been used for volatile organotins, but the
ideal method would speciate both volatile and nonvolatile organotins. Both
high-performance liquid chromatography (HPLC) and thin layer chromatography
(HPTLC) offer compatible speciation methods prior to atomic absorption
analysis. Of the two chromatographic methods, HPLC-AA is the most promising
since it offers the possibility of on-line speciation and quantitation.
Unfortunately, it has not been developed sufficiently to permit routine
analysis. However, the method appears quite promising (Botre et al., 1976;
Jones and Manahan, 1975; Jones and Manahan, 1976a, 1976b; Stoveken and
Fernzndez, no date; Jones et al., 1976).
Due to the time limitations of our program, we decided to pursue the
HPTLC separation method followed by flameless atomic absorption spectro-
scopy. During our program, an article on the use of HPTLC followed by AA
has been published which details the possibility of the method (Issaq and
Barr, 1977). Excellent bibliographies on organotin analysis have been
published detailing most of the methods discussed briefly in this report
(Luijten, 1970; Dillard, 1971). However, the newer, more sensitive me
methods were not available at that time.
26
-------
ANALYTICAL METHODS DEVELOPMENT
After our literature search had been concluded, it became obvious that
there were two serious faults with published methods; the lack of sensi-
tivity and the capability of analyzing for more than one class of organotin
compounds. In most cases the published methods were for high levels of
organotin compounds which are not likely to be encountered in the environ-
ment.
Because of the nature of this program, it was necessary to develop a
method which was capable of quantifying individual organotin species as
well as their transformation products at low levels in environmental samples.
To accomplish this goal, we developed a three-step procedure in which the
organotins are concentrated by dynamic adsorption, speciated by chromato-
graphic methods, and quantified by atomic absorption spectroscopy.
Low levels of organotins can be quantified when concentrated by dynamic
adsorption techniques. Water samples (100 ml) spiked with different con-
centrations of organotins were pumped through a 10-cm column packed with
silica gel, as shown in Figure 8. Organotin compounds are strongly adsorbed
onto the silica gel and are, therefore, easily concentrated. Solid media
such as soil and particulate are first extracted into an aqueous phase,
then concentrated.
After concentrating the organotins, the column is reversed and the
organotins are eluted with a solution of methanol/acetic acid (80/20).
Recoveries of 90 percent or better were achieved on the organotins tested
with elution volumes of 5 ml. The resultant solution is further concentra-
ted by evaporation. We are presently achieving concentraction factors of
100:1 using this technique but this may not yet represent the optimum.
Analytical data concerning recovery studies during the concentration step
are described shortly.
The second step in our analytical procedure is speciation of the
organotins using thin layer chromatography (TLC). We feel that ultimately
the chromatographic method of choice is high performance liquid chromato-
graphy (HPLC), but due to the time limitations of the present program we
were not able to convert our TLC data to HPLC. Nevertheless, the TLC
procedure can be used to successfully speciate the organotins of interest.
Table 6 shows the separation of the major organotin species and their
degradation products. This separation represents the first attempt at
speciating organotins of environmental concern in one chromatographic step.
We have found that compounds having similar aryl/alkyl groups attached to
the tin atom have essentially identical Rf values when chromatographed on
TLC plates of silica gel with eluents containing acetic acid. Presumably,
the organotin compounds are readily hydrolyzed in acetic acid solution and
chromatograph as the corresponding acetates. Thus, similar aryl/alkyl
derivatives of organotin compounds with different anionic groups (e.g.,
dibutyl tin dilaurate, dibutyl tin dichloride) can be analyzed as single
compounds (acetates) thereby significantly simplifying the analytical effort.
Since the anionic moieties of the organotin probably have no effect on its
27
-------
N,
100 mL -
resevoir
Concentration column
10cm X mm ID
5mL
syringe
Figure 8. Concentration and subsequent elution
of organotins.
28
-------
TABLE 6. SEPARATION OF ORGANOTIN CLASSES BY
THIN-LAYER CHROMATOGRAPHY
(Solvent conditions: 20% acetic acid in hexane;
silica gel plates; Quanta/gram LDQ)
Aryl/Alkyltin Derivative
Tricyclohexyltin (
Tributyltin
Dioctyltin
Triphenyltin
Dicyclohexyltin
Dibutyltin
Diphenyltin
Trimethyltin <
Dimethyltin (
Butyltin (
Methyltin «
1 0.7
0.6
0.5
0.4
I 0.3
> 0.2
> 0.1
1 0.0
toxicity, our simplified speciation appears justified. Specific analytical
data concerning thin layer chromatography separations, techniques, and
recoveries will be shown shortly.
The final step in methods development was to develop a sensitive and
specific detection method capable of quantifying the organotin compounds.
Atomic absorption spectrophotometry was chosen mainly because of its unique
specificity for tin. Although it is not the most sensitive technique for
these compounds, lower limits of detection of 5 ppb can be achieved. We
have found that instrument sensitivity is not the limiting factor in
detection limits. The ability to carry low levels of organotins through
the concentration and separation steps with minimum losses is a most
difficult analytical task. Atomic absorption spectroscopy must be capable
of analyzing a 0.1 ppm final solution which represents down to the ppb
level of total tin content in the original environmental sample.
Figures 9 and 10 show atomic absorption calibration curves for bis(tri-
n-butyltin) oxide and methyltin trichloride. In both cases, 0.05 ppm was
easily detected at a scale expansion of IX. Correlation coefficients for
both curves exceed 0.999. We have found that more accurate results are
obtained when calibration curves for each individual organotin compound are
used for quantification. Figure 11 shows the atomic absorption response for
different concentrations of bis(tri-n-butyltin) oxide. We originally were
having problems with the reproducibility of individual injections into the
29
-------
IOC
90
BO
70
5 60
o
50
20
10
100
9O
80
70
60
50
40
30
20
10
Ol
02
PPM
(Tin)
03
04
01
O2
PPM
(Tin)
O3
04
Figure 9. Bis(tri-n-butyltin) oxide
standard curve.*
Figure 10. Methyltin trichloride
standard curve.*
*Perkin-Elmer 305B with a background corrector and a flameless HGA 2000. An electrodeless
discharge lamp was used at 224.6 nm wavelength. The operating parameters of the HGA with a
tube are: dry, 75-30 sec; char, 500-40 sec; atomize, 2300-8 sec; and N2 flow, 50 ml. Injections
of organotin standard solutions containing 10% (by volume) acetic acid and 1% (by volume)
phosphoric acid were used.
-------
Q.
to
0.32 ppm
0.24 ppm
0.16 ppm
0.08 ppm
Time
Figure 11. Atomic absorption response to
bis(tri-n-butyltin) oxide (condi-
tions same as in Figures 9 and 10)
TABLE 7. REPRODUCIBILITY OF ATOMIC ABSORPTION SPECTROSCOPY RESPONSE
FOR METHYLTIN TRICHLORIDE AT DIFFERENT CONCENTRATIONS
(HGA parameters are: dry, 75-30 sec; char, 500-40 sec;
atomize, 2300-8 sec; auto, N£ purge interrupt.)
Concentration,
ppm
No. of
Injections
Scale
Expansion
Mean
Peak Height
Standard
Deviation,
0.025
0.049
0.123
0.245
5
5
5
6
3X 12.6 ± 0.37 2.9
3X 23.2 ± 0.18 0.9
3X 59.3 + 1.4 2.4
IX 38.4 ± 1.0 2.6
31
-------
the sample compartment of the atomic absorption spectrometer. The addition
of phosphoric acid to the solutions to be analyzed significantly improved
our results. Table 7 shows the excellent reproducibility obtained for
methyltin trichloride at different concentrations.
Recovery Studies During the
Sample Concentration Procedure
Water. Two organotin compounds, methyltin trichloride and bis(tri-n-
butyltin) oxide, were chosen as test compounds for initial recovery studies
from water based on their physical properties. Thin layer chromatography
data had shown that methyltin trichloride was the most strongly adsorbed
and bis(tri-n-butyltin) oxide was the least strongly adsorbed of all the
tin compounds tested. Since we were testing adsorption-desorption
properties of these compounds in our recovery study, these two compounds
were logical choices, representing the most difficult cases. In addition,
bis(tri-n-butyltin) oxide, a fairly volatile compound, could be used to
evaluate any losses that might be occurring during sample handling and
evaporation steps.
The procedure used to evaluate the concentration step is as follows:
100 ml of 0.1 ppm organotin solution was pumped
through the silica column at 1 to 2 ml/min.
The column was inverted and the organotin eluted
successively with five 2 ml portions of methanol/
acetic acid (80/20).
The methanol was evaporated under a gentle stream
of nitrogen.
The solution was then diluted with water, resulting
in a 15 percent acetic acid solution which matched
the solutions used to obtain the standard curve in
atomic absorption spectroscopy.
The concentrations of the solutions were determined
by atomic absorption spectroscopy.
The results of these studies are given in Tables 8 and 9. Table 8 shows
the excellent recovery of methyltin trichloride during the concentration step
(99 percent). This result indicated to us that there would be no problem in
desorbing these compounds from the adsorbent, since methyltin trichloride
absorbed the strongest. Also shown are the results obtained when the
methanol evaporation step was not performed. This experiment was performed
simultaneously with the first experiment since we were concerned with the
loss of organotin on evaporation. On the contrary, low recoveries were ob-
served. We have found that these low recoveries are due to the change in atomic
absorption response because of the presence of methanol in the graphite tube.
32
-------
TABLE 8. RECOVERY STUDY OF METHYLTIN TRICHLORIDE FROM WATER
Sample
Spiked Solution
(in 100 ml)
Filtered Solution
Eluent Fraction 1
Eluent Fraction 2
Eluent Fraction 3
Eluent Fraction 4
Eluent Fraction 5
Total
a
Analysis,
ug Sn
6.7
<0.5
5.60
0.42
0.30
0.22
0.10
6.64
(Recovery - 99%)
Analysis,
Pg Sn
6.7
<0.5
4.10
0.22
0.08
<0.05
<0.05
4.50
(Recovery - 82%)
Spiked sample placed on silica gel column and eluted
successively with five 2 ml portions of methanol/
acetic acid (80/20).
3Same as (a) except methanol evaporation step was not performed.
TABLE 9. RECOVERY STUDY OF BIS(TRI-N-BUTYLTIN) OXIDE FROM WATER
Sample '
Spiked Solution
(in 100 ml)
Filtered Solution
Eluent Fraction 1
Eluent Fraction 2
Eluent Fraction 3
Eluent Fraction 4
Eluent Fraction 5
Total
Analysis ,
Ug Sn
10.5
<0.5
5.90
0.95
0.34
0.16
0.10
7.45
(Recovery - 71%)
Analysis,
yg Sn
10.5
<0.5
6.10
1.07
0.26
0.18
0.09
7.70
(Recovery - 73%)
Duplicate determinations.
Spiked sample placed on silica gel column and eluted
successively with five 2 ml portions of methanol/
acetic acid (80/20).
33
-------
It is very important that the solvent matrix of both the samples to be
analyzed and the analytical standards be as similar as possible.
Table 9 shows the recovery of bis(tri-n-butyltin) oxide during the
concentration step. The low recoveries were once again attributed to the
atomic absorption spectroscopy response. Response curves for bis(tri-n-
butyltin) oxide were used for quantitation during recovery studies.
However, we have found that the compound actually analyzed after thin
layer chromatography is tributyltin acetate. The response curve for
tributyltin acetate has a different slope than the oxide and accounted for
this low recovery. Recoveries of greater than 90 percent were obtained
when the response was quantified from the tributyltin acetate curve.
Soil and Particulate Samples. Soil and particulate samples were the
most difficult media to handle with respect to extraction and concentration
of the organotin compounds. Numerous solvents were tried for extracting
bis(tri-n-butyltin) oxide from soil samples. Bis(tri-n-butyltin) oxide was
chosen as the test solute, since we had learned that samples from the sites
selected for monitoring would most likely contain this compound. Table 10
shows the extraction recoveries obtained with four different solvent
mixtures. Soil (Brookstone silty clay loam) samples were spiked at the 1 ppm
level and soxhlet extracted overnight. Only 100 percent acetic acid gave
good enough recoveries for further analytical work. However, at the 0.1 ppm
level a similar extraction gave a recovery of only 40 percent tin recovery.
Therefore, further work needs to be done to improve tin recovery at lower
levels.
TABLE 10. RECOVERIES OF BIS(TRI-N-BUTYLTIN)
OXIDEa FROM SOIL
Recovery,
Solvent %
Methanol 2
50% methanol/50% acetic acid 10
80% water/20% acetic acid 10
Acetic acid 90
a25 g soil spiked with 1.0 ppm bis(tri-n-
butyltin) oxide and soxhlet extracted overnight.
Once the samples were extracted, they needed to be concentrated prior
to speciation and quantitation. To accomplish this, a sufficient amount of
water was added to dilute and neutralize the acetic acid, followed by
34
-------
concentration by the dynamic adsorption technique. Acetic acid does extract a
considerable amount of material which precipitates on addition of water.
High-speed centrifugation (8000 rpm) was needed to separate out the insoluble
material. The resultant solution was then concentrated onto the silica
column as in the work with water samples, and essentially similar recoveries
were obtained (despite the solution's high ionic strength).
The procedure that was adopted for soil analysis was:
Soxlet extraction of 25 grams of soil in 75 ml
acetic acid
Dilution of the extract to 200 ml with water
Centrifugation to remove suspended material
Neutralization to pH 7 with NaOH
Dynamic adsorption on silica gel.
Air Samples. Two methods were tried for the concentration of tins from
air: (1) adsorption on silica gel, and (2) passing air through an acetic acid
impinger. Since we had experience in adsorbing and desorbing organotins from
silica, we tried this method first. We tried to concentrate bis(tri-n-butyltin)
oxide from air onto a silica gel adsorbent tube using the apparatus shown in
Figure 12. The silica gel used was Analabs' gas chromatographic grade
82-003060-00, 30/60 mesh.
Glass -
U-tube
Silica trap
Organotins
Figure 12. Apparatus for adsorbing volatile organotins,
35
-------
Twenty grams bis(tri-n-butyltin) oxide in methanol was placed in a glass
U-tube. The methanol was evaporated and 250 u£ water added. The tube was
attached to the silica trap and nitrogen was used to purge the tube for 11
hours at 1.4 £/min. Analysis showed 11 yg remaining in the U-tube and 9.4
Mg collected on the silica trap for a recovery of about 104 percent. The
silica gel trapping method was chosen on the basis of these results.*
Recovery of tin using the acetic acid impinger method was not satisfac-
tory. In one experiment, air was passed over 20 pg of bis(tri-n-butyltin)
oxide dissolved in a 1:1:1 mixture of 500 vi£ water, acetic acid, and ethanol
and contained in a glass U-tube. Air was passed through the tube for 9 hours
into an impinger containing 25 ml of acetic acid. During this time it was
necessary to add an additional 30 ml of acetic acid to the impinger. At the
end of this period, the impinger analyzed 7.9 yg of tin. The amount of tin
left in the tube was 9.5 ug. Thus, the recovery of tin volatized was 71
percent.
Speciation of Organotins by
Thin Layer Chromatography
A number of different solvent systems were tried in an attempt to
maximize the resolution of the organotin compounds. These included varying
proportions of hexane, methylene chloride, methanol, butanol, octanol, acetic
acid, and water. Table 11 shows the Rf values using two optimized eluents.
The thin layer chromatography plates used were Quanta/gram LDQ which were
activated by heating for 1 hour at 100 C.
The spots were identified using a standard visualization procedure.**
The plates were air dried after development and placed in a tank filled with
chlorine gas for 2 minutes. The plates were then removed and sprayed with
pyrochatechol violet solution. The organotin compounds showed up as blue to
violet spots.
Recovery studies were once again used to determine whether the spots could
be removed from the thin layer chromatography plate and analyzed by atomic
absorption spectroscopy. Varying concentrations of bis(tri-n-butyltin) oxide
in 100 percent acetic acid were spotted onto the thin layer chromatography
plates. One hundred percent acetic acid was used since this would ultimately
*It should be noted that when commercial silica gel tubes (SKC, Inc.,
Pittsburgh, Pennsylvania, were tried for this work, very poor recoveries were
obtained (less than 10 percent). We feel that the silica in these commercial
adsorbents was too active (high surface area) and that desorption of the
organotins is a problem. When trying to analyze for organotins, care must be
taken in choosing a very low surface area silica gel material.
**"Standard Test Methods, TLC-5", used in-house by M&T Chemical Company,
Rahway, New Jersey.
36
-------
TABLE 11. Rf VALUES OF ORGANOTINS ON THIN LAYER
CHROMATOGRAPHY SILICA GEL PLATES
3, c be
Compound Rf ' R^ '
Dibutyltin dilaurate 0.34 0.31
Dibutyltin dichloride 0.33 0.31
Dibutyltin bis(lauryl mercaptide) 0.32 0.30
Dibutyltin bis(isopropyl maleate) 0.32 0.30
Dibutyltin bis(isooctyl maleate) 0.32 0.29
Dibutyltin bis(isooctyl mercaptoacetate) 0.31 0.29
Butyltin trichloride 0.0 0.0
Butyl stannoic acid 0.0 0.0
Bis(tri-n-butyltin) oxide 0.49 0.34
Tributyltin acetate 0.49 0.33
Triphenyltin hydroxide 0.40 0.33
Tricyclohexyltin bromide 0.55 0.40
Trimethyltin hydroxide 0.14 0.26
Dicyclohexyltin dibromide 0.33 0.31
Methyltin trichloride 0.0 0.0
Tri-n-butyltin fluoride 0.50 0.43
Dioctyltin oxide 0.30 0.34
Dibutyltin oxide 0.32 0.31
Dimethyltin oxide 0.06 0.13
aSolvent: 10% acetic acid in hexane.
Solvent: 20% methylene chloride, 5% methanol, 2% acetic acid
in hexane.
n
Silica gel plates: Quanta/gram LDQ (see text).
be the solvent spotted after the concentration and evaporation steps.
Up to 0.5 ml was spotted without loss in chromatographic resolution. This
was made possible by a unique automatic spotting technique involving deposi-
tion of a solution on a thin layer chromatography plate from a motor-driven
syringe in a pulsing manner. After several unsuccessful recovery studies,
we found that the solutions must be spotted directly on the silica gel, and
and not on the preadsorbent present in the quantum plates. Recoveries of
90 percent plus can now be obtained using our method.
In order to speciate organotin compounds in environmental samples,
their concentration should be at least 100 ppb in air and water and 1000
ppb in soil.
37
-------
COLLECTION AND ANALYSIS OF ENVIRONMENTAL SAMPLES
Environmental water, air, and soil samples were collected from two sites:
the Cincinnati Milacron Chemicals, Inc., plant which manufactures organotins in
Reading, Ohio; and Avondale Shipyard in New Orleans, Louisiana, which uses or-
ganotins as an antifouling agent on shipbottoms. The levels of total tin in the
samples from both sites were below that required for speciation. Hence, the con-
centration levels found are expressed in terms of total acetic acid-extractable
tin (inorganic and organic).
Reading Site
Geographical data on the Reading site are shown in Figure 13; a
topographic map of the area is shown in Figure 14. The plant is located in
an industrial area which is bordered very closely to the south and west by
residential areas. Sampling sites are also shown.
Water Samples. Grab samples of 1000 ml water were collected in amber
bottles and immediately made to 0.01 M with citric acid (to prevent adsorp-
tion of tin compounds on the walls of the glass). These samples were pre-
served by refrigeration until analysis. One hundred ml of these samples
were neutralized to pH 7 and concentrated by dynamic adsorption on silica
gel. The concentrate was then analyzed by atomic absorption spectroscopy
for total tin content. The results are shown in Table 12 along with
TABLE 12. ORGANOTIN ANALYTICAL DATA FOR WATER
SAMPLES FROM READING, OHIO
Sampling U.S. Geological
Station Survey Coordinates Description
Organotin Analysis,
ppba
RW
RE
ROIS
ROS
721168 m E
4346144 m N
720575 m E
4345424 m N
720870 m E
4345700 m N
720880 m E
4345700 m N
Upstream of plant on
Mill Creek just
below city dump
Downstream of plant on
Mill Creek
24-hour outfall inte-
grated sample
Grab sample taken
after collection of
outfall integrated
sample
4.1
2.3
4.4
Values reported are for acetic acid-extractable tin.
38
-------
# SITE NUMBER
HIGHWAY
«-H-RAILROAD
PLANT SITE
AREA
RESIDENTIAL
AREA
SOIL/AIR SAMPLE
A WATER SAMPLE
Figure 13. Map of sampling sites for
organotins at Reading, Ohio.
39
-------
# '-. :\n
£*/*»* i*
O-?i f
^OfflC^k sHSBSXv"
f^^.^&OX^^i^K-A-: ="r>
/4> F^^A w > ^"-^A^j f»'»W
iM
F r ti"<
ZS^SBSB^i^*^! 6=
^^^SC^. NR* M n /
/rrvr*'/toJraSFvin^H9^nt£ai>'' . *fl ;.: ::.;
S.feS^^; <^itr >vf/ /r;
A WATER SAMPLE
Q PLANT SITE
SOIL/AIR SAMPLE
KILOMETER
I .5 1
___i^= K=^K=1
) 1/2
MILE
Figure 14. Topographic map of sampling sites for
organotins at Reading, Ohio.
-------
information concerning collection. These values are too low to permit
speciation and are, therefore, reported as total tin values. To ensure
that our method was performing satisfactorily during sample analysis,
actual water samples spiked with bis(tri-n-butyltin) oxide were analyzed
along with real samples; recoveries of 75 percent were obtained.
Air Samples. Approximately 2 m3 air samples were collected by pumping
air through silica traps (made in our laboratory) at a rate of 1 to 1.5
£/min for 24 hours.* These traps were capped and stored refrigerated until
analysis. Filters were placed upstream of the silica traps to collect
airborne particulate separately.
The organotins were concentrated and analyzed as in the procedure
used for water analysis. The results are shown in Table 13. No traces of
tin compounds were observed by atomic absorption analysis. In order to
evaluate our analytical method for air analysis, an experiment was carried
out in the field at Reading in which a silica trap was followed by a U-tube
filled with 20 yg of bis(tri-n-butyltin) oxide which was followed by an
additional silica trap. In this way we could check our system under actual
conditions by analyzing the recoveries of the second trap. However, the
results of this experiment showed essentially that 100 percent of the bis-
(tri-n-butyltin) oxide remained in the glass U-tube. This could be
attributed to adsorption on the glass surface of the tube due to long
residence time, in contrast to our laboratory experiment in which approxi-
mately 50 percent of the bis(tri-n-butyltin) oxide was volatilized in a
stream of nitrogen.
A representative particulate sample was analyzed in order to check for
deposited organotins. The filter was extracted with acetic acid and analyzed
directly. No tin was found.
Soil Samples. Soil samples were preserved by refrigeration prior to
analysis. The soil was air dried for 3 to 4 hours and 25 g of the dried soil
was soxhlet extracted in 75 ml acetic acid overnight. The acetic acid
extracted was then diluted to 200 ml with water. The particulate was
removed by centrifugation and the solution then neutralized to pH 7.
The 200-ml solution was then treated exactly as the water samples by
concentration on silica gel and analysis by atomic absorption spectroscopy.
The results are shown in Table 14. Once again these values, although the
highest obtained in this study, were below the limits at which we could
speciate and, therefore, are reported as total tin extractable with acetic
acid. Experiments on spiked soil samples gave recoveries of 40 to 50
percent, indicating a need for future work in this area.
Avondale Site
Geographical data and sampling locations on the Avondale site are shown
in Figure 15; a topographical map is presented in Figure 16.
*See appendix for description of sampling method.
41
-------
TABLE 13. ORGANOTIN ANALYTICAL DATA FOR AIR SAMPLES FROM READING, OHIO
U.S. Geological
Sampling Survey
Station Coordinates
1 721288 m E
4346000 m N
2 721552 m E
4345961 m N
3 721503 m E
4345796 m N
4 721360 m E
4345681 m N
5 720829 m E
4346196 m N
6 719652 m E
4346196 m N
Weather Conditions
Wind Temp
Organotin Analysis
Parti-
Direction, Speed, Range, culate, Vapor,
degrees m/sec F General Sampling Period I'g/m^ vig/m3
180-290 0.6-1.8 67-78 Sunny, intermittent 5/4/77
rain, winds variable 5/5/77
180-290 0.6-1.8 67-78 Ditto 5/4/77
5/5/77
180-290 0.6-1.8 67-78 " 5/4/77
5/5/77
180-290 0.6-1.8 67-78 " 5/4/77
5/5/77
180-290 0.6-1.8 67-78 " 5/4/77
5/5/77
247-290 0.6-1.8 67-81 " 5/4/77
5/5/77
1342 to <0.1
1342
1292 to <0.1 <0.1
1297
1267 to <0.1
1268
1225 to <0.1
1233
1175 to <0.1
1200
1033 to <0.1
1392
Values reported are for acetic acid-extractable tin.
-------
RESIDENTIAL
AREA
A WATER SAMPLE
Figure 15. Map of sampling sites for organotins at
Avondale Shipyard at New Orleans, Louisiana.
43
-------
Figure 16. Topographic map of sampling sites for organotins
at Avondale Shipyard, New Orleans, Louisiana.
-------
TABLE 14. ORGANOTIN ANALYTICAL DATA FOR SOIL
SAMPLES FROM READING, OHIO
Sampling
Station
1
2
3
4
5
6
U.S. Geological
Survey Coordinates
721288 m E
4346000 m N
721552 m E
4345961 m N
721503 m E
4345796 m N
721360 m E
4345681 m N
720829 m E
4345407 m N
719652 ra E
4346196 m N
Organotin Analysis,
ppba
120
64
15
60
25
60
Values reported are for acetic acid-extractable tin.
Water Samples. Grab samples of 1000 ml of water were collected and
analyzed by the procedures outlined in the Reading site. The results of
these analyses are shown in Table 15 along with information concerning
collection. Just as in the Reading samples, recoveries of 75 percent were
obtained on spiked samples. Essentially no organotin was found in these
water samples down to our detection limits.
Air Samples. Air samples were collected and analyzed by the procedures
outlined in the Reading site. The results of these analyses are shown in
Table 16. No traces of tin compounds were observed in either the vapor or
particulate samples. Spiking experiments at Avondale also showed that bis-
(tri-n-butyltin) oxide remained involatile during the sampling period.
Soil Samples. Soil samples were analyzed by the procedures outlined in
the Reading site. Unlike the Reading site, however, no tin compounds were
observed at the highest sensitivity of our method. These results are shown
in Table 17. Experiments of spiked samples of these soils also showed poor
recoveries ranging from 35 to 50 percent.
45
-------
TABLE 15. ORGANOTIN ANALYTICAL DATA FOR WATER
SAMPLES FROM AVONDALE, LOUISIANA
Sampling U.S. Geological
Station Survey Coordinates
Description
Organotin Analysis,
ppba
AU
AEF
AD
770154 m E
3313038 m N
772115 m E
3314019 m N
773250 m E
3315192 m N
Upstream of source on
Mississippi River
at loading dock
Effluent, composite
sample at launching
site
Downstream of plant
at Audubon Park
Values reported are for total acetic acid-extractable tin.
TABLE 17. ORGANOTIN ANALYTICAL DATA FOR SOIL
SAMPLES FROM AVONDALE, LOUISIANA
Sampling
Station
1
2
3
4
5
6
U.S. Geological Organotin Analysis,
Survey Coordinates ppba
771731 m E
3313058 m N <10
772942 m E
3314808 m N <10
770942 m E
3312884 m N <10
771462 m E
3313615 m N <10
771019 m E
3313038 m N <10
771038 m E
3313365 m N <10
Values reported are for total acetic acid-extractable tin.
46
-------
TABLE 16. ORGANOTIN ANALYTICAL DATA FOR AIR SAMPLES FROM AVONDALE, LOUISIANA
U.S. Geological
Sampling Survey
Station Coordinates
1 771731 m E
3313058 m N
2 772942 m E
3314808 m N
3 770942 m E
3312884 m N
4 771462 m E
3313615 m N
5 771019 m E
3313038 m N
6 771038 m E
3313365 m N
Weather Conditions Organotin Analysis3
Wind Temp Parti-
Direction, Speed, Range, culate, Vapor,
degrees m/sec F General Sampling Period ug/m3 Pg/m3
Calm-135 Calm-3.5 70-87 Clear, some rain 5/23/88 1667 to
latter fourth of 5/24/77 1750 <0.1 <0.1
sampling period
Ditto Ditto Ditto Ditto 5/23/77 1600 to
5/24/77 1925 <0.1
Calm-150 " " " 5/23/77 1850 to
5/24/77 1858 <0.1 <0.1
Calm-135 " " " 5/23/77 1717 to
5/24/77 1775 <0.1
Calm-150 " " " 5/23/77 1800 to <0.1
5/24/77 1825
Ditto " " " 5/23/77 1750 to <0.1
5/24/77 1900
Values reported are for acetic acid-extractable tin.
-------
CONCLUSIONS AND RECOMMENDATIONS
The speciation and analysis of organotins, and organometallics in
general, is a difficult problem. Because of our success in speciating
organotins, we have been contacted by other research groups concerning
our method. The method we have developed is effective but can and should
be improved.
The first improvement should be in the area of on-line analysis. To
do this, high-performance liquid chromatography combined with atomic
absorption spectroscopy is the method of choice. We would have liked to
pursue this method further but time did not permit it. The on-line HPLC-AA
method should have a number of advantages, including better recoveries, higher
sensitivity, and shorter analysis time.
Second, the recoveries from soil samples need to be improved. There
is still a need for a better method of recovering low levels (ppb) of
organotins from soils. If an effective concentration technique can be
established, then on-line HPLC-AA may provide a simple analysis technique
for measuring low levels of organometallics in environmental samples.
48
-------
PART III. ARYL PHOSPHATES
INTRODUCTION
In the past several years, there has been increasing concern about
environmental levels of many organic compounds. Aryl phosphates are a
class of organic compounds receiving careful study because of their known
involvement as neurotoxins. This study was undertaken to obtain data on
the environmental levels of selected aryl phosphates in the vicinity of
chemical facilities involved in the manufacture or use of aryl phosphates.
The specific aim of this study was to detect the presence (at the
ppb level) of selected aryl phosphates in environmental samples. The
aryl phosphates selected included the five aryl phosphates of greatest
commercial interest: tricresyl phosphate, triphenyl phosphate, cresyl
diphenyl phosphate, isopropylphenyl diphenyl phosphate, and 2-ethylhexyl
diphenyl phosphate. The detection method used had to be capable not only
of separating and individually detecting the aryl phosphates of interest,
but also of separating the various isomers present in a given sample.
LITERATURE REVIEW
Aryl phosphates have been analyzed by a number of different methods
which are conveniently classified into three groups: hydrolysis methods,
thin-layer or paper chromatographic methods, and gas chromatographic
methods.
Hydrolysis
Hydrolysis methods involve cleavage of the ester bond and analysis for
the resultant phenols, either spectrophometrically (Jaukari et al., 1973;
American Industrial Hygiene Association Journal, 1969) or by gas chroma-
tography (Murray, 1975). Hydrolysis methods are unacceptable for the
analysis of environmental samples because it is difficult to correlate the
amount of phenol detected with the concentration of aryl phosphate present.
In addition, the ability to analyze for a specific phosphate group is lost
using hydrolysis methods.
-------
Paper and Thin Layer Chromatographv
Both thin layer chromatography (Lamocte et al., 1969; Brun and Mallet,
1973) and paper chromatography (Firmin and Gray, 1974) have been used to
separate and analyze for aryl phospl.-tet,. These methods, however, lack the
sensitivity needed for analysis ur low levels of aryl phosphates in environ-
mental samples.
Gas Chromatography
Gas chromatographic methods have been used to successfully separate and
detect aryl phosphates and is the detection method of choice for environ-
mental samples. Detection of phosphorus-containing compounds has previously
been achieved using flame ionization (Bloom, 1973), mass spectrometry (Mieure
and Dietrich, 1973), and pizoelectric (Scheide and Guibault, 1972) detectors.
Although these detectors have achieved some success, a gas chromatographic
detection system specific for phosphorus compounds has been developed
(Frostling, 1973; Ave, 1971) which will allow highly sensitive and specific
detection of aryl phosphates. Although use of the flame photometric
detector has not been reported for detection of triaryl phosphates, it has
been highly successful for analysis of phosphate insecticides in environ-
mental samples (Ivey and Claborn, 1971; Ivey et al., 1973).
ANALYTICAL METHODS DEVELOPMENT
The analytical method of choice for the detection of tricresyl
phosphate, triphenyl phosphate, cresyl diphenyl phosphate, isopropylphenyl
diphenyl phosphate, and 2-ethylhexyl diphenyl phosphate in environmental
samples at the ppb level is gas chromatography using a flame photometric
detector.
The gas chromatography system enables resolution of tricresyl phosphate,
a complex mixture of aryl groups attached to phosphorus, into individual
peaks. Injection of specific tricresyl phosphate isomers showed that up to
14 compounds are present in commercial-grade tricresyl phosphate, all of
which can be resolved using the capillary GC/FPD system.
The gas chromatography system developed for these aryl phosphate
determinations was a Varian 1700 gas chromatograph equipped with a 25 m SE-30
capillary column and dual flame ionization and flame photometric detectors
(FID/FPD). The flame photometric detector used a 526 nm filter specific for
phosphorus. The gas flow was split 4:1 at the injector and 1:1 at the
detectors. Helium was the carrier gas (2 ml/rain flow through the column),
while nitrogen served as the make-up gas. The column conditions used were
175 C to 250 C at 2 C/min with an attenuation of 8 x 10~9.
By careful adjustment of the flame photometric detector's air, oxygen,
and hydrogen gas supplies, a maximum sensitivity of 50 ng of tricresyl
phosphate injected was achieved. With the flow splits on the gas
50
-------
chromatography system, a 50 ng injection was split 4:1 at the column to give
approximately 10 ng on the column and again split (1:1) at the detectors to
give a detection limit (at the detector) of approximately 5 ng. However, in
order to detect and quantify the individual isomers in commercial-grade
tricresyl phosphate, it was necessary to inject at least 200 ng of sample.
Care must be exercised in keeping the FP detector clean. Sensitivity
was rapidly lost if the mirrored surface of the detector or the glass plate
became clouded. Cleaning these surfaces once a week with a cottom swab was
found to be the most effective means of maintaining the necessary sensiti-
vity.
A dual FID/FPD system was used on the gas chromatograph in order to
monitor the performance of the capillary column. This dual detector system
was used to determine loss of sensitivity for both phosphorus- and
nonphosphorus-containing compounds.
Calibration of Gas Chromatograph
Calibration curves for the aryl phosphates under consideration were
determined using standard solutions of the selected aryl phosphates in
heptane. Figure 17 shows a gas chromatogram for four of the aryl phosphates
plus tris(butoxyethyl) phosphate, the internal standard. Table 18 shows the
retention times for the aryl phosphates.
TABLE 18. RETENTION TIME DATA FOR ARYL PHOSPHATES
Compound Retention Time(s), min
Triphenyl phosphate
2-Ethylhexyl diphenyl phosphate
Isopropylphenyl diphenyl phosphate
and isomers
Tricresyl phosphate and isomers
Tris(butoxyethyl) phosphate
23
25
23
32
32
35
25
.0
.3
.0,
.7,
5,
2,
.5
25
33
33
35
.0,
.2,
.5,
.5,
28
35
33
36
.3, 30.7,
.9
.9, 34.7,
.0, 36.6
aGas chromatograph conditions 175 C to 250 C at 2 C/min, 25 m
SE-30 column; carrier gas helium, 2 ml/min.
51
-------
Ul
ro
:-!._.-!..- Q~1L_J
i ; '.
; ~ : .. .'..f . .
. , g - .-:;/ -
- ! .: :- |
:-- -S .V--,i--:"-; ^ :
[Z ! :--: 33
- .;-- -
..-. i
--
n --'--!.-- - LZ i
- - : 1 ' ' i
l . . . . . . i ,
f
i
1 i^
' i - F
: i r :- : I./'.-..-..
4 ! 4 :
I J..!'!
. ..
S\
:ID[ ,
;
V ,t\.J ; ^_ ^ i i :;
WH*rt » <*Ji*ii"*'
^--.-.ri :...,....
^
i
.! -'^I.-T--.. v^v^.v-^v^-.'. -^ .j-^-. v/r^^ .
f f n ^
Hii
! - - - In f
1 =--- -L
;-^*r
rfn^^
; : :;-].'.. ; iff
< »!, __j _^
inject noise ippdp
1
i
1
-j
P
1
i
JEP
j
i
j
1
i
5
)
: . |
----r. -:: j
: .: ii;
f ;
r
.
f:~^
2eh *
ipp
i* i
i V
I- j:
.._...
-
|"
M
i
.
o ; :
"i cm
H
i »
l?
*'""/ ly
(jp ippdp
-". : :"'>
ippdp '
j
jfi
|l .
i! n^
yvW
i ! .
: i. : '
: ; - ;
- ! :
w5f
^-^ i - -
-
-U
- .::':.-!."::
.I.:-'.
- K : . :-^-
;j- : '""
. :.: :.::-
-: -r-
: .
H I
:WVfe
- -1 - im-
PV^ run * i *
:CP-^ ippdp-: ippdp ippdp
->I .-_ "."
-_
----- -
. . - .. -
- ^-_.ri..
~ ** |
-_-: . .
_.
.
.- ;- ' '-
: . _7^
:-:- . -: "
._ - .
» U,
: - . :
-^. :
T 1 "."" ~:
- . - :. .
- '- :
- '.- :.\ '
/1/>v^T>i'Mkr
:~ T( 71
.-- - . : .
* .:. .
ippdp ~._
-
1 : f.-! --
Figure 17. Gas chromatogram trace of standard triaryl phosphate mixture.
Gas chromatogram conditions were: 25 m SE-30 capillary column
operated from 175 C to 250 C at 2 C/min with a helium flow of
2 ml/min. Nitrogen was used as the make-up gas; attenuation
was 8 x 10~9.
TPP = triphenyl phosphate
TCP = tricresyl phosphate
ippdp = isopropylphenyl diphenyl phosphate
2eh = 2-ethylhexyl diphenyl phosphate
BEP = tris(butoxyethyl) phosphate
-------
The calibration curves for each of the aryl phosphates (triphenyl
phosphate, 2-ethylhexyl diphenyl phosphate, isopropylphenyl diphenyl
phosphate, and tricresyl phosphate) were found to be linear from 50 ng
to 2400 ng injected (5 ng to 300 ng detected).
Extraction Recoveries
Water, soil, and air (Tenax trap) samples known to be free of aryl
phosphates were spiked with known concentrations of triphenyl phosphate.
The water samples were distilled water or water from a well in a rural
area. Soil samples (Brookstone silty loam) came from a rural farm area
where no organo-phosphorus compounds had been used. Tenax traps were
spiked with the aryl phosphates to simulate air samples.
The breakthrough volume of aryl phosphates on Tenax appears to be
almost infinite. No breakthrough of a 1 yg sample of triphenyl phosphate
on 1 g of Tenax was observed as determined by gas chromatography at a
carrier gas (helium) flow rate of 30 ml/min after 24 hours at 60 C, an
additional 24 hours at 100 C and, finally, an additional 24 hours at 250 C.
Extractions were carried out according to the methods outlined in the
Analysis of Environmental Samples section of this report. Two concentra-
tions of triphenyl phosphate were used: 500 ng and 1.2 mg. Sample extrac-
tions were carried out in triplicate. Each sample extract was gas chroma-
tographed under the conditions outlined in the above section. Table 19
lists the extraction recoveries for each of the environmental sample types.
TABLE 19. EXTRACTION RECOVERIES FOR ARYL PHOSPHATES
Triphenyl Phosphate Recovery, %c
Sample Type 500 ng Spike 1.2 mg Spike
Water 82 85
Soil 85 85
Air (spiked Tenax) 83 87
a
Average of three samples.
53
-------
COLLECTION AND ANALYSIS OF ENVIRONMENTAL SAMPLES
Water
Grab samples of water were taken in duplicate at upstream and downstream
sites. Two consecutive 24-hour composite water samples were taken using an
automated sampler downstream as close as possible to the plant outfall.
Samples were collected 2 to 5 cm below the surface of the water. During grab
sampling, care was taken to avoid bubbling as the water entered the bottle.
Samples were collected in amber bottles fitted with Teflon cap liners which
had been previously cleaned with 10 percent hydrochloric acid, distilled
water, acetone (CP), and distilled water again in that order. Sample bottles
and the sampler were rinsed thoroughly in the water to be collected before
the samples were taken. Samples were stored at approximately 4 C until
analysis.
For each extraction, a 500 ml water sample was placed in a 1 H separa-
tory funnel. The internal standard was added [10 ug tris(butoxyethyl)
phosphate], and the water solution nearly saturated with 150 g of sodium
chloride to improve the extractability of water-soluble compounds. Once
the NaCl was dissolved, 15 ml of distilled-in-glass methylene chloride
(CH2Cl2> was added, the funnel was shaken, and the phases were allowed to
separate. The lower or CH2C12 phase was drained off through a cotton plug.
(There was no more than 10 ml of CH2C12 due to its solubility in water.)
The CH2C12 was evaporated using a Kuderna-Danish evaporator and the residue
was taken up in 150 y& of heptane. Heptane was the final solvent for all
environmental samples to prevent destruction of the capillary gas chromato-
graphy column by Cl^C^. The heptane solutions were sealed in vials with
septa caps and stored in a refrigerator until gas chromatography analysis.
Samples which showed none of the selected aryl phosphate peaks at the 150-
\il level were concentrated to 50 y£ and rerun.
Soil
Three 2-inch soil cores were taken in the vicinity of each air sampling
station. The corer was washed and rinsed with distilled water and acetone
between locations and also between soil types at a given location if necessary.
The core samples were transferred to amber glass jars with Teflon cap liners
and stored at 4 C until extraction.
For each extraction a 50 g soil sample was slurried with 100 ml of
CH2C12 to which internal standard [10 ug tris(butoxyethyl) phosphate] had
been added. The slurry was shaken for 30 minutes and then allowed to stand
for 1 hour. The CH2Cl2 layer was filtered off through double thicknesses
of filter paper: first through #44, then through //5. The Ct^Cl^ layer was
allowed to stand between filtrations to allow particulate to settle. Care
had to be exercised in filtering soil samples to ensure complete removal of
particulate. In some soil samples, centrifugation may have to be used to
remove particulate. After the final filtration, the CH2C12 solution was
evaporated to dryness in a Kuderna-Danish evaporator and the residue
54
-------
taken up in 200 \iH of heptane. The heptane solutions were sealed in vials
with septa caps and stored in a refrigerator until gas chromatography
analysis. Samples which showed none of the selected aryl phosphate peaks
at the 200 ufc level were concentrated to 50 yfl, and rerun.
Air
Approximately 3 m of air was pulled through 3/8 x 5 in. stainless steel
traps filled with 0.8 g of Tenax at approximately 1 Vmin f°r 48 hours. The
design of the air sampling system is described in the appendix. The traps
were closed with Swagelok cap fittings and stored at 4 C until extraction.
Particulate filters were stored at 4 C in Teflon-lined screw-cap vials.
Each Tenax trap used for air sampling was emptied into the thimble of a
Soxhlet extractor. Internal standard was added to the Tenax and the sample
was extracted overnight with 80 ml heptane. The heptane was evaporated to
200 y£ using a Kuderna-Danish evaporator. The heptane solutions were sealed
in vials with septa caps and stored in a refrigerator until gas chromatography
analysis. Samples which showed none of the selected aryl phosphate peaks at
the 200-U& level were concentrated to 50 vH and rerun. The particulate
filters were extracted individually in a similar manner and the extracts
concentrated to 50 \iSL.
Results
The results for water, soil, and air samples from the Gallipolis Ferry,
West Virginia, site are presented in Tables 20 through 22. Simulated and
topographic maps of the sampling stations are shown in Figures 18 and 19,
respectively. An aerial map of this site could not be obtained in time to
be included in this report. Duplicate samples were run for all analyses.
As can be seen from the tables, none of the selected aryl phosphates were
detected at levels above the respective detection limits in any of the
samples, including the particulate filters. Standard samples run at the
same time gave responses similar to those exhibited in Figure 17.
The results for water, soil, and air samples from the Lordstown, Ohio,
site are presented in Tables 23 through 25. Simulated, topographic, and
aerial maps of the site are shown in Figures 20, 21, and 22, respectively.
Duplicate samples were run for all analyses. As can be seen from the tables,
none of the selected aryl phosphates were detected at levels above the
respective detection limits in any of the samples, including the particulate
filters. Standard samples run at the same time gave responses similar to
those exhibited in Figure 17.
55
-------
TABLE 20. ARYL PHOSPHATE ANALYTICAL DATA FOR WATER
SAMPLES FROM GALLIPOLIS FERRY, WEST VIRGINIA
Sampling Analysis,
Site Sample Description ppba
A Grab sample, upstream Ohio River <10, <10
B 24-hour composite, plant outfall, <10, <10
April 21-22, 1977
B 24-hour composite, plant outfall, <10, <10
April 22-23, 1977
C Grab sample, downstream Ohio River <10, <10
Duplicate sample analyses.
TABLE 21. ARYL PHOSPHATE ANALYTICAL DATA FOR SOIL
SAMPLES FROM GALLIPOLIS FERRY, WEST VIRGINIA
Sampling
Site
1
2
3
4
5
6
Analysis ,
ppba
<100,
<100,
<100,
<100,
<100,
<100,
<100
<100
<100
<100
<100
<100
a
Duplicate sample analyses.
56
-------
TABLE 22. ARYL PHOSPHATE ANALYTICAL DATA FOR AIR SAMPLES
FROM GALLIPOLIS FERRY, WEST VIRGINIA
Weather Conditions
Wind Temp
Sampling Direction, Speed, Range,
Station degrees mph F
1 225-175 4.8-8.0 70-63
2 Ditto Ditto 78-68
3 " " 80-66
4 " " 84-68
5 " " 85-73
6 " " 83-68
Aryl Phosphate Analysis
Parti-
culate.3 Vapor
General
Calm-haze; drizzle
Light wind-haze; drizzle
Light breeze-partly
sunny; rain
Calm-sunny to rain
Bright to rain
Sunny to rain to cloudy
Sampling Period
4/21/77 0900 to
4/23/77 0910
4/21/77 1005 to
4/23/77 1400
4/21/77 1155 to
4/23/77 1150
4/21/77 1225 to
4/23/77 1200
4/21/77 0100 to
4/23/77 1235
4/21/77 1335 to
4/23/77 1300
yg/m3 ppb
<0.2
<0.2
<0.2
<2 <0.2
<2 <0.2
<2 <0.2
Ug/md
<2
<2
<2
<2
<2
<2
Detection limit: 1.0 ug/m3.
Detection limit: 1.0 ppb.
-------
Ln
CC
IIGHWAY
RAILROAD
PLANT SITE
KILOMETER
SOIL/AIR SAMPLI
WATER SAMPLE
Figure 18. Map of sampling sites for aryl phosphate at
Gallipolis Ferry, West Virginia.
-------
f£ 1 / + W~ is <
lii
jr/jj
s'7 '
;/
1
.'1 ' Oi
§
' S3
O
i
i
1
. 1
, 1
1
i\r
m.
' TA\ >
/ i
+ ^A
r i ^ ^? 5|
i i
\0iilliix M ^
!;;': '/ If /^D^
===.«, n, v ( i JBti
/ "^ :/! ^ < yv>^JJ
/ . V ' I > vC jy
I I LU ^^4=^
\i I tr ^*ic^
#6 V '" ' °-
""^; - ->
^
g
^ _i
- -I
** w
i t"( m ' -"* ' ' ' >
*/..."' . i -/ \
J-'-, n ' i Sut--!at.on' .' ' -. vj -. . -. >
" '! i:
Gravel
I'll A '
.. ,'--,.-=.-. .,= _.Js« ... .". ..,-
i r
. ,
\ :J4
\ - . . * . . '
* \ , -!J
) PLANT SITE r L.N "
\ T I "'n*'
: ^.^ \X'
*". \ ' / ' v. --x . ' x-
VA .', '-, "\" . ;V".-
*r^' ,yHVV-? A
\ | ,. &«V#P 1^\\- / *) 4^"'
\'V-\\ > I Gaiiij!uiis";.:,y ..-'. -vX L-. \--f%'J f
«" \\ \ \ l_ l'""I'ilJ-VJi: -:.'-\ . .,_#1 V:\ r>sf^, ''',/:.-.' ':-\ . .'-"'
». (^
' / ' :'
)
Y ( '-
\ (
^"N. *' .
/
: ('.
0
B. , t - « .-.--" . . . /
. - - " " *l'l-'|](- '» /V ' \ '" /
> '^XXxyv v #2 -''\.- \ \
'""; ' ljbl >s^- x^-. ' "-A'"'- '' \* ' -''--''/'
...; ; cli|"":
-------
TABLE 23. ARYL PHOSPHATE ANALYTICAL UilA FOR
WATER SAMPLES FROM LORDSTOWN, OHIO
Sampling
Site Sample Description
A
B
B
C
D
E
i
Grab sample, surface water of farm pond
24-hour composite, plant effluent,
May 12-13, 1977
24-hour composite, plant effluent,
May 13-14, 1977
Grab sample, surface water of farm pond
Lake inlet
Surface water, lake
Motel tap water
Analysis,
PPba
-------
TABLE 25. ARYL PHOSPHATE ANALYTICAL DATA FOR AIR SAMPLES FROM LORDSTOWN, OHIO
Sampling
Station
1
2
3
4
5
6
Weather Conditions
Wind Temp .
Direction, Speed, Range,
degrees mph F
270-calm 1.7 76-67
Ditto " 73-66
70-63
77-66
73-64
73-64
Aryl Phosphate
Cloudy
Cloudy
Cloudy
Sunny
Partly
Partly
General
to sunny
to sunny
, drizzle
to drizzle
cloudy to drizzle
cloudy to sunny
Sampling
5/12/77
5/14/77
5/12/77
5/14/77
5/12/77
5/14/77
5/12/77
5/14/77
5/12/77
5/14/77
5/12/77
5/14/77
Period
1300 to
1105
1230 to
1045
1330 to
1120
1120 to
10JO
1000 to
1150
1040
1010
Parti-
culate,3
g/m
<2
<2
<2
Analysis
Vapor
ppb
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
g/m
<2
<2
<2
<2
<2
<2
Solvent extracts of air particulate filters at downwind sites 4, 5, and 6 were analyzed. No aryl
phosphates were detected in these samples and for this reason the other particulate filters were
not analyzed.
-------
PREVAILING WINDS
PLANT SITE
SOIL/AIR SAMPLE
A WATER SAMPLE
Figure 20. Map of sampling sites for aryl phosphates at Lordstown, Ohio.
-------
LO
Duck C
>-';/ 'BMruucKurccK i ; . . ^_ .j^^/^o^j. _ n;
."^At7:s^//vJ:" "> -' '.'.. aOA i- i "..'..-...; ' -i "'
/ .
/ ' I
*"»-l 1 -
' I t'
! ^"»
. jr.,
i
f)| PLANT SITE
l'i__
A
f)| PLANT SITE | \ \ /-~\t\^./-
I : ~V~ ^^.* ..^,-J^ --^ --'/,/
c&L' .jl^L^-j
-ioO° ';/
\\^
X
;
;/x^.__.,xt
PREVAILING WINDS
SOIL/AIR SAMPLE
A WATER SAMPLE
KILOMETER
0 .5 1
Figure 21. Topographic map of sampling sites for aryl phosphates at Lordstown, Ohio.
-------
PREVAILING WINDS
SCALE: 1 INCH= 1667'
A WATER SAMPLE
| SOIL/AIR SAMPLE
PLANT SITE BOUNDARY
Figure 22. Aerial map of sampling sites for aryl phosphates
at Lordstown, Ohio (*D and E sampling sites 0.5
mile farther east).
64
-------
REFERENCES
Adcock, L. H. and W. G. Hope. 1970. A Method for the Determination of Tin
in the Range of 0.2 to 1.6 yg, and Its Application to the Determination of
Organotin Stabilizer in Certain Foodstuffs. Analyst (London) £5(1135):868-
874.
Aldridge, W. N. and J. E. Cremer. 1957. Organo-tin Complexes.
Analyst (London) 82_: 37-43.
American Industrial Hygiene Association Journal. 1969. Triphenylphosphate.
30:435.
Ave, W. A. 1971. Pesticide Identification at the Residue Level; a
Symposium. In: Advances in Chemistry Series, 104. R. F. Gould (ed.).
American Chemical Society, Washington, D.C. 182 p.
Beesing, D. W., W. P. Tyler, D. M. Kurtz, and S. A. Harrison. 1949.
Determination of Acrylonitrile and Alpha, Beta-Unsaturated Carbonyl
Compounds. Analytical Chemistry 2.1(9): 1073-1076.
Belcher, R. and B. Fleet. 1965. Submicro-Methods for the Analysis of
Organic Compounds. Part XIX. Determination of Olefinic Unsaturation.
Journal of the Chemical Society (London), Part II. pp. 1740-1746.
Betso, S. R. and J. D. McLean. 1976. Determination of Acrylamide Monomer
by Differential Pulse Polarography. Analytical Chemistry 48(4);766-775.
Bloom, P. J. 1973. Application of Thin-Layer and Gas-Liquid Chromatography
to Qualitative and Quantitative Analysis of Phosphoric and Phosphorous Acid
Esters. Journal of Chromatography 75(2):261-269.
Booth, M. D., M.J.D. Brand, and B. Fleet. 1970. A Fully Automatic
Apparatus for Stripping Voltammetry. Application to the Determination of
Triphenyltin Compounds. Talanta 17(11);1059-1065.
Botre, C., F. Cacace, and R. Cozzani. 1976. Direct Combination of High-
Pressure Liquid Chromatography and Atomic Absorption for the Analysis of
Metallorganic Compounds. Analytical Letters 9_(9) :825-830.
Brinckman, F. E., W. P. Iverson, and W. Blair. 1976. Approaches to the
Study of Microbial Transformations of Metals. Proceedings of the Third
International Biodegradation Symposium. J. M. Sharpley and A. M. Kaplan
(eds.). Applied Sciences Publishers, Ltd., London, pp. 919-936.
65
-------
Brun, G. L. and V. Mallet. 1973. Detection of Organophosphorous Pesticides
by in situ Fluorometry on Thin-Layer Chromatograms. Journal of Chromatography
80(1):117-123.
Cassol, A., L. Magon, and R. Barbieri. 1965. Complexes of Organometallic
Compounds. IX. Investigation of Organotin-Fluoride Systems by Means of
Anion Exchange Paper Chromatography and Paper Electrophoresis. Journal of
Chromatography 19:57-63.
Chemical and Engineering News. 1977. Acrylonitrile Linked to Cancer in
Workers. 55_(22):6.
Chromy, L. and K. Uhacz. 1968. Antifouling Paints Based on Organotin
Compounds. Part I. Colorimetric Determination of Microgram Amounts of
Organotin Compounds in Aqueous Solutions. Journal of Oil and Colour
Chemists Association .51(6) :494-498.
Corbin, H. B. 1973. Rapid and Selective Pyrocatechol Violet Method for
Tin. Analytical Chemistry 45_(3):534-537.
Corbin, H. B. 1970. Separation and Determination of Trace Amounts of Tin
Present as Organotin Residues on Fruits. Journal of the Association of
Official Analytical Chemists 5_3(1) : 140-146.
Croll, B. T., G. M. Arkell, and R.P.J. Hodge. 1974. Residues of Acrylamide
in Water. Water Research 8_:989-993.
Croll, B. T. 1971. The Determination of Acrylamide in Polyelectrolytes
by Extraction and Gas-Chromatographic Analysis. Analyst (London) 96
(1138):67-77.
Croll, B. T. and G. M. Simkins. 1972. The Determination of Acrylamide in
Water by Using Electron-capture Gas Chromatography. Analyst (London) 97;
281-288.
Dillard, C. R. 1971. Analysis of Organotin Compounds. In: Organotin
Compounds. A. K. Sawyer (ed.). Marcel Dekker, New York. pp. 997-1006.
Engberg, A. 1973. A Comparison of a Spectrophotometric (Quercetin)
Method and an Atomic-Absorption Method for the Determination of Tin in Food.
Analyst (London) 98_: 137-145.
Figge, K. 1969. Thin-Layer Chromatography of N-alkyltin Compounds on
Silanized Silica Gel. Journal of Chromatography .39_(1) :84-87 (in German).
Filer, T. D. 1971. Fluorometric Determination of Submicrogram Quantities
of Tin. Analytical Chemistry 43J13) :1753-1757.
Firmin, J. L. and D. 0. Gray. 1974. A Versatile and Sensitive Method for
the Detection of Organic Acids and Organic Phosphates on Paper Chromatograms.
Journal of Chromatography 94_:294-297.
66
-------
Freeland, G. N. and F. M. Hoskinson. 1970. Non-Aqueous Atomic-Absorption
Spectrophotometric Determination of Organometallic Biocides. Analyst
95_(1131): 579-582.
Frostling, H. 1973. Continuous Detection of Vapours and Aerosols of
Organophosphorus Compounds by Flame lonization-Flame Emission. Journal of
Physics E. Scientific Instruments. 6^:863-867.
Gauer, W. 0., J. N. Seiber, and D. G. Crosby. 1974. Determination of
Organotin Residues from Plictran in Fruit Crops by Gas-Liquid Chromatography .
Journal of Agricultural and Food Chemistry 22(2) ;252-254.
Geissler, H. and H. Kriegsmann. 1965. Separation Solvents for the Gas
Chromatographic Investigation of Certain n-Butyltin Compounds. Zeitschrift
fuer Chemie _5(11) :423-424 .
George, G. M. , M. A. Albrecht, L. J. Frahm, and J. P. McDonnell. 1973.
Atomic Absorption Spectrophotometric Determination of Dibutyltin Dilaurate
in Finished Feeds. Journal of the Association of Official Analytical
Chemists 56(6) : 1480-1482.
Goodyear Tire and Rubber Company. 1977. Telephone communication with
Robert Lattimer, Canton, Ohio.
Issaq, H. J. and E. W. Barr. 1977. Combined Thin-Layer Chromatography/
Flameless Atomic Absorption Method for the Identification of Inorganic
Ions and Organometallic Complexes. Analytical Chemistry 49(1) ;189-190.
Ivey, M. D., D. D. Oehler, and H. V. Claborn. 1973. Gas-Liquid Chromato-
graphic Determination of Chlorfenvinphos in Milk, Eggs, and Body Tissue
of Cattle and Chickens. Journal of Agricultural and Food Chemistry 21(5) :
822-824.
Ivey, M. C. and H. V. Claborn. 1971. Gas-Liquid Chromatographic Determina-
tion of Ronnel (0, 0-Dimethyl 0-2,4,5-Trichlorophenyl Phosphorothioate) and
the Oxygen Analog of Ronnel (Dimethyl 2,4,5-Trichlorophenyl Phosphate) in
Tissues of Cattle. Journal of Agricultural Food and Chemistry 19(6) ;1256-
1258.
Jaukari, M. , S. M. Chatterjee, and P. K. Ghosh. 1973. Indian Academy of
Forensic Sciences 12_(1):28.
Jehring, J. and M. Mehner. 1963. Determination of (C^g^SnCl in
SnCl2 by a.-c. Polarography. Zeitschrift fuer Chemie J3(l):34-35.
Jeltes, R. 1969. Determination of Bis(Tributyltin)Oxide in Air by Atomic
Absorption Spectroscopy or Pyrolysis Gas Chromatography. Annals of
Occupational Hygiene 12(4) : 203-207.
Jones, D. R. and S. E. Manahan. 1975. Atomic Absorption Detector for
Chromium Organometallic Compounds Separated by High-Speed Liquid Chromatog
raphy. Analytical Letters j3(8) :569-574.
67
-------
Jones, D. R., and S. E. Manahan. 1976a. Aqueous Phase High-Speed Liquid
Chromatographic Separation and Atomic Absorption Detection of Amino,
Carboxycycliccopper chleates. Analytical Chemistry 48/3):502-505.
Jones, D. R., and S. E. Manahan. 1976b. Detection Limits for Flame Spectro-
photometric Monitoring of High-Speed Liquid Chromatographic Effluents.
Analytical Chemistry 48^(13) : 1897-1899.
Jones, D. R., H. C. Tung, and S. E. Manahan. 1976. Mobile Phase Effects
on Atomic Absorption Detectors for High Speed Liquid Chromatography.
Analytical Chemistry 48(1);7-10.
Lamotte, A., A. Francina, and J. C. Merlin. 1969. Thin-Layer Chromato-
graphy of Organophosphorous Compounds. II. Chromatography of Neutral
Compounds. Journal of Chromatography 44^1):75-80 (French).
Luijten, J.G.A. (comp.). 1970. A Bibliography of Organotin Analysis. Tin
Research Institute. No. 417. Tin Research Institute, Columbus, Ohio. 48 pp.
MacWilliams, D. C., D. C. Kaufman, and B. F. Waling. 1965. Polarographic
and Spectrophotometric Determination of Acrylamide in Acrylamide Polymers
and Copolymers. Analytical Chemistry 37(12):1546-1552.
Mattocks, A. R. 1968. Spectrophotometric Determination of Pyrazolines and
Some Acrylic Amides and Esters. Analytical Chemistry 40_(8) : 1347-1349.
Meinema, H. A., et al. 1977. Determination of Trace Amounts of Butyltin
Compounds in Aqueous Systems by Gas Chromatography/Mass Spectrometry.
Presented at the Second International Conference on the Organometallic
Coordination Chemistry of Germanium, Tin, and Lead. July, 1977.
Meranger, J. C. 1975. A Rapid Screening Method for the Determination of
Di-(n-octyl) Tin Stabilizers in Alcoholic Beverages, Using a Heated
Graphite Atomizer. Journal of the Association of Official Analytical
Chemists 58/6):1143-1146.
Mieure, J. P. and M. W. Dietrich. 1973. Determination of Trace Organics
in Air and Water. Journal of Chromatographic Science 11(11);559-570.
Murray, D.A.J. 1975. Analysis of Tri-aryl Phosphate Esters and the
Determination of Imol S-140 in Fish Tissue and Water Samples by Gas
Chromatography. Journal of the Fisheries Research Board of Canada 32(4);
457-460.
Narita, H., N. Uchino, and S. Machida. 1963. J. Soc. Text. Cellul. Ind.
Japan 19:225.
Norris, M. V. 1967. Acrylamide. In: Encyclopedia of Industrial Chemical
Analysis. F. D. Snell and C. L. Hilton (eds.). Interscience Publishers,
New York, New York. pp. 169-168.
Parris, G. E., W. R. Blair, and F. E. Brinkman. 1977. Chemical and Physical
Considerations in the Use of AA Detectors with AGC. Preprint.
68
-------
Putnam, R. C. and H. Pu. 1965. Retention Indexes of Organotins. II.
Journal of Gas Chromatography J3(9):289-293.
Rapaport, L. I. and N. G. Ledovskikh. 1972. lodochlorometric Determination
of Trace Quantities of Acrylamide and Methyl Methacrylate. GIG SANIT 37(3):
74-77. Russian. Biological Abstracts 56(46418)(1973).
Sawyer, R. 1967. Determination of Dialkyltin Stabilizers in Aqueous
Extracts from PVC and Other Plastics. Analyst 92(1098);569-574.
Scheide, E. P. and G. G. Guilbault. 1972. Piezoelectric Detectors for
Organophosphorous Compounds and Pesticides. Analytical Chemistry 44(11) ;
1764-1768.
Schmoetzer, G. 1971. Determination of Acrylamide and Acrylic Acid in
Acrylamide Polymers. Chromatographia 4^(9):391-395 (German).
Simpson D. and B. R. Currell. 1971. The Determination of Certain Anti-
oxidants, Ultraviolet Absorbers and Stabilizers in Plastics Formulations by
Thin-Layer Chroraatography. Analyst 96(1144):515-521.
Skeel, R. T. and C. E. Bricker. 1961. Spectrophotometric Determination of
Dibutyltin Dichloride. Analytical Chemistry J33 (30:428-431.
Steinmeyer, R. D., A. F. Fentiman, and E. Jack Kahler. 1965. Analysis of
Alkyltin Bromides by Gas Liquid Chromatography. Analytical Chemistry _37/4) :
520-523.
Stoveken J. and F. Fernzndez. Perkin Elmer Separation Report No. 120.
Thomann, P., W. P. Koella, G. Krinke, J. Petermann, F. Zak, and R. Hess.
1974. The Assessment of Peripheral Neurotocivity in Dogs: Comparative
Studies with Acrylamide and Clioquinol. Agents & Actions 4^47-53.
Tonge, B. L. 1965. The Gas Chromatographic Analysis of Butyl-, Octyl-, and
Phenyl-tin Halides. Journal of Chromatography 19:182-184.
Vernon, F. 1974. The Fluorimetric Determination of Triphenyltin Compounds.
Analytica Chimica Acta (Amsterdam) 71(1);192-195.
Williams, A. I. 1973. The Determination of Bis(tri-m-butyltin) Oxide and
Di-n-butyltin Oxide in Preserved Softwood by Atomic Absorption Spectro-
photometry and Polarography. Analyst 98(1165);233-242.
Williams, D. J. and J. W. Price. 1964. The Paper Chromatography of Some
Organo-tin Compounds. Part II. Reversed Phase Systems. Analyst 89:220-222.
Woggon, H., H. Sauberlich, and W. J. Uhde. 1972. Determination of Organotin
Compounds and Their Degradation and Decomposition Products by Anodic Strip-
ping. Zeitschrift fuer Analytische Chemie 260(4);268-274 (German).
69
-------
APPENDIX
AMBIENT AIR SAMPLING SYSTEM
-------
APPENDIX
AMBIENT AIR SAMPLING SYSTEM
The system shown in Figure A-l was used to sample for acrylamide,
organotins, and aryl phosphates in ambient air. The system permits
concurrent sampling with up to four 0.95-cm-O.D. (3/8 in.) x 30-cm (12
in.) U-tube traps or other trap configurations. Two traps may also be
connected in series to check for breakthrough during sampling. Flow
through each trap is controlled by calibrated jewel orifice. A flow
meter is used to monitor total system flow. Sampling system vacuum is
monitored on a gauge and controlled by a vacuum relief valve. A Cast
carbon-vane, continuous-duty pump is used to provide air flow through the
system. The system is contained in a weatherproof housing. Air inlet to
the system is through a filter mounted about 5 feet above ground level.
Six systems were constructed for field use. Necessary spare parts were
available to repair or replace system components in the field, if neces-
sary.
For acrylamide, a midget impinger (EPA Method 6 type) containing 25 ml
water and 10 ml methanol was used to prevent freezing. Make-up solution
was added every 8 hours to replace that lost by evaporation.
Air samples for organotin analyses were collected on silica gel
tubes, 14-cm x 8-mm I.D., containing two sections of 30/60 mesh silica
gel separated by a stainless steel wire screen. The silica gel was
Analabs1 gas chromatographic grade, 30/60 mesh, 82-003060-00. The
absorbing section of the tubes contained 3.5 g and the backup section
1 g of the absorbent.
Air samples for aryl phosphate analyses were collected on 0.8 g of
Tenax packed in 3/8 x 5-in. stainless steel tubes. The flow rate was
approximately 1 £./min.
A-l
-------
TRAP
MANIFOLDS
47mm
FILTER
^ JEWEL OR/FICE
r
THERMOMETER
TRAP
TRAP ASSEMBLY DEfAiL
VACUUM
FLOVME7ER
VACUUfl CONTROL
VALVE
? C,AST PuriP
Figure A-l. Ambient Air Sampling System.
A-2
------- |