EPA-670/4-74-004
June 1974
Environmental Monitoring Series
ESTIMATION OF
POLYCHLORINATED BIPHENYLS
IN THE PRESENCE
OF DOT-TYPE COMPOUNDS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/4-74-004
June 1974
ESTIMATION OF POLYCHLORINATED BIPHENYLS
IN THE PRESENCE OF DDT-TYPE COMPOUNDS
By
J. T. Brownrigg and A. W. Hornig
Baird-Atomic, Inc.
Government Systems Division
Bedford/ Massachusetts 01730
Program Element No. 1BA027 (16020 GIY)
Project Officer
Dwight G. Ballinger
Methods Development and Quality Assurance Research Laboratory
National Environmental Research Center
Cincinnati, Ohio 45268
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National Environmental Research Center—Cincinnati
has reviewed this report and approved its publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommenda-
tion for use.
ii
-------
FOREWORD
Man and his environment must be protected from the
adverse effects of pesticides, radiation/ noise and other
forms of pollution, and the unwise management of solid
waste. Efforts to protect the environment require a
focus that recognizes the interplay between the components
of our physical environment—air, water, and land. The
National Environmental Research Centers provide this
multidisciplinary focus through programs engaged in
• studies on the effects of environmental
contaminants on man and the biosphere,
• the development of efficient means of
monitoring these contaminants, and
• a search for ways to prevent contamination
and to recycle valuable resources.
The investigation reported herein was conducted for
the National Environmental Research Center—Cincinnati to
explore the use of a new technique for the identification
and measurement of polychlorinated biphenyls in the presence
of similar organic compounds. The study established the
experimental conditions necessary for the detection of PCB
compounds, the sensitivity of the determination, and the
applicability of the method to natural water examination.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
iii
-------
ABSTRACT
Earlier studies suggested that the low temperature lumi-
nescence properties of PCB's and DDT compounds could be
used to identify these compounds singly or in mixtures.
The present investigation was undertaken to develop a
relatively simple, rapid method for estimating these com-
pounds in water. The emphasis in this procedure has been
on the inherent sensitivity and specificity of luminescence,
avoiding chemical separation where possible.
The present procedure involves collection of grab samples
followed by extraction, drying, concentration, and redilu-
tion in a second solvent suitable for luminescence measure-
ment at 77°K. Studies included the determination of
recoveries and detection sensitivities for some of the
compounds of interest and also analyses of several environ-
mental waters.
Detection limits for p,p'-DDT and Aroclor 1254 doped in
1-liter samples of pure water were found to be approximately
0.5 and 0.03 ppb respectively. Sensitivities were reduced
by an order of magnitude or more in natural waters having
high levels of dissolved organic material and particulates.
This is due to a combination of poorer recoveries and
increased fluorescence background. Both of these remain
as problem areas deserving further study. Phthalic acid
esters have spectral features resembling certain Aroclors
and may constitute an interference.
This report was submitted in fulfillment of Program
Element No. 16020 GIY and Contract/Grant No. 68-01-0082
by Baird-Atomic, Inc. under the sponsorship of the Environ-
mental Protection Agency. Work was completed as of February,
i «7 / 4t •
IV
-------
CONTENTS
Section Pag<
I Conclusions 1
II Recommendations 3
III Introduction 5
IV Low Temperature Emission Spectroscopy 6
V Aroclor-Pesticide Solutions in
Methylcyclohexane 9
Sources 9
Solvents 9
Solvent Purification 11
Spectra of Pesticides, Aroclors
and their Mixtures 12
VI Aroclor 1254 and p,p'-DDT in Pure Water 21
Sources 21
Extraction Procedure 22
Concentration of Extracts 23
Analysis of Extracts 24
Recovery & Extraction Efficiency 24
Detection Sensitivities 25
VII Environmental Samples 26
Sampling Locations 26
Collection 27
Extraction & Concentration 28
Spectral Analysis 28
River Water Doped with Aroclor
1254 and p,p'-DDT 32
Evidence of PCB/DDT in Natural Waters 35
Background & Interferences 36
Accuracy & Precision 39
VIII Summary
Spectra of Aroclors, Pesticides,
and their Mixtures 42
Aroclor 1254 and DDT in Pure Water 43
Aroclor 1254 and DDT Doped into
Natural Water 44
Environmental Water Samples 44
Methodology 45
IX References 75
X Appendix—Method 78
1. Equipment & Chemicals 78
2. Solvents and Purification 79
3. Intensity Standard & Instrument
Optimization 80
4. Standard Solutions in MCH 83
5. Doped Water Samples 86
6. Environmental Water Samples 87
7- Accuracy, Sensitivity &
Possible Interferences 88
v
-------
LIST OF FIGURES
No. Page
1 Schematic Diagram of Molecular Triplet and
Singlet Energy Levels. 47
2 Biphenyl Emission in Methylcyclohexane
(MCH) and in Heptane. 48
3 Wavelength Calibration Curves for the SF-100
Fluorispec Used in This Study. 48
4 p,p'-DDT, 10 ppm in MCH, 77°K. 49
5 p,p'-DDD, 10 ppm in MCH, 77°K. 49
6 Analytical Curves for p,p'-DDT and
Several Aroclors. 50
7 Aroclor 1016, 1 ppm in MCH, 77°K. 51
8 Aroclor 1048, 1 ppm in MCH, 77°K. 51
9 Aroclor 1254, 1 ppm in MCH, 77°K. 52
10 Aroclor 1254, 100 ppm in MCH, 77°K. 52
11 Aroclor 1254 (0.9 ppm) + p,p'-DDT (0.1 ppm)
in MCH, 77°K. 53
12 Aroclor 1254 (9 ppm) + p,p'-DDT (1 ppm)
in MCH, 77°K. 53
13 Aroclor 1254 (0.5 ppm) + p,p'-DDT (0.5 ppm)
in MCH, 77°K. 54
14 Aroclor 1254 (5 ppm) + p,p'-DDT (5 ppm)
in MCH, 77°K. 54
15 Aroclor 1254 (0.1 ppm) + p,p'-DDT (0.9 ppm)
in MCH, 77°K. 55
16 Aroclor 1254 (1 ppm) + p,p'-DDT (9 ppm)
in MCH, 77°K. 55
vi
-------
No. Page
17a,b Shawsheen River (2/21/73) Extract in MCH,
77°K (Acidified Sample). 56
18a,b Shawsheen River (2/21/73) Extract in MCH,
77°K (Sample Not Acidified). 57
19a,b Shawsheen River (2/21/73) Extract in MCH,
Room Temperature (Acidified Sample). 58
20a,b Shawsheen River (5/24/73) Extract in MCH,
77°K. 59
21a,b Shawsheen River (6/25/73) Extract in MCH,
77°K. 60
22a,b Shawsheen River (9/26/73) Extract in MCH,
77°K. 61
23a,b Concord River (8/23/73) Extract in MCH,
77°K. 62
24a,b Atlantic Ocean (10/2/73) Extract in MCH,
77°K. 63
25a,b Atlantic Ocean (10/2/73 Extract in MCH,
Room Temperature. 64
26a,b Milwaukee River(10/12/73) Extract in MCH,
77°K. 65
27a,b Milwaukee River (10/12/73) Extract in MCH,
77°K. Original Extract Diluted 10-fold
with MCH. 66
28a,b Milwaukee River (10/12/73) Extract in MCH,
77°K. Detail of Structure Resembling
Benzo(a)pyrene. 67
29a,b Charles River (12/6/73) Extract in MCH,
77°K. 68
30 Pacific Ocean (9/12/72). Gelbstoff
Luminescence at Room Temperature (from
Hornig and Eastwood, 1972). 69
31 Shawsheen River (5/24/73) Extract in MCH,
77°K. Chlorophyll Luminescence. 69
vii
-------
No. . Page
32 Pyrene, 1 ppb in MCH, 77°K 70
33 Benzo(a)pyrene, 0.2 ppm in MCH, 77°K. 70
34a,b Shawsheen River (9/26/73) Extract of Water
Doped with 6 ppb Aroclor 1254. 71
35a,b Shawsheen River (9/26/73) Extract of Water
Doped with' 60 ppb Aroclor 1254. 72
36a,b Shawsheen River (9/26/73) Extract of Water
Doped with 60 ppb p,p'-DDT. 73
37 Diisodecyl phthalate, 1 ppm in MCH,
77°K. 74
38 Dibutyl phthalate, 1 ppm in MCH, 77°K. 74
Vlll
-------
ACKNOWLEDGMENTS
The authors would like to thank Mr. D. G. Ballinger of
the Methods Development and Quality Assurance Research
Laboratory, National Environmental Research Center, U.S.
Environmental Protection Agency, Cincinnati, Ohio, for
his interest and encouragement.
Samples of Aroclors and related technical information
were supplied by Mr. W. B. Papageorge and Dr. E. S. Tucker
of the Monsanto Company, St. Louis, Missouri, and their
assistance in this regard is greatly appreciated.
We particularly wish to thank Dr. Bonnie Dalzell, Mrs.
Geraldine Garnick, and Mrs. Geraldine Wiesen for valuable
technical assistance with the laboratory studies and
with the preparation of this report.
IX
-------
SECTION I
CONCLUSIONS
1. The present study indicates that, using an extraction
procedure combined with luminescence measurement at 77°K,
PCB's (Aroclors) and DDT derivatives can be determined
at sub-part per billion levels in water. Detection
sensitivities for DDT-type compounds are poorer than
for Aroclors because of inherently weaker phosphorescence
intensity.
2. Detection limits for PCB/DDT in natural water samples
having high levels of fluorescent materials and sus-
pended particulates are more than an order of magnitude
worse than in pure water. These limitations could
probably be removed by the inclusion of a simple chro-
matographic step and/or the use of improved extraction
techniques.
3. Exploratory studies of some phthalic acid esters
(phthalates) show that these compounds are phosphorescent
at low temperature. The spectral features of these
compounds resemble those of certain Aroclors and might
therefore constitute an interference.
4. The method described, which utilizes commercially avail-
able instrumentation, should be useful as a screening
method in either a mobile laboratory or small field
station.
5. Although the procedure developed here does not include
coupling with either gas or liquid chromatography, low
temperature luminescence measurement could in principle
be used profitably with either procedure. In the
simplest procedure, specific effluents (or groups of
effluents) could be analyzed for confirmatory measure-
ments. Ultimately, it would be very desirable to
develop a low temperature luminescence detector to be
used in conjunction with a gas chromatographic method.
6. The method developed in this study is, in principle,
extensible to other commercial instrumentation. The
double monochromator instrument used in this study is
particularly applicable to the high-scattering geometry
-------
inherent in low temperature dewar systems. Single
monochromator instruments are expected to have higher
optical efficiency. It is difficult to predict whether
overall detection sensitivity will be increased or
decreased because of the offsetting effects of higher
efficiency and increased scatter.
-------
SECTION II
RECOMMENDATIONS
The recommendations for future work involved both the study
of improved collection/extraction methods and means of min-
imizing interferences.
1. Recently developed extraction techniques employing
polymer resins should be evaluated for compatibility
with low temperature luminescence (LTL) analysis.
Since much larger volumes of water can be processed/
detection limits may be substantially reduced.
2. Means of adapting the present method to suspended
particulates should be investigated, since substan-
tial amounts of water pollutants may be solublized
or transported in this form. For the same reasons,
possible extension to the analysis of bottom sediments
should also be considered.
3. Simple separation steps, such as silica gell chroma-
tography, should be tested for ability to remove
interfering luminescent substances present in natural
water. This would result in greatly improved detec-
tion sensitivities for PCB/DDT compounds.
4. Spectra of commonly used phthalic acid esters
(phthalates) should be further documented, and their
phosphorescence lifetimes measured. Preliminary
studies indicate that these compounds have spectra
resembling those of certain Aroclors, and therefore
may constitute an interference. However, the phos-
phorescence lifetimes may be substantially longer
than those of highly chlorinated PCB's and this might
permit temporal separation.
5. Low temperature luminescence spectra of a wide variety
of aromatic compounds should be obtained, preferably
in a common solvent. These data would be of help not
only to establish possible interferences for PCB/DDT
compounds, but would be useful for purposes of compound
identification or confirmation. Compounds studied
initially should be those which have been identified
in water using other methods, such as gas chromatography/
mass spectroscopy.
-------
6. Additional studies are needed to better establish the
accuracy and precision of the present method using a
greater number (and variety) of water samples. The
reproducibility of the low temperature luminescence
measurement could probably be improved by careful se-
lection of the sample tubes (to assure uniformity)
and by rapid rotation of the sample tube. The efficiency
of the extraction procedure would best be determined
using an independent analytical technique such as gas
chromatography.
-------
SECTION III
INTRODUCTION
In an earlier phase of the present program, low temperature
luminescence (LTL) spectra of several PCB isomers and mix-
tures (Aroclors) were compared with spectra of some DDE
and DDT derivatives. This study also included the analysis
of several Aroclor/DDT mixtures, and indicated that PCB's
could be identified in the presence of DDT derivates by
utilizing inherent differences in the excitation and emission
spectra.
The present effort has been directed principally at the
development of a simple method for the estimation of PCB's
and DDT in water samples. A secondary objective has been
to determine limitations of the method and to consider
possible coupling of LTL analysis with existing chromato-
graphic methods.
The procedure employed in the present program consisted of
collecting grab samples of water, extracting these with
dichloromethane to remove organic material, concentrating
the extracts, and finally adding a second solvent (methyl-
cyclohexane) suitable for LTL measurement.
This report begins with a brief discussion of molecular
luminescence, emphasizing empirical rather than theoretical
concepts. This is followed by a discussion of the LTL
spectra and detection sensitivities for several Aroclors
and DDT derivatives in methylcyclohexane.
Analytical techniques are described for both pure water and
natural (river) water samples doped with Aroclor 1254 and
p,p'-DDT. Recoveries and detection sensitivities are also
determined. Finally, analyses of environmental samples
are discussed with respect to PCB/DDT and other luminescent
substances present in natural waters.
-------
SECTION IV
LOW TEMPERATURE EMISSION SPECTROSCOPY
Most of the luminescence data appearing in this report were
obtained with the sample cooled to liquid nitrogen tempera-
ture (77°K). There are several advantages to working at
this temperature, and these will be discussed briefly here.
Of greatest importance to the present study is that the
strongest emission from PCB's and DDT derivatives is phos-
phorescence, which appears only at low temperature.
Phosphorescence results from the return of a molecule in a
triplet electronic state (usually the lowest) to the ground
state. In this respect it differs from fluorescence, wherein
molecular emission occurs between an excited singlet state
and the ground state. These electronic states are shown
schematically in Figure 1, where SQ represents the ground
state, Si and 82 are excited singlet states, and TI and T2
are triplet states. (Note: Figures are grouped at the end
of the main text, beginning with page 47.) Intersystem
crossing is a non-radiative process by which triplet states
are populated following absorption to singlet states.
In a given molecule showing both fluorescence and phos-
phorescence, the phosphorescence will occur at longer wave-
lengths and will show a much longer lifetime. The lifetime,
or decay time, is defined as the time required for the
emission to fall to 1/e of its original intensity upon
terminating excitation. Phosphorescence lifetimes are
generally in the range of one millisecond to several seconds
and can usually be measured using mechanical choppers.
Fluorescence lifetimes, however, are generally in the nano-
second range and measurement requires much more sophisticated
equipment. Lifetime measurement is thus another spectral
parameter which may be used for compound characterization.
The longer lifetimes of triplet molecules make them subject
to non-radiative deactivation (quenching) by collisions with
solvent molecules, oxygen, and other species. Imbedding
the phosphorescent molecules in a solvent matrix at low
temperature reduces the probability of these quenching
processes, enabling observation of phosphorescence. Although
fluorescence is less susceptible to such quenching, lowered
temperatures may also enhance the fluorescence intensity.
-------
As mentioned previously, chlorinated (and in fact, hal-
ogenated) aromatics generally have lower fluorescence
yields, but higher phosphorescence yields, than the parent
hydrocarbon. Halogenation also shortens the phosphorescence
lifetime (McClure, 1949). The mechanism responsible for this
is called the "internal heavy-atom effect." Basically,
addition of halogen substituents favors the rate of triplet
population by depletion of excited singlet states. Molecules
of interest here offer good examples of this effect.
Biphenyl shows approximately equal yields of fluorescence
and phosphorescence, where chlorinated biphenyls show
phosphorescence/fluorescence yields more than 100 times
greater (Dreeskamp et al., 1972). Also, isomers of DDT
and DDT show moderately strong phosphorescence, but fluo-
rescence of these molecules is at least 100 times weaker
(Brownrigg et al., 1972; hereafter, this work is referred
to as the "first report").
Another advantage of low temperature vs. room temperature
analysis is that fine structure may appear at low tempera-
ture. This fine structure represents vibrational transitions
accompanying the electronic transition. In an emission
spectrum, vibration intervals characteristic of the ground
state would be observed, and certain of these vibrations
would also be observed in the infrared or Raman spectra.
The amount of vibrational structure observed varies with
both the particular molecule and the choice of solvent.
Usually solvents which freeze to an ordered crystalline
state produce sharper structure than glassy (amorphous)
media. This is probably because the crystalline medium
allows only a small number of orientations available to the
guest (emitting) molecule, whereas a great number of random
orientations are possible in the glass.
The use of polycrystalline n-alkane matrices to produce
these so-called "quasi-line" spectra was pioneered by
Shpolskii (1960-1963). Sharpest spectra are usually
obtained using solvents having molecular dimensions nearly
the same as those of the guest molecule, and spectral band-
widths decrease strongly with decreasing temperature. As
an example of this effect, Figure 2 shows the spectra of
biphenyl in heptane at both room temperature and at 77°K
(note the absence of phosphorescence at room temperature).
Also included is a spectrum of biphenyl in methylcyclo-
hexane (MCH) at 77°K. Heptane freezes to a highly
scattering "snow," whereas MCH forms a glass.
-------
The scattering properties of heptane would probably make
it a less desirable solvent for quantitative work. Note,
however, the more highly structured emission obtained in
heptane; this suggests the desirability of heptane for
analysis requiring higher sensitivity and selectivity for
biphenyl.
Unfortunately, the more highly chlorinated PCB isomers and
DDT derivatives discussed in the first report show un-
structured emission in both MCH and heptane, so that there
is apparently no advantage to be gained from a Shpolskii-
type solvent in this case. Since the PCB's most likely
to be found in the environment are of this type (highly
chlorinated), the solvent selected was MCH. As mentioned
above, this solvent freezes to a clear, rigid glass, result-
ing in less scattered light.
To summarize, there are several advantages to be gained
from luminescence measurement at low temperature. First,
halogenated aromatics (such as PCB's and DDT) exhibit
moderately strong phosphorescence, but much weaker fluo-
rescence. Phosphorescence, however, is ordinarily not
observed at room temperature. Other molecules which also
show fluorescence will often show enhanced fluorescence
emission at low temperature. Depending upon the particular
molecule and the solvent chosen, lowered temperatures may
produce more highly structured absorption and emission;
this in turn enables both greater sensitivity and. selectivity,
8
-------
SECTION V
AROCLOR-PESTICIDE SOLUTIONS IN METHYLCYCLOHEXANE
Spectra of several Aroclors and DDT-type compounds, both
singly and in mixtures, were studied in methylcyclohexane
at 77°K. The purpose of these studies was to obtain basic
spectral signatures for comparison with natural water lumi-
nescence signatures. These studies also provide a knowledge
of the concentration range where response is linear, and
an estimate of ultimate sensitivity and selectivity.
In the first report it was noted that the spectra of the
o,p'- and p,p'~isomers of DDD and DDT were all very similar.
Also emissions from DDE were found to be very much weaker
than that of DDD or DDT. For these reasons, only p,p'-DDD
and p,p'-DDT have been included in the present study.
The PCB's most likely to be found in environmental waters
have isomer distributions resembling commercial mixtures
such as Aroclor 1248 and 1254. Therefore, these two Aroclors,
plus the newer product Aroclor 1016, have been included in
the present study. Aroclor 1016 has a composition similar
to Aroclor 1242, but with isomers having, five or more
chlorine atoms removed (Nisbet and Sarofim, 1972).
Sources
Pesticide samples were obtained from the EPA Perrine Primate
Laboratory, Perrine, Florida. Samples were designated
"Reference Standard" and had stated purities of 99+%.
These compounds were used as received.
Aroclor samples were obtained from Dr. E. S. Tucker and
Mr. W. B. Papageorge of the Monsanto Company and were used
as received.
Choice of Solvents
For quantative determinations, it is important that the
solvent chosen should freeze in a uniform and reproducible
manner. Most organic solvents would probably form clear
crystals if cooled slowly enough, but this slow cooling
would not be practical for rapid analyses. Also, unless
the solute can easily occupy the solvent crystal lattice,
the slow rate of cooling would probably favor the rejection
of the solute from the lattice. (Fractional freezing is
in fact sometimes used for solvent purification.) Thus,
rapid freezing is usually necessary but in turn will often
result in a highly cracked, opaque polycrystalline mass
-------
having the appearance of snow. Most common organic solvents
such as hexane, carbon tetrachloride, benzene, and acetone
exhibit this behavior. This type of crystal matrix is
generally undesirable for luminescence analysis for several
reasons. Local solute concentrations can be highly variable
depending on the rate of freezing. Macroscopic cracks
often develop which can influence both the luminescence
intensity and the background scatter. This gives rise to
variability in luminescence intensity and also contributes
a large background signal from scattered exciting light.
Another important consideration, which should be more
obvious, is that the solvent itself not be strongly absorbing
or emitting at the analytical wavelengths. Thus benzene
and its derivatives, which begin to absorb at wavelengths
of 270 nm or longer (and emit themselves) would usually be
undesirable solvents.
For these reasons the solvent chosen should be non-aromatic
and should freeze to a clear, rigid glass, but only a
small number of solvents meet these requirements. Winefordner
and St. John (1963) have evaluated many solvents and solvent
mixtures for ability to form glasses, and an expanded tab-
ulation has been given by Winefordner, McCarthy, and St. John
(1967). Westrum and McCullough (1963) also include glass-
forming compounds in their tabulation of thermodynamic
data for organic compounds.
We have had particularly good success with methylcyclohexane
(MCH). The PCB/DDT compounds are soluble at relatively
high concentrations (at least 100 ppm), and the solvent
itself is non-absorbing and non-emitting; ultraviolet
absorption begins at about 230 nm. Sample concentrations
of at least 100 ppm can be frozen quickly (15-30 seconds)
to liquid nitrogen temperature without cracking. Our
experience thus seems to contradict the results of Winefordner
and St. John (1963), who consider MCH "not usable" due to
high frequency of crack formation. The cracking which we
occasionally observe is believed to result from residual
water remaining after sample tubes have been washed.
Commercial grades of methylcyclohexane usually contain
luminescent impurities. The most prevalent of these
impurities appears to be toluene, which has a structured
fluorescence in the 270-320 nm region and a partially
structured phosphorescence in the 350-500 nm region. Because
the excitation spectrum is in the 240-270 nm region, it
would interfere with the determination of DDT-type compounds
and, to a lesser extent, with PCB's.
10
-------
The toluene concentration of commercial MCH ranges from
less than 1 to over 1000 ppm. Practical or technical grade
solvents have the highest toluene concentrations and would
not be particularly suitable without purification.
Solvent Purification
Methylcyclohexane—The solvent used for most of this work
was Matneson,Coleman, and Bell Spectroquality grade. This
solvent typically contains about 0.3 ppm toluene. The
second monthly report erroneously reports that this solvent
contains about 5 ppm toluene. This value was obtained by
comparison of the intensity with a 500 ppm standard, which
was later found to be too strongly absorbing to enable a
linear extrapolation. The resulting phosphorescence inten-
sity is comparable to that of PCB/DDT when the latter are
presented at concentrations on the order of 0.1 ppm. There-
fore standard solutions containing such low concentrations
are best done using toluene-free solvents.
Originally, efforts focused on purification of practical
grade MCH, which is much cheaper than the Spectroquality
material. The toluene concentrations, however, are in the
1000 ppm range, making it unsuitable for most luminescence
work.
Various procedures were used in an attempt to purify the
practical grade material: distillation, sulfuric acid
wash, and column chromatography using silica gel and
activated carbon. None of these procedures reduced the
content appreciably. It is possible that no treatment,
except possibly exhaustive hydrogenation, can appreciably
reduce high levels of toluene. However, it was found that
silica gel chromatography was effective in removing the
much lower levels of toluene present in the Spectroquality
material. This procedure is simple, efficient, and fairly
rapid. A more elaborate variation of this technique was
first developed by Potts (1952).
Woelm silica gel, activity grade I (200 mesh) was first
Soxhlet extracted overnight with pure dichloromethane.
This extraction removes fluorescent contaminants which may
be present. The extracted material was then baked over-
night in an oven at 160-180°C.
A one-inch diameter chromatographic column with Teflon stopcock
was filled to a depth of about 30 cm with silica gel. The silica
11
-------
gel was added as a slurry in MCH with occasional.tapping
to free air bubbles. The level of MCH was kept above the
silica gel at all times.
Methylcyclohexane (Matheson Spectroquality) was allowed to
pass through the column at rates on the order of 1 ml/min.
After discarding the initial 75 ml, five 100 ml fractions
were collected.
Luminescence analysis of the various fractions at 77°K showed
no contaminants other than trace amounts of the toluene
present in the starting material. The first three fractions
appeared to be the best, and contained at least thirty
times less toluene than did the starting material (which
contained about 0.3 ppm toluene). The fourth and fifth
fractions contained about twice as much toluene as the
preceding fractions, but were still about a factor of ten
better than the starting material.
Dichloromethane (Methylene Chloride)—This solvent, obtained
from Fisher Chemical Company as the "Spectranalyzed" grade,
was usually found suitable for use as received, but on one
occasion was found to contain high concentrations of lumi-
nescent impurities.
Impure dichloromethane was purified by distillation, the
fraction boiling at 42°C being retained.
Spectra of Pesticides, Aroclors, and Their Mixtures
All emission/excitation spectra appearing in this report
were obtained on a Baird-Atomic Model SF-100 "Fluorispec"
fluorescence spectrophotometer, using standard optics and
accessories. Optically, this instrument is equivalent to
the earlier Model SF-1, but has improved electronics.
Both instruments employ a 150 watt xenon source, with light
dispersed by dual monochromators for both excitation and
emission. These instruments are single beam types, and the
excitation/emission spectral intensities are uncorrected
for instrument response. Although corrected excitation and
emission spectra could be produced if desired, the present
method, being based on comparisons with standards, does not
require these corrections. A good discussion of how these
corrections may be obtained is given by Parker (1968).
The first report contains many spectra obtained on the SF-1.
Small differences in grating and mirror efficiency and photo-
tube response characteristics give rise to spectral differences
12
-------
obtained with the two instruments. Since the SF-100 was
used exclusively for the present study, some of the more
important pesticide and Aroclor spectra have been repeated.
Wavelength Calibration—The wavelength calibration of the
SF-100 was assumed accurate to ±2 nm over the standard
range of 220-700 nm. Unfortunately, after much of the present
work was completed, a check of the instrument revealed that
the wavelength calibration was beyond accepted tolerances
in certain spectral regions.
The accuracy of the emission monochromator was checked by
comparison of the dial settings with the discrete spectrum
from a low pressure mercury lamp. Wavelengths of mercury
lines may be found in the AIP Handbook (Crosswhite, 1972).
Excitation wavelength corrections were obtained by scanning
the excitation monochromator through the scatter peak at
a given fixed emission wavelength, then applying the ap-
propriate correction for the emission wavelength as obtained
previously.
Curves showing the wavelength correction factors as a
function of the apparent (dial) wavelength are given in
Figure 3. To obtain corrected wavelengths, the appropriate
correction factor is subtracted from the apparent wavelength.
Wavelengths indicated on all spectra in this report rep-
resent actual dial readings and are thus uncorrected.
Excitation spectra of the pesticides and PCB's studied
here are strongest in the 240-290 nm region, and the wave-
length correction factors are relatively small (0-3 nm).
Also, the phosphorescence emissions of these same compounds
are strongest in the 380-500 nm region, and here the
correction factors are within the expected tolerances
(2 nm or less). Correction factors for excitation become
quite large for wavelengths of 300 nm or longer, where
there is significant absorption by natural substances in
water.
The loss of acceptable wavelength calibration in our instru-
ment was quite unexpected and probably resulted from its
previous shipments to field locations and associated rough
handling. It is advisable that the monochromator wave-
lengths be checked occasionally (e.g., every six months),
particularly if the instrument has undergone shipment. In
particular, it must not be assumed that the curves in Figure 3
-are representative of every SF-100, since a standard instru-
ment should be correct within ±2 nm.
13
-------
Instrumental Parameters—Certain instrumental parameters
which remained largely unchanged have been omitted from the
labels on the spectra. Thus, the Hewlett-Packard X-Y
recorder gain was typically 5mv/cm, and the photomultiplier
voltage (RCA IP28 tube) was 750 volts. The SF-100 time
constant was always 0.3 seconds, and the slow wavelength
scan mode (1 nanometer per second) was used exclusively.
When these parameters differ, this has been noted on the
spectrum label.
The instrumental slit widths employed were usually 22/11
or 33/11 for emission spectra, and 11/22 or 11/33 for
excitation spectra. The two numbers on either side of the
slash are mechanical positions of slit controls, referring
to either the excitation (left pair of numbers) or emission
(right) monochromator. There are two slit controls for each
monochromator and three possible positions for each control.
In particular, the combinations 11, 22, and 33 give spectral
bandwidths of about 2, 5, and 26 nm respectively. Our
preferred approach has been to use the narrowest set of
analyzing slits in order to obtain the optimum spectral
resolution (2nm). The particular slit combinations employed
for a given trace appear on the spectrum label. For example,
the designation (22/11; 11/22) means that the excitation
monochromator used to excite emission used 22 slits, and
the emission was analyzed by scanning the emission mono-
chromator using 11 slits. Conversely, the excitation
spectra were recorded with the (scanning) excitation
monochromator slits 11 and the emission monochromator
slits 22.
Other Parameters—Phosphorescence occurs weakly if at all
at room temperature. Since phosphorescence is the emission
of greatest analytical importance for compounds studied
here, all spectra were obtained at 77°K unless otherwise
indicated.
The quartz sample tubes used for low temperature emission
and analysis were optical quality, equivalent to Suprasil
grade. The mean inside diameter of these tubes was 3 mm
(with a variation of up to 15%), and wall thickness averaged
about 0.5 mm. The tubes were typically 10 inches long
and were closed (fused) at one end. In order to avoid
possible transfer contamination, tubes were filled by
pouring directly from the glass vial. With care and
practice, this could be done without spillage. Tubes
were usually filled to a depth of 7-8 cm, or a volume of
14
-------
about 0.5 ml. Sample tubes were closed with an improvised
cap consisting of a short section of 1/8 x 1/4 inch rubber
tubing stopped with a 3/16 inch diameter glass bead at one
end. At room temperature, MCH diffuses through the rubber
at a rate of about 0.05 ml per month so that samples should
not be kept longer than this if initial concentrations are
to be maintained.
After filling and capping the tube, it was lowered directly
into liquid nitrogen in the optical dewar. About 30 seconds
were required to completely immerse the sample into liquid
nitrogen, at which time the sample was usually transparent
and uncracked. Occasionally the sample may crack, even
after repeated attempts to produce a glass; this is usually
caused by residual moisture.remaining in the tube after
washing. Freshly washed tubes can be dried quickly by
flushing several minutes with dry nitrogen gas.
Spectra of Pesticides—Spectra of the o,p'- and p,p'-
derivatives of DDE, ODD, and DDT were discussed in the
first report. It was found that the spectra of the ODD
and DDT compounds are all very similar, and these compounds
would be difficult to distinguish on the basis of lumi-
nescence spectra alone. The emission of DDE is very much
weaker (about 400 times) than that of ODD or DDT, result-
ing in poor sensitivity. Spectra of p,p'-DDT were obtained
in purified MCH at concentrations of 0.1, 1, 10, and 100
ppm. Solutions were prepared by successive ten-fold
dilutions of a 100 ppm stock solution, prepared by dis-
solving 1 mg of DDT in 10 ml of solvent. All solutions
were prepared and stored in screw-cap glass vials (4 dram)
having both aluminum foil and Teflon cap liners.
A representative excitation/emission spectrum of p,p'-DDT
is shown in Figure 4, and that of p,p7-DDT appears in
Figure 5. As indicated above, both excitation and emission
spectra are very similar, which is hardly surprising in
view of the similarity in chemical structure (see Figure 1
of the first report).
At optimum resolution, the detection limit in MCH for DDT
is about 0.03 ppm. Since DDD has a similar phosphorescence
intensity, a similar detection limit is expected. This
limit is largely imposed by the noise contributed by the
light source and photomultiplier tube. Lower detection
15
-------
limits could of course be achieved by using wider analyzing
slits, at the expense of resolution (and specificity).
A graph of phosphorescence intensity (approximated as the
value of the signal maximum) versus concentration is shown
in Figure 6. The analytical curve is linear except near
100 ppm, where departures become significant (about 25%).
Spectra of Aroclors—Monsanto Aroclors 1221, 1242, 1248,
1254, and 1260, along with several PCB isomers, were studied
in the first report. Some of the more important highly
chlorinated Aroclors (1248) and (1254) have been reinves-
tigated here, along with the newer aroclor 1016, a
replacement for Aroclor 1242.
Spectra of Aroclors 1016, 1248, and 1254, in MCH appear in
Figures 7-10. Solutions were prepared by successive
ten-fold dilutions of a 100 ppm standard (1 mg per 10 ml
solvent).
Because Aroclors are complex mixtures of PCB's, excitation/
emission spectra obtained at different wavelengths should
be somewhat different. This is indeed found to be the
case, but much more so in the excitation spectra than in
the emission spectra. Phosphorescence spectra of the more
highly chlorinated Aroclors are broad, with maxima in the
440-470 nm region. An unexpected feature of the Aroclor
1016 phosphorescence is the double emission peaks near
450 and 470 nm (Figure 7). Comparison with excitation
spectra of certain PCB isomers in the first report suggests
that the emission components at short wavelengths (380-
400 nm) are largely due to more highly chlorinated PCB
isomers.
Spectral features of these Aroclors are largely independent
of concentration below 10 ppm. At higher concentrations
excitation intensities become distorted in the short wave-
length region. An example of this is shown in Figures 9
and 10, which compare spectra of Aroclor 1254 at concentra-
tions of 1 and 100 ppm (the 10 ppm spectrum is very similar
to the 1 ppm). The attenuation at short excitation wave-
lengths in the more concentrated solution is due to strong
absorption at these wavelengths. The excitation beam does
not penetrate the sample, and the resultant emission occurs
from a relatively shallow surface layer. Since the emission
monochromator focuses more nearly at the center of the sample,
this surface emission is not collected efficiently-
16
-------
The theoretical dependence of emission intensity on concen
tration can be derived as follows. The emission intensity
Ie is proportional to the intensity of the light absorbed
by the sample Ia and the quantum yield of emission 0. The
quantum yield is defined as the ratio of photons of light
emitted to the number absorbed.
Ie = la ' *
Equation 1 neglects the possible existence of quenching
mechanisms, such as collisional deactivation or energy
transfer. This approximation is probably valid for most
dilute solutions at low temperature. From Beer's Law,
Ia = ID - Jo 10~ebc (2)
where I0 is the intensity of the exciting light, e is the
molar extinction coefficient, b is the path length, c is
the concentration. If the quantity (ebc) is suitably small
(usually achieved by making c small), the exponential can
be approximated as a series expansion giving
Ie = 2.303 0 I0 ebc (3)
Therefore the emission intensity is predicted to vary
linearly with concentration at low concentration. Departures
from linearity at high concentration imply that the series
expansion applied to Beer's Law is no longer valid.
Phosphorescence intensities (approximated as peak heights)
as a function of concentration are shown in Figure 6.
Aroclor 1254 shows a linear response (within 20%) over the
range 0.01 to about 60 ppm. Aroclors 1016 and 1248 were
studied over a more limited range, but also appear to be
linear below about 40 and 90 ppm respectively. The excita-
tion wavelength employed for the Aroclor data of Figure 6
was 290 run, which gives nearly optimum sensitivity for this
Aroclor. The other Aroclors were excited at the same wave-
lengths as Aroclor 1254 for purposes of spectral comparison.
However, the 290 nm excitation wavelength is not the optimum
for Aroclor 1016 and 1248, whose excitation maxima are at
shorter wavelength. In order to compare relative intensities
more closely in Figure 6, intensity data for Aroclors 1016
and 1248 should be increased by about a factor of two.
17
-------
The noise-limited detection sensitivity for Aroclor 1254
in MCH at 77°K is about 0.002 ppm at optimum resolution.
On the basis of relative intensities, detection sensitivities
for Aroclor 1016 and 1248 should be similar to that of
Aroclor 1254.
Spectra of Aroclor/Pesticide Mixtures—A limited number of
mixtures were studied in the first report, and these have
been extended here. In particular, earlier work showed
that Aroclor 1254 could be detected in the presence of DDT
concentrations 100 times greater.
The mixtures studied here were composed of Aroclor 1254
(A) and p, p'-DDT (D) at concentrations ranging from 0.1
to 9 ppm in MCH:
a) 9 ppm A + 1 ppm D
b) 5 ppm A + 5 ppm D
c) 1 ppm A + 9 ppm D
d) 0.9 ppm A + 0.1 ppm D
e) 0.5 ppm A + 0. 5 ppm D
f) 0.1 ppm A + 0.09 ppm D
Spectra of the above mixtures in MCH at 77°K appear in
Figures 11-16. The spectral features of these mixtures are
basically'similar to the superposition of the spectra of
the separate components. DDT is best measured free of
Aroclor interference from an excitation spectrum monitoring
emission near 380 nm. This is an emission wavelength where
the DDT phosphorescence is appreciable (about 2/3 of its
peak value), while the Aroclor 1254 emission is very low
(about 2% of its peak value). Aroclor 1254 can be detected
in the presence of DDT in two ways. First, the excitation
spectrum in the 240-290 nm region could be obtained by
monitoring emission at about 440 nm. Although DDT also has
emission in this region, the Aroclor excitation spectrum
begins at longer wavelength than that of DDT and thus can
be measured free of DDT interference in this region. An
alternate approach would be to excite at wavelengths of
about 290 ran. The resultant emission would be principally
that of Aroclor, depending on the excitation bandwidth
employed; an excitation bandwidth of less than about 20 nm
should give almost pure Aroclor 1254 emission.
18
-------
In Aroclor 1016 and 1248, phosphorescence is also very
weak at 380 nm, and this would remain a good wavelength
to monitor the DDT excitation. However, the Aroclor 1016
or 1248 emission would not be excited free of DDT emission
since the excitation spectra of these Aroclors are closer
to that of DDT. In these cases, it would probably be
preferable to excite Aroclor emission at about 285-290 ran
with relatively narrow slits (less than 10 nm) and compensate
for loss of intensity by using wider emission slits. Since
the Aroclor emissions are intrinsically broad, bandwidths
of up to 30 nm (10 nm for double-peaked Aroclor 1016)
could probably be used without loss of spectral resolution.
Plots of the Aroclor 1254 emission intensity, based upon
the peak height of the narrow excitation peak near 292 nm,
show good linearity for concentrations between 0.1 and
10 ppm. DDT concentrations of this same magnitude thus
appear not to affect linearity significantly.
Similar plots of the DDT emission intensity, based upon the
narrow excitation peak at about 278 nm, show a similar
linear dependence with concentration with one exception.
The exception is the mixture having 1 ppm DDT and 9 ppm
Aroclor 1254, and the intensity is about 50% lower than
expected. The large departure from linearity may be due
to energy transfer. PCB's have longer wavelength absorptions
than DDT, which is equivalent to saying that their elec-
tronic states are at lower energies. The excited DDT
molecules can transfer their energy to neighboring molecules,
such as PCB's, having lower energy states (acceptors),
and the probability of this varies approximately with the
square of the acceptor concentration. The DDT emission is
thus effectively quenched in the presence of high concen-
trations of Aroclor 1254. Energy transfer, however, does
not appear to be important in the solution having 0.1 ppm
DDT and 0.9 ppm Aroclor so that the absolute concentrations
of the components seem even more important than their
relative concentrations.
From the present study and the results of the first report,
Aroclor 1254 could probably be determined in the presence
of DDT concentrations of up to 100 times greater. DDT,
having inherently weaker emission than Aroclor 1254, could
probably be detected in the presence of about 20 times
greater concentrations of Aroclor 1254. However, for
absolute concentrations of Aroclor above about 10 ppm,
there appears to be significant energy transfer from DDT
to Aroclor. This is effectively a quenching mechanism,
19
-------
producing non-linearity in the analytical curve for DDT.
This problem might be overcome by appropriate dilution of
the mixture, provided that signal/noise considerations
permit this.
Sources of Error—The uncertainty in a given intensity
measurement averages about 25% and arises from several
sources. First, quartz sample tubes vary by up to 15% in
their diameters and are not perfectly straight. The latter
condition results in non-uniform scattering of the excita-
tion beam as the tube is rotated; the emission signal thus
varies somewhat with tube rotation. This problem can be
overcome by a rapid rotation of the sample tube during
analysis so that inhomogeneities are averaged. A rotating
sample cell of this type is described by Zweidinger and
Winefordner (1970).
Another source of variation is related to the reproducibility
and homogeneity of solute substitution in the frozen glass.
This can be controlled somewhat by maintaining the same
freezing rates, but variations of up to 10% have been found.
A potential source of error is contributed by the xenon
source lamp. Using the water Raman band as a "standard,"
daily checks of the xenon lamp intensity at 350 nm showed
random variations of up to 15% with the lamp intensity
maximized. Of course slight defocusing of the arc so as to
produce the same relative intensity would be possible. A
better solution would be to ratio the sample emission to a
signal proportional to the excitation light intensity;
instrument gains should then be independent of lamp intensity
fluctuations. This source of error can be minimized by
running an appropriate luminescence standard (see Appendix).
20
-------
SECTION VI
AROCLOR 1254 AND p,p'-DDT IN PURE WATER
These experiments were performed in order to determine
recovery and extraction efficiency for Aroclor 1254 and
p,p'-DDT in pure water.
Sources
Distilled water samples used originally were obtained from
Belmont Springs Water Co., Belmont, Mass. This water comes
in five-gallon plastic bags with a rubber hose outlet.
Methylcyclohexane extracts of this water, when cooled to
77°Kf showed a moderately strong but variable background
emission. Based upon similarities in the UV absorption and
phosphorescence excitation/emission of one of the principal
impurities with that of dibutylphthalate, phthalates appear
to be likely contaminants, with concentrations on the
order of 10 ppb.
Although these contaminants could be removed by repeated
solvent extraction, it was decided to try a different
source of water. It was found that "Ultrapure" water
obtained from Harleco contained negligible amounts of
luminescent contaminants, and this material was used for
subsequent experiments.
Originally, the solvent used for extraction was methycyclo-
hexane, chosen largely because it would function both in
this capacity and as the rigid solvent for low temperature
luminescence measurements.
Use of this solvent for extraction was soon abandoned
because the relatively high boiling point (101°C) required
rather long concentration times. In addition, this solvent
is fairly expensive at purities suitable for luminescence
work.
The solvent selected for extraction,-dichloromethane (DCM),
has a low boiling point (42°C) and is obtainable in high
purity at relatively low cost. Fisher Spectranalyzed DCM
was used for all extractions.
One liter samples of Harleco water were doped with Aroclor
or DDT directly into the original glass container. Dopant
solutions were either 10 or 100 ppm solutions in ethanol,
21
-------
using the Aroclors and pesticides described in Section V.
The ethanol was Graves Extra Fine, distilled prior to use
to remove a phthalate-type contaminant. Glass lambda pipettes
were used to deliver accurate microliter volumes of the
dopant solution into water. The doped water was allowed
to incubate overnight while being stirred continuously
with a Teflon bar magnet activated by a magnetic stirrer.
Extraction Procedure
Extraction was performed in the same container. Seventy-five
milliliters of DCM were transferred to the bottle and the
mixture stirred on a magnetic stirrer for periods ranging
from several hours to one day. The stirring speed was
adjusted to be high enough to break the DCM into a fine
emulsion which appeared to be uniform throughout the water.
After stirring, the bottle was removed from the magnetic
stirrer and the DCM layer allowed to coalesce at the bottom
of the bottle. This bottom layer, along with a few milli-
liters of water, was then transferred to a 125 ml separatory
funnel using a 50 ml pipette. Approximately 20 ml of the
original DCM added was retained by the water.
The DCM extract was dried using 10-20 grams of anhydrous
Na2SC>4. The Na2SO4 was Fisher Reagent grade material which
had been previously heated to 400-500°C for several hours.
Originally, the Na2SO^ was placed in a filter funnelv (100 mm
high, 23 mm I. D.) having a medium porosity glass frit.
The DCM was then allowed to drip from the separatory funnel
through the Na2SC>4 layer into a small flask. This procedure
usually required waiting several hours and was even longer
with extracts of environmental samples containing particulates,
Therefore this procedure was later modified by transferring
the DCM to a 125 ml flask containing Na2S04, adding a
Teflon-covered bar magnet, and stirring the mixture on a
magnetic stirrer for 15 minutes. Successive 75 ml extracts
(two to three) were treated in the same manner and analyzed
separately.
The use of DCM was suggested by the work of Kites and
Biemann (1972). Various solvents and solvent mixtures have
been used by other investigators, and these could probably
be used in place of DCM. For example, Blumer (1970) has
used n-pentane to extract organic compounds from sea water,
Goerlitz and Brown (1972) recommend the use of n-hexane,
and the EPA Method (1971) specifies a mixture of 15% ethyl
ether in hexane. A disadvantage of DCM in standard procedures
22
-------
utilizing gas chromatography with an electron capture
detector is the strong response of this type of detector
to halogenated compounds.
Much higher detection sensitivities can be achieved by
processing large volumes of water. The use of simple
liquid-liquid extraction, however, would no longer be
practical. Not only would large volumes of solvent be
necessary, but even higher purity would be required to
minimize interference from solvent impurities. For this
reason, liquid extraction methods are replaced by flow
systems employing adsorptive materials. The standard
method of this type involves passing water through a column
of activated carbon followed by Soxhlet extraction to re-
move adsorbed organics (Breidenbach et al, 1964).
More recently, Ahling and Jensen (1970) describe a filter
consisting of a mixture of n-undecane and Carbowax 4000
monostearate on Chromosorb W to adsorb PCB's and chlorinated
pesticides from water; the organochlorine compounds are
eluted from the column with petroleum ether. Gesser et al
(1971) and Uthe et al (1972) have reported the use of
porous polyurethane foam plugs to absorb organochlorines,
including PCB's from water. Harvey (1972) has reported the
use of a crosslinked polymer resin (Rohm and Haas Amberlite
XAD series) to adsorb PCB's and DDE/DDT compounds from
marine waters. Although none of these methods was evaluated
in this work, there appears to be no inherent difficulty
in adapting these techniques to the present luminescence
methodology. Of the above methods, that of Harvey (1972)
would appear to be the least complex. It would be very
worthwhile to determine how this procedure might best be
coupled with low temperature luminescence analysis.
Concentration of the Extracts
Initially, extracts were concentrated using a Kuderna-
Danish (K-D) concentrator, obtained from Ace Glass Company,
Model 6707. This consisted of a 500 ml flask, a graduated
10 ml receiver, and a 3-ball Snyder column. The uppermost
portion of the flask and the column were insulated with
glass wool. The extract was put into the K-D, the receiver
immersed in a water bath at 70-90°C, and the extract con-
centrated to approximately 1 ml (which required periods of
several hours). Although a modification of this method
might be desirable for batch processing of many samples, it
was decided that the smaller number of samples analyzed
here could be done more rapidly using a rotary evaporator.
23
-------
The rotary evaporator used here was a Calab Model 5150
with Teflon fittings. During evaporations, the receiver
flask was cooled with an ice water bath and the DCM distilled
at room temperature under reduced pressure provided by a
water aspirator. Typically, about 5 minutes were required
to concentrate a 75 ml extract to less than one ml. A
potential disadvantage of the rotary evaporator, in comparison
with the Kuderna-Danish, is the problem of measuring small
volumes of concentrated extract. No attempt was made to
measure the final volumes of DCM, although as an estimate
these were less than 0.5 ml.
Analysis of Extracts
Each concentrated DCM extract was first brought to a volume
of 5-10 ml with methylcyclohexane. This solvent, unlike
DCM, produces a rigid glass when cooled to 77°K. In addition,
optical transparency is good throughout most of the ultra-
violet. These volumes were originally chosen so that, if
desired, absorbance measurements could be made easily in
1 cm cells, with enough excess to permit repeated measure-
ments in the event of accidental sample loss or contamination.
Sample volumes as small as 0.1 ml can be analyzed with the
same equipment used here, although a more convenient sample
size is approximately 0.5 ml.
Low temperature luminescence analysis was performed in
the same manner as for the Aroclor/pesticide standards
in MCH described in Section V.
Inclusion of about 10% DCM causes the MCH to freeze to a
slightly cloudy glass. Intensities appear to average 2 to
3 times higher in this matrix than in pure MCH, probably
because of increased optical path length due to scattering.
Because of this scatter variability it is desirable to keep
the final DCM volume as small as possible.
Recovery and Extraction Efficiency
Recovery of Aroclor 1254 from 1-liter samples of pure water
was about 70% at the 1 ppb level and about 80% at the 10 ppb
level. At both concentrations, 65-85% of the total amount
extracted was contained in the first extract, and 95% or
more in the first two extracts combined. Similar experiments
with p,p'-DDT gave recoveries of about 50% at the 10 ppb level
and about 70% at the 100 ppb level. Of the total extracted,
80-90% was contained in the first extract and 95% or more in
the first two extracts combined.
The difference in recovery between Aroclor 1254 and p,p'-DDT
is probably significant but is not presently understood.
The observed losses are probably due mainly to adsorption
24
-------
on glass surfaces. Since the water samples were unfiltered/
additional losses (presumably minor) due to adsorption on
particulates may have occurred as well.
Detection Sensitivities
Detection limits for p,p'-DDT and for Aroclor 1254 in MCH
at 77°K were found to be about 30 and 2 ppb respectively.
If recoveries for p,p'-DDT and Aroclor 1254 are assumed
to be 60% and 75% respectively, the detection limits for
these compounds in a 1-liter volume of pure water would
be about 0.5 and 0.03 ppb respectively. This assumes that
the extraction solvent is free of potentially interfering
contaminants.
25
-------
SECTION VII
ENVIRONMENTAL SAMPLES
Grab samples were collected from several natural waters,
mainly in the Boston area. The samples were taken within
one foot of the surface. These were extracted and analyzed
for PCB/DDT in a manner similar to that described for doped
samples of pure water. River water samples were also doped
with Aroclor 1254 and p,p'-DDT to determine recovery and
sensitivity.
Sampling Locations
Samples were taken from the following locations:
Location Date(s)
A. Shawsheen River, Bedford, Mass.
B. Concord River/ Billerica, Mass.
C. Hodgkins Cove, Gloucester, Mass.
D. Milwaukee River, Milwaukee, Wis.
E. Charles River, Cambridge, Mass.
2/21/73
5/24/73
6/25/73
9/26/73
8/23/73
10/2/73
10/12/73
12/6/73
The Shawsheen River is a shallow, narrow river close to the
Baird-Atomic plant. This water does not appear to be
seriously polluted, but is rich in natural organic material,
giving it a yellow-brown appearance. Samples were collected
where the river passes through two viaducts under the Middlesex
Turnpike, a two-lane highway. The sampling bottle was held
manually about three feet from the shore line (roughly half
the width of the viaduct) and the surface water taken.
The Concord River, into which the Shawsheen flows, is also
relatively unpolluted at the location sampled. In this
case, a sample was taken near the approximate center of the
26
-------
river. This was done by lowering the sample bottle into
the water from a bridge on Massachusetts Route 3. The
bottle was contained in a bucket fashioned from an aluminum
cylinder, secured by a rope.
The ocean water sample was taken near the University of
Massachusetts Marine Station at Hodgkins Cove, Gloucester,
Massachusetts. The point of collection was off a floating
platform anchored about ten feet from shore. Several boats
were anchored nearby, and streaks of oil were observed on
the surface about six feet from where the sample was taken.
The wind was calm, however, and no oil streaks were observed
to enter the sample bottles.
Milwaukee River water samples were collected by A. Hornig
on a return trip from that area. Other analyses of water
in this area indicated PCB concentrations of several parts
per billion (G. Veith, private communication). Samples
were taken in Estabrook Park, on the east bank of the river,
about 200 yards south of the northern boundary of the park.
The samples were taken off the bank near the shoreline,
where the water was less than a foot deep. The water did
not appear to be heavily polluted, and no slicks, foams,
or odors were noted. The water samples were yellow-green,
as found for the Shawsheen and Concord River Samples.
The Charles River samples were collected near the center
of the river by lowering bottles into the water from near
the center of the Longfellow Bridge. This bridge carries
moderately heavy car and subway traffic between Cambridge
and Boston. Kites and Biemann (1972) performed extensive
analyses of this river during 1971. Using combined gas
chromatography-mass spectrometry and liquid chromatography,
several pollutants including phthalates and polynuclear
aromatics were identified.
Collection
Samples were collected in half gallon, wide-mouth glass
bottles, with Teflon-lined plastic screw caps. Five milli-
liters of concentrated HC1, approximately 12 N., were added
as soon as possible after collection to prevent possible
biodegradation and to avoid extraction of basic compounds.
Since organic pollutants may also be adsorbed on particulates,
no attempts were made to remove these prior to extraction.
27
-------
Extraction and Concentration
Sample bottles were returned to the laboratory and extracted
with DCM in the manner described for pure water samples
(Section VI). Some difficulties were encountered with
these samples in that, after settling, the DCM extract
consisted of a relatively clear lower layer, and an upper
layer which appeared to consist of small globules suspended
in a yellow-brown froth. This frothy mass was probably an
emulsion of water in DCM, stablized by naturally occurring
surfactants in the water.
Originally the DCM extract was dried by allowing it to drip
through a short column of anyhdrous Na2SC>4. The DCM passed
through slowly until, the frothy layer encountered the NajjSO^
whereupon the flow essentially stopped. This was probably
because water in this layer hydrated the Na2S04, forming a
hard, largely impermeable crust. This procedure was later
abandoned, and the DCM extract stirred with Na2SO^ in a
small flask instead. After stirring, the extract was
decanted off, concentrated, and diluted with MCH as de-
scribed previously. Luminescence measurements, described
in the following section, were performed on these solutions.
The parameters used for each sample are summarized in Table 1.
Spectral Analysis
The natural water samples have a strong background lumi-
nescence, probably due to humic substances. In order to
get representative signatures of this and potential organic
pollutants, excitation and emission spectra were obtained
at wavelengths other than those of analytical importance
for PCB's or DDT. The wavelengths selected appeared suitable
for excitation of the principal background components, as
well as PCB/DDT. Wavelengths ordinarily used to excite
emission were (in nm): 275, 290, 325, and 344. Wavelengths
ordinarily used to monitor excitation spectra were: 380,
405, 430, and 465. Most spectra were obtained at 77°K
using the SF-100 as described in Section V. Representative
spectra are shown in Figures 17 through 29.
The various environmental samples were analyzed over a
period of nearly ten months. During this time, both the
xenon source lamp and the photomultiplier high voltage
supply were changed several times. Also, the SF-100
collection mirror was cleaned occasionally. All of these
modifications can influence the SF-100 gain setting so
that these are not strictly comparable. However, spectra
28
-------
to
TABLE 1
ANALYTICAL PARAMETERS USED FOR ENVIRONMENTAL SAMPLES
Collection
Site
Shawsheen R.
Shawsheen R.
Shawsheen R.
Shawsheen R.
Shawsheen R.
Concord R.
Atlantic O.
Milwaukee R.
Charles R.
Water
Volume
Date in liters
Collected (no. bottles)
2/21/73 1.7(1)
2/21/73
5/24/73
6/25/73
9/26/73
8/23/73
10/2/73 6.8(4)
10/12/73 3.4(2)
12/6/73 1.7(1)
DCM DCM
Volume Volume Volume
HC1 Per bottle, DCM After
Added (ml) Per extract Stirring Concentration Figure
Per bottle (ml) Time (hrs) (ml) No.
3 50 3 1
0 " '1
5 75 11 1
" ' 15 2*
1 d**
96 1
2 d**
10 150 2 d**
5 75 6 d**
17,
18
20
21
22
23
24,
19
25
26-28
29
*1 ml of DCM combined with 9 ml MCH.
**DCM extract evaporated nearly to dryness on rotary evaporator.
Final volume not measured, but probably <_ 0.5 ml. To this was added 10 ml MCH.
-------
of the toluene impurity in MCH were taken routinely throughout
this period to check the cleanliness of the quartz sample
tubes. The relative intensity of this emission can be
taken as rough "standard," although it has several short-
comings, such as the variation in toluene content between
different bottles of the same solvent lot and irregularities
in sample tube geometry. The intensities of the toluene
emission indicate that the relative intensity as inferred
from instrument gain settings are probably comparable within
a factor of two, except for the first Shawsheen River
sample (Figures 17 through 19), for which the instrument
gain settings are about ten times higher.
Room Temperature Absorption Spectra—Room temperature
absorption spectra were taken of a Shawsheen River extract
(collected 2/21/73) in order to determine whether there
was significant spectral information. In a solvent con-
sisting primarily of MCH, spectra in the 240-460 nm region
are very broad and bear some resemblance to those of humic
and fulvic acids appearing in Schnitzer and Khan (1972);
the spectra (in aqueous solution) shown in this work are
diffuse throughout the region 200-400 nm with a broad
maximum near 220 nm.
Although some PCB/DDT absorption could probably be discerned
above this background at high concentrations, much greater
sensitivity is obtainable from low temperature phosphorescence
analysis with fewer potential interferences, since not all
absorbing compounds emit. For these reasons, room temperature
absorption measurements were discontinued.
Room Temperature Fluorescence—Since neither PCB's nor
DDT-type compounds show significant fluorescence but do
phosphoresce strongly at low temperature, room temperature
analysis for these compounds is generally not profitable.
However, some room temperature measurements were included
simply to gauge the spectral behavior of background material.
Also, in those instances where spectral evidence of PCB's
(or DDT) is found, room temperature analyses may help con-
firm whether the observed emission is fluorescence or phos-
phorescence (without the need for a phosphoroscope).
As an example of the room temperature emission, an extract
of Shawsheen River water in MCH is shown in Figure 19. The
principal emission extends from about 300 to 600 nm, with
an apparent maximum in the 400 nm region. The principal
excitation spectrum is largely continuous from 220 to 400 nm,
although several narrow bands appear above the continuum.
30
-------
The continuous excitation spectrum is rather similar to
that observed in the absorption spectrum and is again
probably due to humic material. Shapiro (1957) has found
that lake water contains 2-4 ppm of fluorescent yellow
material. These compounds may be carboxylic acids, with a
mean molecular weight of 456.
Fluorescence spectra obtained in this study are similar to
those found for marine water. In marine water, the water
soluble fluorescent substances are known as "Gelbstoffe"
and are produced from decayed plant material. Representa-
tive spectra of Gelbstoffe appear in Figure 30, from Hornig
and Eastwood (1973). Spectra in this report were obtained
on untreated water samples at ambient temperatures in 1 cm
rectangular curvettes. The principal excitation/emission
peaks (uncorrected) are typically at 350 and 440 run,
obtained with the same instrument used for this work.
These wavelengths differ somewhat for the river water
extract, and these differences may arise from several
sources including the biological history of the sample,
possible fractionation of organic material by DCM, and
spectral shifts induced by the solvent. Despite these
differences, however, the origin of the background fluorescence
in both fresh and marine water is probably due mostly to
plant decomposition products.
Another natural component commonly observed in the water
extracts is chlorophyll. An example of this emission
appears in Figure 31. The emission consists of a single
rather narrow band at about 670 run, with an excitation
maximum near 420 nm. This emission is fluorescence and
would be observed at room temperature as well. The spectra
of chlorophyll are well separated from DDT/PCB and would
not constitute an interference.
Low Temperature Emission—The background features of the
low temperature(77°K)spectra are generally similar to
spectra obtained at room temperature, but with several
important differences. First, the overall emission inten-
sity is approximately three times greater at low temperature.
This increase is too large to be attributed to volume
shrinkage alone (about 20% for MCH) and probably reflects
increased fluorescence yields due to diminished collisional
and/or oxygen quenching.
In addition, much more fine structure appears above the
broad background at low temperature. The most prominent
structure consists of several sharp emission bands beginning
31
-------
at about 370 run (Figure 17). The structure of this system
bears a strong resemblance to that of pyrene (Figure 32),
and the observed compound is probably a pyrene derivative.
When this emission was first observed in Shawsheen River
water, it was thought to possibly originate from auto
exhaust condensate washed into the river from a nearby
highway. Since then, similar structure has been found to
some degree in all natural water samples and may suggest
a natural origin. If this compound has a quantum yield
similar to that of pyrene, its concentration in water is
in the range 0.1-1 parts per trillion (ppt).
Another relatively sharp spectrum was observed in the
Milwaukee River sample (Figures 26-28). The original MCH
solution was deep yellow, and the high fluorescence intensity
suggested that the concentrations were no longer in the
linear region. For this reason the sample was diluted ten-
fold with MCH. Spectra of the diluted sample (Figures 27
and 28) had generally the same features as the undiluted,
except for the appearance of a sharp band system originating
at about 401 nanometers. Comparison of the excitation/
emission spectrum with that of benzo(a)pyrene (Figure 33)
suggests that this may be benzo (a)pyrene or a derivative.
If this is benzo (a)pyrene, the concentration in water is
on the order of 0.01 ppb. Since benzo (a)pyrene is a
potent carcinogen, its possible presence in water is of
interest. In a recent study of several New Hampshire
rivers, Ellis (1972) has found trace (parts per trillion)
concentrations of dibenz(ah)anthracene, benzo(a)pyrene,
benzo(b)fluoranthene, and fluoranthene. The fact that
these compounds were found in relatively unpolluted waters
led the author to suggest a natural origin.
River Water Doped with Aroclor 1254 and p,p'-DDT
These experiments were undertaken in order to establish
recoveries and also to determine possible spectral changes
induced in Aroclor 1254 by natural organic matter; for
example, by differential adsorption or extraction of PCB
isomers. In both cases solutions were doped with ethanol
solutions of Aroclor 1254 and p,p'-DDT as described in
Section VI.
Solutions were mixed thoroughly on the magnetic stirrer
for periods ranging from one hour to one day. Extraction
was done by addition of successive 75 ml aliquots of DCM.
Stirring with DCM was done on the magnetic stirrer, with
32
-------
the solution agitated sufficiently to break the DCM into
a fine mist throughout the water; extraction times were
one hour or longer. The DCM extracts were dried, concentrated,
and diluted with MCH in the manner described previously for
undoped samples.
Spectral Features—Figures 34 and 35 show spectra obtained
for the first extract of 6 and 60 ppb Aroclor 1254 in 1.7
liter samples of Shawsheen River water. These samples
were stirred continuously for one day after doping, then
extracted three times with 75 ml portions of DCM (each
extract stirred for one hour).
Both samples were stirred continuously for one day after
doping, and then extracted three times with DCM (each
extract stirred one hour).
The Aroclor 1254 emission maximum at 440 nm at the 6 ppb
level is largely lost in the background; excitation spectra
monitored at 430 and 465 nm do show the expected Aroclor
peaks near 260 and 290 nm. Spectra of the 60 ppb solution
show distinct Aroclor excitation and emission signatures.
The structure of both excitation and emission does not
appear to be significantly different from that of the
Aroclor 1254 standard in MCH (Figure 9).
A similar experiment was done by spiking p,p'-DDT into
another sample of the same water at concentration of 60 ppb.
Whereas the emission spectrum, peaked at about 400 nm, is
largely obscured by the background, the sharp excitation
peaks at about 240 and 278 nm are clearly visible (Figure 36).
Recovery and Extraction Efficiency—Recovery of Aroclor 1254
from the natural water samples was 19% for the 6 ppb solution
and 27% for the 60 ppb solution. Of the amount extracted,
75-80% was contained in the first extract and an additional
15% or more in the second extract (both concentrations).
A third extract of the 60 ppb sample contained about 1%
of the total.
Recovery for a similar water sample spiked with p,p'-DDT
to a concentration of 60 ppb was about 35%. Of the amount
extracted, about 80% was contained in the first extract
and 15% in the second.
Reducing the incubation period with increased extraction
time did not substantially improve recovery. For example,
33
-------
a 6 ppb sample of Aroclor 1254 in Shawsheen River water
incubated for one hour and stirred with DCM for 15 hours
still gave only .a 24% recovery. Recoveries from river
water are thus much lower (2-4 times) than from pure water.
Possible reasons for this are discussed in a later section.
Detection Sensitivities—The detection limit for Aroclor
1254 spiked into raw river water of the type used here is
about 2 ppb, and for p,p'-DDT is about 10 ppb. These rather
high values are due to low recoveries and to the presence
of large amounts (probably several ppm) of luminescent
background material in natural water. Removal of this
background material prior to analysis should result in much
lower detection limits. Assuming a recovery of about 25%,
the detection limits of Aroclor 1254 and p,p'-DDT should be
on the order of 0.1 and 1 ppb respectively for sample
volumes of about one liter. Larger water volumes (and
higher recoveries) would of course permit better sensitivity.
Discussion—The low recovery obtained for the environmental
samples may be due to adsorption by particulates. Golden
and Sawicki (1973) have investigated ultrasonic extraction
for removal of aromatic hydrocarbons from air particulate
matter collected on glass fiber filter paper. In order
to quickly establish the potential of this technique for
the problem at hand, the particulates from the 60 ppb
Aroclor sample were placed in a small flask along with about
10 grams of Na2SO4 and 100 ml of DCM. The mixture was
then subjected to 50 watts of ultrasonic power from a 1/4"
diameter probe for about ten minutes. The probe was passed
through the slurry to promote better contact. Subsequent
concentration and analysis of the DCM revealed only a very
small (about 1% of the previous total) additional quantity
of Aroclor 1254 had been removed. However, the fact that
some improvement was obtained suggests that this technique
deserves further study. For example, higher power levels
and improved contact between probe and solids might give
better recovery.
Aside from adsorption by particulates, low recovery may be
a consequence of enhanced water solubility of these com-
pounds by humic substances. Wershaw et al (1969) have
found that a 0.5% sodium humate solution in water solubilizes
DDT. The DDT solubility in pure water is normally about
40 ppb (Babers, 1955) but is at least twenty times larger
with sodium humate. The sodium humate may function as a
sulfactant which stablizes a dispersion of DDT in water.
Since filtration of DDT-humate solutions through a 0.45V
34
-------
membrane filter did not appreciably affect DDT concentra-
tion, emulsion particles would have to be smaller than
this.
Still another possibility is that the low recoveries are
only apparent, being a consequence of the measurement
technique. In particular, natural compounds might quench
the DDT or PCB phosphorescence by serving as energy transfer
acceptors. That is, rather than phosphorescence, the
organochlorine may transfer its excitation energy (light
absorbed) to other molecules nearby which subsequently
emit. This mechanism does not appear to be strongly opera-
tive for p,p'-DDT or Aroclor 1254. This was demonstrated
by doping MCH extracts of river water with solutions of
Aroclor and DDT in MCH. In both cases, the intensities
were found to be the same as obtained in MCH alone within
experimental error.
Evidence of PCB/DDT in Natural Waters
None of the environmental samples showed clear evidence of
DDT-type compounds. The only environmental sample showing
evidence of PCB's was the marine sample taken near Gloucester,
Mass. Extracts of this sample show excitation peaks above
the background continuum in the 280-290 nm region. Of the
Aroclors studied, the observed excitation peaks most closely
resemble Aroclor 1254, although even here the resemblance
is not strong; this Aroclor should have a moderately strong
excitation band in the 250-260 nm region which is not
apparent in the environmental sample. If the observed s
structure is assumed to be that of Aroclor 1254, its con-
centration in water would be about 1 ppb. This value
assumes 100% recovery, although experience with spiked
samples suggests that actual recoveries are probably 20-30%.
Harvey et al (1972) have found levels of 1-150 ppt (deter-
mined as Aroclor 1254) in the open North Atlantic, with
higher concentrations occurring nearer the surface. Higher
concentrations, as may be present for the sample collected
here, are expected in coastal areas.
Excitation spectra of the marine sample obtained at several
emission wavelengths show that there are two separate
excitation peaks in the 280-290 nm region. Monitoring emis-
sion at 400-440 nm emphasizes a peak at 284 nm, while
monitoring at longer wavelengths emphasizes a' peak near
290 nm. The later peak is at the wavelength characteristic
of Aroclor 1254, whereas the former is more characteristic
of a phthalate, perhaps di-2-ethylhexyl phthalate (DEHP) or
35
-------
the closely related diisodecyl phthalate (DIP). Spectra of
dibutyl phthalate (DBF) are discussed in more detail in
the following section.
The excitation peak at about 290 nm may in fact be due to
a phthalate, since the secondary absorption of Aroclor
1254 in the 250-260 nm region is not evident. If one or
both of these bands are due to phthalates, their concen-
tration in water would be on the order of 1-10 ppb, assuming
extraction efficiencies of 100%. Kites (1973) and Kites
and Biemann (1972) report finding phthalates in the Charles
River (Boston) at levels of 1-2 ppb. Corcoran (1973)
has reported finding about 0.6 ppm phthalate, probably
DEHP, in the lower Mississippi River.
Background and Interferences
Humic Substances—As discussed earlier, the principal emis-
sion of natural organic matter present in water consists
of a relatively broad background in the 300-600 nm region
for excitation wavelengths below 300 nm. The maximum of
this background emission is in the 400-450 nm region,
which is also the location of the phosphorescence of DDT-
type compounds and the more highly chlorinated Aroclors.
Since the emission of these compounds is also broad, they
are not easily distinguished from background at low con-
centrations. The interference is less severe, however, if
excitation spectra are monitored. This is because the
excitation spectrum of the background material is usually
continuous in the 220-300 nm region where the absorption
of DDT and PCB's occur. The excitation spectra of the
latter compounds, particularly DDT, are relatively narrow
and can therefore be more easily distinguished above the
background continuum.
It would be desirable for future work to first remove as
much of these natural organics as possible, particularly
those emitting in the 400 nm region. Several simple approaches
seem attractive. First, if the majority of the fluorescent
compounds are acids, as suggested by Shapiro (1975), making
the water sample basic should convert the acids to the much
more soluble salts. These salts should have much lower
solubility in DCM than the acids and thus remain primarily
in the aqueous phase. To test this, two identical 1.7
liter samples of Shawsheen River water were extracted with
DCM and the concentrated extracts diluted with MCH and
analyzed in the usual manner. One water sample had been
treated with 5 ml concentrated HC1, and the other with
36
-------
2.5 grams NaOH. Unfortunately.- the alkaline sample showed
only a very slight (about 10%) reduction in background
intensity. The quantity of NaOH used here was chosen to
provide a hydroxide ion molarity equivalent to that of the
acid solution and may not have been adequate. This experi-
ment should probably be repeated at even higher pH values.
Blumer (1970) has used a simple chromatographic operation
on silica gel to separate components of marine water into
three groups: saturated hydrocarbons, olefinic/aromatic
compounds, and polar materials. For luminescence analysis,
it would probably suffice to remove only the polar sub-
stances from the remaining compounds since saturated com-
pounds would have negligible luminescence. In Blumer's
procedure, pentane is used as the extractant and is also
used to elute saturated, olefinic, and aromatic compounds
from the chromatographic column.
A similar procedure has been recommended by Johnson (1971).
In this approach, components are eluted from a silica gel
column using benzene-hexane mixtures. Aromatic chlorinated
pesticides are thus separated from chlorinated aliphatics
and from various polar compounds. The possible disadvantage
of this technique for luminescence analysis is the presence
of benzene, which is itself luminescent. However, the
procedure of Blumer discussed above suggests that hexane
alone might be used to elute all except polar materials.
To summarize, luminescent natural substances (probably
largely polar) should be removed prior to analysis. This
may be possible by proper adjustment of the pH "of the water
preceding extraction. If this proves unsuccessful, the
silica gel cleanup procedure of the type used by Blumer
should be contemplated.
Phthalates—An important class of compounds which can
interfere with PCB determination are phthalates (phthalic
acid esters). These compounds, particularly di-2-ethylhexyl
phthalate (DEHP) are commonly used as plasticizers for
polyvinyl chloride. These compounds may be even more
prevalent in the environment than PCB's. Total domestic
sales of Monsanto PCB's peaked at about 75 million pounds
in 1970 but have declined steadily since then (Nisbet and
Sarofim, ^1972). ( Sales jqf phthalates, however, reached nearly
a 'bi'ilfon pounds *ih~I£70" "and "nave proba'bly risen since then
(Graham, 1973). The water solubility of these compounds
may be greater than that of PCB's. Wallnofer et al (1973)
37
-------
give water solubilities of many PCB isomers at 22-24°C
as being in the range 0.001 to 6 mg/100 ml; higher chlorina-
tion generally gives reduced solubility. In contrast, the
solubility of dibutylphthalate is 40 mg/100 ml (Handbook
of Chemistry and Physics).
Phosphorescence excitation/emission spectra of diisodecyl
phthalate (DIP) and dibutyl phthalate (DBP) are shown in
Figures 37 and 38. The spectrum of di-2-ethylhexyl
phthalate (DEHP), being chemically very similar to DIP,
should strongly resemble that of DIP. The emission spectra
are similar in structure and wavelength to those of the
more highly chlorinated Aroclors. The excitation spectra
resemble that of Aroqlor 1254 in the long wavelength region
but lack strong secondary absorptions in the 250-260 nm
region. The phosphorescence intensities of DIP and DBP in
MCH (77°K) are roughly factors of 7 and 2 less than the
same weight of Aroclor 1254.
If only DIP or DBP (or some mixture of these) were present,
excitation structure monitored at several emission wave-
lengths should be nearly the same. In the case of an
Aroclor, which consists of many isomers, rather different
excitation spectra are obtained depending on the emission
wavelength chosen.
Another parameter which should provide some discrimination
between phthalates and highly chlorinated PCB's is the
phosphorescence lifetime. The mean lifetime of Aroclor
1254 should be similar to that of an "average" biphenyl
isomer having five chlorine atoms. Dreeskamp et al (1972)
have measured the phosphorescence lifetimes of several
chlorinated biphenyl isomers. Lifetimes given for 2,2',
4,4'-tetrachlorobiphenyl and 2,2', 4,4', 6,6'-hexachlorobiphenyl
are 0.11 and 0.05 seconds respectively in EPA glass at 77°K.
(EPA is a mixture of ethanol, isopentane, and ether in the
proportions 2:5:5 by volume.) Dubinskii (1959) has deter-
mined the lifetime of DBF as 0.80 seconds in ethanol at
93°K; the lifetime of DIP is probably similar. Thus the
common phthalates probably have phosphorescence lifetimes
up to ten times longer than the more highly chlorinated
PCB's. This lifetime difference can be utilized as follows:
Suppose one has an arbitrary mixture of Aroclor 1254 and
DBP. If one allows a time delay of 0.80 seconds between
excitation and detection, the intensity of the DBP emission
will have fallen to 0.37 (1/e) of its original value,
whereas the intensity of the Aroclor 1254 emission, assuming
a mean lifetime of 0.08 seconds, will have dropped by a
38
-------
factor of 5 x 10 . Thus, by using time discrimination,
phthalates could probably be determined in the presence
of much higher concentrations of highly chlorinated PCB's.
Time discrimination of this type can be achieved using a
phosphoroscope, which is simply a mechanical light chopper
which introduces a time delay between excitation and detec-
tion. A drawback with devices of this type is that usually
much intensity is lost. Alternatively, the source can be
pulsed, but this is no longer in the realm of a simple
laboratory experiment.
To summarize, certain phthalates may interfere with the
more highly chlorinated PCB isomers. Although intrinsic
differences in phosphorescence lifetimes should permit
some discrimination, this would require more sophisticated
instrumentation than employed in the present study.
Accuracy and Precision
In order to estimate accuracy and precision, separate esti-
mates of these are needed for both the extraction and lumi-
nescence procedure. The luminescence measurement itself
is probably reproducible to about 10% for a given sample
frozen repeatedly to 77°K; this variability arises in the
differences of sample cooling rate. This further assumes
that normal variations in the xenon arc intensity are
properly compensated by use of a standard. If the same
sample is analyzed in different quartz tubes, additional
variations of up to 15% can be expected due to the vari-
ability in quartz tube dimensions. The accuracy of a
given determination, using a randomly selected quartz tube,
would probably be better than 25% presuming that the analyte
concentration were not too high (i.e., beyond the range of
linearity)-
The error introduced by the extraction procedure is less
easily established since recoveries of the compounds of
interest are probably dependent on both the nature of the
container (e.g. adsorption on glass) and the presence of
other materials in water (e.g., particulates and natural
surfactants). Also, determination of recoveries by the
luminescence method are also subject to the inaccuracies
of the latter as discussed earlier. An intensive study of
this subject was not undertaken, but most determinations
of PCB/DDT doped into water were run in duplicate, with
an estimated precision of about 15%.
39
-------
Recoveries from natural waters were as low as 20%, so that
if typical recoveries were completely unknown, concentra-
tions could be too low by up to a factor of five. If
recoveries have been determined for a particular type of
sample, the accuracy should be roughly equal to the
estimated precision (15%).
For the combined extraction/luminescence analysis, the
accuracy and precision are approximately the sum of the
values for extraction and luminescence determination
separately. These estimates have been summarized in
Table 2.
TABLE 2
ESTIMATED ACCURACY AND PRECISION OF
THE LUMINESCENCE METHOD
Accuracy (%) Precision (%)
Luminescence
Measurement* 10-25 ~10
Total Analysis** 25 - 40 -25
(Extraction + Luminescence)
*Estimates based on measurement of standard solutions
in methylcyclohexane at 77°K using a randomly selected
sample tube.
**Estimates based on recoveries of Aroclor 1254 and
p,p'-DDT from doped water samples.
40
-------
Improved estimates of accuracy and precision require that
recoveries be separately determined for the particular
type of water sample under investigation. Since the esti-
mates given here are subject to errors in the luminescence
measurement, recoveries should preferably be checked using
an independent method such as gas chromatography.
41
-------
SECTION VIII
SUMMARY
Basic studies consisted of the documentation of the low
temperature luminescence spectra of several PCB mixtures
(Aroclors) and DDT-type compounds, and the application of
these results to the development of an analytical procedure
for these compounds in water. A detailed description of
the instrumentation and methodology is given in the
Appendix.
Spectra of Aroclors,. Pesticides/ and their Mixtures
Luminescence spectra of several PCB isomers, Aroclors, and
DDT derivatives were given in the first final report of
the present program (Brownrigg et al 1972). Since some
of these data were obtained on a slightly different instru-
ment than used for the present work, spectra of some of the
more important Aroclors and pesticides have been repeated
here. Compounds studied were p,p'-DDD and DDT, and Aroclors
1016, 1248, and 1254.
Luminescence measurements were performed at liquid nitrogen
temperature (77°K) using a standard Baird-Atomic SF-100
Fluorispec. The solvent selected was methylcyclohexane,
which forms a clear glass when frozen rapidly to 77°K.
Trace amounts of luminescence impurities (mainly toluene)
present in the solvent were removed by column chromatography
using silica gel. Removal of these impurities resulted in
improved detection sensitivites over those given in the
first report.
The phosphorescence intensities of pfp'-DDT and the three
Aroclors were found to be linear with concentration below
about 10 ppm. Detection limits for DDD/DDT and the Aroclors
are about 0.03 and 0.002 ppm respectively, using highest
spectral resolution (2nm). Limits are imposed primarily
by source/phototube noise, and could be improved by using
wider slits at the expense of resolution.
Several mixtures of Aroclor 1254 and p,p'-DDT were studied
with concentrations ranging from 0.1 to 9 ppm. The phos-
phorescence intensity of each component was linear over
the concentration range, except for the solution containing
1 ppm DDT and 9 ppm Aroclor 1254, in which the DDT intensity
was almost 50% lower than expected. The observed departure
42
-------
from linearity is thought to be caused by energy transfer
quenching from DDT to PCB molecules. Since a ten-fold
dilution of this solution behaved normally, a possible
solution to problems of this type might be sample dilution,
provided loss of sensitivity were not too large.
Aroclor 1254 and DDT in Pure Water
Distilled water samples of approximately one liter were
doped with Aroclor 1254 and p,p'-DDT (in ethanol solution)
in order to determine recoveries and extraction efficiencies
After allowing the samples to stir for up to a day, samples
were extracted with two or three successive 75 ml volumes
of dichloromethane (DCM). This particular solvent was
selected for extraction because of its low boiling point,
low cost, and generally high purity. Extraction was
accomplished by stirring the mixture vigorously for at
least one hour with a Teflon-coated bar magnet, activated
by a magnetic stirrer. Extracts were dried with 10-20
grams anhydrous sodium sulfate and concentrated almost to
dryness with a rotary evaporator. To the concentrated
extract was then added 5 or 10 ml of MCH, and the lumi-
nescence analyzed at 77°K.
Recovery of Aroclor 1254 was about 70% at the 1 ppb level
and about 80% at the 10 ppb level. Of the total amount
recovered, 75-85% was contained in the first extract and
95% of more in the first two extracts combined. Similar
experiments with p,p'-DDT gave recoveries of about 50%
at the 10 ppb level and 70% at the 100 ppb level. Extrac-
tion efficiencies were similar to that for the Aroclor,
with 80-90% of the total in the first extract and 95% or
more in two combined extracts.
The difference in recovery between Aroclor 1254 and DDT
is probably significant but is not presently understood.
Observed losses are probably due mainly to adsorption on
glass surfaces. In addition, since the water samples
were not filtered, some additional (presumably minor)
losses due to adsorption on particulates may have occurred.
Detection limits for DDT and Aroclor 1254 in one-liter
samples of pure water are estimated at 0.5 and 0.03 ppb,
assuming recoveries of 60 and 75% respectively.
43
-------
Aroclor 1254 and DDT Doped into Natural Water
Samples of water (1.7 liters) taken from a nearby river
were doped with Aroclor 1254 and p,p'-DDT and analyzed in
the same manner as described for the pure water samples.
The river water samples were yellow-green and contained
participates, but were left unfiltered for these studies.
Recoveries obtained for Aroclor 1254 were 19% at the 6 ppb
level and 27% at the 60 ppb level. Of the total extracted,
75-80% was in the first extract and an additional 15% or
more in the second. Recovery for a similar sample doped
with p,p'-DDT at the 60 ppb level was about 35%, with 80%
of the total in the.first extract and 15% in the second.
The recoveries from raw river water are much lower (2 to
4 times) than from pure water. This is thought to be due
primarily to increased adsorption by particulates, but
increased solubilization by humic material may also be
responsible. Possible quenching of PCB/DDT by other sub-
stances present in water does not appear to be significant.
It would have been of interest to determine the relative
importance of these mechanisms by analyzing filtered river
water. However, since a proper definition of natural water
should probably include suspended particulates, no attempts
were made to remove these from any of the environmental
water samples prior to analysis.
Detection limits for Aroclor 1254 and p,p'-DDT in 1.7 liter
volumes of raw river water used here are about 2 and 10 ppb
respectively. These high values are due to low recoveries
and the presence of a strong background luminescence from
natural substances, probably humic material. If these
compounds (presumably polar aromatics) were removed prior
to analysis, detection limits for Aroclor 1254 and DDT
could probably be reduced to approximately 0.1 and 1 ppb
for a one-liter sample volume, assuming 25% recovery.
Improved extraction techniques and analysis of larger
sample volumes could further reduce these limits.
Environmental Water Samples
Grab samples of 1.7 liters were taken from three rivers in
the Boston area and also from the ocean near Gloucester,
Massachusetts. A sample of water from the Milwaukee River
was collected by A. Hornig on a return trip from this area.
All samples were acidified with 5 ml concentrated HC1 and
then extracted and analyzed in the same fashion as described
for the pure water samples.
44
-------
All samples show broad background emission peaking in the
400 run region which is probably due to humic material, par-
ticularly aromatic carboxylic acids and phenols. Narrow
bands are observed superimposed on the broad background
emission. One particular system, probably due to a pyrene
derivative, is observed in all water samples. If this
material were pyrene, its concentration in natural water
would be in the range of 0.1 - 1 ppt. The wide distribu-
tion of this compound suggests a natural origin. Another
sharp spectral system was observed in a sample of Milwaukee
River water. This spectrum resembles that of benzo(a)pyrene
or a derivative; it this is a correct identification, the
concentration in water would be on the order of 10 ppb.
The broad emission contributed by humic substances overlaps
the similarly broad emission of the Aroclors and DDT and
therefore constitutes an interference. However, the excita-
tion spectra of the humic materials appears to be nearly
continuous in the 220-300 run region where the excitation
(absorption) spectra of PCB's and DDT occur. Since the
excitation spectra of the latter compounds, particularly
DDT derivatives, are much narrower, they are more easily
distinguished above background than are the emission spectra.
Even so, the background emission contributed by humic
materials diminishes sensitivity and future studies should
be directed at removing these prior to analysis.
None of the environmental samples studied shows evidence of
DDT-type excitation peaks, implying DDT concentrations of
less than 10 ppb. Only one of the samples, namely a sample
of marine water taken near Gloucester, Massachusetts,
shows evidence of excitation peaks in the region expected
for Aroclors. Two distinct excitation peaks are observed
at approximately 284 and 290 nm. The latter is close to the
value expected for Aroclor 1254, although a secondary Aroclor
peak near 260 nm is not evident. The 284 peak resembles
that obtained for diisodecyl phthalate, and both this peak
and the one near 290 nm may in fact be due to phthalates.
If this is the case, the phthalate concentrations found
would be on the order of 1-10 ppb. In general, the spectra
of phthalates are similar enough to those of Aroclors to
result in possible interference, and further work on this
problem is necessary.
Methodology
The results of the experimental work performed in this study
have been analyzed, and a first cut has been taken in
45
-------
producing a practical "method" which appears as an appendix.
The method considers extraction techniques to prepare a
water sample for instrumental analysis, detailed instruc-
tions for specific instrument settings to be used, and
data interpretation.
Grab samples of water are collected, extracted with
dichloromethane, and the combined extracts dried and con-
centrated. Methylcyclohexane is then added to the con-
centrated extract and luminescence analysis performed at
liquid nitrogen temperature (77°K). Identification and
estimation of Aroclors and DDT derivatives are based upon
comparison of the unknown excitation/emission signatures
and intensities with those of appropriate standards in
methylcyclohexane.
Known samples of Aroclors and DDT in methylcyclohexane are
used to calibrate the fluorescence instrument. Doped
water samples are used to gain familiarity with the extrac-
tion and concentration techniques, and environmental samples
are treated in a similar fashion.
Recommended wavelengths, slit widths, and other operational
parameters for the fluorescence instrument are given.
Aroclor/DDT spectra shown in the main text can be used as
a guide. Apparatus and chemicals necessary for collection
and extraction of water samples are described. Finally,
accuracy, detection sensitivities, and possible interferences
are discussed.
46
-------
Triplet-
System
Singlet
System
Inters ystem
Crossing (b)
Intersystem
Crossing (a)
Intersystem
Crossing (c)
FIGURE 1. SCHEMATIC DIAGRAM OF MOLECULAR TRIPLET
AND SINGLET ENERGY LEVELS
47
-------
500
600
WAVELENGTH ( nm )
FIGURE 2. BIPHENYL EMISSION IN METHYLCYCLOHEXANE (MCH) AND IN HEPTANE
700
SF-100 Wavelength Calibration .=
Apparent ( Dial ) Wavelength ( nm )
FIGURE 3. WAVELENGTH CALIBRATION CURVES FOR THE SF-100 FLUORISPEC USED IN THIS STUDY
48
-------
Sample : p,p'-DDT
'=. Concentration : 10 ppm in MCH
X ex : 275 ( 3/10 )
X em : 400 ( 1/10 )
3 Slits: 33/11; 11/33
400 500
WAVELENGTH (nm )
FIGURE 4. P,P'-DDT, 10 PPM IN MCH, 77-K
700
Sample : p,p' - ODD
Concentration : 10 ppm In MCH
ex : 275 ( 10/5 )
I X em : 400 ( 3/5 )
h Silts: 33/11; 11/33
400 WAVELENGTH ( nm )500
FIGURE 5. P, P"-DDD, 10 PPM IN MCH, 77°K
49
-------
1000
® P/P'-DDT
Q Aroclor 1016
A Aroclor 1248
O Aroclor 1254
i I
i !
10
100
Concentration (ppm) in MCH
FIGURE 6. ANALYTICAL CURVES FOR P, P'-DDT AND SEVERAL AROCLORS
50
-------
Sample : Aroclor 1016
Concentration : 1 ppm in MCH
\ ex: 275 ( 3/10 ) ; 290 ( 3/10 )
X em : 400 ( 30/5 ); 440 ( 3/5 )
465 (3/5 )
400 500
WAVELENGTH ( nm )
FIGURE?. AROCLOR 1016, 1 PPM IN MCH, 77°K
700
Sample : Aroclor 1248
Concentration : 1 ppm in MCH
X. ex : 275 ( 3/10 ) ; 290 ( 3/10 )
m: 400 ( 3/5 ); 440 ( 30/5 )
465 ( 30/5 )
==ii Slits: 33/11 ; 11/33
200
300
400 WAVELENGTH ( nm )500
FIGURE 8. AROCLOR 1048, 1 PPM IN MCH, 77°K
600
700
51
-------
t Sample : Aroclor 1254
j--r^i Concentration : 1 ppm in MCH
X ex : 275 ( 3/10 ) ; 290 ( 3/10 )
X em : 400 ( 10/5 ); 440 ( 3/5 )
465 ( 3/5 )
Slits: 33/11; 11/33
400 500
WAVELENGTH ( nm )
FIGURE 9. AROCUOR 1254, 1 PPM IN MCH, 77°K
Sample : Aroclor 1254
Concentration : 100 ppm in MCH
ex : 275 ( 10/5 ) ; 290 ( 10/5 )
»• em : 400 ( 30/5 ) ; 440 ( 10/5 )
465 ( 10/5 )
Slits: 22/M; H/22
Recorder Gain : 50 mv / cm
400 500
WAVELENGTH ( nm )
FIGURE 10. AROCLOR 1254, 100 PPM IN MCH, 77°K
52
-------
pr,..,
<_ Sample: Arocl or 1254 + p,p'-DDT
Concentration : 0.9 ppm + 0.1 ppm in MCH
X ex : 275 ( 10/5 ) ; 290 ( 10/5 )
= »• em : 380 ( 30/10 ) ; 400 ( 30/5 )
440 ( 10/4 ) ; 465 ( 10/5 )
Slits: 33/11; 11/33
300 "OO WAVELENGTH (nm)500 °°°
FIGURE 11. AROCLOR 1254 (0.9 PPM) + P, P'-DDT (0.1 PPM) IN MCH, 77°K
700
Sample- Aroclor 1254 + p,p'-DDT
Concentration : 9.0 ppm + 1.0 ppm in MCH
ex: 275 ( 1/5 ) ; 290 ( 1/5 )
X em: 380 ( 10/10); 400(3/5)
440 ( 1/3 ) ; 465 ( 1/3 )
Slits: 33/11 ; 11/33
400 500
WAVELENGTH ( nm )
FIGURE 12. AROCLOR 1254 (9 PPM) + P, P'-DDT (1 PPM) IN MCH, 77°K
53
-------
£lr Sample : Aroclor 1254 + p,p' -DDT
Concentration : 0.5 ppm + 0.5 ppm in MCH
ex : 275 ( 10/5 ) ; 290 ( 10/5
X em : 380 ( 30/5 ); 400 ( 10/10 )
440 ( 10/5 ) ; 465 ( 10/5 )
33 Slits: 33/11; 11/33
WAVELENGTH ( nm ) '
FIGURE 13. AROCLOR 1254 (0.5 PPM) + P,P'-DDT (0.5 PPM) IN MCH, 77°K
Sample: Aroclor 1.04 + p,p--L>UI
Concentration : 5.0 ppm 4 5.0 ppm in MCH
X ex : 275 ( 1/5 ); 290 ( 1/5 )
Xem: 380(3/10); 400(3/5)
440(1/5); 465(1/5)
m Slits: 33/11; 11/33
400 500
WAVELENGTH ( nm )
FIGURE 14. AROCLOR 1254 (5 PPM) + P, P'-DDT (5 PPM) IN MCH, 77°K
54
-------
"
— -
_.L-
^
„
_
=-i-
—
_
.V
_
_,.
-
"
"" 1
1
:
Em
X
I
4
==— 1-
-
-c*
J:r—
H"
-r=F
:T|_
-^!
J ~
ngj^
Si'"
J —^
* ' 1
t /
/A
H
' i
381
— ;
1
«
-, f
>T
r
:F
t
1
„
i t
1
t
'*
1
•
_
-
„-
-r.
-
j
f
-
"--
sr
—n
—
--
440 1
I
/
I
-
/ j
ri-
/I!
>J
* -a-
r *
4^
_
-
.
TK
-
1,
-
\
i.
i
-
^^
=
T:J '
— -
— •
-
d- ,
_
.£
'•ff
'-
_1_P1J_^..JJ
: 1
'!
t
t
^ ^J
r-
i?1
1
„
"-
-;•- —
-—j
-.... .
.!:
•-
t
>
i
t.
\^ t
1'
f
'U
i
"-t
4
-y
^
>-U
,
1
•p
f
I_
4
/
^
-
-
---F^j-^-4-
— i-f-
i
Ex 27
! J,
t
I
/
^
/
/
/
,
i
f
u
v
V
•*
/*
- !
-,
- /
J~
-r-
\~
-----
-r-
V:
:\
\
,
Ex 29
5"" —
N
\
t
i
1
I
1
-1
t
f
t-
:
' t "
\
it
o x
-
\
-X
1-
-1—
V
A
S '
s
1
1 *"
I ~~
---
Sample: Aroclor 1254 -h p,p'-DDT
Concentration : 0.1 ppm + 0.9 ppm in MCH
>. ex : 275(30/5); 290(30/5)
X em: 380 ( 30/5 ); 440 ( 10/10)
Slits: 33/11; 11/33
t i
—
j
t
, i
, i
V
, \
—
-j—
I
i
1
I
1
i
v \l
^
i
i
1
. V
V
' \
i-
-
t
I-
I
j
-1
y^
r-v
i
1
~
-"
^=
'
J
i
1
I
I
1
.
I
!
,
>
---.
-
•=
-
t -
-~
-
~=
-
1
- '
,
,
1
'
— :
1
i
- =
i
-±-
-
.1
~ '
_
J
•
_
i -
^H
IS
•=;==
t —
~_-
.
-=•==
-4=
-^=
c-J
j^^
--Ar
i
1
'
! .
-^
"T~:
=--p=
200
300
WAVELENGTH (nn,)
500
60°
700
FIGURE 15. AROCLOR 1254(0.1 PPM) + P, P'-DDT (0.9 PPM) in MCH, 77°K
Sample: Aroclor 1254 + p,p'-DDT
Concentration : 1.0 ppm + 9.0 ppm in
». ex : 275 ( 3/10 ) ; 290 ( 3/10 )
X em : 380 ( 3/10 ) ; 440 ( 3/5 )
Slih: 33/11; 11/33
WAVELENGTH ( nm J
FIGURE 16. AROCLOR 1254 (1 PPM) + P, P'-DDT (9 PPM) IN MCH, 77°K
700
55
-------
Sample : Shawsheen River ( 2/21/73
ex : 275 ( 100/5 ) ; 290 ( 100/5 )
325 ( 100/5 ); 344 ( 100/5 )
22/11
M. : 6 (approx. 800 v. )
ments : Water sample acidified 3 ml
prior to extraction
WAVELENGTH ( nm ) 5°°
600
FIGURE 17A,B. SHAWSHEEN RIVER (2/21/73) EXTRACT IN MCH, 77°K (ACIDIFIED SAMPLE)
522S Sample : Shawsheen River ( 2/21/73 )
; 405(100/5);
430 ( 100/5 ) ; 465 ( 100/5 )i
n Slits: 11/22
i. ; P.M. . 6 ( approx. 800 v. )
££= Comments : Water sample acidified 3ml
£?3 HCI prior to extraction
L!iI±>k:LJ_L i I"--]
200
WAVELENGTH ( nm )
56
-------
INTENSITY
RELATIVE
2C
F=p-
==
n
—
B;
- ~
—.
=
3r-f—
~ :_
= 3
-F-
r=i -
=F=
—
:-;
==L-
3
™
=
—
~
-
L- -
— 1
-_T"
rsr
3£| _~
I
"
-
~_
~
=
—-
~-
-
"
.=
h^:
=--
"—
~~L~
==j_
1 _
L
1
I
"
:=-
==
-^.
_
"r
!
~*
I ~
_
-
-
~
"•H
TH
--
--
3S
;-_•
1
,
!
r
_—
-
I
hn
i~
r
-=,
-
~
.tH
Si
=4-2
>^:
lr
i -
- =
-t-
---•
•=~
-Jr_
- t_ -
....j .
!
VW\
r
10
*
=^r_"
H
_.u.
j-
t -
••i
— :--"
^
3
X.
i
|
\
\
==
Pi1
,
t
-JZ
=•
-.--
/
t^
~
i
_^
f
-
" ~T
i
i
.
j_;
-
-
_ -_a
=Sf~'
-~
H
-V
/
4
;
1
1 Ir- Ex 344
S il J I
ii/i
1
liB
r
nrr
-- I B j i
=jf !
TUT
^P
1
i7
s
i
j
300
-
!
i
'
t
>
\
\
\
•m ,
- ,
i^>
^-f
i
i
400
S
s
_
L
[=
=^=
—
"v--
-^
~
A
s
=s!i
-
-
i
-
-
r_r _
*e
s=
x290
H
275
r
~^h
"4^
t~
>s
I
1
i
i
,
*
r=
\
>\
•r "
S
N
~ t _
i
,
Sample Shawsheen River ( 2/21/73 )
X ex 275 ( 100/5 ) ; 290 ( 100/5 )
325 ( 100/5 ) ; 344 ( 100/5 )
Slits 22/11
P.M. 6 (approx. 800 v. )
Comments Water sample not acidified prior
to extraction
' "]
V
\
V
s,
s.
1-r
-.
-i
1
|
1 '
^±-
p4=-
\
r
A:
< 1 *•
H -
-v
500
*
_
=?
J-
=^
=
-=r=-
H-
SJ>|
s->2<
_=
-—
-
_
-_?
r=
•^
-=
•
~
"-=
^
^
1
_
-
_ I -
t '
u_i-_
—
1 ~
- -f~
i
~ ~J ~~—
~r*^
%•
-
-
_
•-
=
—
^-_' -
"tr—
- r.
j
~*i
-,
=_-
-=;-=-
~ i^
-
i
HH
i
j....
t
-j-
-f-T
-
T\
^
=.=
- -=
—
~
_! 1
-S-_
— =
=^
i
=s
—
--£=
_,.-
. " "
A —
•== S
--{•••"JT^S
600 700
WAVELENGTH ( nm )
FIGURE 18A,B. SHAWSHEEN RIVER (2/21/73) EXTRACT IN MCH, 77°K (SAMPLE NOT ACIDIFIED)
=g Sample: Shawsheen River ( 2/21/73 )
J *.em: 380(100/5); 405(100/5)
430 ( 100/5 ); 465 ( 100/5 )
11/22
6 ( approx. 800 v. )
Comments : Water sample not acidified prior
to extraction
WAVELENGTH ( nm )
400
500
57
-------
Sample : Shawsheen River ( 2/21/73)
X ex : 275 (300/5 ); 290 ( 300/5 )
325 (300/5 ); 344 ( 300/5 )
H Slits: 22/11
H
f P.M. : 6 (approx. 800 v. )
Temperature : Room
"J7 Comments : Extract of acidified water sample/
jj*i spectra taken in 1 cm rectangular cuvette
600
FIGURE 19A,B. SHAWSHEEN RIVER (2/21/73) EXTRACT IN MCH, ROOM TEMPERATURE (ACIDIFIED SAMPLE)
LU
h-
LU
>
<
200
Sample : Snawsham mve
X em : 380 ( 300/5 ) ; 405 ( 300/5 )
430 (300/5 ); 470 ( 300/5 )
rr Slits: 11/22
| P.M. -. 6 (approx. 800 v. )
I! Temperature : Room
Comments: Extract of acidified water
sample, spectra taken in 1 cm rectangular
cuvette.
400 300
200 WAVELENGTH ( run )
58
-------
400 500
WAVELENGTH ( nm )
600
700
FIGURE 20A,B. SHAWSHEEN RIVER (5/24/73) EXTRACT IN MCH, 77°K
Sample : Shawsheen River ( 5/24/73 )
X em : 380 ( 10/10 ) ; 405 ( 10/10 )!
430 ( 10/10 ) ; 465 ( 10/10 )
200
400 300
200 WAVELENGTH ( nm )
59
-------
I: •} T:: Sample : Shawsheen River ( 6/25/73 )
"T X. ex : 275 ( 30/5 ); 290 ( 30/5 )
325(30/5); 344(30/5)
22/11
400 500
WAVELENGTH ( nm )
FIGURE 21A, B. SHAWSHEEN RIVER (6/25/73) EXTRACT IN MCH, 77°K
bhawsheen River ( 6/25/73 )
380(30/5); 405(30/4)^
430(30/4); 465(30/4)
11/22
200
WAVELENGTH (nm )
60
-------
400 500
WAVELENGTH ( nm )
FIGURE 22A,B. SHAWSHEEN RIVER (9/26/73) EXTRACT IN MCH, 77°K
; Em 405-S
pi Sample : Shawsheen River ( 9/26/73 ) M
I X em : 380 (10/5 ); 405 (10/5 ) fO!
430 (10/5 ); 465 (10/5 ) |i
| Slits: 11/22 r|"
=K 1 lllll''jLllrrrrT: ^HiZI! I:li£gr '~-fl^" ~'~!±- =—•
rll"Lm^
:^1:IS
S^
:;|^H '£->~-
:=lgi
;^i;£iEH,
UJ
I—
sgjnn,
li
HI
gl^lll
;pg
x-
1—r
200
300
) WAVELENGTH (nm)
300
400
500
61
-------
Sample : Concord River ( 8/23/73 )
400 500
WAVELENGTH ( nm )
FIGURE 23A,B. CONCORD RIVER (8/23/73) EXTRACT IN MCH, 77°K
r-'
p=t- t
= —
:•: I-
= k.
=3—
-— =
iL t.
:-_=;
==te
F=T_|
=J~
^
==_
J£L
l=d-
=f=-
^rr-
=3?
- 1
==!-
=-4~
z=L
_-
-
-^—
r
j__
i
~
-
- -
- i -
_
- j-^:
t-
^
r
'•
' .. •=
-
_
-
~-^ -
-1
-
__;
i
1
=
r
--t
-
-i=
~fc
=T—
. I
-,
1 —
~
_:
r-JT-
-
3^
-^,
y
^
i
"-!~
_ -t-
„.!
1
i
i
1
. ,"
h=
-r*
t_
:-.
_-J
_
^
y
1
!
TUIS
_Ir
I:
-f-
=
~
—
'-'
==;
^
j r
.J,
fr-
y
^
y
/
i
=rr
- , "
•
Em
j
• —
r
,'
i
:r
st
s:
T
/
lr
f
_;
L
,
1
,
.
i
-^
I
- t 1 -
' 1 V.*
— • —
- •
-
_
V
-V
-v
\
-
—
__|tH
• r
\
\
;\
=~=-
— —
i
'
1
/
-~~-^
—
.1
:
1 -
-
1
-----
~-^=j=_
--I"
t-:
_-
1H
-ir
i:
1-
!_:
=
-- -
-• =
=
• r *-
1
_
;
H
:.J
.13
.. ::::
I: .-:
I 5=
200
300
I WAVELENGTH ( nm )
300
400
500
62
-------
Sample : Atlantic Ocean ( 10/2/73 )
X ex : 275 ( 3/10 ) ; 290 ( 3/10 )
325 ( 3/10 ); 344 ( 3/10 )
Slits: 22/11
WAVELENGTH ( nm )
FIGURE 24A,B. ATLANTIC OCEAN (10/2/73) EXTRACT IN MCH, 77°K
1 Sample: Atlantic Ocean ( 10/2/73)
Xem: 380(10/5); 405(10/5)
430 ( 10/5 ) ; 465 ( 10/5 )
Slits: 11/22
200
300
200 WAVELENGTH ( nm )
63
-------
Sample : Atlantic Ocean ( 10/3/73 )
ex : 275 ( 10/10 ); 290 ( 10/10 )
325 ( 10/10 ) ; 344 ( 10/10 )
Slits: 22/11
Temperature : Roan
E: " Comments : Same sample in quartz tube
as analyzed at low temperature
400 500
WAVELENGTH ( nm )
FIGURE 25A,B. ATLANTIC OCEAN (10/2/73) EXTRACT IN MCH, ROOM TEMPERATURE
Sample : Atlantic Ocean ( 10/2/73 )
em : 380 ( 10/10 ) ; 405 ( 10/10 )
430 ( 10/10 ) ; 465 ( 10/10
Slits: 11/22
Temperature : Room
(Comments : Same sample In quartz tube
igas analyzed at low temperature.
D ~*» — ,|, i ,,,1 " • i - i i - i i - i
200
300
400
200 WAVELENGTH ( nm )
64
-------
Sample : Milwaukee River ( 10/12/73 )
X ex : 275 ( 1/5 ) ; 290 ( 1/5 )
325 ( 1/5 ) ; 344 ( 1/5 )
Slits: 22/11
700
FIGURE 26A,B. MILWAUKEE RIVER (10/12/73) EXTRACT IN MCH, 77°K
UJ
I—
200
5-.J—.gag Sample ; Milwaukee River (10/12/73 )
= V em : 380 ( 1/5 ) ; 405 ( 1/5 )
300
300
400
500
WAVELENGTH ( nm )
65
-------
Sample: Milwaukee River ( 10/12/73 )
X em : 380 ( 10/5 ) ; 405 ( 10/5 )
430 (10/5 ); 465 (10/5 )
11/22
Comments : Original extract ( in MCH ) diluted
A ^10-fold with MCH.
300
400 300
200 WAVELENGTH ( nm )
400
500
FIGURE 27A,B. MILWAUKEE RIVER (10/12/73) EXTRACT IN MCH, 77°K ORIGINAL EXTRACT DILUTED 10-FOLD WITH MCH
Sample: Milwaukee River ( 10/12/73 )
ex : 275 ( 10/5 ) ; 290 ( 10/5 )
325 (10/5); 344 (10/5); 370 (10/5)
: 22/11
Comments : Original extract ( in MCH ) diluted
fold with MCH.
400 500
WAVELENGTH ( nm )
66
-------
UJ
LLJ
1
200
-I -i-1 4-
Somple : Milwaukee River ( 10/12/73 )
Xex: 301(30/10); 392(30/10)
Slits: 11/11
Comments : Original extract ( in MCH )
diluted 10 - fold with MCH . Arrows
indicate bands resembling benzo ( a ) pyrene.
300
WAVELENGTH ( nm )
500
600
700
FIGURE 28A,B. MILWAUKEE RIVER (10/12/73) EXTRACT IN MCH, 77°K - DETAIL OF STRUCTURE RESEMBLING BENZO(A) PYRENE
INTENSITY
RELATIVE
r
=.-:
=H
==;
.=_
S
F
?
s
T**
•=
Si.
:5I
^_-
E:E
£i±
=^
Ir"
.
— -
~
^n
=
_
-
:':'
• j .
1
I
I .
:=
.-
_r
---
-
-
--
r~-
—
^
..
i
I
1
'
l
1""
1
A
N
i
i
i
~
'
I
/
/
* !
E
-
~
~
1
t
J
ft
Ml
-
1
t
/"
v
j
:
4-
T
/IA
1 1^ l\
f\
t _j
S
i
\i
Ji
,_ j
i
f
i
T ~
_,
I
1
i.
1
.
t
ff
• I
y
.,.
: i
JvJ
(r
"t
-.1 .
-
» 300
V
-
t
r1
|
i
-
t.
,
•i"
t
m
\ V
j\
i
—
—
V
-
1
^
u
1
,
-----
—
i 401
Af
™ .
:...
i i. >
' 1
- i-
•- , i
--fi;-
-: ;....
. — (.-^.i_M —
-
1
Tn4?
•*<-
; -
• - -
•--
•--
.
r
400
_l
.
Sample :
Milwaukee River ( 10/12/73 )
X em : 401 ( 30/10 ) ; 426
Slits: 11/11
_
Comments : Original extract (
dil
• in
1 ,
i
Jted 10 - f
dicate bane
i
! .
.
i
—
'
I
old w
s resc
i:
-•--.
•-H-+
: : T j
500
th M
mblir
— »—
1 «
H-
i
,
:H
9>
e.
>en
-
f
i
600
Ai
zo
-
( 30/10 )
in MCH )
TOWS
( a ) pyrene.
1
...
f=-
-
._.
--
~
==
~
-
=
•
-
--•—:
B
1
•1
70
WAVELENGTH ( nm )
67
-------
II Somple Charles River ( 12/6/73 )
X ex 275 ( 3/5 ); 290 (3/5 )
325 (3/5 ); 344 ( 1/10)
22/11
400 500
WAVELENGTH ( nm )
FIGURE 29A,B. CHARLES RIVER (12/6/73) EXTRACT IN MCH, 77°K
!jjf Somple : Charles River ( 12/6/73 )
380(3/5); 405(3/5)
430(3/5); 465(3/5)
O Slits: 11/22
400
200 V/AVELENGTH (i
500
68
-------
CODE: G4BXM
DATE: 9/li/72
PACIFIC OCEAN:
SOUTHERN CALIFORNIA
(E. B. Scripps) Station 4
EMISSION SPECTRA Excited
(A) ^90 nm
(B) 350 nm
EXCITATION SPECTRUM Monitored
at 440 nm
ZOO
300
400 500
WAVELENGTH(NANOMETERS)
600
700
FIGURE 30. PACIFIC OCEAN (9/13/72) GELBSTOFF LUMINESCENCE AT ROOM TEMPERATURE (FROM HORNIG AND EASTWOOD, 1972)
Sample : Shawsheen River
(5/74/73)
X ex : 414 ( 100/5 )
X em : 671 ( 100/7 )
Slih: 33/11; 11/33
P.M.: 650 v.
Comments : Chlorophyll
200
300
400 500
WAVELENGTH ( nm )
700
FIGURE 31. SHAWSHEEN RIVER (5/24/73) EXTRACT IN MCH, 77°K - CHLOROPHYLL LUMINESCENCE
69
-------
•=- , 1
- I -
n
_
-
-
-~
i -
! -
- ,
.;
r j
=.' T
!r~
1=—
n_!
— - _
Sr1.-
=y--
=r=g_
^. ~
=
-
—
. J--
-n-~
— - -
q
" 1 '
_< :
1
I-
_
r
- '
-
1
--
-
-
-
,
1
1
r~-
,
i-
_
_
t:
j"
i
-
T =
=!-=
-si ~
'"V
_-.
- |
l~
,
-
S
;
.tL
__-
•=
J
"~*^
-
-
i
- I :
.1."
1
•E«
-3t-
R~
/n
^i \
,
= !
r j
J
-t—
i
~
--
_.
—
~ -*
!
"t
-'l
^
"*
-
Em 4
1 _
H
- -
— 1
02-
i
-
. _ _
Jin \
/¥ T-
.t-
- 1
i
1
- -r
i :
i
[- -
|~ '
\
"
J
^
' -
j|-Ex
I
- 141114-
411
\
N
i
1
\
it
II'
x-m
\r
I
\
-,.,..4-.'
'
:-.;.:
i : - ']
-• :
-
•
4
1
!
._.
11V
i
i
i
-
--
-•1
1
l
i
i - :-
_
; j '
v
Sv
-4-
i
I
— i—
i
:
-
i
'N'
— j-
B__
-
-
J
ns
-
-
^X
-
liijJlS
=r=: i-::
1
~~~
i
1
X
-
--
^
!
1
Si
-i
^
-
1
==
-4
:
Sampl e : Pyrene
Concentration : 1 ppb in MCH
X ex : 340 ( 100/5 )
>.em: 402(100/5)
Slits: 22/11; 11/22
P.M. : 7( approx. 900 v. )
Temperature : 77^ K.
Comments : Solution degassed by
bubbling N2 for 2 minutes.
|
1
-
t
—
—
h=
--
•
HU
(
_t
~
^M
~
......
j
rj
—
-==
-
-
':-..[ .;
„.
1
—
-
""1
")=""
l=-
.
-
—
—
™
nn
:HI
"
„
~
==j=
nnnr
i
"I
-l
-
-
=====
iron
.— ^:
s=
^i
•H
=
-~
==
:§
«
3
:3
: ad
rf
200
300
400 WAVELENGTH ( nm )5°°
FIGURE 32. PYRENE, 1 PPB IN MCH, 77°K
600
700
^
-
_.
.-
^
rH
^
^
~
~!
rH
i
~.-\
•• \
• :
-i
— -
_...
ob"
-» —
a
-_
--j
-_-
"_5
r~
;te
=^=
=p
-
-
•H
-=F|
—
—
~i.:
_-;
~T=S
==p=
=F=
T
• -3 "
1-
...... .,
i
==)=£
J
- 1
— ~
£T
=¥
=-^
.
_-
:=
=
t "
J-
'
•1
i
1
...|_
i
I
i-
i
=P
---
t-
f
-i •
j
/^
-
•^T
i
i
t -
-4-
i
—
• ;
^.
s=j==
-4
!
t .
1
1
-r-
i
i
.....
j
i
''
--=):-
r
' "-
,
- —
A
._.^1/.....
300
1 .;..
-H
^rr-
— r
-i
1
Em 4254
I It
-4
A
TV
/ (
fill
f\
.
^
1
— i —
•
==
1 -
^1
.
=
.-=1=
~~=
r
i
-4-
i- _
N
1
I
•j- -
--'
'-
r
1
t
I
Ex 370 !
1 - r -J
; 1
rfr
.. ,. .
=4=
i
-
-
---
-
--
-
>
-
t
.. .1. _
I !ll : :
•U
K|
-.:.! .:
400
I
^
V
i
ni:
-
^
_|
-J
-
-
^
-t
t-
-
—
-
- u
1 -
i
.j
—
_ _ .-
—
"I -
500
=~
S Sample .
ii:
S
-"^5
=f=
=-
=
-
Benzo (a) pyrene
Concentration : 0.2 ppm In MCH
X ex : 370 ( 30/5 )
X em : 425 ( 30/5 )
Slits: 22/11; 11/22
P.M. : 6 (approx. BOO v. )
Temperature : 77° K
i
t.
. ... .L. .
I
1 i
-..
-
-|-
- -
—
i
6
l~
—
~T=-
•r
1
-
i
1
. ;...
30
-
1
--
-
l.
-
I
I
1
'
»±
~
=s
~
'•=?.
_
^^
_
^=
1
— r^~
7«
2 :^-
< ^=£ -
WAVELENGTH ( nm )
FIGURE 33. BENZCXA)PYRENE, 0.2 PPM IN MCH, 77°K
70
-------
UJ
f-
LU
P
"
- -
-
5J--
~ r ~
=TX
^ _-"_
— _ .
tH.
.=ir_
SP
S=="
ks —
Ii
s~
^ i'
r^ ~-
=1 ~
=-K=
===-
B=hf
= =
T7*r|----
3= -_.
an
jg~.
gfS:
1:
- 1
1
I
-j
-" -
-
-_
- —
I/
r~1-_
i^~
Z1L
~-
=TT
1 .
-
==
--•
, •
:-=
~
- r_-'
r L
=
-
;_
i=
Silt
---
r_
m.
if4
^
->
\~
_j -
|r •
t
"'
i
= —
-T^
i ~
-
•=.
-
-
_
^~
"
-=1
~'l
=
„
zni
=-
-.
_•
-
-.
—
-~
— .
-r-j-
r
I
i
1
i.
- ,
i
-(
i
t
— i —
,
j=
-
~~
--
~
i
\
^
-a;
-
r_js=
- !•
1
i
!
J
J
1
. .
i
1
__
-
--
-
~
_-
-
'b
.=
"
i.
1
i
!
i
-
-
-
-
,\
1
t
V
>iV
1 -
1
—
— j —
-1-
I
t
I
1
1
f,
J
3
f
•V
I
•
~
'"•)-
,
i
-1
1
s*
,
f
1
t
aLfx
"-
,
=r
i*
!
1 K
-^
,1
U
1
tl
~n
J'
ifr
T"
I
[
—
~~jS
t
y
t
275
-
•a-
-
fcp
•I
1
^
/
/
£_
^
^
1
i
1 1"
Ex 344
r~
\
\
p
v
\
«325
V
>
—
\
\
•
j
|
1
<29C
!
1
-
I
l
1-
.1
S
s
\
"\
i •
i
71
i*-«i
/
t
i
-t-
.1
. — , —
!
— •
-
\
yJ
\
1
__
^
S*
\
i
V.,
, s
X
1
1
1
t
1
j
-)
-
--•
:.
i
4-
i
S
>K
N1"
N
^ i
X
K
V, s
>
I
--
i
.1
i
-
v '
\
\
s
\
S
'
XL
X
\
|
-
—
,
1
l_
-
i
j
"1
1
I
1
I
I
\
\\
v.\^
\\
V
--
,-
,
-i
-i
,
-!—
,
1
i
1
i
!
t
^
\
\ \
\\
S,V
K,
i
-1
•ata
Sample
X
Slits
Com
1254
t"
I
,
I
1
!
'
t
1
1
Si '•
sk
V
s\
jN
ex
ments
to a
luu
5-
I -
i
• !
i
i -
t
i
\K
H*
Shawsheen River ( 9/
275 ( 3/10 ) ; 29C
325 ( 3/10 ) ; 34<
22/11
Water spiked with
concentration of 6 ppl
-; r
R
— j—
i
i
t
t
i
"^^
S^
J
1
_ ,
'
-
t
--
'
t
—
«
EH
-
~
_
-
^
_~
_
i
1
'
i
„
-
3S
--
'—
^
d=
-
-
1
1
t
1 .
.
=
=-
—
m
"-
^26/7
(3/
t(3/
Aroc
>
4-
-r
-i
'
=
!
—
j
-J-
„
=."
at
3)
0)
10)
or
—
I *-*
S3
- JJH
rS
-l~-
i
t
- =
1 —
; •
~~ps
—
-
_
A
^
JH
H:
.H
3=
•K
tr
'I
200
300
400 500
WAVELENGTH ( nm )
600
700
FIGURE 34A, B. SHAWSHEEN RIVER (9/26/73) EXTRACT OF WATER DOPED WITH 6 PPB AROCLOR 1254
* 1
&
to
LLI
P
d
0£
20
i
k
§
!=
K
i
^
T5
~_
==
=.
rrr
=
HH
^
~
-
,
—.
. -j
i
=
~ 3
s*-
=-=
-=-_
-r-
. __
-----
....
32
r^
--:
-.
^
==
-"
w
-
-
?--
^
-
... ..
-
,1'
_
r1
P*
-
-
=-
_
-
—
-
m
-
•-
—t
/ /
f*
/
f
s?
r=
^
"_
-
•-
1 '
,'
I/
''
/ /
'
f-
~fr
!
i
• i
^
p
Em
1
-
-
™=H
405 -i
i
l
I'J.
1
J
I/ !
-
—
t-.
4^
T~
f-
^
/-
I
1 1
— f-
-
,
0 300
_
.
Kinn
l
n
— i-*-
.1
i
i
4—
-\~
"j=
i
i
i
,
!
i Sample : Shawsheen River ( 9/26/73 ) —
^ cm
Slih :
Com
-125-
i
1.
l
i
:P
IP
>
1
•
: 380 ( 3/10 ) ; 405 ( 3/10 )
430 ( 3/10 ) ; 465 ( 3/10 )
11/22
ments
t too
-----
— .
' =
-
!
!
i
: Water spiked with Arc*
concentration pf 6 ppb.
-
Are
--4
|
j
400
!
j
, ,
- ^
. ... ..
-.
'
/
'•
/
/
=lwj
.i <
\ t,
'
/
f
/
rrj'
T-
• \
1 .
!
,/
f
~
-
C.
:Ior ,
j
A
)
r 1 '
.1
1
J
r
t
f
1
=S
~
r
M-
.1
1
— "—
•
H~
l_
!
'
300
i^
-
5.
_
\
_
1
~~i
1
i
i
Em
4;
v,
-
-4-
i
i
i
i
i
-~
.
~l
-
ri.
-
\
,
--
-
!*
\
\
1
— i
[
i
i
I
r=i
~
=
rrr-
l
1
|\
"
I
1
-H
rrr
--
—
,
^i
\
\
\
\
i
~
-j-
400
rrrr
-
—
•
-
—
.
•
,
\j )
r~~
r
-
-
-H
I
i
E
t
*r*
—
'
=
=s
^
—
~^s
^J
^
~
:==
~
50
200 WAVELENGTH ( nm )
71
-------
1
:I
-_
b£
=
r-i "
- M
-II
-l
--
-1
-i
.•~t
ix
br
BH
J~£.-
s-
p
- - -
,
— !
._,_
= •
--
-
„.
: —
i
1
t
t
~ i
~ i
t
-i
i
-77
-T-
4
_
1 '
-_,
L"
1
H-_-
t
s
>~r
-.
1
i
1
- - - -: 4 --.-!
-
: •
-
l
1
i
it*
n"
._-
^1
K*
-
^J
t1
_.
==
5
3^
Wl
-'
^^
—
-
=ii
t
I
*\
LJ!
:Z
iu-
./
^
^
.
. .
f
-
_
J
/
r..
/
3-,
+
„-
,i«
—
JAroc or
1 i
• -
1254
T
...
/ NrEx 290
k
)
^
1
-,
c
1
N
fl
' i-
^
~z.
S:
^.
r=
(
^
-
^_
N
N
•j
-i
_~
"\
Fx'?7
\
N
i
_L_.
\
5-
Hr
\
v
— ^
:
: ; Sample: Shawsheen River ( 9/26/73 )
X ex : 275 ( 10/5 ); 290 ( 10/5 )
325 ( 10/5 )
i Slits : 22/11
_!. I-.-
-(-
V
-\
t
! V
1
i
i
i
\
v,
"s,
s
i
l
-
Comments : Water spiked with Aroclor
254 to a concentration of 60 ppb.
S'
\
\
-
k
Is
1
1
-
-
n
-^
"hi
<„_
--
r=
—^
i
!:
,
r~
-
^
-
^
W^
„
J
rd
f?
^
„
— »
„_
'
.
.=•
H
r^r
-»
!
r.=
:~
f
SK
--=
•2
-
S
nr
.•r:
i
t
^
•-
™
„ ;
"-
~
—
e
'-
\
_=
*
-T—
;
E
"•
^;
^.
JT
-~
~
^S.
=r
=s
=^
=c
~*r
-
—
=r
-
A f
200
300
400 500
WAVELENGTH ( nm )
600
700
FIGURE 35A, B. SHAWSHEEN RIVER (9/26/73) EXTRACT OF WATER DOPED WITH 60 PPB AROCLOR 1254
UJ
h-
UJ
1
Sample : Shawsheen River ( 9/26/73 )
25, >• em : 380 ( 10/5 ) ; 405 ( 10/5 ) p=
m 430 ( 10/5 ); 465 ( 10/5 ) ]
Slits: 11/22
Comments : Water spiked with Aroclor!
H- )254 to a concentration of 60 ppb. ])£
400 300
200 WAVELENGTH ( nm )
400
500
72
-------
Sample : Showsheen River ( 9/26/73 )
X ex : 275(3/5) , 290 ( 3/5 )
325 ( 3/5 ) , 344 ( 3/5 )
Slits:
Comments : Water spiked with p,p*-DDT
to a concentration of 60 ppb.
400 500
WAVELENGTH ( nm )
FIGURE 36A, B. SHAWSHEEN RIVER (9/26/73) EXTRACT OF WATER DOPED WITH 60 PPB P, P'-DDT
ample : Shawsheen River ( 9/26/73 )
em : 380 ( 3/5 ) ; 405 ( 3/5 )
) ; 465 ( 3/5 )
Comment!: Water spiked with p,p'-DDT_|
to a concentration of 60 ppb.
400 300
200 WAVELENGTH ( nm )
73
-------
Sample : Oiiiodecxl phthalate
Concentration : 1 ppm in MCH
A ex : 275 ( 30/10 )
\ em: 440 ( 30/5 )
Slits: 33/11; 11/33
400500
WAVELENGTH ( nm )
FIGURE 37. DIISODECYL PHTHALATE, 1 PPM IN MCH, 77°K
jg Sample : Dibutyl phthalate
Concentration : 1 ppm In MCH
>- ex : 275 ( 10/5 )
A. em : 440 ( 3/10 )
Slih: 33/11; 11/33
WAVELENGTH ( nm )
FIGURE 38. DIBUTYL PHTHALATE, 1 PPM IN MCH, 77°K
74
-------
SECTION IX
REFERENCES
Ahling, B. and Jensen, S. (1970). Anal. Chem. 42; 1483.
Babers, F. J. (1955). J. Am. Chem. Soc. 77: 4666.
*
Blumer, M. (1970) . "Dissolved Organic Compounds in Sea
Water." Symposium on Organic Matter in Natural
Waters (D. W. Hood, ed. ) Ins tit. of Marine Science ,
U. of Alaska, College, Alaska, 153-167.
Breidenbach, A. W. , Lichtenberg, J. J. , Henke, C. F.,
Smith, D. J., Eichelberger,- J. W. , Jr., and Stierli, H.
(1966). "The Identification and Measurement of
Chlorinated Hydrocarbon Pesticides in Surface Waters."
U.S. Dept. of the Interior, Federal Water Pollution
Control Administration, Washington, D.C., 70 pages.
Brownrigg, J. T. , Eastwood, D. and Hornig, A. W. (1972).
"Identification of Polychlorinated Biphenyls in the
Presence of DDT-Type Compounds," U.S. Government
Printing Office, Publication EPA-R2-72-004, Washington,
D.C. , 62 pages.
Corcoran, E. F. (1973). Environ. Health Perspectives;
(3) 13.
Crosswhite, H. M. , ed. (1972). American Institute of
Physics Handbook, Third Edition, McGraw Hill Book
Company, New York, HI 92-96.
Dreeskamp, H. , Hutzinger, O. and Zander M. (1972). Zeit.
fur Naturforsch. 27A: 756.
Dubinskii, I. B. (1959). Bull. Acad. Sci. USSR, Phys.
Serv. (Eng.) 2jJ: HI.
Ellis, D. W. (1972). "The Analysis of Aromatic Compounds
in Water Using Fluorescence and Phosphorescence,"
Project No. A-009-NH. Water Resource Research Center ,-
U. of N. Hamp., Durham, N. Hamp. (Project Completion
Report), 73 pages.
Gesser, H.D., Chow, A., Davis, F. C., Uthe, J. F. and
Reinke, J. (1971). Anal. Letters 4: 883.
75
-------
Goerlitz, P. F. and Brown, E. (1972). "Methods for Analysis
of Organic Substances in Water." Techniques of Water-
Resources Investigations of the United States Geological
Survey, Book 5, Chapter A3, U.S. Government Printing
Office, Washington, B.C., 40 pages.
Golden, C. and Sawicki, E. (1973). "Ultrasonic Extraction
of Total Aromatic Hydrocarbons from Airborne Particles
.at Room Temperature." J. Environ. Chem; in Press.
Graham, P. R. (1973). Environ. Health Perspectives; (3) 3.
Harvey, G. R. (1972). "Adsorption of Chlorinated Hydro-
carbons from Seawater by a Crosslinked Polymer."
National Technical Information Service Publication
PB-213-954, Springfield, Va., 30 pages.
Harvey, G. R., Steinhauer, W. G. and Teal, J. M. (1972).
Science 180; 643.
Kites, R. A. (1973). Environ. Health Perspectives; (3) 17.
Kites, R. A. and Biemann, K. (1972). Science 178; 158.
Hornig, A. W. and Eastwood, D. (1973). "A Compendium of
Marine Luminescence Signatures." NASA Contract
No. NAS2-6408.
Johnson, L. G. (1971). Bull. Environ. Contam. and Toxicol.
_5: 542.
Lichtenberg, J. J., Chairman (1971). Methods for Organic
Pesticides in Water and Wastewater, U. S. Environmental
Protection Agency. National Environmental Research
Center, Cincinnati, Ohio.
McClure, D. S. (1949). J. Chem. Phys. 3/7: 905.
Nisbet, I. C. T. and Sarofim, A. F. (1972). Environ.
Health Perspectives; (1) 21.
Parker, C. A. (1968). Photoluminescence of Solutions,
Elsevier, Amsterdam, 246-261.
Potts, W. J., Jr. (1952). J. Chem Phys. 20; 809.
Schnitzer, M. and Khan, S. V. (1972). Humic Substances in
the Environment, Marcel Dekker, New York, 65.
76
-------
Shapiro, J. (1957). J. Limnol. and Oceanog. 11: 161.
Shpolskii, E. V. (1960). Sov. Phys. Usp. _3_: 372'
Shpolskii, E. V. (1962). Sov. Phys. Usp. _5: 522.
Shpolskii, E. V. (1963). Sov. Phys. Usp. j>: 411.
Uthe, J. F., Reinke, J. and Gesser, H. (1972). Environ.
Letters _3_: 117-
Wallnofer, P. R., Koniger, M. and Hutzinger, O. (1973).
Analabs Inc. Research Notes 13; (3) 14-16.
Wershaw, R. L., Burcar, P. J. and Goldberg, M.C. (1969).
Environ. Sci. and Techn. _3_: 271.
Westrum, E. P., Jr. and McCullough, J. P. (1963). "Thermo-
dynamics of Crystals" in Physics and Chemistry of the
Organic Solid State (D. Fox, M. M. Labes and A. Weissberger,
eds.) Vol. T~r Inter science, New York.
Winefordner, J. D., McCarthy, N. J. and St. John, P. A.
(1967). "Phosphorimetry as an Analytical Approach"
in Methods of Biochemical Analysis (D. Clock, ed.)
Vol~15,Interscience, New York.
Winefordner, J. D. and St. John, P. A. (1963). Anal.
Ghent. _3_5: 2211.
Zweidinger, R. and Winefordner, J. D. (1970). Anal.
Chem. 42: 639.
77
-------
SECTION X
APPENDIX
METHOD FOR ESTIMATION OF PCB/DDT IN WATER
The method described herein can be summarized briefly
as follows:
Grab samples of water are extracted with dichloromethane
and the combined extracts concentrated nearly to dryness
using a rotary evaporator (or the equivalent). The residue
is then diluted with methylcyclohexane and the luminescence
analyzed at 77°K. Identification and quantitation is
based upon comparison of the observed spectra with those of
Aroclor/pesticide standards. Since the analysis of water
samples doped with these compounds may aid in the interpre-
tation of the results (or in testing possible method
modifications), these procedures are also described.
1. Equipment and Chemicals
The fluorescence instrumentation and accessories required
are:
SF-100 Fluorispec.
X-Y recorder and graph paper.
Optical dewar and dewar positioner.
Suprasil quartz sample tubes (one or two dozen)
with caps (caps can be fashioned as described
in Section V of the main text).
Liquid nitrogen. The optical dewar capacity is
about 40 cc and will hold nitrogen for one to
two hours; this is usually adequate to complete
four to eight analyses. A glass storage dewar
is useful for refilling the optical dewar.
It is strongly recommended that the user become familiar
with the basic instrument as described in the Fluorispec
User's Manual. In particular, the user should practice
obtaining spectra of common fluorescent compounds such
as anthracene or quinine sulfate at room temperature
before attempting the low temperature analyses described
here.
78
-------
Equipment and chemicals used for the collection and
treatment of water samples are:
Collection bottles. Wide mouth glass bottles of
2 to 4 liter capacity are recommended. These
should have screw-type lids with Teflon liners.
Magnetic stirrer. Several Teflon-covered stirring
bars are also required.
Rotary evaporator (with flasks). If desired, a
Kuderna-Danish concentrator can be substituted.
In either case, the device should have a capacity
of at least 100 ml.
Glass pipette, 50 ml, and pipetting bulb. This or
an equivalent device is used to remove the solvent
extract from the water sample.
Erlenmeyer Flask, 125 ml. This or an equivalent
glass container is used to stir the extract with
sodium sulfate. If the container volume is
calibrated, it can also be used to deliver the
extraction solvent to the water sample.
Screw-cap glass vials, 4 dram. These are used
for final sample storage, and should be equipped
with Teflon cap liners.
Sodium sulfate, anhydrous, reagent grade. This
is used to remove traces of water from the extract.
Solvents. These are discussed in the next section.
Silica gel, activity grade I, 200 mesh. This is
used for solvent purification.
2. Solvents and Purification
Methylcyclohexane (MCH)—Spectroquality grade or an
equivalent purity should be used. This solvent generally
contains trace amounts of toluene as an impurity, which
can be removed by column chromatography using silica
gel (see Section V). Using the slit combinations
suggested in Section 3 of this Appendix, the purified
solvent should be free of background luminescence at
77°K in the spectral region 300-500 nm when excited
between 220 and 290 nm.
79
-------
Dichloromethane (PCM)—Spectrograde material or an
equivalent purity should be used although, since the
boiling point is quite low (42°C), it may be possible
to adequately purify poorer grades of material by dis-
tillation. To check the purity, 100 ml of this solvent
should be evaporated nearly to dryness and then diluted
with 5 ml of purified MCH. This solution is then
analyzed for contaminants in the manner described for
MCH. If necessary, this solvent can be purified by
distillation.
Ethanol (95%, 5% water)—This is a useful solvent for
doping PCB's and DDT-type compounds into water. This
solvent, like MCH, should be checked for luminescence
at 77°K. If necessary, the material should be distilled.
3. Intensity Standard and Instrument Optimization
Gain Linearity—The linearity of the SF-100 gain should
be checked using an appropriate standard. In this
case, a water Raman band is satisfactory (see User's
Manual, Section II and Figure 5). With the recorder
gain at the appropriate value (e.g., 0.1 volt/cm),
the Raman peak is recorded at a fine gain of 10, the
coarse gain being varied from 1 to 1000. The fine gain
is checked by keeping the coarse gain fixed and record-
ing the Raman band at the ten positions of the fine
gain. If the instrument used in this study is typical,
the coarse gain should be linear except for setting
number 1, which gives a gain about 20% above the
expected linear value. The fine-gain is very non-linear,
and in fact appears to be nearly quadratic. It is
often desirable to use fine gain settings of either
5 or 10, where the gains differ by about a factor of
two.
Low Temperature Standard—Spectra of a standard should
be run daily to monitor instrument sensitivity. Since
the instrument is a single beam type, long-term deteri-
oration of the lamp and optics will result in a loss
of intensity. Use of a standard thus allows the
correction of instrument response to compensate for
possibility sensitivity variations. The standard
tentatively recommended is a 10 ppm solution of p,p'-DDT
in MCH. This solution has been checked for photodecom-
position at 77°K, exciting at the two principal excitation
bands near 240 and 275 nm, with maximum slit widths
80
-------
(33 combination). A nearly exponential decay was
found at both wavelengths, and the decay rates appeared
nearly identical. The emission intensity fell by 10%
of its initial value after five hours, which is adequate
time for about 50 complete excitation/emission scans
(or 50 days if the procedure were done daily). However,
since the volume irradiated is only about one-tenth of
the total (0.5 ml) sample volume, the same solution
could probably be used for even longer periods without
significant decomposition.
About 0.5 ml of this solution is transferred to a
quartz sample tube. If possible, the solution should
be degassed in a clean vacuum system using several
pump-freeze-thaw cycles and the tube sealed under
vaccum. If this is not possible, the tube can be
capped in the usual way. Since MCH slowly diffuses
through the rubber cap, it is necessary to replace
the solution in the tube every two or three weeks.
After the dewar positioner and optical dewar are
installed and the emission wavelengths set on the graph
paper (Section III of the User's Manual), the dewar
is filled with liquid nitrogen, and the standard solu-
tion is lowered into the dewar. Lowering the sample
completely into the dewar should require about 30
seconds. The frozen sample should be perfectly trans-
parent with no cracks. (The sample can be observed
shile in the dewar by opening the small door on the
sample compartment.) If cracks do appear, the sample
should be removed quickly and the solvent warmed with
the fingers from the top down. (If the sample warms
from the bottom, confinement of the warming solution
by frozen material above it can result in breakage of
the quartz tube.) The tube is then lowered again until
a clear glass is obtained. If several attempts fail
to produce a glass, the tube or solvent may contain
water. If this happens, the tube should be emptied
and flushed throughly with dry nitrogen gas before
refilling (or a dry tube substituted). It may also
help to keep light from striking the sample during the
freezing process. This can be done by pulling up on
the lever on the SF-100 labeled Source Filter (which
inserts a pyrex glass plate into the optical path)
and setting the excitation monochromator to 270 nm
or lower.
81
-------
Optimizing Dewar Position and Lamp Intensity—The slit
combination 33/11 is used for the emission spectrum.
Discussions of this notation are given in Section V
and also in the User's Manual. The recorder vertical
gain is set at 0.1 volt/inch and the time constant is
set at 0.3 seconds. The excitation monochromator is
set at 275 nm and the emission monochromator at 400 nra.
The emission wavelength is then moved manually to shorter
wavelength until the recorder pen begins to rise due to
scattered exciting light (about 290-300 nm). With the
room lights off, the knob on the dewar positioner
which affects forward and backward movement of the
dewar is adjusted until the scattered light intensity
is minimized (there will probably be two such positions;
the one giving least scatter is selected).
After minimizing scatter, the emission monochromator
is then returned to the maximum of the DDT phosphorescence
near 400 nm. The instrument gains are adjusted such
that the intensity is brought to about 2/3 of the height
of the paper. The Lateral Adjustment and Focus Controls
for the xenon arc (see User's Manual, Figure 3) are then
turned until the emission intensity is maximized; the
control nearer the operator should be more influential
in this respect. If these adjustments cause the pen
to move off the chart paper, the instrument gains should
be reduced accordingly. Proper execution of this pro-
cedure will result in minimum scattered light and maximum
source intensity.
Running the Standard—The emission spectrum should then
be scanned in the 300-500 nm region using the "slow"
scan rate (1 nm per second). The pen will go off scale
near 500 nm since the emission monochromator is now
sensing the second order of the 275 nm exciting light.
An appropriate filter (such as Corning 0-53 or equivalent)
can be used on the emission side of the monochromator
to eliminate this if desired; otherwise, the scan should
be stopped when this point is reached. If the filter
is used, the instrument fine gain can be increased
slightly to compensate for the loss of intensity. The
excitation spectrum is obtained next, monitoring at
about 400 nm using an 11/33 slit combination. The
excitation spectrum can be placed on the same page as
the emission spectrum, but the wavelength calibration
should be checked out and reset if necessary. The
coarse gain setting will now have to be reduced to
82
-------
maintain equivalent peak height. This spectrum is then
scanned from 220 to 300 run. Analyzing slits used for
these spectra are the narrowest available and give a
spectral resolution of about 2 nm.
A record should be kept of relevant instrumental para-
meters, such as the wavelengths, slit combinations,
recorder and instrument gains, and time constant.
4. Standard Solutions in MCH
The recommended standard solutions should consist of
p,p'-DDT and Aroclors 1016, 1248, and 1254; however,
other Aroclors, pesticides, etc., could be added.
Stock solutions at nominal concentrations of 100 ppm
are prepared by dissolving 1 mg of the appropriate
standard in 10 ml MCH. Successive ten-fold dilutions
are prepared by combining 1 ml of a given concentra-
tion with 9 ml of solvent. The basic standard solutions
should range from 100 to 0.01 ppm in MCH.
The glass containers used for storage of these solutions
should have screw-type caps; a convenient size is 4
dram, which has a capacity of about 15 ml. The caps
should be provided with Teflon liners to prevent possible
contamination by cap materials.
Luminescence Analysis—It is suggested that the quartz
sample tubes be checked for possible contaminants by
filling them with MCH and exciting with the wavelengths
recommended below. The tubes should then be filled
to a depth of 7 to 8 cm (about 0.5 ml) with the
appropriate standard solution and capped. It is perhaps
best to begin with the highest concentrations in order
to form a clear understanding of the nature of the
luminescence signatures. The quartz tube is lowered
into liquid nitrogen contained in the optical dewar.
The rate of freezing should be similar to that described
in the previous section and the frozen sample should be
free of cracks.
With a piece of graph paper properly positioned on the
recorder, the recorder vertical gain is set at 0.1
volts/inch, and the Fluorispec time constant is set at
0.3 seconds. The emission spectra should be run first,
using the slit combination 33/11. The instrument gains
83
-------
should be adjusted to bring the peak of the emission
to nearly the full heicrht of the chart paper. The arc
adjustment controls should be checked as described in
the previous section to insure that the intensity is
maximized. The emission monochromator is moved manually
to shorter wavelengths until the pen begins to rise in
response to the excited light. The emission spectrum
is then scanned ("slow" speed)to longer wavelengths
until the exciting light is sensed in second order.
The scan should be stopped there unless the exciting
light is filtered out. Similar instrument settings are
used to record the excitation spectra, except the slit
combination 11/33 is used; the instrument gain will
have to be reduced if peak heights approximating those
of the emission spectra are desired.
The wavelengths suggested for emission and excitation
of DDT-type compounds and Aroclors are given below:
Emission Excitation
DDT 275 380
Aroclor 275, 290 400, 440, 470
It should be noted that excitation spectra monitored at
440 and 470 nm will show peaks near 220 and 235 nm,
since the emission monochromator is sensing these
excitation wavelengths in second order. These peaks
can be eliminated by use of a filter as discussed in
Section 3 of this Appendix.
Spectra of DDT and several Aroclors appearing in the
main text may be used for comparison. Small wave-
length differences (± 2 nm) may be found from those
given here, and relative intensities may show some
variation due to differences in lamp brightness and
spectral distribution. Since the composition of Aroclors
are somewhat variable, spectra obtained of other samples
may show intrinsic differences.
Reduction of Data—Intensities should show a nearly
linear dependence on concentration for concentrations
below 10 ppm. Ideally, the integrated band intensities
(areas) should be obtained; this, however, can be a
very time-consuming process. Instead, peak maxima are
used as an approximation to areas. Either emission or
84
-------
excitation peaks could be measured, but excitation
peaks appear to be ultimately more useful in the analysis
of environmental samples where background emission
occurs. In general, the strongest (uncorrected)
excitation band is selected for measurement. For DDT,
this is the sharp origin band near 278 nm. The peak
heights are measured relative to the underlying back-
ground, which should be due primarily to scattered
light at low concentrations. The nature of this back-
ground can be determined by comparison with the pure
solvent (MCH) spectrum at the same wavelengths and gain
settings. The background should rise smoothly and
continuously as the scatter peak is approached.
Band heights are conveniently measured in millimeters
using the chart paper grid if desired. These are
converted to approximate intensities by dividing the
peak height by the product of the Fluorispec coarse
and fine gain reading. This of course presumes that
the gain settings are strictly linear over the range
selected; if this is not the case, corrected gain values
as discussed previously must be used. If the recorder
gain has been varied, this gain must also be included.
Finally, changes in the arc intensity as reflected in
the intensity of the standard (Section 3 of this
Appendix) should also be accounted for in the intensity
determination. This is most easily accomplished by
using the ratio of sample peak height to the peak
height of a standard as a measure of normalized intensity:
Hx Gs
I = _ • _
x Hs Gx
where
I = normalized sample intensity.
X,
H = sample peak height measured at a gain of G .
«t X
H = standard peak height measured at a gain of G
s s
G ,G = linearized instrument gain settings for
measuring sample and standards respectively.
In general, G is a product of the instru-
ment coarse and fine gain setting, and
would also include the recorder gain
if this is varied.
85
-------
The normalized sample intensity will then depend only
on the sample concentration, independent of overall
instrument sensitivity.
5. Doped Water Samples
Although this section is optional, it is recommended
for several reasons. First, it enables the analyst
to become familiar with extraction and concentration
techniques prior to studies of environmental samples.
Alternative extraction solvents or other new techniques
may be evaluated with respect to recoveries of the
compounds of interest. Finally, potential sources of
contamination are more easily uncovered.
Water samples used initially should be distilled or of
comparable purity. Purity can be checked by first
extracting and analyzing an undoped sample. Doped
samples of natural water can be analyzed in a similar
fashion. The recommended procedure discussed below
assumes a one-liter volume of water.
Dopant Solutions—Solutions of Aroclor and DDT compounds
are prepared in purified ethanol at concentrations of
1, 10, and 100 ppm. Microliter pipettes are used to
deliver the desired volumes into the water sample. The
doped water sample should be stirred thoroughly using
a Teflon or glass covered stirring bar activated by
a magnetic stirrer.
Extraction and Concentration of Extracts—Two successive
75 ml aliquots of purified DCM are added to the bottle
and the mixture stirred for at least an hour on the
magnetic stirrer. Stirring speed should be fast enough
to break the DCM layer into a fine emulsion dispersed
throughout the water.
The combined extracts are dried with 10 to 20 grams of
anhydrous Na2SO4- This can be done by stirring the
extract with the drying agent in a small flask or
beaker for about fifteen minutes. If the entire mass
of desiccant becomes caked (hydrated), additional ^2864
should be added. The dried DCM is then transferred to
a round bottom flask and brought almost to dryness using
a rotary evaporator or other evaporative concentrator.
The reduced pressure required for the rotary evaporator
may be provided by a water aspirator. At 77°K, residual
amounts of DCM in MCH cause the normally clear MCH
86
-------
glass to become slightly hazy, which contributes to
scatter. For this reason the final DCM volume should
be kept on the order of 0.1 ml.
The flask containing the DCM residue is then thoroughly
rinsed with 5 ml MCH and the MCH solution stored in a
screw-cap vial. This solution is ready for luminescence
measurements.
Analysis—Low temperature luminescence analysis is
performed in the manner described for the standard
solutions in MCH. Measured peak heights are converted
to approximate relative intensities as discussed in
the previous section and the concentration in MCH
determined from the appropriate analytical curves.
Concentrations in water are obtained by multiplying
the concentration in MCH by the ratio of the volume
of MCH to that of the water sample.
6. Environmental Water Samples
Known sample volumes of two to four liters should be
collected in glass bottles with Teflon-lined screw
cpas. These should have been previously rinsed several
times with purified DCM to remove residual impurities.
Approximately 2.5 ml of concentrated HC1 per liter are
added to the sample as soon as possible after collection.
The water sample is extracted with DCM and the extracts
dried and concentrated as described in Section 5 of
this Appendix. Five ml of purified MCH are added to
the combined extracts and the resulting solutions are
ready for luminescence analysis.
Luminescence Analysis—Luminescence analysis is performed
using the instrumental parameters discussed in Section 4.
Recommended wavelengths (in nm) are:
Emission: 275, 290
Excitation: 380, 400, 440, 470
The emission wavelengths are sufficient to excite both
Aroclors and DDT derivatives. The excitation wavelengths
380 and 400 will produce little Aroclor excitation but
will result in nearly maximum DDT intensity. Excita-
tion spectra monitored at 440 and 470 nm will show
reduced DDT intensity but nearly maximum Aroclor intensity.
87
-------
For identification, spectral signatures should be
compared with those of the Aroclors and DDT compounds
in MCH. Better still, they should be compared with
signatures obtained from natural water samples doped
with these compounds, since natural water will contain
a high background emission in the Aroclor/DDT region.
It is likely that this background emission will obscure
the broad emission signatures of the Aroclors and DDT
when present at low ppb levels in environmental waters.
Since the excitation spectra of the Aroclors and DDT
are much sharper than the natural background absorption,
lower concentrations are observable in the excitation
spectra. For this reason, quantitation is based upon
excitation peak heights measured relative to the back-
ground absorption, which should be nearly continuous
in the region pf interest.
Reduction of Data—Peak heights are converted to rela-
tive intensities and thus to concentrations as described
in Section 4. The value obtained would be the actual
value if 100% of the material were recovered. The
actual recoveries will probably be less than this. If
recoveries have been estimated from doped water samples,
the concentration obtained here can be divided by the
fraction recovered.
7. Accuracy, Sensitivity, and Possible Interferences
The accuracy and precision of the determination are
influenced by both the luminescence measurement and by
the extraction procedure. These sources of error are
discussed in detail in Section VII of the main text. For
the analysis of natural water samples for which recoveries
have been determined, the maximum percentage error
should be in the range of 25-40%. The estimated precision
of the determination is approximately 25%.
The detection limits for DDT and 'Aroclor 1254 in MCH
at 77°K should be approximately 0.03 and 0.002 ppm
using the recommended instrument parameters. Similar
sensitivities are expected for DDD and the more highly
chlorinated Aroclors. Since these values were determined
using optimum resolution (2 nm), lower limits might be
achieved using wider analyzing slits, but this would
result in loss of resolution and therefore specificity.
88
-------
Detection limits for DDT and Aroclor 1254 in one-liter
samples of pure water should be approximately 0.5 and
0.03 ppb respectively using the extraction and spectral
method described. For natural waters rich in dissolved
fluorescent organic matter and suspended particulates,
detection sensitivities may be from 10 to 100 times
higher depending on the extent of adsorption on par-
ticulates and the fluorescence background level.
Phthalic acid esters (phthalates) may be encountered
both in water samples and as laboratory contaminants.
Since the phosphorescence excitation/emission spectra
may mimic certain Aroclor spectra (and thus interfere),
it is strongly recommended that spectra be obtained of
several of the more commonly used phthalates, such as
dibutyl phthalate and di-2ethylhexyl phthalate. Excita-
tion spectra of phthalates should show less variation
in structure than Aroclors when monitored at several
different emission wavelengths, and this feature should
permit some discrimination. However, at the present
time phthalates should be regarded as potential inter-
ferences for Aroclors.
89
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-670/4-74-004
2.
3. RECIPIENT'S ACCESSIOP+NO.
. TITLE AND SUBTITLE
5. REPORT DATE
Estimation of Polychlorinated Biphenyls
in the Presence of DDT-type Compounds
June 1974; Issuing Date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.T. Brownrigg and A.W. Hornig
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG "\NIZATION NAME AND ADDRESS
Baird-Atomic, Inc.
125 Middlesex Turnpike
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO. (16202GIY
1BA027;ROAP 09ABZ;TASK013
11. CONTRACT/GRANT NO.
68-01-0082
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Renort. February 1974
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
6. ABSTRACT
Earlier studies suggested that the low temperature luminescence properties of PCB's
and DDT compounds could be used to identify these compounds singly or in mixtures.
The present investigation was undertaken to develop a relatively simple, rapid
method for estimating these compounds in water. The errphasis in this procedure
has been on the inherent sensitivity and specificity of luminescence, avoiding
chemical separation where possible.
The present procedure involves collection of grab samples followed by extraction,
drying, concentration, and redilution in a second solvent suitable for luminescence
measurement at 77°K. Studies include the determination of recoveries and detection
sensitivities for some of the compounds of interest and also analyses of several
environmental waters.
Detection limits for p,p'-DDT and Aroclor 1254 doped in 1-liter samples of pure water
were found to be approximately 0.5 and 0.03 ppb respectively. Sensitivities were
reduced by an order of magnitude or more in natural waters having high levels of
dissolved organic material and particulates. This is due to a combination of poorer
recoveries and increased fluorescence background. Both of these remain as probler
areas deserving further study. Phthalic acid esters have spectral features reserrib-
ling certain Aroclors and may constitute
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*DDT, *Fluorometers, Spectroscopy,
*Water analysis, Aromatic compounds
Chemical analysis, Water pollution,
Chlorine aromatic compounds,
Polyphenyl compounds, Luminescence
*Aroclor, *Analyt-
ical techniques,
*Pollutant analysis,
Chlorinated hydro-
carbon pesticides,
*Polychlorinated
biphenyls, *Low tem-
perature luminescence
7C
13B
18. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
100
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
90
•ft 111. GOVERNMENT PRINTING OFFICE, 1974- 757-S84/53Z4
------- |