EPA-600/1-77-052
NOVEMBER 1977
Environmental Health Effects Research Series
METHOD DEVELOPMENT AND MONITORING OF
POLYNUCLEAR AROMATIC HYDROCARBONS IN
SELECTED U.S. WATERS
Health Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. “Special” Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/1-77-052
November 1977
METHOD DEVELOPMENT AND MONITORING OF POLYNUCLEAR AROMATIC
HYDROCARBONS IN SELECTED U.S. WATERS
by
J. Saxena, D.K. Basu, and J. Kozuchowski
Syracuse Research Corporation
Syracuse, New York 13210
Grant No. R803977
Project Officer
Herbert R. Pahren
Field Studies Division
Health Effects Research Laboratory
Cincinnati, Ohio 45268
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Health Effects Research Laboratory,
US. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation for use.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of in-
creasing public and government concern about the dangers of pollution to
the health and welfare of the American people. Noxious air, foul water, and
spoiled land are tragic testimony to the deterioration of our natural envi-
ronment. The complexity of that environment and the interplay between its
components require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions. The primary mission of the Health Effects Research
Laboratory in Cincinnati (HERL) is to provide a sound health effects data
base in support of the regulatory activities of the EPA. To this end,
HERL conducts a research program to identify, characterize, and quantitate
harmful effects of pollutants that may result from exposure to chemical,
physical, or biological agents found in the environment. In addition to
valuable health information generated by these activities, new research
techniques and methods are being developed that contribute to a better
understanding of human biochemical and physiological functions, and how
these functions are altered by low-level insults.
This report describes the development and testing of a new and more
sensitive analytical technique for monitoring polyriuclear aromatic hydro-
carbons in waters. With the ability to measure very low levels, we will
have a better understanding of the degr e o insult from these potentially
harmful materials.
R. . G er
Director
Health Effects Research Laboratory
111
-------
ABSTRACT
Flexible polyurethane foam plugs have been successfully used for con-
centration of trace quantities of six representatives of the polyTluclear
family -- fluoranthene, benzo(k)fluorantherie, benzo(j)fluoranthene, benzo-
(a)pyrene, benzo(ghi)perylene, and indeno(l,2,3-cd)pyrene -- from drinking
waters and their raw water sources. After studying a variety of parameters
affecting retention, the conditions best suited for field sampling were
defined. Recoveries of PAH were quantitative when water was heated to
62 + 2°C prior to passage through foam columns. A portable sampler allow-
ing control of optimum conditions for collection of PAH at water distri-
bution/treatment sites was assembled. The collection of PAH on foam plugs
was followed by their elution with organic solvent, purfication by parti-
tioning with solvents, and column chromatography on Florisil. The puri-
fied concentrate was analyzed by two dimensional thin layer chromatography
on cellulose acetate-alumina plates and PAl -I quantitated directly on the
plates by fluorometry. Gas liquid chromatography - FID was also employed
in the studies but failed to detect PM! in most drinking water samples and
in some instances did not provide adequate resolution.
Using the method developed, analyses were performed for the 6 PM! in
finished and raw waters at 10 selected water supplies in the eastern United
States. PAl-I were detected in the ppt range in all water supplies sampled.
In many cities all six representative polynuclears, as well as several un-
known compounds, were detected. Although the concentration (sum of the
6 PM!) in drinking waters was small (0.9 to 15 ppt), the values found in
raw water were as high as 600 ppt. whether the PM! are actually removed or
transformed to some other compounds during treatment in unclear. The health
significance to man of the presence of the above levels of PM! in drinking
waters is not understood.
This report was submitted in fulfillment of Grant R 803977 by Syracuse
Research Corporation under the sponsorship of the U.S. Environmental Pro-
tection Agency.
iv
-------
CONTENTS
Foreword .
Abstract iv
Figures V1
Tables viii
Abbreviations and Symbols x
Acknowledgment xi
I. Introduction 1
II. General Conclusions 3
III. Recommendations 4
IV. Literature Review 6
V. Preconcentratlon Method 7
Basis for developing preconcentration method 7
Evaluation of foam plugs for extraction and recovery
of benzo(a)pyrene . . 8
VI. Method of PA l-I Analysis 23
Basis for developing analytical method 23
Selection of PAR for study 24
Experimental methods and results 26
VII. Evaluation of Foam Plugs for Collection Efficiencies
of Six PAIl 46
Collection efficiency of six PAl-I evaluated by
gas—liquid chromatography 46
Collection efficiencies of six PAl- I evaluated by
thin—layer chromatography—f luorometry . 48
VIII. Field Monitoring 52
Fabrication of sampling unit for field monitoring . . . 52
Determination of PAl-I in selected water supplies . . . . 54
IX. Laboratory Evaluation of Activated Carbon for Addition!
Removal of PAll 68
References 69
Appendix — Standard Operating Procedure 74
V
-------
FIGURES
Number Page
1 Set—up for flow system extraction . 11
2 Sorption of BaP by various foam plugs under static conditions . 12
3 Retention of BaP from spiked tap water at various flow rates . 15
4 Influence of water pH on BaP retention 16
5 Effect of water temperature on recovery of BaP on foam plugs . 16
6 Gas chromatogram of the standard PAR mixture on Dexsil 300
packed column 30
7 Calibration curves for two representative PAR determined
by GLC—PID method 30
8 Thin—layer chromatograni of standard PAR mixture 31
9 Fluorescence emission and excitation of model PAR compounds
obtained directly on the plate 34
10 Calibration curves for six reference compounds determined by
fluorimetric method directly on the plate 37
11 Purification efficiency of the clean—up procedure 40
12 Recovery of PAR in the clean—up procedure 42
13 Flow chart of the clean—up method 43
14 Thin—layer chroinatogram of 12 foam blanks 44
15 Retention of 6 PAR on polyurethane foam determined by
gas chromatography 47
16 Portable unit assembled for concentrating PAR from
water in the field 53
17 Flow chart of the method of PAR analysis in water 57
18 Fluorescence emission and excitation spectra of model PAR
and those Identified in Huntington, W.Va. water samples . . . 64
vi
-------
Number Page
19 Fluorescence emission and excitation spectra of
selected unknown compounds 65
vii
-------
TABLES
Number Page
1 Characteristics of various types of foams used 8
2 Recovery of BaP from spiked water by continuous flow system . 13
3 Effect of diameter of column used for holding the foam
plug of BaP retention from tap water 15
4 Benzo(a)pyrene retention from tap water with foam plugs
coated with chromatographic phases 18
5 Recoveries of BaP from spiked tap water at various
concentrations 19
6 Mass balance of 14 C—activity added to water 20
7 Effect of sample volume on the recovery of BaP with
a single foam plug 21
8 Stability of BaP on foam plugs stored at room temperature
andat4°C 22
9 PAM compounds studied 25
10 RB values of PAR in two solvent systems 32
1]. Fluorescence colors of PAR 32
12 Excitation and emission wavelengths used for running spectra . 33
13 Recovery of the overall clean—up method determined by
gas liquid chromatography 41
14 Detection limit of six PAR by thin layer and gas chromatography 45
15 Foam retention efficiencies of six PAR from treated water . . . 48
16 Amount of six PAM unaccounted for as a result of mixing
with water in a glass bottle 49
17 Foam retention efficiencies of six PAR from treated water . . . 50
vi i i .
-------
Number Page
18 Foam retention efficiencies of six PAR from raw water 50
19 Details of the water supply systems used for sampling 56
20 Results of analyses of field samples 59
21 Frequency of occurrence of various PAR in drinking waters . . . 62
22 Unidentified luminescent spots from each sample 62
23 Efficiency of removal/transformation of PAR in water treatment 67
ix
-------
ABBREVIATIONS ANI) SYMBOLS
A Wavelength
PAIl Polynuclear aromatic hydrocarbon
BaP Benzo(a)pyrene
BjF Benzo(j )fluoranthene
BkF Benzo (k) fluoranthene
IP Indeno(l, 2 ,3—cd)pyrene
BPR Benzo(ghi)perylene
PCB Polychlorinated biphenyl
W Watt
ppb Parts per billion
ppm Parts per million
ppt Parts per trillion
TLC Thin layer chromatography
GLC Gas liquid chromatography
ng Nanogram
pg Microgram
FID Flame ionization detector
WHO World Health Organization
ND Not detected
x
-------
ACKNOWLEDGEMENT
Syracuse Research Corporation wishes to acknowledge the invaluable
assistance and cooperation of the personnel of all the water supplies
monitored in the project. The generous guidance and assistance of the
EPA Project Officer, Mr. Herbert R. Pahren, is also appreciated.
xi
-------
SECTION I
INTRODUCTION
Chemicals in the work place, in the environment, and in diet may be the
single most important cause of human cancer (34). It has been estimated that
at least 60% or perhaps as much as 90% of the total cancer cases in the U.S.
this year will have been caused by environmental factors, mostly chemicals.
Polynuclear aromatic hydrocarbons (PAH) are of particular concern in
the environment because of their demonstrated carcinogenic activity (23, 49),
wide distribution and persistence in the environment (42). The potential
hazards from the occurrence of PAH in man’s water supplies have been noted
by the World Health Organization’s Committee on the Prevention of Cancer (55),
which recommends evaluation of the treated surface waters for PAH. The
organization has proposed that collective concentration of six PAH which
serve as representatives of the whole group should not exceed 0.2 .ig/i.
The natural background of PAM in the environment is provided either by
an endogenic synthesis of these compounds by microorganisms, phytoplankton,
algae and highly developed plants or by natural pyrolytic processes, namely,
forest fires and volcanic activities (50). The major source of PAH in the
environment, however, arise from technological sources, such as heat and
power generation, refuse burning, miscellaneous industrial processes and
emission from vehicular transportation media. These carcinogenic substances
may enter natural water and thereby public water supplies in a variety of
ways including the release of industrial effluents, direct fall out from the
atmospheric particulate matters, road run—of fs, discharge of urban and
domestic sewage and run—off or leaching from soils.
The studies on the incidence of PAM in the water environment have been
predominantly carried out in Europe and are pertinent to European waters
only. A number of PAM have been detected in European natural and treated
waters at levels which are alarming. In one instance, the concentration of
benzo(a)pyrene alone was 0.01 mg/i in drinking water and as high as 6 mg/i
in strongly polluted surface water (52). Data regarding their levels in
U.S. water is virtually nonexistant. In view of the fact that polynuclear
compounds have been detected in the air of U.S. urban and non—urban sites
(36) and in soil and marine sediments (4, 5), it is suspected that they may
also be present in U.S. waters. The evaluation of hazards to the public
from the presence of PAM in water and the implementation of the remedial
measures requires the knowledge of the levels and the nature of PAM in these
waters.
1
-------
The objectives of the present study were to determine the concentration
of PAH in a selected raw water source used for drinking purposes and their
degree of removal as a result of the treatment processes. Since no suitable
method was available for concentration of PAH from water, the first phase
of the project was devoted to developing a rapid and efficient method for the
preconcentration of trace quantities of PAH. The second phase consisted of
developing an analytical scheme for the quantitation of collected PAH, and
monitoring of PAR in drinking waters and their raw water sources.
2
-------
SECTION II
GENERAL CONCLUSIONS
Polyurethane foams under optimum conditions can be used to concentrate
trace quantities of polynuclear aromatic hydrocarbons from water. The
inability of the earlier investigators to recognize and use the sorbent
properties of polyurethane foam was due to the absence of a systematic study
dealing with various parameters affecting retention. The method developed
enables sampling of larger volumes of water, and is more efficient and
practical than cumbersome batch extraction methods normally used. The method
of concentration can be conveniently used at the sampling site and thus
eliminates the need for sample preservation and transport, and problems
arising from adsorption on the container. These advantages render this
method very desirable for PAH analysis in water.
The chemical clean up of the foam eluate utilizing solvent partitioning
and column chromatographic separation was found to be necessary prior to
subjecting to analysis. Final purification and resolution of the sample
concentrate on two dimensional thin layer chromatography was found preferable
over GLC columns. For identification and quantitation, fluoroinetric analysis
was more suitable than flame ionization detection. The former method was
more sensitive and selective and possessed a much larger sample capacity,
allowing a lower detection limit. With overall preconcentration (6O2 volume),
clean up and detection procedure, concentrations of PAH as low as 0.1 ppt can
be determined.
Judging from the limited information generated, PA l - I appear to be wide-
spread in drinking waters and their raw water sources, although concentra—
tions, particularly in drinking waters, are low. Current technology of
water treatment is able to substantially reduce the levels of PAH in treated
waters. Whether the reduction is due to removal or transformation to other
products could not be concluded from this study.
3
-------
SECTION III
RECOMMENDATIONS
It is recommended that further research in the following areas be under-
taken to fully understand the implications of the presence of polynuclear
aromatic hydrocarbons in drinking waters and their raw intake waters:
A. Our results of single samples from a few water supplies reveal
that PAR are widespread in raw and drinking waters although their
levels are generally low. More monitoring work using the sensi-
tive method developed must be undertaken to derive more definite
conclusions.
B. Work in Germany has shown that PAH are introduced in water during
distribution through paint, asphalt and other such material used
for coating the pipes. Thus, studies should be undertaken to
evaluate the additional contribution of PAR, if any, arising from
the supply network.
C. Our determinations of the levels of PAR in finished and raw waters
have shown that considerable reduction in the concentration of PAR
takes place in water treatment. Whether this decrease reflects a
true removal or transformation of PAIl to other and perhaps more
hazardous products must be ascertained.
D. The cumulative cancer risk to humans posed from the presence of
low levels of PAR in drinking waters can not be assessed by con-
ventional animal tests. Thus, epidemiological studies must be
undertaken. Alternatively, studies employing more sensitive
in vitro test methods must be carried out. One must keep in mind
that these compounds may possess a very high bioaccumulation
potential.
E. It is well known that the particulate PAR emission is maximum
during winter months because of home heating. This, coupled with
the fact that microbial breakdown of PAIl, if any, will be lower
during winter months, may result in increase In PAR concentration
during winter months. On the other hand, concentration during
summer months may rise due to increased solubility of PAR. Thus,
the fluctuations in the levels of PAR due to seasonal variations
must be studied.
4
-------
F. To establish a “cause and effect” relationship, simultaneous moni-
toring of effluents from discharge sites and the drinking waters
derived from raw waters receiving such discharges should be per-
formed.
G. As regards the monitoring of PAH in drinking waters, although
routine monitoring at every location may not be necessary, it is
recommended that at least one center be set up to carry out the
monitoring of PAH using the present method. Training of the
treatment plant personnel to use the concentration procedure will
help reduce the overall cost of such a program.
It is also recommended that an average ratio of benzo(a)pyrene to
other carcinogenic PAll be established from the analysis of a large
number of drinking waters. Subsequently, it may be possible to
monitor for benzo(a)pyrene alone, and derive the overall concen-
tration of carcinogenic PAll by multiplying with the factor.
H. The ability of polyurethane foam plugs to collect PAIl quantitatively
upon heating of the water, also opens up a variety of new areas for
further research. Retention of contaminants such as PCB’s, pesti-
cides, phthalate esters and others will probably also be enhanced
from water at elevated temperatures, and should be examined. The
high sorbent ability and low cost suggests the feasibility of poly-
urethane foam filtration as an effective method of water treatment.
Polyurethane foam may also serve as an alternate to the carbon
adsorption method for monitoring trace organic contaminants in
water. Efficient collection of non—particulate polycyclic organic
matter from air may also be possible with polyurethane foam.
Further evaluation of polyurethane foam is necessary for these
purposes.
S
-------
SECTION IV
LITERATURE REVIEW
The first phase of the project was devoted to updating the literature
relating to the occurrence of PAll in environmental waters, their degree of
removal and/or destruction by various treatment processes, and the method
of collection and analysis of PAll. The review of collection methods
included various techniques of concentration of trace organics from water
with particular emphasis on polyurethane foam method. Analytical methods
for separation and identification of PAH from different sources, including
atmospheric particulate matters, water environment, soils and sediments,
were extensively reviewed. Some areas were searched by computerized
literature search services, whereas in other cases manual review was under-
taken. The result of the review work is reflected by the different
pertinent references cited in various sections of this report.
6
-------
SECTION V
PRECONCENTRATION METHOD
BASIS FOR DEVELOPING PRECONCENTRATION METHOD
The difficulty in identifying and quantitating trace quantities of PAH
of ten encountered in drinking water necessitates a preconcentration step to
meet detection limit of the analytical method. The methods available at the
present time are unattractive because of poor recovery, and their inability
to handle large sample volumes. Therefore, the initial study of the moni-
toring phase was devoted to the development of a suitable preconcentration
method.
Liquid—liquid extraction is the most widely used method for the precon—
centration of organic compounds from water. Although the method is employed
for the analysis of PAH in water samples, it is rendered unfeasible when the
volume of the sample becomes large. Instruments employing on—site continuous
liquid—liquid extraction (2) have doubtful practical value for routine
analysis of PAH because of slow flow rates (20—30 mi/mm) necessary to
establish an equilibrium distribution between the aqueous and organic phases.
The disadvantage of both direct and continuous solvent extraction methods is
that many good PA1 solvents (e.g. benzene) cannot be used in the system
because of their relatively high water solubility.
Concentration of PAH on a suitable sorbent offers a viable alternative.
The two most promising sorbents presently available, e.g., XAD—resin (11,
53) and Tenax (31) have limited value because of their incapability of
allowing water flow rates greater than 50 ml/min. Passage of large volumes
of water at such flow rates is very time consuming.
The discovery of the ability of polyurethane foams to retain a number of
compounds including PCB’s and organochiorine pesticides from water (3, 8, 15)
produced a surge of interest in using this technique to concentrate other
compounds as well. Cough and Gesser (17) used polyurethane foams for the
recovery of phthalate esters from water. These authors and others (51, 35)
used foam plugs coated with gas chromatographic liquid phases with varying
degrees of success. Bedford (3), however, reported that polyurethane foams
cannot be used reliably to extract PCB’s from turbid natural waters. The
studies conducted by EPA (53) with paper mill wastewater components, fuel
oil, and textile dyes showed that both coated and uncoated foams are very
limited in their extraction ability.
7
-------
Detection of very low concentrations of PAR in drinking water requires
that the preconcentration method devised should be able to deal with a large
volume of samples in a relatively short time. Preconcentration using flexible
polyurethane foams meets the above characteristics and is the method des-
cribed in this report. The optimum conditions for PAR retention by foam
plugs were first determined with radioactive benzo(a)pyrene for ease of
detection. The method was then extended successfully to five other PAR to
demonstrate its general applicability as a method of preconcentration of PAR.
EVALUATION OF FOAM PLUGS FOR EXTRACTION AND RECOVERY OF BENZO(A)PYRENE (BaP)
The initial evaluation of foam plugs for the suitability of extraction
of various PAll from large volume of water was conducted with BaP, one of
the representative PAR. Radiolabelled BaP (7—lO— 1 - 4 C) was used in these
studies because of ease and greater sensitivity of 1 - 4 C--detection. This
also eliminated the necessity for determination of background levels of Ba?
in water. It was assumed that if the preconcentration method proved success-
ful with Ba?, the method could be extended to other PAR as well.
Materials and Reagents
Radiolabelled BaP (7—lO-- 14 C): Radiolabelled Ba? was obtained from
California Bionuclear Corporation with a manufacturer’s claimed purity of
98Z. Storage of this solution showed an impurity spot on benzene developed
cellulose TLC plates as evidenced by scanning for radioactive spots on a
Nuclear Chicago Actigraph III. Purification by partitioning between
benzene—water (3:7) phase removed this impurity.
Polyurethane Foam Plugs: Flexible polyurethane foam plugs, and sheets
from which appropriate size plugs were cut, were obtained from cotmnercial
sources. The types of foams used in the study and their sources are
summarized in Table 1.
TABLE 1. CHARACTERISTICS OF VARIOUS TYPES OF FOAMS USED
Trade Chemical Plug Dimension Foam Referred to
Source Name Nature Color (diameter x Density in text as
length, mm) (kg/mi)
Scientific DiSPo Polyester White 50 x 38 24 A
Products, inc. plugs
(Batch purchased
in 1974)
Scientific DiSPo Polyester White 50 x 38 22 B
Products, Inc. plugs
(Batch purchased
in 1976)
VWR Scientific Identi Polyester White 45 x 45 25 C
plug
Thomas P. Forrest UU 4 Polyester Green 45 a 45* 24 0
C.: , In(
* Plugs were cut from 45—mis thick flexible polyurethane foam sheet.
8
-------
Sources of Water: Tap water used in these studies was obtained from
laboratory tap and is derived from Skaneateles Lake. The only treatments the
water received prior to distribution were chlorination and fluoridation (37).
Raw water was collected from Onondaga Lake (Syracuse, N.Y.) in the month of
December. Large floating and suspended particles were removed by filtration
through a fine wire screen prior to use.
Chromaflex Columns: Extender type columns of i.d. 20, 25, 40 and 50 mm
with t1011 rings, adapter and clamps were purchased from Kontes Class Co.
Oscillating—type pump, Gilmont flowmeter, and Haake Model FE thermo—
stated circulator with 1000 W. heater: These were purchased from Scientific
Products, Inc.
Radioactive Counter: The radioactive counting was done with the aid of
a Nuclear—Chicago 720 series automatic liquid scintillation system.
Radioactive Scanner: Scanning of radioactive spots on a strip of TLC
plate was performed by a Nuclear Chicago Actigraph III.
Chemicals: All solvents used were A.R. grade and purchased from
Mallinckrodt Chemical Co. The nematic liquid crystal [ N,N’bis(p—methoxy—
benzylidene)—cL,c ’—bi—p—toludine] was obtained from Eastman Kodak Co. and gas
chrotnatographic phases DC—200 and SE—30 from Analabs, Inc. Insta—gel and the
chemicals for preparing scintillation fluid were purchased from Packard
Instrument Co.
Procedure
The ability of various foam plugs to retain BaP from aqueous solution
under static conditions was determined by allowing preweighed foam plugs to
equilibrate for 4 hours in 150 ml volumes of 1 - 4 C—BaP solution of different
concentrations. The activity of the initial and final solution was deter-
mined by counting a 5 ml aliquot of the water in Insta—gel. From the
difference in the - 4 C—activity in initial and final water sample, the amount
of 1 - 4 C—BaP sorbed by foam plugs was calculated. The adsorption of BaP on
the container was taken into consideration by allowing equilibration of the
solution with the glass container prior to introducing the foam plugs and
taking the residual activity in water as the initial BaP concentration.
In flow system experiments, a foam plug was wetted with distilled water,
squeezed to expel air and placed in a Chromaflex column. Each plug was
washed with 20 ml acetone, 50 ml benzene, again 20 ml acetone, and finally
with 250 ml distilled water. In a typical experiment, four liters of water
was spiked with 1 - 4 C—BaP to a concentration of 0.1 ppb and allowed to stand
for 30 minutes to allow equilibration with glass surface. The equilibrated
solution was drawn through the column at a constant flow rate of 250 ±
10 ml/min by means of two oscillating type pumps connected in series and con-
trolled by a Variac. The flow rate was continuously monitored with the help
of a Gilmont flowmeter. Prior to reaching the foam column, the water was
brought to a desired temperature by passing it through a custom made glass
coil (25 cm x 6 mm) which was immersed in a Haake thermostated circulator.
9
-------
The coil was placed inside the reservoir housing through the opening of the
cover plate. For continuously monitoring the temperature of water during
passage through the column, a water trap equipped with a thermometer was
introduced in the line between the exit end of the column and the pump. The
complete set—up is shown in Figure 1.
The concentration of radioactive BaP in spiked water and effluent after
passage through the foam plug was determined by extracting a known aliquot
(100 and 500 ml, respectively) with 20 ml benzene and counting the 14 C—
activity in the solvent layer. To account for change in the concentration
of Ba? in spiked water with time, two samples, one at the beginning and the
other at the end of the run, were taken and the values were averaged.
The BaP retained on the foam was eluted with 20 ml acetone followed by
75 ml benzene. In experiments where two foams were used in the same column,
the volume of eluent acetone and benzene was increased to 30 and 125 ml,
respectively. Soxhiet extraction of the plugs did not elute more Ba? than
that recovered by batch column extraction and, therefore, its use was not
considered necessary. The calculation of the efficiency of retention of
BaP on the foam plugs was based on the concentration of BaP detected in
spiked water.
The amount of Ba? retained on the glass bottle, heating coil, and
connecting tubings was determined by washing them individually with enough
acetone necessary to avoid emulsion formation and enough benzene to leach
out all the activity.
In every case a 5 ml aliquot of the extracted solution was used for
racLioactive counting with 10 ml of scintillation fluid. The scintillation
fluid was composed of 5 g of 2,5—diphenyloxazole (PPO) and 0.15 g of 2,2’—
phenylenebis—(5—phenyl)oxaZole (POPOP) per liter of toluene. Water samples
were counted, when necessary, in Insta—gel.
In several experiments the foam plugs were coated with selective chrotna—
tographic phases according to the procedure of Uthe etal. (51),and tested
for their sorption ability towards Ba?.
The stability of Ba? on foam plugs was studied by passing 2 liters of
Ba? spiked tap water (0.1 ppb) through the foam plugs and storing the plugs
with sorbed Ba? for various lengths of time. In order to prevent photo—
degradation of Ba?, the plugs were stored in Chromaflex columns covered with
aluminum foil. Following the storage periods, BaP from the foam plugs was
eluted as before and counted for 14 C—activity. The eluent from the storage
experiments was concentrated and subjected to thin layer chromatography for
determining if any degradation of BaP had occurred. Cellulose plates were
spotted with the concentrate and developed with benzene as solvent. The
plates were scanned for radioactive spots on a Nuclear Chicago Actigraph III.
Most experiments were repeated 2—3 times and results are expressed as
mean of these values. The deviations from the mean were usually in the range
of 3—6%.
10
-------
Figure 1. Set—up for flow system extraction: 1 reservoir, 2 thermostated water bath, 3 foam column,
4 water trap for measuring temperature, 5 pump, 6 variac, 7 flowmeter.
-------
Results and Discussion
Sorption Characteristics Under Static and Flow Condition——
Preliminary assessment of retention characteristics of various foam plugs
was based on the ability of the plugs to retain BaP from aqueous solution
under static conditions. Musty and Nickless (35) used methylene blue adsorp-
tion from aqueous solution for determining the relative effectiveness of foam
plugs for removing pesticides. A more direct method, such as used in the
present case, has been used by Lawrence and Tosine (29).
The retention of BaP by various foam plugs increased linearly with
increase in BaP concentration as shown in Figure 2. The foam plugs differed
only to a small extent in their sorption characteristics under static con-
ditions. It was found that type D foam sorbed the greatest amount of BaP,
followed by foams B, C, and A. The polymer linkage in the foam-ester or
ether, and foam density did not appear to be related to the sorption
properties.
0
C M
E
0
U-
4 -
C
0
N
C
600
Figure 2. Sorption of BaP by various foam plugs under static conditions.
200 300 400
Benzo(a)pyrene Concentration (Mg/ I)
12
-------
The results presented in Table 2 show the recoveries of Ba? from 4 liters
of spiked—distilled and tap water obtained with various foam plugs in a con-
tinuous flow system. The results are in agreement with the recoveries obtain-
ed under static conditions. Foam type D was most effective in retaining BaP
and again, the retention abilities of various foam plugs were within a narrow
range. The recovery from tap water with various foam plugs at ambient temper-
ature ranged between 58—66%. Under similar conditions the recovery from
spiked—distilled water was considerably higher (89—96%). The decreased
retention from tap water was probably not due to competition for sites by
contaminants since the retention of Ba? from distilled water was not affected
by prior exposure to tap water. The lower efficiency appeared to be linked
to the presence of suspended particles since the tap water dosed with BaP
following Millipore filtration (O.45p) gave retention values equivalent to
distilled water.
TABLE 2. RECOVERY OF BaP FROM SPIKED WATER BY CONTINUOUS
FLOW SYSTEM: WATER TEMPERATURE, 23°C; VOLUME,
49 ; FLOW RATE, 150 ± 10 ml/min; COLUMN DIAMETER,
25 Thin.; BaP CONCN., 0.1 ppb
% Recovery from
Foam Tap water Distilled water
plug
A 62 89
B 58 91
C 65 91
D 66 96
Other Considerations in Selection of Foam Plugs—--
The water flow rates from foam plugs of type B were found not to be
uniform from plug to plug. With tap water the maximum flow rates did not
exceed 150 nil/mm. An effort was made to increase flow rates of these plugs.
The foam plugs were subjected to treatment in alkaline solutions according to
the procedure of Buist and Gudgeon (10) to clean foam cells by removing face
membranes. Such treatment of type B plugs, although increasing water flow
rates, resulted in reduction in BaP sorption capability. The possibility of
using shredded foam in place of foam plugs was also investigated. The foam
plugs were shredded with the help of a Virtis tissue homogenizer, packed
loosely in a 25 mm column, and used in the retention studies. The shredded
foam column allowed high flow rates and gave a good Ba? retention efficiency.
The efficiency, however, depended heavily on the technique of the packing of
the column, and a slight variation resulted in marked decrease in BaP reten-
tion efficiency. In view of this difficulty, the use of shredded, foam was
not considered further. 13
-------
The type D plugs were colored green and the possibility that the colored
material may be eluted and interfere with the determination of PAll led to
their exclusion from further use. Plug type A were no longer commercially
available. Based on the above considerations, type C plugs were selected for
further studies. These plugs showed good water flow rates and sorption
characteristics for BaP from spiked tap water.
Parameters Affecting Sorption Characteristics——
The following parameters have been studied with respect to their affect
on sorption characteristics.
BaP retention as a function of flow rates——The efficiency of foam
plugs to retain BaP at water flow rates in the range of 130 to 520 mi/mm was
examined. As shown in Figure 3, the recovery remained unaltered with change
in flow rates with both distilled and tap water. This is particularly encour-
aging since the adjustment of the flow rates will not be critical under most
experimental conditions, as long as the rates stay within broad limits.
Even though BaP recovery was independent of flow rates in the range
examined, a flow rate of 250 ± 10 mi/mm was preferred for further studies.
At flow rates exceeding this value, the foam plug quite often began to slide
down and rest at the bottom of the glass column during the runs. Foam plugs
pressed against the bottom of glass columns in this manner retained Ba? less
effectively.
A further study was carried out to determine the desorption of Ba? from
foam with tap water. Various volumes of tap water were passed over spiked
foam plugs, and the amount of Ba? remaining on the foam was determined. The
findings revealed no significant removal of already sorbed BaP from foam
regardless of the volume of water passed and the flow rates.
Effect of column diameter on BaP retention from tap water——The results
of increasing the diameter of the column holding the plugs on BaP recovery
from tap water is shown in Table 3. As the column diameter is increased,
(20—50 mm), the recovery of BaP on the foam plug steadily decreased; the
value with water at ambient temperature changing from 53% for a 50 m is column
to 73% for a 20 mis column. Squeezing of the foam into a small column probably
results in the reduction of foam pore size, and subsequently in more effec-
tive retention of the particulate—sorped BaP.
Although the column with 20 mm diameter allowed the highest recovery of
BaP from tap water, the squeezing of the foam in a smaller diameter column
produced difficulty in attaining flow rate of 250 mi/mm. An approximate
ratio of 2:1 between the plug and column diameter is considered the best
compromise between retention efficiency and flow rate. Accordingly, the foam
plugs of 45 mm diameter were held in a 25 mm column In the studies described
In this report.
Effect of pH on BaP retention from tap water——The effect of pH of the
spiked water on the recovery of BaP on the foam plugs is shown in Figure 4.
As can be seen, a significant increase in the retention efficiency of foam
occurs with an increase in pH. At ambient temperature the retention at tap
14
-------
Table 3. EFFECT OF DIAMETER OF COLUMN USED FOR HOLDING THE
FOAM PLUG ON BaP RETENTION FROM TAP WATER: WATER
TEMPERATURE 23°C; VOLUME, 42 ; FLOW RATE,
250 ± 10 mi/mm; BaP CONCN., 0.1 ppb
Column diameter % Recovery of
(mm) BaP
20
73
25
65
40
62
50
53
I I
90- a A
80 —
70 -
- 0 0
— 0 0 -o
60
w
450
2 40 —
C
— a Distilled Water
- o Tap Water
20 —
10 —
U I I I
0 100 200 300 400 500 600
Water Flow Rates ( mi/mm.
Figure 3. Retention of BaP from spiked tap water at various flow rates:
water temperature, 23°C; volume, 49 ; BaP concentration, 0.1 ppb.
15
-------
70 —
60-
50 —
10 --
0 1 2 3 4 5 6
Water pH
7 8 9 10 11 12
Influence of water pH on BaP
23°C; volume, 49 ,; flow rate,
tration, 0.1 ppb.
retention: water temperature,
250 ± 10 mi/mm; BaP concen—
80 -
70 -
60 -
I I I
0 tO 20 30 40 50 60 70 80
Temperature of Water Passed Through Foam (°C)
Figure 5.
Effect of water temperature on recovery of BaP on foam plugs:
volume of spiked water, 42W; BaP concentration, 0.1 ppb; flow
rates, 250 ± 10 mi/mm. o = tap water (unfiltered), . =
filtered tap water, E = distilled water.
eq
eq
8
C
C.
0
N
Jill l -t—I ‘I
A
I I I i I i_ __________________
Figure 4.
C
41
C
a-
C
U
e
C
C
0
C
16
-------
water pH (6.7) was between 62—65%, but upon increasing the pH to 10.0, the
retention efficiency increased to 76%. Lowering the water pH below 6.7
resulted in a decrease in BaP retention. A pH—linked increase in the recov-
ery of chlorinated hydrocarbons on polyurethane foam plugs has been noted by
Musty and Nickless (35).
The increase in retention with increase in pH may be attributed to the
desorption of BaP from the suspended particles and subsequent sorption on
the foam plug, or to coagulation of suspended particles and minerals in water
at higher pH enhancing retention through filtration. Alternatively, a
pH—dependent increase In the sorption process in the foam may be responsible
for increased retention of Ba?. If a charge—transfer complex formation is
involved between the foam polymer and Ba?, such an increase is not improbable.
Effect of water temperature on Ba? recovery——The most dramatic effect on
the recovery of benzo(a)pyrene from spiked tap water was observed when the
temperature of the water was varied (Figure 5). The relationship between BaP
recovery from tap water and temperature was found to be biphasic. The per-
cent retention steadily increased with increase in temperature up to 40°C,
but decreased with further increase in temperature. When the temperature was
increased beyond 50°C, the increase in Ba? retention was resumed until a
plateau was reached starting at 60°C. The recovery of BaP at temperature
> 60°C was approximately 87%.
The effect of temperature on BaP recovery from tap water is complex and
probably the consequence of many interacting factors. The initial increase
appears to be linked to the presence of suspended particles in water. This
is due to the fact that such an increase was not seen with spiked distilled
water, and that the increase was less pronounced when tap water was
Millipore—filtered prior to spiking. It is suggested that heating of tap
water causes desorption of BaP from suspended particles and thus prevents
loss of Ba? which otherwise occurs due to passage of particle—sorbed BaP
through the porous foam.
The increased retention of Ba? beyond 50°C is observed with tap water,
distilled water, as well as filtered water and, therefore, appears to be
linked to the foam itself. Such an increase could be attributed to a possible
change in the conformation of the polymer at higher temperature, thereby
increasing sorption of BaP.
Subjecting of polyurethane foam to the action of hot water or steam
causes hydrolysis of residual isocyanate and ethyl silicate (47). The
increased retention of BaP at higher temperature may be linked to these
chemical changes in the foam. However, this did not appear to be the case,
since the pre—heated or steam—treated plugs gave recovery of BaP similar to
untreated foam. The data suggests that the increase in Ba? sorption at higher
temperature is not due to chemical change in the foam, but probably to a
reversible conformational change.
Effect of coating foams on Ba? recovery——The retention ability of foam
plugs for many pesticides has been shown to be improved by coating the plugs
17
-------
with selective sorbents (51). Studies were undertaken to determine if coat-
ing of the plugs with suitable chromatographic phase can also improve the
retention of BaP from water. The chromatographic phases tested included
nematic liquid crystal, SE—30 and DC—200. Both nematic crystal and SE—30
have been used earlier as gas chroinatographic phases for separation of PAR
compounds (24, 45). DC—200 was also tested because Uthe etal. (51) reported
good recoveries of pesticides with foam plugs coated with this phase. With
the foams coated to the extent of 5—10% by weight, an increase in BaP recov-
ery of 4—9% over untreated foam was observed at a water temperature of
62°C ± 2 (Table 4). The eluate from coated—foam plugs contained large quan-
tity of the coating material and its concentration to smaller volume (which
would be necessary in the later stages of development) was difficult.
Furthermore, since only a small increase in BaP retention was noted in the
presence of the chromatographic phases, the coating of the foam plugs was not
considered further.
TABLE 4. BENZO(a)PYRENE RETENTION FROM TAP WATER WITH FOAM PLUGS
COATED WITH CHROMATOGRAPHIC PHASES: WATER VOLUME 42 ;
FLOW RATE, 250 ± 10 mi/mm; BaP CONCENTRATION, 0.1 ppb
Concentration of Temperature of
Chromatographic coating on plug spiked water % Retention
phase (% of foam, w/w) (°C) on foam
Uncoated 23 62.0
Uncoated 62 85.3
DC—200 5 62 87.5
DC—200 10 62 92.6
SE—30 10 62 91.2
Nematic liquid
crystal 3.7 23 66.0
(N ,N —bis(p—methoxybenzylidine) —
c ,a ‘—bi—p—toiudine]
18
-------
Effect of BaP concentration on_recovery——In order to become an effective
method of preconcentration, the foam plugs should demonstrate high and con-
sistent recovery at varying concentrations of BaP usually encountered In
treated and raw waters. In view of this, the recovery studies with the foam
plugs were carried out at different concentrations. The results of the study
are shown in Table 5. The retention of BaP did not vary significantly with
change in BaP concentration in the range examined (0.002—25 ppb). This can
be interpreted to mean that the polyurethane foam plugs of the dimension and
chemical characteristics used in this work can effectively concentrate BaP
from tap water over a broad concentration range.
TABLE 5. RECOVERIES OF BaP FROM SPIKED TAP WATER AT VARIOUS
CONCENTRATIONS: WATER VOLUME, 4 ; FLOW RATE,
250 ± 10 mi/mm; TEMPERATURE, 62 ± 2°C
Concn. of
(ppb)
*
BaP
% Retention of foam
25
84.0
0.1
86.0
0.05
84.0
0.02
83.0
0.002
87.0
*
On the basis of the amount added to water
14
Mass Balance of C—Activity——
Benzo(a)pyrene, like other PAR, has a tendency to stay adsorbed on
solid surfaces. Considerable amount of BaP added to water can be expected
to be adsorbed to the wall of the reservoir, glass coil and connecting tube,
etc., employed in the experimental set—up. Experiments were carried out to
determine complete mass balance of the BaP added to water to account for
losses due to sorption on various surfaces. The studies were carried out at
62 ± 2°C with tap as well as distilled water.
19
-------
As can be seen in Table 6, nearly 25% of the total 14 C—added to tap
water was recovered from the bottle surface. From distilled water, however,
the loss to the bottle was only 15%. The extractability of BaP with benzene
from the two types of water also varied to a significant extent. All the BaP
was extractable with benzene from spiked distilled water. The amount recover-
able from tap water was only 92%. Whether this is due to transformation of
BaP to some other form not extractable with benzene or due to non—extractable
particle—adsorbed portion of BaP is not clear. Acheson et al. (1) have
suggested that adsorption of PAB upon suspended solids may lead to a change
in extraction efficiency.
The calculation of the percent retention of BaP by foam plugs in this
report has been based on the amount of - 4 C—detected in spiked water by ben—
zene extraction. This eliminates the necessity for quantitation of the BaP
adsorbed on the reservoir surface. The total loss of BaP due to adsorption
on the glass coil used for heating of the water, and on connecting tubes was
small, and ignored in the calculation of the percent retention efficiency.
TABLE 6. MASS BALANCE OF 14 C-ACTIVITY ADDED TO WATER;
WATER VOLUME, 49W, TAKEN IN A 52 BOTTLE;
FLOW RATE, 250 + 10 mi/mm; TEMPERATURE,
62 ± 2°C; BaP CONCENTRATION, 0.1 ppb
Tap Water
Z’ C—distribution
Material of the amount of the amount
Tested added to water detected in water*
Distilled Water
%1 C—distribution
of the amount of the amount
added to water detected in water*
Foam plug +
glass 56.0
column
Glass Bottle 24.4
Glass coil and
connecting 2.1
tubes
Effluent 8.0
lL C Non 8 0
extractable
86.0
—
2.8
11.0
800
15.0
1.8
3.2
95.0
—
2.1
3.3
Total 98.5
99.8
100
100.4
*
by extraction with benzene
20
-------
Determination of BaP Breakthrough Volume With Tap Water——
Ef forts were made to determine the volume of water from which a single
foam plug could concentrate BaP efficiently. Increasing volumes of spiked
water were passed through individual foam plugs and the efficiency of BaP
retention was determined in each case. The results showed that the eff i—
ciency of retention steadily declined as the sample volume increased; the
efficiency of retention with 42 of tap water was nearly 86%, however, when
the sample volume was increased to 4O2 the efficiency fell to nearly 50%
(Table 7). The efficiency of retention from distilled water, however,
remained > 95% even at sample volume of 40i.
TABLE 7. EFFECT OF SAMPLE VOLUME ON THE RECOVERY OF BaP WITh
A SINGLE FOAM PLUG. FLOW RATE, 250 ± 10 mi/mm;
TEMPERATURE, 62 + 2°C; BaP CONCN., 0.05 ppb
Descripti
of water
on
Sampl
volume
e
(I)
% BaP retention
Tap water
4
86.5
5
84
10
73
20
67
40
49
Distilled
water
4
98
40
95
In an attempt to increase the efficiency of retention with larger volumes
of water, the number of foam plugs in the column was increased. With 202, of
spiked—tap water, 4 plugs — two each in two different columns — gave a recov—
ery of 85.5%. The distribution of BaP retained on individual plugs was as
follows: 65% on the first plug,4% on the second, 13% on the third (first
plug on the second column), and 3.5% on the fourth.
BaP Recovery from Spiked Surface Waters——
It was of interest to determine if polyurethane foam plugs can effectively
concentrate BaP from raw water as well. Onondaga Lake water was chosen as a
case of raw water as it would represent a worst possible case of raw drinking
water source (total suspended solids in water = 102 ing/l; total solids =
2.4 gIl). The retention from 42 of spiked raw water (0.1 ppb) with a single
foam plug was found to be 69%. This indicates that breakthrough occurs
earlier for raw water than tap water. When the number of plugs in the column
was increased to two, the efficiency of retention resumed to normal value.
Therefore, twice as many total foam plugs (not exceeding two plugs per column)
should be sufficient to effectively concentrate BaP from the same volume of
21
-------
raw water as with tap water. Since the concentration of BaP in raw water can
be expected to be hither, the number of foam plugs require J can be cut down
by decreasing the sample size.
Stability of BaP on Foam Plugs——
Prior to considering foam plugs for field monitoring, it is important to
assess the stability of BaP on foam plugs. The effect of storage at room
temperature and in refrigerator was compared over a 7 day period. Almost all
the 14 C—activity was recoverable from foam plugs after 7 days storage at room
temperature (Table 8). Thin—layer chromatography of the eluates from foam
plugs stored at room temperature and at 4°C, revealed no measurable
activity in any spot other than BaP. The data suggests that BaP is sufficient-
ly stable on foam plugs for transportation to the laboratory for analysis.
However, cooling of the foam plugs to 4°C during transportation is suggested.
TABLE 8. STABILITY OF BaP ON FOAM PLUGS STORED AT ROOM
TEMPERATURE M D AT 4°C. EACH PLUG WAS SPIKED
WITH APPROXIMATELY 0.2 g BaP
Storage
temperature
% recovery of BaP at different
(days)
storage periods
1
2
4
7
4°C
100
91
103
95
Ambient
98
90
87
82
Conclusion
Polyurethane foam plugs have been found to be excellent sorbent for
benzo(a)pyrene from treated and untreated waters. For sampling 209. of drink-
ing water, 4 foam plugs — two each in two different columns — should be used.
The same number of foam plugs should be used for 109. raw water. The water
should be heated to 60—65°C prior to passage through the foam column, and
flow rate should be maintained at nearly 250 mi/mm. Foam plugs following
sampling should be shipped in ice to prevent loss of benzo(a)pyrene.
The increase in the efficiency of retention of BaP by heating of water
appears to be linked to the desorption of benzo(a)pyrene from suspended
particles in water, and to a possible change in the conformation of the
polymer at higher temperatures. It is felt that the retention of other PA l-I
as well as other compounds, e.g. PCB’s, pesticides, on polyurethane foam
will also be enhanced if water temperature is increased, and should be
Investigated.
22
-------
SECTION VI.
METHOD OF PAR ANALYSIS
BASIS FOR DEVELOPING ANALYTICAL METHOD
The previous section describes the preconcentration of BaP from water
samples using polyurethane foam plugs as sorbent. The use of radiolabelled—
BaP eliminated the need for any chemical separation procedure for its isola-
tion from impurities and the need for any selective identification method for
its estimation. Establishing the general validity of the method towards
other PAN requires the determination of the efficiencies of retention of
other PAR and thus the development of an analytical procedure involving clean-
up of foam extract and a selective identification procedure for their quan—
titation. The analytical procedure is required to provide separation to the
extent necessary for their interference—f ree detection. The problem in the
development of an analytical method for the determination of PAR which
usually occur as a small quantity in a large matrix of impurities is three-
fold. First, the compounds of interest must be adequately separated from
impurities concentrated from water along with PAR and those which are eluted
from foam, second, the PAR must be separated from each other, and third, a
high sensitivity of detection is required to quantitate small amounts of PAR.
The choice of procedure for the separation of PAIl from the interfering
classes of compounds depend largely on the type of sample to be analyzed.
Although the literature concerning PAR separation in various other types of
samples is abundant (8,13,14,28), it is not so in the case of treated and raw
waters. Irrespective of the nature of sample, no single analytical separa-
tion procedure to date is capable of providing complete separation and
resolution of the PAR fractions. Therefore, analyses of PAR mixtures have
typically entailed various partitioning sequences followed by column, thin
layer or paper chromatography prior to detection.
Aches on et al. (1) reported that TLC procedure for separation of PAR
compounds was highly effective and less susceptible to interference by back-
ground organic materials. TLC has been successfully used for separation of
PAR isomers (40). In recent years, a number of researchers have used TLC in
combination with other methods for the separation and identification of PAR
in complex mixtures (9,26).
Gas chromatographic separation of PAR compounds using Dexsil—300 (18,27)
column is another versatile method which combines simplicity of operation
with relatively high sensitivity. Strosher and Hodgson (48) have separated
PAR compounds from lake waters and associated sediments using this column.
23
-------
However, the resolution of some of the closely related compounds has not been
achieved. Janini etal. (24) have claimed that a new G.C. column using
nematic liquid crystal can accomplish the above separation. But column bleed
at higher temperatures still remains a problem with this packing material.
Although surface coated, open tubular (SCOT) columns (27,28) and high
efficiency glass capillary columns (30) have been used in the past for
separation of PAR, their limited total sample load capability restricts the
detection limit to an undesirable value. High pressure liquid chromatography
(hplc) with bonded octadecylsilyl phases of microparticle size (16) is a
particularly promising technique for the separation of PAR compounds. But
the cost of the instrumentation makes it a restrictive alternative to other
techniques available for routine analysis.
With regard to detection, fluorescence, spectroscopy has become well
established as a sensitive and selective analytical technique for PAM.
Several groups (12,46) have found it to be at least ten times more sensitive
than U.V. method. It is also more sensitive and less expensive than mass
spectral detectors. Mass spectrometers normally have a nanogram detection
limit (21,41), although integrated ion current techniques reduce this limit
to the subpicogram range (33,39).
This section describes a method for the separation of PAR from impuri-
ties, and their qualitative and quantitative determination. The clean—up
procedure utilizes the ability of the PAR compounds to form charge—transfer
complexes for separating these compounds from the bulk of impurities.
Further purification which was found necessary was performed on a short
Florisil column. The identification and quantitation was done by GLC—f lame
ionization techniques, and by spectrofluorophotometry following two dimen-
sional thin layer chromatography.
SELECTION OF PAR FOR STUDY
The number of PAR which have been detected in environmental samples is
considerable. Since it is impossible to detect all of them it was found
practical to choose several specific and representative compounds and limit
the monitoring phase of the project to these.
The six PAM included in this study are: benzo(a)pyrene, fluoranthene,
benzo(k)fluoranthene, benzo(j )fluoranthene, indeno(l ,2 , 3—cd)pyrene, benzo(ghl)—
perylene. With the exception of benzo(j)fluoranthene, WHO (5) recommends
analysis of these PAR in drinking waters. Benzo(b)fluoranthene, which is
reconmended for analysis by WHO, has been replaced with benzo(j)fluoranthene
in our analysis due to its non—availability. The group of the 6 PAR is
regarded as representative of the whole polynuclear family.
The structure and carcinogenic properties of the selected PAR are given
in Table 9.
24
-------
TABLE 9. PAR COMPOUNDS STUDIED
Symbol Compound Synonym Structure Emperical Carcinogenic
formula Potency*
BaP benzo(a)pyrene 3,4 benzpyrene C 20 H 20
BkF benzo(k)fluoranthene 8,9—ben fluoranthene C H
20 12
BjF benzo(j)fluoranthene 7,8—benzfluoranthene - C 2 H 12 -4—f
FL fluoranthene C H
1 ñ 10
IP indeno(1,2,3—cd)pyrene 0—phenylenepyrene C 22 H 12 +
BPR benzo(ghi)pery lene 1,L2—benzpery lene
--- - -
* National Academy of Sciencc , 1972 (36)
-------
EXPERIMENTAL METHODS AND RESULTS
Necessary Precautions
In the analysis of samples containing PAR in ppb—ppt range, the necessity
of preparation of scrupulously clean glassware cannot be overemphasized. All
the glassware should be free from grease, and Teflon stopcock should be used
in place of glass. No smoking should be permitted in the laboratory. PA l- I are
light sensitive and therefore all the glassware used in transportation,
analysis and storage of PAR must be wrapped with aluminum foil and all the
work be done in subdued light. Appropriate precautions must be taken in work-
ing with these compounds because of their carcinogenic nature.
Materials and Reagents
PAR Standards : Fluoranthene, Benzo(ghi)perylene, Indeno—(l , 2, 3—cd)pyrene
and Benzo(a)pyrene were obtained from Aldrich Chemical Co.; Benzo(j)— and (k)
fluoranthene were supplied by Mr. J.L. Monkman, Air Pollution Control
Directorate, Ottawa, Canada.
Materials for gas liquid chroinatographic analysis :
Gases for GLC: N 2 and 112, prepurified grade; air
breathing grade: Linde Division, Union Carbide Corp.
Aluminum foil backed septa (Netasep), GLC 6 ft x 1/8 inch
matched columns packed with 3% Dexsil 300 on chromosorb W
(A.W,) 100/120 mesh from Altech Associates.
GLC with dual flame ionization detector: Hewlett—Packard,
Model No. 573OA.
Materials for thin layer chromatograph :
Acetylated cellulose (40%), Aluminum Oxide G, type E,
8 x 8 cm. glass plates, TLC Tank, Desaga template,
Desaga Brinkman spreader: Brinkinan Instrument, Inc.
Spectrophoto—fluorimeter with thin film scanner;
American Instrument Co.
U.V. lamp (Mineralight) for visualization of TLC spots:
Scientific Products, Inc.
Other Materials :
Solvents, Distilled in glass: Mallinckrodt Chemical Co.
and Burdick Jackson Lab, Inc.
Florisil (60—100 Mesh) chromatographic grade, Matheson
Coleman and Bell, Inc.
Rotary vacuum evaporator with thermoregulator and immersion
heater, Buchler Model, Scientific Product, Inc.
26
-------
Calibrated tubes (6.5 ml): Fisher Scientific Co. Each of
the last four tenth ml is calibrated to one—hundreth ml.
Source of other materials and reagents not described here — see
Section V., p. 8.
Standard PAR Solution
Stock solutions of each of the six PAR are prepared by dissolving
100 mg in 100 ml benzene in a volumetric flask (1000 ppm).
In order to prepare standard PAR mixture for gas chromatographic analysis,
mix 1 ml each of the six stock solutions in a 10 ml volumetric flask and make
up the volume with benzene (finalconcn. 100 ppm). Standard PAH mixture for
TLC is prepared by mixing 5 ml of the standard fluoranthene solution and 1 ml
each of the five other PAR in a glass stoppered bottle (final concn. 50 ppm
fluoranthene, 10 ppm others).
PAR Separation and Analysis
PAR analysis was performed using two techniques — gas liquid chroma-
tography — FID detection, and thin layer chromatography—fluorescence detection.
The relative merits of these and other methods of PAH analysis have been dis-
cussed by many researchers (1,44). The TLC procedure used was of Borneff (6)
and recommended by the World Health Organization (55). The method was
employed by us after slight modifications. It permits qualitative as well as
quantitative determination of PAH directly on the TLC plate, and thus losses
associated with removal of PAR for fluorescence measurement are eliminated.
The details of the two methods and their sensitivity and scope under our
experimental conditions are given below.
Gas Liquid Chromatography — FID Detection——
The gas chromatograph used was equipped with a dual flame ionization
detector and a linear temperature programmer. The experimental conditions
used were as follows:
Column: 6 ft x 1/8” stainless steel packed with 3%
Dexsil 300 on chromosorb W, 100—120 mesh (columns
preconditioned at 325°C for 24 hrs. and operated in
differential mode)
Carrier gas (N 2 ) flow rate: 30 mi/mm.
H 2 gas flow rate: 30 mi/mm.
Air flow rate: 300 mi/mm.
Detector temp.: 300°C
Injection Port temp.: 250°C
27
-------
Column oven temp. programming:
Initial temp.: 200°C
Initial delay: 2 mm.
Program rate: 4°C/mm.
Final temp.: 290°C
Final delay: 8 mm.
The maximum temperature of 290°C used in the present study for GLC
temperature programming completely eluted all the six components. This is in
conformity with the work of Lao etal. (27). The temperature programming,
however, caused a stepwise increase in the baseline up to the maximum pro-
grammed temperature and then the baseline fell rather sharply during the
cooling cycle. To compensate for the column bleeding effect, an identical
second column was used in a differential mode. Even under this condition,
the bleeding from septa remained as a problem. With most septa used, at least
seven ghost peaks were evident in the chromatograms. The use of aluminum—foil
backed septa (Metasep) which had been preconditioned for 12 hours at 250°C
gave satisfactory results.
To avoid the error due to dead volume in the syringe, the following
solvent flush injection technique was used:
Draw about one i.cl of solvent into the syringe and withdraw the plunger
to introduce an air gap of about 1 iii. Insert the syringe needle into the
solution to be analyzed and withdraw the plunger till the top of the air gap
differentially moves to the desired volume to be sampled. Withdraw the
syringe needle from the solution and withdraw the plunger till an air gap
shows at the bottom of the barrel. Precisely read the volume of the liquid
trapped between the two air gaps (see diagram below).
c i )
0
ci )
ci)
z
Solvent
I -i
CID
Air
Sample
Air
28
-------
Fig. 6 illustrates the gas chromatograms of the standard PAR mixture.
The identification of the PAR on the chromatogram was achieved by comparison
of the relative retention times (RRT) of each PAR determined under identical
conditions. The 6 PAR injected resulted in 5 peaks, the isomers — benzo(k)—
fluoranthene and benzo(j)fluoranthene — could not be separated on this column.
Quantitative estimation of the individual PAR concentration was per-
formed by peak height measurement for the peaks which were sharp and symmetri-
cal. PAR showing wide peaks were quantitated by evaluating their area with
the method of multiplication of the peak height by the width at half height.
Standard curves for each compound were constructed to find out the linearity
of responsewith concentration. Fig. 7 shows two such curves obtained from
fluoranthene and benzo(ghl)perylene, the first and last eluent from the
GLC column.
As revealed earlier, the gas chromatographic column used was unable to
separate the isomers benzo(j)— and benzo(k)fluoranthene. From the relative
retention times compiled by Lao et al. (27) for 12 ft x 1/8” Dexsil 300
column it can be predicted that present GLC conditions will not separate the
following groups of compounds: benzo(b)—, (j)— and (k)-fluoranthene;
benzo(a)— and (e) pyrene/perylene; indeno(l,2,3—cd)pyrene/benzo(b)—chyrsene/
picene; and benzo(ghi)perylene/anthranthene. The incomplete resolution will
result in higher values for the estimated PAR concentrations in environmental
samples than actually present. The other limitation of GLC analysis using
liquid injection technique is its inherent limit on the sample volume
(<5p1) that can be injected. This makes the detection of low levels of PAR
impossible. In view of these disadvantages of the GLC method, efforts were
directed to evaluate the scope of the TLC—fluorometric method of PAH analysis.
Thin—layer Chromatography—Fluorescence Detection——
Preparation of Plates : Mix 28g aluminum oxide C, type E, 12g 40%
acetylated cellulose and 2 g CaSO 4 .2H 2 0 (200 mesh) with 83 ml 95% ethanol with
a magnetic stirrer for 5 mm. The resultant slurry was spread to a thickness
of 250 irn on eight 20 x 20 cm glass plates using the spreader. The plates
were dried for 1/2 hour and approximately 2 nun of adsorbent was scraped from
all sides to prevent “edge effects.” Activate the plates in an oven for
30 minutes at 80°C. Store the plates in a desicator containing silica gel.
The adsorbent layer on the plates is extremely fragile and should be handled
with special care.
Sample Application : With the help of a syringe, appropriate aliquot of
the concentrate or standard was applied on one corner about 2 cm away from
both sides of the plate. Samples were applied in successive small doses and
dried by passing prepurified grade N 2 to minimize spot size. For best results,
the amount of individual PAR should remain between 20—100 ng/spot.
Development of Plates : The first dimensional development of the plates
was done in a n—hexane: benzene (4:1) in a tightly sealed development tank.
The solvent front was allowed to run about 2 cm below the dry end of the
plate. The development time was approximately 30 mm. After drying, the
plate was rotated by 90° and developed in the second direction with methanol:
ether:water (4:4:1). It takes approximately 2 hours for the solvent front to
29
-------
Figure 6. Gas chromatogram of the standard PAIl mixture on Dexsil 300
packed column.
LU
N
a,
C
0
a,
C,
‘a
0,
0
‘S
0
.0
0
C
=
‘ S
0 ?
0.1
Amount of Compound In pg
0.2
Figures 7.
Calibration curves for two representative PAM deter iined
by GLC—FID method.
0
FL
B(j)F *B(k)F
IF BRP
BaP
0 4 8
T,rne(m,n.)
20
24
28
100 -
50 —
0
0
o FL
• BPR
0.3
30
-------
run about 2 cm below the end of the plate. The plates were then air dried
and luminescent spots visualized by illuminating the plate in dark with a low
intensity U.V. lamp. The boundaries of the spots were marked with a clean,
sharp stainless steel needle. If necessary, developed plates can be stored
in the dark in a desicator.
The excellent resolution obtained with the TLC system (Fig. 8) for
compounds not separated by the GLC method prompted further use of the method.
The RB values defined as the ratio of the distance travelled by a specific
compound to that of benzo(a)pyrene appear in Table 10 for the two solvent
systems. The values range from 0.5 — 1.8 in the first system and 1 — 3.1 in
the second system, showing good separation for the 6 PAR.
0
0
U,
0
BPR cI FL
EII
C Q
BaP
o Origin
1st Development —
Figure 8. Thin—layer chromatogram of standard PAR mixture (100 ng
fluoranthene and 20 ng each of the other 5 PAR).
31
-------
TABLE 10. RB VALUES OF PMI IN TWO SOLVENT SYSTEMS
Solvent system
RB Value
1
Solvent system
2
FL
1.87
3.15
BjF
0.94
1.46
BkF
0.99
2.13
BaP
1.0
‘
1.0
IP
0.60
1.62
BPR
0.57
3.19
Solvent system 1: n—hexane—benzene (4:1)
Solvent system 2: methanol—ether—water (4:4:1)
Identification of Spots on TLC plate : Identification was based on 3
criteria: (1) fluorescence color, (Ii) RB values, and (iii) characteristic
bands in the excitation and fluorescence spectra. The fluorescence colors
of the six PAR are given in Table 11.
TABLE 11. FLUORESCENCE COLORS OF PAR
Compounds Color of fluorescence
FL light blue
BjF yellow—orange
BkF dark blue
BaP violet
IP yellow—green
BPR violet
32
-------
The emission and excitation spectra of all the spots were run directly
on the plates with the help of spectrophotofluorometer equipped with a thin—
film scanner. The excitation and emission sweep power of the monochromotor
was synchronized between 200 and 800 m i of the X—Y recorder chart paper.
The photomultiplier slit was so adjusted that it provided the fine structure
in the spectra. Emission spectra of each individual spot except the suspected
fluoranthene spot was obtained by setting the excitation wavelength at 300 nm
and scanning each spot between 350 and 550 nm. In case of fluoranthene, the
excitation wavelength was set at 280 nm and the spot was scanned from 400 to
550 nm. Similarly, the excitation spectra of each spot was obtained by set-
ting the emission wavelengths at one of the emission maxima exhibited in the
emission spectra and scanning the spot from 220 to 400 nm. The excitation
and emission wavelengths used for running spectra are summarized in Table 12.
TABLE 12. EXCITATION AND EMISSION WAVELENGTHS USED FOR RUNNING SPECTRA
Compound
Emission spectra
A for excitation
(nm)
Excitation spectra
x for emission
(nm)
FL
280
458
BjF
300
427
BkF
300
428
BaP
300
427
IP
300
467
BPR
300
430
The identification on the basis of fluorescence color is at best tenta-
tive. In a complex chromatogram, the identification on the basis of RB values
which may vary by as much as 15% can be misleading (25,43,54). The best
method of identification is the matching of the shape and characteristic bands
between the unknown and known spots of the standard compounds. Although the
emission bands of some of the compounds, e.g., benzo(a)pyrene and benzo(k)—
fluoranthene are very similar, their excitation bands are quite distinctive.
Similarly, the lack of fine structure in the emission spectra of both fluor—
anthene and benzo(j)fluoranthene poses no identification problems because of
their distinctive excitation spectra. The excitation and fluorescence spectra
of the six PAR appear in Figure 9.
33
-------
600
450 500 550 600
I I I
B(I)F BIj)F
/ - I I
200 250 300 350 400 350 400 450 500 550 600
W. eIength I no, I
Figure 9. Fluorescence emission and excitation spectra of model PM-i
compounds obtained directly on the plate.
continued
Co
0
0
“C
200
250 300
350 400
350 400 450 500 550
Wave l.oqth (nm)
BIkIF
B(k)F
I I
200 250 300 350 400 350 400
Wa s4ength nmI
34
-------
Figure 9. continued
FL
400 350 400
Wavelength ( nm)
450 500 550 600
‘P
I I
I
. \
I
- _J_ _i____
200 250 300 350 400
‘P
I I I
I I
350 450
Wavelength I nm)
BPA
(
/
BPR
• III I -_
200 ,5 JOL 50 400 350 400 450 500 550 600
W.v. l.nqthlnm)
c c
0
U,
FL
S
0
=
200 250 300 350
I I I I
I I
400 500 550 600
35
-------
Quantitative Analysis : Quantitation was performed by scanning each spot
directly on the plate for fluorescenc.e intensity. The following experimental
conditions were used:
Recorder (50 my) chart speed: 4”Imin.
Scanner cycle time: 2 mm.
Excitation wavelength: 365 nm
Emission wavelength in run
FL: 458
BjF: 427
BkF: 428
BaP: 427
IP: 467
BPR: 416
For each marked spot, the position of the light beam on the spot was
adjusted such that the maximum fluorescent signal was recorded. Each spot
was, then, scanned for fluorescence intensity at a photomultiplier slit
width of 2.0. The output from the photomultiplier tube (IP 28) was applied
to a strip chart recorder to obtain a trace of each spot. The direction of
scanning was selected in a way such that interference from adjoining spots
was minimum.
The concentration of PAM was determined from the calibration curve
obtained for each individual compound using three different concentrations.
The area under the fluorescence peak was determined with the help of a
planimeter. The range of linearity for the six PAM is represented in
Figure 10.
The variation in the intensity of the light source causing error in the
quantitative values, was studied by measuring the intensity regularly with a
standard 100 ng quinine sulfate spot. Any variation of intensity was correct-
ed for in the quantitative values of the PAM concentrations. Studies by
Keegan (25) showed that the fluorescence intensity of individual PAH spots
decreased significantly as the moisture content of the plates decreased.
Therefore, thorough drying of the plates after development was necessary for
reproducible results.
Clean—up Procedure for Removing Impurities of Water and Foam Origin
During the concentration of PAM from water on foam plugs, several other
contaminants also get concentrated and some get eluted during PAH elution.
In addition, several impurities belonging to the foam are also leached during
the elution process. Figure 11 shows a gas chromatogram of the water concen-
trate prepared with the help of foam plugs. These impurities interfered with
the analysis of PAM. Pre—cleaning of the plugs with organic solvent by batch
or soxhlet extraction failed to remove the trace impurities from foam. In
fact, soxhieting of the foam increased the levels of impurities extracted.
Efforts were, therefore, directed to devise a clean—up procedure for removing
impurities originating from the foams and those derived from water.
36
-------
B{k)F
‘P
B(j)F
70 —
so -
7—.
6—
5— 25- 50-
u.
4— 20- 40-
I::
1— 5—
0— — I
Figure 10. Calibration curves for the six reference compounds determined by fluorimetric method
directly on the plate.
1’1 1•1• I
0 10 20 30 40 50
Amount oi PAH Nanograms)
10 -
I 0-
60 70 0
I I I
10 20 30 40
SaP, BPH
1 I ‘ I
50 60 70
I I ‘I
0 50 100 150 200 250 300 350
FL.
Nanogranit
BaP
BPR
35-
30 -
0
-------
The efficiency of the clean—up procedure and the recovery of PAR was
first evaluated using gas chromatographic method of PAR analysis. Studies
were then undertaken to determine if interferring substances could be
detected by thin—layer chromatography which allowed spotting large volumes
of the concentrate.
Earlier studies (Section V, p. 21) showed that PAR from 2O2 of tap water
could be effectively concentrated with the help of 4 foam plugs placed in two
different columns. Assuming that a sample volume of 609 of tap water should
be adequate to detect PAR in drinking waters, efforts were directed to devise
a clean—up procedure which will remove interfering substances introduced from
602, of tap water plus those leached from 12 foam plugs.
The details of the clean—up procedure and its efficiency is covered in
this section.
Preparation of Sample Containing Interfering Substances——
Twelve foam plugs are placed in 6 Chromaflex columns and washed as
described before (Section V, p. 9). Sixty liters of unspiked tap water are
passed through the plugs maintaining a temperature at 62 ± 2°C and flow rate
at 250 ± 10 mi/mm. Elute each column with 30 ml acetone and 125 ml of cyclo—
hexane. Mix the eluate and spike it with 10 ug each of the 6 PAR. The
contribution of PAIl from the tap water passed over the foam plugs was deter-
mined to be insignificant in comparison to the amount added and was, therefore,
ignored.
Experimental Details of Clean—up Procedure——
(i) Solvent Partitioning — Add 50 ml distilled water to the eluate,
shake the contents thoroughly and let it stand till the layers separate.
Discard the bottom aqueous/acetone layer and transfer the organic layer with
two 20 ml washings of cyclohexane into a round bottom flask. Concentrate the
contents to about 10 ml with a rotary evaporator at a temperature of 40°C.
Transfer the extract along with two 25 ml cyclohexane washes into a 250 ml
separatory funnel. Wash the cyclohexane layer twice with 60 ml 4:1 methanol—
water and twice with 60 ml distilled water. Add 20 ml dimethylsulfoxide
(DM80) to the cyclohexane layer and shake the contents for 3 minutes. Let it
stand and when the layers separate, withdraw the DMSO layer to another clean
250 ml separatory funnel. Repeat the DMSO extraction two more times and
dilute the combined DMSO extract with 120 ml distilled water.
Extract the PAR from the aqueous DMSO phase by shaking it with 40 ml
cyclohexane for 5 minutes. Repeat the cyclohexane extraction one more time.
If emulsion formation poses any problem at this stage, addition of approxi-
mately 0.5 g of cyclohexane washed arihydrous Na 2 SO4 should help breaking the
emulsion. Wash the combined cyclohexane layers twice with 60 ml distilled
water and dehydrate by passing it through an approximately 15 g Na 2 SO 4
bed supported on glass wool and prewashed with cyciohexane. Wash the
separatory funnel twice with 10 ml cyclohexane and add the washi ,ngs to the
Na 2 SO 4 bed. Wash the Na 2 SO 4 bed with an additional 20 ml cyclohexane and
collect the cyclohexane and the washings in a 200 ml round bottom flask.
Concentrate the extract to 5 ml by rotary evaporation. At no time the PAR
mixture should be allowed to proceed to complete dryness. This has been
shown to result in a loss of PAR (14).
38
-------
(ii) Column Chromatographic Clean—up — Further purification of the
extract is achieved by column chromatography on a Florisil column. Make a
slurry of 8 g of preactivated Florisil in methanol and transfer it into a
1.5 cm i.d. and 30 cm length glass column fitted with teflon stopper. Wash
the bed with 100 ml methanol and 100 ml 1:1 hexane—benzene mixture. Activate
the column (without the teflon stopper) by placing it in an oven for at least
four hours at 130°C. Following activation, cool the Florisil bed to room
temperature and wash it with 100 ml benzene. Transfer the PAR contain-
ing cyclohexane layer on the Florisil bed with the help of a Pasteur pipet.
Wash the round bottom flask thrice with 5 ml benzene and add the washings to
the Florisil column. Elute the PAR from the bed with 125 ml benzene at a
flow rate of 5 ml/min. and receive the eluate in a 200 ml round bottom flask.
Concentrate the eluate to about 2—3 ml by rotary evaporation and transfer the
contents with adequate washings to a calibrated tube. The benzene layer is
further concentrated to 0.1 ml by passing purified nitrogen and subjected to
quantitation.
Results and Discussion——
During the development of the clean—up method, considerable effort was
devoted to justify the necessity and verify the reliability of each step.
The extraction and partition steps which precede the chromatographic purif i—
cation of PAR, have been described by other authors (38) for different kinds
of material and appear largely empirical. With high resolution gas chroina—
tograph, Novotny, etal. (38) have shown that the chromatogram obtained from
airborne particulate samples were vastly dominated by alkane. Since water
samples should bear some resemblence with airborne particulate samples, a
selective enrichment of the PAR fraction was deemed necessary for water
samples as well. Since PAR are known to form charge—transfer complexes with
suitable compounds, DMSO was elected to obtain an aromatic hydrocarbon
enriched mixture by this method. The partition co—efficient for PAR between
cyclohexane and DMSO is high compared to other partitioning agents (19).
The isolation of the extracted PAR by dilution with water and back extraction
into cyclohexane prevented subsequent evaporation losses due both to vola-
tilization and thermal degradation. These characteristics render DMSO espec-
ially attractive. Acheson et al. (1) have shown that DMSO extraction effi-
ciency for the six PAR varies between 90—100%.
The chromatogram of the concentrate purified by solvent partitioning
alone is shown in Figure 11. It is evident that several impurity peaks are
present suggesting the necessity of further purification of the extract.
Column chromatographiC purification involving ticlassicalti adsorbents, such as,
silica and alumina have several disadvantages — (1) their adsorptivity often
resulting in losses of trace constituents and (ii) desorption properties
depending strongly on the amount of moisture in the column. Both these
factors cause irreproducible results. Gel permeation chromatography (38), on
the other hand, is a very time consuming process and is usually used for
fractionation of the PAR components in addition to their purification. Such
fractionation was found unnecessary in the present method. Chromatography on
a short Florisil column was tried for further purification of the PAR. A
remarkable visually observable clean—up of the sample is effected in this
step. As in all column chromatographic purification techniques, the selection
of proper eluting solvent and the volume of the eluent Is important to attain
39
-------
C
3
It
Figure 11. Purification efficiency of the clean—up procedure. Precleaned
foam plugs were exposed to drinking water, eluted with organic
solvent, eluate concentrated and subjected to GLC.
Solvent partitioned concentrate
after column rhrematographv on
tlortnjj, olotion volume 125 t!
After solvent partitioning
Use la ned
0 8
40
-------
desirable results. It was found that 125 ml benzene was r quir d to elute
all the 6 PAH from the Florisil column without eluting background contam-
inants (Fig. 11). The PAll recoveries from the complete clean—up step were
essentially quantitative (Fig. 12, Table 13) and the extract was sufficiently
purified to be readily amenable to GLC analysis. An increase in the eluent
volume resulted in elution of impurities from the column.
TABLE 13. RECOVERY OF THE OVERALL CLEAN—UP METHOD DETERMINED
BY GAS LIQUID CHROMATOGRAPHY
cpd.
amt. of
added in
std.
ig
amt. of std.
recovered in ig
%
recovery
FL
10.0
8.97
89.7
B(j+k)F
20.0
20.2
101.0
BaP
10.0
10.0
100.0
IP
10.0
8.93
89.3
BPR
10.0
9.15
91.5
The complete clean—up as used in further studies is illustrated by a
flow chart in Figure 13.
The clean—up procedure devised was able to eliminate contaminants to the
extent that they did not interfere with PAN peaks in gas liquid chromatography.
Since the volume of the extract usually subjected to thin layer chromatography
is much larger (see Section VIII, p. 52), it appeared likely that the impur-
ities may become visible and interfere with thin layer chromatography. Thus
the efficiency of the clean—up procedure was evaluated using thin layer
chromatography. The extract containing contaminants of water and foam origin
(60Z water, 12 foam plugs), was prepared as described before except that it
was not spiked with the standard mixture of PAll. The extract was subjected
to the clean—up procedure, concentrated to 0.1 ml and subjected to thin layer
chromatography. The chrotnatogram revealed from none to a maximum of 5 fluor-
escent spots depending upon the volume of the concentrate spotted. TLC of
the concentrate prepared from foam plugs without exposure to water showed
that 3 spots had originated from foam plugs (Fig. 14). The other spots were
identified to be the PAR concentrated from water (not detectable by GLC).
The solvents employed in elution and clean—up were not the source of the
impurities, since the concentrate prepared from solvents alone failed to
show these spots.
The fluorescent impurities eluted from the foam plugs were studied
further to determine if they will cause interference with the analysis of
41
-------
Figure 12.
C
I ’,
0
=
0
S
Recovery of PAIL in the clean—up process. Precleaned foam plugs
following exposure to drinking water were spiked with the PAIL
mixture. The foam plugs were eluted, eluate concentrated,
cleaned up and subjected to gas—liquid chromatography.
A. Standard PAR mixture, B. Eluate from water exposed and PAIL
spiked foam plugs after clean—up.
FL
B(j)F + B(k)F
B
BaP
BPR
FL
F
A
B aP
BPR
0 4 8 12 16 20 24
Tirne(non.?
28
42
-------
Concentrated foam extract (10 ml)
1.) Add 11 C BaP and dilute to 60 ml with cyclohexane
2.) Wash with 2 x 60 ml 4:1 methanol:water
Cyclohexane layer Methanol/Water Discard
Wash with 2 x 60 ml distilled water
Cyclohexane layer Aqueous layer Discard
Extract with 3 x 20 ml DMSO
P
Combined DMSO layer Cyclohexane layers Discard
1.) Add 120 ml distilled water
2.) Extract with 2 x 40 ml cyclohexane
Cyclohexane layers DMSO/Water Discard
1.) Pass it through anhydrous Na 9 SO,
2.) Concentrate to 5 ml & pass it t1 rough Florisil column
+
Elute with 125 ml benzene
Concentrate eluate to 0.1 ml and subject to quantitation
Figure 13. Flow chart of the clean—up method.
43
-------
0
U,
I
0 Origin
1st Development —
Figure 14. Thin—layer chroinatogram of 12 foam blanks. The signs inside the
spots indicate relative intensity: ++ moderate, + weak,
± very weak.
44
-------
6 PAH. This was accomplished by qualitating and quantitating the standard
PAR mixture by thin layer chromatography in the presence and absence of the
foam extract. Spot 1 (Fig. 14) did not interfere with any of the 6 PAR spots.
Of the other two spots, spot 2 remained unresolved with benzo(ghi)perylene
and spot 3 with fluoranthene. The fluorescence emission and excitation spec—
tra revealed that spot 2 was benzo(ghi)perylene. Similar studies of spot 3
revealed that it was a composite of an unidentified compound superimposed on
a spot recognized as fluoranthene.
The presence of trace amounts of PAR in foam plugs necessitated running
a foam blank with each batch of foam plugs to determine the background levels
of fluoranthene and benzo(ghi)perylene and correcting the values of these PAR
detected in water. The levels of PAR in the foam plugs used for the present
study are as follows:
Total amount (ng) detected in the concentrate
PAR prepared from 12 foam plugs
BPR 10.0
FL 70.0
The unidentified compound found close to the fluoranthene spot exhibited
its emission minima at the fluoranthene emission maxima and thus presented no
problem in quantitation of fluoranthene.
Overall PAll Detection Limit of the Method
Table 14 shows the detection limit for the 6 PAR compounds by gas liquid—
and thin layer chromatography. The detection limit for fluoranthene and
benzo(ghi)perylene by TLC is restricted by the background level of PAR intro-
duced from the foam plugs. Assuming that the detection limit is twice the
background level, the detection limit of these PAR has been derived. The
detection limit for the GLC method is based on a minimum output response of
five times the background values obtained at an output attenuation of 2 x 10
and a maximum sample loading volume of 5 p1 from a total of 100 p1 concentrate.
TABLE 14. DETECTION LIMIT FOR SIX PAR BY THIN LAYER- ARD GAS CHROMATOGRAPHY
PAH
TLC—fluorometric Detection
absolute limit limit in 60L
(ng) water (ppt)
absolute
(ng)
GLC—FID
limit
Detection
limit in 60
water (ppt)
FL
140 2.3
13.6
4.5
BjF
BkF
7.5 0.1
5.0 0.1
10.1
3.4
BaP
10.0 0.2
11.9
4.0
IP
10.0 0.2
14.7
4.9
BPR
20.0 0.3
14.9
5.0
45
-------
SECTION VII
EVALUATION OF FOAM PLUGS FOR COLLECTION EFFICIENCIES OF SIX PAR
It is demonstrated in Section V that polyurethane foam plugs effectively
retain trace quantities of benzo(a)pyrene from water. The extension of the
method to other representatives of the polynuclear family requires determin-
ation of the collection efficiencies of those PAR This section deals with
the procedure and results concerning the collection efficiencies of six in-
dividual PAIl from a representative treated water, and a raw water represent-
ing the worst case of raw drinking water source.
Initially, the collection efficiency of the foam plugs was evaluated at
PAR concentrations which could be detected by GLC. Experiments were later
conducted to determine collection efficiency at lower PAR concentrations and
concommitantly large sample volumes. The detection of lower PAR concentra-
tion required analysis by thin—layer chromatography and fluorometric detection.
COLLECTION EFFICIENCY OF SIX PAR EVALUATED BY GAS-LIQUID CHROMATOGRAPHY
These studies were carried out with 4 liters of tap water which had been
spiked with the standard mixture of six PAH to give a concentration of 25 ppb
for each PAR. The water was drawn over a prewashed foam plug maintaining
water temperature at 62 ± 2°C and flow rate at 250 ± 10 ml/min. PAR were
eluted from the foam plug using the procedure described earlier (Section V,
p. 10). In order to account for the PAR adsorbed to the walls of the reser-
voir, the reservoir was washed with benzene—acetone and the washings were
combined with the foam eluate. The combined extract was concentrated and
subjected to gas—liquid chromatography. Purification of the concentrate was
found to be unnecessary because of the presence of high concentrations of PAR
and low levels of impurities. The background levels of PAR in the tap water
were non—detectable by gas chromatographic method and thus did not affect the
results.
Gas chromatogram of the PAR added to water and those recovered from the
combined foam and bottle extract are shown in Figure 15. Collection eff 1—
ciency for each PAR is calculated by comparing the two peak areas. The
results shown in Table 15, confirm that polyurethane foam plugs under suitable
conditions not only effectively concentrate benzo(a)pyrene but other PAR
as well.
46
-------
a
C
0
a
Time(n n.)
B
28
Figure 15. Retention of 6 PAll on polyurethane foam determined by gas
chromatography. A. PAR added to water; B. PAR retained on
foam plugs.
FL
B(j F * 8(k)F
BPR
BaP
FL
BaP
A
BPR
‘P
0 4 8 12 16 20 24
47
-------
TABLE 15: FOAM RETENTION EFFICIENCIES OF SIX PAR FROM
TREATED WATER. WATER SOURCE: LABORATORY TAP
WATER; VOLUME: 4i; CONCN. OF EACH PM!: 25 ppb
Compound % Retention
FL 100
BJF. 88
BkF f
BaP 81
IP 89
BPR 91
COLLECTION EFFICIENCIES OF SIX PM! EVALUATED BY ThIN-LAYER CHROMATOGRAPHY-
FLUOROMETRY
The concentration of PAll in treated water can be expected to be low and
thus it was considered important to evaluate the ability of foam plugs to
concentrate six representative PAR from large sample volumes at low concen-
trations. The studies were carried out using thin—layer chromatography—
fluorometry as the method of PAR analysis. A concentration of 100 ppt for
each PM! (except for fluoranthene which was 500 ppt) and sample size of 6O2
for treated water and 3O for raw water was used.
Procedure
As treated water source, laboratory tap water was used. Onondaga Lake
water was used as an example of a worst possible raw water source (see
Section V, p.21). Spike tap or raw water taken in a twelve gallon pyrex
glass bottle with PAR standard mixture to a concentration of 500 ppt for
fluoranthene, and 100 ppt for all others. Mix the solution well for about
an hour with a Teflon magnetic stirrer.
As shown earlier (Section V, p.20), a large portion of the added PM!
becomes adsorbed to the walls of the reservoir resulting in a much lower con-
centration of the individual PM! in the aqueous phase. Since it is impossible
to quantitate the loss of PM! by adsorption to 12 gallon glass bottles, the
actual concentration of PM! in the aqueous phase was determined by extracting
an aliquot of the spiked sample.
48
-------
Two one—liter aliquots of spiked water were withdrawn, one at the begin-
ning and the other towards the end of the sampling procedure. Each was
extracted with 100 ml cyclohexane. The extracts were combined and concentrated
to 0.1 ml. Following clean—up, the extracts were subjected to thin—layer
chromatography—fluorometric analysis for separation and quantitation of the
six PAH. Combining of the two aliquots produced average values of the initial
PA R concentration in water.
The spiked water was drawn over four plug system consisting of two pre—
cleaned foam plugs in two different columns. In the case of finished water,
both columns were changed after every 201 of water, whereas with raw water,
the columns were changed after every 101. Thus, 12 foam columns were required
for concentrating PAH from 601 of tap water or 301 of raw water. Foam plugs
are then eluted in the usual manner, the eluate concentrated and subjected to
the full clean—up procedure, and analyzed by thin—layer chromatography—f luoro—
metry as described earlier.
Results and Discussion
The present method utilizes a direct determination of initial concentra-
tion of the PAR in the spiked water phase. This not only takes care of the
problem of reduction in concentration due to different kinds of losses, but
eliminates the necessity of determination of the background PAR concentrations
in the unspiked water samples. The amount of each PAR actually added and
that recovered from the aqueous phase is shown in Table 16. The losses of
PAN due to adsorption to reservoir surface appear to be somewhat related to
the molecular wt. of the compound, for example, in case of BPR (LW., 274),
as much as 77% of the added amount remained adsorbed, whereas only 44% of
FL (M.W. 202) was lost due to adsorption. An increase in the adsorption of
BaP in these studies compared to the results reported in Section V (p. 20).
can be attributed to increase in the size of the reservoir.
TABLE 16. AMOUNT OF 6 PAR UNACCOUNTED FOR AS A RESULT OF MIXING
WITH WATER IN A GLASS BOTTLE. WATER VOLUME: 601,
BOTTLE CAPACITY: 12 GALLONS
Concn. on the
Co
ncn. found
% P
AR adsorbed
Molecular
basis of
amt.
in
aq. phase
to
reservoir
weight
Compound
added to
water
(ng/ 1)
surface
FL
(ng/i)
202
500
278.6
44.3
BjF
100
48.3
51.7
252
BkF
100
51.7
48.3
252
BaP
100
36.4
63.6
252
1P
100
25.5
74.5
276
BPR
100
22.6
77.4
274
The retention efficiencies of the 6 PAll from spiked laboratory tap water
and Onondaga Lake water are shown respectively in Tables 17 and 18. From
both these waters polyurethane foam plugs concentrated PAR almost quantita-
tively. The efficiency of retention will actually be somewhat higher because
49
-------
some loss of the PAR sorbed on polyurethane foam plugs occurs during the
elution and clean—up (see Section VI, p.41) and such loss has not been
corrected for in the data illustrated in the tables.
TABLE 17. FOAM RETENTION EFFICIENCIES OF SIX PAR FROM TREATED WATER,
WATER SOURCE: LABORATORY TAP WATER; WATER VOLUME: 60i;
CONCN. OF FLUORANTHENE: 500 ppt; ALL OTHERS, 100 ppt
Compound
FL
Concn. present Amt. retained by
in water foam from a liter
(ng/Z) of water (ng)
% Retention
278.6 260.4
93.5
BjF
48.3 47.4
98.1
BkF
51.7 50.6
97.9
BaP
36.4 33.6
92.3
IP
25.5 23.9
93.7
BPR
22.6 19.8
87.6
TABLE
18. FOAM RETENTION EFFICIENCIES OF SIX PAll FROM
WATER SOURCE: ONONDAGA LAKE; WATER VOLUME:
OF FLUORANTHENE: 500 ppt; ALL OThERS, 100
RAW WATER.
3O9 ; CONCN.
ppt
Compound
Concn. present Amt. retained by
in water foam from a liter
% Retention
FL
(ng/Z) of water (ng)
578.1 687.5
118.9
BjF
77.6 94.0
121.1
BkF
66.1 55.6
84.1
BaP
74.5 59.7
80.1
IP
85.2 61.2
71.8
BPR
23.9 28.3
118.4
The retention data given in the tables is for 0.5 ppb for fluoranthene and
0.1 ppb for others. It was not possible to undertake recovery studies with
the mixture of 6 PAM at concentrations lower than this for the following
reasons. The susceptibility of PAR to be lost by adsorption to reservoir
and other surfaces, necessitated determination of the concentration of each
PAM in the spiked sample. This involved extraction of PAll from an aliquot of
water and their quantitation. At PAM concentration lower than 0.1 ppb, the
amount extracted will be difficult to detect accurately. Experiments with
radio—labelled BaP described in Section V (p.l 8 ) showed that at concentration
as low as 2 ppt, foam plugs were able to retain BaP with 87% efficiency. In
view of these findings, it is not unlikely that other PAR will also be
efficiently retained at such low concentrations.
50
-------
CONCLUSION
The polyurethane foam retention efficiencies of the 6 PAIl from drink-
ing water have been investigated and found to be almost quantitative. Studies
conducted with a worst possible raw water for a likely source of drinking
water indicate that the retention values for six PAIl are quite high. It is
concluded that polyurethane foam plugs can be used as a preconcentration
method for all of the six PAH from large volumes of both treated and raw
water.
51
-------
SECTION VIII
FIELD MONITORING
The results reported in earlier sections clearly show the potential of
polyurethane foam plugs as a convenient method for concentration of PAR from
treated and raw water. The next step was to apply this method under field
conditions to demonstrate the feasibility of maintaining optimum conditions.
Initially, efforts were devoted to fabricate a portable sampling unit which
would allow maintaining of temperature and flow rate of water optimum for
PAN retention. After evaluation of the unit in the laboratory, sampling of
ten drinking water supplies in the Eastern U.S. was undertaken.
FABRICATION OF SAMPLING UNIT FOR FIELD MONITORING
A field sampling equipment was assembled from a combination of custom—
made and off—the—shelf hardware available commercially. It consisted of the
following five parts as shown in Figure 16: (1) variable water pumping unit,
(2) thermostated water circulator, (3) unit containing the column system,
(4) temperature monitoring device and (5) flowmeter. The sources of all
commercially available hardware are described in Sections V and VI.
The purpose of the pumping unit was to pump water from the sample source
through the foam columns at a controlled rate. The unit consisted of two
oscillating type pumps connected in series with a minimum amount of thick—
walled tygon tubing. The electrical input of the pumps was applied through
a Variac. By controlling the output from the Variac, the water pumping rate
was controlled at 250 ± 10 ml/min. Haake thermostated circulator with a
custom—made glass coil (25 cm x 6 mm) immersed inside the reservoir housing
through the cover plate was used to bring the water to 62 ± 2°C. One end of
the coil was connected to the sampling source and the other end to the column
system. The temperature of the water passing through the coil was controlled
by means of a 1000W heating element and the thermostating device of the
Haake circulator. Two 25 mm i.d. Chromaflex extender type columns were
mounted on two column stands which were held on wooden blocks. The columns
were connected In series by means of a custom—made double—bent Pyrex tubing
to each end of which was attached an adapter. The adapter and Chromaflex
columns were connected by means of “0” rings and clamps. For continuously
monitoring the temperature of water during passage through the columns, a
water trap equipped with a right angle thermometer and an inlet and outlet
was introduced in the setup. The inlet was connected to the exit end of the
column system and the outlet to the pumping unit. A Gilmont flowmeter cali-
brated for measuring flow rate of 62°C water was placed at the end of the
sampling system.
52
-------
Figure 16. Portable unit assembled for concentrating PA}I from water in
the field.
-------
The column stands, pumps, Variac, water trap and flow meter were firmly
fixed on an 18” x 12” x 1/2” plywood piece as shown in Figure 16. The con-
nection up to the foam columns were made as much as possible with glass
tubing and only where necessary with tygon tubing. The type of tubing used
for connections beyond this point was immaterial. The overall connection was
such that water travelled through the sampling system in the order: sampling
source - - thermostated circulator ± Chromaflex columns - water trap - pumps -÷
flowmeter. This whole unit weighing about 10 lbs. was transported to the
sampling site along with the thermostated circulator, an electric timer, a
202 . capacity Jerrican and other detachable items, such as Chromaflex columns,
thermometers, clamps, “0” rings and connecting tubes. The purpose of the
calibrated 2O9 Jerrican was to collect effluent from the sampler to obtain the
volume of water passed through the foam columns. The use of an electric
timer for running the unit for appropriate amount of time was optional.
DETERMINATION OF PAR IN SELECTED WATER SUPPLIES
The new method developed was applied to monitoring PAR in ten water
supplies in the eastern United States. In some instances, raw intake waters
as well as finished waters were examined and the information on the levels
and nature of PAR was related to the raw water source, waste/discharges con-
taminating raw water, and treatment provided. Our sample size was 602. in
the case of finished water and 3OP in the case of raw water. It was assumed
that levels of PAR in raw water will be high and, therefore, concentration
of 302. will be adequate.
Selection of Sampling Sites
In selecting sampling locations, several factors which are expected to
have an impact on the levels of PAR in finished waters were taken into con-
sideration. These included category of raw water source, waste/discharges
entering raw waters, treatment processes provided, etc. The quality of
surface waters is dependent on the various sources of pollution that they are
subjected to. The selection of sampling sites using surface waters as a raw
water source was, therefore, based on the nature of the discharges entering
the surface water. Ground water in general is expected to be relatively free
of pollution and this prompted inclusion of a water supply using ground water
as a raw water source. Several sites were selected which use activated carbon
as a treatment process. EPA research (32) indicates that activated carbon
removes general organic compounds before attaining its breakthrough. Compar-
ison of the level of PAR in raw and treated water at the above sites, and
with the data on the water supply systems which do not use activated carbon
treatment, may provide an insight into the effectiveness of the treatment
process in removing PAR. Also included in the selected sites are water
supplies of 5 major urban centers, each of which serve a large number of
consumers.
Treated drinking waters were sampled for all the 10 water supplies but
only Pittsburgh, Huntington, Buffalo and Philadelphia were sampled for raw
intake waters. Non—availability of the raw intake water source at the other
treatment plants forced the exclusion of raw water sampling at these sites.
All the monitoring work except the New Orleans sampling was conducted during
54
-------
the period beginning December 1976 through March 1977, and samples were taken
at the treatment distribution sites. New Orleans sampling was done in a motel
in the old part of the city in the month of May, 1977.
Table 19 identifies the raw water sources, type of waste/discharge (if
any) entering the raw waters, treatment provided and date of sampling for the
water supplies selected for this study.
Sampling Procedure
The sampling unit, 6 precleaned foam columns for finished water and 6
for raw water (if applicable) and two empty Chromaflex columns were trans-
ported to the sampling site via automobile. The unit was assembled with two
empty columns on a counter top or table near the water source. A clean, one
liter beaker was placed under the water tap and water from the beaker was
pumped through the unit at a flow rate of 250 ± 10 mi/mm. The thermostat in
the Haake Circulator was adjusted until the temperature of the flowing water
was at 62 + 2°C. The setting will vary depending upon the temperature of the
water to be sampled and, therefore, it cannot be set in the laboratory.
Following this adjustment, the empty Chromaflex columns were replaced with
columns packed with foam plugs and 2O2 of finished water or lO2 of raw water
was passed over, maintaining the flow rate and temperature as described above.
The volume of the water passed was measured by collecting effluent from the
sampler in a graduated Jerrican. Both foam columns were changed every 209 in
the case of finished water and every l0i in the case of raw water and the
sampling was continued to the desired volume of water. The columns were
brought to the laboratory for analysis following sampling. No special hand-
ling and storage except wrapping of the columns with aluminum foil was nec-
essary if the transit period was less than 2 days, otherwise the wrapped foam
columns were cooled with reusable ice packs in styrofoam containers.
PAH Elution and Analysis
The procedure for elution of PAR from foam plugs, clean—up and quantita—
tion has been described in earlier sections. A flow chart of the entire
procedure including sampling and shipping is shown in Figure 17.
Addition of Internal Standard and Determination of Recovery Factor
The efficiency of the PAR elution and purification procedure for each
analysis was evaluated by addition of an internal standard to the foam plugs
upon arrival in the laboratory and prior to initiating any sample work—up.
Carbon—l4 labeled benzo(a)pyrene was employed as internal standard. A known
amount of 14 C (approximately 600 dpm) was added to foam plugs, PA l- I eluted,
purified and concentrated as described before. A 10 p1 aliquot of the con-
centrate was assayed for radioactivity in a liquid scintillation system (see
Section V, p.10). Knowing the 1 - 4 C—activity originally added to the foam
plugs allows calculation of the recovery factor.
Results and Discussion
The results of analysis of ten treated and four raw water samples are
presented in Table 20. No data regarding the precision and accuracy of the
55
-------
U,
TABLE 19. DETAILS OF ThE WATER SUPPLY SYSTEMS USED FOR SAMPLING
Type of
*
locat ion
inppiy System
Water S,,,r,
P c lot on,
1, ., , t mr’, Pr, ’v Ided
Do (s)
ii
Sampled
Syracuse, N.Y
City of Syracuse
Water Works,
Skaneateles, NY
Lake
Skaneateles
ilnc cntaninaI
lake walor
Copper sal fate addit ion, hl mat in amd
fluoridation
1—16—76
Hutfile, N.Y.
Ward’s
Pumping
Station
Lake Erie
ilononga—
bela River
Contaminated with
Ind ostrial
discharge
Contaminated with
coke oven effluent
Coagulation, activated carbon addition,
chlorination and fluoridation
l.ime, lerric sulfate addition, activated
carbon addition (two stages: (1) powdered
carbon (2) granular carbon), chlorination
and fluoridation.
l2—ct ,1 27 - 7 6
1—19—77
Pittsburgh, Pa.
Hays Mine and
EM. Aldrich
Purification
Station
Huntington,
Huntington
Ohio River
Downstream from
Lime, ferric sulfate and granu1a carbon
1—20—77
W. Va.
Water Corp.
coke oven plants
addition, chlorination and fluoridation.
Fndicott, N.Y.
Endicott
Village, Dept.
of Public
Works
Ground water
Keuka lake
Uncontaminated
ground water
Contaminated with
Chiortnat ton and fluoridation.
Chlorination
1 — 1 1 /7
2—28—7’
Hamtnondsport,
Hantmondsport
N.Y.
Village, Dept.
of Public
agricultural and
vinery waste
Works
Philadelphia,
Torresdale
Delaware
Contaminated with
Ferric chloride, lime, activated carbon
3—S/6—77
Pa.
Water
Treatment
Plant
River
municipal waste
Uncontaminated
ammonia addition, . hiorination and
fluoridat Ion.
Copper sulfate addition, aeration,
3—17—77
New York,
Dept. of
Croton
N.Y.
Waist
Resources
Reservoir
upland water
Contamination fron
corrosion control, chlorination and
fluoridation.
i—26—77
Chlorination
Lake George,
Lake George
Lake George
N.Y.
Village,
Dept. of
Public Works
recreational sources
New Orleans,
Tap Water from
Mississippi
Downstream from
Chemical treatment to control alkalinity,
5—1—77
La.
a Motel in
Downtown New
Orleans
River
industries on
Mississippi River
hardness and organics; coagulation,
ammonia addition, and chlorination.
* Not necessarily in the proper order ci treatment, & filtration steps wherevet used during the treatment are not shown.
Used temporarily
-------
PM in raw and
N -
rinished water
U,
to
U
Concentrate PAR ut
site with prewashed
polyurethane foam plugs
Ship foam plugs
to the laboratory
Spike foam plugs with bonus amouot
of C—BaP (approo 600 Ape)
as internal standart.
Elate with acetone—
cyclohexane
Clean—up
(a) partitioning with solvents
(b) column chromatography n
Florisil
Extract concentrated
to 100 -
Liquid Sciotillatioe Gas—liquid chromatography—PlO 1 TWO dimens)onal TLC on
counting, 10 i Colams,: 3% Dec.11 300 on alumina—cellulose acetate
aliquot in 10 m l chroeosorb C coupled with fluorometric
L scintillation fluid Inject 2—S -, assay .aith mc ocanoer.
___________ I Amount spotted: In accordance
with GLC—oalculated results.
Retention J Peak area ] Relative position Pluoroscence Excitation I resceoce intensity
time j of spots. Flue’— Spectra Spectra measurement
I rec m ti escence color. —J / - —
Identiuin.tion F yuanritarion
ect r;
Figure 17. Flow chart of the method of PM! analysis in water.
57
-------
results can be given since a single analysis for each sample was performed.
Caution must be exercised in making interpretations from these results. The
recovery factors given in the tables are derived from the recovery of the
added radioactive benzo(a)pyrene internal standard after the full clean—up
procedure. The values of the six PAR determined have been corrected using
this recovery factor. It has been assumed that the recovery of all the six
PAR will be the same as benzo(a)pyrene. The values given for fluoranthene
and benzo(ghi)perylene have been corrected for the trace amounts of these PAH
contributed by the foam plugs. Benzo(a)pyrene values have been corrected for
the amount of - 4 C—benzo(a)pyrene added as internal standard. Correction has
not been made for the retention efficiency of the foams which has been
assumed to be 100%. Since in reality, the retention efficiency of foam plugs
will always be lower than 100%, the actual amount of PAR in water should be
slightly higher than shown in the table.
The BaP used in this investigation has a specific activity of 8.1 sic!
millimole. The addition of 600 dpm as internal standard will amount to
8.4 ng increase in the amount of BaP concentrated from water. Since a tenth
of the total final concentrate is used for radioactive counting, this will
amount to 60 dpm which is equivalent to 48 cpm with counting efficiency of
80%. A substantial change in the amount of 14 C added may present problems.
Lowering the amount of added radioactivity will result in statistically
insignificant difference between the counts determined and background counts.
On the other hand, any substantial increase in the amount of radioactivity
and hence BaP, might make the detection of the low levels PAH from water
inaccurate.
The elution and clean—up efficiency for the treated and raw water samples
average 80% and 69%, respectively. The lower average value In the case of
raw water is not unexpected since the recovery factor is dependent upon the
quality of water samples. However, evaluation of individual recovery factor
eliminated this uncertainty in the reported values.
The final concentration of the PAR presented in Table 20 is based
exclusively on the TLC—spectrofluorometric value. The GLC—FID method in most
cases failed to detect the PAR. Furthermore, the concentrations of PAH in
water derived from the GLC values are susceptible to an error due to large
multiplication factor since only 2 p1 volume is injected out of a total of
100 p1 concentrate. In spite of this, whenever the GLC values could be
obtained, they served as a crosscheck of the TLC—value.
PAR were detected in the ppt range in both raw—and—finished water at all
the locations sampled. While the concentrations of PAR (Sum of the 6 PAR) in
drinking waters were small, the values found in raw water were as high as
600 ppt. In many cities, all the six representatives of the PAN family were
detected. The polynuclear compounds — benzo(a)pyrene, benzo(ghi)perylene and
indeno(l,2,3—cd)pyrene were among the most frequently occurring PAR (Table 21).
Of interest was the finding that fluoranthene, a PAIl with relatively higher
water solubility (265 ppb at 25°C compared to 10 ppt for BaP, ref. 20,22), was
not widely detected. The sum of the six PAR in the finished waters analyzed
ranged from 1—15 ppt which are well below the World Health Organization’s
recommended upper limit of 200 ppt. Water samples derived from Buffalo with
58
-------
TABLE 20. RESULTS OF ANALYSES OF FIELD SAMPLES
Data presented below is from analysis of a single sample from the selected location
Location and
Type of
Water
Compound
Total Aist.
Detected by
TtC (ug)
Total Aint.
Detected by
GLC (ng)
Conc. in
Water by
TLC (ppt)
Syracuse
Finished
FL
S.D.
N.D.
S.D.
BIF
S.D.
N.D.
S.D.
BkF
21.5
ND.
0.4
BaP
16.3
N.D.
0.3
IP
S.D.
S.D.
S.D.
BPR
2 .0
ND.
0.4
Total of 6 PAR
1.1
Recovety factor:
0.80
FL
N.D.
S.D.
S.D.
BjF
N.D.
S.D.
S.D.
BkF
S.D.
N.D.
S.D.
BaP
9.6
S.D.
0.2
IF
S.D.
S.D.
S.D.
BPR
40.1
S.D.
0.7
Total of 6 PA l- I
0.9
Buffalo
Finished
Raw
Pittsburgh
Finished
Raw
Recovery
factor:
0.81
FL
S.D.
S.D.
S.D.
B1F
S.D.
S.D.
S.D.
BkF
17.1
S.D.
0.6
BaP
7.6
S.D.
0.3
IP
S.D.
S.D.
S.D.
BPR
112.8
S.D.
3.8
Total of
6 PAR
4.7
Recovery
factor:
0.61
FL
S.D.
S.D.
S.D.
BjF
B1CF
18.0
11.0
S.D.
S.D.
0.3
0.2
gaP
22.1
N.D.
0.4
IF
73.3
S.D.
1.2
BPR
43.6
S.D.
0.7
Total of 6
PAll
2.8
Recovery factor:
0.82
FL
12250.0
12179.1
408.3
BjF
BWF
1070.0
572.5
l507.5
19.1
BaP
1262.9
2313.4
42.1
IF
1812.5
2417.9
60.4
BPR
1032.5
1492.5
34.4
Total of 6
FAIl
600.0
Recovery factor:
0.67
continued
59
-------
Table 20. (cont’d) Results of Analyses of Field Samples
Location and Total Arnt. Total Amt. Conc. in
Type of Detected by Detected by Water by
Water Compound TLC (ng) GLC (ng) TLC (ppt)
Huntington
Finished FL 144.8 N.D. 2.4
BjF 20.5 N.D. 0.3
BkF 11.0 N.D. 0.2
BaP 27.4 N.D. 0.5
IP 72.6 N.D. 1.2
BPR 147.5 N.D. 2.5
Total of 6 PAE 7.1
Recovery factor: 0.85
Raw FL 704.9 1140.0 23.5
BjF 150.3 N.D. 5.0
BkF 109.2 N.D. 3.6
BaP 169.2 N.D. 5.6
iP 285.1 N.D. 9.5
BPR 322.2 N.D. 10.7
Total of 6 PAH 57.9
Recovery factor: 0.75
Endicott
Finished
FL
259.4
331.5
4.3
BjF
9.7
N.D.
0.2
BkF
N.D.
N.D.
N.D.
SaP
13.7
N.D.
0.2
IP
42.6
N.D.
0.7
BPR
171.3
N.D.
2.9
Total
of 6 PAH
8.3
Recovery factor:
0.89
Hamrnondsport
Finished FL S.D. S.D. N.D.
SjF 17.0 N.D. 0.3
BkF 8.3 N.D. 0.1
SaP 17.9 N.D. 0.3
IP 56.7 S.D. 0.9
BPR 115.8 S.D. 1.9
Total of 6 PAIl 3.5
Recovery factor: 0.83
continued
60
-------
Table 20 (cout’dl Results of Analyses of Field Samples
Location and Total Anti. Total tort. Coot, in
Type of Detected by Detected by Water by
Cater Compound TLC (sg) GLC (rtg) TLC (ppt)
adelhia
Finished FL 532.6 541,1 8.9
BjF S.D. 9.0. ND.
B6F ND. ND. 0.0.
NaP 17.4 *
IF 103.6 ND. 1.7
SF8 237.4 S .D. L.A
Total of 6 FAN 13.9
Recovery actor: 0.73
* the ANT value was slightly different than RaP
suggestIng a mixture of Ba? with name other compound.
The amount quantituted using Ba? standard was 1692 sg.
Row FL 3450.0 3743.1 114. i
B3F 1277.5 2138.9
Bk? 990.0 33.0
OaF 1232.9 * 41.
iF 2172.5 3166.7 72.4
SF8 1452.5 1986.1 48...
Total of 6 PAN 351.8
Reccmerv factor: 0.72
*The RRT oa ue was sltghtlv differerr rhuo Rat’
suggesting a mixture of BaF wltlt some other compound.
The amount nuantitated using BaF standard ttas 7279 ng.
N ew Fork C
Fittished
FL
0.0.
5.0.
N.y.
8 3 F
Bk?
70.2
39.1
8.0.
S.D.
1 .7
0.
NaP
32.5
S.D.
0.5
IF
130.5
S.D.
2.2
SF8
106.8
5.4.
1.8
Total of
6 PAN
6.4
Recovery
factor:
0.79
eere
Fittished
FL
9.0.
S.D.
8.0.
BjF
Bk?
SaP
IF
ANN
Total of
Recovery
6 PAN
factor:
20.6
8.0
16.4
54.5
158.4
0.66
5.0.
N.D.
S.D.
8.0.
S.D.
0.3
0.1
0.3
0.9
2.6
s.2
New Orleas
Finished?
FL
BJF
BkF
Ba?
IF
BPR
Total of
6 PAIl
so.
S.D.
S.D.
37.8
97.0
S.D.
S.D.
S.D.
S.D.
ND.
S.D.
ND.
S.D.
S.D.
0.6
1.6
ND.
2.2
Recovery
B Water
factor: 0.66
sampled at a motel in
old section of New Orleans.
N.D. t Not Detected
7 : Values subjected to error because the sorbent layer on
developed plates was occidentally scraped.
61
-------
TABLE 21. FREQUENCY OF OCCURRENCE OF VARIOUS PAR IN TEN DRINKING WATERS EXAMINED
PAR
Number of
waters sho
detectable
of PAR
drinking
wing the
levels
% Drinking waters showing
the presence of each PAR
FL
3
30
BaP
10
100
BkF
7
70
BJF
6
60
IP
7
70
BPR
9
90
TABLE 22. UNIDENTIFIED LUMINESCENT SPOTS FROM EACH SAMPLE
Type of sample
Total No
unidenti
spots
. of
fled
No. of unidentified
spots with substantial
luminescence
Syracuse finished water
1
None
Buffalo finished water
1
None
Buffalo raw water
3
1
Pittsburgh finished water
3
1
Pittsburgh raw water
5
1
Huntington finished water
2
1
Huntington raw water
7
4
Endicott finished water
3
2
Hammondsport finished water
2
1
Philadelphia finished water
2
2
Philadelphia raw water
5
2
New York finished water
6
5
Lake George finished water
5
3
New Orleans finished water
4
3
62
-------
Lake Erie as the source showed surprisingly lower level of PAH. It is
difficult to explain the reason for the lower level of PAll at this location
than those determined in ground water at Endicott. In this regard it should
be mentioned that the Buffalo samples were collected during the severe
snowstorm period in that area.
Figure 18 represents the spectra of PAll identified in selected water
supplies along with the spectra of the model compounds. In each case, the
solid line represents the model compound, whereas the dashed lines represent
the TLC spot having the same RB value and fluorescence color as the model
compound. The superimposed spectra for most samples were significantly
similar. The absence of the fine structure compared to the model spectra
noted in some instances may be due to low extinction coefficient of these
line(s) and low concentration of the compound in the spot. This was proven
with model compounds with succeedingly lower concentrations. The fluorescence
excitation and emission peaks were not obliterated or distorted confirming
the presence of a single compound in each spot. The fact that the fine
structure did not even flatten out is suggestive of the absence of any
unseparated alkylated compounds in the spots. It should be pointed out that
in quantitating the results by fluorescence, the error will be comparatively
higher when the compounds possess low fluorescence quantum yield as in the
case of IP.
In the Philadelphia raw—and—finished water samples, the GLC peak
corresponding to benzo(a)pyrene was slightly shifted from the RRT value of
the model compound. This is an indication of the presence of a mixture of
compounds with very close RRT values and the inability of the GLC column to
accomplish the proper separation. Consequently, the amounts of benzo(a)pyrene
calculated from GLC values, are erroneous. The separation of this mixture
was complete on the TLC plate and the values derived from spectrofluorometric
analysis are reliable.
In addition to the 6 PAH identified and quantitated, a number of other
unidentified luminescent spots appeared on the TLC plates of various samples.
Table 22 lists the total number of unidentified luminescent spots obtained
from each sample. Spectrofluorometric identification of these spots could
not be made because no matching spectra of known compounds could be located.
Examination of the fluorescence emission— and excitation spectra of the
unknowns with substantial luminescence revealed that in many instances the
same compound was present in many waters. Fluorescence emission— and excita-
tion spectra of selected unknown compounds are given in Figure 19. Selected
spots were scrapped off from the plates and subjected to electron impact mass
spectroscopic analysis. The results were, however, inconclusive which was
attributed to the presence of a large number of background impurities.
Either the samples were contaminated during scrapping or impurities were
picked up from TLC plate coating during solvent elution. Further efforts to
identify these compounds could not be carried Out because of the nonavail—
ability of more samples.
The data on the levels of PAM in finished and raw water show that a con-
siderable reduction in the concentration of PAM occurs during water treatment.
It is unclear if the decrease is due to actual removal, or transformation of
PAM to other products. The treatment process at Hays Mine Treatment Plant at
63
-------
[ TT1T -t T r- r- i
— I I
400
I,
f’J
‘P
/
i’
300 400 460 500 460
200 250 300 350 400
r
BP S
‘ill ,j
L_J_ J_ I - _L i - L I
200 250 300 360 400 360 400 450 500 460
Figure 18. Fluorescence emission and excitation spectra of model PAR and
those identif led in Huntington, W. Va. water samples.
64
I SaP
Ii ,
II
I I I
360 100 450 500 500000
— S I
200 250 300 350
I J i
350 400 450 500
550 000
250 250 300 360
fl ’ -Y T
5 1 6S F
[ \ rj
H
I /
il l
I ; ‘
I I
/ I!
I- J
360
I S I I -
360 100 150 500 560
-------
1
A -_ -
200 250 300 350 400 450 500
7
0
I
500 250 300 350 400 450 500 550 600
W . .60n 5thI n4’
I
//
I
—__ I I A
200 250 300 000 4.30 050 x 560 250 300 350 400 450 500 550 400
L : 1
I A 1
200 250 300 350 400 450 500 300 040
4,. Detected in drtnkiflg voter at Philadelphia
Laic Georga, Pitt .bcrgh. Hwnt ington.
4 a ndBp0rt, tndicott. and in raw voted
Ot Huntington and Philadelpt3id
Deterted In drinking water at PhiLadelphia.
Lake George, New Tori. ittabsorgh, Huatingt ’fl,
and raw voter at Philadelphia.
C. Detected in ran water at 4,ntingtOo
Det ctad is, raw water at Hcntingt on.
Detected in drinking water at )*unttngtoo.
‘4ew York and Suffalo.
)ecectad In raw ta or at Pitt sb urgh.
Detected in raw votes at Philadelphia
Figure 19. Fluorescence emission and excitation spectra of selected unknown
compounds.
65
-------
Pittsburgh appeared to have the highest PAll removal/transformation efficiency
(Table 23 A). This may be attributed to the fact that the water treatment at
this plant involved two stages of activated carbon treatment: 1st stage—
powdered carbon; 2nd stage—granular carbon, whereas treatment at other plants
involved only one stage activated carbon treatment. Comparison of the removal!
transformation efficiency of individual PAH suggested that fluoranthene.
indeno(l,2,3,—cd)pyrene and benzo(ghi)perylene were not as effectively
removed/transformed as benzo(a)pyrene, and benzo(j)— and (k)fluoranthene
(Table 23 B). No correlation between the efficiency of removal/transforma-
tion and molecular weight of compound was evident. Correlation between
removal/transformation efficiency and water solubility could not be made
because solubility data for all the PAR were not available.
Conclusions
The sampling method developed f or the collection of six PAM from raw
and finished waters and the analytical method used for quantitation have been
successfully employed for field monitoring. It provides an excellent routine
method for the analysis of PM!. The recovery of the six PAN by the
method is almost quantitative. The addition of 1 - 4 C—BaP as internal standard
to the foam plugs upon arrival in the laboratory allows determination of the
recovery factor for each analysis and thus account for any losses of PAM
during elution, clean—up and analysis. Whenever possible, quantitation of
each sample should be performed both by GLC-FID and TLC—spectrofluorometric
method. This will provide a crosscheck of the values derived. However, in
samples containing PAN below the detection limit of FID, only TLC—spectro--
fluorometric method can be applied for quantitation. This method is capable
of detecting PAR at sub—ppt levels from both raw and treated drinking
waters. The detection limit of the method can be further improved by increas-
ing the sample volume.
With the application of this method, six representatives of the poly—
nuclear family and several unknown compounds were detected in ppt range at
all the water supplies sampled. The levels detected were well below the
World Health Organization’s recommended limit of 200 ppt. Health hazards to
man from the presence of such low levels of PM! in drinking water are not
clearly understood. A considerable reduction in the concentration of PAH was
noted as a result of water treatment. It is unclear if PAH are actually
removed, deactivated or transformed to more carcinogenic product.
-------
TABLE 23. EFFICIENCY OF REMOVAL/TRANSFORMATION OF PAR IN WATER TREATMENT
A. REMOVAL/TRANSFORMATION EFFICIENCY OF VARIOUS TREATMENT PLANTS
Total concn. of the
Sampling 6 PAR (ppt ) % Removal!
location transformation
Raw water Finished water
Pittsburgh, Pa. 600.0 2.8 99.5
Philadelphia, Pa. 351.8 14.9 96.0
Huntington, W.Va. 57.9 7.1 88.0
Buffalo, N.Y. 4.7 0.9 81.0
B. REMOVAL/TRANSFORMATION EFFICIENCY FOR EACH INDIVIDUAL PAR
PAR
Pittsburgh
% Removal/transformation
Mean
Philadelphia
Huntington
FL
100
92
90
94.0
RaP
99
99
91
96.3
BkF
99
100
94
97.6
BjF
99
100
94.5
97.8
IP
98
98
87
94.3
BPR
98
92
77
89.0
67
-------
SECTION IX
LABORATORY EVALUATION OF ACTIVATED CARBON FOR ADDITION/REMOVAL OF PAN
Since the use of granular activated carbon beds for general organic
removal in water treatment is fairly widespread, studies were undertaken to
determine the ability of activated carbon to remove and/or add PAR. A
Barnstead organic removal cartridge (catalog No. 0812) was evaluated at
manufacturer’s recommended flow rate and life time. Sixty liters of the
laboratory tap water was passed through the cartridge and PAM were determined
in influent and effluent using the polyurethane plug procedure. The findings
showed that no detectable quantities of PAN were leached from the cartridge.
The experiment failed to provide information on the ability of the cartridge
to remove PAR because the level of PAH in the influent was very low. Further
work is necessary for assessment for activated carbon as a method of treat-
ment for removal of PAH.
68
-------
REFERENCES
1. Acheson, M.A., R.M., Harrison, R. Perry and R.A. Welling. Factor
affecting the extraction and analysis of polynuclear aromatic hydro-
carbons in water. Water Res., 10:207, 1976.
2. Ahnoff, M. and B. Josef sson. Simple apparatus for on—site continuous
liquid—liquid extraction of organic compounds from natural waters.
Anal. Chein., 46:658, 1974.
3. Bedford, J.W. The use of polyurethane foam plugs for extraction of
polychiorinated biphenyls (PCB’s) from natural waters. Bull. Environ.
Cont. Toxicol., 12:622, 1974.
4. Blumer, M. Benzopyrenes in soil. ScIence, 134:474, 1961.
5. Blumer, N. and W.W. Youngblood. Polycyclic aromatic hydrocarbons in
soil and recent sediments. Science, 188:53, 1975.
6. Borneff, J. Polycyclic aromatic hydrocarbons in water, Summary of the
German work on PAH, received from Dr. Borneff.
7. Bowen, H.J.M. Use of polyurethane foam in extraction of substances from
fluid media. J. Chem. Soc. (A.), 1082, 1970.
8. Brass, H.J., W.C. Elbert, M.A. Feige, E.M. Glick and A.W. Lington.
United States Steel, Lorain, Ohio Works, Black River Survey: Analysis
for hexane organic extractable and polynuclear aromatic hydrocarbons.
Report U.S. Environmental Protection Agency, Cincinnati, Ohio,
Oct. 1974, 68 pp.
9. Brocco, D., V. Cantuti and G.P. Cartoni. Determination of polynuclear
hydrocarbons in atmospheric dust by a combination of thin—layer and gas
chromatography. J. Chromatog., 49:66, 1970.
10. Buist, J.N. and B.G. Woods. Flexible foam: manufacture and properties.
In: Adv. Polyurethane Foam Technology, J.M. Buist and H. Gudgeon, eds.
John Wiley and Sons mc,, N.Y. 1968, pp. 141—186.
11. Burnhan, A.K. G.V. Calder, J.S. Fritz, G.A. Junk, H.J. Svec and R. Willis.
Identification and estimation of neutral organic contaminants in potable
water. Anal. Cheni., 44:139, 1972.
12. Conlon, R.D. Liquid chromatography detectors. Anal. Chem., 41:107A,
1969.
69
-------
13. Dong, M. and D.C. Locke. High pressure liquid chromatographic
method for routine analysis of major parent polycyclic aromatic
hydrocarbons in suspended particulate matter. Anal. Chem., 48:368, 1976.
14. Fox, M.A. and S.W. Staley. Determination of polycyclic aromatic
hydrocarbons in atmospheric particulate matter by high pressure
liquid chromatography coupled with fluorescence technique.
Anal. Chem., 48:992, 1976.
15. Gesser, H.D., A. Chow, and F.C. Davis. The extraction and recovery
of polychiorinated biphenyls (PCB) using porous polyurethane foams.
Anal. Letters, 4:883, 1971.
16. Gold, A. Carbon black adsorbates: Separation and identification of
a carcinogen and some oxygenated polyaromatics. Anal. Chem., 47:
1469, 1975.
17. Gough, K.M. and H.D. Gesser. The extraction and recovery of phthalate
esters from water using porous polyurethane foam. J. Chromatgr.,
115:383, 1975.
18. Greinke, R.A. and I.C. Lewis. Development of a gas chromatographic—
ultraviolet absorption spectrometric method for monitoring petroleum
pitch volatiles in the environment. Anal. Chem., 47:2151, 1975.
19. Haenni, E.0., J.W. Howard and F.L. Joe, Jr. Dimethyl sulfoxide:
A superior analytical extraction solvent for polynuclear hydrocarbons
and for some highly chlorinated hydrocarbons. J. Assoc. Off ic. Agr.
Chem., 45:67, 1962.
20. Harrison, R.M., R. Perry and R.A. Welling. Effect of water chlorin-
ation upon levels of some polynuclear aromatic hydrocarbons in water.
Water Research, 9:331, 1974.
21. Horning, E.C., D.I. Carroll, I. Dzidic, K.D. Haegele, M.C. Horning
and J. Stilwell. J. Chromatogr., 99:13, 1974.
22. Ilnitskii, A.P., D.P. Ershova, A.Y. Khesina, L.G. Rozhkova, V.G.
Klubkov and A.A. Korolev. Stability of carbinogens in water and
various methods of decontamination. Gig. Sanit., 36(4):8, 1971
(Russian); (Eng. trans. Hyg. Sanit., 36:9, 1971).
23. International Agency for Research on Cancer, World Health Organi-
zation, Geneva, Switzerland. “Mongraph on the Evaluation of
Carcinogenic Risk of the Chemical to Man: Certain Polycyclic
Aromatic Hydrocarbons and Heterocyclic Compounds,” 3, 1973.
24. Janini, G.M., K. Johnston and W.L. Zielinski, Jr. Use of nematic
liquid crystal for gas—liquid chromatographic separation of
polyaroinatic hydrocarbons. Anal. Chem., 47:670, 1975.
70
-------
25. Keegan, R. The trace fluorometric determination of polynuclear aromatic
hydrocarbons in natural water. Ph.D. Thesis, University of New Hampshire,
1971, 144 pp.
26. Kubota, H., W.H. Griest and M.R. Guerin. Determination of carcinogens
in tobacco, stoke and coal—derived samples — Trace polynuclear aromatic
hydrocarbons. NTIS Report, Conf. —760603— 3, 1975, 17 pp.
27. Lao, R.C.,, R.S. Thomas, H. Oja and L. Dubois. Application of a gas
chromatograph—mass spectrometer—date processor combination to the
analysis of the polycyclic aromatic hydrocarbon content of airborne
pollutants. Anal. Chem., 46:908, 1973.
28. Lao, R.C., R.S. Thomas and J.L. Monkman. Computerized gas chromato—
graphic—mass spectrometric analysis of polycyclic aromatic hydro-
carbons in environmental samples. J. Chromatog., 112:681, 1975.
29. Lawrence, J. and H.M. Tosine. Adsorption of polychlorinated biphenyls
from aqueous solutions and sewage. Environ. Sd. Technol., 10:381, 1976.
30. Lee, M.L., K.D. Bartle and M.V. Novotny. Profile of polynuclear
aromatic fraction from engine oils obtained by capillary—column
gas—liquid chromatography and nitrogen selective detection.
Anal. Chem., 47: 540, 1975.
31. Leoni, V G. Puccetti and A. Grella. Preliminary results on the use
of Tenax for the extraction of pesticides and polynuclear aromatic
hydrocarbons from surface and drinking water for analytical purposes.
J. Chromatogr., 106:119, 1975.
32. Love, 0.T., K.J. Carswell, A.A. Stevens and J.M. Symons. Evaluation
of activated carbon as a drinking water treatment unit process.
EPA Progress Report, 1975, 17 Pp.
33. Najer, J.R. and R. Perry. Mass spectrometric and chromatographic
techniques for determination of polycyclic compounds. Pure Appi.
Chem., 24:685, 1970.
34. Maugh, T.H. Chemical carcinogenesis: A long—neglected field blossoms.
Science, 183:940, 1974.
35. Musty, P.R. and C. Nickless. The extraction and recovery of chlorinated
insecticides and polychlorinated biphenyls from water using porous
polyurethane foams. J. Chromatogra., 100:83, 1974.
36. National Academy of Sciences, Biological Effects of Atmospheric
Pollutants, Particulate Polycyclic Organic Matter, Washington, D.C.,
1972, 361 pp.
71
-------
37. New York State Department of Health, Albany, N.Y. Inventory, Community
Water Systems with Sources,” 1974, 182 pp.
38. Novotny, M., M.L. Lee and K.D. Bartle. The method for fractionation,
analytical separation and identification of polynuclear aromatic
hydrocarbons in complex mixtures. J. Chromatog., 12:606, 1974.
39. Perry, R. Mass spectroinetry in the detection and identification of
air pollutants. Intr. Symp. Ident. Measur. Environ. Pollut. (Proc.),
p. 130, 1971, C.A. 80, 86978m, 1972.
40. Pierce, R.C. and N. Katz. Determination of atmospheric isomeric poly—
cyclic arenes by thin layer chromatography and fluorescence spectrophoto—
metry. Anal. Chem., 47:1743, 1975.
41. Popi, N., N. Stejskal and J. Mostecky. Determination of polycyclic
aromatic hydrocarbons in white petroleum products. Anal. Chem.,
47:1947, 1975.
42. Radding, S.B., T. Mill, C.W. Gould, D.H. Liu, H.L. Johnson,
D.C. Bomberger and C.V. Fojo. The Environmental Fate of Selected
Polynuclear Aromatic Hydrocarbons. EPA—560/5--75/009, Office of
Toxic Substances, Environmental Protection Agency, Washington, D.C.
1976, 121 pp.
43. Sawlcki, E., S.S. McPherson and M. Morgan. Use of gas—liquid
chromatography and thin layer chromatography in characterizing air
pollutants by fluorometry. Talanta, 13:619, 1966.
44. Sawicki, E., T.W. Stanley, W.C. Elbert, J. Meeker and S.P. McPherson.
Comparison of methods for the determination of benzo(a)pyrene in
particulates from urban and other atmospheres. Atmos. Environ.,
1:131, 1967.
45. Searle, T.D., F.J. Cassidy, W.H. King and R.A. Brown. An analytical
method for polynuclear aromatic compounds in coke oven effluents by
combined use of gas chromatography and ultraviolet absorption
spectrometry. Anal. Chem., 42:954, 1970.
46. Skoog, D.A. and D.M. West. Principles of Instrumental Analysis,
Holt, Reinhart and Winston, Ltd., N.Y., 1971, 222 pp.
47. Strickman, R.L. Particles of polyurethane foam. U.S. Patent 3,618,618,
assigned to the Robert L. Stricknian Foundation, Inc., Nov. 9, 1971.
48. Strosher, M.T. and G.W. Hodgson. Polycyclic aromatic hydrocarbons
in lake waters and associated sediments: Analytical determination
by gas chromatography — mass spectrometry. Water Quality Parameters,
ASTM STP 573, Mierlcan Society for Testing and Materials, 1975,
pp 259—270.
72
-------
49. Suess, M.J. Presence of PAIl in coastal waters and the possible health
consequences. Rev. mt. Oceanogr. Med., 18:181, 1970.
50. Suess, M.J. The environmental load and cycle of polycyclic aromatic
hydrocarbons. The Science of the Total Environment, 6:239, 1976.
51. Uthe, J.F., J. Reinker and H.D. Gesser. Extraction of organochiorine
pesticid es from water by porous polyurethane coated with selective
absorbent. Environ. Letters, 3:117, 1972.
52. Water Quality Control Data Book, Organic Chemical Pollution of Fresh
Water, EPA Stock No. 501-0144, U.S. Government Printing Office,
Washington, D.C., 1970.
53. Webb, R.G. Isolating organic water pollutants: XAD resins, urethane
foams, solvent extraction. EPA—660/4—75—003, U.S. Environmental
Protection Agency, Athens, Georgia, 1975, 20 pp.
54. White, R.H. and J.W. Howard. Thin layer chromatography of polycyclic
aromatic hydrocarbons. J. Chrornatog., 29:108, 1967.
55. World Health Organization, International Standards for Drinking
Water, 3rd Edition, Geneva, 1971.
73
-------
APPENDIX
STANDARD OPERATING PROCEDURE
Subject: Monitoring of Polynuclear Aromatic Hydrocarbons (PAR) in Water with
Polyurethane Foam Plugs
A. Applicability of the Method
The method described Is applicable to drinking waters and their raw water
sources. Raw waters containing as much as 102 mg/i of suspended solids and
2.4 g/l dissolved solids have been successfully analyzed. The efficiency of
retention has been determined to be >90% in the concentration range of
0.002—25 ppb. The sample volume generally is 602. for drinking water, and 302.
f or raw water. The method described is for these volumes of water. However
depending upon the concentration of PAR suspected, sample volume can be
increased or decreased.
Since it is impossible to determine all PAR, analysis has been restrict-
ed to the following 6 representatives of the whole group. The World Health
Organization recommends analysis of these PAR In drinking waters.
Fluoranthene
Benzo(a)pyrene
Benzo (ghi) perylene
Benzo (k) f luoranthene
Benzo (j ) fluoranthene
Indeno (1,2,3—cd) pyrene
74
-------
B. Flow Chart
PM in raw and
finished water
Concentrate PAH at
site with prewasbed
polyurethane foam plugs
Ship foam plugs
to the laboratory
“Jr
Spike foam plugs with known amount
of ‘C—BaP (approx 600 dpm)
as internal standard
Elute with acetone—
cyclohexane
‘4,’
Clean—up
(a) partitioning with solvents
(b) column chromatography on
Florisll
Ext tact c ,ncent, at
tol00 i
V
Two dimen ’sional TLC ott
alumina—cellulose acetate
coupled with fluorometric
assay with TLC scanner.
Amount spotted: In accordance
with GLC—calculated results.
75
-------
C. Experimental
1. Materials & Reagent: The sources of the materials and reagents
needed for PAR concentration and analysis are given below.
PAll Concentration
Foam plugs (Trade name: Identiplugs)
45mm x 45mm: VWR Scientific
Chromaflex columns (25mm) & adapters: Kontes Glass Co.
Oscillating type pumps: Scientific Product Inc.
}Iaake Model FE thermostated circulator: Scientific Product Inc.
Gilmont flowmeter: Scientific Product, Inc.
Electric timer: local store
Linear propylene Jerrican: Scientific Product Inc.
Right angle thermometer: New Brunswick Scientific Co.
Water trap for introducing right angle thermometer for
continuous monitoring of water temperature: custom made
Glass coil (10 ft. x 6 imn): custom made
PAR Analysis by TLC—Fluorometry
PAR Standards: Fluoranthene, Benzo(ghi)perylene,
Indeno(l,2,3—cd)pyrene, and Benzo(a)pyrene from Aldrich
Chemical Co.; Benzo(j )fluoranthene and Benzo(k)fluoranthene
from Dr. J.L. Monkman, Air Pollution Control Directorate,
Ottawa, Canada.
Benzo(a)pyrene (7,lO— 14 C): California Bionuclear Corp.
Acetylated cellulose (40%), Aluminum oxide G, type E:
Brinkman Instruments Inc.
Florisil (60—100 mesh): Matheson Coleman and Bell Inc.
Spectrophotofluorimeter with Thin Film Scanner: American
Instrument Co.
Solvents, Distilled in Glass: Mallinkrodt Chemical Co. and
Burdick Jackson Lab Inc.
Chemicals for preparing scintillation fluid: Packard
Instrument Co.,
76
-------
6.5 ml calibrated tubes (each of the last four tenth ml is
calibrated to one hundredth ml): Fisher Scientific
PAH Analysis by Gas—liquid Chromatography
6 ft. x 1/8 in. Stainless steel column packed with
3% Dexsil 300 on chromosorb W 100/120 mesh: Altech
Associates Inc.
Aluminum foil backed septa (Metasep): Altech Associates Inc.
Other reagents and material: Same as described under
TLC—Fluorometric analysis.
2. Sampling System
The sampling system consists of four parts (see Figure)
(a) Thermostated circulator with glass coil einmersed inside
the reservoir housing through the opening of the cover plate.
(b) A 18” x 12” x 1/2” plywood piece to which are mounted
two column stands, water trap equipped with a thermometer, two
oscillating type pumps and a variac to control pumping speed, and
flowmeter. The components are connected such that water travels
in the order: column -‘ water trap -‘- pumps - flowmeter. The
connections up to the foam columns are made with glass tubing
as much as possible and only where necessary with tygon tubing.
The type of tubing used for connections from this point onward
is immaterial.
(c) Jerrican, 20 liter capacity.
(d) Electric Timer.
3. Procedure
(a) Prewashing of foam plugs: Introduce two foam plugs in each
Chromaflex column, and prepare 6 such columns for concentrating PAIL from
602 . of drinking water or 302. raw water. Wash each column with 30 ml acetone,
125 ml beuzene, again with 40 ml acetone and finally with 250 ml distilled
water. Wash these foams with an additional 500 ml distilled water at 60°C
and squeeze the plugs free of any organic solvents by application of suction.
The columns are now ready for passing water for concentration PA}I.
(b) Concentration of PAH: The sampling units and washed foam
columns are transported to sampling site. Two foam columns are clamped to
the column stands in part ‘b’ of the sampler (see Fig. 1), and are connected
in series with the help of glass tubing. The water from the tap which is to
be sampled is continuously run, water comes into a beaker from where it is
pumped into the sampler. Water is brought to 62 ± 2°C in the thermostated
circulator prior to passing through foam columns.
77
-------
Apparatus used for concentrating polynuclear aromatic hydrocarbons from
water: 1 thermostated circulator, 2 foam columns, 3 oscillating pump,
4 water trap with thermometer, 5 flowmeter, 6 timer, 7 graduated 2O2
Jerrican, 8 glass beaker, 9 removable column stand.
78
Direction of water flow
-------
A water flow rate of 250 ± 20 mi/mm is maintained by adjust-
ing the pumping speed with the help of a variac. The thermometer and f low—
meter readings are routinely checked throughout the sampling period and
proper adjustments in the pumping speed and thermostat are made when necessary.
•The effluent from the sampler is collected in a graduated
container (209. capacity Jerrican) for determining the volume of water sampled.
The direction of flow of water through the complete set—up is depicted in
Figure 1. Foam columns are changed every 202w in case of finished water, and
every 109. in case of raw water. The unit may be connected to a timer which
has been set for appropriate time, f or convenience.
(c) Shipment of Foam Columns: PAR are sufficiently stable on foam
plugs that no special handling except wrapping columns with aluminum foil is
necessary for transportation of foam columns to the laboratory following
sampling. However, it is recommended that transit time not exceed 7 days and
foam columns be shipped in styrofoam containers cooled with reusable ice packs.
(d) Addition of Internal Standard: Carbon—14 labeled benzo(a)
pyrene is employed as internal standard in the method of analysis. A known
amount of 14 C (approx. 600 dpm) is added to one of the foam columns and then
elution of PAR is carried out as described below. All the work from this
step on should be done in subdued lighting.
(e) Elution of PAR: Wash each column with 30 ml acetone and
125 ml cyclohexane at a flow rate of 5—10 ml/min into a separatory funnel.
Add 50 ml distilled water to the eluate, shake the contents thoroughly and
let it stand till the layers are well separated. Disregard the lower aqueous
layer and transfer the organic layer into a round bottom flask. Concentrate
the layer to about 10 ml with a rotary evaporator maintaining the temperature
at 40°C and controlling the vacuum such that no bubbling takes place.
(f) Clean—up: Transfer the extract along with two 10 ml cyclo—
hexane washes into 250 ml separatory funnel. Wash it twice with 60 ml 4:1
methanol—water and twice with 60 ml distilled water. Disregard the bottom
methanol water phase in each case. Shake the cyclohexane layer with 20 ml
dimethylsulf oxide and let it settle, withdraw the DMSO layer to another
250 ml separatory funnel. Repeat the DMSO extraction two more times and the
combined dimethylsulfoxide extract is diluted with water (120 ml). PAR from
this phase are extracted with cyclohexane (2 x 40 ml). Wash the combined
cyclohexane phase with water (2 x 60 nil) and dehydrate it by passing it
through anhydrous sodium sulfate which had been prewashed with cyclohexane.
Wash the separatory funnel with cyclohexane (3 x 5 ml) and add the washings
to the sodium sulfate bed. Collect the cyclohexane in a 200 ml round bottom
flask and concentrate it to 5 ml with a rotary evaporator.
Further purification of the extract is achieved by column
chromatography on Florisil. Make a slurry of 8 gms of preactivated 60—100
mesh Florisil in methanol and transfer into a 1.5 x 30 cm column. The bed
is further washed with methanol (100 ml) and 100 ml 1:1 n—hexane and benzene.
Activate the column by placing it in an oven at 130°C for at least 4 hours.
Following activation wash the bed with 100 ml benzene. Transfer the PAR
79
-------
containing cyclohexane layer on Florisil bed with the help of a Pasteur pipet.
Wash the round bottom flask three times with 5 ml benzene and add the washings
to the Florisil column. Elute PAH with 125 ml benzene at a flow rate of
5 mi/mu. Concentrate the eluate to about 2—3 ml by rotary evaporation and
transfer the contents with adequate washiüg to a calibrated tube. Concentrate
the layer further to 0.1 ml with prepurif led grade nitrogen. The extract is
now ready for analysis by gas—liquid or thin layer chromatography.
(g) Determination of Recovery Factor: A 10 p1 aliquot of the
concentrate Is assayed for radioactivity in a liquid scintillation system.
Scintillation fluid is prepared by mixing 5g of 2,5—diphenyloxazole (PPO) and
0.15 g 2,2’—phenylene bls—(5—phen ”l)oxazole (POPOP) in a liter of scintilla-
tion grade toluene. Knowing the 4 C—actlvity originally added to the foam
plugs allows calculation of the recovery factor.
(h) Analysis: It Is recommended that analysis be carried out using
both the gas liquid chromatography with FID detection and thin layer chroma-
tography with spectrophoto—fluorometric detection. The limitation of the gas
chromatographic analysis of PAR are: (1) With the column used in the present
study and most other columns, it fails to separate isomeric PAR compounds (for
example, benzo(k)fluoranthene and benzo(j)fluoranthene give one peak), (ii)
The limit on the sample volume that can be injected makes detection of low
levels of PAll impossible. These problems are overcome when the analysis is
carried out by thin layer chromatography coupled with fluorometry. On the
other hand, TLC—fluorometric analysis is more time consuming and probably not
as quantitative as gas liquid chromatography.
(1) Analysis by Gas—Liquid Chromatography: Analysis is performed using
a gas chroinatogram equipped with a dual flame ionization detector system and
linear temperature programmer. The experimental conditions used In gas
chroinatographic analysis of PAR are as follows:
Column: 6 ft x 118” stainless steel packed with 3% Dexsil 300 on
Chromosorb W, 100—120 mesh (two columns operated in
differential mode).
Carrier gas (nitrogen) flow rate: 30 ml/min.
Detector temperature: 300°C
Injection port temperature: 250°C
Column oven temperature programmed as follows:
Initial temperature: 200°C
Initial delay: 2 miii.
Program: 4°C/mu.
Final temperature: 290°C
Final delay: 8 miii.
Standard mixture: 100 ppm of each PAM in benzene (inject 1—2 pls).
The septa used must be aluminum foil backed (Metasep) and preconditioned
for 12 hours at 250°C.
(ii) Analysis by Thin Layer Chromatography coupled with Fluorometry:
Preparation of Plates : Mix 2 8g aluminum oxide G type E, 12g 40%
acetylated cellulose and 2g CaSO 4 .2H 7 0, 200 mesh (activated for 2 hours at
80
-------
130°C) with 83cc of 95% ethanol on a magnetic stirrer for 5 mm. The resul—
tant slurry is spread to a thickness of 250 pm on eight 20 x 20 cm glass
plates using a coating apparatus. The plates are air dried for 1/2 hr and
activated for 30 mm. at 80°C. Plates are stored in a desicator.
Sample Application : Appropriate aliquot of the concentrate is
applied with the help of a syringe in one corner of the TLC plate about 2 cm
from the two sides. The spot is dried with prepurif led grade nitrogen. The
concentration of PM! as revealed by GLC analysis may serve as a guide to
determine the amount that needs to be spotted. Best resolution is obtained
in the concentration range of 20—100 ng compound/spot.
Deve1op’ nent of Plates : Plates are developed in the first direction
with n—hexane:benzene (4:1) (developing time approx. 30 mm). After drying
the plate is rotated by 90°C and developed in the second direction with
methanol:ether:water (4:4:1) (developing time approx. 2 hr).
Location of Spots : The dried chromatogram is taken into a dark
room and fluorescent spots are visualized with low intensity U.V. illuminators.
The boundaries of spots are marked with a clean, sharp needle.
Identification of Spots : Tentative identification of the spot is
based on (a) fluorescence color and (b) relative position on the plate and
comparison with reference chromatogram.
The fluorescence colors of the 6 PAM are given below:
Benzo(ghi)perylene violet
Benzo(j )fluoranthene yellow—orange
Benzo(k)fluoranthene dark blue
Benzo(a)pyrene violet
Fluoranthene light blue
Indeno(l, 2, 3—cd)pyrene yellow—green
The identity of each spot is confirmed by obtaining fluorescence
emission— and excitation spectra. The spectra are run directly on the plate
with the help of spectrofluorometer equipped with a thin film scanner and a
X—Y recorder. The position of the scanner is adjusted to obtain the maximum
fluorescence signal at a desired spot. Excitation and emission spectra are
then obtained using the wavelengths given below:
Compound Emission spectra Excitation spectra
A for excitation (run) A for emission (urn )
Benzo(ghi)pery lene 300 430
Benzo(j)fluoranthene 300 427
Benzo(k)f luoranthene 300 428
Benzo(a)pyrene 300 427
Fluoranthene 280 458
Indeno(1,2,3—cd)pyrene 300 467
81
-------
The fluorescence emission and excitation peaks of the WHO recommend-
ed polynuclear aromatic hydrocarbons on TLC plates are given below:
Compound
Fluorescence
spectra, wave
excitation
length, tim
Fluoresc
spectra,
ence emission
wavelength, tim
Benzo(a)pyrene
263,
293,
363, 382
403,
427,
453
Benzo(ghi)perylene
295,
362,
380
404,
416,
426
Benzo(j)fluoranthene
293,
347,
365
427,
450
Benzo(k)fluoranthene
240,
300,
376
403,
428,
457
Fluoranthene
278,
342,
354
435,
458
Indeno(l,2,3—cd)pyrene
297,
362,
375
467,
497
Quantitation : Quantitative analysis of the separated compounds is
performed by scanning each spot for fluorescence intensity with the help of
a strip chart recorder. The area of the recorded peak is proportional to the
amount of substance present. Measurement of peak areas is made with a plani—
meter. The concentration of PAR in unknown sample is determined from the
calibration curve.
The experimental conditions used in fluorescence intensity measure-
ment are as follows:
Recorder (50 my) chart speed: 4”/min.
Scanner cycle time: 2 miii.
Excitation wavelength: 365 tim
Fluorescence wavelengths (tim)
Benzo(ghi)perylene 416
Benzo(j )fluoranthene 427
Benzo (k) fluoranthene 428
Eenzo(a)pyrene 427
Fluoranthene 458
Indeno(l, 2, 3—cd)pyrene 467
Preparation of Reference Chromatogram and Calibration Curve :
Prepare a 100 ppm stock solution of each of the six PAR. Mix 5 ml of the
stock solution of fluoranthene and 1 ml each of the other stock solutions in
a small glass stoppered flask. The test solution now contains 50 ng of fluor—
anthene, and 10 ng each of the other PAM in 1 p1.
Prepare three reference chromatogranis with 2, 5and 10 i.il of the
test solution. Scan spots for fluorescence intensity. For each PAIL, a cali-
bration curve is set up by plotting emitted fluorescence (area of the peak)
at the three concentrations as a function of the amount of the compound
applied. F nploying one of the above chromatograms obtain reference emission
and excitation spectra for each of the PAR.
TLC Foam Blank : During elution of PAM from foam plugs a number of
impurities belonging to the foam are also eluted. The clean—up procedure
eliminates these impurities to the extent that they are not seen in the gas
82
-------
chromatographic analysis of the extract. However, since the volume of the
extract subjected to thin layer chromatography is much larger, traces of
fluorescent impurities (0—3 spots, depending upon the volume of the concen-
trate spotted) show up in the thin layer chromatography. This necessitates
running a TLC—foam blank and with its help marking the spots of foam origin
in the test chroinatogram and correcting for background fluorescence. To
prepare TLC—foam blank, 12 prewashed foam plugs are extracted, the extract
cleaned—up and subjected to thin layer chromatography as described above.
(iii) Correction of the Values Determined by Analysis: In order to
calculate the actual amount of PAH present in water samples, the amount
determined by GLC or TLC analysis should be corrected for the recovery
factor as determined earlier. In addition, the benzo(a)pyrene values should
be further corrected for the amount of - 4 C—benzene(a)pyrene added as internal
standard.
83
-------
TECHNICAL REPORT DATA
(Please read f,is_ ructjopz on the reverse before corn pleting
I. REPORT NO. 2.
EPA—600/l—77—052
3. RECIPIENT’S ACCESSIO NO.
4. TITLE AND SUBTITLE
Method Development and Monitoring of Polynuclear
Aromatic Hydrocarbons in Selected U.S. Waters
5. REPORT DATE
Noventhe r 1977 iss.uinci date
6.PERFORMINGORGANIZATIONCOOE
?. AUTP-IOR S)
J. Saxena, D.K. Basu, and J. Kozuchowski
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Syracuse Research Corporation
Syracuse, N.Y. 13210
10. PROGRAM ELEMENT NO.
l C6l4
11.CONTRACT/GRANTNO.
R803977
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory—Cin.,OH
Off ice of Research & Development
U.S. Environmental Protection Agency
Cincinnati,_Ohio_45268
3. TYPE OF REPORT AND PERIOD COVERED
14 SPONSORING AGENCY CODE
600 10
15. SUPPLEMENTARY NOTES
16. ADD, r MCT
A method for concentration of trace quantities of the six representatives of poly—
nuclear aromatic hydrocarbon (PAR) family has been developed and successfully applied
to PAR monitoring in finished and raw waters.
PAR are collected by passing water through polyurethane foam plugs. Water is heated
to 62 ± 2°C prior to passage and flow rate is maintained at approximately 250 mi/mm
to obtain quantitative recoveries. The collection is followed by elution of foam plugs
with organic solvent, purification by partitioning with solvents and column chromato—
graphy on Florisil, and analysis by two dimensional thin layer chromatography—f luoro—
metry and gas liquid chromatography—FID.
Employing this method and a sample volume of 609, PAR have been detected in all the ten
water supplies sampled. Although the sum of the six representative PAR in drinking
waters was small (0.9 to 15 ppt), the values found for raw waters were as high as
600 ppt. It is unclear if PAR are removed during treatment or transformed to another
product and escape detection.
17. KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group -
Polynuclear Aromatic
Water Supply Hydrocarbon, PAR, Poly—
Water Treatment urethane Foam Plugs,
68D
Potable Water Recovery from Water, Raw
Monitors Waters, Concentration
technique, High Volume
Sampler, Clean—up Proce-
dure, PAR Analysis
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGt,S
96
20. SECURITY CLASS (This page)
lJnclassif ied
22. PRICE
EPA Form 2220-1 (0-73)
84
US.GOVEmMNTPR!NrINGOfFICE1,77_ 757-14O/66c 3
------- |