NBS
EPA
U S Department
of Commerce
National
Bureau of
Standards
Center for Analytical
Chemistry
Washington DC 20234
United StatM
Environmental Protection
Agency
Office of Environmental Engineering
and Technology
Washington. DC 20460
EPA-600 7 80 031
February 1980
Research and Development
Development of an
Aqueous Trace
Organic Standard
Reference Material
for Energy Related
Applications
Interagency
Energy/Environment
R&D Program
Report
-------
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 INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports m this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
THE DEVELOPMENT OF AN AQUEOUS TRACE ORGANIC
STANDARD REFERENCE MATERIAL FOR ENERGY RELATED APPLICATIONS:
INVESTIGATION OF THE AQUEOUS SOLUBILITY BEHAVIOR
OF POLYCYCLIC AROMATIC HYDROCARBONS
by
Willie E. May
Center for Analytical Chemistry
National Bureau of Standards
Washington, DC 20234
Interagency Agreement No. EPA-IAGD5-E684
Project Officer
J. Stemmle
Environmental Protection Agency
Washington, DC 20460
This study was conducted
as part of the Federal
Interagency Energy/Environment
Research and Development Program
Prepared for
OFFICE OF ENERGY, MINERALS, AND INDUSTRY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 20460
-------
CONTENTS
Abstract, , . ., .,..,..,..,..,..,.....,..,.,, . iii
Figures iv
Tables v
Acknowledgment , vii
1. Introduction 1
2. Summary and Conclusions 4
3. Methods for Measurement of the Aqueous Solubility
of Aromatic Hydrocarbons 7
4. Experimental Procedures 11
References 46
Appendices
A. Processes That Affect the Accuracy of Dynamic Coupled
Column Liquid Chromatography. , 49
B, Variation of Solubility with Temperature...... 53
-------
DISCLAIMER
This report has been prepared and reviewed by the Center for Analytical
Chemistry and the Office of Environmental Measurements, National Bureau
of Standards, and reviewed by the U. S. Environmental Protection Agency,
and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency. In order to adequately describe materials and experimental
procedures, it was occasionally necessary to identify commercial products
by manufacturer's name or label. In no instance does such identification
imply endorsement by the National Bureau of Standards or the U. S.
Environmental Protection Agency nor does it imply that the particular
products or equipment is necessarily the best available for that purpose.
-------
FOREWORD
The role of the National Bureau of Standards (NBS) in the Interagency
Energy/Environment R&D program, coordinated by the Office of Research
and Development, U. S. Environmental Protection Agency, is to provide
those services necessary to assure data quality in measurements being
made by a wide variety of Federal, state, local, and private industry
participants in the entire program. The work at NBS is under the
direction of the Office of Environmental Measurements and is conducted
in the Center for Analytical Chemistry, the Center for Radiation Research
and the Center for Thermodynamics and Molecular Science. NBS activities
are in the Characterization, Measurement, and Monitoring Program category
and addresss data quality assurance needs in the areas of air and water
measurement methods, standards, and Instrumentation. NBS outputs in
support of this program consist of the development and description of
new or improved methods of measurement, studies of the feasibility of
production of Standard Reference Materials for the calibration of both
field and laboratory instruments, and the development of data on the
physical and chemical properties of materials of environmental importance
in energy production. This report is one of the Interagency Energy/Environment
Research and Development Series reports prepared to provide
detailed information on an NBS measurement method or standard development.
C. C. Gravatt, Chief
Office of Environmental Measurements
National Bureau of Standards
11
-------
ABSTRACT
The development of Standard Reference Materials (SRMs) for aqueous
solutions of known concentration of polynuclear aromatic hydrocarbons
(PAH) is a high priority but extremely difficult undertaking. This
paper is one of a seriae discussing the development of a "Generator
Column" technique for the production of SRMs for PAH's in water. In
addition to providing the basis for SBM development, aqueous solubility
is a fundamental parameter in assessing the extent and rate of the
dissolution of energy based PAHs and their persistence in the aquatic
environment. The extent to which aquatic biota are exposed to a toxicant
such as a PAH is largely controlled by the aqueous solubility of the
toxicant. These solubilities are also of thermodynamic interest, and
give information on the nature of these highly non-ideal solutions.
In this study, saturated solutions of individual PAHs were prepared
by pumping distilled or salt water through a "Generator Column" maintained
at constant temperature. This column was packed with glass beads containing
one percent by weight of the PAH compound of interest.
Extraction of the generated PAH solutions was through an "Extractor
Column". This column was packed with a superficially porous bonded Cis
stationary phase (Bondapak C^g). This column was found to provide
better than 99 percent extraction efficiency for less than 25 ml volumes
of aqueous PAH solutions.
After extraction, a water-acetottitrile solvent blend was passed
through the "Extractor Column" to elute the adsorbed PAH. This eluate
was then passed through a microparticulate analytical column (yBondapak
C18) for separation of the PAH from non-analyte interferences. Individual
response factors were determined by measuring the UV detector signal
produced by injecting a known volume of an acetonitrile solution of the
PAH of interest. This report describes the dynamic coupled column
liquid chromatographic (DCCLC) procedure. The DCCLC technique was
found to be a rapid, accurate, and precise method for measuring PAH
aqueous solubilities.
The DCCLC technique was used to obtain solubility data on twelve
aromatic hydrocarbons. The aqueous solubility at 25 °C was determined
for each compound. The precision of replicate solubility measurements
was better than ±two percent. The variation of solubility of each
compound with temperature is expressed in the form of either a quadratic
or cubic equation on the basis of a least squares fit of the solubility
and temperature measurements. These fits have no theoretical significance,
but they can be used to interpolate the solubility to within ±two
percent of the experimentally measured values between 5 and 30*C.
Enthalpies of solution, AHs, were then calculated in order to describe
the effect of salinity upon solubility. This system was also used to
investigate the partitioning of several PAHs between aqueous solutions
and sediment samples.
Ill
-------
FIGURES
Number Page
— O —
1 Dynamic coupled column liquid chromatographic flow diagram 12
2 Effect of impurities on analyte solubility measurement 18
3 The dependence of the solubility of some PAHs on temperature .... 28
4 Solubility of phenanthrene as a function of sodium chloride
concentration 30
5 Variation of solubility with carbon number 32
6 Variation of solubility with molar volume ... 33
7 Variation of solubility with molecular length 34
8 Concentration dependence of components of a ternary solution on
temperature 44
B-l Temperature dependence of the aaueous solubility of benzene .... 54
B-2 Temperature dependence of the aqueous solubility of naphthalene . . 56
B-3 Temperature dependence of the aqueous solubility of fluorene .... 58
B-4 Temperature dependence of the aqueous solubility of anthracene ... 60
B-5 Temperature dependence of the aqueous solubility of phenanthrene . . 62
B-6 Temperature dependence of the aqueous solubility of
2-methylanthracene 64
B-7 Temperature dependence of the aqueous solubility of
1-methylphanthrene 66
B-8 Temperature dependence of the aqueous solubility of fluoranthene . . 68
B-9 Temperature dependence of the aqueous solubility of pyrene 70
B-10 Temperature dependence of the aqueous solubility of
1,2 benzanthracene 72
B-ll Temperature dependence of the aqueous solubility of chrysene .... 74
B-12 Temperature dependence of the aqueous solubility of triphenylene . . 76
-------
TABLES
Number Page
1 Phenanthrene Solubility Dependence on the Aqueous Flow Rate
Through a "Generator Column" 14
2 A Demonstration of the Equilibrium-Reversible Nature of the
Solution Generating Process with "Generator Column" 15
3 Solubility Dependence on the Supply of Phenanthrene on a
"Generator Column" 16
4 Effect on an Added Impurity, Phenanthrene, on the Concentration
of 2-Methylanthracene Solution Generated at 25.4 8C 17
5 Surface Adsorption Characteristics of PAHs Aqueous Solutions .... 19
6 Losses of Various PAHs from Aqueous Solutions Containing
Caffeine as a Complexing Agent 20
7 Adsorption of Some Aromatic Hydrocarbons from Aqueous Solution
to Stainless Steel Sample Loop at 25 °C 22
8 Precision with which Anthracene Solutions may be Generated and
Measured at 25.4 °C 23
9 The Aqueous Solubilities of Some Aromatic Hydrocarbons as
Determined by Several Investigators 25
10 Variation of Aqueous Solubility with Temperature 26
11 Enthalpies of Solution of Some Aromatic Compounds Between
5 and 30 °C 27
12 Setschenow Constants for Some Aromatic Hydrocarbons at 25 °C . . . .29
13 Correlations of Solubility with Molecular Parameters 31
14 The Partitioning of Phenanthrene Between Silica Gel and Distilled
Water as a Function of Nitrogen Surface Area 37
15 The Partitioning of Some FAHs Between Zipax and Water
at Room Temperature 37
-------
TABLES (Continued)
Number Page
16 The Partitioning of Some PAHs Between Porasil (250) and
Water at Room Temperature 38
17 The Partitioning of Some PAHs Between Water and Alaskan
Sediment "A" 38
18 The Partitioning of Some PAHs Between Water and Alaskan
Sediment "B" at Room Temperature 39
19 Comparison of (LOG) Retention Indices (I) of Phenanthrene,
Anthracene and Methylated Homologs on Several Liquid
Chromatographic Packing Materials 39
20 Effluent Stability of "Generator Column" G-l 42
21 Effluent Stability of "Generator Column" G-2 42
22 Effluent Stability of "Generator Column" G-3 43
B-l Benzene 55
B-2 Naphthalene 57
B-3 Fluorene 59
B-4 Anthracene 61
B-5 Phenanthrene 63
B-6 2-Methylanthracene 65
B-7 1-Methylphenanthrene 67
B-8 Fluoranthene 69
B-9 Pyrene 71
B-10 1,2-Benzanthracene 73
B-ll Chrysene 75
B-12 Triphenylene 77
Vl
-------
ACKNOWLEDGMENT
The author is grateful to the U. S. Environmental Protection Agency for
partial support of this work under the Interagency Energy/Environment Agreement
EPA-IAG-D5-E684. This work is from a dissertation submitted to the Graduate
School, University of Maryland, (December 1977) by Willie E. May, in partial
fulfillment of the requirements for a Ph.D. degree in Chemistry.
Vli
-------
1. INTRODUCTION
The development of a Standard Reference Material (SRM) for aqueous
solutions of polynuclear aromatic hydrocarbons (PAH) in a water matrix
is an extremely important but difficult undertaking. The need for such
SRMs to provide data quality assurance in the measurements and monitoring of
energy related organic effluents has been identified in a number of studies,
including the series of Workshops sponsored by the National Bureau of Standards
and the Environmental Protection Agency, the reports of the Proceedings of
which are contained in other publications of this Energy/Environment Series.
One of the major SRM recommendations developed by those Workshops was for
energy-based PAH in a water matrix. In response, the National Bureau of
Standards instituted a project to seek to develop SRMs for PAH's. This report
is one of several providing details on the development of that technique, which
is based on the "generator column" concept. An understanding of the
solubility behavior of PAH's in aqueous systems is imperative before any
preparation of SRMs can be performed.
An understanding of the solubility behavior of polycyclic aromatic
hydrocarbons (PAHs) in aqueous systems is important in several fields. In
water pollution control, such information is helpful in devising abatement
processes (1), in modeling natural water systems (2), in designing toxicity
experiments, and in developing analytical techniques. In petroleum research,
aqueous solubilities are useful in understanding how hydrocarbons might
migrate and accumulate to form oil fields (3). In biology, a knowledge of
how hydrocarbons behave in water is important for understanding the effects of
hydration on the configuration of biopolymers (4). And in chemistry,
solubility data are needed for testing models concerned with the behavior of
these compounds in aqueous solution (5). For example, Kites (6) has suggested
that the natural mechanism that modifies PAH homolog distribution on sediments,
following initial deposition, is the differential water solubility of the
various alkyl homologs.
The aqueous solubility is a fundamental parameter in assessing the
extent and rate of the dissolution of polycyclic aromatic hydrocarbons and
their persistence in the aquatic environment. The extent to which aquatic
biota are exposed to a toxicant such as a PAH is largely controlled by the
aqueous solubility of the toxicant. These solubilities are also of thermo-
dynamic interest, and give information on the nature of these highly non-
ideal solutions.
Data on the aqueous solubility of PAHs in the presence of a third
component, such as an electrolyte, are also very important. The practical
implications are that the presence of this third component may substantially
change the solubility, an example being the "salting out effect" of sodium
chloride present in seawater (7). Methods for determining PAH aqueous
solubilities are subject to errors associated with both the preparation,
extraction, and quantitative analysis of saturated solutions. There is
-------
no one method that has addressed the problems associated with each of
these processes. Systematic errors associated with quantitative analyses
of saturated solutions should be reduced in methods where selective
analytical measurement techniques are used. Chromatographic methods allow
separation of nonanalyte signals-in-time from those of the analyte.
Fluorescence spectroscopy allows greater selectivity than ultraviolet
spectroscopy though less than gas or liquid chromatography.
Efficient extraction of saturated aqueous solutions of PAHs with
organic solvents is usually not a problem. Problems are however associated
with the transfer of aliquots of the saturated solution to extraction vessels.
Adsorptive losses of PAHs on the surfaces of transfer tools (pipettes,
beakers, etc.) are possible. These errors can be eliminated by rinising
the transfer tools with the extracting solvent or by employing methods
that do not involve transfer steps.
The source of systematic error that remains, and that is present in
all methods that have been used to measure PAH aqueous solubilities, is
associated with the preparation of saturated solutions. In all methods
that have been reported, saturated solutions were prepared by adding
excess quantities of the PAH to water and mechanically mixing the solution
for at least 24 hours.
Peake and Hodgson (35, 37) have shown that when hydrocarbons are
dissolved by mechanical means, the resulting solutions are often supersaturated.
They called this phenomenon accommodation. They showed that accommodated
hydrocarbons are not in equilibrium with the water and that their concentrations
are a function of hydrocarbon supply, settling time, and mode of introduction.
Wasik and Brown (34) have shown that accommodation is prevented
when saturated solutions are prepared by equilibration of hydrocarbon vapors,
rather than liquids or crystalline hydrocarbons with water. The methods
that employ this approach (33, 34), however, do not have the sensitivity or dynamic
range necessary to measure PAH aqueous solubilities accurately.
Although there are values for the aqueous solubility of many PAHs in
the literature, they have been reported at only one temperature. The
agreement between values determined by different methods is sometimes
poor. Furthermore, there have been very few determinations of the
aqueous solubility of PAHs in seawater. Because of the increasing need
for information about these systems, a study was undertaken to investigate
"The Solubility Behavior of Polycyclic Aromatic Hydrocarbons in Aqueous
Systems."
In order to begin this investigation, it was necessary to develop a
method capable of accurately measuring PAH solubilities. In this method
saturated solutions are prepared by an equilibrium process and extracted
almost instantaneously. Quantitative analyses of the extract is done by
reverse-phase HPLC. Preparation, extraction, and analysis of the saturated
solutions all occur within the same system.
-------
In this investigation, this method was used to measure:
(1) The aqueous solubility of some aromatic hydrocarbons;
(2) The effect of temperature on solubility;
(3) The effect of salinity on solubility;
(4) The partitioning of some PAHs between water and sediments.
This methodology is also being used in the development of an aqueous
PAH Standard Reference Materials (SRM).
-------
2. SUMMARY AND CONCLUSIONS
The National Bureau of Standards currently issues over 900 Standard
Reference Materials (SRM's), with various groups being represented, such as:
clinical laboratory standards, trace element standards, nuclear materials,
glass viscosity standards, rubber materials, color standards, and coating
thickness standards. We are now endeavoring to add to this list an
additional group, namely trace organic Standard Reference Materials.
One of the first SRM's from this new group of materials will be an
aqueous polynuclear aromatic hydrocarbon (PAH) Standard Reference Material.
There are several problems associated with the preparation, storage and handling
of aqueous solutions of PAH, that previously prevented the development of this
SRM. Now through the use of a dynamic coupled-column liquid chromatographic
method developed in this laboratory, we have been able to circumvent these
problems. The use of this technique for the preparation and certification of an
aqueous PAH SRM will be discussed.
A. Problems Associated with the Preparation and Stabilization of a
PAH/Water SRM
Preparation of aqueous PAH solutions of known concentration by gravimetric
procedures is difficult because of the extremely low aqueous solubilities of PAH.
As shown in Table 9 the aqueous solubilities of many PAH are less than
500 ug/kg (ppb). Preparation of aqueous solutions of known concentration by
serial dilutions of a more concentrated organic solution is both hazardous and
wasteful. After small aliquots are taken, large volumes of organic solvent
containing toxic and expensive chemicals remain to be disposed of.
Preservation of stable aqueous solutions of PAH is hampered by adsorptive
losses of the PAH to the surfaces of containers and transfer tools. The
magnitude of the adsorptive effect is variable and is a function of the
manner in which the solutions are handled. The adsorptive properties of three
PAH on four different surfaces are shown in Table 5. These results show that
losses of PAH from static solutions to surfaces occur in short periods of
time. Stirring such solutions only slightly reduces such losses.
B. Preparation and Quantisation of Dilute PAH Solutions by Dynamic
Coupled-Column Liquid Chromatography
Recently, we have developed a dynamic coupled-column liquid chromatographic
(DCCLC) method for investigating the aqueous solubility behavior of PAH (39).
In the DCCLC method, saturated aqueous solutions of PAH are generated by pumping
water through a column packed with glass beads that have been coated with the
compound of interest ("generator column"). The concentration of the desired
compound in the effluent of the "generator column" is measured by a modification
of the coupled-column liquid chromatographic process that has been previously
described by May et_^ a^., (38). A flow diagram of this system is shown in
Figure 1. Since the aqueous solubility of a compound is a well-defined
thermodynamic quantity, saturated aqueous solutions produced by "generator
columns" may also be defined as standard solutions at a given temperature.
-------
Stable saturated aqueous solutions are eluted from "generator columns"
after an initial aqueous purge volume of approximately 1000 mL. After this
initial conditioning, equilibrium is obtained and the PAH concentration at
constant temperature has been shown to be independent of flow rate between
0.1 and 5 mL/min. Equilibrium can be re-established after a change in
temperature by passing 10 mL of water through the column under the new
conditions. A purge volume of only 100 mL is necessary to re-equilibrate a
"generator column" after a shelf storage period of as long as three months.
Extraction of the generated PAH solutions is accomplished by passing a
measured volume of that solution through an "extractor column" (see Figure 1).
This 60 x 0.6 cm column is packed with a superficially porous bonded C-.Q
stationary phase (Bondapack C.R, Water Associates, Milford, MA) and provides
better than 99 percent extraction efficiency for less than 25 mL volumes of
aqueous PAH solutions.
After extraction, a water-acetonitrile solvent blend is passed through
the "extractor column" to elute the adsorbed PAH. This elute is then passed
through a microparticulate analytical column (yBondapak C.g) for separation
of the PAH from non-analyte interferences. Individual response factors
are calculated by replacing the "extractor column" with a calibrated sample
loop, and injecting known amounts of the PAH of interest dissolved in
acetonitrile.
C, Evaluation of DCCLC as a Method for the Preparation and Certification
of an Aqueous PAH SRM
There are several factors that make the DCCLC approach ideal for the
preparation and analysis of very dilute aqueous solutions of individual
PAH:
1. Saturated solutions are prepared by an equilibrium process. See
Section 4, Tables 1 and 2.
2. The use of "generator columns" to produce aqueous PAH solutions
circumvents the problems that are usually associated with storing such
solutions, since the solutions need not be generated until they are needed.
3. The concentration of the generated aqueous solutions are a function
of temperature, and may be expressed in terms of least squares fits of the
concentration to temperature. Such equations can be used to interpolate
the concentration as a function of temperature to within +2% of the experimentally
determined values between 5 and 30 °C.
4. Both the short and long term precison with which aqueous PAH solutions
can be generated appear to be <2%.
5. Shelf storage of "generator columns" does not seem to present a
problem. They have been shown to be stable for longer than 1 1/2 years and
through more than 100 liters of aqueous purge.
-------
6. DCCLC is a rapid and accurate method for analyzing the generated
dilute aqueous PAH solutions. Analytical errors due to adsorption are minimized
in DCCLC because the solution is extracted and concentrated, on line, in less
than 500 ms after generation. It has been estimated that this analytical method
has an uncertainty of less than 2 percent .
-------
3. METHODS FOR MEASUREMENT OF THE AQUEOUS SOLUBILITY OF AROMATIC HYDROCARBONS
The aqueous solubilities of benzene, the alkylbenzenes, and the
napthalenes have been measured by several investigators (7-22). All of these
investigators prepared saturated solutions by adding an excess
quantity of the solute to water and mechanically stirring the mixture for
at least twenty four hours. These solutions were usually allowed to settle
prior to filtration, extraction, and quantitative analysis by ultraviolet
spectroscopy.
This method worked well for the fairly soluble compounds that are
being studied. Interlaboratory precision was good. Therefore, this basic
methodology, with minor modifications, such as the use of gas
chromatography or fluorescence spectroscopy for analysis of the aqueous
solutions has been generally adopted for the determination of the aqueous
solubilities of PAHs (23-32) even though the solubilities of some PAHs differ
from that of benzene by a factor of more than 10 . In the remainder of this
section some of the modifications that have been made on this basic method,
and some new approaches to the measurement of the aqueous solubility of aromatic
hydrocarbons will be briefly reviewed.
Measurement of Solubility of Nephelometry
In 1942, Davis and co-workers (16) determined the aqueous solubility of
30 PAHs at 29 °C. The solubilities varied from 1600 yg/kg. In this
procedure, the test substance was first dissolved in a water miscible
solvent such as ethanol or acetone. Dilutions of increasing amounts of
this solution with relatively large volumes of water gave a series of
turbid suspensions. The turbidity was measured nephelometrically and
the relative intensity of scattered light was plotted against the
concentration of the test substance. Extrapolation of this standard curve
to the relative intensity of a reagent blank gave the experimental
solubility.
The precision of replicate analyses was reported to be +10 percent.
The major source of uncertainty was the narrow time frame between saturation
of the solution, and coalescence of the dispersed crystals to an extent that
would alter the nephelometric behavior of the solution. This approach
was also nonselective. It was impossible to discriminate against nonanalyte
signals arising from crystalline impurities, dust particles, etc.
Measurement of Solubility by Ultraviolet Spectroscopy
In 1972, Wauchope and Getzen (29) studies the temperature dependence
of the aqueous solubility of some PAHs between 25 and 75°C. Saturated
solutions were prepared by adding 20 grams of each solid to a 250 mL glass
stopped flask containing distilled water. The flasks were suspended in an
open water bath and shaken gently for one to three weeks between measurements.
-------
Temperature control was maintained to within K>.5°C. Samples of the solutions
for measurement were withdrawn with pipets through glass wool plugs and
emptied into volumetric flasks containing measured amounts of cyclohexane for
extraction of the aqueous solutions. The volume of cyclohexane was
chosen so that the measured signal fell in the range between 0.5 and 1.5
absorbance units. Although the temperature of the equilibrated aqueous
solutions varied from 25 to 75°C, this technique allowed the analyses and
the Beer-Lambert coefficient determinations to be performed at room temperature.
This method was more selective and it gave better precision 0+5%) than the
nephelometric method because quartitative analysis was performed at the
X maximum for each individual compound.
Measurement of Solubility by Gas Chromatography
Although this method was not applicable for PAHs, McAuliffe (8) was
the first investigator to use chromatography as an analytical tool in the
determination of aqueous hydrocarbon solubilities. He measured the
solubility of some light alkanes, olefins, and aromatics by making direct
50 uL injections of mechanically-prepared saturated solutions into a gas
chromatograph. This chromatographic step allowed him to eliminate (via
separation in time) nonanalyte signals contributed by dissolved impurities
associated with the analyte. For example, he found that the gas chromatogram
of cyclopentane revealed an associated 0.2 percent impurity when injected
neat. The size of this impurity peak increased to 25 percent of the size
of the cyclopentane peak after euqilibration with water.
Sutton and Calder (9) have also measured the solubilities of several
alkylbenzenes in distilled water and in sea water by a method based on
gas chromatography. Saturated solutions were prepared by equilibrating
water with aromatic vapor in an all-glass apparatus consistine of a one-
liter Erlenmeyer flask with an insert tube. The insert tube was used to
store the compound. It was capped with a ground-glass stopper. The
liquid hydrocarbon did not come into contact with the water except
through a perforation in the insert, which allowed hydrocarbon vapors to
enter the headspace above the water in the flask. The flask was placed
in a constant temperature-shaking bath controlled at 25.0 + 0.1°C. The
water was equilibrated for 48 hours prior to analysis. The solubilities
were determined by solvent extraction of the saturated solutions with
subsequent analyses of the extracts by gas chromatography.
The solubilities reported by Sutton and Calder were 5 to 20 percent
higher than those determined by McAuliffe. This is not surprising since
the McAuliffe method is susceptible to serious losses due to adsorption
of the hydrocarbons on the walls of the syringe. (See Section 3, pages
16-18, for a discussion of the phenomenon.) Sutton and Calder injected
concentrated organic extracts rather than dilute aqueous solutions. This
reduced the chances for adsorptive losses during injection and increased
the sensitivity of the method. However, neither McAuliffe's nor Sutton
and Calder's method is sufficiently sensitive for determination of the aqueous
solubility of PAHs.
-------
Determination of Solubility by Headspace Analysis
Some researchers have determined the aqueous solubility of several
benzenes and naphthalenes using a headspace analysis method. In 1971,
McAuliffe (33) reported hydrocarbon solubilities determined by the following
method. A 50 mL glass hypodermic syringe was filled with equal volumes
of an aqueous hydrocarbon solution and helium. The mixture was vigorously agitated
for 20 minutes on a wrist-action shaker to establish equilibrium between the
phases. The gas phase was then flushed through a sample loop of a gas sampling
valve and injected into a gas chromatograph. A fresh supply of helium was
added to the syringe and equilibrated with the aqueous solution and treated as
before. A plot of log peak area in the gas phase versus equilibration number
(gas volume) produced a straight line. The parition coefficient (CICg) between
the gas and liquid phase can then be calculated from the slope of the plot and
the gas and liquid volumes. The solubility is the product of the partition
coefficient and the saturated vapor pressure.
Wasik and Brown later (34) modified McAuliffe's static headspace method.
They constructed a dynamic headspace analysis unit where hydrocarbons were
introduced into the apparatus as vapors to avoid the danger of emulsion
formation (35), which can occur when liquid hydrocarbons are mixed with water.
They allowed the gas mixtures to circulate through the liquid phase for at
least 1/2 hour to insure that equilibrium was attained. After the first
equilibration, a small portion of the headspace was sampled via a gas
sampling valve and measured by gas chromatography, while the remainder was
vented to waste. Fresh helium was then added, equilibrated with the
aqueous solution and analyzed as before. The partition coefficients were
calculated as in McAuliffe's method.
Determination of solubility by headpsace analysis offers several
advantages over spectrophotometric techniques. First, because of the
selectivity of chromatographic analysis, compound purity is not a critical
factor; second, absolute calibration of the gas chromatographic detector is
not necessary if the response is linearly related with concentration over
the range necessary for the measurements; and finally, this method does not
require the preparation of saturated solutions, since a partition coefficient,
not a solubility, is actually measured. However, headspace methodology would
probably not be applicable for determining PAH solubilities for three reasons.
First, there is little data in the literature on the vapor pressures of PAHs.
Second, the aqueous solubilities of most PAHs are too low to be measured by this
procedure. Finally, adsorptive losses of PAHs to glass surfaces from the
vapor phase would cause errors.
Measurement of Solubility by Fluorescence Spectroscopy
Recently, Schwarz (31) has measured the temperature dependence of
the solubilities of several PAHs in aqueous solutions by a fluorescence
method. Saturated solutions were prepared and measured in situ in
modified fluorescence cells. Each cell contained 5 ml of an aqueous
-------
solvent with an excess amount of the PAH to be studied. The cells were
rotated for at least 24 hours at each temperature before the solution was
allowed to settle and be measured. The fluorescent intensity of these
saturated aqueous PAH solutions was found to be proportional to
concentration between 8 and 30°C. The fluorescence measurements were put on
an absolute scale by ultraviolet spectroscopic (UV) measurements at 25°C.
This fluorescence method has two major advantages over the methods
previously described for measuring the solubilities of PAHs in aqueous
solutions. First, no transfers of the saturated solutions are made so that
systematic errors arising from adsorption on the transfer tools is eliminated.
Fluorescence is also inherently more selective for PAHs than UV and other
nonluminescence spectroscopic techniques. Nonanalyte signals may be reduced
through both selective excitation and selective monitoring of fluorescence
(.emission). This latter advantage is partially negated, however, since the
absolute measurement is made by UV. One of the disadvantages of the method
lies in the fact that the solutions cannot be filtered prior to analysis to
remove suspended microcrystals and other particulate matter.
Estimations of Solubility by High Performance Liquid Chromatography
Locke (36) has used a reverse-phase high performance liquid chromatographic
(HPLC) method to estimate the aqueous solubilities of some PAHs. He
made the following postulates: selectivity in reverse-phase HPLC is governed
by the solubilities of similar solutes in the mobile phase; group
selectivity is determined by both the stationary and mobile phases;
selectivity toward individual components within a group is set by the eluent.
The eluent used in reverse-phase HPLC is a blend of water and an organic
solvent that is miscible with water. Acetonitrile or methanol is most
often used as the organic component. PAHs (up to 5 condensed rings) are
fairly soluble in both acetonitrile and methanol. The aqueous solubilities
of PAHs are low and vary between 1 and 2000 yg/kg. Selectivity then is based
on the differences in the aqueous solubility of the individual PAHs. Locke
suggested that a linear relationship should exist between log retnetion
volume and log solubility of unsubstituted PAHs. He used the solubilities of
benzene and fluoranthene to define the slope of the line, and determined the
solubilities of other aromatic hydrocarbons from this relationship.
Locke's method provides a rapid and simple means for estimating the
aqueous solubilities of PAHs. Individual solubility values can be
estimated to within 200 percent. However, these extrapolated values differ
too greatly from measured solubilities for this procedure to be categorized
as a valid method of determining the aqueous solubility of PAHs,
10
-------
4. EXPERIMENTAL PROCEDURES
DEVELOPMENT OF DYNAMIC COUPLED COLUMN LIQUID CHROMATOGRAPHY - AN ACCURATE
METHOD FOR THE DETERMINATIONS OF THE AQUEOUS SOLUBILITY OF POLYNUCLEAR AROMATIC
HYDROCARBONS
Any method for determining the aqueous solubilities of PAHs must
provide a means of preparing saturated solutions of the compounds of interest,
and performing quantitative analysis of those solutions. Some of the methods
previously reported in the literature for determining PAH aqueous solubilities
were reviewed in Section 1. The accuracy and precision of these methods can
be adversely affected by a number of processes including: incomplete
equilibration of the solid hydrocarbons with the aqueous medium; dispersion
rather than true dissolution of the solid hydrocarbons (accommodation); failure
to remove suspended microcrystals from the solution; losses of the PAHs from
solution to the surfaces of containers, filters and transfer tools due to
adsorption.
The dynamic coupled column liquid chromatographic (DCCLC) technique
which will be discussed in this section circumvents all the problems started
above. In addition to a description of this technique this section
includes a critical evaluation of the method in terms of the saturated
solution preparation, transfer, extraction, and quantitative analysis
processes. The section is concluded with a discussion of the precision
and potential accuracy of this method.
Methodology
This method for determining aqueous solubility is based on generating
saturated solutions by flowing water through a column packed with glass beads
coated with the compound to be measured (generator column). The
concentration of the desired compound in the effluent from the generator
column is measured by a modification of the coupled column liquid chromatographic
procedure developed by May et al. (38). A flow diagram of the system is
shown in Figure 1.
Reagents—
The water used in this study was distilled from a potassium permanganate-
sodium hydroxide solution and passed through an XAD-2 column (38). The
acetonitrile and sodium chloride were spectro and reagent grades, respectively.
All of the PAHs used were obtained commercially and were reported to contain
11
-------
Generator
Pump
Extractor
Column
Six Port Switching Valve
Generator
Column
UV Detector
Waste
Constant
Temp * Waste
Bath
Programmer
Digital Strip
Integrator Chart
Recorder
H20 CH3CN
Pump Pump
Detail of Six Port Switching Valve
Extractor Column
Extractor Column
From
Generator
Column
To
Analytical
Waste
Column
From
HPLC Unit
From
Generator
Column
Waste
From
HPLC Unit
Extract
Position
Analyze
Position
Figure 1. Dynamic coupled column liquid chromatographic flow diagram.
12
-------
less than three percent impurities by their respective manufacturers. Gas
and liquid chromatographic analyses of these materials supported these
claims.
Preparation of Saturated Solutions Using "Generator Columns"—
Columns for the generation of saturated solutions of PAHs were prepared
by packing 60 x 0.6 cm stainless steel columns with 60-80 mesh glass beads
coated with 1 percent (wt/wt) of the compound of interest. The beads were
coated by adding 20 grams of the beads to 200 ml of a 0.1 percent methylene
chloride solution of the PAH of interest and stripping the solvent with a
rotary evaporator.
Columns for the generation of saturated solutions of liquid aromatic
hydrocarbons such as benzene were prepared by pumping 50 mL of the liquid
through a 60 x 0.6 cm stainless steel column with uncoated glass beads.
Excess amounts of benzene were purged from these columns with 50 mL of
water.
Saturated solutions were generated by pumping distilled water or
saline solutions through these columns at flow rates ranging between 0.1
and 5 mL/min. These columns were thermostated by means of a water bath
controlled to ±0.05 °C.
Extraction of Generated Solutions—
Extraction of the PAHs from the generated solutions was accomplished
by flowing a measured volume of the solution through a 6 x 0.6 cm stainless
steel "extractor column" packed with a 37-50 ym superficially porous support
with a bonded CIS stationary phase (Bondapak CIS, Waters Associates, Milford,
MA) and fitted on both ends with 2 ym frits. This column has been found to
provide greater than 98 percent extraction efficiency for less than 25 ml
volumes of aqueous PAH solutions. (See Appendix A-l) Generated solutions
of benzene were not passed through the "extractor column", but were injected
directly into the chromatographic system via a loop sample injector. Reasons
for this modification are discussed later in the section and in Appendix A-l.
Chromatographic Transfer and Analysis of Extract—
After extraction, flow from the "generator column" was diverted to
waste, while an acetonitrile - water solvent blend from a HPLC gradient
pumping system (Waters, Associates, Milford, MA) was simultaneously routed
through the "extractor column" to elute the adsorbed compounds. The eluate
then passed through a 30 x 0.6 cm microparticulate analytical column (yBondapak
C18, Waters Associates) for separation of individual PAHs from nonanalyte
interferences such as those caused by crystalline impurities. Isocratic
elution conditions were usually employed, resulting in some band broadening
in the transfer of extracted components from the extractor to the analytical
column. The band broadening resulting from this transfer process was
eliminated when necessary by employing a linear solvent gradient from water
to acetonitrile. This caused elution focusing of the components at the
head of the analytical column. A detailed discussion of this phenomenon
is given elsewhere (39).
13
-------
The Integrated UV detector signal produced by each of the aromatic
hydrocarbons was determined to be proportional to its concentration.
Individual response factors were found for each compound by first replacing
the "extractor column" with a calibrated sample loop and then injecting
acetonitrile solutions with known concentration of the individual hydro-
carbons. Details concerning the loop calibration technique and preparation
of the acetonitrile solutions are given in Appendices A-3 and A-4 respectively,
Critical Evaluation of the DCCLC Method
The development of an accurate method for determining the aqueous
solubility of PAHs is contingent upon the ability of that method to prepare,
maintain, transfer and analyze saturated solutions. In this section some
of the problems associated with PAH solubility methodology in general are
discussed along with explanations of how these problems are circumvented by
the DCCLC approach.
Production of Saturated Solutions Via an Equilibrium Process--
Stable saturated solutions were eluted from generator columns after an
initial aqueous purge volume of between 100 and 500 mL. After this initial
conditioning, equilbrium was obtained and the PAH concentration at constant
temperature became independent of flow rate between 0.1 and 5 mL/min (see
Table 1). Equilibrium could be re-established after a change in temperature
or salinity by passing 10 mL of water through the column under the new
conditions. A purge volume of approximately 50 mL was necessary to recon-
dition a "generator column" after a shelf storage period of three months.
TABLE 1. PHENANTHRENE SOLUBILITY DEPENDENCE ON THE AQUEOUS FLOW RATE
THROUGH A "GENERATOR COLUMN"
Flow Rate
(mL/min.)
5.0
0.4
0.1
Temperature
Concentration
(yg/kg)
865±7
868±4
86613
22.0 eC
aThe concentration reported represent the averages of five measurements at each
temperature. The uncertainties represent the standard deviation of the mean
at each temperature.
14
-------
The reversible nature of this process was demonstrated by the following
experiment. Distilled water was pumped through two independently thermostated
generator columns that were connected in series. The temperature of the
first column, "A", was maintained at 24.3 °C. Concentration measurements
were made on the effluent from the second column in the series, "B", at
temperatures of 25.3, 12.8 and 6.6 °C. The concentration of the solution
that eluted from column "B" of the series, at each temperature, was identical,
within experimental error, to the concentration that had been obtained from
column "B" alone. The data from this experiment is presented in Table 2.
TABLE 2. A DEMONSTRATION OF THE EQUILIBRIUM-REVERSIBLE NATURE OF THE SOLUTION
GENERATING PROCESS WITH "GENERATOR COLUMNS"
Generator Column Generator Column Generator Column
"A" "B" "A" + "B" in series
Temp Cone. Temp Cone. Temp Cone.
(°C) (yg/kg) (°C) (yg/kg) (°C) (yg/kg)
24.3 42.7 25.3 45.5±0.2 25.3 45.4±0.1
12.8 21.3±0.2 12.8 20.710.1
6.6 14.0±0.1 6.6 14.2±0.2
The flow rate through these anthracene "Generator Columns" was 5.0 mL/min.
Normally saturated PAH solutions were generated by pumping distilled
water through "generator columns". The results of this experiment show that
identical results were obtained when the opposite situation was imposed,
that is the PAH concentration in the feed solution was greater than that
expected in the effluent. This could only happen through an equilibrium
process.
Accommodation—
It is well established that hydrocarbons can exist in colloidal,
micellar or particulate form in appreciable quantities. This was first
noted by Peake and Hodgson (35,37) who found that filtration reduced the
apparent solubility or "accommodation" measured for hydrocarbon solutions
prepared by mechanical means. They also found that the degree of accommoda-
tion (difference between the measured and equilibrium amounts of hydrocarbons
in the water) was a function of hydrocarbon supply, settling time, filtration
pore size, and the mode by which crystalline hydrocarbons were introduced
into the solutions.
15
-------
The saturated solutions produced by "generator columns" are independent
of hydrocarbon supply. Phenanthrene solutions of identical concentrations
were generated by a "generator column" prepared by the procedure described
earlier (coated glass beads) and one prepared by packing a 60 x 0.6 cm
column with crystals of phenanthrene (see Table 3). The concentrations of
the solutions produced by "generator columns" are also stable through large
volumes of aqueous purge, where the supply of hydrocarbon on the column is
being steadily depleted (see Table 8). This depletion however represents only
a small percentage of the total amount of compound on the column.
TABLE 3. SOLUBILITY DEPENDENCE ON THE SUPPLY OF PHENANTHRENE
ON A "GENERATOR COLUMN"
1% Phenanthrene on Crystalline
Glass Beads Column Phenanthrene Column
(60 x 0.6 cm) (60 x 0.6 cm)
Phenanthrene Concentration in Generator Column Effluent
(vg/kg) (yg/kg)
959 938
947 935
958 943
955+7 Average 939±4
Aqueous Flow Rate - 5 mL/min.
Temperature - 24.3 °C
Differential settling times are not a problem because the saturated
solutions are generated and measured almost instantaneously, all within
the same system.
Filtration of the saturated solutions does not alter the concentration
of the solutions generated by "generator columns" because the solutions
are actually filtered during, and not after preparation. The filters
(snubbers) which are fitted on both ends of these columns are saturated
with the PAH of interest during the initial column conditioning period
and are therefore an integral part of the solution generating system.
Nonanalyte Signals—
One of the problems that reduces the accuracy of methods that use spectro-
scopic techniques for the measurement of solubility is the presence of
16
-------
dissolved impurities which contribute to the spectroscopic signal. The most
direct method for eliminating this problem is by the use of ultrapure chemicals.
Ultrapure PAHs, however, are difficult to obtain. Thus selective analytical
techniques must be used to minimize such interferences.
The method discussed in this report employs a reverse phase liquid
chromatographic procedure which separates the signals produced by nonanalyte
constituents, in time, from the signal produced by the analyte. The use
of this technique can also relax the purity requirements for the individual
PAHs. The following experiment was performed to illustrate this point:
A generator column coated with a 1 percent blend of a 90 percent 2-methyl-
antracene, 7 percent phenanthrene, and 3 percent of several other PAHs
was prepared, conditioned and purged with 4500 ml of water at 25.4 °C.
The results of this experiment are presented numerically in Table 4 and
graphically in Figure 2. It should be noted that impurities are eluted
from the column along with the analyte. After initial conditioning, the
concentration of the 2-methylanthracene is stable while the concentrations
of the impurities continue to diminish. The measured concentration of
2-methylanthracene produced by the impurity containing column at 25.4 °C
was 21.8±0.4 yg/kg. The concentration of the solution produced by a
"generator column" packed with glass beads coated with only 2-methylanthracene
and run under identical conditions was 22.6±0.5 vg/kg.
TABLE 4. EFFECT OF AN ADDED IMPURITY, PHENANTHRENE, ON THE CONCENTRATION
OF 2-METHYLANTHRACENE SOLUTION GENERATED AT 25.4 °C
Volume Pumped Through Phenanthrene 2-Methylanthracene
"Generator Column" Cone. Cone.
(yg/kg) (yg/kg)
0
75
225
375
525
675
825
975
4500
<1215
1215
1219
1198
1194
1196
1193
1183
161
21.6
21.6
22.3
21.4
22.1
22.8
21.8
21.9
21.7
Note: Measured 2-Methylanthracene concentration - 21.8±0.4 vg/kg. Concentra-
tion calculated from a least square fit of solubility and temperature
data obtained using a "generator column" packed with 99% pure
2-Methylanthracene - 21.9±0.2 yg/kg.
17
-------
INITIAL
B
oo
POST
500 cc
D
E
A
, L
J
*
D
U
POST
975 cc
A) UNIDENTIFIED IMPURITY
B) UNIDENTIFIED IMPURITY
C) PHENANTHRENE
D) 2-METHYLANTHRACENE
POST
4,100 cc
Figure 2. Effect of impurities on analyte solubility measurement.
-------
Adsorption—
The most elusive analytical interference associated with the measurement
of PAH aqueous solubilities is that of adsorptive losses of these compounds
from solution to the surfaces of containers and transfer tools. PAH
adsorption causes reductions in the measure analyte signal, whereas the
other interferences discussed up to this point cause positive errors.
The magnitude of the adsorption effect is variable and a function
of the manner in which the solutions are handled. The adsorptive properties
of three PAHs on four different surfaces are shown in Table 5. These
results show that losses of PAHs from static solutions to surfaces occur
in short periods of time. The use of caffeine as a PAH complexing agent
(32) reduces these losses, but does not eliminate them (see Table 6).
Stirring the solutions causes only small reductions in these losses. The
PAH losses shown in Table 6 for stirred solutions are less than those
reported in Table 5 for static solutions.
TABLE 5. SURFACE ADSORPTION CHARACTERISTICS OF PAHs
AQUEOUS SOLUTIONSa
Percent Loss from Solution
Phenanthrene Chrysene Benz(a)Pyrene
1 Hr 13 Hrs 1 Hr 13 Hrs 1 Hr 13 Hrs
Glass
Silanized Glass
Platinum
Aluminum
8
5
35
5
73
87
76
46
64
66
50
71
81
85
89
53
73
57
67
82
93
93
95
aThe concentration of each PAH was 1 yg/kg.
Surface to volume ratios are approximately equal to approximately 1 cm2/cm^ for
all materials studies. All measurements are made by the coupled column LC
technique described by May et^ al. in Reference 38.
The dynamic coupled column liquid chromatographic system was also designed
to circumvent problems due to adsorption. After preparation and filtration,
saturated solutions are transferred and extracted by a process that minimizes
PAH contact with surfaces. The volume between the "generator column" and the
"extractor column" is approximately 6 pL. The time required for transfer of
the saturated solutions at a flow rate of 5 mL/min is less than 75 micro-
seconds. Furthermore, the walls of the transfer lines are pre-saturated with
the compound being studied during the column conditioning process. This further
reduces the possibility of adsorptive losses of the PAHs during the brief time
that the saturated solutions remain in these lines.
19
-------
TABLE 6. LOSSES OF VARIOUS PAHS FROM AQUEOUS SOLUTIONS
CONTAINING CAFFEINE AS A COMPLEXING AGENT3
Exper-
iment
1
2
3
Complexing
Agent
0.1% Caffeine
None
0.1% Caffeine
None
0.1% Caffeine
None
Experimental
Conditions
Stir 4 hr,
then analyze
Stir 4 hr,
then analyze
Stir 16-20 hr,
then analyze
Stir 16-20 hr,
then analyze
Stir 40 hr,
then analyze
Stir 40 hr,
then analyze
Percent Loss
3,4
Pyrene Chrysene Benzpyrene
10 10 10
29 61 68
25 29 32
74 72 86
12 22 25
96 93 96
1,2,5,6-
Dibenzanthracene
44
78
61
85
61
97
0.5 pg of each PAH present in 500 mL of distilled water (1 pg/kg per compound).
-------
Direct injection of the generated saturated solutions into the
chromatographic system via a sample loop was initially attempted. PAH
adsorptive effects, however, restricted the use of this more simple and
rapid measurement approach to only relatively soluble aromatic hydrocarbons
such as benzene (solubility 1800 ppm at 25 °C). The results presented
in Tables 5 and 6 indicate that the magnitude of the adsorption problem
is inversely related to aqueous solubility. Therefore, substitution of
a stainless steel sample loop for the "extraction column" would be
expected to cause systematic positive errors. The results presented in
Table 7 indicate that the size of these errors would be a function of
both the solubility of the hydrocarbon and the volume of saturated solution
passed through the loop prior to injection into the chromatographic system.
The data presented in Table 7 also shows that adsorptive errors may be
eliminated from the aqueous solubility determination of benzene by limiting
the volume of saturated solution passed through the loop to 50 yL prior
to injection (one stroke of the pump used to pass water through the
generator column delivers 50 yL). The magnitude of the errors caused by
adsorption on the walls of the sample loop were much greater for naphtha-
lene (see Figure B-2 in Appendix B) and the less soluble PAHs, and were
not eliminated by limiting the volume of saturated solution passed
through the loop to 50 yL prior to injection. Because of these adsorptive
errors, saturated solutions of all the compounds studied in this investigation,
with the exception of benzene, were extracted via an "extraction column"
and transferred to the chromatographic system via the coupled-column
process that was described in this section.
The Potential Accuracy of the DCCLC Method—
The preparation of saturated solutions by "generator columns" has
been shown to be the result of an equilibrium process. It has been
demonstrated that the "extractor column" can provide quantitative extraction
of naphthalene and the PAHs from aqueous solutions. Problems associated
with the stability of aqueous PAH solutions are avoided because the
solutions are instantaneously analyzed after they are generated. Nonanalyte
signals are separated from those of the analyte by chromatographic
means. The potential accuracy of the DCCLC method is then limited by
the liquid chromatographic analysis of the generated PAH solutions.
The accuracy of the liquid chromatographic analysis of the generated
solutions is limited by the uncertainties involved with the calibration
of the sample loop, preparation of standard acetonitrile solutions of the
PAHs and the volumetric measurement of the amount of saturated solution
sampled for a given analysis. The random errors that are associated with
each of these processes have been estimated to be less than ±1.2 percent,
±0.1 percent, and ±1.0 percent, respectively. An explanation of how each of
these estimates was made is presented in Appendix A. Quadratic addition of
these random errors yields a minimum uncertainty of 1.6 percent for the
quantitative analysis of the generated saturated solutions and hence a poten-
tial accuracy of greater than 98 percent for the method.
21
-------
TABLE 7. ADSORPTION OF SOME AROMATIC HYDROCARBONS FROM AQUEOUS SOLUTION
TO STAINLESS STEEL SAMPLE LOOP3 AT 25 CC.
Volume of Saturated Solution
Passed Through Loop
(mL)
Apparent
Concentration
(mg/kg)
.05
1.00
10
25
.05
0.20
.0
.0
2.
5.
10
.05
4.0
9.2
11.0
Benzene
Naphthalene
Phenanthrene
1795
1800
2273
2728
32.0
32.2
35.0
36.4
37.8
0.97
1.36
1.61
1.70
The concentrations of these generated solutions as determined by solvent
extraction followed Liquid Chromatographic analysis (Benzene) of DCCLC were:
Benzene - 1800 ±50 (mg/kg)
Naphthalene - 31.8 ±1
Phenanthrene - 0.95± .02
aSample Loop Volume - 23.1 uL
Generator Column
Flow Rate - 0.4 mL/min. in all cases.
22
-------
Reproducibility—
Both the short and long term precision with which saturated solutions
can be generated and measured are better than ±3 percent. The results
presented in Table 8 demonstrate this fact.
TABLE 8. PRECISION WITH WHICH ANTHRACENE SOLUTIONS MAY
BE GENERATED AND MEASURED AT 25.4 °C
Volume of
Eluted
Distilled Water
Through Column
(mL)
1
190
230
500
775
910
1175
Concentration
Measured
(Vg/kg)
45.2
47.0
46.6
46.1
46.9
44.6
45.5
Average Measured Concentration - 46.0±0.9
(6-21-77)
Concentration Calculated from
Calibration Curve - 45.7
(12-20-76)
Concentration Measured by
Solvent Extraction Followed
By GC Analysis - 45.8±1.0
(6-21-77)
THE APPLICATION OF DYNAMIC COUPLED COLUMN LIQUID CHROMATOGRAPHY TO THE
DETERMINATION OF THE AQUEOUS SOLUBILITY AND OTHER RELATED PARAMETERS OF
SOME AROMATIC HYDROCARBONS
Aqueous solubility data for the twelve aromatic hydrocarbons studied
in this investigation are reported in this section. The solubilities deter-
mined spanned a range of 106. The solubilities measured at 25 °C are compared
23
-------
to values reported by other investigators and are correlated with molecular
parameters such as carbon number, molar volume and molecular length.
The variations of solubility with both temperature and salinity are
reported. Enthalpies of Solution, AHs, are calculated from the temperature
dependence of the solubility, and the effect of salinity on solubility,
expressed in terms of the Setschenow constant, is reported for each compound.
Aqueous Solubilities at 25 °C
Table 9 compares the solubilities determined by DCCLC with some values
reported by other investigators. Of the twelve values reported, there are
only two cases of gross disagreement with the consensus literature value.
Those are the values for anthracene and triphenylene.
The solubilities reported for anthracene are clustered about two
values. A possible reason for this phenomenon lies in the fact that most
commercial preparations of anthracene contain at least two percent phenanthrene.
Though the two compounds are structural isomers, phenanthrene is approximately
twenty times more water soluble than anthracene. The presence of phenanthrene
in solution would contribute a positive systematic error to methods that
employ nonspecific analytical measurement techniques, such as ultraviolet
spectroscopy (29) and nephelometry (24). The value reported by Schwarz (31),
who employed a more selective analytical technique (fluorescence) agrees with
that determined by DCCLC.
The aqueous solubility of triphenylene as determined by DCCLC is in
gross disagreement with all other literature values. The reason for this
discrepancy is unknown. DCCLC values obtained through use of triphenylene
prepared by different commercial manufacturers were identical. Neither
variations in the length of, nor triphenylene supply on, the "generator
column" had any effect on the solubility determined.
Variation of Solubility with Temperature
Solubility data for benzene, naphthalene and ten other PAHs are presented
in Table 10 in the form of either a quadratic or cubic least squares fit of the
solubility to temperature for each compound. These fits have no theoretical
significance, but can be used to interpolate the solubility as a function
of temperature to within ±2 percent of the experimentally measured values
between 5 and 30 °C. The experimental values from which these equations
were derived are given in Appendix B.
The data presented in Appendix B-l show that the DCCLC method has
both sufficient precision and accuracy to detect a minimum in solubility
profile of benzene in the region between 17.5 and 18.0 °C although the
solubility never varies by more than 2.5 percent between 5 and 25 8C. This
effect has been previously observed and reported by several other investigators
(10,11,16,34).
24
-------
TABLE 9. THE AQUEOUS SOLUBILITIES OF SOME AROMATIC HYDROCARBONS
AS DETERMINED BY SEVERAL INVESTIGATORS
M
Compound
Benzene
Naphthalene
Fluorene
Anthracene
Phenanthrene
2-Methylanthracene
01 1-Methylphenanthrene
Fluoranthene
Pyrene
1 , 2-Benzanthracene
Chrysene
Triphenylene
olecular
Weight
78.1
128.2
166.2
178.2
178.2
192.3
192.3
202.3
202.3
228.3
228.3
228.3
25
1791 ±
31.69 t
1.685 t
.0446 ±
1.002 ±
.0213 ±
.269 ±
.206 ±
.132 ±
.0094 ±
.0018 ±
.0066 ±
*C 29 *C 29 *C
(1942)
10
.23
.005
.0002 .0570 ± .003 .075 ± .005
.011 1.220 ± .013 1.600 ± .050
.003
.003
.002 .264 ± .002 .240 ± .020
.001 .162 ± .001 .165 ± .007
.0001 .0122 ± .0001 .011 ± .001
.00002 .0022 ± .00003 .0015 t .0004
.0001 .038 ± .005
MacKay & Shui Schwarz° Wauchope
25 *C 25 *C & Getzen Others
(1977) (1977) (1972)
1780(34). 1796(22),
1755(13), 1780(12)
31.7 ± .2 30.3 ± .3 31.2 34.4(11)
1.98 ± .04 1.90
.073 ± .005 .041 ± .0003 .075 - .075(28,30)
1.290 t .070 1.151 ± .015 1.180 .994(32)
.260 i .020 0.265 0.240(32)
.135 ± .005 .129 ± .002 .148
.014 ± .0002 .010(28)
.002 ± .0002 .006(28)
.043 ± .001 .043(28)
a Solubilities determined by the nepholometric method described in Reference (24).
b Solubilities determined by method described in Reference (26).
0 Solubilities determined by fluorimetrlc method described in Reference (31).
d Solubilities determined by UV method described in Reference (29).
-------
TABLE 10. VARIATION OF AQUEOUS SOLUBILITY WITH TEMPERATURE
Compound
Solubility Dependence on Temperature
(pg/kg)a
Correlation
Coefficient2
Benzene
Naphthalene
Fluorene
Phenanthrene
1-Methylphenanthrene
Pyrene
Fluoranthene
Anthracene
2-Methylanthracene
1,2 Benzanthracene
Chryaene
Thiphenylene
(0.0247t3 - 0.6834t2 + 0.3166t + 1833) x 103
(0.0189t2 + 0.2499t + 13.66) x 103
0.0185t3 + 0.4543t2 + 22.76t + 543
0.0025t3 + 0.8059t2 + 5.413t + 324
O.OOSOt3 - 0.1301t2 + 6.802t +55.4
-O.OOllt3 + 0.2007t2 - l.OSlt +50.2
0.0072t3 - 0.1047t2 + 4.322t + 50.4
0.0013t3 - 0.0097t2 + 0.8861t + 8.21
O.OOllt3 - 0.0306t2 + 0.8180t +2.79
0.0003t3 - 0.0031t2 + 0.1897t +1.74
0.0024t + 0.0144t + 0.609
-0.002t3 + 0.0250t2 - 0.4250t + 4.89
.9443
.9987
.9999
.9992
.9994
.9997
.9988
.9998
.9988
.9991
.9982
.9990
3These equations and correlation coefficients were obtained by fitting solubility vs. temperature data
to a Polynomial Regression (2° or 3°) program supplied with a Hewlett Packard 9830A Calculator.
-------
Calculation of the Enthalpy of Solution
The solubility data presented in the preceding section can be used
to calculate the heat of solution, AH , from the following relationship:
8
where T is the absolute temperature, R is the ideal gas constant, and S
is the molar solubility at T. Values of AH at 298 °K were calculated
from a least squares fit of the integrated form of equation (1), namely
InS - -AH /RT + b, where b is a constant.
8
(2)
The results of these least squares fits are presented graphically in
Figure 3. Values of AH for eleven of the compounds studied in this investi-
gation are found in Table 11 along with values for AH determined by other
investigators. A single value of AH cannot be reported for benzene because
the enthalpy of solution varies with temperature.
TABLE 11. ENTHALPIES OF SOLUTION OF SOME AROMATIC COMPOUNDS
BETWEEN 5 AND 30 °C
Compound
Naphthalene
Fluor ene
Anthracene
Phenanthrene
1-Me thylphenanthrene
2-Methylanthracene
Fluoranthrene
pyrene
1 2-Benzanthracene
Chrysene
Triphenylene
This Work
6.30 ± .12
7.88 ± .15
10.46 ± .10
8.32 ± .14
9.34 ± .18
10.10 ± .17
9.52 ± .38
8.47 ± .36
10.71 ± .25
9.86 ± .24
10.71 ± .39
Enthalpy of Solution
(Kcal/Mol)
Schwarza
5.29 ± .10
8.32 ± .38
8.68 ± .23
11.4 ± .2
Wauchope
& Getzen
6.10 ± .
6.99 ± .
10.4 ± .
7.7 ± .
7.3 ± .
28
16
8
9
2
Reported in reference 31.
^Calculated from solubility data reported in Reference 29 in the 0-30° temp.
range.
27
-------
20.00
CC
LLJ
t 16.00
LLJ
_J
o
Z 12.00
8.00
11
3.25 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65
TH (KELVIN) x103
Figure 3. The dependence of the solubility of some PAHs on temperature.
28
-------
Variation of Solubility with Salinity
Determination of the solubility of PAHs in water is of both thermodynamic
interest and of practical relevance in assessing the environmental fate and
effects of oil present in rivers, lakes, ground water, and oceans. Although
the aqueous solubilities of a large number of specific PAHs have been
measured and correlated, few data exist on the solubility in the presence
of a third component such as an electrolyte. The practical implications
are that the presence of this third compound may substantially change
the solubility, an example being the "salting out" effect of sodium chloride
present in sea water.
Setschenow (40) derived an empirical relationship for the magnitude
of the "salting out" effect, that is, the dependence of solubility on
the salt concentration of saline solutions:
log So/Ss - K C
s s
where So and Ss are the concentrations of the solute in fresh and salt
water, respectively, KS is the Setschenow constant for the sodium chloride
solution and C is the molar salt concentration. The solubility of
phenanthrene as a function of ionic strength and temperature is illustrated
graphically in Figure 4. The value of the Setschenow constant is shown
to be independent of temperature over the range studied. The Setschenow
constants calculated in the same fashion for the compounds studied in
this investigation are presented in Table 12.
TABLE 12. SETSCHENOW CONSTANTS FOR SOME AROMATIC HYDROCARBONS
AT 25 °C
Compound
Benzene
Naphthalene
Fluorene
Anthracene
Phenanthrene
2-Methylanthracene
1-Me thy Iphenanthrene
Pyrene
Fluoranthene
Chrysene
Triphenylene
1 , 2-Benzanthracene
K (liters /mole)
S
0.175
0.213
0.267
0.238
0.275
0.336
0.211
0.286
0.339
0.336
0.216
0.354
.006
.001
.005
.004
.010
.006
.018
.003
.010
.010
.010
.002
29
-------
3.50
3.20
J
CO
2.30
2.00
NaCI CONCENTRATION
(MOLES/LITER)
Figure 4. Solubility of phenanthrene as a function of sodium chloride concentration.
-------
The theory developed by McDevitt and Long (35) for the "salting out"
of liquid hydrocarbons from aqueous solutions predicts an increase in K
with increasing liquid molar volume for nonelectrolyte solutes. This
relationship seems to also be valid for this series of crystalline compounds
with trlphenylene again giving anamolous results. (Molar volumes presented
in Table 13.)
TABLE 13. CORRELATIONS OF SOLUBILITY WITH MOLECULAR PARAMETERS
Molar Volume3 Molecular Length
Compound Carbon No. (mL) (A) -ln(S)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
a.^— -^—
Benzene
Naphthalene
Fluorene
Phenanthrene
Anthracene
Pyrene
Fluoranthene
1,2 Benzanthra-
cene
Triphenylene
Chrysene
6
10
13
14
14
16
16
18
18
18
189
125
153
159
160
172
175
194
190
194
5.5
8.0
9.5
10.5
9.5
9.4
11.8
9.5
11.8
3.77
8.30
11.5
12.1
15.2
14.2
13.8
17.0
17.4
18.7
aReported by Davis and Goltbieb in reference 52.
^Reported by Klevens in reference 25.
Correlations of Solubility with Molecular Parameters
The aqueous solubility of aromatic hydrocarbons has been shown by
Klevens (25) to be related to carbon number, molar volume and molecular
length. These parameters along with the molar solubilities (expressed as
-InS) of the compounds studied are presented in Table 13. Figures 5 through
7 demonstrate the relationship between each of these parameters and
solubility. These data figures show that there are several compounds
whose anamolous behavior makes accurate extrapolations of solubility from
these relationships impossible. For example, anthracene and phenanthrene
are structural isomers. They, therefore, have identical carbon numbers
31
-------
20.00
^ 15.00
cc
(0
111
o 10.00
N)
)
5.00
0.00
5.00
10.00 15.00
CARBON NUMBER
20.00
Figure 5. Variation of solubility with carbon number.
-------
20.00
75.00
100.00
125.00 150.00
MOLAR VOLUME (ml)
175.00
olO
200.00
Figure 6. Variation of solubility with molar volume.
-------
20.00
Crt
7.00
MOLECULAR LENGTH
(A)
Figure 7. Variation of solubility with molecular length.
10.00
-------
and very similar molar volumes. However, their aqueous solubilities differ
by more than a factor of twenty. Phenanthrene, fluoranthene, pyrene and
triphenylene all have very similar molecular lengths; but their respective
aqueous molar solubilities at 25 °C are, 5.6x10% l.Oxlcf6, 6.8x10 7,
2.8xlO~8.
Tsonopoulos and Pransnitz (27) have reported that the hydrocarbon infinite
dilution coefficient, y00, is the appropriate quantity for correlating the
aqueous solubilities of hydrocarbons. They, along with Leinonen et al. (23)
and Pierotti et al. (28) have successfully correlated Y" with carbon number,
molar volume, and degree of branching. Recently Mackay and Shiu (26) have
correlated the hydrocarbon infinite dilution coefficients of 32 aromatic
hydrocarbons (using the super cooled standard state) with carbon number. From
this relationship, they derived a parabolic equation from which individual
solubilities could be calculated. A comparison of the correlated and
experimental solubility values showed that solubilities could be estimated
only to within a factor of 3.
The DCCLC technique is a very rapid, precise and accurate method
for determining aqueous solubilities and related parameters for polynuclear
aromatic hydrocarbons. The agreement with solubilities and calculated
solubility parameters, such as AH , for which values have been previously
reported in the literature help to confirm the validity of this approach.
In all cases replicate solubility measurements of the generated PAH
solutions at constant temperature were better than ±3 percent and in
most cases better than ±1 percent. Quantitative analyses of hexane
extracts of the generated solutions by both GC and HPLC have agreed with
the DCCLC values to within ±3 percent in all cases.
The lack of an appropriate physical parameter for accurately extrapolating
PAH solubilities further enhances the utility of the DCCLC approach. PAH
aqueous solubilities can be accurately measured in about the same time that
it would take to make an estimation via correlation with a physical
parameter.
PAH SORPTION ON SEDIMENTS
Introduction
Recent studies conducted in this laboratory (NBS) and elsewhere have
shown that sediments taken from areas that were thought to be pristine and
those taken from areas which have been exposed to low levels of petroleum
both contain PAH mixtures of similar distribution. The alkyl homologs
predominate over the unsubstituted species. Different possible sources of
the PAHs in these and other marine sediments have been suggested. These
include biosynthesis; petroleum spillage; combustion from mobile and
stationary sources, including refuse burning, forest fires, and industrial
activity.
35
-------
There have been suggestions that at least some PAHs originate from
biosynthesis of algae (6), plants (41), or various bacteria (42). Kites
and Hase (43) have shown, however, that bacteria accumulate but do not
biosynthesize PAHs.
The PAHs found on marine sediments may also originate from petroleum.
In petroleum, the three and four carbon alkyl homologs predominate. Petroleum
derived PAH mixtures are usually very deficient in the unsubstituted parent
PAH.
Combustion-produced PAHs transported by air is another likely source
of these compounds in marine environment. Low temperature combustion (such
as 1100 °K, as in a cigarette) will yield a soot with an abundance of the
alkyl substituted PAHs. Blumer and Youngblood (44) have concluded that
PAHs found in several recent sediments are likely to have originated from
forest fire particulates. Kites (6), however, refutes this conclusion and
offers instead the belief that PAH mixtures are initially deposited by
anthropogenic sources and then undergo an in situ modification. Kites
postulates that the differential water solubility of the higher alkyl
homologs versus the unsubstituted species is the mechanism that is responsible
for the PAH distributions found on recent sediments. He proposes the
following model: After airborne particulate deposition on soil or in water,
the lowest homologs continuously fractionate into the water phase to an
extent proportional to their solubility (increasing alkylation lowers
aqueous solubility), the remaining species, which accumulate on the particu-
late matter and on sediment, are, therefore, devoid of the lowest homologs,
thus increasing the relative abundance of the higher homologs.
Measurement of PAH Sorption by DCCLC
The DCCLC system is also capable of providing a means of characterizing
the sorption characteristics of PAHs on solid adsorbents. Since the concen-
trations of generated PAH solutions are known functions of temperature, PAH
sorption data for the adsorbent was determined by placing a column packed
with the material to be characterized between the "generator column" and
the switching valve. (See Figure 1). A plot of the concentration measured
at the exit of the adsorbent column versus time (or volume) provided data
pertaining to the kinetics of the sorption process (interpretation of such
data are beyond the scope of this work). Equilibrium was defined as that
point at which the PAH concentration at the adsorbent column inlet and
outlet were the same. At that point the equilibrium uptake was measured by
removing the adsorbent from the column and extracting it with diethyl
ether. An aliquot of that extract was then analyzed by liquid chromatography
Results
Table 14 presents some results obtained when using this method for
measuring the equilibrium partition coefficient between distilled water and
several silica gels. The sorption of phenanthrene is shown to be a function
of surface area for these materials. Tables 15 through 18 present the
results obtained for the partitioning of a PAH mixture of phenanthrene,
1-methylphenanthrene, anthracene, and 2-methylanthracene on two silica'gels
36
-------
and two Alaskan sediment samples. The absolute values of the partition
coefficients (K) vary, but the preferential sorption of the methylated
PAHs is evident in all four experiments. These partition coefficients are
not simply inverse functions of solubility. The partition coefficient for
anthracene is less than that for 1-methylphenanthrene, although 1-methyl-
phenanthrene is more soluble by a factor of 5.
TABLE 14. THE PARTITIONING OF PHENANTHRENE* BETWEEN SILICA GEL AND
DISTILLED WATER AS A FUNCTION OF NITROGEN SURFACE AREA
Surface Area
Concentration Sorbed
on Silica
Partition Coefficient,
K
, cone, sediment N
2
1 2
22
200
400
(mg/kg)
8 ± 2
50 ± 4
95 ± 7
130 ± 3
v cone, water '
8
52
99
135
^henanthrene aqueous concentration - 0.96 mg/kg.
Surface areas used are those reported by the manufacturer.
All measurements were made at room temperature.
TABLE 15. THE PARTITIONING OF SOME PAHS BETWEEN ZIPAX
AND WATER AT ROOM TEMPERATURE
Concentration
in Water
Concentration on
Sediment at
Equilibrium
Cone. Sediment
Phenanthrene
1-Me thy Iph enanthr ene
Anthracene
2-Me thylanthracene
(Pg/kg)
710
114
15.9
8.1
(yg/kg)
1790
816
25.7
94.0
v Cone. Water '
2.5
7.2
1.6
12
aZipax is a superficially porous adsorbent marketed by the DuPont DeNemours Co.,
Wilmington, Delaware.
37
-------
TABLE 16. THE PARTITIONING OF SOME PAHS BETWEEN PORASIL (250)
AND WATER AT ROOM TEMPERATURE
Concentration
in Water
.(Jig/kg)
Phenanthrene
1-Me thy Iphenanthr ene
Anthracene
2-Me thy lanthracene
809
92.4
12.0
9.18
^orasil (250) is a totally porous silica
Associates, Milford, Massachusetts.
Concentz
Sedime
EquiH
.(jig/
2.93
1.00
5.85
1.46
material
ation on
mt at
.brium f (
kg) (
xlO4
xlO3
x 102
xlO3
manufactured
K
-otic. Sediment
Cone. Water '
36
108
49
159
by Waters
TABLE 17. THE PARTITIONING OF SOME PAHS BETWEEN WATER
AND ALASKAN SEDIMENT "A"
Phenanthrene
1-Me thy Iphenanthr ene
Anthracene
2-Me thy lanthracene
Concentration
in Water
(yg/kg)
710
114
15.9
8.1
Concentration on
Sediment at
Equilibrium
(yg/kg)
7.3 x 103
2.3 x 103
3.0 x 102
2.9 x 102
K
, Cone. Sediment
Cone. Water '
10
20
19
36
38
-------
TABLE 18. THE PARTITIONING OF SOME PAHS BETWEEN WATER AND
\LASKAN SEDIMENT "B" AT ROOM TEMPERATURE
Concentration
Concentration on
Sediment at
K
Phenanthrene
1-Methylphen
Anthracene
2-Methylanthracene
in Water
(yg/kg)
874
ithrene 94.8
11.7
acene 9 . 80
Equilibrium
(yg/kg)
3930
1185
127
246
.. Cone. Sediment ..
Cone. Water
4.5
13
11
25
These trends in PAH sorption in aqueous systems are not predicted by
chromatographic retention on liquid/solid (adsorption) or normal bonded phase-
partition chromatographic systems. Table 19 compares the relative chromato-
graphic retention of these four compounds on silica, alumina, and yBondapak
NH2 using hexane as a mobile phase and on yBondapak C\Q using a water-acetonitrile
mobile phase. There is little difference shown in the chromatographic
retention of these three condensed ring PAHs on the two classical adsorbents,
silica and alumina, or yBondapak NH2, a new "bonded phase" chromatographic
material. However, the trends in the retention of these three condensed ring
PAHs on yBondapak C\Q (a bond phase usually used in the reverse phase mode)
closely resembles the trends of the partition coefficients measured for the
sorption of these compounds from water onto silica gels and sediments.
TABLE 19. COMPARISON OF (LOG) RETENTION INDICES (I)a OF PHENANTHRENE, ANTHRACENE,
AND METHYLATED HOMOLOGS ON SEVERAL LIQUID CHROMATOGRAPHIC PACKING MATERIALS
Silica (45) Alumina (64)
yBondapak yBondapak (48)
NH2 (48) C18
Phenanthrene
1-Me thylphenanthr ene
Anthracene
2-Me thy lanthracene
3.00
3.26
2.97
•te_»
3.00
3.18
3.00
»
3.00
2.98
2.95
2.92
3.00
3.43
3.02
3.63
aThe retention indices were calculated as previously described by Popl et al. (45),
with the hydrocarbon standards being assigned the following retention indices:
benzene 10, naphthalene 100, phenanthrene 1000, benzo(a)anthracene 10,000, and
benzo(b)chrysene 100,000. The retention index of an aromatic hydrocarbon was
then calculated in a manner analogous to the calculation of KovSts indices for
gas chromatography.
39
-------
Summary
The results of this brief study indicate that there should be much more
work directed toward the understanding of the mechanisms of the sorption (par-
titioning, adsorption, etc.) of slightly soluble components from water onto
particulate matter. The DCCLC technique is an excellent tool for conducting
such studies.
DEVELOPMENT OF AN AQUEOUS POLYNUCLEAR AROMATIC HYDROCARBON STANDARD REFERENCE
MATERIAL
Introduction
The National Bureau of Standards currently issues over 900 Standard
Reference Materials (SRMs), with various groups being represented, such as:
clinical laboratory standards, trace element standards, nuclear materials,
glass viscosity standards, rubber materials, color standards, and coating
thickness standards. We are now endeavoring to add to this list an additional
group, namely trace organic chemical Standard Reference Materials.
The first SRM from this new group of materials will be an aqueous poly-
nuclear aromatic hydrocarbon (PAH) Standard Reference Material. There are
several problems associated with the preparation, storage and handling of
aqueous solutions of PAH that previously prevented the development of this SRM
Now, through the use of a dynamic coupled-column liquid chromatographic method"
developed in this laboratory, we have been able to circumvent these problems.
The use of this technique for the preparation and certification of an aqueous
PAH SRM will be discussed.
Background
In the fall of 1975, the National Bureau of Standards (NBS) and the
Environmental Protection Agency (EPA) jointly sponsored a series of workshops
entitled "Standards and Reference Materials for Environmental Analysis
Associated with Energy Development". The objective of these workshops was to
obtain input to NBS on the methodology and certified standards needed for the
accurate analysis of environmental samples associated with the production of
alternate fuels.
At the conclusion of these workshops, a number of Standard Reference
Materials (SRMs) were recommended by the participants for NBS consideration.
One of the SRMs recommended was a polynuclear aromatic hydrocarbon (PAH) in" a
water matrix. Although many PAH have demonstrated mutagenic properties this
SRM was given a low priority because of the presumed difficulties associated
with the preparation and stabilization of such a material.
Problems Associated with the Preparation and Stabilization of a PAH/Water SPM
Preparation of aqueous PAH solutions of known concentration by gravimetri
procedures is difficult because of the extremely low aqueous solubilities of
PAH. As shown in Table 9 many PAH have aqueous solubilities of less than
500 vig/kg (ppb). Preparation of aqueous solutions of known concentration by
40
-------
serial dilutions of a more concentrated organic solution is both hazardous and
wasteful. After small aliquots are taken, large volumes of organic solvent
containing toxic and expensive chemicals remain to be disposed of.
Preservation of stable aqueous solutions of PAH is hampered by adsorptive
losses of the PAH to the surfaces of containers and transfer tools. The magni-
tude of the adsorptive effect is variable and is a function of the manner in
which the solutions are handled. The adsorptive properties of three PAH on
four different surfaces were shown in Table 5. These results show that losses
of PAH from static solutions to surfaces occur in short periods of time.
Stirring such solutions only slightly reduces such losses.
preparation of Standard Aqueous Solutions of Individual PAHs
With the development of the DCCLC technique, accurate preparation of
standard aqueous PAH solutions is possible. The aqueous solubility of a com-
pound is a well defined thermodynamic quantity. "Generator columns" loaded
with single PAHs produce saturated solutions. The use of generated solutions
circumvents the problems that are usually associated with maintaining aqueous
PAH solutions, since the solutions need not be generated until they are needed.
Preparation of Aqueous Solutions Containing Several PAH Using "Generator Columns"
Although they do not generate saturated solutions, "generator columns"
coated with three or more compounds are of more practical value. The initial
attempt to prepare such a e-.olumn met with only limited success. A column that
was composed of fifty milligrams each of phenanthrene, 1-methylphenanthrene,
anthracene, and 2-methylanthracene coated on 20 grams of glass beads, was
prepared. The results of this experiment are reported in Table 20. After
6 liters of water had been pumped through this "Generator Column" (G-l),
equilibrium had not been achieved.
The same four compounds were used in a second attempt to prepare a
"generator column" that would be capable of generating a stable solution of
these four components. The amounts of each compound comprising the one per-
cent coating on the glass beads was made to be proportional to its aqueous
solubility. The results reported in Table 21 demonstrate the feasibility of
this approach. After initial conditioning, over six liters of solution was
generated with the maximum change being less than 10 percent over that range.
A third "generator column" (G-3), loaded with three compounds was pre-
pared by the same technique used to prepare G-2. The column was purged with
5000 mL of water at 25 °C after which a state of equilibrium was reached.
Although the concentrations generated by this column were not equal to the
equilibrium saturation concentrations for the individual components, they did
not ever change by more than ±5% while more than 6000 ml of water was purged
through the column (see Table 22). The concentrations of the individual com-
pounds in the effluent of this column over this interval at 25 °C were:
Anthracene 62.1 ± 1.3 yg/kg
2-Methylanthracene 13.3 ± 0.2
1,2-Benzanthracene 10.8 ± 0.3
41
-------
TABLE 20. EFFLUENT STABILITY OF "GENERATOR COLUMN" G-l
Water Purge
Volume Phenanthrene
Cone.
(ml) (yg/kg)
100
500
860
1330
4500
5350
550
6400
x (1300-6499 ml)
610
541
520
506
405
388
380
377
411
± 53
Maximum Range Between
High And Low Value 29%
TABLE 21
Water Purge
Volume
(ml)
25
550
2250
3700
5200
6320
7100
8100
x (2250-8100 ml)
. EFFLUENT
Phenanthrene
Cone.
(yg/kg)
913
897
890
906
898
888
903
876
894
± 11
1-Methyl-
Phenanthrene
Cone.
(yg/kg)
167
155
148
150
157
162
164
170
161
± 7.6
13%
Anthracene
Cone.
(ug/kg)
22.1
26.0
26.6
29.1
34.9
35.4
36.1
35.2
34.1
± 2.9
19%
STABILITY OF "GENERATOR COLUMN"
1-Methyl-
Phenanthrene
Cone.
(ug/kg)
96.3
73.6
71.8
69.9
70.8
74.4
71.0
78.1
72.7
± 3.1
Anthracene
Cone.
(wg/kg)
12.1
20.1
18.4
16.6
16.4
16.0
17.0
16.5
16.8
± 0.9
—"-••••• ,._
2-Methyl
Anthracene
Cone.
(ug/kg)
14.6
15.9
15.1
15.7
15.5
15.8
15.9
17.7
16.1
± 8.9
12%
C,-1)
2-Methyl
Anthracene
Cone.
(Pg/kg)
12.2
9.6
11.1
10.2
10.0
8.94
10.3
9.24
10.0
± 0.8
Maximum Range Between
High and Low Value
l.J
8.1%
5.1
19.5%
42
-------
TABLE 22. EFFLUENT STABILITY OF "GENERATOR COLUMN" G-3
Water Purge Volume Anthracene 2-Methyl Anthracene 1,2 Benzanthracene
200
2000
5000
6000
7200
8100
9300
9990
11500
x (5000-11,500 mL)
63.6
58.9
62.0
64.1
62.8
62.2
61.7
60.6
61.0
62.1
± 1.3
13.3
11.9
13.3
12.9
13.4
13.4
13.2
13.1
13.4
13.3
± 0.2
15.8
14.0
10.9
11.0
10.7
11.0
10.7
10.4
11.1
10.8
± 0.3
Minimum Range Between
High and Low Value 5.5% 3.0% 6.3%
Figure 8 shows that the relationship between In solubility and temperature is
linear for each of the three compounds over the 0-30 °C temperature range.
This, indicates that this solution generating process is thermodynamically
well behaved and allows one to predict the concentration of each component in
the ternary solution, given the temperature.
Preparation of Dilute Organic Standard PAH Solutions
Small volumes of dilute organic solutions may also be accurately prepared
by extracting the generated aqueous solutions with the desired volume of an
Immiscible organic solvent, such as hexane. Since PAH hexane/water distribu-
tion coefficients are very large, the concentration of these dilute solutions
can be calculated if the volume ratios are known. For example, 5 mL of a
10 (vg/fcg) PPD solution of chrysene in hexane could not be prepared gravi-
metrically. Such a solution would have to be prepared by serial dilutions of
a more concentrated solution. This process is both wasteful and hazardous.
After an aliquot is taken from the concentrated solution, a large volume of
toxic solution will remain to be discarded. However, such a solution can be
prepared by extracting 25 mL of a saturated aqueous solution (^2ppb) of
chrysene with 5 ml of hexane. This is only a hypothetical example, but it
does demonstrate the utility of using "Generator Columns" to prepare organic
PAH solutions of known concentrations indirectly.
43
-------
18.00
£ 17-50
(X 17<0°
LU
O 16.50
S 16.00
15.50
15.00
14.50
3.25
3.35 3.45
T1 (KELVIIM)xlO3
3.55
Figure 8. Concentration dependence of components of a ternary solution on temperature.
-------
evaluation of DCCLC as a Method for the Preparation and Certification of an
PAH SRM ~
There are several factors that make the DCCLC approach ideal for the prep-
aration and analysis of very dilute aqueous solutions of individual PAHs. Many
of the factors have been discussed earlier in other contexts and will be only
referred to here.
1. Saturated solutions are prepared by an equilibrium process. This
process has been shown to be reversible and PAH concentrations are independent
of the rate of flow through the "generator columns" between 0.1 and 5.0 mL/min.
2. The use of "generator columns" to produce aqueous PAH solutions
circumvents the problems that are usually associated with storing such solutions,
since the solutions need not be generated until they are needed.
3. The concentration of the generated aqueous solutions are a function
of temperature, and may be expressed in terms of least squares fits of the
concentration as a function of temperature to within ±2% of the experimentally
determined values between 5 and 30 °C (see Appendix B) .
4. Both the short and long term precision with which aqueous PAH solu-
tions can be generated appear to be <2% (see Table 8) .
5. Shelf storage of "generator columns" does not seem to present a
problem. They have been shown to be stable for longer than one and a half
years and through more than 100 liters of aqueous purge.
6. DCCLC is a rapid and accurate method for analyzing the generated
dilute aqueous PAH solutions. Analytical errors due to adsorption are mini-
mized in DCCLC because the solution is extracted and concentrated, on line,
in less than 500 ms after generation. It has been estimated that this
analytical method has an uncertainty of less than two percent.
45
-------
REFERENCES
1. Tsonopoulos, C., and J. Prausnitz, Ind. Eng. Chem. Fundam., 10:503, 1971.
2. LeFeuvre, A. Water and Water Pollution Handbook. Vol. 1, L. L. Ciaccio,
Ed. Marcel Dekker, Inc., New York, 1971. p. 263.
3. Peak, E., and G. Hodgson. J. Amer. Oil Chemists Soc., 215, 1966.
4. Ben-Nairn, A. J. Chem. Phys., 57:5257, 1972.
5. Ben-Nairn, A. Water and Aqueous Solutions. R. A. Home, Ed., (Wiley,
New York, 1972. p. 425.
6. Kites, R. Proceedings of the "Source, Effects and Sinks of Hydrocarbons
in the Aquatic Environment" Symposium, American University, Washington, DC
Aug. 1976.
7. MacKay, D., and W. Shiu. Can. J. Chem. Eng., 53:239, 1975.
8. AcAuliffe, C. J. Phys. Chem., 70:1274, 1966.
9. Sutton and Clader. J. Chem. Eng. Data, 20:320, 1975.
10. Arnold, D., C. Plank, and E. Erickson. Chem. Eng. Data Serv., 3:253
1958.
11. Bohon, R., and W. Claussen. J. Amer. Chem. Soc., 73:1571, 1951.
12. Vermillion, H. Ph.D. Thesis, Duke University, Durham, NC, 1939.
13. Stearns, R., H. Oppenheimer, E. Simon, and W. Harkins. J. Chem. Phys
14:496, 1947.
14. Hill, A. J. Amer. Chem. Soc., 44:1163, 1922.
15. Franks, F., M. Gent, and H. Johnson. J. Chem. Soc., 2716, 1973.
16. Hayashi, M., and T. Sasaki. T. Bull. Chem. Soc. Japan, 29:857, 1956.
17. Booth, H., and H. Everson. Ind. Eng. Che, 40:1491, 1948.
18. Morrison, T., and F. Billet. J. Chem. Soc, 3819, 1952.
46
-------
19. Horiba, S. Mem. Coll. Eng. Kyoto Imp. Univ. 1, 49, 1914.
20. Ftihner, H. Che. Ber..., 57:510, 1924.
21. Andrews, L., and R. Reefer. J. Am. Chem. Soc., 72:5034, 1950.
22. MacKay, D., W. Shiu, and A. Wolkoff. Gas Chromatographic Determination of
Low Concentrations of Hydrocarbons in Water, in Water Quality Parameters.
ASTM STP 573, Philadelphia, Pennsylvania, 1975.
23. Leinonen, P., D. MacKay, and C. Philips. Can. J. Chem. Eng., 49:288, 1971.
24. Davis, W., M. Krohl, and G. Glower. J. Am. Chem. Soc., 64:108, 1942.
25. Klevens, H. J. Phys. Colloid Chem., 54:283, 1950.
26. MacKay, D., and W. Shiu. Accepted for Publication in J. Chem. Eng. Data.
27. Tosonopoulos, C., and J. Prausnitz. Ind. Eng. Chem. Fundam., 10:593,
1971.
28. Pierotti, G., C. Deal, and E. Derr. Ind. Eng. Chem., 51:95, 1959.
29. Wauchope, R., and F. Getzen. J. Chem. Eng. Data, 17:38, 1972.
30. Weimer, R., and J. Prausnitz. J. Chem. Phys., 42:3643, 1965.
31. Schwarz, F. J. Chem. Eng. Data, 22:273, 1977.
32. Eisenbrand, J., and K. Baumann. Z. Lebensm. Unters. Forsch., 144:312, 1970.
33. McAuliffe, C. Chem. Tech., 1:46, 1971.
34. Brown, R., and S. Waskik. J. Res. NBS (US), 78A, 453, 1974.
35. McDevit, W., and F. Long. J. Amer. Chem. Soc., 74:1773, 1952.
36. Locke, D. J. Chromatog. Sci., 12:433, 1974.
37. Peake, E., and G. Hodgson. J. Amer. Oil Chemists Soc., 44:696, 1967.
38. May, W., S. Chesler, S. Cram, B. Gump, H. Hertz, D. Enagonio, and S. Dyszel.
J. Chromatogr. Sci., 13:535, 1975.
39. May, W., Ph.D. Dissertation, University of Maryland, College Park, MD, 1977,
pp. 34-40.
40. Setschenow, J. Z. Physik Chem., 4:117, 1889.
41. Hancock, J., H. Applegate, and J. Dodd. Atoms. Environ., 4:363, 1970.
47
-------
42. Niaussat, P., C. Auger, and L. Mallet. C. R. Acad. Sci. (PARis),
2700:1042, 1970.
43. Base, A., and R. Kites. Identification and Analysis of Organic Pollutants
in Water. Ann Arbor Science Pub., Ann Arbor, Michigan, 1976.
44. Blumer, M., and W. Youngblood. Science, 188:53, 1975.
45. Popl, M., V. Dolansky, and J. Mostecky. J. Chromatogr., 117:117, 1976.
46. Popl, M., V. Dolansky, and J. Mostecky. J. Chromatogr., 91:649, 1974.
47. Popl, M., V. Dolansky, and J. Coupek. J. Chromatogr., 130:195, 1977.
48. Wise, S., S. Chesler, H. Hertz, L. Hilpert, and W. May.
Anal. Chem. 49_, 2306, 1977.
49. Instruction Manual #4930 for CAHN Electrobalance Model "4700".
50. NBS Circular C434 (1941), US Government Printing Office.
51. The International System of Units (SI), NBS Special Publication 330.
52. Davis, H., and S. Gottlieb. Fuel, 8:37, 1962.
48
-------
APPENDIX A
PROCESSES THAT AFFECT THE ACCURACY OF DYNAMIC
COUPLED COLUMN LIQUID CHROMATOGRAPHY
This section presents a brief discussion of the major processes that affect
the accuracy of the DCCLC technique. They are:
A-l "Extractor Column" Extraction Efficiency;
A-2 Measurement of Sample Volume;
A-3 Calibration of Sample Loop;
A-4 Preparation of Standard Solutions.
This chapter also includes:
A-5 Calculation of Detector Response Factors;
A-6 Calculation of Concentration.
£_1 "Extractor Column" Extraction Efficiency
Extraction of the PAHs from the generated solutions was accomplished by
pumping volumes varying between 5.0 and 25.0 mL through the "extractor column."
Over this range, extraction efficiencies are quantitative for eleven of the
twelve compounds studied in this Investigation. Benzene was not extracted
efficiently by the extractor column. The solubility of benzene was, therefore,
determined by .direct injection of 23.2 uL of the generated solutions via a
sample loop.
The extraction efficiencies for 5, 10, and 25 mL volumes of benzene,
naphthalene, fluorene, and phenanthrene are given below.
Extraction Efficiency (%)
Compound
Benzene
Naphthalene
Fluorene
Phenanthrene
5 mL
60.0 ± 5.0
101.4 ±2.2
99.5 ± 1.0
99.5 ± 0.6
10 mL
4.3 ± 0.9
98.8 ± 1.0
98.4 ± 1.1
98.1 ± 1.5
25 mL
3.1 ± 1.2
98.5 ± 0.5
98.6 ± 1.2
98.7 ± 0.9
It has been shown elsewhere (39) that retention on, and hence the extraction
efficiency of CIQ packed columns is a function of carbon number. Extraction
of the eight larger PAHs studied in this investigation are, therefore, also
quantitative.
-------
A-2 Measurement of Sample Volume (See Figure 1)
The volume of saturated solution extracted was determined by and placing
the sample valve in the "Extract" position collecting a designated volume of the
effluent from the "Extractor Column" in a class "A" 5, 10, or 25 mL volumetric
flask. After the volumetric flask had been filled, the sample valve was
manually switched to the Analyze position, thus diverting the saturated solu-
tion to waste.
The uncertainty associated with this process was dependent on human
reflexes. One stroke of the generator pump (Milton Roy Controlled Volume mini-
Pump) is .05 mL. If we assume that the volumetric flasks could be reproducibly
filled to within one stroke of the generator pump, then the maximum uncertainty
associated with the measurement of sample volume is ± 1.0 percent (.05/5.0 mL) .
This is a very liberal estimate and is equivalent to a very slow reaction time
of 0.6 seconds at the maximum flow rate used, 5 mL/min.
A-3 Calibration of the Sample Loop
The volume of the sample loop was determined by an indirect gravimetric
procedure. The loop was initially filled with mercury. The mercury was swept
from the loop into a tarred weighing dish with approximately 1 mL of pentane.
Since mercury and pentane are not miscible, the bulk of the pentane was decanted
and the remainder was allowed to evaporate. This process was repeated four
times, yielding the following results:
Mass of Mercury displaced from loop
0.3097 grams
0.3167 grams
0.3128 grams
0.3178 grams
x 0.3143 ±1.2 percent grams.
The density of mercury at room temperature is 13.538 g/mL. The volume of the
loop was then determined to be 23.2 ± 0.3 uL. Since determination by mass is
a definitive method, the errors associated with this determination are random
and associated with the ability to reproducibly fill the loop. The accuracy is
then the precision with which the mass measurements can be made and is equal to
1.2 percent.
A-4 Preparation of Standard Solutions
Standard solutions of the aromatic hydrocarbons studied in this investiga-
tion were prepared by the Direct method. Solutes were weighed on a Cahn
density of water varies from 0.9960 g/mL at 30° to 1.0000 g/mL
at 5 °C. This represents a maximum change in mass/volume of
0.4 percent. In this investigation, it is assumed that one liter
of water has a mass of one kilogram.
50
-------
Electrobalance, dissolved In acetonitrile and made up to a 500 ml volume in a
class "A" volumetric flask.
The uncertainty involved ln_weighing the solutes was small. The balance
has a certified accuracy of SxlO"1* or 0.05 percent (49). Class "A" 500 mL
volumetric flasks have tolerances of less than 0.05 percent (50). The uncer-
tainty associated with the preparation of these solutions is estimated to be
less than 0.1 percent.
A_5 Calculation of Response Factors
Detector response factors for the 12 compounds studied in this investiga-
tion were determined by injecting 23.2 uL of standard solutions of the compounds
into the liquid chromatographic system. These response factors are presented
below. a
Response Factor
Compound , area units v
tig
Benzene 20.14 ± 0.22
Naphthalene 184.0 ±1.18
Fluorene 1071 ± 10
Anthracene 7123 ± 66
Phenanthrene 2890 ± 14
2-Methylanthracene 10647 ± 95
1-Methylphenanthrene 3042 ± 32
Fluoranthene 659.7 ± 8
Pyrene 463.0 ± 3
1,2-Benzanthracene 1479 ± 10
Crysene 2063 ± 15
Triphenylene 3081 ± 10
Calculated from areas obtained when using Perkin Elmer "System I" electronic
integrator with a chromatographic flow rate of 2.77 mL/min.
Area
Response Factor = — —
(23.2 uL) (Cone. Std.) (Purity of PAH)
A-6 Calculation of Concentration
The concentration of the aromatic hydrocarbons in the aqueous solutions was
calculated from the following equation:
51
-------
Concentration (jjg/kg or ppb) - rea
(R.F.K Sample Volume)
Area - Area under the chromatographic peak as determinded by the electronic
integrator.
R.F. - Response Factor (area/ng)
Sample Volume - Size of aqueous sample analyzed in mL (grams).
52
-------
APPENDIX B
VARIATION OF SOLUBILITY WITH TEMPERATURE
The equations relating aqueous solubility to the temperature that were
presented in Section 3 were derived from the data presented in this section.
Experimental values of the solubilities between 5 and 30 °C are given in Tables
B-l through B-12 along with the standard deviations associated with each
measurement. Each table is accompanied by a plot of a least squares fit of the
solubility to temperature.
The SI unit for concentration is moles per cubic meter (mol/m3). Mole
(mol) is the base SI unit for the amomrt of substance. Cubic Meter (m3) is a
derived unit for volume. The use of a regularly formed multiples such as cubic
centimeter and cubic decimeter are also allowed. The special name, liter (51)
has been approved for use instead of cubic decimeter, but it's use is restricted
to the measurement of liquids and gases.
The PAH solubilities reported in the tables are expressed in both the SI
unit mol/liter and metric units mg/kg or yg/kg. The concentration axes for the
least squares fits of the solubility to temperature are labeled as either parts
per million (ppm) or parts per billion (ppb). Since the density of water
between 4 and 30* is essentially constant and equal to 1.000, one liter of
water has a mass of one kilogram (kg). Thus yg/kg equals ppb and mg/kg equals
ppm.
53
-------
1,750
5.00
10.00 15.00 20.00
T {CENTIGRADE)
25.00
30.00
FIGURE B-1. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF BENZENE.
-------
TABLE B-l. BENZENE
Solubility - 0.0247t3 - 0.6838t2 + 0.3166t + 1833 (mg/kg)*
Concentration
t Measured
(°C) (mol/liter x 10~2) (mg/kg)
0.2
6.2
11.0
14.0
16.9
18.6
25.0
25.8
2.35
2.31
2.30
2.27
2.26
2.26
2.29
2.33
1836 ± 12
1804 ± 2
1799 ± 13
1770 ± 13
1762 ± 10
1767 ± 5
1791 ± 10
1819 ± 21
Concentr at ion*
Calculated
(mg/kg)
1833
1815
1787
1771
1762
1761
1799
1810
55
-------
Ul
OS
45.00
40.00
- 35.00
o_
~ 30.00
DO 25.00
O
w 20.00
15.00
10.00
A-DETERMINED BY EXTRACTION OF GENERATED SOLUTION
WITH "EXTRACTOR COLUMN"
B-DETERMINED BY DIRECT LOOP INJECTION
OF GENERATED SOLUTION
0.00
5.00
10.00 15.00 20.00
T (CENTIGRADE)
25.00
30.00
FIGURE B-2. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY
OF NAPHTHALENE.
-------
TABLE B-2. NAPHTHALENE
Solubility = 0.0189t + 0.2499t 4- 13.66 (mg/kg)*
Concentration
Measured
(mol/liter x 10'") (mg/kg)
2/.0
25.0
23.4
19.3
15.1
13.4
11.5
8.2
2.66 34.15 ± 0.47
2.49 31.91 ± 0.30
2.30 29.47 ± 0.21
2.01 25.79 ± 0.31
1.68 21.48 ± 0.11
1.59 20.37 ± 0.05
1.50 19.23 ± 0.20
1.32 16.91 ± 0.26
Concentration*
Calculated
(mg/kg)
34.16 *
31.69
29.83
25.51
21.73
20.39
19.02
16.97
57
-------
2,500
2,000
GO
QL
QL
5 1,500
E5
en
oc
O
CO
1,000
500
0.00 5.00
CH
10.00 15.00 20.00 25.00 30.00
T (CENTIGRADE)
FIGURE B-3. TEMPERATURE DEPENDENCE OF THE AQUEOUS
SOLUBILITY OF FLUORENE.
-------
TABLE B-3. FLUORENE
Solubility - 0.0185t3 + 0.4543t2 + 22.76t + 543.3 (yg/kg)*
Concentration
t Measured
(°C) (mol/liter x 10~6) (yg/kg)
31.1
27.0
24.0
18.0
13.2
6.6
13.5
11.1
9.72
7.24
5.82
4.32
2248 ± 5.0
1845 ±4.9
1616 ± 6.0
1203 ±1.1
967.3 ± 9.7
718.4 ± 0.7
Concentration*
Calculated
(wg/kg)
2246
1853
1607
1208
965.4
178.6
59
-------
0\
o
65.00
55.00
45.00
g 35.00
OQ
§ 25.00
15.00
5.00
0.00 5.00
10.00 15.00 20.00
T (CENTIGRADE)
25.00 30.00
FIGURE B-4. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF
ANTHRACENE.
-------
TABLE B-4» ANTHRACENE
Solubility = 0.0013t3 - 0.0097t2 + 0.8886t +8.21 (ug/kg)*
Concentration
Measured
Cool/liter x 10~7) (ug/kg)
28.7
24.6
22.4
18.3
14.1
10.0
5.2
3.13
2.44
2.09
1.63
1.25
0.98
0.71
55.7 ± 0.7
43.4 ± 0.1
37.2 ± 1.1
29.1 ± 0.6
22.2 ± 0.1
17.5 ± 0.3
12.7 ± 0.4
Concentration*
Calculated
(ug/kg)
55.8
43.1
37.5
29.0
22.3
17.4
12.7
-------
1,500
300
0.00
10.00 15.00 20.00
T (CENTIGRADE)
25.00 30.00
FIGURE B-5. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY
OF PHENANTHRENE.
-------
TABLE B-5, PHENANTHRENE
Solubility = 0.0025t3 + 0.8059t2 4- 5.413t + 324 (yg/kg)*
Concentration
t Measured
(°C) (mol/liter x 10"6) (ug/kg)
29.9
24.3
21.0
20.0
15.0
12.5
10.0
8.5
4.0
7.16
5.36
4.58
4.42
3.37
2.87
2.63
2.37
1.97
1227 ± 11
955 ± 1
816 ± 8
787 ± 2
601 ± 7
512 ± 1
468 ± 2
423 ± 4
361 ± 1
Concentration*
Calculation
(yg/kg)
1274
967
816
775
595
523
461
430
359
63
-------
35.00
30.00
_ 25.00
0.
Q.
ON
D
_j
O
CH,
20.00
15.00
10.00
5.00
0.00 ' J--
0.00 5.00
i
10.00
l__
15.00 20.00
T (CENTIGRADE)
25.00
30.00
FIGURE B-6. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF
2-METHYLANTHRACENE.
-------
TABLE 3-6. 2-METHYLANTHRACENE
Solubility - O.OOllt3 - 0.0306t2 + 0.8180t -1-2.79 (vg/kg)*
Concentration
Measured
(raol/liter x 10"7) (yg/kg)
31.1
27.0
23.1
18.3
13.9
10.8
9.1
6.3
1.67
1.26
0.994
0.754
0.575
0.490
0.441
0.367
32.1 ± 0.3
24.2 ± 0.1
19.1 ± 0.6
14.5 ± 0.1
11.1 ± 0.3
9.43 ± 0.37
8.48 ± 0.09
7-06 ± 0.18
Concentration*
Calculation
(ug/kg)
32.0
24.4
19.0
14.3
11.2
9.45
8.52
7.00
65
-------
400.00
350.00
_ 300.00
GO
Q_
~ 250.00
K-
QQ 200.00
_i
O
w 150.00
100.00
50.00
0.00
5.00
CH.
10.00 15.00 20.00
T (CENTIGRADE)
25.00
30.00
FIGURE B-7. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF
1-METHYLPHENANTHRENE.
-------
TABLE B~7. 1-METHYLPHENANTHRENE
Solubility • 0.0074t3 - 0.0858t2 + 5.785t 4- 62.9 (yg/kg)*
Concentration
Concentration*
t Measured
(CC) (tool/liter x 10~6) (ug/kg)
29.9
26.9
24.1
19.2
14.0
8.9
6.6
1.85
1.58
1.32
1.01
0.765
0.594
0.495
355
304
255
193
147
114
95.2
± 2
* 1
± 5
± 1
± 1
± 4
± 0.2
Calculation
(Vg/kg)
357
302
257
195
147
111
97.3
67
-------
ex
320.00
200.00
_ 240.00
CO
a.
~ 200.00
160.00
120.00
80.00
40.00
0.00
5.00
10.00 15.00 20.00
T (CENTIGRADE)
25.00
30.00
FIGURE B-8. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF
FLUORANTHENE.
-------
TABLE B-8. FLUORANTHENE
Solubility - 0.0072t3 - 0.1047t2 + 4.322t + 50.4 (yg/kg)*
Concentration
Concentration*
t Meas
(°C) (mol/liter x 10"
29.9
24.6
19.7
13.2
, 8.1
13.8
10.0
7.33
5.29
4.05
urea
7> (lig.
279.3
202.7
148.3
107.0
82.0
/kg)
± 5.9
± 0.2
± 0.1
± 0.4
± 2.1
calculation
(lig/kg)
280
201
150
106
82.3
69
-------
200.00
160.00
o_
o_
120.00
m
80.00
40.00
00.00 5.00
10.00 15.00 20.00
T (CENTIGRADE)
25.00 30.00
FIGURE B-9. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF PYRENE.
-------
TABLE B-Q. PYRENE
Solubility - -O.OOllt3 + 0.2007t2 - l.OSlt + 50.2 (vg/kg)*
Concentration
Measured
(mol/liter x l(f7) (vg/kg)
29.9
25.5
21.2
18.7
14.3
9.5
4.7
8.39
6.73
5.37
4.61
3.56
2.89
2.43
170 ± 1
136 ± 2
109 ± 1
93.3 ± 1.0
72.0 ± 1.0
58.5 ± 0.6
49.2 ± 0.1
Concentration*
Calculate .1
(yg/kg)
170
136
108
93.7
73.1
57.4
49.5
71
-------
13.00
10.00
05
CL
CL
7.00
ffi
O
CO
4.00
1.00
0.00 5.00
i
10.00 15.00 20.00
T (CENTIGRADE)
i
25.00
30.00
FIGURE B-10. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY
OF 1, 2 BENZANTHRACENE.
-------
TABLE B-1CL 1,2-BENZANTHRACENE
Solubility = O.OOOSt3 - 0.0031t2 + 0.1897t + 1.74 (yg/kg)*
Concentration
Concentration*
t Measured Calculation
(°C) (mol/liter x 10~8) (wg/kg) (ug/kg)
29.7
23.1
19.3
14.3
10.7
6.9
5.58
3.67
2.77
2.10
1.66
1.31
12.7 ± 0.2
8.37 ± 0.03
6.33 ± 0.02
4.79 ± 0.02
3.78 ± 0.05
2.99 ± 0.10
12.8
8.28
6.47
4.72
3.79
3.00
73
-------
2.25
2.00
o.
a.
O
CO
1.50
1.25
1.00
0.75
0.50
0.00
5.00
25.00
30.00
10.00 15.00 20.00
T (CENTIGRADE)
FIGURE B-11. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY
OF CHRYSENE.
-------
TABLE B-ll. CHRYSENE
Solubility = 0.0024t - 0.0144t + 0.69 (yg/kg)*
28.7
25.3
24.0
20.4
11.0
6.5
Concentration
Measured
(mol/liter x 10"9) (vg/kg)
9.68
8.28
7.36
6.13
3.50
3.10
2.21 ± 0.02
1.89 ± 0.03
1.68 ± 0.03
1.40 ± 0.02
0.80 ± 0.02
0.71 ± 0.02
Concentration*
Calculation
(yg/kg)
2.23
1.84
1.71
1.38
0.82
0.69
75
-------
9.00
15.00
20.00
T (CENTIGRADE)
25.00
30.00
FIGURE B-12. TEMPERATURE DEPENDENCE OF THE AQUEOUS SOLUBILITY OF TRIPHENYLENE.
-------
TABLE B-12. TRIPHENYLENE
Solubility = -0.0002t3t + 0.0250t2 - 0.4250t + 4.89
28.2
27.3
20.5
14.8
12.0
8.0
Concentration
Measured
(mol/liter x ICf8) (yg/kg)
3.55
3.35
2.14
1.49
1.33
1.18
8.11 ± 0.11
7.65 ± 0.09
4.89 ± 0.05
3.39 ± 0.06
3.03 ± 0.02
2.99 ± 0.08
Concentration*
Calculation
(yg/kg)
8.10
7.67
4.88
3.40
3.03
2.98
77
-------
TECHNICAL REPORT DATA
(Please read Instructions on tin1 m-ene ticjorc completing)
\. REPORT NO.
EPA-600/7-80-031
The"
3. RECIPIENT'S ACCESSION NO.
«. TITLE AND SUBTITLE j no Development of an Aqueous Trace
Organic Standard Reference Material for Energy Related
Applications: Investigation of the Aqueous Solubility
Behavior of Polycyclic Aromatic Hydrocarbons
5. REPORT DATE
1980
. PERFORMING ORGANIZATION CODE
1. AUTHOR(S)
Willie E. May
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
National Bureau of Standards
Washington, DC 20234
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U. S, Environmental Protection Agency
Office of Research § Development
Office of Energy, Minerals § Industry
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Interim - Nov. 76- Mav 78
14. SPONSORING AGENCY CODE
16. SUPPLEMENTARY NOTES
This project is part of the EPA-planned and coordinated Federal Interagency
Energy/Environment R§D Program.
16. ABSTRACT
The development of a Standard Reference Material for aqueous
solutions of knov.Ti concentration of polynuclear aromatic hydrocarbons
is an extremely difficult procedure. This paper is one of a series
discussing the development of a generator column technqiue at NBS
for the production of Standard Reference Materials for PAH's in
water. In addition to providing the basis for SRM development
the aqueous solubility is a fundamental parameter in assessing the
extent and rate of the dissolution of energy based polynuclear
aromatic hydrocarbons and their persistence in the aquatic environment.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSVTI } icIJ/GroUp
18. DISTRIBUTION STATEMENT
19. SECURITY CLAl>S i InisHepurt)
21. NO. OF PAG(-S
87
20. SECURITY CLASS (This page)
22. PRICE
$8.00
EPA Form 2220-1 (» 7J)
------- |