A'"
f i
-'.' ' I
ENVIRONMENTAL ASSESSMENT OF POLYMERS
UNDER THE U.S. TOXIC SUBSTANCES CONTROL
ACT
Robert S. Boethling1* and J. Vincent Nabholz2
1 Chief, Environmental Fate Section, Exposure Assessment Branch, Economics, Exposure and
Technology Division-7406; and 'Senior Biologist, Environmental Effects Branch, Health and
Environmental Review Division-7403, Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency, 401 M St., S.W., Washington, D.C. 20460-0001
"To whom correspondence should be addressed
-------
-------
BACKGROUND
In the U.S. the safety of specific chemical substances is evaluated
primarily under three statutes. Substances used as food additives,
drugs, and cosmetics are registered by the Food and Drug
Administration (FDA) under the Federal Food, Drug, and Cosmetic Act
(FFDCA). Chemical substances proposed for use as pesticides are
registered by the Environmental Protection Agency (EPA) under the
Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), which
imposes a host of data requirements for any submitter seeking to
register the substance as an active ingredient. For the vast
majority of polymers, however, the Toxic Substances Control Act
(TSCA) is the applicable statute. TSCA (Public Law 94-469) was
enacted by Congress in 1976, in response to a perceived need to
limit exposure to industrial "environmental chemicals" such as
polychlorinated biphenyls (PCBs). As stated in the Act, its
primary purpose was
"...to assure that...innovation and commerce in...
chemical substances and mixtures do not present an
unreasonable risk of injury to health or the envi-
ronment." (TSCA, section 2(b))
As defined in TSCA, "chemical substances" specifically exclude
substances already regulated under FFDCA and FIFRA (unless they
have non-FFDCA or non-FIFRA uses) as well as alcohol, tobacco, and
certain other materials; but all other substances are included.
Thus, most polymers used in water treatment, coatings, household
laundry products, and manufactured goods are subject to the
requirements of TSCA.
These requirements are different for existing substances and
substances not yet in production, i.e., new chemicals. One of the
first tasks of the newly created Office of Toxic Substances (now
the Office of Pollution Prevention and Toxics or OPPT) was in fact
to assemble and publish a list of chemical substances already in
commerce. This was accomplished by July, 1979 as the TSCA Chemical
Substance Inventory, commonly referred to simply as the Inventory,
which listed approximately 50,000 substances then in production or
being imported into the U.S. Since that time the Inventory has
grown to include over 70,000 substances by the addition of new
chemicals.
Anyone who wishes to manufacture or import into the U.S. for \
commercial purposes a substance not listed on the Inventory and not
otherwise excluded by TSCA (pesticides, drugs, etc.) must submit
formal notification of their intent to do so. Such a submission is
called a Premanufacture Notice (PMN), and for most new chemicals it
must be submitted to EPA at least 90 days prior to manufacture or
import. New chemicals not regulated by EPA are then added to the
Inventory and become existing chemicals when EPA receives a
required Notice of Commencement (NOC). The NOC declares the
-------
submitter's intent to commence manufacture or import. Since 1979
EPA has received more than 30,000 valid PMNs (Fig.
submissions currently average well over 2,000 per year.
THE NEW CHEMICAL REVIEW PROCESS
1) , and
Receipt of a PMN sets in motion a process of review that has
evolved since July 1979 to meet the unique requirements established
by TSCA. The fundamental purpose of PMN review is stated in the
law and is to determine whether
"...the manufacture, processing, distribution in com-
merce, use, or disposal [of a new chemical substance]
or any combination of such activities presents or may
present an unreasonable risk of injury to health or
the environment..." (TSCA section 5(b))
However, TSCA imposed serious challenges to EPA's ability to
accomplish this. The most significant are, first, that PMN
submitters are only required to furnish health and safety studies
already in their possession (if any) , and are not required to
conduct any testing as a precondition for notification; and
second, that the review must be done within 90 days. The '90-day
review period may be extended to 180 days under certain
circumstances. If no action to regulate the substance is taken by
EPA, the submitter is free to commence manufacture or import after
expiration of the PMN review period. The burden of proof that a
substance presents or may present an unreasonable risk rests on the
Agency's shoulders. EPA must make sound decisions based on few or
no submitted test data within 90 days. This situation is obviously
unlike that for pesticides in the U.S., but is also unlike that for
new commercial chemicals in the European Union, where a series of
tests that supply the "minimum premarket data set" are prescribed
by law.
Risk is defined as the probability of occurrence of an adverse
health or environmental effect associated with exposure to the
substance. OPPT's new chemical review program (NCP, USEPA 1995d.)
is, therefore, at its core a risk assessment process. According to
the National Research Council (NRC) risk assessment consists of
four components: hazard (or effect) identification, dose-response
assessment, exposure assessment, and risk characterization (NRC
1994) . The NCP in OPPT includes these steps, but generally does
not adhere to the exact standards set by the NRC due to the
aforementioned lack of data. In addition, OPPT's determination of
which risks are "unreasonable" also includes assessment of relative
risk, i.e., comparison of relative hazards of the PMN substance and
similar existing substances, and certain non-risk factors.
Foremost among these are economic factors such as the costs or
benefits of the new chemical, the cost of any additional testing
that may be required, and the economic impact of testing or
regulation on the submitter; and the pollution prevention
-------
potential (if any) associated with manufacture and use of the new
chemical.
Fig. 2 outlines the NCP in OPPT. More detailed information
about the NCP can be obtained from Nabholz' et al. 1993a, Wagner et
al. 1995, and Moss et al. 1996. As a result of the review a new
chemical substance may be:
Dropped from review because there is low concern with
respect to its toxicity towards humans and the
environment;
Dropped from review because there is low risk with
respect to its manufacture, distribution, use, or
disposal;
Dropped from review, but with specified concerns
related to its manufacture, distribution, use, or
disposal which are communicated to the submitter via
letter and/or a significant new use rule (SNUR,
USEPA 1995a);
Regulated due to either potential risk or potential
significant release to the environment after the
results of an initial review;
Subjected to either (!) detailed review ("standard
review") or (2) an immediate request to the submitter
for testing and/or additional information to determine
if the substance should be regulated. Immediate
testing and/or information can only be requested if the
chemical belongs to a Sec. 5(e) chemical category
(Moss et al. 1996); or
Regulated or dropped from further review as a result of
the standard review or the additional information
and/or testing.
Section 5(f) of TSCA grants EPA the authority to take
immediate action if there is a reasonable basis to conclude that a
new chemical's manufacture, processing, use," or disposal will
present an unreasonable risk. However, in practice, 5(f) findings
are seldom if ever made. Regulation under a sec. 5(e) Consent
Order, issued when it is found that a substance may present an
unreasonable risk, is much more common. Through a Consent Order,
the manufacturer or importer of a new chemical consents to the
order's requirements in exchange for being permitted to manufacture
or import the substance. Section 5(e) findings are always
predicated on insufficient information to adequately assess risk,
and typically include requirements such as:
Use of protective equipment; e.g., gloves, respirators,
-------
and goggles for workers;
Use restrictions;
Restrictions to certain manufacturing, processing, use,
and disposal sites;
* Warning labels and MSDS statements; and
Testing to resolve uncertainties regarding toxicity or
exposure.
Testing is a frequent requirement under sec. 5(e), but more
specific information about releases and uses are also requested.
If exposures can be controlled and potential risks sufficiently
reduced, then the testing is generally triggered such that it is
only required if and when a predetermined production volume
{identified in the economic analysis) is reached. If exposures
and, thus, potential risk cannot be controlled, then "upfront"
testing (i.e., not triggered) is generally required.
Section 5(e) findings may be either risk-based or exposure-
based. Prior to 1988 nearly all regulations under 5(e) were mainly
risk-based, which in practice means that more emphasis is placed on
toxicity than on exposure in the "may present" finding. In 1988
OPPT implemented an alternative strategy, exposure-based review
(XBR), that emphasized substantial or significant human exposure or
substantial release to the environment. Criteria that defined the
terms substantial and significant were developed for production
volume, worker exposure, consumer exposure, ambient and general
population exposure, and environmental release both total releases
to the environment and releases to surface waters (Table 1). EPA
estimates these parameters based on data submitted in the PMN as
well as its own databases and models. If appropriate criteria are
met, the outcome is usually an exposure-based 5(e) Consent Order
for health effects, environmental toxicity, and/or environmental
fate testing.
Full or partial exemption from review may also be granted
for new chemicals meeting certain requirements. EPA has limited
reporting requirements for:
Substances manufactured in small quantities solely for
research and development, as long as special procedural
and recordkeeping requirements are met;
Substances submitted as Test Market Exemption (THE) re-
quests. TMEs undergo expedited (45-day) review, since
the exposure assessment generally considers only the
volume, number of customers, and period of time speci-
fied in the notice;
-------
6
Substances to be manufactured in quantities of 10,000
kg or less per year, which are submitted as Low Volume
Exemption (LVE) requests (USEPA 1995b, see 40 CFR part
723). Low Volume PMNs undergo an abbreviated (30-day)
review and, like TME requests, the LVE is either granted
or denied. LVEs are not added to the Inventory but are
maintained in a separate list;
Substances expected to have low release to the
environment and low exposures (LoREx exemption, USEPA
1995b, see 40 CFR part 723) , which also undergo an
abbreviated (30-day) review and are either granted or
denied; and
Certain classes of polymers that are not chemically
active or bioavailable (USEPA 1995c and 40 CFR part
723); these are discussed in more detail in the
following section.
Chemistry review
The PMN review process can be separated conceptually into four
phases: chemistry review, toxicity evaluation, exposure
assessment, and risk assessment/risk management. The chemistry
review phase is the first step and begins as soon as the PMN has
been checked for completeness of required information, which
includes all available data on chemical identity, production
volume, byproducts, proposed use(s), environmental release,
disposal practices, and human exposure. The first step in the
chemistry review is to check the adequacy of the submitted chemical
name. EPA now requires the chemical name to be consistent with
Chemical Abstracts Service (CAS) nomenclature policies as well as
names used for similar substances already on the Inventory.
Consistency between chemical name, chemical structure, and the
chemical's manufacturing process is assured. A full report on the
chemistry of the new substance, containing submitted information on
the chemical identity, route of synthesis, impurities or byproducts
of the synthesis, and some physical/chemical properties, is then
prepared.
For most PMNs, however, the Agency is unable to conduct a
meaningful review based on submitted data alone. In large part
this is due to TSCA's stipulation that only data "known to...or
reasonably ascertainable by" (i.e., already in the possession of)
the submitter must be provided to EPA. The problem of missing data
is manifested in the earliest phases of review, as even the most
basic information on properties such as melting point and boiling
point temperatures and vapor pressure is often absent. In one
study of submitted PMN data, for example, only 300 chemicals with
submitted data for any of several physical/chemical properties
important in environmental assessment were identified from 15,000
PMNs for the period 1979-1989 (Lynch et al. 1991). Although many
-------
of the 15,000 were polymers, at least half of these submissions
were for class 1 substances, i.e., single compounds composed of
particular atoms arranged in a definite, known structure.
Polymers are class 2 substances, meaning that they have
variable compositions and/or are composed of complex combinations
of reactants, such that only a representative molecular structure
can be drawn. Properties such as melting point and boiling point
temperature are relevant for discrete (class 1) substances but,
except for water solubility or water dispersibility, have little
meaning for polymers. For polymers the chemistry review focuses
on:
The monomers of which the polymer is composed and their
mole percentages;
The molecular weight (MW) distribution, including the
number-average molecular weight (MWn) , how it was
determined, and the oligomer content of the polymer,
i.e., percentage of oligomers with a molecular mass
less than 1,000 and 500 daltons;
The equivalent weight of any reactive functional
groups and/or cationic charge density; and
Properties, such as, physical form, particle size
distribution, swellability, water solubility, and water
dispersibility.
\
Water solubility, water dispersibility, HW, charge, and cationic
charge density are the most important physical properties for
aquatic toxicity assessment, but basic information is often missing
or is not reported in a way that is most useful for risk
assessment. For example, PMN submitters are not required to report
typical MWn values, but this is often quite useful, especially if
the typical and lowest values are far apart. In addition, MW
values given as greater than or less than some MW are not very
helpful unless the number indicates the actual MW (e.g., >10,000
daltons is not helpful if the actual MW is more accurately
described as >100,000 daltons). Finally, monomer composition is
sometimes incomplete or the structural diagram fails to show the
most likely types of linkages between monomers. For example, a
random reaction between monomers rather than a blocked reaction
between monomers is a very important distinction that most
submitters fail to make explicit. Monomer linkage is particularly
important information for polymers claimed to be biodegradable.
Based on its experience from reviewing over 10,000 PMNs for
polymeric substances, EPA has identified a group of polymers that
it believes poses low to no unreasonable risk of harm to human
health or the environment due to, generally, low toxicity. As of
30 May 1995 (USEPA 1995b and 40 CFR part 723) these polymers became
-------
fully exempt from reporting. To qualify for this exemption the
polymer (USEPA 1996) must:
Belong to one of 12 acceptable polymer classes: poly-
esters, polyamides and imides, polyacrylates, poly-
urethanes and ureas, polyolefins, aromatic polysul-
fones, polyethers, polysiloxanes, polyketones, aro-
matic polythioethers, polymeric hydrocarbons, and
phenol-formaldehyde copolymers;
Oligom'er content must be less than 25 percent by weight
below 1,000 daltons and less than 10 percent by weight
below 500 daltons;
Have no more than the permissible level of cationic
character which is a functional group equivalent
weight for cationic groups j>5,000 daltons;
Have no reactive functional groups, or only reactive
functional groups specifically allowed based on OPPT's
risk assessment experience, e.g., blocked isocyanates,
or a reactive functional group equivalent weight no
less than a defined threshold, e.g., for pendant meth-
acrylates, the equivalent weight threshold is 5,000
daltons; and
For polymers with MWn >10,000 daltons, must not be
capable of absorbing their weight of water.
Since the majority of polymers with aquatic uses, e.g., polymeric
flocculants used in water treatment, are cationic in character,
they are not eligible for exemption under the polymer exemption
criteria.
t
Environmental assessment; exposure
For more than 99 percent of new chemicals the focus of
environmental assessment under TSCA is on the aquatic environment.
Contamination of surface water is possible for landfilled chemicals
having certain properties, e.g., high water solubility and low
biodegradability, if the landfill is in hydrologic contact with
ground water. Since ground water constitutes a pathway for
potential human exposure via drinking water, it is routinely
evaluated along with.other potential exposure pathways. However,
for polymers it is generally assumed that releases to landfills and
deep well injection do not result in significant aquatic- or
terrestrial-ecological exposures.
Environmental risk assessment under TSCA is accomplished using
the quotient method (Nabholz 1991, Nabholz et al. 1993, Rodier and
Mauriello 1993) . The quotient method compares a concern
concentration (CC; see Environmental assessment; hazard) or
-------
9
effective concentration (EC) to an actual or predicted
environmental concentration (PEC). PECs are determined from the
amount of chemical released, the extent of removal (if any) in
wastewater treatment, the extent of dilution of the wastestream by
the receiving environment, and the fate of the substance post-
release. Depending on the specific substance, release to the
environment may result from the manufacture, processing, or
industrial use of the substance; from commercial use; or from
consumer use of household products disposed of down the drain.
Except for release from consumer uses, available release data are
summarized in an Engineering Report prepared prior to the
environmental exposure assessment. Releases are expressed in terms
of kilograms (kg) of the chemical substance released per day over
a specified number of days per year, for each specific site of
release if the sites are known and for generic sites otherwise.
Release data transmitted via the Engineering Report do not include
consideration of wastewater treatment either on-site or in
publicly-owned-treatment works (POTWs or sewage treatment plants),
which is factored into the assessment in a subsequent step.
The proper starting point for any realistic assessment of
environmental exposure is a sound understanding of the substance's
physical and chemical properties. To fill the gaps created by
missing data OPPT uses the Estimation Programs Interface (EPI) to
access computerized estimation methods for melting point, boiling
point and vapor pressure (MPBPVP); water solubility (WSKOW; Meylan
et al. 1995) ; octanol/water partition coefficient (LOGKOW; Meylan
and Howard 1995); Henry's Law constant (HENRY; Meylan and Howard
1991) ; and soil/sediment sorption coefficient (PCKOC; Meylan et al.
1992). Also used to predict the octanol/water partition
coefficient is the computer program, CLOGP (Leo and Weininger
1985). These methods are based on 'fragment contribution and
fragment-based correction factor approaches, and provide state-of-
the-art estimates for most class 1 substances.
For both class 1 and class 2 substances it is generally
assumed that water releases are subjected to primary treatment by
gravitational settling and activated sludge secondary treatment, at
a minimum. If additional or specialized treatment processes are
employed, this information is also considered in the assessment if
sufficient data are provided. A frequent shortcoming of PMK
submissions is that they contain detailed information on the
treatment system, but only general information on efficiency. For
example, removal of biochemical oxygen demand (BOD) in on-site
treatment may demonstrably exceed 98% in order to satisfy water
permit requirements, but usually there is no way to determine
actual removal of the PMN substance itself from this information.
Whatever the characteristics of the wastewater treatment system,
efficiency of removal nearly always must be estimated for new
chemicals. For class l substances a fugacity-based multimedia
model (STP; Clark et al. 1995) is often used to provide removal
estimates that integrate the key removal processes of sorption, air
-------
10
stripping, and biodegradation. However, the model is neither
useful nor is it needed for most polymeric substances, since
sorption and/or precipitation are likely to be the only significant
processes.
The final step in the exposure assessment process is to
calculate a PEC based on post-treatment release of the PMN
substance. In initial review, if discharge is to a water body with
unidirectional turbulent flow, i.e., a river or stream, surface
water concentrations are usually calculated using a simple dilution
model and fate is considered qualitatively. More complex aquatic
ecosystem models such as EXAMS II (USEPA 1985), in which site-
specific ecosystem characteristics as well as transport and
transformation parameters can be quantitatively incorporated, are
employed only in detailed review ("standard review"; Fig. 2 and
USEPA 1994a). If the location of the release site is known, the
stream flow for the specific river reach to which the facility
discharges is used. PECs are calculated for mean and low flow
conditions using U.S. Geological Survey (USGS) data. If the
location is not known or if multiple release sites are possible,
mean and low flows from the stream flow distribution for industrial
facilities in the most relevant Standard Industrial Classification
(SIC) codes are used (50th and 10th percentile flows for any
distribution) . If this analysis results in PECs that exceed the
CC, the number of days of exceedance per year is estimated using
the Probabilistic Dilution Model {PDM; USEPA 1988). In addition to
the CC the model requires input data on kg released per day, number
of days of release per year, and either location of specific
site(s) or relevant SIC code. PDM performs a simple dilution
calculationi.e., does not account for transport or transformation
processes after releasebut does account for natural variability
in stream and effluent flow by incorporating a probability
distribution, i.e., flows are assumed to be log-normally
distributed, into the calculations.
Environmental assessment: toxicity
The purpose of environmental toxicity assessment is (1) to j
identify as many of the potential effects of a chemical substance
towards organisms in the environment as possible and (2) to predict
the potency of each effect. Environmental toxicity data are
usually expressed as effective concentrations (EC) of the
substance, which indicate both the type and seriousness of the
effect. The compilation of available ECs constitutes the hazard
profile or the toxicity profile of the substance, and for new
chemicals it contains, at a minimum, six effects with effective
concentrations based on 100% active ingredients (ai):
Fish acute value (96-h LC50) ;
Aquatic invertebrate (usually Daphnia macma) acute
value (48-h LC5Q) ;
Green algal toxicity value (96-h EC50) ;
US EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue N\ V
Washington DC 20460
-------
11
Fish chronic value (ChV) from an early life stage
toxicity test;
Aquatic invertebrate ChV from the daphnid partial life-
cycle toxicity test; and
Green algal ChV value from the 96-h toxicity test.
Since OPPT considers that the 96-h algal toxicity test also
constitutes a test for chronic toxicity, the geometric mean of the
96-h lowest-observed-effect concentration (LOEC) artd the 96-h no-
observed-effect concentration (NOEC) is considered to be the algal
ChV if the LOEC is less than the 96-h EC50. The ECs in a
substance's toxicity profile can be either measured or predicted
using structure activity relationships (SAR). Measured values are
preferred, but since less than 4.8% of PMNs contain submitted
environmental toxicity data (Nabholz et al. 1993a) , ECs must be
predicted for the great majority of substances (Wagner et al.
1995). Over 95% of the toxicity profiles for new chemicals under
TSCA consist entirely of predicted ECs. When ecotoxicity data are
submitted with PMNs, they are mostly for acute toxicity to fish.
The need to predict the toxicity to chemicals based on their
structure and properties has led to the development of a variety of
methods collectively referred to as structure activity
relationships (SAR). SARs used in OPPT are always quantitative in
that they are used to predict the toxic potency of a chemical or
the EC. Predictions are obtained through the use of two general
SAR methods: nearest analog analysis or mathematical equation.
SAR analysis via nearest analog analysis compares a chemical with
one or more analogs. A prediction is obtained by either
interpolation between analogs, i.e., geometric mean, or
extrapolation from one or more of the analogs. SAR analysis via
mathematical equation generally involved correlation of chemical
structure or physical properties as surrogates for structure to
measured toxicity values via statistical regression analysis. The
chemicals in.a SAR are homologous chemicals, generally, from the
same chemical class. SARs for environmental toxicity now exist
(Clements and Nabholz 1994, Clements 1988, and Clements et al.
1993) for many classes of substances though by no means all
substances subject to TSCA review, and they have been immensely
useful in assessing new chemicals. Most SARs for ecotoxicity have
been developed from data on class 1 substances; SARs for class 2
substances were more difficult to develop. SARs are available for
all major classes of polymers. Validation of these SARs is an
ongoing process within OPPT and has been successful (Nabholz et al.
1993b, and USEPA 1994b).
Determination of a CC is the last step in the toxicity
assessment process for aquatic toxicity. A CC is that
concentration of the substance which, if exceeded in ' the
environment, may cause a significant risk of harm. If the exposure
assessment does not yield PECs that exceed the CC, it is assumed
that the probability of significant environmental risk from
-------
12
exposure to the substance is too low to warrant direct regulatory
action, for example to limit release. The CC for aquatic life is
homologous to the Organization for Economic Co-operation and
Development's (OECD) and European Union's (EU) predicted-no-effect
concentration (PNEC) and analogous to a reference dose (RfD) in
risk assessment for human health (Barnes and Dourson 1988). The CC
is expressed in the same units as for PECs; i.e., milligrams or
micrograms per liter. It is calculated from the ECs in the
toxicity profile by the application of assessment factors (AsF),
which are homologous to uncertainty factors (Health Council of the
Netherlands 1989) but less similar to margins of safety or exposure
(USEPA 1990).
In general, AsFs are used to adjust ECs downward in order to
account for the amount and quality of data in the toxicity profile
and their relevance for assessment of chronic toxicity under actual
environmental conditions. For example, an AsF of 1,000 is used
when the toxicity profile contains only one acute toxicity value.
This EC is divided by 1,000 to derive the CC. If data are
available from several acute but no chronic toxicity tests a factor
of 100 is used, and a factor of 10 is'used when the ChV is known
from laboratory test data for the presumed most sensitive species,
but field data are not available to confirm the ChV. The CC is set
equal to the ChV, i.e., the AsF is \, only when the ChV is derived
from field data or, in some cases, microcosms and/or mesocosms.
Currently, if the exposure assessment identifies endangered species
that exist in the area where release is expected (OPPT has linked
its database for river flows for specific streams and reaches with
an endangered species database), then the CC is adjusted downward
by another factor of 10 for comparison to the PEC.
TESTING
Environmental toxicity
Reliable test data are the foundation of risk assessment and
should always be considered preferable to predictions based on SARs
or other estimation methods. In environmental toxicity assessment,
the issue of which tests should constitute a minimum data set
("base set") or first tier of testing surfaced at an early stage in
the implementation of TSCA (Zeeman and Gilford 1993). For aquatic
exposures the base set of tests corresponds to the acute toxicity
attributes in the toxicity profile for new chemicals, and consists
of:
Fish acute toxicity test;
Aquatic invertebrate acute toxicity test;
Green algal toxicity test.
and
For terrestrial exposures the base set consists of a rodent acute
oral toxicity test, plant early seedling growth test, an earthworm
acute toxicity test, a soil microbial community toxicity test, and
-------
13
an avian acute oral toxicity test. The aquatic base set is
frequently invoked under TSCA as tier one testing for both new and
existing chemicals, and some or all of the tests may be mandated
depending upon the specific data requirements for the substance
under review. Higher tier tests consist of subchronic and chronic
toxicity tests, toxicity tests using contaminated sediments,
microcosms, and field studies (Smrchek et al. 1993).
OPPT's aquatic toxicity test guidelines as well as those of
other groups such as the OECD, EU, Standard Methods, and the
American Society for Testing and Materials (ASTM) were designed to
generally apply to class l substances. These guidelines typically
employ "clean dilution water" test systems with low levels of
dissolved organic carbon (DOC) and total suspended solids (TSS) in
fish and invertebrate tests and growth media with low water
hardness in algal toxicity tests. However, studies (Gary et al.
1987 and Nabholz et al. 1993a) have shown that the toxicity of
certain types of polymers may be affected by water hardness or DOC
concentrations, which also vary with environmental conditions.
Toxicity testing for polymers has therefore been modified to
address the effects of DOC and hardness and the realistic
conditions encountered in the natural environment. This is
discussed more thoroughly in ENVIRONMENTAL CONCERNS FOR POLYMERS.
Environmental fate
The environmental behavior of chemical substances is highly
variable and ultimately reflects the controlling influence of
molecular structure. Although it is true that among the many
physical properties of chemicals that might be measured some are
more important than others, the most Important parameter(s) to
measure is not necessarily the same for all substances. Thus from
an environmental perspective vapor pressure may be the most
critical property for one substance, whereas for another it may be
the substance's water solubility. Transformation processes, e.g.,
biodegradation and hydrolysis, are even more varied and in general
are also substantially affected by environmental conditions. For
these reasons it has not been possible to develop a fixed yet still
affordable set of tests for physical/chemical properties and
environmental fate under TSCA.
Testing to obtain data on key properties and transformation
processes is sometimes required in risk-based 5(e) Consent Orders.
In such cases the nature of the testing generally depends upon the
exposed populations and routes of exposure, and the outcome of
required testing is often used as a trigger for toxicity testing.
For example, if a high concern for aquatic toxicity suggests a need
for base set testing to resolve uncertainty regarding ecotoxicity,
the latter nevertheless may be triggered only if the PMN substance
is not readily biodegradable as determined by an appropriate OECD
biodegradation test. As another example, if a compound's water
solubility or vapor pressure strongly influences the likelihood of
-------
14
exposure via some specific route, health or environmental effects
testing may be made contingent upon determination of relevant
properties.
A fate testing strategy designed to account for the complexity
of environmental transport and transformation phenomena was
developed for application in exposure-based PMN review (XBR).
Since most polymers used in cleaning products, as wastewater
flocculants, and in industrial water treatment are expected to be
released to water in significant volumes, it is likely that the XBR
fate testing strategy will apply to them. The strategy borrows
some aspects of the "base set" approach in ecotoxicological testing
but is also intended to be flexible. Aspects relevant to aquatic
exposures are outlined in Table 2. An important feature is that
EPA generally limits required fate testing to one or two tests for
substances with estimated production volumes >100,000 kg/year,
whereas if the volume is >l,000,000 Ib/year three or more tests may
be required if the data are deemed essential.
Existing OPPT test guidelines in the principal areas that the
XBR fate strategy addresses are primarily intended to apply to
class 1 substances, not polymers. Nevertheless, the strategy is
interpreted broadly by OPPT and acceptable test protocols do not
need to be based strictly on existing published guidelines. For
example, data from an appropriate jar test (ASTM 1994) or other
laboratory test designed to measure performance of sedimentation
aids in wastewater clarification (Halverson and Panzer 1980) may be
more relevant for determining fate of a polymeric substance in
wastewater treatment. It is prudent for the chemical manufacturer
or importer to submit the protocol to OPPT for review and approval
before testing begins.
FATE ASSESSMENT FOR POLYMERS
Electronic charge, MW, and solubility/dispersibility in water
are the most important properties of polymers for fate assessment.
Charge density, i.e., number of charges per unit of MW, may also be
important but probably plays a greater role in toxicity assessment
(see ENVIRONMENTAL CONCERNS FOR POLYMERS). Polymers are
conveniently and pragmatically divided into four classes based on
charge: anionic (negative charge), cationic (positive charge),
amphoteric (positive and negative charges present on the same
molecule) and nonionic (electronically neutral).
The vast majority of synthetic polymers are essentially
nonbiodegradable, a fact that has been known for many years (for
example, see Alexander 1973). Even modified natural polymers such
as carboxymethylcellulose and cellulose acetate butyrate having an
appreciable degree of substitution are not significantly
biodegradable. Some modified natural polymers and synthetic
polymers with ester linkages or other labile groups incorporated
into the main chain of the polymer are biodegraded under favorable
-------
15
conditions. Examples include poly(/3-hydroxybutyric acid), a
naturally occurring bacterial energy storage product, poly(6-
caprolactone), polyfglyoxylic acid) and poly(tetramethylene
succinate) (Bailey and Gapud 1986, Park et al. 1989, Pranamuda et
al. 1995). However, rates of biodegradation that may be
significant relative to typical residence times in terrestrial
environments (especially landfills) may not be so for wastewater
treatment systems, where retention times are much shorter. Thus
even so-called biodegradable polymers are not necessarily
significantly biodegraded in sewage treatment.
All PMN substances, including polymers, are evaluated for
their biodegradability based on molecular structure if submitted
data are insufficient. In practice, the few biodegradable polymers
among the many recalcitrant ones are easily identified because the
polymer's biodegradability is usually a major reason for the
submitter's desire to market the substance. In these cases the PMN
generally contains experimental data on biodegradability and this
characteristic figures prominently in a Pollution Prevention (PP)
Claim (see THE NEW CHEMICAL PROCESS).
Since air stripping is obviously not a major loss mechanism
for polymers the assessment of removal of polymers in wastewater
treatment essentially reduces to an assessment of their likely
removal by sorption and/or precipitation. In review of new
chemicals under TSCA, treatability of a polymer discharged to
wastewater treatment is inferred from the polymer's charge, MW, and
solubility/dispersibility unless actual data accompany the PMN
(rarely the case). It is generally assumed that the sequence of
treatment processes is limited to primary clarification followed by
activated sludge secondary treatment, which is typical for POTWs.
Nonionic, cationic, and amphoteric polymers with MWn >1,000 daltons
are assumed to partition mainly to the solids phase and be 90%
removed relative to total influent concentration. The 90% figure
was selected because it represents a typical level of solids
removal in POTWs (USEPA 1982) . The remaining fraction (10%) is
thus assumed to be discharged to receiving waters, although of
course it is likely that this material is in the form of polymer
sorbed to sludge solids (organic matter). For polymers of 500 <
MWn < 1,000, a lower removal rate (typically 50 to 90%) may be
assumed depending on the polymer's structure and properties.
Anionic polymers with negligible water solubility and
dispersibility are assumed to behave similarly to nonionic
polymers, but lower removal rates are assumed for anionic polymers
having appreciable solubility or dispersibility. Removal rate
varies with MW but values are typically assigned as follows:
MWn < 5,000
5,000 < MWn < 20,000
20,000 < MWn < 50,000
MWn > 50,000
0 to 50%
50%
75%
90%
-------
16
This scheme is consistent with data for elimination of
poly(carboxylic acids) in the Semi-Continuous Activated Sludge
(SCAS) and OECD Confirmatory tests (Table 3; unpublished Procter &
Gamble data summarized in Opgenorth 1992), and it is consistent
with the expectation that removal of anionic polymers will be lower
than for cationic or neutral polymers due to the net negative
charge of microbial cell surfaces. Polycarboxylates are widely
used in detergents and other cleaning products and are generally
either homopolymers of acrylic acid or copolymers of acrylic and
maleic acids. Their environmental properties and toxicity have
been extensively reviewed by Opgenorth (1992).
As a general rule, for purposes of estimating environmental
concentrations polymers discharged from wastewater treatment are
assumed not to undergo further removal by in-stream fate processes.^
Precipitation of anionic polymers by alkali metal cations, e.g, Ca
and Mg ; polymer binding to dissolved organic carbon (DOC); and
other sorption phenomena control the bioavailability of the
discharged material (Gary et al. 1987 and Nabholz et al. 1993a).
These factors affecting bioavailability are considered when setting
the CC for a polymeric substance. These areas are addressed in the
following section.
ENVIRONMENTAL CONCERNS FOR POLYMERS
In toxicity assessment, distinction is made between polymers
with minimal low molecular weight (LMW) material, i.e., MWn>1000
with <25% <1000 and <10% <500 daltons, and polymers with either
MWn<1000 or polymers with significant amounts of LMW material,
i.e., >25% <1000 and >10% <500 daltons (USEPA 1996). Water-soluble
or dispersible polymers with either MWn <1000 or significant
amounts of LMW material are of concern because of their
toxicologically similarity to polymers with MWn >1000, but they may
also be absorbed through biological membranes and cause systemic
effects. Environmental toxicity assessments for these polymers are
generally based on both the type of polymer and the type of
functional group(s) in the LMW components (Nabholz et al. 1993aJ.
Polymers with MWn >1000 are not absorbed through the respiratory
membranes of aquatic organisms, and thus toxicity is manifested
either through direct surface-active effects on outer membranes of
aquatic organisms or indirectly via chelation of essential
nutrients, or both (Nabholz et al. I993a). All polymers are
assesses as polymers, but, in addition, polymers with MWn <1000 and
polymers with significant amounts of LMW oligomers are also
assessed as monomers.
Insoluble polymers are not expected to be toxic unless in the
form of finely divided particles. The toxicity of insoluble
particles does not depend upon the chemical structure of the
polymer and results from occlusion of respiratory organs such as
gills. In this case toxicity occurs only at concentrations which
are considered of low concern (Wagner et al. 1995): acute toxicity
-------
17
values are generally >100 mg/L and chronic values
>10 mg/L.
Anionic polymers
Polyanionic polymers with MWn >1000 that are water soluble or
dispersible in water are of concern for aquatic toxicity from
either direct or indirect toxicity. Anionic polymers are divided
into two subclasses for assessment purposes: poly(aromatic
sulfonates/carboxylates) and poly(aliphatic acids).
Poly(aromatic sulfonate/carboxvlatel polymers with MWn >1000
are of moderate concern for toxicity to aquatic organisms, i.e.,
base set LC5g/EC50 values between 1 and 100 mg/L (Wagner et al.
1995), depending on the monomers they contain (Table 4, polymers 1
thru 15) . Dominant monomers associated with polymer toxicity thus
far identified are: carboxylated diphenolsulfemes, sulfonated
diphenolsulfones, sulfonated phenols, sulfonated cresols,
sulfonated diphenylsulfones, and sulfonated diphenylethers.
Monomers of low concern are sulfonated naphthalene and sulfonated
benzene. The strongest evidence for direct toxicity of
poly(aromatic sulfonates/carboxylates) polymers is based on data
for polymers of carboxylated biphenolsulfone (1). This is the only
poly(aromatic sulfonate/carboxylate) polymer which had an MWn
clearly >1000 and nil percentages of LMW oligomers. In addition,
the green algal toxicity test was repeated three times with
growth/test medium with moderate hardness, i.e., 46, 152, and 160
mg/L as CaCO,, to determine if the observed toxicity was due to the
indirect effect of over-chelation of nutrient divalent elements.
There was no reduction in toxicity with increased hardness, i.e.,
algal 96-h EC50 values of 24, 20, and 47 mg/L, respectively, and the
polymer was observed to be algicidal. The mode-of-toxic action is
unknown but it appears to be a direct acting toxicant.
The base set toxicity values for the other toxic polymers (2
to 13) ranged to a low value of 2.0 mg/L (8). The toxicity of
these polymeric substances is undoubtedly a combination of the
direct toxicity from the polymer itself and systemic toxicity from
the LMW oligomers which are capable of being absorbed through
respiratory membranes.
It appears that any of the monomers identified with polymer
toxicity can either be substituted with carboxylic acid or sulfonic
acid. Although there are no data for polymers with substituted
phosphoric acid monomers, toxicity of poly(aromatic phosphate)
polymers should be assumed to be equivalent to poly(aromatic
sulfonates/carboxylates) until test data can be obtained. It also
appears that the most common chemical moiety for this subclass is
phenol. Thus, other polymers similar to the polymers in this
subclass but differing only in the type of phenolic moiety should
be assumed to have similar toxicity. The nearest analog SAR method
is used by OPPT to predict the toxicity of new polymers which
-------
18
belong to this subclass.
Poly(aliphatic acids); Toxicity. Poly(aliphatic acid)
polymers are a subclass of polyanionic polymers and have been shown
to contain three acids: carboxylic acid, phosphinic acid, and
sulfonic acid. Polymers can be either homopolymers relative to one
acid or can contain mixtures of acids. The polymers at
pH 7 are of concern only for their indirect toxicity to green
algae. Toxic.ity to algae as indicated by the 96-h EC50 for growth
inhibition is considered moderate. Toxicity towards fish and
aquatic invertebrates (Table 5) is consistently low, i.e., LC50
values >100 mg/L (Wagner et al. 1995). The average toxicity value
for polyanionic polymers in Table 5 towards fish and aquatic
invertebrates is >225.0 mg/L. Measured acute toxicity values for
an additional 43 polyanionic polymers submitted to the NCP under
sec. 5 of TSCA had an average fish acute value >780.0 mg/L (n = 43)
and a average daphnid acute value >560.0 mg/L (n = 22) at pH near
7 and 100% active ingredients (ai) or 100% polymer solids.
Test data suggest that the mechanism of toxicity is over-
chelation of nutrient elements needed by algae for growth (Nabholz
et al. 1993a), especially calcium and magnesium, and probably iron.
Test data further suggest that the potency of a polymer to cause
indirect toxicity is directly related to the distances between
acids. The distance between acids control the strength of the
chelate formed between two acids and polyvalent metals.
Homopolymers of acrylic acid (polyacrylates) which contain
carboxylic acid substituents on every other carbon in the main
carbon backbone of the polymer have been the most toxic
poly(aliphatic acids) towards green algae (Table 5, polymers 17 to
22} with a geometric mean 96-h EC50 value of 8.6 mg/L. In contrast,
if you move the acids closer together or farther apart relative to
their position along the polymer backbone, then toxicity decreases
based on test data for PMN polymers whose chemical identity is
confidential under TSCA. Test data for homopolymers of maleic acid
(16), which have COOH on every carbon of the polymer backbone, (a)
indicate low toxicity to algae (96-h EC50 = 560 mg/L) , (b) 65 times
less toxicity relative to poly(acrylic acid), and (c) suggesting a
weak ability to chelate nutrient elements. Other test data for
polymers which have acids farther apart than 1 carbon, i.e., 1.5
carbon separation (23) and 2.0 carbon separation (24), showed 8
times and 17 times less toxicity, respectively, relative to
poly(acrylic acid) (17 to 22).
In addition, the toxicity of a polyanionic polymer can be
reduced relative to the toxicity of poly(acrylic acid) by (a)
adjusting the distance between acids by moving some acids farther
from the polymer backbone or (b) randomly diluting a monomer which
chelates, e.g., acrylic acid, with a monomer which does not, e.g.,
acrylamide and maleic acid. Polymer 25 is a polymer in which
poly(acrylic acid) acid has been randomly reacted with another
monomer in which still has a carboxylic acid on every other carbon
-------
19
but the carboxylic acid is pendant to the backbone by a distance of
3 carbons. The green algal 96-h EC50 of polymer 25 is 57.0 mg/L?
7 times less toxic than poly(acrylic acid). Test data for polymer
26 (Table 5) shows the effect on toxicity of diluting a homopolymer
of acrylic acid by just one third with a non-chelating or weak
chelating monomer. The algal 96-h EC50 of this polymer was >500.0
mg/L which is a toxicity reduction of over 60 times relative to
poly(acrylic acid).
Therefore, based these data, poly(aliphatic acid) polymers
have been grouped by the average distance between the acids on the
polymer (Table 5) and the toxicity of these polymers ^ can be
predicted by using the nearest analog SAR method and knowing
(a) the monomer composition of a polymer; and (b) the reaction
sequence, i.e., random reaction of monomers or blocked reactions.
In addition, the toxicity of poly(aliphatic acids) towards green
algae is (a) an indirect effect via over-chelation of nutrient
elements, (b) is sensitive to changes in the chemical structure
which result in changing the distance between acids, and (c) will
only be observed in soft natural waters which have hardness of less
than 30.0 mg/L as CaC03. The hardness of test/growth media as
specified in EPA, OECD, EU, and ASTM test methods is 15 to 24 mg/L.
The toxicity measured in the standard green algal toxicity tests is
an overestimate of toxicity for most natural surface waters.
Poly(aliphatic acids): Mitigation of Toxicitv. If the Ca2*
salt of a poly(aliphatic acid) polymer is tested in standard algal
media or if the polymer is tested in media with a hardness of near
150.0 mg/L CaC03, then the toxicity is mitigated (Table 6). OPPT
began to recommend toxicity mitigation testing for polyanionic
polymers in 1989 after toxicity mitigation testing for polycationic
polymers was initiated. PMN test data for polyacrylates (19, 20,
and 22) demonstrated mitigation factors from 14 to 600 times as the
amount of Ca in the testing environment increased. Test data for
two polymers intrinsically less toxic than polyacrylate polymers
(23 and 25) indicated less mitigation with mitigation factors of 12
and 8.9, respectively. However, the mitigated effective
concentrations were similar. Green algal 96-h EC5o values v^ere all
in the range of 500.0 to 950.0 mg/L in the presence of Ca . All
of this mitigation testing was done near pH 7.5. Toxicity
mitigation of poly(aliphatic acids) fails when the acid is'tested
un-neutralized. The toxic effects of the H*, i.e., low pH, cannot
be overcome by adding Ca either to the polymer or to the media and
dominate all testing results.
The manifestation of toxicity via over-chelation of nutrient
elements may be largely confined to laboratory conditions, +such as,
release of the polymer as the soluble salt, e.g., Na or K ,
directly to naturally soft surface waters without sewage treatment.
Standard algal toxicity tests generally seek to maximize algal
growth and measure intrinsic or baseline toxicity and, thus, employ
media with a hardness of about 15 to 24 mg/L, which is soft. Yet
-------
20
many poly(aliphatic acids) are used as scale inhibitors in
industrial and commercial applications and are released to the
environment already chelated with Ca* and Mg". Moreover, other
major uses such as in detergents and cleaning products should lead
to release of the polymers to wastewater treatment, where both
chelation and removal by sorption occurs. For any polymer released
in sewage treatment effluent, mitigation of toxicity may also occur
from divalent cations in receiving waters, since the average
hardness of freshwater in the U.S. is about 120 mg/L (CaC03
equivalent; Nalco 1988).
Recommended algal toxicity testing for poly(aliphatic acid)
polymers under sec. 5 of TSCA generally consists to three tests:
(1) neat polymer at pH 7.5 tested in standard algal growth medium;
(2) polymer with stoichiometrically equivalent Ca + added to the
polymer stock solution at pH 7.5 and tested in standard growth
medium; and (3) neat polymer at pH 7.5 tested in modified algal
growth medium. Modified growth medium contains calcium alone or
calcium and magnesium added to attain'a measured hardness of about
150 mg/L. If Ca * and Mg are added together they should be present
in a 2:1 (Ca:Mg) ratio. The purpose of this scheme is to simulate
various conditions of release. For example, polymers used as scale
inhibitors are expected to be released to the environment chelated
with Ca and Mg and should therefore be tested with all three
tests. Test (1) measures intrinsic toxicity. If the 96-h EC50 is
>100.0 mg/L and the Chv is >10.0 mg/L, then no further testing need
be done because of low concern for intrinsic toxicity. Test (2)
simulates the release from use as a scale inhibitor and test (3)
simulates the release of the neat polymer from manufacturing and
processing. For polymers not expected to be used as scale
inhibitors or released as the Ca+ and/or the Mg+, test (2) can be
eliminated. Of course, if the site of release is known and the
hardness of receiving water is known at the site, then the hardness
of the medium can be specified to match the site-specific
conditions. In some cases, PMN submitters have combined tests (2)
and (3) into one test, i.e., testing the calcium salt of the
polymer in moderately hard medium. This is acceptable but keep in
mind that the.green algal 96-h EC50 due to hardness alone is equal
to 1140.0 mg/L and the ChV =80.0 mg/L.
Nonionic polymers
Nonionic polymers with MWn >1000, <25% <1000, and <10% <500
are generally of low concern for ecotoxicity because they have
negligible water solubility. If a nonionic polymer is water
soluble or dispersible and has monomers reacted via random order,
then aquatic toxicity is still low with base set LC50/EC50 values
expected to be >100.0 mg/L. However, if monomers are blocked in
order to use the polymer as a surfactant or dispersant, then the
polymer could be toxic to aquatic organisms through a surface-
active detergent-type mode-of-toxic action and should be tested.
Nonionic polymers with significant amounts of oligomer content,
-------
21
i.e., >25% <1000 and >10% <500, may be of concern based on the
bioavailability and toxicity of the LMW material and the LMW
oligomers are assessed as monomers. And finally, nonionic polymers
with MWn <1000 are assessed as monomers.
Cationic polymers; Toxicitv
Cationic polymers of concern for environmental toxicity
includes polymers that contain a net positively charged atom or
that contain groups that can reasonably be anticipated to become
cationic in water (USEPA 1996). Atoms with a net positive charge
includes, but are not limited to, quaternary ammonium, phosphonium,
and sulfonium. Groups anticipated to become cationic in water
include, but are not limited to, aliphatic primary, secondary, and
tertiary amines. Forms of nitrogen not included are (a) aromatic
nitrogens, unless they are quaternarized; (b) nitrogens which are
directly substituted to benzene, such as, aniline; (c) amides;
(d) nitriles; and (e) nitro groups. As with polyanionic
polymers, polycationic polymers of concern have to be either water
soluble or dispersible in water as either micro- or macro-
dispersions, cationic polymers that are solids and are only to be
used in the solid phase are of low concern. specifically,
dispersed beads of cationic polymers are of low concern.
Polycationic polymers are assessed according to their type of
polymer backbone: carbon-based, silicone-based, e.g., [Si-O], or
natural, e.g., chitin, starch, and tannin, because the toxicity
data for these polymers have suggested that the type of backbone
can influence toxicity and some physical/chemical properties.
Therefore, the toxicity data for cationic polymers have been
grouped according to polymer backbone ib Table 7.
The intrinsic aquatic toxicity of polycationic polymers in
clean water, i.e., water with a total organic carbon (TOG) content
of <2 mg/L, increases exponentially with increasing charge density
until toxicity becomes asymptotic (Table 7) . Polycationic polymers
have been shown to elicit acute toxic effects in aquatic organisms
by physically disrupting respiratory, e.g., gill membranes, thus
interfering with O2 exchange (Biesinger and Stokes 1986). It is
presumed that polycationic polymers strongly adsorb to all
biological membranes which are negatively charged or anionic. It
is also presumed that chronic toxicity occurs thru the same
mechanism for polymers with minimal amounts of LMW oligomers, i.e.,
<25% <1000 and <10% <500 daltons. For cationic polymers with KWn
<1000 and polymers with significant amounts of LMW oligomers, i.e.,
>25% <1000 and >10% <500, systemic toxicity is also possible
(Nabholz et al. 1993a).
The aquatic toxicity of polycationic polymers is most strongly
influenced by cationic charge density and type of polymer backbone
(Table 7) rather than (a) if the charge is permanent, such as,
quaternary ammoniums, or dependant on pH, e.g., aliphatic 1°, 2',
-------
22
or 3° amines; (b) the position of the cation relative to the
polymer backbone, such as, in the backbone or pendant from the
backbone; or (c) MW. For example, for polymers with silicone-
based backbones (polymers 62 to 70, Table 7), the largest
difference between daphnid 48-h LC50 values due to charge density
is about 2300 times, i.e., 62 compared to 69, but the difference
between toxicity due to converting tertiary amines . (65) to
quaternary ammoniums (66) is only 5 times. Another example
compares the apparent effect of MW between two polymers with the
same charge density. Polymers 28 and 29 have the same charge
density of 0.7% amine-N but differ greatly in MNn, i.e., 1800
versus 8 million. The average difference between fish acute
toxicity values (n = 3) is <2 times and the average difference
between daphnid acute values (n = 2) is 9 times. The differences
between green algal 96-h EC50 and chronic values are >160 times and
150 times, respectively.
Cationic charge density is based on percent amine-nitrogen
(%a-N) because more than 99.9% of all polymers that have been
submitted under sec. 5 of TSCA have had their cationic group based
on nitrogen. However, %a-N, cation equivalent weight (EQWT), and
number of cations per 1000 MW (#C/K) are all equivalent expressions
of cationic charge density. To convert from %a-N to EQWT or #C/K
the following equations can be used:
1400 * %a-N = N-EQWT or cation EQWT; and
%a-N x 0.714286 = #C/K.
To convert cation EQWT to #C/K:
1000 + EQWT = #C/K.
The structure activity relationships (SAR) for polycationic
polymers are grouped according to type of toxic effect and type of
polymer backbone. All SARs are based on toxicity data for polymers
listed in Table 7; 100% active ingredients (ai) of polymer or 100%
polymer solids; test dilution water with less than 2.0 mg/L of
total organic carbon (TOC) and hardness <180.0 mg/L as CaCO3; pH
near 7; and nominal concentrations.
A. Fish acute toxicity: 96-h LC50
(1) Carbon-based backbone
(a) %a-N <3.5
log fish 96-h LC50 (mg/L) = 1.209 - 0.462 %a-N;
where n = 19 and R = 0.66;
(b) %a-N >3.5%
fish 96-h LC50 = 0.280 mg/L; where n = 34;
-------
23
(2) Silicone-based backbone
(a) %a-N <3.5
log fish 96-h LC5Cfe (mg/L) = 2.203 - 0.963 %a-N;
where n = 4 and R =0.73;
(b) %a-N >3.5%
fish 96-h LC50 = 1.17 mg/L; where n = 1; and
(3) Natural-based backbone
(a) %a-N <3.5
These polymers will have either similar toxicity to carbon-based
backboned polymers or less toxicity. Polymers based on tannin (77,
78) have similar toxicity; polymers based on chitin or glucosamine
(79) are 2000 times less toxic; and polymers based on starch (71
thru 76) are either 10 times more toxic or 80 times less toxic.
SAR analysis for cationic polymers with natural-based polymer
backbones will require the nearest analog method to predict fish
acute toxicity.
(b) %a-N >3.5%
The only datum available (79) indicate less toxicity than predicted
using SARs for carbon-based polymer backbones.
B. Daphnid acute toxicity: 48-h LC,
(1) Carbon-based backbone
50
(a) %a-N <3.5
log daphnid 48-h LC50 (mg/L) = 2.839 - 1.194 %a-N;
where n = 7 and R =0.90;
(b) %a-N >3.5%
daphnid 48-h LC50 = 0.100 mg/L; where n = 13;
(2) Silicone-based backbone
Data for silicone-based polymers (Table 7) indicate that acute
toxicity towards daphnids will be either similar to carbon-based
backboned polymers or less toxic. SAR analysis will require the
nearest analog method to predict daphnid acute toxicity: use the
most toxic nearest analog.
(3) Natural-based backbone
(a) %a-N <4.3
log daphnid 48-h LC50 (mg/L) = 2.77 - 0.412 %a-N;
where n = 6 and R =0.82;
-------
24
(b) %a-N >4.3%
daphnid 48-h LC50 < 11.0 mg/L; where n = 1;
Cationic polymers with natural-based backbones are generally less
toxic than predicted relative to carbon-based polymers with the
same charge density.
C. Green algal toxicity: 96-h EC50
(1) Carbon-based backbone
(a) %a-N <3.5
log green algal 96-h EC50 (mg/L) = 1.569 - 0.97 %a-N;
where n = 5 and R =0.54;
(b) %a-N >3.5%
green algal 96-h EC50 = 0.040 mg/L; where n = 12;
(2) Silicone-based backbone
Data for silicone-based polymers (Table 7) indicate that toxicity
towards green algae will be either similar to carbon-based
backboned polymers or less toxic. SAR analysis will require the
nearest analog method to predict algal 96-h EC50 toxicity value:
use the most toxic nearest analog.
(3) Natural-based backbone
Cationic polymers with natural-based backbones are less toxic than
predicted relative to carbon-based polymers with the same charge
density. SAR analysis will require the nearest analog method: use
the most toxic nearest analog.
D. Fish Chronic Toxicity: Only one polymer (82) has been tested
for fish chronic toxicity using a fish early-life-stage toxicity
test. The chronic value (ChV) was 0.018 mg/L which resulted in an
acute-to-chronic ratio (ACR) of 18.0.
E. Daphnid Chronic Toxicity: Only one polymer (82) has been
tested for daphnid chronic toxicity using the daphnid 21-d
reproductive inhibition toxicity test. The ChV was 0.022 mg/L
which resulted in an acute-to-chronic ratio (ACR) of 14.0.
F. Green algal chronic toxicity: 96-h ChV
(1) Carbon-based backbone
(a) %a-N <3.5
log green algal ChV (mg/L) = 1.057 - 1.0 %a-N;
where n = 5 and R =0.53;
(b) %a-N >3.5%
i
-------
25
green algal 96-h EC50 = 0.020 mg/L; where n = 11;
(2) Silicone-based backbone
The single datum for silicone-based polymers (Table 7, polymer 62)
indicated that the algal ChV was similar to the ChVs for carbon-^
based backboned polymers. SARs for carbon-based polymers can be
used to predict the ChV for silicone-based polymers until more test
data are obtained.
(3) Natural-based backbone
Cationic polymers with natural-based backbones are less toxic than
predicted relative to carbon-based polymers with the same charge
density. SAR analysis will require the nearest analog method: use
the most toxic nearest analog.
Amphoteric polymers: Toxicity
Polyamphoteric polymers contain both cationic and anionic
moieties, e.g., carboxylic acids, in the same polymer. The
toxicity of polyamphoteric polymers is largely determined by the
cationic charge density and the cation-to-anion ratio (CAR). As
the cationic charge density increases the toxicity to aquatic
organisms increases, and, when charge density is constant, toxicity
also increases with increases in the CAR. And vice-versa, as the
CAR decreases, toxicity decreases. The toxicity of polyamphoteric
polymers can be predicted in four steps: (1) The cationic charge
density, i.e., percent amine-nitrogen (%a-N), and the cation-to-
anion ratio (CAR) is calculated from the polymer chemical
structure. (2) The toxicity of the polymer is predicted assuming
that it is a polycationic polymer using the charge density and the
SARs for cationic polymers with carbon-based backbones. (3) A
toxicity reduction factor (TRF) is calculated for the polymer for
each effect, e.g., fish acute toxicity and daphnid acute toxicity,
using SARs between the TRF and the CAR. (4) Finally, the predicted
toxicity values (based on the assumption that the polyamphoteric
polymer is a polycationic polymer) are multiplied by the TRF.
The SARs between TRF and CAR were developed by (a) predicting
the toxicity of each polymer in Table 8 using the SARs for carbon-
based backboned cationic polymers; (b) dividing the actual
measured toxicity value for the polyamphoteric polymer (Table 8) by
the predicted toxicity value and calculating a TRF for each
polymer; and (c) regressing the TRFs against the CARs for each
effect. The SARs between the TRF and the CAR are:
A. Fish acute toxicity: 96-h LC50
log TRF - 1.411 - 0.257 CAR; where n = 3 and R2 = 0.86;
B. Daphnid acute toxicity: 48-h LC50
-------
26
log TRF = 2.705 - 0.445 CAR; where n = 2 and R2 = 1.0;
C. Green algal toxicity: 96-h EC50
re
D. Green algal chronic toxicity: 96-h ChV
log TRF = 1.544 - 0.049 CAR; where n = 2 and R2 = 1.0; and
log TRF = 1.444 - 0.049 CAR; where n = 2 and R2 = 1.0.
Cationic and amphoteric polymers: Mitigation of Toxicity
From 1970 thru 1984 many polycationic polymers submitted under
sec. 5 of TSCA as new chemicals were assessed as having the
potential of presenting an unreasonable risk to the aquatic
environment. As a result, many polycationic polymers were
regulated by restricting uses and/or release sites, requiring on-
site treatment in addition to sewage treatment, and requiring
toxicity and fate testing. The impact on the polymer manufacturing
industry became so great that, in late 1984, the Synthetic Organic
Chemicals Manufacturers Association (SOCMA) began discussions with
the Environmental Effects Branch (EEB) of OPPT about the
environmental risk and fate of cationic polymers. SOCMA claimed
(and it was generally known and accepted by OPPT) that polycationic
polymers were designed to react with dissolved organic carbon (DOC)
in water and formed neutral insoluble complexes (or floe) which
were assumed to settle on sediment surfaces and accumulate in
sediment. There was general information in the literature (1960 to
1970s) that the toxicity of. cationic polymers towards fish was
reduced in the presence of suspended solids in water. In addition,
it was proposed that cationic polymers should not be toxic to
aquatic organisms as long as there was sufficient DOC in natural
waters to satisfy the exchange capacity of the polymer. It quickly
became apparent to SOCMA and OPPT that fate studies would be of
little use in supplying adequate test data to refine the PEC in the
environment because (1) CCs were in the low M9/L (Ppb) range, (2)
analytical methods for cationic polymers had detection limits in
the 1 mg/L (ppm) range, and (3) even the best analytical method
could not distinguish between dissolved ancl DOC-reacted polymer.
OPPT wanted adequate rigorous testing data using standardized test
guidelines to prove that cationic polymers were of low toxicity in
the presence of DOC. In addition, OPPT wanted to known if cationic
polymers in sediment were bioavailable to organisms which fed on
sediment.
In 1986, EEB recommended that OPPT consider developmental of
a TSCA sec. 4 test rule for existing polycationic polymers because
discussions with SOCMA had resulted in no significant progress on
testing and it was known that there were many unpublished toxicity
studies for cationic polymers on the Inventory. A test rule would
also automatically generate sec. 8(a) and 8(d) rules which would
obtain industry data on the manufacturing of cationic polymers and
-------
27
those unpublished studies, respectively. In the same year, OPPT
begin to design a standardized test guideline, which would be
acceptable to everyone, to measure the mitigation of toxicity of
cationic chemicals in the presence of DOC.
The two most important factors to be decided were (1) what DOC
to use in the test and (2) what was the mean concentration and
distribution of DOC in U.S. surface fresh waters. Humic acid was
selected as the surrogate DOC based on research done by Gary et al.
1987. Cary et al. measured the mitigation of 4 suspended solids
and five types of purified dissolved-organic-carbon compounds on
the acute toxicity of four cationic polyelectrolytes to freshwater
fish and aquatic invertebrates. Humic acid was about average in
its ability to mitigate toxicity. Analysis of Table 4 in Gary et
al. indicated that the mitigation factors (MF) for humic acid were
closest to the mean MF for all of the DOCs tested. The mean MF
factor was calculated by OPPT for each of the polymer/species
combinations. The MF of each DOC was compared to the mean KF and
the absolute value of the difference was averaged for that DOC.
Humic acid had the lowest average difference or, in other words,
the MFs for humic acid were closest to the mean MF for each
polymer/species combination, i.e., lignin > tannic acid > fulvic
acid > lignosite > humic acid. In addition, humic acid was easily
available from chemical supply companies.
OPPT decided that 10 mg/L of total organic carbon (TOG) was to
be the amount of DOC at which it would assess the risk of cationic
polymers to water-column aquatic organisms. This decision was
based on three facts: (a) concentrations of humic acid in natural
waters were rarely measured, (b) the average measured amount of TOC
in natural freshwater of the U.S. was about 6.79 mg TOC/L (Lynch
1987) , and (c) 10 mg TOC/L was a round number close to 7 which errs
on the side of safety. Lynch (1987) analyzed the USEPA Office of
Water's STORET Data Base for measured amounts of TOC in U.S.
waters. Lynch found 67,994 measurements of TOC taken from 1977
through 1987 from all over the U.S., i.e., 19 of 23 major river
basins. These TOC measurements were lognormally distributed and
skewed toward larger concentrations of TOC. , The geometric mean of
these data was 6.79 mg TOC/L. Since OPPT does generic risk
assessments for most chemicals, at least the first time they are
assessed, it was decided to use the average amount of TOC in
natural waters as the benchmark amount of DOC.
By November 1988, negotiations between SOCMA and OPPT resulted
in the OPPT test guideline, "Fish Acute Toxicity Mitigated by Humic
Acid (OPPTS 850.1085)." The guideline was distributed in December
1988 and by October 1989 seven acute toxicity and toxicity
mitigation tests using fish had been submitted to OPPT and
validated by EEB. As a result, EEB recommended to OPPT that the
proposed sec. 4 test rule, sec. 8 (a), and sec. 8 (d) rules for
cationic polymers be withdrawn. There is no current activity on
these polymers within the OPPT Existing Chemical Program.
-------
28
The toxicity of polycationic polymers is mitigated in the
presence of DOC {Table 9) . Fish acute toxicity data for 16
polycationic polymers have been evaluated and validated by EEB.
All of these polymers were submitted under sec. 5 as new chemicals
and all were dropped from further review after the mitigation
factor (MF) was integrated into the risk assessment. The MF is
defined as the amount of toxicity mitigation due to 10 mg TOC/L
relative to fish acute toxicity as a 96-h LC50. Mitigation factors
for cationic polymers which are (a) random reactions of monomers
and (b) have minimal oligomer content, i.e., <25% <1000 and <10%
<500, are directly correlated with charge density, i.e., percent
amine-N (%a-N). The structure activity relationships (SAR) are:
A. For charge densities >3.5 %a-N:
MF = 110.0
where n = 7;
B. For charge densities between 3.5 to 0.7 %a-N:
log MF = 0.858 + 0.265 %a-N
where n = 4 and R =0.61; and
C. For charge densities <0.7 %a-N, MFs have not yet been measured,
but are expected to be <7.0.
Cationic polymers which have significant oligomer content,
i.e., >25% <1000 and >10% <500, can have MFs significantly lower
than predicted for polymers with minimal oligomer content. For
example, polymers 47 and 48 have predicted MFs of 110, however,
their measured MFs were 21 and 26, respectively. In addition, MFs
for blocked polymers cannot, at this time, be accurately predicted
using these SARs. Polymer 27 is a blocked polymer with the
cationic group attached to the polymer tips and is atypical of the
other polymers in Table 9 which are randomly reacted polymers. The
measured MF was 130 while the predicted MF was 11; a difference of
12 times.
The toxicity of polycationic polymers is also mitigated
when mixed with sediment. Toxicity testing with natural sediment
contaminated with cationic polymers and with species which ingest
sediment has shown that cationic polymers with charge densities of
£4.2% a-N (or >3 cations/1000 MW or a N-equivalent weight <333.0)
are not bioavailable to cause toxicity, i.e., 48-h no-effect-
concentrations (NEC) >100.0 mg/kg dry weight sediment (Rogers and
Witt 1989) , and thus are of low concern once transported to
sediments. A word of caution is needed here. No cationic polymer
with significant oligomer content, i.e., >25% <1000 and >10% <500,
has been tested. The oligomers are expected to be more
biologically available for uptake due to the lower MFs observed for
polymers with significant oligomer content (Table 9) . Another
possible exception occurs when a cationic polymer is formulated
-------
29
with acid in excess of the amount needed to neutralize the polymer,
such that the product's pH is approximately 2. Under these
conditions the polymer fails to flocculate DOC in the water column
and DOC apparently does not mitigate acute toxicity to fish
(Nabholz et al. 1993a).
The testing scheme for cationic polymers under TSCA generally
requires 5 tests: (1) fish acute toxicity with clean dilution
water, i.e., TOC <2 mg/L; (2) daphnid acute toxicity with clean
dilution water; (3) algal toxicity test with standard growth
medium; (4) fish acute toxicity with 10 mg of humic acid/L added
to clean dilution water; and (5) fish acute toxicity with 20 mg
humic acid/L added to dilution water. All tests are done with the
static method; nominal concentrations of polymer based on 100% ai;
and TOC concentrations are measured in the controls of all five
tests just prior to test initiation. When the testing results are
validated by EEB, the three fish 96-h LC50 values are regressed
against TOC concentration. The fish 96-h LCc0 due to the addition
of 10 mg TOC/L is predicted using the regression equation, and this
value is divided by the fish LC50 measured in clean water to obtain
the toxicity mitigation factor (MF) due to 10 mg TOC/L. The MF is
then used to decrease the toxicity or adjust upward all effective
concentrations (EC) in the polymer's toxicity profile that are
based on the intrinsic toxicity of the cationic polymer in clean
water. A new CC is then determined from the ECs for the polymer in
the presence of 10 mg TOC/L and a new risk assessment is done. If
there is still a potential risk to aquatic organisms in the water
column, chronic toxicity testing with fish and invertebrates in
clean dilution water is recommended using flow-through or static-
renewal methods (with renewals every 24 h) and nominal
concentrations.
SUMMARY AND CONCLUDING REMARKS
In the U.S. the environmental safety of chemical substances
with non-food, non-drug and non-pesticidal uses is evaluated by EPA
under the authority granted by the Toxic Substances Control Act
(TSCA). For polymers the scope of TSCA's applicability is very
broad and includes such use categories as water and wastewater
treatment, coatings, household, and industrial cleaning products
and manufactured goods. TSCA clearly distinguishes between
existing chemical substances and new substances not on the TSCA
Inventory. To date most of the environmental assessments for
polymers have focused on new polymers, for which prospective
manufacturers or importers must file a Premanufacture Notice (PMN)
under sec. 5 of TSCA. The PMN process essentially involves.
weighing potential risks and benefits for each new substance.
Polymers that meet certain requirements are exempt from
reporting under sec. 5, but this does not include most polymers
with aquatic uses such as in water treatment. Non-exempt polymers
are subject to the normal PMN review process, which includes
-------
30
chemistry review, toxicity evaluation, exposure assessment, and
risk assessment/risk management phases. A unique feature of this
process is its high level of reliance on structure activity
relationships (SAR) to predict missing data. SAR development has
become a welcome indirect benefit as a result of the absence of any
specific or implied requirement in TSCA that PMN submitters perform
testing as a precondition for approval. Spurred by the need to
make informed judgments often with few or no experimental data,
estimation methods have been developed for most physical and
chemical properties, some transformation processes, e.g.,
biodegradation, and many toxic effects relevant to aquatic
exposures. Unfortunately, these methods, although on balance very
successful for discrete (class 1) substances, are of little benefit
for predicting the fate of polymers. The result is that
experimental data continue to play a critical role in environmental
fate assessment of polymers. SAR methods for environmental aquatic
toxicity are available for all of the broad classes of polymers.
Most polymers in commerce are essentially nonbiodegradable.
Environmental fate assessment therefore is reduced to an evaluation
of their potential for sorption or precipitation under various
conditions. Rules of thumb are available to inform judgments about
these processes, but in general they are based on a heavy dose of
faith and relatively few data. Clearly it is in industry's best
interest to furnish EPA with relevant fate data.
Environmental toxicity assessment are based on polymer charge:
anionic, nonionic, cationic, and amphoteric. Anionic; Polyanionic
polymers are divided up into two classes: poly(aromatic
sulfonates/carboxylates) and poly(aliphatic acids). Many
poly(aromatic sulfonates/carboxylates) are moderately toxic to
aquatic organisms thru an unknown direct acting mode-of-toxic
action. The nearest analog-SAR method is used to predict the
toxicity of these polymers by identifying the dominant monomer(s)
of the polymer. Poly(aliphatic acids) show toxicity only towards
green algae. These polymers indirectly effect algal growth by
over-chelating the nutrient elements needed for growth. The
potency of a polymer to cause toxicity is directly related to the
ability of the polymer to chelate divalent metals. The distance
between acids on the polymer controls the strength of the chelate.
The toxicity of these polymers in soft water environments can be
predicted by determining the average distance between acids. The
toxicity of poly(aliphatic acids) is mitigated by changing their
salt to a divalent salt, e.g., Ca *, or by releasing them to
moderately hard water. Environmental risk from these polymers will
most likely be observed in soft natural water which have hardness
of <30.0 mg/L as CaC03. Nonionic: Nonionic polymers are generally
of low concern for ecotoxicity, however, if hydrophobic and
hydrophilic monomers are blocked in order to use the polymer as a
detergent or dispersant, then it could be toxic to aquatic
organisms. Likewise, polymers with significant amounts of
oligomers whose MW <1000 may be of concern due to the toxicity of
-------
31
the oligomers and the oligomers should be assesses as monomers.
Cationic; Cationic polymers can be highly toxic to aquatic
organism and should be used with caution. Cationic polymers can be
grouped by the type of polymer backbone: carbon-based, silicone-
based, and backbones using a natural polymer such as starch. The
aquatic toxicity of cationic polymers is primarily related to their
cationic charge density, and, if soluble or dispersible as micro-
or macro-emulsions, can be reasonably predicted by SARs based on
regression equations. The toxicity of Cationic polymers based on
silicone or a natural polymer can be less toxic relative to a
polymer based on carbon, but not always. Amphoteric:
Polyamphoteric polymers contain both cationic and anionic moieties
in the same polymer. The toxicity of these polymers is determined
by the cationic charge density and the cation-to-anion ratio.
Toxicity increases as charge density increases and, when charge
density is held constant, toxicity increases with increases in the
cation-to-anion ratio. Polyamphoteric polymers can also be highly
toxic to aquatic organisms. SARs for polyamphoteric polymers are
based on calculating the cationic charge density and the cation-to-
anion ratio from polymer chemical structure.
The toxicity of cationic and araphoteric polymers can be
mitigated by dissolved organic carbon (DOC) which occurs naturally
in surface waters. Toxicity mitigation in the presence of 10 mg
total organic carbon/L is related to cationic charge density and
can be predicted by SARs. If cationic and amphoteric polymers are
used according to current standard practices, then they are of low
environmental risk.
Environmentally safe manufacture and use of polymers is best
assured if EPA and industry cooperate to the extent and from the
earliest phase of product development practicable for the PMN
submitter. Much more use of the New Chemical's Program's Pre-
Notice Communication process could be made than is the case now,
and the effect can only be beneficial for all stakeholders. The
most important information needed by EPA assessors is polymer
chemical structure. Specific data about polymer structure is
critical to the whole assessment process. Data needs include, but
are not limited to, monomer identity; polymer composition, e.g.,
mole ratios of monomers; reaction sequence of monomers: random or
blocked; average-number molecular weight; molecular weight
distribution; and oligomer content, i.e., percent <1000 and
percent <500 daltons.
If testing is considered necessary, either by voluntary
agreement or Consent Order, to resolve uncertainties, test
protocols should be reviewed by EPA before any testing commences.
Further, what is being tested should be understood and agreed to by
stakeholders; obviously this is particularly important for class 2
substances. Through a combination of (1) testing tailored to
specific substances subject to PMN review and
(2) more fundamental research on the mechanisms of polymer
-------
32
sorption, precipitation, and toxicity and its mitigation, the
database needed for truly informed ris.k assessment can be realized.
DISCLAIMER
This document has been reviewed by OPPT, EPA and approved for
publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency nor does
mention of tradenames or commercial products constitute endorsement
or recommendation for use.
-------
-------
33
LITERATURE CITED
Alexander, M. 1973. "Nonbiodegradable and other recalcitrant
molecules,1* Biotechnology and Bioengineering. 15:611-647.
American Society for Testing and Materials (ASTM). 1994.
"Standard practice for coagulation-flocculation jar test of
water," D 2035-80, in 1994 Annual book of ASTM standards.
Section 11. Water and environmental technology, volume
11.02, pp. 751-753. Philadelphia, PA: ASTM.
Bailey, W.J. and B. Gapud. 1986. "Synthesis of new
biodegradable polymers," in Biodeterioration 6. Proceedings
of the sixth international biodeterioration symposium, S.
Barry and D.R. Houghton, eds., pp. 605-610. Slough, United
Kingdom: CAB International.
Barnes, D.G. and M.L. Dourson. 1988. "Reference dose (RfD):
description and use in health risk assessment," Regulatory
Toxicology and Pharmacology. 8:471-486.
Biesinger, K.E. and G.N. Stokes. 1986. "Effects of synthetic
polyelectrolytes on selected aquatic organisms," Journal of
the Water Pollution Control Federation. 58:207-213.
Gary, G.A., J.A. McMahon and W.J. Kuc. 1987. "The effect of
suspended solids and naturally occurring dissolved organics
in reducing the acute toxicities of cationic
polyelectrolytes to aquatic organisms," Environmental
Toxicology and Chemistry. 6:469-474.
Clark, B., J.G. Henry and D. Mackay. 1995. "Fugacity analysis
and model of organic chemical fate in a sewage treatment
plant," Environmental Science and Technology. 29:1488-1494.
Clements, R.G. (ed) . 1988. .Estimating toxicity of industrial
chemicals to aquatic organisms using structure activity
relationships, EPA report no. 560/6-88-001 (NTIS no. PB89-
117592). Washington, DC: Office of Pollution Prevention and
Toxics.
Clements, R.G. and J. V. Nabholz. 1994. ECOSAR: A computer
program for estimating the ecotoxicity of industrial chemicals
based on structure activity relationships, EPA report no. 748-
R-93-002 (NTIS no. PB94-500485 [program] and PB94-104668
[user's guide]). Washington, DC: Office of Pollution
Prevention and Toxics.
Clements, R.G., J.V. Nabholz, D.E. Johnson and M.G. Zeeman.
1993. "The use of quantitative structure-activity
relationships (QSARs) as screening tools in environmental j
assessment," in .Environmental toxicology and risJc
assessment: 2nd volume, ASTM STP 1216, J.W. Gorsuch, F.J.
Dwyer, C.G. Ingersoll and T.W. LaPoint, eds. Philadelphia,
PA: American Society for Testing and Materials.
Halverson, F. and H.P. Panzer. 1980. "Flocculating agents," in
Kirk-Othmer Encyclopedia of Chemical Technology, volume 10,
3rd edition, pp. 489-523. New York: Wiley. j
Health Council of the Netherlands (HCN). 1989. Assessing the
risk of toxic chemicals for ecosystems. The Hague, The
Netherlands: HCN. U.S. EPA Headquarters Ubrarv
_Ma" code 3201
Pennsylvania Avenue NW
Washington DC 20460
-------
34
Leo, A. and D. Weininger. 1985. "CLOGP Version 3.3: Estimation
of the n-octanol/water partition coefficient for organics in
the TSCA Industrial Inventory." Computer Program. Pomona
College, Claremont, CA.
Lynch, D.G. 1987. "Summary of STORET data on dissolved organic
carbon (DOC) levels in surface waters," Memorandum dated 15
June. Washington, DC: Engineering, Economics, and Technology
Division, Office of Pollution Prevention and Toxics, USEPA.
Lynch, D.G., N.F. Tirado, R.S. Boethling, G.R. Huse and G.C.
Thorn. 1991. "Performance of on-line chemical property
estimation methods with TSCA Premanufacture Notice
chemicals," Ecotoxicology and Environmental Safety. 22:240-
249.
Meylan, W.M. and P.H. Howard. 1991. "Bond contribution method
for estimating Henry's Law constants," Environmental
Toxicology and Chemistry. 10:1283-1293.
Meylan, W.M. and P.H. Howard. 1995. "Atom/fragment contribution
method for estimating octanol-water partition coefficients,"
Journal of Pharmacological Science. 84:83-92.
Meylan, W.M., P.H. Howard and R.S. Boethling. 1992. "Molecular
topology/fragment contribution method for predicting soil
sorption coefficients," Environmental Science and
Technology. 26:1560-1567.
Meylan, W.M., P.H. Howard and R.S. Boethling. 1995. "Improved
method for estimating water solubility from octanol/water
partition coefficient," .Environmental Toxicology and
Chemistry, in press.
Moss, K., D. Locke, and C. Auer. 1996. "EPA's New Chemicals
Program," Chemical health & Safety. Jan/Feb:29-33.
Nabholz, J.v. 1991. "Environmental hazard and risk assessment
under the United States Toxic Substances Control Act (TSCA),"
The Science of the Total Environment. 109/110:649-665.
Nabholz, J.V., P. Miller and M. Zeeman. 1993a. "Environmental
risk assessment of new chemicals under the Toxic Substances
Control Act (TSCA) section 5," in Environmental toxicology
and risk assessment, ASTM STP 1179, W.G. Landis, J.S. Hughes
and M.A. Lewis, eds., pp. 40-55. Philadelphia, PA: American
Society for Testing and Materials.
Nabholz, J.V., R. G. Clements, M. G. Zeeman, K. C. Osborn, and R.
Wedge. 1993b. "Validation of structure activity
relationships used by the USEPA's Office of Pollution
Prevention and Toxics for the environmental hazard assessment
of industrial chemicals," in Environmental toxicology and risk
assessment: 2nd volume, ASTM STP 1216, J. w. Gorsuch, F. J.
Dwyer, C. G. Ingersoll, and T. W. LaPoint, eds., pp. 571-590.
Philadelphia, PA: American Society for Testing and Materials.
Nalco Chemical Company. 1988. The NALCO water handbook. New
York, NY: McGraw-Hill Book Company.
National Research Council (NRC). 1994. Science and judgment in
risk assessment. Washington, DC: National Academy of
Sciences.
Opgenorth, H.J. 1992. "Polymeric materials polycarboxylates," in
-------
35
The Handbook of Environmental Chemistry, 0. Hutzinger, ed.,
Volume 3 Part F, Anthropogenic compounds. Detergents, N.T. de
Oude, ed., pp. 337-350. Berlin: Springer-Verlag.
Park, J.K., D. Jenkins, T.M. Holsen, T.W. Warnock and W.E.
Gledhill. 1989. "Fate of the detergent builder, sodium
polyglyoxylate, in wastewater treatment," Journal of the
Water Pollution Control Federation. 61:491-499.
Pranamuda, H., T. Tokiwa and H. Tanaka. 1995. "Microbial
degradation of an aliphatic polyester with a high melting
point, poly(tetramethylene succinate)," Applied and
Environmental Microbiology. 61:1828-1832.
Rodier, D.J. and D.A. Mauriello. 1993. "The quotient method of
ecological risk assessment and modeling under TSCA: a
review," in Environmental toxicology and risk assessment,
ASTM STP 1179, W.G. Landis, J.S. Hughes and M.A. Lewis,
eds., pp. 80-91. Philadelphia, PA: American Society for
Testing and Materials.
Rogers, J. R. Jr. and W. T. Witt. 1989. "Effects of sediments
flocculated with cationic polyelectrolytes when fed upon by
Daphnia magna. Denton, TX: Department of Biological
Sciences, University of North Texas. Available from SOCMA,
1330 Connecticut Avenue, N.W., Washington, DC 20036.
Smrchek, J., R. Clements, R. Morcock, and W. Rabert. 1993.
"Assessing ecological hazard under TSCA: methods and
evaluation of data," in Environmental toxicology and risk
assessment, ASTM STP 1179, W.G. Landis, J.S. Hughes and
M.A. Lewis, eds., pp. 22-39. Philadelphia, PA: American
Society for Testing and Materials.
United States Environmental Protection Agency (USEPA). 1982.
Fate of priority pollutants in publicly owned treatment
works, EPA report no. 440/1-82-303. Washington, DC: Office
of Water.
United States Environmental Protection Agency (USEPA). 1985.
.Exposure analysis modeling system (EXAMS): user manual and
system documentation, EPA report no. 600/3-82-023 (NTIS no.
PB82-258096). Washington, DC: Office of Research and
Development.
United States Environmental Protection Agency {USEPA). 1988.
User's guide to PDM3. Washington, DC: Office of Pollution
Prevention and Toxics, USEPA.
United States Environmental Protection Agency (USEPA). 1990.
Statistical methods for estimating exposure above the
reference dose, EPA report no. 600/8-90-065 (NTIS no. PB90-
261-504AS). Washington, DC: Office of Health and
Environmental Assessment, USEPA.
United States Environmental Protection Agency (USEPA). 1994a.
A review of ecological assessment cas studies from a risk
assessment perspective. Volume II, Part II, Section 1: New
chemical case study, EPA report no. 630-R-94-003. Washington,
DC: Office of Research and Development, USEPA.
United States Environmental Protection Agency (USEPA). 1994b.
U.S. EPA/EC joint project on the evaluation of (quantitative)
-------
36
structure activity relationships, EPA report no. 743-R-94-001.
Washington, DC: Office of Prevention, Pesticides, and Toxic
Substances, USEPA.
United States Environmental Protection Agency (USEPA). 1995a.
Amendment for expedited process to issue significant new use
rules for selected new chemical substances; final rule,
Federal .Register. 60:16311-16316.
United States Environmental Protection Agency (USEPA), 1995b.
Premanufacture notification for exemption; revision of
exemption for chemical substances manufactured in small
quantities; low release and exposure exemption; final rule,
Federal .Register. 60:16336-16351.
United States Environmental Protection Agency (USEPA). 1995c.
Premanufacture notification exemptions; revisions of
exemptions for polymers; final rule, Federal Register.
60:16316-16336.
United States Environmental Protection Agency (USEPA). 1995d.
New chemicals program. EPA report no. 743-F-95-001.
Washington, DC: Office of Pollution Prevention and Toxics,
USEPA.
United states Environmental Protection Agency (USEPA). 1996.
Polymer exemption guidance manual. Washington, DC: Office of
Pollution Prevention and Toxics, USEPA. Available through the
TSCA Assistance Information Service at 202-554-1404 (voice) or
202-554-5603 (facsimile).
Wagner, P.M., J. V. Nabholz, and R. J. Kent. 1995. "The new
chemicals process at the Environmental Protection Agency
(EPA): structure-activity relationships for hazard
identification and risk assessment," Toxicology Letters.
79:67-73.
Zeeman, M. and J. Gilford. 1993. "Ecological hazard evaluation
and risk assessment under EPA's Toxic Substances Control Act
(TSCA): an introduction," in .Environmental toxicology and
risk assessment, ASTM STP 1179, W.G. Landis, J.S. Hughes and
M.A. Lewis, eds., pp. 7-21. Philadelphia, PA: American
Society for Testing and Materials.
-------
FIGURE LEGENDS
FIGURE 1. Total number of valid Premanufacture Notices received
by EPA for the period 1979-1994.
FIGURE 2. Premanufacture Notice review process.
-------
-------
Number of Submissions
i
W
30
(D
o
2.
a
-------
-------
Duusaj.
IU9SUOO (a)s
I
bHNS
UJ80UOO
JO dOJQ
(*9IAajAep-&) SUOUdUiaxliaveHl&l*
ajnpejnuew oj aajj BJB
Aam jBgj pa^iioN Auedutoo
uo punoj
«u
snooj
wjodapofaftpsseg
sasBajay PUB
ajnsodjg p samoj
-------
-------
TABLE 1. Criteria for substantial or significant exposure in
Exposure-Based Review (XBR) of Premanufacture Notices
Type of exposure
Exposure criterion
Worker exposure:
substantial or significant
High number of workers
Acute worker inhalation
Chronic worker exposure
Inhalation
Dermal
Non-worker exposure
Consumer exposure via
direct contact with
consumer products
General population:
significant exposure
Surface drinking
water
Air
Ground drinking
water
Environment:
substantial exposure
Aggregate ambient
exposure through
surface water, air
and ground water
Substantial release
to water
>. 1,000 workers exposed
>_ 100 workers exposed by
inhalation to >. 10 mg/day
> 100 workers exposed to 1-10
mg/day for > 100 days/yr
.> 250 workers exposed by
routine dermal contact
> 100 days/yr
Presence of the chemical in any
consumer product where the
physical state or manner of use
make exposure likely
> 70 mg/yr
>. 70 mg/yr via ambient air
> 70 mg/yr
Total release to environmental
media > 10,000 kg/yr
Release to surface water
> 1,000 kg/yr after treatment
U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
-------
-------
TABLE 2. Fate testing strategy for aquatic exposures in
Exposure-Based Review (XBR)
Exposure criterion
Recommended tests
Production volume > 100,000
kg/yr and at least one other
XBR criterion8
Production volume > 454,000
(1,000,000 lb/yr) and at
least one other XBR criterion
Wastewater treatment removal:
Semi-Continuous Activated
Sludge (SCAS), Porous Pot, or
Activated Sludge Sorption
Isotherm
-and-
One of the following: water
solubility, vapor pressure,
soil/sediment sorption iso-
therm, ready biodegrada-
bility, or hydrolysis
Wastewater treatment removal:
as above
-and-
At least two of the following:
water solubility, vapor
pressure, soil/sediment sorption
isotherm, ready biodegrada-
bility, anaerobic biodegrada-
bility, or hydrolysis
General population: exposure via surface drinking water of
> 70 mg/yr; and environment: release to surface water of
> 1,000 kg/yr.
-------
-------
TABLE 3. Biodegradation and elimination of polycarboxylates in
tests with activated sludge8
Polycarboxylate5 Ultimate SCAS test
biodegradation (DOC
(TC02), %c loss), %d
P(AA) 1,000
P(AA) 2,000
P(AA) 4,500
P{AA) 10,000
P(AA) 60,000
P(AA-MA) 12,000
P(AA-MA) 70,000
45
20
9
16
ND
31
20
45
21
40
58
93
83
95
Confirmatory
test
(removal) , %
NDe
ND
27
ND
ND
70
82
8 Source: Opgenorth (1992).
b p = poly; AA = acrylic acid; MA = maleic acid; numbers given
are approximate MWn.
c Percent of theoretical C02 formation after contact with
activated sludge for 30-90 d.
d SCAS = Semi-Continuous Activated Sludge; DOC = dissolved
organic carbon.
e ND = not done.
-------
-------
1
Table 4. Aquatic toxicity data for polyanionic polymers:
poly{aromatic sulfonates) and poly(aromatic carboxylates).
Polymers are grouped by dominant monomer. Data elements are/ from
left to right: ID No. = Polymer identification number;
MWn/1/40/15 5.0 145.0 20.0
1/7/15 5.1 188.8 44.0
>0. 4/100/10 9.7 136.0 9.3
1/7/15 23.0 950.0 12.0
Biphenol sulfone-SO,H and cresol-SO,H
GAChV
<12 .5
7.0
18.0
<12. 5
<7. 5
6.2
7
7
8
9
10
11
1.1/0/0 3.2
Biphenol sulfone-S03H and biphenvl ether-S03H
120.0
0.6/95/40
0.6/95/40
12/7/0
1
80.0
32.0
2.0
510.0
3.0
<2.3
Phenol-S03H
Cresol-SO3H
-------
Table 4. Continued.
ID MWr/500.0 340.0
Benzene-S03H
600.0 900.0 800.0
44.2
-------
Table 5. Aquatic toxicity data for polyanionic polymers:
poly(aliphatic acids). Polymers are grouped by distance between
acids which is measured by counting the average number of polymer-
backbone-carbons between acids* Acids (A) include carboxylic acid,
phosphinic acid, and sulfonic acid. Data elements are, from left
to right: ID No. = Polymer identification number; KWn/339.0 560.0
16
16
0.5
0.5
1140.0
2500.0
Distance between acids; 1 carbon: [C(A)C],
17 3
18 3.5
19 2.5/5/3
20 0.9/49/19
21 10/35/5
22 1.4/41/8
330.0
>500.0
>225.0
>440.0
>500.0
>1000.0
7.44
3.13
37.4
11.0
7.6
5.5
Distance between acids; 1.5 carbons: [C(A)C(A)C]X
23 0.6/88/69 500.0 66.0
Distance between acids; 2 carbons; [C(A)CCC(A]X
24 3/0/0 150.0
280.0
24.0
4.7
0.500
0.540
93.0
Distance between acids; l carbon; distance from backbone;
3 carbons; [C(A)C]X + [C(CCCA)C]X
25 1/15/5 >1000.0 1800.0 57.0 36.0
Distance between acids; 1 carbon;chelating monomer diluted bv
non-chelatina monomer; [C(COOH)C]3 + [C(CONH2)C]1
26 >1/10/1 >40.0 >40.0 >500.0 >100.0
-------
-------
Table 6. Mitigation of toxicity towards green algae from
polyanionic polymers by calcium ions (Ca *). Calcium is added by
either increasing the hardness of the algal growth medium or
testing the Ca * salt of the polymer. Data have been grouped by the
average distance between acids in terms of polymer-backbone-
carbons. Data elements are, from left to right: ID No. = polymer
identification number; MWr/
-------
-------
Table 7. Aquatic toxicity data for polycationic polymers grouped
by type of polymer backbone. Data elements are, from left to
right: ID No. = Polymer identification number; A-N = cationic
charge density in percent amine nitrogen; MW(/5.0
>5.0
>5.0
1/38/12
1/32/23
Pll
P4
P4
P4
B
B
P2
P2
P5
P5
P4
P2
PI
P3
P4
P4
P4
P4.
P5
P4
B
B
B
B
B
3
4
4
4
3
2
3,4
3,4
3
3
3
3
4
2,3
4
4
4
5 4
4
4
4
4
4
4
2
4
9
.
.
8
3
3
10
10
30
17
1
0
0
1
.
0
2
0
1
0
0
0
.
0
0
0
0
0
0
0
0
*
6
2
.5
.9
.3
.61
.0
.0
0
.970
.3
.640
.2
.840
.900
.320
5
.600
.300
730
.310
.900
.760
.150
.160
.290
720
0
300
.0
310.0
28.2
40.0
16
0
0
0
2
.0
1.7
.090
.180
.073
.9
1.3 0.160
2.2 0.880
>360.0 130.0
0.520 0.270
0.035 0.006
0.025 0.013
0.014 0.006
0.015 0.007
-------
Table 7. continued.
ID
NO.
A-N MW,/ Cat. C
(%)
-------
Table 7. Continued.
ID A-N MW,/ Cat. Cat. F96
No. (%) 850
>1000
>1000
>1000
1
0
0
1.
570
*
*
m
*
5
*
0
0
0
0
18
370
860
0
117.
>1000.
>1000.
130.
26.0
11.
>1000.0
0
0
0
0
0.620
0 >480.0
>1000
.0
0.390
>480
.0
-------
-------
Table 8. Aquatic toxicity data for polyamphoteric polymers. Data
elements are, from left to right: ID No. = Polymer identification
number; A-N = cationic charge density in percent amine nitrogen;
MWr/
-------
-------
Table 9. Mitigation of aquatic toxicity from polycationic and
polyamphoteric polymers by humic acid. Data elements are, from
left to right: ID No. = Polymer identification number; A-N =
cationic charge density in percent amine nitrogen; MWr/
-------
Table 9. Continued.
ID
No.
44
44
44
44
47
47
47
82
82
82
82
82
82
82
82
82
48
48
48
83
83
83
54
54
54
54
54
56
56
56
56
56
58
58
58
58
58
58
A-N
(%)
4
4
4
4
6
6
6
6
6
6
6
6
6
6
6
6
7
7
7
8
8
8
11
11
11
11
11
14
14
14
14
14
14
14
14
14
14
14
*
*
*
m
*
*
*
»
*
»
*
*
*
*
*
*
6
6
6
6
4
4
4
6
6
6
6
6
6
6
6
6
8
8
8
4
4
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
MW,/
2
>2
>20
>20
0
0
3
, 5
10
0
0
0
2
5
9
*
*
*
*
*
*
*
900
0
0
0
720
2
9
280
380
600
6
6
1
2
0
0
0
0
0
0
0
0
200
0
0
0
0
100
160
0
36
2
072
084
160
2
4
6
Mitigation
Factor
76.0
21.0
290.0
26.0
27.0
>180.0
140.0
170.0
------- |