Priority Substances List Assessment Report for Nonylphenol and its Ethoxylates

2.0 Summary of Information Critical to Assessment of "Toxic" Under CEPA 1999

2.1 Identity and physical/chemical properties

2.1.1 Nomenclature

NP is a chemical intermediate composed of a phenol ring attached to a lipophilic straight or, more usually, branched nonyl group. NPEs belong to the larger group of compounds known as APEs and have the following general formula: C₁₅H₂₄O+(CH₂CH₂O)n. The predominant positional isomer of monoalkylphenols is the para isomer, which usually comprises =90% of industrial formulations, while the ortho isomer comprises = 10%. In the United States, the U.S. Environmental Protection Agency and the Chemical Manufacturers Association's Alkylphenols and Ethoxylates Panel have agreed that the commercial product that best represents "nonylphenol" is a chemical substance composed of branched C₉-alkylphenols with Chemical Abstracts Service (CAS) registry number 84852-15-3 (Hellyer, 1991). There may also be small amounts of 2,4-dinonylphenol in commercial nonylphenol preparations. CAS registry numbers for a variety of alkylphenols, ethoxylates and other derivatives are given in Talmage (1994). Considering the branching of the C₉-chain, there may be scores, if not hundreds, of individual NPE isomers in an industrial NPE formulation. Each NPE is conventionally described by its average ethoxylate (EO) chain length, which ranges between 1 and 100 for different formulations.

Structures of NPEs and associated degradation products are shown in Figure 1.

Figure 1 Chemical structures for NP, NPE, NP1EC and NP2EC

2.1.2 Physical and chemical properties

Some physical and chemical properties that have a bearing on the environmental persistence of NP and NPEs with average chain lengths of four (NP4EO) and nine (NP9EO) are summarized in Table 1. The properties of NP4EO and NP9EO are considered to be representative of NPEs and are presented because the available data set for these two compounds is the most complete.

Table 1 Properties of NP, NP4EO and NP9EO ¹

Property/specification	NP	NP4EO	NP9EO
CAS registry number	84852-15-3,	7311-27-5,	26027-38-3
	25154-52-3	9016-45-9
Synonyms	4-nonylphenol,	Nonoxynol-4	Nonoxynol-9,
	p-nonylphenol		Tergitol NP-9
Molecular formula	C15H24O	C15H24O- [C2H4O]4	C15H24O- [C2H4O]9
Molecular weight (g/mol)	220.3	396.2	617.6
Melting point (°C)	-8 2,3	-40 4	2.8 4
Boiling point (°C)	295-320 1,3
Physical characteristics	colourless to pale	white to light amber	almost colourless
	straw (liquid) 1,3	liquid 5	liquid 6
Specific gravity	0.953 7	1.020-1.030 (25°C) 6	1.057 (20°C) 4
pKa	10.7 8
Vapour pressure (Pa)	0.004 55 ± 0.003 5 8
	0.3 3
Solubility (mg/L)	5.4 9	7.65 9	"soluble" 5
log Kow	4.2-4.48 10,11,12	4.2410	3.5910
Henry's law constant (Pa·m3/mol)	11.02 3,13		0.000 245

Table 2 Amount of NP and its ethoxylates produced, imported, exported and available for use in Canada in 1995 and 1996 (from Environment Canada, 1998)

	Amount (tonnes NP/NPE)
	1995	1996
Produced in Canada	32 700	25 600
Imported to Canada	3 700	4 500
Exported from Canada	12 600	11 100
Available for use in Canada 1,2	23 800	19 000

Specific gravity, viscosity and aqueous solubility increase with EO chain length. NPEs with a chain length greater than six are readily soluble in water. It should be noted, for example, that the pK_a of NP is 10.7 (Romano, 1991), which indicates that in most natural waters, virtually all NP is present in the undissociated form. The Henry's law constant and vapour pressure of NP and especially NPEs are low; therefore, partitioning to air is extremely limited.

NPECs are likely to be substantially, if not almost completely, ionized at the pH values of many natural waters (e.g., the pK_a of unsubstituted phenoxyacetic acid has been estimated as 5.12; NTP, 1998), and their log K_ow values are expected to be much lower than those of the corresponding ethoxylates (e.g., the log K_ow value of unsubstituted phenoxyacetic acid has been estimated as 1.34; Syracuse Research Corp., 1998).

2.2 Entry characterization

2.2.1 Uses, production and market trends in Canada

NP is produced by reacting phenol and mixed nonenes in the presence of a catalyst. It is used in the production of NPEs, as a monomer in polymer production and as an additive in polymer processing (U.K. Environment Agency, 1998). NPEs are manufactured by the ethoxylation of NP with ethylene oxide and used in the production of phenol/formaldehyde resins; in the production of tris(nonylphenyl)phosphite (TNPP) antioxidant for rubber and manufacture of lube oil additives; as a catalyst in the curing of epoxy resins; and in the manufacture of phenolic oximes (used in the extraction of copper from ores). The length of the EO chain is varied by controlling the reaction time or the ratio of NP to ethylene oxide.

Over 4500 companies operating in Canada were surveyed in 1997 under authority of Section 16 of the Canadian Environmental Protection Act (CEPA) to determine the uses of priority chemicals (Environment Canada, 1997b). Data were collected on the amount of NP and NPEs produced, imported, exported, shipped, acquired and used in Canada. A total of 189 companies responded that they were involved with NP and NPEs above the trigger quantity of 1000 kg per year (Environment Canada, 1997c).

The amount of NP plus NPEs available for use in Canada (domestic production plus imports minus exports) was 23 800 and 19 000 tonnes in 1995 and 1996, respectively (Table 2). It is not known how much of those totals refer to NP and how much to NPEs. NPEs were manufactured at three facilities in Canada in 1995 and 1996. They are used, in descending order, as feedstock, formulation, articles, chemical aid, manufacturing aid and containers.² As reported in surveys of Canadian industry carried out under authority of Section 16 of CEPA, the total amount of NP used by industry in Canada in 1996 was 5000 tonnes, with the majority being used as chemical intermediates. The total reported use of NPEs in the same year was also 5000 tonnes (Environment Canada, 1997c).

In 1989, domestic demand for NP in Canada was 4500 tonnes: 3700 tonnes were from domestic production, 1800 tonnes were imported and 1000 tonnes were exported (Camford Information Services Inc., 1990). In that year, 1000 tonnes were used in ethoxylated textile specialties, 1600 tonnes in ethoxylated pulp mill specialties, 500 tonnes in miscellaneous ethoxylates, 500 tonnes for TNPP and 900 tonnes for miscellaneous uses, including pesticides and lube oil (Camford Information Services Inc., 1990).

Although forecasts for demand of NP and NPEs in Canada were not available, the growth for NP in the United States was 2% per year in the period 1988-1997 and was forecast at 1-2% per year from 1998 through 2002 (Anonymous, 1998).

Because of the surfactant properties of NPEs, NPE-containing products have many industrial, commercial, institutional and household uses in Canada, including lubrication, defoaming, assisting in dyeing, as emulsifiers, controlling deposits and cleaning machinery and materials, scouring fibres, as wetting and de-wetting agents and in product finishing. NPE-containing products are used in many sectors in Canada, including textile processing, pulp and paper manufacturing, metal processing, petroleum refining, oil and gas recovery, power generation, food and beverage processing, plastics manufacture, and the building and construction industry, as well as in various cleaning products, paints, resins and protective coatings (Talmage, 1994; Maguire, 1999). As well as being used in industry, a variety of cleaning products, degreasers and detergents are also available for institutional and domestic use. Consequently, there are many routes of entry into the environment for these substances during the course of their manufacture, use and disposal.

Two hundred and eleven pesticide products containing NP and/or NPEs were identified for current use in Canada. Forty percent of these products contain less than 1% NP or NPE, 85% of these products contain less than 10% NP or NPE and 95% of these products contain less than 20% NP or NPE. The NP and NPEs appear only as formulants in these pesticides, primarily as emulsifiers, surfactants, wetting agents, etc. The application rate of these products varies considerably, depending on their use (Moore, 1999).

It is not known if NP is used to mark fuel oil for taxation purposes in Canada; however, this has been identified as a use pattern in the United States (Reed, 1978).

NPEs have been prohibited since 1997 as an active ingredient in soil supplements that are regulated under the Fertilizers Act (Webster, 1998).

A range of other applications, some of which appear to have limited potential for release to the environment, nonetheless represent direct sources of exposure for humans. Uses that may result in residues of NP and NPEs in foods include the use of NP and its ethoxylates as a "spreader" in several pesticide formulations, as a dispersant/emulsifier in vegetable and fruit waxes, in detergents and disinfectants used on foods and in various food packaging applications (World Wildlife Fund Canada, 1996). NP is also a contaminant and breakdown product of TNPP, used as a polymer resin in food contact packaging (The Society of the Plastics Industry, Inc., 1998a), and has been reported to leach from polyvinyl chloride (PVC) polymer tubing into water passed through the tube (Junk et al., 1974) and from PVC bottles into food simulants (Gilbert et al., 1986). A wide range of consumer products contain NP and NPEs. Cosmetic products such as skin creams, eye and face makeup, hair care products, deodorants and bath products, as well as cleaning products and paints, may be direct sources of exposure to these substances. Nonoxynol-9 (NP9EO) is also used as a spermi-cide in contraceptive foams, jellies and creams (McIntyre, 1996; World Wildlife Fund Canada, 1997; WHO, 1998). Specific information on the percent composition of NP/NPEs in various consumer products is presented in Tabl e 3.

2.2.2 Sources and releases

2.2.2.1 Natural sources

There are no known natural sources of NP and NPEs. Their presence in the environment is, therefore, solely a consequence of anthropogenic activity.

2.2.2.2 Anthropogenic sources

Releases of NP and NPEs to the environment can occur at various points in the product life cycle - namely, during primary production of NPEs, manufacture of NPE-containing products, product use and disposal of the product to wastewater treatment, septic system or landfill.

Over 4500 companies operating in Canada were surveyed in 1997 under authority of Section 16 of CEPA to determine their releases of priority chemicals to the environment (Environment Canada, 1997b). Ranges of releases in 1996 from 65 companies in Canada involved in primary NPE production, manufacture of NPE-containing products or industrial use are shown in Table 4. It is not known how much of these totals refers to NP and how much to NPEs. The total release of NP and NPEs combined in 1996 from industrial manufacture and use was 96.5 tonnes (Environment Canada, 1997c). The largest industrial releasers were (1) formulators and distributors of surfactants and (2) industrial users of cleaning products, degreasers and detergents, which each released between 25 and 60 tonnes of NP and NPEs in 1996. Together, these two groups of industries accounted for the majority of total releases from industrial sources. Producers of paints, protective coatings, resins and adhesives released between 5.000 and 9.999 tonnes per year. Releases of between 0.100 and 4.999 tonnes of NP and NPEs in 1996 were reported for each of the following industries: formulators of industrial, institutional and domestic cleaning products, degreasers and detergents; pulp and paper mills; oil and gas recovery; producers of wastewater treatment products; formulators and distributors of products for the pulp and paper industry; and miscellaneous industries.

Table 3 Concentrations of NP and its ethoxylates in consumer products

Product	Average ethoxylate chain lengths	Range of concentrations1	Reference
Antiwrinkle preparation	8	>1 to 3%	McIntyre, 1996
Bath preparation	10, 12, 20, 40	>1 to 3%	McIntyre, 1996
Deodorant	10, 12, 14, 18	>1 to 3%	McIntyre, 1996
Eye makeup	Nonylphenol, 10, 15	>3 to 10%	McIntyre, 1996
Face makeup	10	0.1% or less	McIntyre, 1996
Fragrance	12, 14	>3 to 10%	McIntyre, 1996
Genital lubricant	9	>1 to 3%	McIntyre, 1996
Hair bleach	4, 6, 9, 49	>30 to 100%	McIntyre, 1996
Hair conditioner	10, 14, 23	>3 to 10%	McIntyre, 1996
Hair dye	1, 2, 4, 6, 9, 10, 49	>30 to 100%	McIntyre, 1996
Hair grooming products	4, 9, 10, 11, 12, 15, 23	>3 to 10%	McIntyre, 1996
Hair removal products	10	>0.1 to 0.3%	McIntyre, 1996
Hair shampoo	Nonylphenol, 4, 10, 12, 15	>10 to 30%	McIntyre, 1996
Hair straightener	10	>1 to 3%	McIntyre, 1996
Hair waving preparation	4, 9, 10, 11, 12, 14, 15, 23, 30	>3 to 10%	McIntyre, 1996
Manicure preparation	7, 14, 100	>30 to 100%	McIntyre, 1996
Skin cleanser	4, 8, 9, 10, 12, 14, 15	>3 to 10%	McIntyre, 1996
Skin moisturizer	4, 5, 6, 9, 10, 12, 14, 15	>3 to 10%	McIntyre, 1996
All-purpose spray cleaner	not specified)2	<0.2 to 5%	World Wildlife Fund 1997
Stain remover	not specified 2	<0.2 to 11%	World Wildlife Fund 1997
Liquid laundry detergent	not specified 2	<0.2 to 28%	World Wildlife Fund 1997
Paints	8-9, 15-20	0.6 to 3%	WHO, 1998

It is important to note that industrial releases do not indicate the total release of NP and NPEs to the Canadian environment. For instance, individual households and institutions using NPE-containing products were not contacted under the Section 16 survey, and these releases are, therefore, not reported. Nevertheless, these products are generally disposed of "down the drain" and are released to municipal water treatment facilities or septic systems. This source is likely to be significant to the NPE loading at wastewater treatment facilities.

2.3 Exposure characterization

2.3.1 Environmental fate

With regard to the behaviour of a chemical in the environment, it should be noted that there are many factors that influence its persistence, including its physical and chemical properties and ecosystem-specific properties, such as (for aquatic ecosystems) the nature and concentration of microbial populations, the nature and concentration of dissolved and suspended material, temperature, degree of insolation, etc. In general, important physical, chemical and biological removal mechanisms for chemicals in aquatic ecosystems are, (i) volatilization and adsorption to suspended solids and sediment, (ii) chemical and photochemical degradation or transformation and (iii) uptake and transformation by microorganisms, respectively. The variation in the physical/chemical properties of NP/NPEs and their rapid conversion to other metabolites make their environmental fate extremely complex.

Table 4 Releases of NP and its ethoxylates to various environmental media, by industry sector in Canada in 1996 (from Environment Canada, 1998) ^1,2

Industry sector	No. of sites 3	Total (kg) released to				Total (kg, range) released from all sites
		Air	Stream	Waste-water	Landfill/or deep well
Formulators and distributors of surfactants	4					25 000-60 000
Industrial users of cleaning products, degreasers and detergents	3					25 000-60 000
Producers of paints, protective coatings, resins and adhesives	19					5000-9999
Formulators of industrial, institutional and domestic cleaning products, degreasers and detergents	22					100-4999
Pulp and paper mills	3					100-4999
Oil and gas recovery	2					100-4999
Production of wastewater treatment products 4	2					100-4999
Formulators and distributors of products for the pulp and paper industry	6					100-4999
Miscellaneous	4					100-4999
Total	65

2.3.1.1 Air

NP and NPEs are not expected to readily volatilize into air and are expected to degrade rapidly in the atmosphere. Dachs et al. (1999) detected NP (sum of 11 isomers) in all atmospheric samples collected from the urban and coastal regions of the Lower Hudson Estuary and predicted that NP may volatilize out of water into the air in areas where NP concentrations are elevated in surface waters, although the Henry's law constant is low. The U.K. Environment Agency (1998) has estimated a half-life of 0.3 days for the reaction of hydroxyl radicals with NP in the atmosphere, indicating that it would be unlikely for any NP in air to be transported to remote regions. NPEs are far less volatile than NP, and thus it is expected that they would not partition to the atmosphere. Because of the presence of NPEs in aerially applied pesticide formulations, however, there is a need to determine their atmospheric chemistry, photochemistry and fate.

2.3.1.2 Water and sediment

2.3.1.2.1 Degradation in water in laboratory tests

Although there are some conflicting reports in the literature, in general NPEs and NP are not readily biodegradable using standard test methods. Substantial biodegradation will occur after a period of acclimation. NPEs are, therefore, inherently biodegradable, and the mechanism involves stepwise loss of ethoxy groups to lower NPE congeners, followed by the production of NPEC and NP, depending upon experimental conditions (Rudling and Solyom, 1974; Maki et al. 1994). The degradation pathway is shown in Figure 2. This pathway is an oversimplification because it does not include NPnEC where n > 2 or NPEs with carboxyl groups attached to the nonyl chain. The intermediate and final products of metabolism are more persistent than the parent NPEs, but it is believed that such chemicals will also be ultimately biodegraded. Branching of the nonyl group in NP and NPEs retards biodegradation, as does increase in length of the EO chain. APs and APEs are more persistent than alkylbenzene sulfonates and alcohol ethoxylates (Kravetz et al., 1991; Maguire, 1999). It should also be noted that the use of high concentrations of chemicals in biodegradability tests may result in artificially high persistence data if the chemical poisons the test organisms. This possibility has been suggested to account for some differences in results for the biodegradability of NPEs (e.g., U.K. Environment Agency, 1997).

2.3.1.2.2 Degradation in municipal wastewater treatment plants

It has been noted that full-scale municipal wastewater treatment plants (MWWTPs) can provide greater efficiencies for the removal of NPEs than can bench-scale systems, which may be due to a greater variety of microbial populations and nutrients in the former (Holt et al., 1992). In general, primary biodegradation of NPEs in MWWTPs is readily achievable, but ultimate biodegradation is not. Substantial differences in treatment efficiencies for NPEs and their degradation products exist among MWWTPs. These differences have been attributed to the load of NPEs in influent streams and MWWTP design and operating conditions, including temperature of treatment. In some locations, more persistent products such as NP and lower-chain NPEs have been observed in MWWTP final effluents and receiving waters. In addition, substantial concentrations of NP and lower-chain NPEs are found in sludges from MWWTPs. The application of NP-containing sludges to agricultural land may result in potential exposure in terrestrial environments.

In general, primary biological degradation of NPEs is the major pathway and occurs more rapidly in MWWTPs than in natural environments because of the higher concentration of microorganisms in MWWTPs compared with natural environments. Most municipalities in Canada have some type of wastewater treatment. MWWTPs play a significant role in the transformation and degradation of NP and NPEs before their entry into the environment. More than 60% of the higher-chain APEs that enter MWWTPs exit as stable metabolites (e.g., APs and short-chain APEs) in either their effluents or their sludges (Ahel et al., 1994a).

Figure 2 Biological degradation pathway for NPEs

In general, once APEs, including NPEs are released to municipal wastewater systems, several transformations can occur. APEs with more than eight EO units (most common commercial products) are readily degraded in effluent treatment systems with >92% efficiency (Brunner et al., 1988; Kubeck and Naylor, 1990; Ahel et al., 1994a,b; Naylor, 1995). Under aerobic and anaerobic treatment conditions, the biodegradation mechanism involves an initial loss of ethoxy groups, leading to the production of NP1EO and NP2EO and their carboxylate derivatives NP1EC and NP2EC (as well as NPnEC, where n > 2, and CAPECs, alkyphenols with carboxylate groups on both the alkyl and ethoxylate chains and CAPEs, alkylphenols with carboxylated alkyl chain) and the final product, NP. The wastewater treatment, therefore, results in chemical transformation to compounds that are more persistent, toxic and estrogenic than the parent NPEs. Structurally analogous metabolites are formed by the degradation of OPEs and other APEs in MWWTPs. Additionally, there is evidence for the halogenation of some of these products in MWWTPs that use chlorine for disinfection (Maguire, 1999).

Di Corcia et al. (1998) studied CAPECs experimentally and in effluents. These APE metabolites were found at higher concentrations than CAPEs and persisted in experimental media for more than five months. CAPECs represented 63% of the total APE metabolites present in MWWTPs at concentrations of 58 µg/L (Di Corcia et al., 1998). Ding et al., (1996) reported CAPECs in MWWTPs effluents at concentrations ranging from 0.9 to 1.1 µg/L.

Discharges from MWWTPs provide the two major routes for environmental release of NPEs and their degradation products. The first route is by discharge of the final treated effluent to nearby receiving waters. The second major environmental release route for nonylphenolics is via sewage treatment processes and adsorption onto sludge. NP (in particular) and NP1EO and NP2EO are more lipophilic than the parent NPEs and tend to accumulate in sludges and sediments, while NPECs (which are more water soluble and can substantially, or completely, ionize at the pH of most natural waters) are generally found in the final effluents, sometimes at much higher concentrations than other nonylphenolic compounds. NP, NP1EO and NP2EO, however, also have been found in effluents and receiving waters. Most Canadian MWWTPs employing secondary or tertiary treatment utilize the activated sludge process (an anaerobic digestion process), which results in the sorption of NP, the dominant nonylphenolic substance, onto sludge. Up to 95% of the nonylphenolic composition of digested sludge may be attributed to NP. Sludge is disposed of in three ways - by incineration, by landfilling and by spreading on agricultural soils. Although there has generally been little research on the fate of nonylphenolics in sludge disposed of by any of these three techniques, some studies have examined the fate of NP in landfills (Maguire, 1999).

The final effluent composition is dependent on the treatment process(es) used in the facility. Where only primary treatment is used, the effluent composition reflects the short hydraulic retention time, and ethoxylated products (i.e., NPnEO, where n = 3-20) are dominant (82%), with minor components of NP (3%), NP1EO and NP2EO (12%), and NP1EC and NP2EC (3%). Secondary-treated effluent composition is substantially different from that of primary-treated effluent (Figure 3). Higher-chain NPEs make up only 28% of the nonylphenolic compounds in secondary-treated effluent, whereas meta bolites make up the rest (Figure 3). Carboxylic acid metabolites (i.e., NP1EC and NP2EC) account for about 46% of the secondary-treated effluent composition, while NP1EO and NP2EO make up 22% and NP accounts for only 4% (Ahel et al., 1994a) (Figure 3). Buildup of NP, however, has been observed in activated sludge and digested sludge from MWWTPs utilizing secondary or tertiary treatment systems (Giger et al., 1987) (Figure 4). Additionally, some production of NP1EO and NP2EO was observed in the digested sludge.

Birch (1991) and Watkinson and Holt (1991) noted that a critical control parameter for the treatment of NPEs in MWWTPs with activated sludge plants is the sludge retention time (SRT). This parameter dictates the necessary growth rate for the competent organisms within the total microbial population. When the growth rate of the organisms is less than the SRT, the competent organisms are washed out of the system, and little treatment of the specific substance takes place. The growth rate of organisms is influenced by temperature, and thus a combination of decreasing SRT and decreasing temperature will result in a less effective biodegradation system. Watkinson and Holt (1991) noted that the normal range of SRTs for activated sludge plants would appear to be in the range 6-20 days. Ahel et al. (1994b) noted that the highest NPE elimination rates were achieved in the MWWTPs characterized by low sludge loading rates and nitrifying conditions. This was confirmed in a limited study of two Canadian MWWTPs (Water Technology International Corp., 1998b). It should also be noted that there may be substantial differences in treatment efficiencies of NPEs between dedicated industrial wastewater treatment facilities and MWWTPs. Field and Reed (1996) reported that industrial wastewater treatment can be characterized by higher temperatures, increased hydraulic residence times and greater degrees of acclimation than that of MWWTPs. Because MWWTPs operate at ambient temperatures, more seasonal variation in effluent composition would be expected from MWWTPs than from industrial effluents.

Figure 3 Concentration of NP, lower-chain NPEs and NPECs in various types of municipal wastewater treatment plant effluents (Windsor = primary treatment; Burlington = secondary treatment; Galt, Guelph and Edmonton = tertiary treatment) (data from Bennie, 1998a)

Figure 4 Distribution of NP, NPEs and lower-chain NPECs in effluent and sludge from a tertiary-treated municipal wastewater treatment plant. Sample taken September 1997, representing the mean of eight 24-hour composites (Water Technology International Corp., 1998b)

2.3.1.2.3 Degradation in water and sediment

Primary biodegradation of higher-chain NPEs is generally faster than ultimate degradation of more persistent products, such as NP1EO, NP2EO, NP1EC, NP2EC and NP (Ahel et al., 1994b). Microbial acclimation to such chemicals is required for optimal degradation efficiencies (Maguire, 1999). Photodegradation of NP and lower-chain NPEs is also expected to be important. In aquatic ecosystems, it appears that parent NPEs are not persistent, although some degradation products may have moderate persistence, especially under anaerobic conditions. It should be noted that the U.K. Environment Agency (1997) estimated a biodegradation half-life of about 150 days for NP in surface water. Based on the limited data available, NP and the lower NPEs and NPECs are expected to be persistent in groundwater. The recent results of Heinis et al. (1999) indicate that NP can be moderately persistent in sediments. It is expected that the more water soluble (and ionized) carboxylate derivatives NP1EC and NP2EC will largely remain in the aqueous phase.

2.3.1.3 Soil

Although there are relatively few studies of NP and NPEs in soil, NP has been found to persist in landfills under anaerobic conditions; however, it does not appear to be persistent in soil under aerobic conditions (Marcomini et al., 1991). Based on results of laboratory biodegradation studies, Hughes et al., (1996) reported that NP9EO would be expected to biodegrade in soil under aerobic conditions. The U.K. Environment Agency (1997) estimated a half-life of about 30 days for primary biodegradation of NP in soil and of 300 days for ultimate mineralization. Studies conducted by Water Technology International Corp. (1998a) demonstrated similar results in Canadian soils. When sludge was added to soil, the concentration of NP initially increased, followed by a decrease to below detection limits within 120 days.

Sewage sludges are commonly applied to soils in Canada. Studies on the persistence of NP in soils indicate that NP can be rapidly degraded to carbon dioxide by soil microorganisms (Topp, 1999). NP, at concentrations as high as 250 mg/kg, was rapidly mineralized by soil organisms in cultivated agricultural soils at 4°C, temperate, non-cultivated soils and arctic soils. The lack of a lag phase in the mineralization indicated that the soil contained active microflora, conditioned to mineralize other natural phenols in soils. A study conducted at the Guelph Turfgrass Institute by Bennie et al. (1998) demonstrated a rapid disappearance of initial concentrations of 5.5 mg NP/kg soil in sludge-treated soil plots. NP concentrations were undetectable after 90 days. The NPEs may be degraded to NP in the soils, therefore resulting in the non-linear disappearance of NP in the soils after sludge application. Bokern et al. (1998) concluded that the uptake of NP from soil was slow and that NP was quickly mineralized by soil microorganisms.

2.3.2 Environmental distribution

Fugacity modelling was carried out to provide an overview of key reaction, intercompartment and advection (movement out of a system) pathways for NP and its overall distribution in the environment. Fugacity modelling for NPEs was not considered appropriate, given the tendency of surfactants of this type to accumulate at media interfaces. A steady-state, non-equilibrium model (EQC Level III fugacity model, Version 1.01; May 1997) was run using the methods developed by Mackay (1991) and Mackay and Paterson (1991). Values for input parameters were as follows: molecular weight, 220 g/mol; aqueous solubility, 6 mg/L; vapour pressure, 0.3 Pa; log K_ow, 4.3; melting point -8°C. A "low half-lives" scenario and a "high half-lives" scenario were constructed to "bracket" the half-lives suggested by data referred to in this Assessment Report. The following values were used in the modelling: half-life in air, 5-17 hours; half-life in water, 1700-5500 hours; half-life in soil, 550-1700 hours; and half-life in sediment, 17 000-55 000 hours.

The results of this modelling indicate that NP partitions differently depending on the medium to which it is released. For example, if emitted into air only, more than two thirds of the NP that remains at steady state is predicted to be present in air (67-76%), with lower fractions in water (12%), sediment (8-12%) and soil (3-8%). When NP is released to water, the model predicts that most of it is present in water (49-59%) and, to a lesser extent, sediment (41-50%), with a negligible proportion (<1%) in air and soil. If released to soil only, virtually all (>99%) of the NP is predicted to be present in the soil compartment.

2.3.3 Environmental concentrations

Concentrations of NP and NPEs meas ured in environmental samples are summarized in Table 5, based on Canadian data when such data are available. A limited number of studies have reported the environmental occurrence of NP and NPEs in Canada. Together with available unpublished data, the reported environmental concentrations of NP and NPEs in effluents, sludges, surface waters and aquatic sediments are listed in Appendix A of the environmental supporting document (Servos et al., 2000). Concentrations of NP and NPEs found in Canadian sediments, effluents and sludges in Canada are similar to those found in other countries (Servos et al., 2000).

2 .3.3.1 Air

No data were identified on concentrations of NP and its ethoxylates, or related compounds, in Canadian air. Dachs et al. (1999) detected NP (11 isomers) in all samples of ambient air from urban and coastal areas of the Lower Hudson Estuary. Concentrations of NP in air of the New York-New Jersey Bight ranged from 2.2 to 70 ng/m³. No data on the levels of NPEs in ambient air were identified, although, based on the fact that they are far less volatile than NP, it is expected that they would not partition to the atmosphere.

2.3.3.2 Water and effluents

In general, NPEs are found at high concentrations (maximum concentration 8811 µg/L) in untreated or partially treated industrial (e.g., textile mills) and municipal effluents in Canada. Untreated effluents typically have elevated NP3-17EO concentrations and relatively high levels of NP and NP1,2EO (Table 5). Treatment significantly reduces the concentration of NP3-17EO in final effluents. The levels of NP3-17EO, NP1,2EO and NP in final effluents can, therefore, vary dramatically, depending on the type and degree of treatment. Well-treated effluents typically have very low levels of NP3-17EO. As higher-chain-length NPEs move through the treatment system, they are degraded to lower-chain-length NPEs and NPECs and ultimately to NP, which itself can be further degraded or sorbed to particles or sludges. Although NP1EO and NP2EO are created during treatment, concentrations of these transformation products are generally reduced in well-treated effluents. In contrast, NP1EC and NP2EC can increase in concentration with increased degree of treatment (Figure 3). The nature of the inputs and type and degree of treatment strongly influence the concentrations and relative proportions of NPEs released in final effluents. The relative distribution and concentrations of NPEs in influent, final effluent and sludges are, therefore, very different (Figure 4).

As the EO chain length decreases, a corresponding decrease in water solubility is observed. NP is, therefore, generally associated with organic particles, sludges in the treatment system and ultimately sediments in the environment. NPECs, however, are considerably more water soluble than the corresponding NPEs and are present in the aqueous phase of final effluent.

In Canadian fresh water, concentrations of NP ranged from non-detectable (<0.02 µg/L) to 4.25 µg/L (mean 0.20 µg/L; median <0.02 µg/L) (42 sites; n = 126) (Bennie et al. 1997; Bennie, 1998a) (Table 5). Highest freshwater concentrations of NP were observed in areas in close proximity to MWWTP discharges, pulp mill discharges, large population centres or regions of heavy industry. The different types of sites sampled included rivers, lakes (primarily Great Lakes) and harbours. In rivers across Canada, the concentrations of NP ranged from <0.02 to 4.25 µg/L, although Carey et al. (1981) reported values up to 2600 µg/L in Canagagigue Creek in Elmira, Ontario. These latter values were not considered representative because they were associated with an industrial spill into this small creek. NP concentrations in lakes ranged from <0.02 to 0.06 µg/L, and NP concentrations in harbours were between <0.02 and 0.98 µg/L. The maximum NP1EO concentration in rivers (2.30 µg/L) was lower than those observed in lakes (5.07 µg/L) or harbours (10.3 µg/L); however, the maximum NP2EO level in rivers (2.45 µg/L) was higher than that in lakes, which was below the detection limit (<0.02 µg/L) (Table 5). The maximum NP2EO concentration observed in harbours was 10.4 µg/L. Levels of NP3-17EO detected in two rivers in southern Ontario ranged from 0.11 to 17.6 µg/L (mean 1.41 µg/L; median 0.39 µg/L; n = 27 at three sites).

Table 5 Ranges of concentrations of NPEs in the Canadian environment (total number of sites, total number of samples)

Enlarge table

Concentrations of NP in untreated textile mill effluents ranged from 2.68 to 13.3 µg/L (Table 5). NP levels in effluent from on-site treated textile mills ranged from 0.09 to 3.56 µg/L. NP concentrations in textile mill effluents that discharge through a municipal treatment facility were between 0.23 and 25.6 µg/L (Bennie, 1998a). NP1EO and NP2EO concentrations in final effluents from textile mills were dependent on type of effluent treatment. Highest concentrations were observed in effluents that were not subject to treatment (Bennie, 1998a). Higher-chain NPEs were found at higher concentrations (2.07-8811 µg/L) than NP or the lower-chain NPEs in textile mill effluents (Table 5).

Pulp mill effluent samples taken in the years 1990-1993 showed that NP concentrations were quite variable (Bennie, 1998a). Due to recent changes to reduce the use of NPEs in Canadian pulp and paper mill processes, NP concentrations in final effluent from pulp and paper mills were divided into those values obtained prior to 1998 (<0.02-26.2 µg/L) (Bennie, 1998a) and those obtained more recently (<0.10-4.3 µg/L) (Lee and Peart, 1999) for this assessment. Lee and Peart (1999) reported NP1EO, NP2EO and NP3-17EO concentrations in pulp and paper mill effluents ranging from <0.10 to 6.90 µg/L, from <0.10 to 35.6 µg/L for NP2EO and from 5.90 to 28.8 µg/L, respectively (Table 5).

NP was below method detection limits (<0.02 µg/L) in two effluent samples from a major Canadian oil refinery (Bennie, 1998a).

MWWTPs equipped with primary, secondary and tertiary treatment systems have been sampled across Canada (Bennie, 1998a; Lee et al., 1998; Water Technology International Corp., 1998b). Final effluents contained concentrations of NP that ranged from <0.02 to 62.1 µg/L, from 0.12 to 4.79 µg/L and from <0.02 to 3.20 µg/L for primary, secondary and tertiary treatment systems, respectively (Table 5). MWWTPs that use a lagoon system had NP effluent concentrations ranging from 0.75 to 2.15 µg/L (Bennie, 1998a; Water Technology International Corp., 1998b). Bennie et al. (1998) measured NP concentrations in raw sewage from nine communities in eastern Canada (0.69-156 µg/L). The highest concentrations (>100 µg/L) were associated with two cities where textile mill inputs to the municipal treatment system were significant. NP1EO concentrations in MWWTP effluents ranged from <0.02 to 56.1 µg/L, with highest concentrations observed in effluents from primary treatment systems. NP2EO concentration trends were similar (<0.02-36.3 µg/L), with the highest maximum concentrations observed in primary effluents. NP3-17EO concentrations in final effluents from primary treatment plants were greater than secondary, and, similarly, a decrease in concentration in final effluent from tertiary MWWTPs was observed relative to secondary treatment effluents. NP1,2EC concentrations ranged from 1.01 to 75.0 µg/L in effluents from MWWTPs, with maximum concentrations occurring in effluents from facilities with secondary or tertiary treatment.

2.3.3.3 Sediments

The water-soluble NPEs and NPECs are present in the aqueous phase (Table 5). In contrast, the hydrophobic transformation products of NPE degradation - NP and NP1,2EO - are generally sorbed to sediments. Most of the APE concentration data in Canadian sediments are for NP. Very few data for the NPEs are available in the literature. A recent study by Shang et al. (1999) on the distribution of NPEs in marine sediments from the Strait of Georgia observed a shift from the dominance of NPEs with 8-10 EO units in commercial products to NP and NP1EO in sediments. These authors also concluded that NPEs were relatively persistent in aquatic sediments.

NP concentrations in sediments from the Great Lakes basin and the upper St. Lawrence River ranged from below detection levels (<0.02 µg/g dry weight [dw]) to 72.2 µg/g dw (Lee and Peart, 1995; Bennie et al., 1997; Bennett and Metcalfe, 1998; Bennie, 1998a) (Table 5). NP concentrations measured in sediment from the upper and lower reaches of the Fraser River and the Thompson River sub-basin ranged from <0.02 to 0.57 µg/g (Brewer et al., 1998). The highest NP concentrations in the Great Lakes data set are associated with Hamilton Harbour samples taken near the discharge of the MWWTP at Burlington, Ontario. The mean Canadian NP sediment conce ntration was determined to be 4.46 µg/g, with a median value of 0.21 µg/g (n = 58 at 23 sites).

NP1EO concentrations in sediment from the upper St. Lawrence River and Great Lakes basin ranged from <0.02 to 38.12 µg/g dw, with mean levels of 3.13 µg/g dw and a median of <0.03 µg/g dw (n = 14 at six sites) (Bennie et al., 1997; Bennie, 1998a). In the same study, NP2EO concentrations in sediment ranged from <0.02 to 6.02 µg/g dw, with a mean of 0.51 µg/g dw and a median of <0.02 µg/g dw (n = 14 at six sites). NP3-17EO sediment concentrations measured at one site in Ontario ranged from <0.02 to 0.17 µg/g dw (mean 0.05 µg/g dw; median 0.02 µg/g dw; n = 4) (Bennie, 1998a). No NPEC concentration data were identified in Canadian sediments.

2.3.3.4 Sludges

Lee and Peart (1995), Lee et al. (1997, 1998), Bennie (1998a), Bennie et al. (1998) and Water Technology International Corp. (1998a) determined NP concentrations in sludge samples from MWWTPs across Canada. Levels ranged from 0.74 to 1260 µg/g dw, with a mean concentration of 299.28 µg/g dw (n = 107 at 30 sites); the median value was 217.27 µg/g dw. The highest values were associated with MWWTPs that utilize anaerobic secondary sludge digestion processes. NP1EO and NP2EO were reported in sludge at concentrations ranging from 2.9 to 1825 µg/g dw and from 1.5 to 297 µg/g dw, respectively. NP2EO in sludge is usually present at much lower concentrations than NP1EO, but in some samples concentrations of NP2EO were slightly higher. NP3-17EO concentrations ranged from 0.43 to 215 µg/g dw (mean 49.58 µg/g dw; median 47.60 µg/g dw; n = 90 at 28 sites). NP1EC concentrations were reported to be between <0.30 and 8.70 µg/g dw (mean 2.53 µg/g dw; median 2.26 µg/g dw; n = 66 at 17 sites), while the range of NP2EC concentrations was reported to be <0.30-26.0 µg/g dw (mean 9.27 µg/g dw; median 9.56 µg/g dw; n = 66 at 17 sites).

Lee et al. (1997, 1998) reported concentrations in digested sludge from nine MWWTPs ranging from <0.5 to 25 µg/g dw for NP1EC and from <0.5 to 38 µg/g for NP2EC. Detectable levels of NP1EC (2.8-6.6 µg/g dw) and NP2EC (7.1-23 µg/g dw) were found in sludge from two Canadian MWWTPs (Water Technology International Corp., 1998b).

2.3.3.5 Soil

There are essentially no data available for APE concentrations in Canadian soils. A reference sample collected during a sludge addition study was found to have concentrations below detection limits (<0.03 µg/g dw) (Water Technology International Corp., 1998a). Bennie (1998b) reported a concentration of NP of 2.72 mg/kg and traces of NPEs in sludge-amended soil. Following the aerial application of 0.47 L NP/ha in a pesticide formulation to 40 ha of forest, concentrations in all soil samples collected for up to 62 days were below the limit of detection (0.1 ppm) (Sundaram et al., 1980).

The concentrations of NP, NP1EO and NP2EO in sludge-amended soil in Switzerland immediately after the application of sewage sludge were 4.7, 1.1 and 0.1 mg/kg, respectively. After 320 days, residual concentrations were 0.5, 0.1 and 0.01 mg/kg for NP, NP1EO and NP2EO, respectively (Marcomini et al., 1989).

2.3.3.6 Biota

There are no published data on NP levels in fish or other aquatic biota in Canada; however, NP levels in a limited number of specimens have been determined in an unpublished study (Bennie, 1998a). Two carp (Cyprinus carpio) samples from Hamilton Harbour had non-detectable levels (<0.02 µg/g) of NP, while a third carp sample contained 0.02 µg/g (whole tissue wet weight). Nine rainbow trout (Oncorhynchus mykiss) taken from western Lake Ontario had non-detectable levels (<0.02 µg/g) of NP, but a tenth fish contained 0.043 µg/g (whole tissue wet weight). Liver and fat samples from five different beluga whales (Delphinapterus leucas) collected on the St. Lawrence River shore were analysed for NP. NP levels in all five liver samples were below detection (0.02 µg/g), but three of the five fat samples contained detectable NP concentrations (0.02-0.12 µg/g wet weight) (Bennie, 1998a). Concentrations of NP, NP1EO and NP2EO were as high as 1.6, 7.0 and 3.0 mg/kg, respectively, in four composite samples of fish (chub [Leuciscus (Squalus) cephalus], barbel [Barbus barbus] and rainbow trout) collected from the Glatt River in Switzerland; concentrations in a single sample of wild duck (Anas boscas) ranged from not detected to 1.2, 2.1 and 0.35 mg/kg dw for NP, NP1EO and NP2EO, respectively (Ahel et al., 1993). In five samples of mussels, Mytilus edulis, exposed in situ for 7 weeks to water from a surfactant manufacturer's wastewater outlet in Sweden, levels of NP, NP1EO, NP2EO and NP3EO were up to 0.40, 0.28, 0.13 and 0.04 mg/kg fresh weight, respectively (Wahlberg et al., 1990).

2.3.3.7 Drinking water

Clark et al. (1992) measured the concentration of NP1-7EO in a single drinking water sample in the United States. The concentration of each NPE measured was between 0.062 and 0.129 µg/L (Clark et al., 1992). Three drinking water samples in Italy contained concentrations of NPEs of 0.061-0.120 µg/L (Crescenzi et al., 1995). Concentrations of NP and NP1-3EO in an unspecified number of tap water samples collected in Barcelona, Spain, ranged from below the limit of detection (not specified) to 0.25 µg/L (Guardiola et al., 1991).

2.3.3.8 Food

The concentration of NP in one sample of fresh pork loin purchased from a local market in Toronto, which was cooked, was 0.53 mg/kg, while the level in one sample of cured pork was 0.34 mg/kg (Ramarathnam et al., 1993).

2.3.3.9 Consumer products

NP and NPEs are components in a wide range of consumer products, including cosmetics, cleaners and paints. The NP/NPE concentrations and EO chain lengths present in a number of these products are summarized in Table 3. Cosmetics notifications submitted by manufacturers to Health Canada (McIntyre, 1996) indicate that NP and NPEs are used in a large number of cosmetic products applied to the skin and hair. While the concentrations reported are approximate, some cosmetic products are reported to contain in excess of 30% by weight (Table 3). The World Wildlife Fund Canada (1997) analysed 31 common brand-name soaps and cleaning products available on the Canadian market (laundry detergents, stain removers, dishwashing soap, surface cleaners, bathroom cleaners, fabric softeners, oven cleaners and shampoo/hand soap) for NPEs with a chain length between 4 and 10 inclusive. Seven of these products contained detectable concentrations of NPEs (the limit of detection was 0.2% w/w); two were liquid laundry detergents, two were stain removers and three were surface cleaners (Table 3). NPEs are typically found in paints at concentrations between 0.6% and 3% (WHO, 1998). NPEs are also the active ingredient in vaginal spermicides (Talmage, 1994).

2.3.4 Levels in human tissues and fluids

In a study conducted in Switzerland, adipose samples from 25 human cadavers thought to be non-occupationally exposed (4 collected in 1983-84 and 21 in 1994) were analysed for NP and NP1EO and NP2EO. The tissue concentrations of NP (which ranged from 19.8 to 84.4 ng/g lipids) and NP1EO and NP2EO (which were below the limit of detection [5 ng/g lipids] in all samples) were all within the range of background contamination found in the analytical "blank" samples. Müller (1997) indicated that all reasonable precautions had been taken to minimize contamination during analysis. NPEs with a chain length of 7-10 were identified but not quantified in urine samples from three non-occupationally exposed Canadian human subjects (Charuk et al., 1998).

2.4 Effects characterization

2.4.1 Ecotoxicology

Although the studies reported in the literature have used many species, different test methods and different chemicals, there is a consistent pattern in the toxicity reported for NP and NPEs. NP is acutely toxic to fish (LC₅₀ values 17-1400 µg/L), invertebrates (LC₅₀ values 20-3000 µg/L) and algae (LC₅₀ values 27-2500 µg/L). Chronic toxicity values (No-Observed-Effect Concentrations, or NOECs) are as low as 6 µg/L in fish and 3.9 µg/L in invertebrates. There is an increase in the toxicity of NPEs with decreasing EO chain length. NPECs are less toxic than the corresponding NPEs and have acute toxicities similar to those of NPEs with 6-9 EO units. NP and NPEs have bee n reported to cause a number of estrogenic responses in a variety of aquatic organisms. The relative estrogenic potency determined in several different in vitro systems is in the order NP > NP1EO = NP2EO > NP1EC = NP2EC > NP9EO. APEs bind to the estrogen receptor, resulting in the expression of several responses in both in vitro and in vivo systems, including the induction of vitellogenin. The threshold for vitellogenin induction in fish is 10 µg/L for NP. The estrogenic responses appear to be at least additive and should, therefore, be considered as a group. APEs also affect the growth of testes, alter normal steroid metabolism, disrupt smoltification and cause intersex (ova-testes) in fish.

2.4.1.1 Toxicity via atmospheric exposure

No data were identified on the toxicity of NP and its ethoxylates, or of other APs and their ethoxylates, to organisms via atmospheric exposure in Canada.

2.4.1.2 Toxicity via aquatic exposure

Most of the data in the literature have examined the effects of NP, although there are some data on the toxicity of NPEs and NPECs to freshwater organisms. There are relatively few toxicity data for marine organisms.

To assist in the assessment and interpretation of the data, a level of confidence (I-III) was associated with each of the studies based on the methodologies used and reported, the availability of supporting information (e.g., measured concentrations, water quality, etc.) and the availability of the original reports. Emphasis has been placed on studies that used individual compounds rather than mixtures or commercial preparations. This review of toxicity is focused on NP and its associated polyethoxylates (NPEs) and carboxylates (NPECs). The review presented by Talmage (1994) included many commercial products and mixtures of a wider variety of APEs.

Figure 5 Acute toxicity of NP and NP9EO to various fish species

Early studies were focused on commercial products containing a variety of APEs. The majority of recent studies in the literature have focused on NP, OP or their respective polyethoxylates with 9 or 10 EO units (e.g., NP9EO, OP10EO). Although there are numerous studies on APEs in the literature, few have specifically examined the relationship between structure (EO chain length) and toxicity. In general, the toxicity of APEs in most organisms increases as the length of the EO chain decreases.

NPE toxicity increases as EO chain lengths decrease. During biodegradation, a reduction of NPE chain length is observed; however, this process simultaneously reduces total APE concentrations and thereby lowers toxicity. In a die-away study, Yoshimura (1986) also observed a net reduction in toxicity despite the degradation of NPEs to more toxic, lower-molecular-weight constituents (e.g., NP1EO). It is, therefore, important to measure the specific chemical composition of the matrix of interest to be able to adequately evaluate the potential toxicity of APEs.

2.4.1.2.1 Nonylphenol

The 96-hour LC₅₀ for NP has been determined for at least 18 different species of fish, with reported values ranging from 17 to 1400 µg/L, although most of the values range from 100 to 300 µg/L (Figure 5). The 96-hour LC₅₀s for fathead minnow (Pimephales promelas), reported in several validated studies, ranged from 128 to 300 µg/L (Holcombe et al., 1984; Ward and Boeri, 1991b; Brooke, 1993; Naylor, 1995; Weeks et al., 1996) (Figure 6). Reported 96-hour LC₅₀s for rainbow trout were similar, 190-920 µg/L (Brooke, 1993; Dwyer et al., 1995; Naylor, 1995). Ceriodaphnia dubia had a 96-hour EC₅₀ of 69 µg/L and a 7-day NOEC of 134 µg/L, based on reproduction, for NP (Weeks et al., 1996). The NOEC (growth) in mysid shrimps, Mysidopsis bahia, was found to be 3.9 µg/L (Ward and Boeri, 1991c). Forty-eight- hour LC₅₀s for Daphnia magna ranged from 93 to 470 µg/L (Ankley et al., 1990; Brooke, 1993; Comber et al., 1993; Naylor, 1995) (Figure 6). Twenty-one-day NOECs of 24 µg/L (Comber et al., 1993) and 116 µg/L (Brooke, 1993) have been reported for Daphnia magna, based on reproduction. Additionally, 96-hour LC₅₀s of 20 µg/L (Brooke, 1993) and 170 µg/L (England and Bussard, 1994) have been reported for the freshwater amphipod, Hyalella azteca. LC₅₀s for NP in dragonflies, Ophiogomphus sp., and snails, Physella virgata, were also in the same range, >768 µg/L and 774 µg/L, respectively (Brooke, 1993).

Figure 6 Relative toxicity of NP, NPEs and NPECs in fathead minnow (Pimephales promelas, 96-hour LC₅₀), killifish (Oryzias latipes, 48-hour LC₅₀), mysid shrimp (Mysidopsis bahia, 96-hour LC₅₀), Daphnia magna (48-hour LC₅₀) and Ceriodaphnia dubia (7-day LC₅₀)

NPE toxicity to algae is similar to results observed for other organisms. Ninety-six-hour EC₅₀s based on growth for NP in Selenastrum capricornutum (410 µg/L) and for the marine alga, Skeletonema costatum (27 µg/L), were reported by Ward and Boeri (1990a,b). A 72-hour EC10 value of 500 µg/L in Scenedesmus subspicatus was reported by Hüls, AG (1996).

Weinberger and Rea (1981, 1982) calculated a 24-hour LC₅₀ for Chlorella pyrenoidosa of 1500 µg NP/L, while they saw effects on growth at concentrations as low as 25 µg/L. The photosynthetic activity (Moody et al., 1983) and the ultrastructure of cell membranes (Weinberger and Rea, 1981) of Chlamydomonas reinhardii were inhibited by 500 µg NP/L. Prasad (1989) observed inhibition of frond production after 2 days of exposure to >500 µg/L in pond weed (Lemna minor). Reduced growth and photosynthetic activity were reported at NP concentrations between 125 and 500 µg/L. Similar effects also were reported for Salvinia molesta exposed to NP (Prasad, 1989).

Studies conducted in sediment-water exposure systems with NP determined a 14-day LC₅₀ for the midge, Chironomus tentans, of 75 µg/L based on interstitial water concentrations and a NOEC (growth and survival) of 20 mg/kg based on sediment concentrations (England and Bussard, 1993). Kahl et al. (1997) reported a NOEC of 42 µg/L and a Lowest-Observed-Effect Concentration (LOEC) of 91 µg/L for life cycle tests with the midge, which evaluated survival, growth, emergence and fecundity. Tadpoles, Rana catesbiana, had a 30-day LC₅₀ of 260 mg/kg and a NOEC of 155 mg/kg in sediment (Ward and Boeri, 1992; Weeks et al., 1996). The LC₅₀ was similar after 10, 20 and 30 days of exposure.

Studies have been performed to examine effects of NP on bacteria. An EC10 for oxygen consumption by the bacterium, Pseudomonas putida, was >10 000 µg NP/L (Knie et al., 1983), while an EC₅₀ for Photobacterium phosphoreum (Microtox) occurred at 60 600 µg NP/L (Dorn et al., 1993).

Results from large littoral enclosure studies were conducted in triplicate with NP at nominal concentrations between 3 and 300 µg/L (Liber et al., 1998a,b; O'Halloran et al., 1998; Schmude et al., 1999). There were no effects on zooplankton observed in the enclosures treated with NP at the lowest exposure concentration, 5 µg/L (O'Halloran et al., 1998). Periphyton growth was not affected at any treatment level (O'Halloran et al., 1998). Snails and clams (Pisidium) were the most affected macroinvertebrates in exposed enclosures, with significantly reduced abundances (up to 100%) in the 243 µg/L treatment during the 2-year duration of the study (Schmude et al., 1998). Oligochaetes and chironomid midges also were reduced in the 243 µg/L treatment but recovered within 6 weeks. Only minor effects on snails and oligochaetes were observed in the 76 µg/L treatment, and no effects were seen on macroinvertebrates in the 5 or 23 µg/ L treatment (Schmude et al., 1999). Juvenile bluegill sunfish (Lepomis macrochirus) added to the enclosure had reduced survival in the highest treatment, 243 µg/L.

2.4.1.2.2 Nonylphenol ethoxylate, diethoxylate and polyethoxylates

The toxicity of NPEs decreases with increasing EO chain length in a wide variety of species, including fish, invertebrates, algae and soil microorganisms (Figure 6). The LC₅₀s and EC₅₀s for NP9EO are much higher than those reported for NP in fish, invertebrates and algae. LC₅₀ values ranging from 2500 to 12 500 µg/L have been reported for the higher EO chains in fathead minnows and rainbow trout (Marchetti, 1965; Calamari and Marchetti, 1973; Unilever Research Laboratories, 1977; Dorn et al., 1993).

In invertebrates, the 48-hour LC₅₀ of NP9EO in Daphnia magna was reported as 14 000 µg/L by Dorn et al. (1993). The 48-hour LC₅₀ for the marine amphipod, Mysidopsis bahia, was 900-2000 µg/L for NP9EO (Hall et al., 1989; Patoczka and Pulliam, 1990), 2570 µg/L for NP15EO and >100 000 for NP40EO and NP50EO (Hall et al., 1989). The 96-hour LC₅₀ for NP10EO was determined in a number of crustaceans and clams and was generally >10 000 µg/L (Swedmark et al., 1971, 1976). Low toxicity (19 300->100 000 µg/L) relative to NP was observed for NP12EO in shrimp, crabs and molluscs (Portmann and Wilson, 1971; Van Emden et al., 1974; Waldock and Thain, 1991). Eggs and larvae of the mussel, Mytilus edulis, were more sensitive than adults. Collyard et al. (1994) also demonstrated a 2- to 3-fold decrease in toxicity in the amphipod, Hyalella azteca, with age of the organisms exposed to NPE.

Twelve species of marine algae were tested using branched NPEs (Igepal). All showed total or some growth inhibition at concentrations above 100 000 µg/L (Ukeles, 1965). In the algae, the reported 96-hour EC₅₀ of NP9EO for Selenastrum capricornutum ranged from 12 000 to 50 000 µg/L (Lewis, 1986; Dorn et al., 1993).

A few studies have also shown effects of NPEs on bacteria, although generally bacteria appear to be less sensitive than other biota to APs and APEs. Photobacterium phosphoreum toxicity (EC₅₀) decreased with increasing EO chain length for NPEs (Ribosa et al., 1993). Cserhati et al. (1991) tested several species of soil bacteria in agar cultures and found that at high concentrations, NPEs inhibited growth, while at low concentrations, NPEs stimulated the growth of some bacteria.

2.4.1.2.3 Nonylphenoxyacetic acid and nonylphenoxyethoxyacetic acid

Yoshimura (1986) reported 48-hour LC₅₀s in killifish (Oryzias latipes) for NP1EC and NP2EC of 9600 and 8900 µg/L, respectively. These values are slightly lower than the values reported for NP8.4EO/NP8.9EO (11 200-14 000 µg/L) but much higher than that reported for NP (1400 µg/L) (Yoshimura, 1986). Similar results for NP1EC were observed in fathead minnows. In another study, LC₅₀s for NP1EC (2000 µg/L) and NP9EO (6600 µg/L) were reported in fathead minnows (Williams et al., 1996). A similar trend was seen in Daphnia magna, Mysidopsis bahia and Ceriodaphnia dubia for NP1EC in recent studies by the Chemical Manufacturers Association (Naylor et al., 1997). Maki et al. (1998) measured 48-hour LC₅₀s in Daphnia magna for NP2EO (115-198 µg/L) and NP2EC (990 µg/L). These data suggest that the NPECs are much less toxic than the corresponding NPEs.

2.4.1.3 Toxicity to terrestrial plants and animals

There are only limited data available on the toxicity of NP to plants, and there are no data in the published literature on other APs and APEs. The concentration causing 50% growth reduction in cell suspension cultures of 14 species ranged from 0.05 mM (11 mg/L) to more than 1.00 mM (220 mg/L) (Bokern and Harms, 1997). NP was also toxic to plant roots. Lupinus hartwegii showed a 50% growth reduction at 0.1 mM (22 mg/L) (Bokern et al., 1998). The growth of Lupinus polyphyllus root cultures also was inhibited but did not reach 50% growth reduction at 1 mM (220 mg/L) NP. The uptake of NP from soil was slow, and NP was quickly mineralized by soil microorganisms. NP accumulated in several species of plants and was metabolized to hydroxylated and conjugated derivatives.

The earthworm, Apporectodea calignosa, tested by Krogh et al. (1996) and reported by the U.K. Environment Agency (1998), had a 21-day EC10 (reproduction) of 3.4 µg/g in soil for NP. There are no toxicity data available for soil-dwelling organisms for the other NPE metabolites.

NP accumulation in several species of plants was minimal, and NP was metabolized to hydroxylated and conjugated derivatives. Terrestrial plants appear to be relatively insensitive to the effects of NP and NPEs (Bokern et al., 1998).

2.4.1.4 Effects of alkylphenols and alkylphenol polyethoxylates on endocrine function

APs and APEs have been reported to cause a number of estrogenic responses in a variety of aquatic organisms. These responses occur at concentrations similar to those at which chronic effects are reported in aquatic biota. Experiments in several different in vitro systems have indicated similar relative potencies among NPEs. NP was found to be ~100 000 times less potent than estradiol (E₂).

NP2EO and NP1EC were only slightly less potent than NP in inducing vitellogenin in trout hepatocytes. Addition of EO units to NPEs reduced the potency, such that NP9EO was an order of magnitude less potent in vitro (Jobling and Sumpter, 1993) (Table 6). APEs bind to the estrogen receptor, resulting in the expression of several responses, including the induction of vitellogenin in both in vitro and in vivo systems.

One of the functions of endogenous estrogens in fish is to stimulate the liver to produce vitellogenin, a large phospholipoprotein (Chen, 1983). It is released into the bloodstream and sequestered by developing oocytes for production of egg yolk (Wallace, 1985; Tyler et al., 1988a,b; Tyler, 1991). In maturing female fish, vitellogenin is a major constituent of blood proteins; in male fish, it is not normally present in appreciable amounts. If male fish are exposed to estrogens, however, vitellogenin can be produced at similar levels to those found in maturing females. Although the implications of the induction of vitellogenin for the reproductive function of fish are not fully understood, it has been used as a very sensitive indicator of exposure of fish to exogenous estrogens. Jobling et al. (1996) determined the potency of NP2EO and NP1EC to be only slightly less than that of NP in rainbow trout. Jobling et al. (1996) also demonstrated that NP2EO and NP1EC had similar potency for in vivo induction of vitellogenin in rainbow trout. The threshold for vitellogenin induction in fish is 10 µg/L for NP (Jobling et al., 1996). The induction of mRNA in rainbow trout was recently reported at 1 µg NP/L (Fent et al., 1999). The estrogenic responses appear to be at least additive (Soto et al., 1994; Sumpter and Jobling, 1995) and should, therefore, be considered as a group. The threshold for expression of intersex (ova-testes) in killifish was <50 µg NP/L (Gray and Metcalfe, 1997). APEs also affect the growth of testes in fish, alter normal steroid metabolism and disrupt smoltification (Madsen et al., 1997; Ashfield et al., 1998; Fairchild et al., 1999). There is currently considerable debate resulting from the inconsistency in relative potency reported for estradiol receptor binding, yeast estrogen screen (YES) assay and vitellogenin induction in trout hepatocytes. Additional research is required to fully understand the potential estrogenic effects of APs and APEs on the environment. The significance of estrogenic responses to the individual or population is also not known. A thorough discussion of current research into effects of APEs on endocrine function was presented in the environmental supporting documentation (Servos et al., 2000; Servos, 1999b).

Table 6 Summary of relative toxicity and relative estrogenicity based on endocrine disrupting effects

Enlarge table

2.4.2 Bioaccumulation in the environment

Bioaccumulation of APs and APEs has been studied in a number of algae, plants, invertebrates and fish species, both in the laboratory and under field conditions. Bioconcentration factors (BCFs) of APs and APEs determined in the laboratory and bioaccumulation factors (BAFs) measured in the field are similar and represent a low to moderate tendency to bioaccumulate (Table 7). This is expected based on a measured log K_ow of 4.48 (Ahel and Giger, 1993b) for NP. OECD (1997) predicted a theoretical BCF of 1280 based on K_ow.

Metabolic rate and excretion could alter the actual value considerably from the theoretical value, resulting in lower BAFs measured in both laboratory and field studies. The available literature suggests that the ability of NP and NPEs to bioaccumulate in aquatic biota in the environment is low to moderate. BCFs and BAFs in biota, including algae, plants, invertebrates and fish, range from 0.9 to 4120 for NP. There are relatively few data available for NPEs, but, based on their structure, they are not expected to bioaccumulate (Table 7).

2.4.3 Effects in experimental mammals and humans

Identified information on effects of NP and NPEs in laboratory animals and humans is summarized in this section. As noted in Section 1.0and in Appendix A, relevant studies were identified primarily from several recent reviews (Talmage, 1994; U.K. Environment Agency, 1998; WHO, 1998), as well as searches of on-line databases. For endpoints such as acute toxicity and genotoxicity (i.e., those that were not critical to the limited objectives of this screening assessment), the information presented in this section was derived principally from the above reviews; in contrast, the effect levels for those effects resulting from repeated exposures to NP/NPEs that were considered potentially relevant to development of the margins of exposure were confirmed from the primary sources. Weight of evidence for and adversity of effects were generally not considered in this screening exercise.

In view of the limited objective of this screening assessment, presentation in the sections below is limited to an overview of the nature of the identified data on toxicity of NP/NPEs, with emphasis on the magnitude of the effect levels from repeated-dose studies potentially relevant to development of the margins of exposure, rather than full descriptions of protocol and results of available studies. More detailed information is presented in tabular form in the health-related supporting documentation. Information included herein is also restricted principally to that considered to be directly relevant to determination of the margin of exposure.

Table 7 Bioaccumulation of NP and NPEs in aquatic organisms

Enlarge table

For various endpoints, the toxicity of this class of substances generally decreases as the number of EO units increases (Talmage, 1994). In the following sections, information on NP, NP4EO, and NP9EO for which there are a considerable amount of data, is presented separately, followed by information on the remaining NPEs.

2.4.3.1 Effects in laboratory animals and in vitro

2.4.3.1.1 Nonylphenol

Identified data for NP include studies of acute toxicity, several repeated-dose toxicity studies in rats by the oral route, several genotoxicity tests in bacteria and mammalian cells, a multigeneration reproductive study in rats and several in vitro and in vivo assays of estrogenic activity.

The acute toxicity of NP is relatively low, with oral LD₅₀ values in the rat between 580 and 1620 mg/kg-bw. Dermal LD₅₀ values in rabbits were >2000 mg/kg-bw. NP was moderately to severely irritating to rabbit skin and eye (U.S. EPA, 1992a,b,c; WHO, 1998).

In short-term and subchronic studies with rats, toxic effects reported following oral exposure to NP included histopathological changes in the liver and kidney. The Lowest-Observed-Effect Level (LOEL) for NP in male rats exposed for 28 days was 25 mg/kg-bw per day, based on increased relative liver weight; in the same 28-day study, a No-Observed-Effect Level (NOEL) of 400 mg/kg-bw per day (highest dose tested) was reported for females (Richards, 1989). In a 90-day study in rats exposed by the oral route, absolute ovary weight and mean body weight were reduced in females and relative kidney weight was increased and mean body weight was reduced in males at 2000 ppm (approximately 129-149 mg/kg-bw per day) (Cunny et al., 1997).

No chronic toxicity studies with NP were identified, with the exception of the multigeneration study described below.

Genotoxicity data identified were restricted to a small number of in vitro studies. NP was consistently negative in bacterial tests of mutagenicity (WHO, 1998), although it induced DNA damage in human sperm, lymphocytes and MCF-7 breast cancer cells exposed in vitro (Banerjee and Roy, 1996; Anderson et al., 1997).

In a multigeneration study in which rats were exposed to NP in the diet, the LOEL was 200 ppm in diet (equivalent to a mean dose of approximately 12-18 mg/kg-bw per day in males, 16-21 mg/kg-bw per day in non-lactating females or 27-30 mg/kg-bw per day in lactating females), based on an increase in renal medullary tubular dilation and cyst formation in males in all generations (F0-F3) and in F3 females. There were also increases in gestation length and in percent abnormal sperm morphology observed in the F2 generation at this dietary level, as well as at the 650 ppm and at 2000 ppm, but these were probably not treatment-related. In both cases, the increase was small, not clearly dose-related, and within the range of control values from other generations and from historical controls. As well, these effects were not observed in other generations and the F2 control values were unusually low. No developmental effects were reported at any dietary level; however, a range of effects on endocrine-regulated endpoints, including delayed vaginal opening, was observed at 650 and 2000 ppm (NTP, 1997; Chapin et al., 1999).

In reproductive toxicity studies, histological changes in the seminiferous vesicles of the testes of rats were observed following oral exposure to 100 mg NP/kg-bw per day for 10 days (de Jager et al., 1999a,b), though this was accompanied by compound-related mortality at doses that did not cause deaths in several other studies. Reductions in relative testis, epididymis, seminal vesicle and prostate weights were reported in rat pups exposed to 0.8 mg NP/kg-bw per day intraperitoneally in the first 15 days after birth (Lee, 1998); however, this information is not considered directly relevant to the margin of exposure, owing to the lesser relevance of this route of administration.

In a number of in vivo and in vitro studies, NP has been weakly estrogenic. NP increased uterine weight in immature or ovariectomized rats and in mice following oral administration of 50 mg/kg-bw per day and above and following subcutaneous and intraperitoneal administration (Lee and Lee, 1996; Shelby et al., 1996; CMA, 1997; Coldham et al., 1997; Laws and Carey, 1997; Odum et al., 1997). Several other effects indicative of estrogenic activity have been observed in rats following the subcutaneous administration of NP in vivo, including endometrial proliferative response (Soto et al., 1991; Cotroneo et al., 1997) and stimulation of uterine vascular permeability (Milligan et al., 1998). Colerangle and Roy (1996) reported an increase in cell proliferation in the mammary gland of rats exposed to 0.01 mg NP/day by subcutaneous minipump; however, this effect was not reproducible in two subsequent studies (Odum et al., 1999a,b). NP was 1000-100 000 times less potent than estradiol in stimulating estrogenic activity (Lee and Lee, 1996; Milligan et al., 1998). In in vitro studies, NP activated the estrogen receptor with a potency 5000-7000 times less than that of 17ß-estradiol (Routledge and Sumpter, 1996; Gaido et al., 1997; Odum et al., 1997). In MCF-7 human breast cancer cells, cell proliferation was stimulated by NP at concentrations between 0.1 and 10 µM (22 and 2203 mg/L) (White et al., 1994; Villalobos et al., 1995; Blom et al ., 1998).

2.4.3.1.2 Nonylphenol-4-polyethoxylate

Identified data for NP4EO (Nonoxynol-4) include studies of acute toxicity, repeated-dose toxicity studies in rats and dogs by the oral route, several genotoxicity tests in bacteria and mammalian cells in vitro and in mice in vivo and two in vivo assays of estrogenic activity in rats.

The acute toxicity of NP4EO is low, with oral LD₅₀ values in the rat between 4290 and 7400 mg/kg-bw and an oral LD₅₀ of 5000 mg/kg-bw in guinea pigs. The dermal LD₅₀ in rabbits is greater than 2000 mg/kg-bw. NP4EO was non-irritating to severely irritating to rabbit skin following exposure to undiluted compound. Irritation of the eye was reported to be minimal to severe (corrosive) following exposure of rabbits to undiluted NP4EO and slight following exposure to 10% diluted compound (Talmage, 1994; WHO, 1998).

In subchronic (90-day) studies with rats and dogs, toxic effects observed following oral exposure to NP4EO included an increase in liver to body weight ratio and a decrease in body weight gain in the first 4 weeks only. The LOEL was 200 mg/kg-bw per day, with a NOEL of 40 mg/kg-bw per day (Smyth and Calandra, 1969).

In a 2-year oral chronic study with rats, a LOEL of 200 mg/kg-bw per day was reported based on reduced weight gain, which the authors concluded was due to decreased food consumption. In a similar study with dogs, the LOEL was 200 mg/kg-bw per day based on an increase in alkaline phosphatase activity in serum and an increase in relative liver weight. In both cases, the NOEL was 40 mg/kg-bw per day (Smyth and Calandra, 1969).

No evidence of genotoxicity was reported in tests of reverse mutation at the histidine locus in bacteria or in unscheduled DNA repair studies in rat primary hepatocytes with NP4EO. NP4EO did not induce micronuclei in the bone marrow cells of mice following intraperitoneal injection in one study (WHO, 1998).

No evidence of estrogenic activity was observed in rats in vivo as evidenced by a lack of the stimulation of uterine growth following oral exposure of ovariectomized females to NP4EO at doses up to 1000 mg/kg-bw per day for 4 consecutive days in two studies (Berke and Mitchell, 1995; Williams et al., 1996).

2.4.3.1.3 Nonylphenol-9-polyethoxylate

Identified data for NP9EO (Nonoxynol-9) include studies of acute toxicity, repeated-dose toxicity studies in rats and dogs by the oral route and in rats by the intraperitoneal and intravaginal routes, an immunotoxicity test in mice by the intraperitoneal route, several genotoxicity tests in bacteria and mammalian cells in vitro and in rats and mice in vivo, reproductive and developmental studies in rats following oral, intrauterine, intravaginal and dermal exposure and an in vivo assay of estrogenic activity.

The acute toxicity of NP9EO is relatively low, with oral LD₅₀ values between 1410 and 5600 mg/kg-bw in the rat and between 620 and 4400 mg/kg-bw in rabbits, mice and guinea pigs. Dermal LD₅₀ values in rabbits were =2830 mg/kg-bw. NP9EO was reported to be minimally to severely irritating to rabbit skin and moderately to severely irritating to the rabbit eye (Smyth and Calandra, 1969; WHO, 1998).

Oral and dermal LD₅₀ values for NP9.5EO were greater than 3000 mg/kg-bw in the rat and rabbit. The compound was slightly irritating to rabbit skin and minimally to severely irritating to the rabbit eye (WHO, 1998).

In subchronic (90-day) studies with rats and dogs, toxic effects reported following oral exposure to NP9EO included reduced polysaccharide in the liver, increased relative liver weight and decreased weight gain, which may have been related to a decrease in food intake (Smyth and Calandra, 1969). The LOEL for NP9EO was 50 mg/kg-bw per day in the rat based on reduced polysaccharide in the liver. Effects on the liver were observed following the administration of 50 mg/kg-bw per day to female rats intraperitoneally for 5 days or intravaginally for 5-20 days (Chvapil et al., 1982a). When NP9EO was administered intravaginally, effects were also seen in the kidney. In a study with mice, NP9EO did not affect thymus-dependent humoral immunity or leukocyte counts when administered intraperitoneally for 24 days (Caren and Brunmeier, 1987).

No evidence of carcinogenicity was reported in 2-year chronic oral toxicity studies of NP9EO with rats and dogs. The only effect reported was an increase in relative liver weight in dogs at 88 mg/kg-bw per day (Smyth and Calandra, 1969).

In studies of genotoxicity, NP9EO did not induce mutation in bacteria or mammalian cells, although it did increase cell transformation in mammalian cells in one of three studies (Long et al., 1982). In single studies, intraperitoneal exposure to NP9EO did not induce cell proliferation in peritoneal cells in rats or abnormalities in germ cells in mice (Buttar et al., 1986; Jinxi et al., 1992).

In rats exposed orally to NP9EO on gestational days 6 through 15, litter size was decreased and pre-implantation loss and incidence of skeletal tissue deformities were increased, but only at maternally toxic doses (i.e., 250 and 500 mg/kg-bw per day, based on decreased maternal weight gain) (Meyer et al., 1988).

There have been several investigations of effects following intravaginal and intrauterine administration, presumably because of NP9EO's common use as an active ingredient in spermicidal formulations. Irritation and inflammation of the vaginal epithelium were observed in rats, rabbits and monkeys following intravaginal exposure to NP9EO (Talmage, 1994; Patton et al., 1999). Effects including reduced number of pregnancies, reduced number of viable embryos and live fetuses and increased resorptions were reported in the absence of maternal toxicity when 0.50 mg was injected directly into the uterus of pregnant rats (Stolzenberg et al., 1976). The number of live fetuses was significantly reduced when pregnant rats were exposed to 25 mg NP9EO/kg-bw per day intravaginally on gestational day 4, 5, 8 or 9 (Buttar, 1982), and the number of implantations per uterus was significantly reduced following exposure to 50 mg/kg-bw per day on gestational day 3 or 7 (Tryphonas and Buttar, 1982). No teratogenic effects were reported at intravaginal doses up to 25 mg/kg-bw (Buttar, 1982).

No dose-related reproductive or teratogenic effects were reported in rats following dermal exposure to up to 500 mg NP9EO/kg-bw per day administered on gestational days 6-15 (Meyer et al., 1988).

When immature female rats were administered NP9EO for 3 days by gavage, there was no reported effect on uterine weight, indicating a lack of estrogenic activity, at doses up to 1000 mg/kg-bw per day (Williams et al., 1996).

2.4.3.1.4 Other nonylphenol polyethoxylates

Identified data for other NPEs include acute toxicity studies, repeated-dose toxicity studies in rats and dogs by the oral route for NP6EO, NP15EO, NP20EO and NP30E, a small number of genotoxicity studies (mostly in bacteria) for NP5EO, NP10EO, NP12EO and NP20EO, a reproductive and developmental study for NP10EO in mice and for NP30EO in rats by the oral route, and an in vitro estrogenic activity study for NP2EO and NP12EO.

In NPEs with chain lengths up to 40 (excluding 4 and 9), acute oral LD₅₀s in the r at range from 1300 to 15 900 mg/kg-bw; acute dermal LD₅₀s in rabbits were above 1800 mg/kg-bw. Skin irritation in rabbits ranged from non-irritating to severely irritating, with lower chain lengths generally being more irritating. Eye irritation to rabbits was minimal to severe in most studies with NPEs, with NP30EO and NP40EO being non-irritating. NP6EO was non-sensitizing in guinea pigs (Younger Laboratories 1961a,b; Union Carbide, 1992; WHO, 1998).

In subchronic studies with NP6EO and NP15EO, the LOEL was 40 mg/kg-bw per day in rats and 1000 mg/kg-bw per day in dogs following oral exposure, based on an increase in relative liver weights. No effects were noted in rats following oral exposure to NP20EO and NP30EO at doses up to 5000 mg/kg-bw per day or in dogs following oral exposure to NP30EO at doses up to 1000 mg/kg-bw per day. The LOEL in dogs for NP20EO was 40 mg/kg-bw per day based on an increase in the incidence of focal myocardial necrosis or degeneration (Smyth and Calandra, 1969).

In studies of genotoxicity, NP5EO, NP10EO and NP20EO did not induce mutation in bacteria (CIR, 1996; WHO, 1998). In single studies, NP12EO did not induce unscheduled DNA repair in primary rat hepatocytes in vitro or produce micronuclei in bone marrow cells of mice following intraperitoneal injection (WHO, 1998). < /p>

No reprodu ctive or developmental effects were observed following oral exposure during gestation to 600 mg NP10EO/kg-bw per day in mice (Hardin et al., 1987) or up to 1000 mg NP30EO/kg-bw per day in rats (Meyer et al., 1988).

NP2EO stimulated the transcription of the estrogen receptor and cell proliferation in human breast cancer cells in vitro (White et al., 1994) and activated the estrogen receptor in yeast with a potency 500 000 times less than that of estradiol (Routledge and Sumpter, 1996). NP12EO did not demonstrate any estrogenic activity in an estrogen-inducible strain of yeast (Routledge and Sumpter, 1996).

2.4.3.2 Effects in humans

2.4.3.2.1 Nonylphenol

No data were identified on the effects of NP in humans.

2.4.3.2.2 Nonylphenol-4-polyethoxylate

Nonoxynol-4 application (10% in mineral oil) to the skin on the back resulted in faint to moderate erythema in 36 of 111 volunteers. Three of these reactions were classified as allergic contact dermatitis; however, in a 30-minute retest, there was evidence of a mild allergic response in only one of these three subjects (Jordan, 1995).

No further data were identified on the effects of NP4EO in humans.

2.4.3.2.3 Nonylphenol-9-polyethoxylate

Data in humans are limited to studies of effects following exposure to spermicides containing NP9EO (Nonoxynol-9).

In several studies, the use of NP9EO-containing spermicides has been reported to cause vaginal irritation and/or burning and genital ulceration in some females; irritation of the urinary tract has also been reported in some males and females following exposure (Chvapil et al., 1982b; Rekart, 1992; Roddy et al., 1993; Weir et al., 1995; Saborio et al., 1996, Stafford et al., 1998). Allergic contact dermatitis reactions to NP9EO in an antiseptic preparation and a condom were also reported (Dooms-Goossens et al., 1989; Fisher, 1994).

Following the intravaginal application of 150 mg NP9EO for 14 consecutive days by 10 women, the only significant effect reported was a reduction in serum cholesterol; no effects on liver function or hematological parameters were observed (Chvapil et al., 1982b). In another study, 12 women applied 2.5 g of cream containing 5.0% NP9EO intravaginally (yielding an applied dose of 125 mg) for 14 consecutive days. There were no significant differences in levels of proteins, lipids, triglycerides or serum enzymes in blood samples collected before, during (day 8) and after (day 15) exposure (Malyk, 1981).

The association between spermicide use and congenital malformations has been examined in a number of historical cohort and case-control studies (reviewed in Manjuck, 1989; Talmage, 1994; WHO, 1998). While there were statistically significant increases in overall malformations or in trisomies in relation to spermicide use in a small number of the available studies, no increase was observed in the majority of studies (which were generally larger and had better characterization of exposure and/or better adjustment for possible confounders). Further, the relative risks in most of the positive studies were relatively low (i.e., less than 2). Hence, based on these reviews, the weight of evidence for congenital malformations appears to be quite limited, with almost no indication of consistency, specificity or strength of association. There is also little evidence of an exposure-response relationship (although a gradient of exposure was not investigated in most studies, and the exposure characterization was somewhat crude in all of the available studies).

2.4.3.2.4 Other nonylphenol polyethoxylates

Contact dermatitis and contact photosensitivity have been reported in humans following exposure to NP6EO, NP10EO and NP12EO in consumer products (Nethercott and Lawrence, 1984; Meding, 1985; Michel et al., 1994; Wilkinson et al., 1995).

No further data were identified on the effects of NPEs of chain lengths other than four or nine in humans.

2.4.4 Abiotic atmospheric effects

Ozone Depletion Potential, Global Warming Potential and Photochemical Ozone Creation Potential were not calculated for NP/NPEs. NP/NPEs are not expected to readily volatilize into air and are expected to degrade rapidly in the atmosphere.