3,3'-Dichlorobenzidine - PSL1

2.0 Summary of Information Critical to Assessment of "Toxic"

2.1 Identity, Properties, Production and Uses

3,3'-Dichlorobenzidine is a chlorinated primary aromatic amine with the Chemical Abstracts Service registry number 91-94-1 and the molecular formula C₁₂H₁₀N₂C1₂. Synonyms for 3,3' -dichlorobenzidine include 4,4'-diamino-3,3'-dichlorobiphenyl, o,o'-dichlorobenzidine, 3,3'-dichlorobiphenyl-4,4'-diamine and 3,3'-dichloro-4,4'-diamino(1,1-biphenyl). 3,3' -Dichlorobenzidine is usually available as the dihydrochloride salt, 3,3'-dichlorobenzidine dihydrochloride (C₁₂H₁₀N₂C1₂·2HCl). 3,3' -Dichlorobenzidine is a grey to purple crystalline solid (Banerjee et al., 1978) with a melting point of 133°C (Ferber, 1978). It has a relatively low vapour pressure (1.33x10^-3 Pa at 22°C) (Mabey et al., 1982), a low water solubility (4 mg/L at 20°C as the dihydrochloride salt) (Banerjee et al., 1978) and a log n-octanol/water partition coefficient of 3.02 (Callahan et al., 1979).

3,3'-Dichlorobenzidine can be produced by the alkaline reduction of o-chloronitro-benzene and rearrangement of the resulting hydrazo compound (Ferber, 1978).

3,3'-Dichlorobenzidine is not produced in Canada (Environment Canada, 1980; 1991a; 1991b) but has been imported on a regular basis (Statistics Canada, 1990). The reported volumes of importation figures since 1986 are as follows: 1986 (21 tonnes); 1987 (60 tonnes); 1988 (not available); and 1989 (109 tonnes). The destinations in Canada are generally not known, except for the year 1989, in which 80 of the 109 tonnes imported went to Ontario.

In Canada, 3,3'-dichlorobenzidine (and some of its derivatives) are used primarily as intermediates in the manufacture of pigments for printing inks, textiles, paints, plastics and crayons (Environment Canada, 1991a; 1991b). It can also be used as a curing agent in the synthesis of polyurethane elastomers, and in the analytical determination of gold (Ferber, 1978; Budavari, 1989).

2.2 Entry into the Environment

3,3'-Dichlorobenzidine can enter the Canadian environment from any stage in the production, storage, transport, use and disposal of 3,3'-dichlorobenzidine-containing materials in Canada, or possibly by atmospheric and water-borne transport from other countries. While there are no supporting data, the largest releases would likely be through direct emissions from plants that manufacture 3,3'-dichlorobenzidine-containing materials, and from the degradation of 3,3'-dichlorobenzidine-containing pigments. In polymers, the residual 3,3'-dichlorobenzidine content would be generally so low that loss from this source would not constitute a significant route of entry. For those 3,3'-dichlorobenzidine-containing materials disposed by incineration, it is likely that this process would destroy unreacted 3,3'-dichlorobenzidine.

Data on the release of 3,3'-dichlorobenzidine into the Canadian environment were not identified. Total industrial emissions of 3,3'-dichlorobenzidine into the environment in the United States in 1988 were estimated to be 6 tonnes (U.S. EPA, 1990). If it is assumed that 3,3'-dichlorobenzidine use in the United States amounted to 5x103 tonnes in 1988, losses would be estimated as 0.1% of total use. Based on data of Canadian uses in 1989, losses would be estimated as 0.1 tonne.

2.3 Exposure-related Information

2.3.1 Fate

Owing to the relatively short half-life (i.e., less than a few weeks) of 3,3'-dichlorobenzidine in water, soil and air, this substance is not expected to persist in the environment. Photolysis, photooxidation, partitioning to soil, biota and sediment, and slow microbiological degradation are expected to be the main pathways of distribution and transformation of 3,3'-dichlorobenzidine in the environment (Sikka et al., 1978; Callahan et al., 1979; Howard, 1989).

The half-life for volatilization of 3,3' -dichlorobenzidine from surface water (1 m deep, flowing at 1 m/s, with wind velocity of 3 m/s, at 200C) to the atmosphere was estimated to be 72 days (Thomas, 1982). In most natural waters, 3,3'-dichlorobenzidine will be present almost entirely as the free base, based on the pKa's of 4.5 and 3.3 for this substance (Korenman and Nikolaev, 1974). In water, 3,3'-dichlorobenzidine may be degraded by photooxidation (Mill and Mabey, 1985), photolysis (Banerjee et al., 1978) and biodegradation (Sikka et al., 1978). In an aqueous medium, 3,3' -dichlorobenzidine is very rapidly photodegraded by sunlight, producing monochlorobenzidine, benzidine and a number of brightly coloured, water-insoluble compounds (Banerjee et al., 1978); the half-life of 3,3'-dichlorobenzidine in water is less than 10 minutes. The photodegradation of 3,3'-dichlorobenzidine may involve its sequential dechlorination, yielding benzidine.

3,3'-Dichlorobenzidine is fairly resistant to degradation by naturally occurring aquatic microbial communities (Sikka et al., 1978) and by a sewage sludge inoculum (Brown and Laboureur, 1983) over a 4-week period. Half-lives of 4-26 weeks and 16-101 weeks have been estimated for the biodegradation of 3,3'-dichlorobenzidine in surface water and anaerobic ground water, respectively (Syracuse Research Corp., 1989).

3,3'-Dichlorobenzidine is strongly bound to soil and is, therefore, highly immobile. The formation of covalent linkages between 3,3' -dichlorobenzidine and soil humic components may be the predominant fate of this substance in the soil. 3,3'-Dichlorobenzidine quickly became unextractable in soil after application in municipal sewage sludge, with none recovered by organic solvent extraction after 4 months (Demirjian et al., 1987). In soil, 3,3'-dichlorobenzidine is mineralized very slowly under aerobic and anaerobic conditions (Boyd et al., 1984; Chung and Boyd, 1987). The half-life for the aerobic degradation of 3,3'-dichlorobenzidine in soil has been estimated to range from 4 to 26 weeks (Syracuse Research Corp., 1989). 3,3'-Dichlorobenzidine may also be oxidized by metal ions such as iron (III) present in clay. The products formed by the degradation of 3,3'-dichlorobenzidine in soil have not been identified.

Half-lives of 1.5-5 minutes and 0.9-9 hours have been estimated for the photolysis and photooxidation of 3,3'-dichlorobenzidine in air, respectively (Syracuse Research Corp., 1989).

2.3.2 Concentrations

No data on the levels of 3,3'-dichlorobenzidine in drinking water, surface water, ground water, biota, soil or sediment, foodstuffs or products containing pigments or dyes derived from this substance within Canada were identified.

No information on the levels (or occurrence) of 3,3'-dichlorobenzidine in drinking water or foodstuffs in the United States was identified. In the United States, 3,3'-dichlorobenzidine was not found in the ambient air surrounding two dye-stuff production facilities, at the detection limits of 0.1 and 5 mg/m3 (Narang et al., 1982; Riggin et al., 1983, both cited in ATSDR, 1989). 3,3'-Dichlorobenzidine was not detected in samples of biota or sediment within the United States, although it was detected (but not quantitated) in 1% of 1 239 samples of industrial effluent and 0.1% of 863 samples of natural water (Staples et al., 1985).

3,3'-Dichlorobenzidine can accumulate in aquatic biota. Appleton and Sikka (1980) reported a bioconcentration factor of approximately 500 for bluegill sunfish (Lepomis macrochirus), based on a study in which the fish were exposed to 5 or 100 µg/L [¹⁴C]3,3'-dichlorobenzidine; equilibria were achieved within 96-168 hours. Freitag et al. (1985) reported a 3-day bioaccumulation factor in fish (golden orfe, Leuciscus idus melanotus) of 610, a 5-day bioaccumulation factor in activated sludge of 3 100, and a 1-day bioaccumulation factor in algae (Chlorella fusca) of 940.

Owing to the absence of data on fate and concentrations in the Canadian environment, the likely distribution of 3,3'-dichlorobenzidine in the environment was predicated based on the level III fugacity computer model of Mackay and Paterson (1991) applied to southern Ontario using worst-case assumptions. It was assumed that all the 3,3'-dichlorobenzidine imported into Canada in 1989 was for use only in southern Ontario, and that it would be released into the water at a rate of 0.05 mol/h (based on a loss of 0.1 tonne (see Section 2.2)). The results indicated that, at steady-state, the proportion of released 3,3' -dichlorobenzidine found in the environment would be as follows: air (< 0.001%); surface water (99.75%); sediment (0.254%); and soil (< 0.001%). This would result in predicted steady-state concentrations of 7.6x10^-16 µg/m3 in air, 3.44x10^-7 µg/L in water, 1.1xl0^-16 µg dry weight in soil and 3.1x10^-12 µg/g dry weight in sediment. Since this model does not address the possibility of formation of bound residues in sediment, the concentrations in sediment may be underestimated, while the concentrations in water may be overestimated.

2.4 Effects-related Information

2.4.1 Experimental Animals and In Vitro

It is generally believed that the hepatic metabolism of 3,3'-dichlorobenzidine in rodents (i.e., rats) involves its oxidation to highly reactive N-oxygenated intermediates, by the microsomal cytochrome P450d and flavin monooxygenase enzyme complexes (reviewed in Iba, 1990). Although the identity of the 3,3'-dichlorobenzidine derived N-oxygenated intermediates has not been unequivocally identified, they are believed to be responsible for the mutagenic and genotoxic effects (i.e., related to DNA binding) of 3,3'-dichlorobenzidine in bacterial and mammalian systems (Iba, 1990; Iba and Thomas, 1988).

No quantitative information on the metabolism of 3,3'-dichlorobenzidine in humans was identified; however, small amounts (approximately 1-2%) of free and (glucuronide)-conjugated N-hydroxyacetyl-derivatives of 3,3' -dichlorobenzidine were excreted in the urine of volunteers administered (orally) dichlorobenzidine (Belman et al., 1968). 3,3'-Dichloro-N-acetylbenzidine and 3,3'-dichloro-N,N'-diacetylbenzidine, as well as conjugated metabolites (the identities of which were not confirmed), have been detected in the urine of rats administered 3,3'-dichlorobenzidine orally (Tanaka, 1981; Hsu and Sikka, 1982). The covalent binding of 3,3'-dichlorobenzidine (metabolites) to haemoglobin (Birner et al., 1990), hepatic lipids (Iba and Lang, 1988; Iba et al., 1990) and DNA in the intestinal and bladder epithelium or liver (Ghosal and Iba, 1990), has been reported in experimental animals (i.e., rodents) exposed in vivo.

The LD₅₀ for the oral administration of 3,3'-dichlorobenzidine to rats, either as the free base or dihydrochloride salt, Was reported to be 7 070 and 3 820 mg/kg b.w., respectively (Gerarde and Gerarde, 1974). Reported LD₅₀ values for the oral administration of 3,3'-dichlorobenzidine to male and female mice were 676 and 488 mg/kg b.w., respectively (Gaines and Nelson, 1977, cited in U.S. EPA, 1980).

No compound-related histopathological effects were observed in Sprague-Dawley rats (n = 20) administered 3,3'-dichlorobenzidine dihydrochloride (30 mg, by gastric intubation) once every 3 days over a 30-day period, and subsequently maintained for a further 9 months; 5/132 animals administered vehicle alone (negative control) had mammary tumours (carcinomas and fibroadenomas), while 100% of surviving animals (n = 29) administered a single dose of dimethylbenz[a]anthracene (positive control) had mammary tumours (carcinomas and fibroadenomas) (Griswold et al., 1968).

The incidence of mammary carcinomas in ChR-CD rats administered diets containing 0 or 1 000 ppm (0.1% w/w) 3,3'-dichlorobenzidine for up to 488 days was 3/44 and 26/44, and 0/44 and 7/44, in the females and males, respectively (Stula et al., 1975); in male rats, the incidence of Zymbal gland tumours was 0/44 and 8/44, and the incidence of granulocytic leukemias was 2/44 and 9/44, in the control and 3,3'-dichlorobenzidine-exposed animals, respectively.

Stula et al. (1978) administered 3,3'-dichlorobenzidine (100 mg in gelatin capsules) 3 times/week for 6 weeks, and then 5 times/week for as long as 7.1 years (the mean dose for all exposed animals was approximately 10.4 mg/kg b.w./exposure) to a group of six female beagle dogs; six unexposed animals (controls) were sacrificed after 8.3 to 9.0 years on study. The incidence of bladder carcinomas and hepatocellular tumours in 3,3'-dichlorobenzidine-exposed animals (that survived longer than 6.6 years) was 5/5 and 4/5, respectively. No liver or bladder tumours were found in the unexposed (control) animals, although 4/6 were reported to have developed mammary carcinomas (a result attributed to the greater age of the control animals rather than to an effect related to 3,3'-dichlorobenzidine) (Stula et al., 1978).

In the first reported study on the carcinogenicity of 3,3'-dichlorobenzidine in rats, tumours (in the Zymbal and mammary glands, bladder and the hematopoetic system) were observed in 79% (23/29 male and female Rappolovo rats) of animals surviving at the time (not specified) the first tumours were detected following administration of a diet containing 3,3'-dichlorobenzidine (0.5 to 1.0 mL of a 4.4% solution) for 6 days/week for 12 months; no tumours were detected in a group of 130 controls (Pliss, 1959). Subsequently, Pliss (1963) reported that 86% of Rappolovo rats (number not stated) administered a diet containing an amount (not clearly specified) of 3,3'-dichlorobenzidine for 10-13 months had tumours (in bone, mammary glands, bladder, "etc."), while only one of 50 rats in a control group had developed a tumour. The studies by Pliss (1959; 1963) were limited by inadequate protocols (e.g., small group sizes, lack of appropriate controls) and incomplete reporting of data.

The incidence of "hepatomas" in male ICR/JC1 mice administered a diet containing 0 or 0.1% (w/w) (1 000 ppm) 3,3'-dichlorobenzidine for 12 months was 2/21 and 18/18 (p < 0.001), in the control and 3,3'-dichlorobenzidine-exposed groups, respectively (Osanai, 1976). In the first reported study on the carcinogenicity of 3,3'-dichlorobenzidine in mice, in which male and female strain D mice were administered 3,3'-dichlorobenzidine (0.1 mL of 1.1% solution) in the diet for 10 months, amongst 18 animals surviving at the time at which the first tumour appeared (i.e., after 18.5 months), 2 animals had hepatomas and 2 had hemangiomas; however, information on the nature of, and results in, an appropriate group of controls was not presented (Pliss, 1959).

Saffiotti et al. (1967) reported no carcinogenic or histopathological effects in the bladder of male and female Syrian Golden hamsters administered a diet containing 0.1% (w/w) (1 000 ppm) 3,3'-dichlorobenzidine (technical grade; 40% dihydrochloride and 60% free base) for their entire lives, compared to unexposed controls. In a subsequent publication (Sellakumar et al., 1969), however, these authors reported the administration of a diet containing 0.3% (w/w) (3 000 ppm) 3,3'-dichlorobenzidine to male and female hamsters "induced 4 transitional cell bladder carcinomas, some liver-cell and cholangiomatous tumors and diffuse chronic intrahepatic obstructing cholangitis (63%)," although little additional information was provided. No evidence of histopathological effects in either the liver or bladder of male Wistar rats (n = 18) was observed following administration of diets containing 0% or 0.03% (w/w) (300 ppm) 3,3'-dichlorobenzidine for a period of 40 weeks (Tsuda et al., 1977), although the small group sizes and limited period of exposure limit the significance of these results.

Based on the results of a number of studies in bacterial and mammalian cells exposed in vitro and experimental animals exposed in vivo, there is convincing evidence to indicate that 3,3'-dichlorobenzidine is genotoxic. 3,3'-Dichlorobenzidine was mutagenic (i.e., the frequency of his+-revertants was increased) in a variety of strains of Salmonella typhimurium, in the absence (Savard and Josephy, 1986; Lazear et al., 1979; Messerly et al., 1987; Garner et al., 1975) or presence (Savard and Josephy, 1986; Lazear et al., 1979; Messerly et al., 1987; Reid et al, 1984; Anderson and Styles, 1978; Garner et al., 1975) of metabolic activation. Iba et al. (Iba, 1987; Iba and Thomas, 1988) concluded that more than 50% of the mutagenic activity of 3,3'-dichlorobenzidine in Salmonella typhimurium strain TA98 (incubated in the presence of metabolic activation) was dependent upon cytoclirome P450 activity. The frequency of sister chromatid-exchange in human B-lymphoblastoid cells was increased following incubation in vitro with 3,3'-dichlorobenzidine (Shiraishi, 1986); unscheduled DNA synthesis in HeLa S3 cells was increased following exposure in vitro to 3,3'-dichlorobenzidine (Martin et al., 1978). 3,3'-Dichlorobenzidine was reported to "morphologically transform" rat embryo cells (Freeman et al., 1973). The oral administration of 3,3'-dichlorobenzidine (1 000 mg/kg b.w.) to male ICR-SPF mice increased the frequency of bone marrow cells with micronuclei, compared to controls administered vehicle alone (Cihak and Vontorkova, 1987). The oral administration of 3,3'-dichlorobenzidine (500 or 1 000 mg/kg b.w.) to male Alpk:AP rats increased unscheduled DNA synthesis in liver cells, compared to animals administered vehicle alone (Ashby and Mohammed, 1988).

3,3'-Dichlorobenzidine has the potential to adversely affect developing embryos, based on studies in which the incidence of "hyperplastic changes" in kidney explants obtained from embryos exposed in utero to 3,3'-dichlorobenzidine was increased compared to embryos exposed (in utero) to vehicle alone (Shabad et al., 1972), and the incidence of "lymphoid leukemias" in offspring born to female mice administered five (subcutaneous) injections of 3,3'-dichlorobenzidine during the last week of pregnancy was increased (significance unspecified), compared to animals administered vehicle alone (Golub et al., 1975, cited in U.S. EPA, 1988). Data on neurotoxic effects are limited to the observation in the study by Stula et al. (1978) of convulsions in one female beagle dog; at sacrifice, slight neuronal degeneration was noted after histopathological examination. No other information concerning the reproductive, developmental, neurological or immunological effects of 3,3'-dichlorobenzidine in experimental animals was identified.

2.4.2 Humans

Quantitative data on the toxicological effects of 3,3'-dichlorobenzidine in humans were limited to the incidence of tumours or death due to cancer associated with exposure to 3,3'-dichlorobenzidine in three limited studies in which the health of production workers was examined. No increase in the incidence or death due to bladder cancer was reported in studies involving groups of 109 (Maclntyre, 1975), 35 (Gadian, 1975) or 207 (Gerarde and Gerarde, 1974) workers occupationally exposed to 3,3'-dichlorobenzidine.

2.4.3 Ecotoxicology

There are very few data on the acute toxicity of 3,3'-dichlorobenzidine to aquatic organisms. An IC₅₀ value of 0.06 mg/L was reported for bacteria in the Microtox assay (Dutka and Kwan, 1981). Sikka et al. (1978) reported a 48-h LC₁₀₀ value for bluegill sunfish (Lepomis macrochirus Raf.) of 2 mg/L; following exposure to 0.5 mg/L 3,3'-dichlorobenzidine for 96 and 120 h. one half of the test group died. Kaiser (1992) estimated the following 96-h LC₅₀ values: > 3 mg/L for fathead minnow (Pimephales promelas); 3 mg/L for rainbow trout (Oncorhynchus mykiss); and 1.5 mg/L for golden orfe (Leuciscus idus melanotus), based on quantitative structure-activity relationships.

No data were identified for the toxicity of 3,3'-dichlorobenzidine to wild mammals, terrestrial organisms, birds, sediment or soil biota.