Appendices of the Screening Assessment

Aromatic Azo and Benzidine-based Substance Grouping
Certain Azo Direct Dyes and Azo Reactive Dyes

Environment Canada
Health Canada
April 2015

Table of Contents

• Appendix A: Supplementary Data Tables
• Appendix B: Exposure Assessment
• Appendix C: Health Effects Assessment Information
• Appendix D: Azo Direct Dyes with Effects of Concern
• Back to the Screening Assessment

Appendix A: Supplementary Data Tables

Top of Page

Appendix B: Exposure Assessment

Appendix B1. Dermal and Oral Exposure via Contact with Textile Materials

Summary B1: Exposure factors and algorithms for estimating exposure from textile materials

Dermal exposure from textile materials

Exposure estimate =

Dermal exposure was estimated assuming full (100%) body coverage from wearing clothing to account for exposures from multiple pieces of apparel that cover the entire surface area of the body.

Oral exposure from textile materials

Exposure estimate =

Oral exposure was estimated for an infant mouthing a textile object (e.g., blanket, textile toy) on a daily basis.

SA: Total surface area 

For dermal exposure (Health Canada 1998) = 18 200 cm² (adult; personal apparel); = 3020 cm² (infant; baby sleeper)

For oral exposure = 20 cm² (Zeilmaker et al. 2000)

AW: Area weight of textile = 20 mg/cm² (US EPA 2012)

SCF: Skin contact factor = 1

C: Concentration =0.01 (unitless) (BfR 2007)

Based on the default model developed by the “Textiles” Working Group established at the German Federal Institute for Risk Assessment (BfR 2007), assuming that a standard textile garment of 100 g/m² is dyed with 1% active dye ingredient.

M: Migration fraction =0.0005 (BfR 2007)

The migration of azo dyes from textiles varies considerably depending on the type of fibre, the type of dye used, the dye load, dyeing technology and colour intensity and after treatment. The exposure from textiles is partly dictated by the amount of dye that migrates from textile material onto human skin (ETAD 1983) or via mouthing. The “Textiles” Working Group (BfR 2007) uses a peak initial migration of 0.5% to estimate exposure to dyes from newly bought unwashed garments, and the chronic migration rate is assumed to be one tenth of the value measured for the first migration to reflect exposure after initial washes. It is assumed that the sweat migration rate is similar to the salivary migration rate; this is consistent with observations of leaching behaviours of dyes from textiles reported by Zeilmaker et al. (1999). Accordingly, the fraction of dye that migrates from a textile material per wear is assumed to be 0.0005 for both dermal and oral exposure.

UF: Uptake Fraction = 1

In the absence of dermal absorption data, the dermal uptake fraction of these substances was conservatively assumed to be 1. Although the dermal uptake of these substances is likely less than 1, potentially there could be substantial absorption if skin bacteria cleaved azo bonds, thereby releasing aromatic amines, which could then be more readily absorbed

F: Frequency = 1×/day

BW: Body weight = 7.5 kg for infant, 70.9 kg for adult (Health Canada 1998)

P: Probability that a given Azo Direct Dye is present in textile =10%

In the RIVM risk assessment of azo dyes and aromatic amines from garments and footwear (Zeilmaker et al. 1999), the authors derived a chance of 8% for the appearance of carcinogenic azo dyes and aromatic amines in garments based on four European studies. Presumably, there would be a higher prevalence in the use of non-EU22 amines and their dyes, compared to EU22 amines and related dyes, since the former are not prohibited. None of the Azo Direct Dyes used to dye textiles in Canada (i.e., Direct Green 28, Direct Orange 26, Direct Orange 39, Direct Red 81 triethanolamine salt, Direct Red 31, Direct Violet 51, Direct Yellow 12, Direct Yellow 28, Direct Yellow 50, Direct Black 56, CAS RN 28706-21-0, CAS RN 71033-21-1, CAS RN 83221-56-1 and CAS RN 84878-17-1) derive from EU22 amines; the prevalence of these dye is not clear because there is relatively limited product testing and monitoring on non-EU22 amines and associated dyes. Based on data available (Danish EPA 1998; Kawakami 2012; Health Canada 2013), the prevalence of certain non-EU22 amines was found to range from 0% to 23.7% (aniline). Since several dyes can derive from a given aromatic amine, the prevalence of an associated dye would be lower. Given the conservatism used in other parameters in this exposure scenario (e.g. full body coverage), the probability that a given Azo Direct Dye is present in a textile is assumed to be 10% in this Screening Assessment based on professional judgement. This is considered reasonable since the chances of an individual’s outfit containing a given Azo Direct Dye every day are low.

Appendix B2: Dermal Exposure via Contact with Leather Products

Summary B2: Exposure factors and algorithms for estimating exposure from leather products

Dermal exposure from leather products

Exposure estimate =

Prolonged skin contact with articles of leather can result in dermal exposure to dyes used in leather dyeing. Of all the leather products considered, the potential drivers for exposure are presented below: furniture, apparel (e.g., jackets, trousers and gloves), footwear (e.g., shoes and boots) and toys, where it is assumed that prolonged contact with the infant’s palms can occur when playing with the toy. As a conservative approach, exposure is assumed for all products. The exposure estimates presented below are based on conservative assumptions, as well as not taking into account a final application of a polyurethane sealant coating, which would further reduce the consumer’s dermal exposure to the leather dye.

SA:Surface area of skin contact (Health Canada 1998; Therapeutic Guidelines Ltd. 2008)

• Shoes: 1275 cm² (adult feet)
• Boots: 4185 cm² (adult legs and feet)
• Gloves: 455 cm² (adult hands)
• Jackets and coats: 8920 cm² (adult trunk and arms)
• Trousers: 5820 cm² (adult lower body)
• Furniture: 5005 cm² (adult back, buttocks and back of thighs)
• Toys: 92.5 cm² (infant palms)

AW: Area weight of leather = 0.15 g/cm² (Danish EPA 2012)

SCF: Skin contact factor

• Shoes: 1
• Boots: 0.1
• Gloves: 0.1
• Jackets and coats: 0.19
• Trousers: 0.19
• Furniture: 0.1
• Toys: 1

When the entire leather article is in direct contact with the skin, SCF is assumed to be 1. When the leather article is in indirect contact with the skin (e.g., shielding due to interior lining), SCF is assumed to be 0.1, which is a default value used to account for exposure due to diffusion of sweat-extracted dye from the leather material through the shielding fabric onto the skin (Zeilmaker et al. 1999). When a portion of the leather article is in direct contact and the remaining portion is in indirect contact, a weighted SCF is calculated: [(SAdirect × 1) + (SAindirect × 0.1)]/(SAtotal).

C: Concentration = 0.02 (unitless weight fraction) (Øllgaard et al. 1998)

M: Migration fraction = 0.39 over 365 days

The dermal exposure to dyes from leather is partly dictated by the amount of dye that migrates from leather material onto human skin. Zeilmaker et al. (1999) measured the experimental leaching of azo dyes from leather footwear material to be 15% and 39%. The leaching was determined by extracting from 1 g of unwashed material from the upper side of a newly bought leather shoe with 100 mL sweat stimulant (extraction conditions: 16 hours at 37°C while shaking). These extraction conditions are expected to overestimate the migration of dyes from sweat. In estimating exposure to dyes from leather articles, it is assumed that 39% of the dye content leaches over 1 year and is available for dermal exposure.

BW: Body weight = 7.5 kg for infant, 70.9 kg for adult (Health Canada 1998)

Top of Page

Appendix C: Health Effects Assessment Information

Summary C-1: Summary of available data on carcinogenicity and genotoxicity for the postulated azo bond reductive cleavage products of the 18 Azo Direct Dyes

4-Nitroaniline (CAS RN 100-01-6)

The health effects of 4-Nitroaniline were assessed along with Cartain Aromatic Amines in a separate Screening Assessment (Environment Canda and Health Canada 2014b). 4-Nitroaniline did not exhibit carcinogenicity in male and female rats and female mice via 2-year oral (gavage) exposure. It exhibited equivocal evidence of carcinogenicity in male mice (2-year gavage). The incidence of liver hemangiosarcoma and the combined incidence of hemangioma and hemangiosarcoma at all sites were marginally increased at the high dose in male mice. In both cases, a trend test showed significance, but no pairwise comparisons were significant. All in vivo genotoxicity assays were negative, including the sex-linked recessive lethal test in Drosophila, unscheduled DNA synthesis in rat hepatocytes (gavage) and micronucleus induction in bone marrow of male and female mice (intraperitoneal injection). In vitro, 4-nitroaniline was positive for chromosomal aberration in mammalian cells with metabolic activation (S9). Mixed results were observed for chromosomal aberration without S9 and for sister chromatid exchange with and without S9. A forward mutation assay in mouse lymphoma cells was positive without S9 and negative with S9. The results were negative for unscheduled DNA synthesis in primary rat hepatocytes. The majority of Ames assays were negative with and without S9.

p-Phenylenediamine (CAS RN 106-50-3)

p-Phenylenediamine (PPD) was classified by the International Agency for Research on Cancer (IARC 1987) as a Group 3 substance--“Not classifiable as to its carcinogenicity to humans.”

The toxicity of PPD has been reviewed by the Scientific Committee on Consumer Products (SCCP 2006). PPD did not exhibit evidence of carcinogenicity in mice, rats or rabbits following chronic oral or dermal exposure (SCCP 2006). In vivo, PPD was negative in various genotoxicity tests, including micronucleus induction in rats and mice via oral gavage or intraperitoneal injection and unscheduled DNA synthesis and the DNA damage assay (comet assay) in rats via oral gavage (SCCP 2006). In vitro genotoxicity test results for PPD were mixed. Overall, SCCP (2006) considered that PPD alone is not genotoxic, whereas positive findings were reported from genotoxicity studies when PPD was tested in combination with couplers and/or hydrogen peroxide.

Sulfanilic acid (CAS RN 121-57-3) and sodium sulfanilate (CAS RN 515-74-2)

Sulfanilic acid and its sodium salt are considered toxicologically equivalent. No empirical cancer bioassay or in vivo genotoxicity testing data have been identified for these substances. Sulfanilic acid was negative in all available in vitro genotoxicity testing, including gene mutation (Ames tests) in various Salmonella strains with and without metabolic activation and DNA damage/repair assays (SOS response and DNA strand breaks) in Escherichia coli (European Commission ©2000; Ben Mansour et al. 2009).

2-(4-Aminophenyl)-6-methylbenzothiazole-7-sulfonic acid (CAS RN 130-17-6)

No empirical cancer bioassay data for 2-(4-aminophenyl)-6-methylbenzothiazole-7-sulfonic acid have been identified. The REACH dossier for this substance (REACH 2013^e) reported some in vivo and in vitro genotoxicity testing data. This substance did not induce micronucleus formation in mice in a 90-day dietary study. It induced gene mutation in bacteria in one Ames test in Salmonella strains TA98 and TA100 with metabolic activation at high dose levels (greater than 1000 µg/plate), but it was negative in another Ames test in Salmonella strains TA98, TA100, TA1535 and TA1537 with and without metabolic activation. The purity of the test material in these studies was not reported.

p-Phenetidine (CAS RN 156-43-4)

The health effects of p-Phenetidine were assessed along with Cartain Aromatic Amines in a separate Screening Assessment (Environment Canda and Health Canada 2014a). No empirical cancer bioassay data for p-Phenetidine have been identified. p-Phenetidine was classified by the European Commission as a Category 2 mutagen--“Suspected of causing genetic defects” (European Commission 2008). p-Phenetidine induced micronucleus formation in mice following intraperitoneal injection and in vitro gene mutation in Salmonella strains TA98 and TA100 with metabolic activation and in mammalian cells with and without metabolic activation (OECD 2002). p-Phenetidine also induced chromosomal aberration and DNA strand breaks in mammalian cells in vitro (Japan MHLW; Nordernskjöld and Moldéus 1983; Andersson et al. 1982). However, it did not induce mammalian cell transformation (Patierno et al. 1989).

1,4-Naphthalenediamine (CAS RN 2243-61-0)

No empirical cancer bioassay data for 1,4-naphthalenediamine have been identified. For the genetic effects, only two Ames studies were identified for 1,4-naphthalenediamine, and positive results were observed in Salmonella strains TA98 and TA100 only in one study with metabolic activation (Mortelmans et al. 1986; Zeiger et al. 1992).

4-Amino-2-methoxy-5-methylaniline (CAS RN 5307-00-6)

No empirical cancer bioassay data for 4-amino-2-methoxy-5-methylaniline have been identified. For the genetic effects, in the only available Ames study, 4-amino-2-methoxy-5-methylaniline was negative in all Salmonella strains tested, with and without metabolic activation (Shahin 1994).

Aniline (CAS RN 62-53-3)

The human health risk assessment for aniline was conducted previously by Health Canada (Canada 1994; Health Canada 2011a), and it summarized that “Acute or short-term exposure to aniline has been reported to cause reversible methaemoglobin formation in experimental animals and humans. There is limited evidence of carcinogenicity of aniline in laboratory animals exposed to high doses. The in vitro or in vivo genotoxicity data were mixed; however, there is no evidence to support the direct genotoxicity potential of aniline” (Health Canada 2011a).

Based on the toxicological data reviewed by Health Canada (Canada 1994; Health Canada 2011a), chronic dietary exposure to aniline has significantly induced splenic tumours in male rats, but not in female rats or mice. In vivo, aniline induced DNA damage and micronuclei via oral exposure, but not chromosomal aberration via intraperitoneal injection. In vitro, aniline induced chromosomal aberration, but not gene mutation in bacteria (Ames test). Aniline did not induce mammalian cell transformation, and the tests results for micronucleus induction were mixed.

4,4′-Diamino-2,2′-stilbenedisulfonic acid (CAS RN 81-11-8)

4,4′-Diamino-2,2′-stilbenedisulfonic acid did not exhibit evidence of carcinogenicity in rats or mice in 2-year dietary studies (NTP 1992). This substance did not induce gene mutation in bacteria (Ames test) or chromosomal aberration or sister chromatid exchanges in mammalian cells, with and without metabolic activation (Zeiger et al. 1987; Loveday et al. 1990; OECD 2004).

Amido-G-Acid (CAS RN 86-65-7)

The carcinogenic potential of Amido-G-Acid has not been investigated via conventional routes of exposure (oral, dermal or inhalation). In a short-term study, this substance was given to A/St mice via intraperitoneal injection at 325, 625 or 1250 mg/kg-bw per injection, 3 times per week for 5 (high-dose group) to 8 (two lower-dose groups) weeks, and the surviving animals were observed for 24 weeks. Significantly increased incidences of pulmonary adenoma were observed in the lowest dose group only (Theiss et al. 1981). No dose–response relationship was observed. No other health effects were examined. The findings of this study are considered to be inadequate evidence for the potential carcinogenic effects of Amido-G-Acid. Amido-G-Acid was negative in the Ames test (Jung et al. 1992).

Mesalamine (CAS RN 89-57-6)

The health effects of mesalamine were critically reviewed by the US Food and Drug Administration (US FDA 2008). Mesalamine did not exhibit evidence of carcinogenicity in rats or mice in 2-year dietary studies. It did not induce micronucleus formation in mice via oral administration or sister chromatid exchanges in hamsters via intraperitoneal injection. It did not induce gene mutation in bacteria (Ames test), with and without metabolic activation (US FDA 2008).

The remaining postulated azo bond reductive cleavage products

Empirical carcinogenicity and genotoxicity data for 4-amino-1,3-Benzenediol (CAS RN 13066-95-0), 1,3-bis(4-amino-2-methoxy-5-methylphenyl)urea, 1,3-bis(4-amino-3-methylphenyl)urea and 4-amino-N-(4-aminophenyl)benzenesulfonamide have not been identified. These substances are not sulfonated, and their carcinogenic and genotoxic potential remains unknown.

4-Amino-4’-nitro-2,2’-stibenedisulfonic acid (CAS RN 119-72-2), 3-aminonaphthalene-1,5-disulfonic acid (CAS RN 131-27-1), Amsonic acid sodium salt (CAS RN 25394-13-2), 2-amino-3,5-xylenesulfonic acid (CAS RN 88-22-2) and the remaining reductive cleavage products are sulfonated aromatic amines. These substances are considered to have low potential to be carcinogenic and genotoxic.

Top of Page

Appendix D: Azo Direct Dyes with Effects of Concern

Some of the Azo Direct Dyes in this assessment have effects of concern based on potential carcinogenicity. The details for supporting the potential carcinogenicity for these substances are outlined in section 7.2 Health Effects Assessment (see specific sub-sections), and generally based on one or more of the following lines of evidence:

• Classifications by national or international agencies for carcinogenicity (may be a group classification).
• Evidence of carcinogenicity in animal studies and/or human epidemiology based on the specific substance.
• Potential to release one or more of the EU22 aromatic amines by azo bond cleavage.
• Read-across to related substances for which one or more of the above lines of evidence apply.