Sage Journals: Discover world-class research

Abstract

Flexible polyurethane foams (PUF) are used in many consumer products. PUF may contain trace levels of aromatic diamine impurities that could represent a potential health risk. The risk associated with sleeping on a PUF mattress was evaluated. Toxicity benchmarks for sensitization and non-cancer endpoints were derived from the respective points-of-departure using standard assessment factors. For the cancer endpoints, toxicity benchmarks were derived from the 25th-percentile values of animal studies. Recently published emission and migration data allowed to link exposure with the CertiPUR^TM voluntary quality limits of ≤5 mg.kg⁻¹ for 2,4-toluene diamine and 4,4’-methylene dianiline in PUF. Using conservative exposure scenarios, lifetime-average daily internal doses from the combined inhalation and dermal exposures were calculated. Margins of safety for non-cancer and sensitization endpoints were >10⁴. The theoretical excess cancer risk was ≤1.5 × 10⁻⁷. It is concluded that sleeping on a mattress that satisfies the CertiPUR limit value does not pose undue risk to consumers.

Keywords

Polyurethane (PU) foam aromatic diamines toluene diamine (TDA)methylene dianiline (MDA)emission migration risk assessment CertiPUR

Introduction

Flexible Polyurethane Foam (PUF) is produced from the reaction of long-chain polyols with diisocyanates in the presence of water, catalysts, and additives. The polyols and diisocyanates make up ca. 95% of the formulation (Brereton et al., 2019; Szycher, 2012). The polyol is typically a polyether polyol and the diisocyanates are either toluene diisocyanate (TDI) or methylenediphenyl diisocyanate (MDI). PUF is manufactured in industrial settings for subsequent use in consumer products such as furniture, automotive interiors, bedding, and carpet underlay.

The main polymerization reaction takes place between the hydroxyl groups of the polyols and the isocyanate (NCO) groups of TDI or MDI, forming the polyurethane backbone of the PUF. Some of the NCO groups react with water to produce CO₂, which facilitates foaming. This hydrolysis process converts the NCO groups into transient amino groups. Since the reactivity of NCO with amino groups is approximately 10⁴ times greater than with hydroxyl groups (Schupp and Plehiers, 2022), various polyurea species are formed and become part of the PUF.

Because of the greater reactivity of NCO groups with amino groups compared to hydroxyl groups, it was once assumed that no residual unbound diamines would be present in PUF (Hall et al., 2001; Hillier et al., 2003). However, it is now generally accepted that, under certain circumstances, residual trace quantities of toluene diamine (TDA) and methylene dianiline (MDA) may be present in fresh PUF at detectable (low mg.kg⁻¹) levels. The residual diamine concentrations have been found to decay over time in normal conditions (Krämer-Lucas et al., 2001).

Since 2,4-TDA and 4,4’-MDA meet the classification requirements in the United Nations Globally Harmonized System (GHS) (UN, 2021) of Class 1B carcinogens (EU Commission, 2008), the CertiPUR^TM voluntary industry certification scheme was introduced and set a limit for the presence of these diamines in PUF at ≤5 mg.kg⁻¹ (EUROPUR, 2021). This limit was based on risk assessment (RA) work done in the early 2000s. Analytical methodology has been developed that enables establishing relationships between the diamines content and emission and migration rates from PUF (Karlsson, 2023; Karlsson et al., 2023).

In this work, the RA approach taken by Arnold et al. (2012) was followed. Their work addressed concerns of potential health effects from residual TDI in PUF associated with sleeping on a PUF mattress. In that work, toxicity benchmarks for key cancer and non-cancer endpoints (e.g., sensitization) were derived from human and animal data and were used to evaluate migration and emission of TDI from PUF based on data by Vangronsveld et al. (2013) in an “extreme case” exposure scenario for sleeping on a PU foam mattress. Upper-bound inhalation exposure to TDI as well as a lifetime-average daily dose to TDI from dermal exposure were calculated. Arnold et al. (2012) concluded that sleeping on a PU foam mattress does not pose TDI-related health risks to consumers.

Using the same principles of deriving toxicological benchmark values, using current RA approaches, and assessing experimental emission and migration data, this work provides a rigorous evaluation of the 5-mg.kg⁻¹ limits for MDA and TDA in the CertiPUR PUF voluntary certification scheme.

Methods

Risk is a function of both hazard and exposure. RAs are generally performed in four steps (NRC, 1983): hazard, dose-response, exposure, and risk assessment. The sections delineate each of these steps, the data used, and the assumptions made for the RA scenario.

Substances and hazard classification

TDI for manufacturing PUF is in most cases an 80:20 mixture of the 2,4- and 2,6-TDI isomers. In rare instances, a 65:35 mixture is used. Within the scope of this investigation, the corresponding diamines are 2,4- and 2,6-TDA, both of which are classified as skin sensitizers (ECHA, 2023a; ECHA 2023b). For 2,4-TDA (CASRN 95-80-7), a chronic toxicity study by oral administration in rats and mice is available (Cardy, 1979; NCI, 1979), on the basis of which 2,4-TDA can be classified as Carcinogen Category 1B (BAuA, 2008). 2,4-TDA is treated as a genotoxic carcinogen. In another chronic study by oral administration in rats and mice (NCI, 1980), 2,6-TDA (CASRN 823-40-5) was found not to be carcinogenic. Tumors were found but their incidence was not significantly increased compared to controls. Although 2,6-TDA displayed both negative and positive results in in vivo genotoxicity studies, based on the 2-year cancer study, 2,6-TDA is taken to be non-carcinogenic (EPA, 2005a).

MDI used in flexible polyurethane foam production may contain the three isomers (4,4’-, 2,4’-, and 2,2’-MDI). The corresponding diamines of relevance in this investigation are 4,4’-, 2,4’-, and 2,2’-MDA. The 4,4’-MDA isomer (CASRN 101-77-9) is classified as a skin sensitizer (ECHA, 2023c). For lack of data, the 2,4’- and 2,2’-MDA isomers are assumed to exhibit the same property. A chronic toxicity study by oral administration in rats is available for 4,4’-MDA (Lamb et al., 1986; NTP, 1983; Weisburger et al., 1984), on the basis of which 4,4’-MDA was classified as Carcinogen Category 1B (ECB, 2001). 4,4’-MDA is treated as a genotoxic carcinogen. No chronic study for 2,2’-MDA (CASRN 6582-52-1) is available and its hazard potential was conservatively assumed to be the same as that of 4,4’-MDA. There is no chronic study available for 2,4’-MDA (CASRN 1208-52-2) either. Even though 2,4’-MDA may not be genotoxic based on negative bacterial reverse mutation (OECD, 2020), chromosomal aberration (OECD, 2016a), and in vitro gene mutation (HPRT) (OECD, 2016b) tests (Unterberger-Henig, 2022), the hazard profile of 4,4’-MDA was conservatively assumed to apply to 2,4’-MDA as well.

Dose-response assessment and derivation of toxicological benchmarks

Carcinogenicity

The most relevant and sensitive systemic effect for 2,4-TDA and 4,4’-MDA is carcinogenicity. The derivation of the toxicological benchmark was based on the T25 from the respective 2-year oral chronic animal studies. T25 is the dose at which 25% of test animals would exhibit the effect, that is, tumor formation. The oral T25 can be taken as the basis for estimating risk provided it is compared with the total absorbed dose from all routes of exposure. An allometric scaling factor of 4 (rat) was applied, assuming a human body weight of 70 kg. The resulting human equivalent T25 values were then adjusted to an excess cancer risk of 10⁻⁶ for consumers (general population) using linear extrapolation to zero, as commonly done for a presumed genotoxic mode of action (BAuA, 2008, 2010; ECHA, 2015).

The use of benchmark dose software (BMDS V.3.3.2, US Environmental Protection Agency, Washington DC, USA) as an alternative to the T25-based approach was explored but ultimately abandoned after careful consideration of the results of the chronic studies and the objections formulated by BAuA (2010: 12) and DECOS (2015: 19). While no overt signs thereof were apparent in the MDA study with rats, the high-dose groups in both the MDA study with mice and the TDA studies with either species were clearly overdosed (early body weight loss, intermediate dosing corrections, premature sacrifice of the animals, significant hepato-nephro-toxicity as well as liver degeneration were noted). With that many uncertainties regarding one of only two non-zero dose levels, the results of the low-dose groups appeared to be the more reliable starting point for deriving benchmark values.

Repeated dose toxicity

The 2,6-TDA isomer was shown not to be carcinogenic in a chronic animal study (NCI, 1980). Reduced body weight was identified as the lowest point-of-departure (POD) and used to derive a toxicity benchmark for this RA. The standard uncertainty factors (UF) for database deficiencies (10), interspecies (10) and intraspecies variability (10) were applied (aggregate UF = 1000; ECHA, 2012; EPA, 2005b).

Skin sensitization (local effects)

For skin sensitization, the Local Lymph Node Assay (LLNA) is the preferred method for deriving quantitative dose-response data (Gerberick et al., 2007; Roberts et al., 2016). Where LLNA data were available for the substances considered, they were used to calculate a human-equivalent No Expected Sensitization Induction Level (NESIL) according to equation (1) (Safford, 2008)

\log (N E S I L [µ g ∙ {c m}^{- 2}]) = 1.16 \times \log (E C 3 [µ g ∙ {c m}^{- 2}]) - 0.64

(1)

The EC₃-value is the concentration that led to a stimulation index of 3 compared to controls. In equation (1), it is converted into a surface loading (EC₃-potency). In the absence of LLNA results, the GHS classification boundary of 1% for Category 1B dermal sensitizers in mixtures (UN, 2021), corresponding to an EC₃ of 250 μg.cm⁻², was used as a surrogate input to equation (1) for estimating the NESIL. The NESIL was then converted to an Acceptable Exposure Level (AEL) using UFs of 1 for interspecies (Basketter et al., 2018) and 10 for intraspecies variability, and 3 for matrix effects (Basketter and Safford, 2016) (aggregate UF = 30).

Experimental data and exposure assessment

To link the emission (inhalation) and migration (dermal) exposure potentials with the CertiPUR^TM quality limit (EUROPUR, 2021), data are required from foams that have defined concentrations of TDA or MDA. This exposure assessment made use of the consistent set of emission and migration data generated by Karlsson et al. (2023).

A brief summary is provided here. Both a representative TDI- and MDI-based PUF were thermally treated to generate extractable levels of diamines close to the CertiPUR limit value of 5 mg.kg⁻¹. Extractable diamines were determined using a method that mirrors the CertiPUR guidance (EUROPUR, 2021) but using 2 instead of the prescribed 4 consecutive extractions with 1% aqueous acetic acid. This could have resulted in a slight underestimation of the extractable diamine content in each foam. When coupled with the fact that measured results were normalized to the 5 mg.kg⁻¹ foam content limit value (as described in the Migration rates section below), the measured extraction values could result in a slight overestimation of the risk and would therefore be considered conservative for the risk assessment evaluation.

Emission rates (inhalation exposure)

The foam samples for emission testing contained up to 9.2 and 6.0 mg.kg⁻¹ of 2,4- and 2,6-TDA, respectively, and up to 27 mg.kg⁻¹ of the summed MDA-isomers (ΣMDA). Emissions from the foams were measured during 8 h using a specifically-designed and validated emission cell (Karlsson, 2023). The limit-of-quantification (LOQ) for the emission rate was 8.5 ng.m⁻².h⁻¹ for 2,4-TDA, 28 ng.m⁻².h⁻¹ for 2,6-TDA, and 71 ng.m⁻².h⁻¹ for ΣMDA. All measured emission rates were less than the respective LOQ.

In the RA, emission for the entire period of use of the mattress was assumed to take place from all surfaces of the mattress at an average rate corresponding to one-half of the LOQ (EPA, 1991) linearly pro-rated to the 5-mg.kg⁻¹ limit value. The rate thus determined was converted into the average external daily inhalation dose as described under “Exposure scenario.”

Migration rates (dermal exposure)

Migration was studied by 15-h tests and by 35-day time-course migration experiments consisting of 2 sequential 3-h tests each on Days 1, 2, 3, 4, 5, 10, and 35. The TDI-based foam samples contained 3.7 mg.kg⁻¹ of 2,4-TDA and 1.4 mg.kg⁻¹ of 2,6-TDA. The MDI-based foam contained a total of 14.1 mg.kg⁻¹ of ΣMDA. Measurable migration of TDA was only detected during the first 2 days of testing and rapidly declined to values less than the LOQ. For MDA, detectable migration only occurred during the first 3 days of testing. The respective aggregate LOQ were 1.7 μg.m⁻².d⁻¹ for each of the TDA-isomers and 3.4 μg.m⁻².d⁻¹ for ΣMDA¹. Migration rates matched with a theoretical model based on Fick’s diffusion law (Karlsson et al., 2023; OECD, 2009), whereby migration rates were linear in the initial concentration of the substance and evolved with the inverse square root of time (t^−0.5). This theoretical basis allowed for a well-founded extrapolation of the experimental data to the anticipated period of use of a mattress. The migration rates were summed over the use period as summarized in Table 1, and then divided by 1826.25 days (5 years) to obtain average daily migration rates. The five-year service life of a mattress is a conservative lower estimate of how long consumers use a foam mattress (see discussion below).

Table 1.

Data from Karlsson et al. (2023) and adjustments made for estimating the average daily migration rate. LOQ = limit of quantification; ΣMDA = sum of the three methylene dianiline isomers; TDA = toluene diamine.

Period	2,4-TDA	2,6-TDA	ΣMDA	Karlsson et al. (2023) data basis and adjustments
	μg.m⁻²
Day 1	7.2	6.6	5.0	2,4-TDA and 2,6-TDA: migrated quantities during 15-h preliminary test were 11.5 and 10.5 ng, respectively, from a surface area of 0.00159 m². The quantities were assumed to have migrated during the first 8 h of contact
Day 1	7.2	6.6	5.0	ΣMDA: migrated quantities during the first day of time-course testing were 8 ng over the two 3-h periods
Days 2-4	3 × 1.7	3 × 1.7	3 × 3.4	Aggregate LOQ adjusted from a 3-h to an 8-h exposure
Days 5 thru 1826.25	137	137	273	Migration rate using the theoretical diffusion model integrated over a 5-year timeframe: LOQ x ∫ t^−0.5 dt integrated between 4 and 1826.25 = 2 LOQ (√1826.25 – √4) = 81.5 LOQ.
Total	149	148	288	Sum of the three preceding rows
Daily average	0.082	0.081	0.158	Total divided by number of days in a 5-year period
	ng.m⁻².h−1
Hourly average	13.8	36.2	7.0	Daily average divided by 8 h and scaled to the 5-mg.kg⁻¹ CertiPUR limit, using 5/3.7 for 2,4-TDA, 5/1.4 for 2,6-TDA, and 5/14.1 for ΣMDA.

In the RA, migration rates were linearly normalized to the 5 mg.kg⁻¹ foam content limit value from the actual level in the foam sample used and adjusted to an 8-h exposure duration as needed. A mitigation factor of 50% was applied to take into account the presence of a sheet (EPA, 2012) for determining the average external daily dermal dose.

Risk assessment

Exposure scenario

An exposure scenario was constructed representing the reasonable-worst-case scenario of a person sleeping on a foam mattress for 8 h per day, based on the same exposure scenario used by Arnold et al. (2012).

The average daily external and internal dermal doses (D_derm,ext and D_derm,int) were calculated using equations (2) and (3)

D_{d e r m, e x t} = M \times t \times A_{m} \times T P F

(2)

D_{d e r m, int} = D_{d e r m, e x t} \times {A b s}_{d}

(3)

The average daily external and internal inhalation doses (D_inh,ext and D_inh,int) were calculated using equations (4) and (5)

D_{i n h, e x t} = \frac{E \times A_{e}}{V \times A C H} \times B R

(4)

D_{i n h, int} = D_{i n h, e x t} \times {A b s}_{i}

(5)

Table 2 explains and summarizes the symbols, parameters, and underlying assumptions. Summation of the internal dermal and inhalation doses resulted in the daily average total internal dose as a basis for comparison with the toxicological benchmark values. Because there is solid evidence that their repeated-dose toxicities are different, doses were calculated separately for 2,4- and 2,6-TDA. Since no chronic repeated-dose studies are available for 2,2’- and 2,4’-MDA, from a precautionary perspective, these two isomers were presumed to be equipotent to 4,4’-MDA and the three isomers were conservatively combined in a single dose calculation.

Table 2.

Summary of symbols, parameter values, and reference sources used as input for dermal and inhalation exposure scenarios. Symbols refer to equations (2)–(5). LOQ = limit of quantification; MDA = methylene dianiline; TDA = toluene diamine.

Parameter	Symbol	Value	Units	Explanation
	Dermal exposure
External dermal dose	D_derm,ext	Substance-specific	ng.d⁻¹	Calculated with equation (2)
Internal dermal dose	D_derm,int	Substance-specific	ng.d⁻¹	Calculated with equation (3)
Migration rate	M	Substance-specific	ng.m⁻².h⁻¹	Derived from migration experiments of Karlsson et al. (2023)
Time	t	8	h.d⁻¹	Standard period of sleep
Contact surface area	A_m	1.075	m²	Half the body surface area of an adult 40 - 50 years old male as described in the US EPA exposure factor handbook (EPA, 2011)
Textile penetration factor	TPF	0.5	—	Factor taken to represent the effect of bed sheets used on mattresses in mitigating exposure. Based on a 50% factor that is used by the EPA office of pesticides for the protection factor offered by a single layer of clothing for pesticides handlers (EPA, 2012)
Dermal absorption	Abs_d	Substance-specific	—	50% for the MDA isomers (El-Hawari et al., 1986; ECHA, 2015)25% for the TDA isomers (Marzulli et al., 1981; BAuA, 2008)
	Inhalation exposure
External inhalation dose	D_inh,ext	Substance-specific	ng.d⁻¹	Calculated with equation (4)
Internal inhalation dose	D_inh,int	Substance-specific	ng.d⁻¹	Calculated with equation (5)
Emission rate	E	Substance-specific	ng.m⁻².h⁻¹	½ LOQ of the emission experiments of Karlsson et al. (2023)
Emission surface area	A_e	10.4	m²	King-size mattress (2 m × 2 m × 0.3 m)
Room volume	V	30	m³	Moderately-sized bedroom (EN 16516:2017)
Air exchange rate	ACH	0.5	h⁻¹	Less than low rate for mechanically ventilated bedroom (Te Biesebeek et al., 2014)
Breathing rate	BR	0.5	m³/h	Extreme case assumption covering the majority of the reported ages/percentile breathing rates in Tables 6–17 of the US EPA exposure factor handbook (EPA, 2011)
Inhalation absorption	Abs_i	100	%	Default factor for inhalation exposure, previously used for risk assessments of 4,4’-MDA and 2,4-TDA (ECHA, 2015; BAuA, 2008)

Risk characterization

Deterministic risk evaluations are typically expressed either as the margin of safety (MOS), where MOS equals the toxicological benchmark divided by the average daily total internal dose, or as its inverse, the risk characterization ratio (RCR). A value of MOS >1 or RCR <1 indicated that there is no unacceptable risk to human health.

Sensitivity analyses

It is plausible that an individual may have additional contact with PUF during waking hours, for instance, from upholstered furniture or automotive vehicles. To estimate the effect of this potential additional exposure, a sensitivity analysis was performed whereby the time for the “sleeping on a mattress” scenario was extended to 24 h.

One of the purposes of this work was to probe the adequacy of the 5-mg.kg⁻¹ CertiPUR quality limit for MDA. The main analysis considered the sum of the three MDA isomers. In addition, the effects of a ΣMDA limit of 15 mg.kg⁻¹ and a 5-mg.kg⁻¹ limit for 4,4’-MDA alone were assessed.

Results

Derivation of toxicological benchmarks

Carcinogenicity

4,4'-MDA

The key study for carcinogenicity was performed in F344 rats and B6C3F1 mice (Lamb et al., 1986; NTP, 1983; Weisburger et al., 1984). Both liver and thyroid tumors were observed. The T25_{(oral, rat)} value derived from this study was 9.01 mg. kg_bw⁻¹.d⁻¹ (ATSDR, 1998; ECHA, 2015), based on liver tumors in male rats², following correction for control incidence and study duration (actual 103-week vs normal 104-week). From this, a systemic toxicological benchmark of 9 ng. kg_bw⁻¹. d⁻¹ was derived. Since no isomer-specific carcinogenicity studies were identified, the systemic toxicological benchmark of 4,4’-MDA was also used for the 2,4’-MDA and 2,2’-MDA isomers

2,4-TDA

The key study for 2,4-TDA was performed in F344 rats and B6C3F1 mice (Cardy, 1979; NCI, 1979). Target organs in rats included both liver and mammary glands. Although the EU RA noted uncertainty in the biological relevance of the mammary gland tumors, they were taken forward as a conservative basis for developing the toxicity benchmark (BAuA, 2008). The T25_{(oral, rat)} valued based on tumorigenesis of the mammary glands was 2 mg.kg_bw⁻¹.d⁻¹ (BAuA, 2008), for liver tumors 14 mg.kg_bw⁻¹.d⁻¹. From this, a systemic toxicological benchmark of 2 ng. kg_bw⁻¹.d⁻¹ was derived.

Repeated dose toxicity

2,6-TDA

2,6-TDA was shown not to be a carcinogen based on a 2-year carcinogenicity study (NCI 1980). Although tumors were observed in both rats and mice dosed with 2,6-TDA hydrochloride, the incidences were not statistically increased from controls (NCI 1980). This study provided adequate data to derive benchmarks based on non-cancer endpoints that can be used for RA. Specifically, reduced body weight in rats was not considered adverse, so the high dose of 25 mg. kg_bw⁻¹.d⁻¹ in male rats (32 mg. kg_bw⁻¹.d⁻¹ for females) could be used as POD (EPA, 2005a). Applying the standard uncertainty factor of 1000, a toxicological benchmark of 25,000 ng. kg_bw⁻¹.d⁻¹ was obtained, in agreement with the published reference dose (EPA, 2005a).

Sensitization

TDA

For 2,4-TDA, a reliable LLNA study (Vanoirbeek et al., 2009) reported an EC₃ value of 19%. The corresponding NESIL was then calculated to be 4216 µg.cm⁻². The choice of a safety assessment factor for sensitization is somewhat case-specific as matrix effects that may affect dermal absorption could vary widely depending on the application under consideration. For dermal contact with foams a safety factor of 30 was used, which yielded an AEL of 141 µg.cm⁻².

Since no LLNA study was found for 2,6-TDA, the same AEL was applied.

MDA

For 4,4'-MDA, no LLNA study has been reported. Based on a human repeated insult patch test (ECHA 2023d), 4,4'-MDA was classified as a Category 1B dermal sensitizer. Taking the GHS reporting threshold of 1% as a surrogate input, a NESIL of 139 µg/cm² was derived. Applying the assessment factor of 30 yielded an AEL of 4.6 µg.cm⁻². Since no isomer-specific sensitization studies were identified, the AEL of 4,4-MDA was also used for the 2,4’-MDA and 2,2’-MDA isomers.

Exposure estimation

Emission rates (inhalation exposure)

ΣMDA

The emission test was conducted with a 150-mm diameter exposed foam surface (0.0177 m²), a test time duration of 8 h, and a ΣMDA LOQ of 71 ng.m⁻².h⁻¹. No emission was detected under these circumstances. The emission rate based on half the LOQ (½ × 71) was 35 ng.m⁻².h⁻¹. Since the foam samples contained up to 27 mg.kg⁻¹ of ΣMDA, scaling to the 5-mg.kg⁻¹ CertiPUR limit value resulted in an adjusted emission rate of (35 × 5/27) 6.5 ng.m⁻².h⁻¹.

TDA

The TDI-based foam samples contained up to 9.2 and 6 mg.kg⁻¹ 2,4-TDA and 2,6-TDA, respectively. No emission was detected at the respective LOQ of 8.5 and 28 ng.m⁻².h⁻¹. Emission rates based on half the LOQ were 4.2 and 14 ng.m⁻².h⁻¹ for 2,4- and 2,6-TDA, respectively. Scaled to the 5-mg.kg⁻¹ limit, the adjusted emission rates were 2.3 ng.m⁻².h⁻¹ for 2,4-TDA and 12 ng.m⁻².h⁻¹ for 2,6-TDA.

Migration rates (dermal exposure)

The results of Karlsson et al. (2023) showed that migration occurred primarily during Day 1 and that little to no replenishment of TDA and MDA from the interior to the surface of the foam took place. Average daily migration rates were calculated as shown in Table 1. The migration tests were conducted with a 45-mm diameter exposed foam surface (0.00159 m²). For Day 1, the maximum measured value was taken into account. For Days 2 through 4, the aggregate LOQ was taken as a basis. For the remainder of the mattress use period, the migration rate was integrated using the theoretical t^−0.5 model (Karlsson et al., 2023). The average hourly values were scaled to the 5-mg.kg⁻¹ CertiPUR limit using the measured concentration of diamine in the foam samples.

ΣMDA

For ΣMDA, an average adjusted hourly migration rate of 7.0 ng.m⁻².h⁻¹ was obtained.

TDA

For 2,4-TDA, an average adjusted hourly migration rate of 13.8 ng.m⁻².h⁻¹ was calculated. For 2,6-TDA, the adjusted hourly migration rate was 36.2 ng.m⁻².h⁻¹.

The adjusted average hourly emission and migration rates were carried forward into the exposure scenario and RA scheme as described in Methods (equations (2)–(5) and Table 2). Detailed calculations and the risk characterization are available in Supplemental Information.

Risk characterization

Carcinogenicity

2,4-TDA

A total systemic uptake of 21 ng.d⁻¹ and a corresponding internal dose of 0.3 ng. kg_bw⁻¹. d⁻¹ were calculated. Compared to the toxicological benchmark of 2.0 ng. kg_bw⁻¹. d⁻¹, the MOS against a 10⁻⁶ risk level was 6.6.

ΣMDA

A total systemic uptake of 33 ng.d⁻¹ and a corresponding dose of 0.5 ng.kg_bw⁻¹.d⁻¹ were calculated. Compared to the toxicological benchmark of 9.0 ng.kg_bw⁻¹. d⁻¹, the MOS against a 10⁻⁶ risk level was 19.

For both 2,4-TDA and ΣMDA, sleeping on a PUF mattress that satisfies the CertiPUR quality standard of ≤5 mg.kg⁻¹ does not constitute an unacceptable risk to consumers.

Repeated dose toxicity

2,6-TDA

A total systemic uptake of 72 ng.d⁻¹ and a corresponding internal dose of 1.0 ng. kg_bw⁻¹. d⁻¹ were calculated. Compared to the toxicological benchmark of 25,000 ng. kg_bw⁻¹. d⁻¹, the MOS was 2.4 × 10⁴, indicating that there is no systemic adverse health risk resulting from exposure to 2,6-TDA.

Sensitization

For local effects, the relevant metric is the highest single dermal dose, which would occur on the first day. The Day 1 migration rates were 33 μg.m⁻² for ΣTDA and 2 μg.m⁻² for ΣMDA. The AEL were 141 μg.cm⁻² for the ΣTDA and 4.6 μg.cm⁻² for ΣMDA. This resulted in MOS of 8.4 × 10⁴ for the ΣTDA and of 5.2 × 10⁴ for ΣMDA. The sensitization risk resulting from exposure to the diamines is very low, even when considering “peak” exposures during the first day of mattress use.

Sensitivity analyses

When the contact and residence time for the “sleeping on a mattress” exposure scenario was extended to 24 h (discussed in more detail below), the combined internal dose increased to 0.9, 3.1, and 1.4 ng.kg⁻¹. d⁻¹ resulting in MOS of 2.2, 8100 and 6.3 for 2,4-TDA, 2,6-TDA, ΣMDA, respectively.

When the RA was conducted considering only the 4,4’-MDA isomer (i.e., assuming that the other two isomers would not be carcinogens), an MOS of 9.0 resulted (Supplemental Information). When the ΣMDA concentration in the foam was increased to 15 mg.kg⁻¹, the combined internal dose became 1.4 ng.kg ⁻¹. d ⁻¹ resulting in an MOS of 6.3.

Discussion

An overall summary of the risk characterization is given in Table 3. Complete details can be found in the Supplemental Information. The sensitization and non-cancer endpoints have very large MOS and slight variations in uncertainty factors or assumptions would not significantly affect the RA outcome. Therefore, in the subsequent paragraphs, the discussion focuses on the uncertainties in the RA for the systemic effects of 2,4-TDA and ΣMDA.

Table 3.

Summary of diamine risk characterization for “sleeping on a foam mattress” exposure scenario. LOQ = limit of quantification; ΣMDA = sum of the three methylene dianiline isomers; TDA = toluene diamine, MoS = margin of Safety, EC3 = effective concentration for a stimulation index of 3, NESIL = no expected sensitization induction level, AEL = acceptable exposure level.

Substance		2,6-TDA	2,4-TDA	ΣMDA
	Units
Toxicological Effect		Repeated dose	Cancer	Cancer
Emission
½ LOQ emission	ng.m⁻².h⁻¹	14	4.2	35
Target-to-test concentration	—	= 5/6	= 5/9.2	= 5/27
Emission rate at 5 mg.kg⁻¹	ng.m⁻².h⁻¹	11.8	2.3	6.5
Daily internal inhalation dose	ng.d⁻¹	33	6.4	18
Migration
Day 1 migration	μg.m⁻²	10.5/0.00159 = 6.6	11.5/0.00159 = 7.2	8.0/0.00159 = 5.0
Days 2-4 migration	μg.m⁻²	1.7 (LOQ) × 3 (d) = 5.0	1.7 (LOQ) × 3 (d) = 5.0	3.4 (LOQ) × 3 (d) = 10.1
Days 5-1826 migration (integral of LOQ)	μg.m⁻²	137	137	273
Average migration rate	μg.m⁻².d⁻¹	(6.6 + 5.0 + 137)/1826 = 0.081	(7.2 + 5.0 + 137)/1826 = 0.082	(5.0 + 10.1 + 273)/1826 = 0.158
Target-to-test concentration	—	= 5/1.4	= 5/3.7	= 5/14.1
Daily internal dermal dose	ng.d⁻¹	39	15	15
Combined internal dose	ng.kg_bw⁻¹.d⁻¹	(33 + 39)/70 = 1.03	(15 + 6)/70 = 0.30	(18 + 15)/70 = 0.47
Toxicological benchmark dose	ng.kg_bw⁻¹.d⁻¹	25,000	2	9
MoS	—	24,000	6.6	19.0
Toxicological Effect		Sensitization		Sensitization
		ΣTDA		ΣMDA
Day 1 external load	μg.cm⁻²	33 (μg.m⁻²) x 0.5 (TPF)/10000 = 1.7 × 10⁻³		2 × 0.5/10000 = 8.9 × 10⁻⁵
EC3/NESIL		4750/4217		250/139
AEL		141		4.6
MoS	—	84,000		52,000

The calculated systemic internal inhalation dose was approximately 30% of the total internal dose for 2,4-TDA and 55% for ΣMDA. One-half of the respective LOQ was used as model input. No decrease in emission rates was assumed over the lifetime of the mattress. This is a conservative assumption since it is expected that emissions of substances from PUF would decrease with time, either by depletion of the substance by emission itself (Hillier et al., 2003; OECD, 2009) or—in the case of diamines—by reaction with the PU matrix (Krämer-Lucas et al., 2001).

The scenario parameters (Table 2) can generally be considered conservative as well. A king-size mattress (2 m × 2 m x 0.3 m) was considered. The room volume (30 m³) was in the range indicated by regulatory bodies (16 m³ (Te Biesebeek et al., 2014) to 41 m³ (Jennings et al., 1987) and matched the reference room size (4 m × 3 m × 2.5 m) used for evaluating construction products (EN 16516, 2017; Oppl, 2014). The ventilation rate (0.5 h⁻¹) was less than the smallest factor published by the Dutch National Institute for Public Health and Environment (RIVM) (Te Biesebeek et al., 2014) (0.6 h⁻¹ for a bedroom with mechanical ventilation). The assumed breathing rate while sleeping was greater than the 95th-percentile reported for all age categories by EPA (2011) and the standard rate used by the Netherlands National Institute for Public Health and the Environment (RIVM) (0.45 m³.h⁻¹, Te Biesebeek et al., 2014).

On balance, the parameters and emission rates used for estimating the internal inhalation dose can be considered a reasonable-worst-case model for sleeping on a mattress. However, it can be argued that 8 h of sleep per day is not necessarily conservative and that longer residence times in the room would be possible or that exposure could occur in other places that contain PUF. This was the subject of the sensitivity analyses discussed below.

The calculated internal dermal dose accounted for approximately 70% of the total internal dose for 2,4-TDA and 45% for ΣMDA. The calculation was based on the measurements performed by Karlsson et al. (2023). For Day 1, the largest measured migration rate was used, that is, for TDA isomers the results of the 15-h experiments and for MDA an aggregate of the measured migration over the two first day 3-h measurements of the time course series. For Days 2 thru 4, while most measurements were <LOQ, migration was assumed to correspond to the aggregate LOQ. For both periods, these are conservative assumptions. For the remainder of the mattress use life, a theoretical model based on Fick’s diffusion law was used (Karlsson et al., 2023). This model is similar to the one described by OECD (2009) for modeling emissions.

The scenario parameters (Table 2) can generally be considered conservative as well. Mattress replacement is assumed to take place every 5 years. Foley (2023) recommends mattresses be replaced every 6–8 years. In The Netherlands, ca. 1.5 million mattresses are discarded every year on a population of 17.4 million (Michel Baumgartner, 15 March 2021; personal communication), suggesting a service life exceeding 11 years. The 8-h contact time is more conservative than for inhalation: while the residence time in the room may be longer, dermal contact is less likely to exceed 8 h.d⁻¹. The contact area of 1.075 m² (half the body skin surface area) assumes the person lies flat and sunk into the mattress for the entire 8 h of sleeping. The textile protection factor (0.5) is the one recommended by US EPA for the protection offered by a single layer of clothing against pesticides applied directly onto the mattress cover (EPA, 2012). Finally, it should be noted that Karlsson et al. (2023) observed that 80% of the migrated MDA and 35% of TDA were retained on the cloth that was in direct contact with the foam in the presence of an extraction medium more acidic than sweat. In balance, the parameters and migration rates used for estimating the internal dermal dose represent a reasonable-worst-case model.

The risk assessment results reflect adult use of PU foam mattresses. However, it is possible that the use of PU mattresses could start in infancy. Thus, it is reasonable to examine the potential impact of these earlier life exposures on estimated risk from exposure to the primary aromatic amines. The US EPA has developed guidance on how to account for different (higher) susceptibility to cancer from early lifetime exposures (EPA 2005b). That guidance advises that an additional factor of 10 be applied for exposures from birth to age 2 years and a 3-fold factor be applied for the 14 years between age 2 and 16. The overall impact of assuming lifetime exposure (i.e., including the first 16 years of life, with the noted adjustment factors) is to increase RCRs by a factor of approximately 1.66 or to decrease MOS by about 1.66. For 2,4-TDA, for example, the MOS for carcinogenicity (adults' exposure only) of 6.6 would be adjusted to 4.1 for lifetime exposure. This indicates that cancer risk is unlikely to be elevated to unacceptable levels when considering lifetime exposures, and the same reasoning applies to the MDA isomers.

While the purpose of this work was to provide a health-based evaluation of the diamine quality limits included in CertiPUR rather than to provide a definitive lifetime risk assessment of contact with all potential PUF applications, several sensitivity analyses were conducted to assess the impact of additional considerations.

Consideration of prolonged exposure or exposure to a variety of other PUF applications would require making many additional assumptions. However, dermal exposure from sleeping on a mattress (largest contact area, minimal cover considered) is conservative compared to that of other applications. On the other hand, inhalation exposure during waking time (balance of breathing rate and better ventilation) may be greater. To provide an indication of full-day exposure to PUF, the scenario was extended to 24 h, resulting in an MoS of 2 or greater for all substances.

The increase of the ΣMDA concentration to 15 mg.kg⁻¹ resulted in an MOS of 6.4, and conducting the RA solely using the 4,4’-MDA isomer resulted in an MOS of 9.0, demonstrating that the CertiPUR diamine limits remain adequately protective.

Summary and conclusion

The sensitization and non-cancer endpoints lifetime risk for consumers by sleeping on a PUF mattress was found to be very low. For the presumed carcinogens (2,4-TDA and 4,4’-MDA), the CertiPUR voluntary certification program requires their concentrations to be ≤5 mg.kg⁻¹. For 2,4-TDA, this limit was found to be protective (MOS = 6.6). Our results showed that ΣMDA ≤5 mg.kg⁻¹ led to a MOS of 19. It was also shown that the limit of 4,4’-MDA ≤5 mg.kg⁻¹ was conservative (MOS = 9). However, pending further studies on 2,2’- and 2,4’-MDA, it would be scientifically more correct to use a ΣMDA limit ≤15 mg.kg⁻¹ (MOS = 6.4).³

Supplemental Material

Supplemental Material - Risk assessment - based verification of the CertiPUR ^TM limit values for toluene diamine and methylene dianiline in flexible polyurethane foam

Supplemental Material for Risk assessment - based verification of the CertiPUR^TM limit values for toluene diamine and methylene dianiline in flexible polyurethane foam by Patrick De Kort, Elke Jensen, Mark W Spence, and Patrick M Plehiers in Toxicology and Industrial Health.

Footnotes

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Patrick de Kort is employed by EUROPUR. The trade association is funded by producers of flexible polyurethane foam. Mark W Spence and Patrick M Plehiers are employed by the International Isocyanate Institute, Inc. The Institute is funded by producers of TDI and MDI. Elke Jensen is employed by Dow, a producer of raw material for polyurethane.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the International Isocyanate Institute, Inc. and EUROPUR. Any opinions expressed are those of the authors, not necessarily of the International Isocyanate Institute Inc., EUROPUR, or their member companies.

Ethical statement

ORCID iDs

Patrick De Kort

Mark W Spence

Patrick M Plehiers

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Arnold

Collins

Graham

, et al. (2012) Risk assessment for consumer exposure to toluene diisocyanate (TDI) derived from polyurethane flexible foam. Regulatory Toxicology and Pharmacology 64(3): 504–515.

ATSDR (Agency for Toxic Substances and Diseases Registry) (1998) Toxicological Profile for Methylenedianiline. Atlanta, GA, USA: ATSDR. Available at: https://www.atsdr.cdc.gov/ToxProfiles/tp122.pdf (accessed 01 May 2023).

Basketter

Safford

(2016) Skin sensitization quantitative risk assessment: a review of underlying assumptions. Regulatory Toxicology and Pharmacology 74: 105–116.

Basketter

Natsch

Ellis

, et al. (2018) Interspecies assessment factors and skin sensitization risk assessment. Regulatory Toxicology and Pharmacology 97: 186–188.

BAuA (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin) (2008). European Union Risk Assessment Report 4-methyl-m-phenylenediamine (toluene-2,4-diamine). Dortmund, Germany: BAuA. Available at: https://echa.europa.eu/documents/10162/c666021a-4985-488e-a066-071cf9513129 (accessed 01 May 2023).

BAuA (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin) (2010) Expositions-risiko-Begründung zu 4,4'-Methylendianilin [Translation: Justification of Exposure Risk of 4,4'-Methylene Dianiline. Dortmund, Germany: BAuA. Available at: https://www.baua.de/DE/Angebote/Regelwerk/TRGS/pdf/910/910-4-4-methylendianilin.pdf (accessed 12 October 2023).

Brereton

Emanuel

Lomax

, et al. (2019) Polyurethanes. In: Ullmann's Encyclopedia of Industrial Chemistry. Weinheim, Germany: Wiley-VCH Verlag. https://onlinelibrary.wiley.com/doi/10.1002/14356007.a21_665.pub3 (accessed 02 May 2023).

Cardy

(1979) Carcinogenicity and chronic toxicity of 2,4-toluenediamine in F344 rats. Journal of the National Cancer Institute 62(4): 1107–1116.

DECOS

(Dutch Expert Committee on Occupational Safety of the Health Council Of The Netherlands) (2015) 4,4’-Methylenedianiline. Den Haag, NL: Health Council of the Netherlands. https://www.gezondheidsraad.nl/documenten/adviezen/2015/11/17/44-methyleendianiline (accessed 13 November 2023).

10.

ECB (European Chemicals Bureau) (2001). European Union Risk Assessment Report 4,4’-methylenedianiline. Brussels Belgium: European Commission Joint Research Centre. https://echa.europa.eu/documents/10162/ba567fa3-282c-4b37-bcba-f82fd9ae52bc (accessed 9 May 2023).

11.

ECHA (European Chemicals Agency) (2012) Guidance on Information Requirements and Chemical Safety Assessment Chapter R.8: Characterisation of Dose [Concentration]-Response For Human Health. Helsinki, Finland: ECHA. https://echa.europa.eu/documents/10162/17224/information_requirements_r8_en.pdf/(accessed 09 June 2023).

12.

ECHA (European Chemicals Agency) (2015) Application for Authorisation: Establishing a Reference Dose Response Relationship for Carcinogenicity of Technical MDA. Helsinki, Finland: ECHA. https://echa.europa.eu/documents/10162/13641/rac_32_notes_moca_en.pdf/0bd6ee3a-8227-4fe9-82e7-d409461c0940 (accessed 30 April 2023).

13.

ECHA (European Chemicals Agency) (2023a) Substance infocard for 2-methyl-m-phenylenediamine. https://echa.europa.eu/substance-information/-/substanceinfo/100.011.376 (accessed 9 May 2023).

14.

ECHA (European Chemicals Agency) (2023b) Substance infocard for 4-methyl-m-phenylenediamine. https://echa.europa.eu/substance-information/-/substanceinfo/100.002.231 (accessed 9 May 2023).

15.

ECHA (European Chemicals Agency) (2023c) Substance infocard for 4,4'-methylenedianiline. https://echa.europa.eu/substance-information/-/substanceinfo/100.002.705 (accessed 9 May 2023).

16.

ECHA (European Chemicals Agency) (2023d) Registration dossier for 4,4’-methylene dianiline. https://echa.europa.eu/registration-dossier/-/registered-dossier/15201/7/5/1 (accessed 30 April 2023).

17.

El-Hawari

Stoltz

Czarnecki

, et al. (1986) Dermal Absorption of ¹⁴C-Labeled 4,4'-Methylenedianiline (MDA) in Rats, guinea Pigs, and Monkeys. Washington, DC, USA: Environmental Protection Agency. EPA-560-56/011.

18.

EN 16516 (2017) Construction Products: Assessment of Release of Dangerous Substances. Determination of Emissions into Indoor Air. Brussels, Belgium: European Center for Standardization (CEN), 16516.

19.

EPA (US Environmental Protection Agency) (1991) Regional Guidance on Handling Chemical Concentration Data Near the Detection Limit in Risk Assessments. Philadelphia, PA, USA: Environmental Protection Agency. https://www.epa.gov/risk/regional-guidance-handling-chemical-concentration-data-near-detection-limit-risk-assessments (accessed 01 May 2023).

20.

EPA (US Environmental Protection Agency) (2005a) Provisional Peer Reviewed Toxicity Values for 2,6-Toluenediamine (CASRN 823-40-5). Cincinnati, OH, USA: EPA. https://cfpub.epa.gov/ncea/pprtv/documents/Toluenediamine26.pdf (accessed 29 April 2023).

21.

EPA (US Environmental Protection Agency) (2005b) Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens. Cincinnati, OH, USA: EPA. EPA/630/R-03/003F March 2005) . https://www3.epa.gov/airtoxics/childrens_supplement_final.pdf (accessed 28 June 2023)

22.

EPA (US Environmental Protection Agency) (2011) Exposure Factor Handbook. 2011 Edition (Final Report, EPA/600/R-09/052F). Washington, DC, USA: EPA. Available at: https://www.epa.gov/expobox/about-exposure-factors-handbook (accessed 5 January 2023).

23.

EPA (US Environmental Protection Agency) (2012) Standard Operating Procedures for Residential Pesticide Exposure Assessment: 183-184. Washington, DC, USA: EPA. https://www.epa.gov/sites/default/files/2015-08/documents/usepa-opp-hed_residential_sops_oct2012.pdf (accessed 29 April 2023).

24.

EU Commission (2008). EU Regulation (EC) 1272/2008. Annex VI: 647. https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32008R1272 (accessed 09 June 2023).

25.

EUROPUR (European Association of Flexible PU Foam Blocks Manufacturers) (2021) CertiPUR^TM label for polyurethane foams: 16. Available at: https://europur.org/wp-content/uploads/2022/04/CertiPUR_Technical_Paper.pdf (accessed 29 April 2023).

26.

Foley

(2023) When should you replace your mattress? Available at: https://www.sleepfoundation.org/mattress-information/when-should-you-replace-your-mattress (accessed 25 March 2023).

27.

Gerberick

Ryan

Dearman

, et al. (2007) Local lymph node assay (LLNA) for detection of sensitization capacity of chemicals. Methods 41(1): 54–60.

28.

Hall

van Parys

FAL

Young

(2001) Toluenediamine in polyurethane foams: hazard or artefact? Plastics, Rubber and Composites 30(9): 426–433.

29.

Hillier

Schupp

Carney

(2003) An investigation into VOC emissions from polyurethane flexible foam mattresses. Cellular Polymers 22(4): 237–259.

30.

Jennings

Hammerstrom

Coleman Adkins

, et al. (1987) Methods for Assessing Exposure to Chemical Substances. Report No. EPA 560/5-85-007. Washington, DC, USA: US Environmental Protection Agency, Vol 7. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1007I8Y.txt (accessed 09 June 2023).

31.

Karlsson

(2023) Sampling chamber with minimal wall surface for simultaneous emission testing of diisocyanates and diamines from polyurethane products. Toxicology and Industrial Health 39(6): 313–324.

32.

Karlsson

Spence

Plehiers

(2023) Emission and time-resolved migration rates of aromatic diamines from two flexible polyurethane foams. Toxicology and Industrial Health 39(6): 298–312.

33.

Krämer-Lucas

Raffel

Seel

(2001) Fate of MDA in Flexible Moulded Foams. Manchester, UK: International Isocyanate Institute. https://www.bl.uk/onlinegallery/

34.

Lamb

Huff

Haseman

, et al. (1986) Carcinogenesis studies of 4,4'-methylenedianiline dihydrochloride given in drinking water to F344/N rats and B6C3F1 mice. Journal of Toxicology and Environmental Health 18: 325–337.

35.

Marzulli

Anjo

Maibach

(1981) In vivo skin penetration studies of 2,4-toluenediamine, 2,4-diaminoanisole, 2-nitro-p-phenylenediamine, p-dioxane and N-nitrosodiethanolamine in cosmetics. Food and Cosmetics Toxicology 19: 743–747.

36.

NCI (National Cancer Institute) (1979) Bioassay of 2,4-diaminotoluene for Possible Carcinogenicity. National Cancer Institute Carcinogenesis Technical Report Series No.162. Bethesda, MD, USA: National Institutes of Health. https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr162.pdf?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=tr162 (accessed 30 April 2023).

37.

NCI (National Cancer Institute) (1980) Bioassay of 2,6-toluenediamine Dihydrochloride for Possible Carcinogenicity. National Cancer Institute Carcinogenesis Report Series No.200. Bethesda, MD, USA: National Institutes of Health. https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr200.pdf?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=tr200 (accessed 30 April 2023).

38.

NRC (National Research Council) (1983) Risk Assessment in the Federal Government: Managing the Process. Washington, DC, USA: National Academies Press.

39.

NTP (National Toxicology Program) (1983) Carcinogenesis Studies of 4,4'-methylenedianiline Dihydrochloride (CAS No. 13552-44-8) in F344/N Rats and B6C3F1 Mice (Drinking Water Studies). NIH Publication 83-2504; NTP-81-143. Bethesda, MD, USA: National Institutes of Health. https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr248.pdf?utm_source=direct&utm_medium=prod&utm_campaign=ntpgolinks&utm_term=tr248 (accessed 30 April 2023).

40.

OECD (Organisation for Economic Cooperation and Development) (2009) OECD Series on Emission Scenario Documents Number 3. Emission Scenario Document of Plastic Additives: 136-141. Paris, France: OECD. https://www.oecd-ilibrary.org/environment/plastic-additives_9789264221291-en (accessed 27 April 2023).

41.

OECD (Organisation for Economic Cooperation and Development) (2016a) Test no. 473: in vitro mammalian chromosomal aberration test. In: OECD Guidelines for the Testing of Chemicals, Section 4. Paris, France: OECD Publishing.

42.

OECD (Organisation for Economic Cooperation and Development) (2016b) Test No. 476: in vitro mammalian cell gene mutation tests using the Hprt and Xprt genes. In: OECD Guidelines for the Testing of Chemicals, Section 4. Paris, France: OECD Publishing.

43.

OECD (Organisation for Economic Cooperation and Development) (2020) Test guideline no. 471: bacterial reverse mutation test. In: OECD Guidelines for the Testing of Chemicals, Section 4. Paris, France: OECD Publishing.

44.

Oppl

(2014) New European VOC testing method CEN/TS 16516 and CE marking of construction products. Gefahrstoffe – Reinhaltung der Luft 74(3): 62–68.

45.

Roberts

Api

Patlewicz

, et al. (2016) Chemical applicability domain fo the local lymph node assay (LLNA) for skin sensitization potency. Part 1, Underlying physical organic chemistry principles and the extent to which they are represented in the LLNA validation data set. Regulatory Toxicology and Pharmacology 80: 247–254.

46.

Safford

(2008) The dermal sensitisation threshold - a TTC approach for allergic contact dermatitis. Regulatory Toxicology and Pharmacology 51(2): 195–200.

47.

Schupp

Plehiers

(2022) Absorption, distribution, metabolism, and excretion of methylene diphenyl diisocyanate and toluene diisocyanate: many similarities and few differences. Toxicology and Industrial Health 38(9): 500–528.

48.

Szycher

(2012) Szycher’s Handbook of Polyuethanes. Second Edition. Boca Raton, FL, USA: CRC Press, 193.

49.

Te Biesebeek

Nijkamp

Bokkers

BGH

, et al. (2014) General Fact Sheet: General Default Parameters for Estimating Consumer Exposure: Updated version 2014. RIVM Report 090013003/2014. Bilthoven, The Netherlands: National Institute for Public Health and the Environment. https://www.ncbi.nlm.nih.gov/books/NBK563027/(accessed 29 April 2023).

50.

UN (United Nations) (2021) Globally harmonized system of classification and labelling of chemicals (GHS). Ninth Revised. edition. New York, NY, USA and Geneva, Switzerland: United Nations, 156–159. https://unece.org/sites/default/files/2021-09/GHS_Rev9E_0.pdf (accessed 01 May 2023).

51.

Unterberger-Henig

(2022) Comparative evaluation of three methylene dianiline isomers in the bacterial reverse mutation assay, the in vitro gene mutation test, and the in vitro chromosomal aberration test. Toxicology and Industrial Health 38(9): 529–543.

52.

Vangronsveld

Berckmans

Spence

(2013) Toluene diisocyanate emission to air and migration to a surface from a flexible polyurethane foam. Annals of Occupational Hygiene 57(5): 650–661.

53.

Vanoirbeek

JAJ

De Vooght

Synhaeve

, et al. (2009) Is toluene diamine a sensitizer and is there cross-reactivity between toluene diamine and toluene diisocyanate? Toxicological Sciences 109(2): 256–264.

54.

Weisburger

Krishna Murthy

Lilja

, et al. (1984) Neoplastic response of F344 rats and B6C3F1 mice to the polymer and dyestuff intermediates 4,4'-methylene-bis(N,N-dimethyl)- benzenamine, 4,4'-oxydianiline, and 4,4'-methylenedianiline. Journal of the National Cancer Institute 72(6): 1457–1463.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

Risk assessment—based verification of the CertiPUR TM limit values for toluene diamine and methylene dianiline in flexible polyurethane foam

Abstract

Keywords

Introduction

Methods

Substances and hazard classification

Dose-response assessment and derivation of toxicological benchmarks

Carcinogenicity

Repeated dose toxicity

Skin sensitization (local effects)

Experimental data and exposure assessment

Emission rates (inhalation exposure)

Migration rates (dermal exposure)

Risk assessment

Exposure scenario

Risk characterization

Sensitivity analyses

Results

Derivation of toxicological benchmarks

Carcinogenicity

4,4'-MDA

2,4-TDA

Repeated dose toxicity

2,6-TDA

Sensitization

TDA

MDA

Exposure estimation

Emission rates (inhalation exposure)

ΣMDA

TDA

Migration rates (dermal exposure)

ΣMDA

TDA

Risk characterization

Carcinogenicity

2,4-TDA

ΣMDA

Repeated dose toxicity

2,6-TDA

Sensitization

Sensitivity analyses

Discussion

Summary and conclusion

Supplemental Material

Supplemental Material - Risk assessment - based verification of the CertiPUR TM limit values for toluene diamine and methylene dianiline in flexible polyurethane foam

Footnotes

Declaration of conflicting interests

Funding

Ethical statement

ORCID iDs

Supplemental Material

Notes

References

Supplementary Material

Supplemental Material - Risk assessment - based verification of the CertiPUR ^TM limit values for toluene diamine and methylene dianiline in flexible polyurethane foam