Technical Field

The present disclosure is directed to halogen-based flame retardants, their preparation and their use in polymers to impart flame retardancy.

Background

Commercial halogen-based flame retardants are mainly based on bromine. However, such bromine-carbon bond containing compounds are being banned due to a large negative sustainability impact. More specifically, these bromine based flame retardants are often persistent, bioaccumulative and toxic.

Chlorinated flame retardants are known in the art but are also very persistent and very bioaccumulative. Therefore, there is a demand for flame retardants that are desirable in view of sustainability, i.e. are biodegradable, that are not persistent or bioaccumulative and non-toxic, but still provide the required flame retardancy.

In the prior art several attempts were made to develop more sustainable flame retardants.

In US 3932541 brominated pentaerithritols are prepared by reacting pentaerithritol with HBr or HCI in the liquid phase using an aliphatic carboxylic acid as catalyst. Because of their relative low halogen-density their flame retardancy will be inadequate. Also, they will contain carbon-halogen bonds, most likely resulting in non-biodegradable compounds. In Journ. of Pol. Science, Part A: Polymer Chemistry, Vol 34, 1455-1464 (1996), of Hongsoo Park, Janghyoun Keun and Kisay Lee, a two component polyurethane flame- retardant coating was prepared by blending chlorine containing modified polyester and polyisocyanate. Said chlorine-containing modified polyester was prepared by polycondensating dichloroacetic acid with butane diol, trimethylolpropane and adipic acid. This concept was described by Hongsoo Park, Hyun-Sik Hahm and Eun-Kyung Park in Journ. of Appl. Polym. Science, Vol 61 , 421 -429 (1996) using bromine instead of chlorine containing modified polyester and polyisocyanate. In Journ. of Islamic Academy of Sciences 4:3, 200-202, 1991 of A.M. Hassan a flame retardant was prepared for paper and paper board. Said flame retardant was a polyurethane of halogenated fatty acid and toluene diisocyanate. This was prepared by mixing coconut oil and castor oil, glycerol, trichloroacetic acid and/or bromoacetic acid. This mixture was then reacted with toluene isocyanate.

In US2003/009003, B. Kohler, K. Horn and R.-V. Meyer describe the preparation of polyether carbonates produced with the alpha-halocarboxylic (e.g. monochloroacetic acid) esters of polyols as branching agent, whereby the branching agent is incorporated in the polycarbonate by means of ether bridges.

In general, polymerized flame retardants (such as those described in the Hassan and Park references above) are not desired because they may compromise the compatibility with the polymers to be rendered flame retardant and may detrimentally affect the chemical and physical properties of the polymer.

In Journ. of Islamic Academy of Sciences 5:3, 149-152, 1992 of A.M. Hassan and E.M. Sadek fire-resistant compounds from diglyceryl borate trichloroacetate were prepared. Boron- containing compounds are less desired as flame retardant for their toxicity and their price. There is still a need in the art for biodegradable flame retardants that have adequate flame retardant properties.

Summary

The disclosure is directed to a process for the preparation of a boron-free halogen- based flame retardant which is prepared by reacting a halogeno-alkylcarboxylic acid containing at least two halogen atoms with an unsaturated and/or hydroxyl containing carrier molecule having at least three bonding sites for the halogeno-alkylcarboxylic acid per carrier molecule so as to form a biodegradable boron-free halogen-based flame retardant.

The carrier molecule may comprise triallyl isocyanurate (TAIC), glycerol (Glyc), pentaerythritol (PTT), trimethylol propane (TMP), triallyl cyanurate (TAC), or trimethylol ethane (TME) according to the following structures I, II, III, IV, V and VI.

(I) (II) (III) (IV)

The halogeno alkyl carboxylic acid may be di- or tribromoacetic acid or most preferably di- or trichloroacetic acid.

The boron-free halogen-based flame retardant preferably has a molecular weight of at most 7000, more preferably at most 5000.

Preferably, the resulting flame retardant comprises 1 to 4 carrier molecules per compound.

The resulting flame-retardant preferably has at least 4 halogen atoms per molecule, preferably between 6-12 halogen atoms per molecule so as to ensure satisfactory flame retardancy. Preferably, the flame-retardant according to the present invention has a halogen content of at least 25wt%, more preferably at least 30wt%, most preferably at least 40wt%, and preferably lower than 90wt%, based on the total weight of the flame-retardant.

The disclosure is further directed to a biodegradable halogen-based flame retardant obtainable by the process according to the disclosure comprising an unsaturated and/or hydroxyl containing carrier unit and a halogeno-alkylcarboxylate unit. Said unsaturated and/or hydroxyl-containing carrier unit optimally originates from one of the following structures I, II, III, IV, V and VI:

(II) (III) (IV)

Said halogen-based flame retardant comprises a carrier molecule unit that has been bound to at least two halogeno-alkylcarboxylic acid residues, and preferably at least three halogeno-alkylcarboxylic acid residues.

The disclosure is further directed to the use of the halogen-based flame retardant according to the disclosure to render a polymer flame retardant and to a method for the preparation of a flame-retardant polymer wherein the flame retardant according to the present invention is blended into a polymer, optionally together with a synergist compound (e.g. antimony(lll) oxide, Perkadox® BC (i.e. dicumyl peroxide) or Perkadox® 30 (i.e. 2,3-dimethyl-2,3-diphenyl butane)). Synergist compounds, sometimes denoted as flame-retardant auxiliaries, are compounds which enhance the efficiency of the flame-retardant and these compounds are commonly known in the art (see e.g. US 2016/0340498). Suitable polymer classes include but are not limited to polyolefins (e.g. polypropylene, polyethylene), styrenics (e.g. expanded polystyrene, extruded polystyrene, high impact polystyrene), polyesters, polycarbonates, polyamides, polyacrylates, polyisocyanurate, epoxies/phenolics, natural and synthetic rubbers (eg EPDM), polyethylene terephthalate, polyvinyl chloride, starch and cellulose based polymers, and blends thereof. The most preferred polymers are polyolefins, styrenics and polyesters. Detailed description

The various aspects of the present disclosure will be elucidated further below.

As indicated above, the disclosure is directed to a process for the preparation of a boron-free halogen-based flame retardant which is prepared by reacting halogeno- alkylcarboxylic acid containing at least two halogen atoms with an unsaturated and/or hydroxyl containing carrier molecule containing at least three bonding sites for the halogeno-alkylcarboxylic acid per carrier molecule so as to form a biodegradable boron-free halogen-based flame retardant.

With bonding site is meant here a double carbon-carbon bond or a hydroxyl-group which can react with the halogeno-carboxylic acid under formation of an ester bond.

The wording "unsaturated and/or hydroxyl containing carrier molecule containing at least three bonding sites" therefore denotes a carrier molecule which comprises at least three unsaturated bonds (i.e. carbon-carbon double bonds) or a carrier molecule which comprises at least three hydroxyl groups, or a carrier molecule which comprises unsaturated bonds and hydroxyl groups.

With the term biodegradable is meant that when subjecting the flame retardant to a closed bottle test according the Test guidelines of OECD 1992 (this is Method 301 D: Closed Bottle of the OECD Guideline for testing of Chemicals, as adopted by the Council on 17 ^th July 1992, "Ready Biodegradability"), the flame retardant can be categorized as being not vPvB (Very Persistent, Very Bioaccumulative) and also not PBT (Persistent, Bioaccumulative, Toxic). The categorization of biodegradability test data, either generated using the above-mentioned test or data taken from literature, is performed by comparing said data with the criteria as set out in REACH Annex XIII to assess the Persistence of substances (ECHA Technical Guidance Document (2017) Guidance on information requirement and chemical safety assessment Chapter R.1 1 ; PBT/vPvB assessment version 3.0).

With the term boron-free is meant with less than 5 wt% of boron, preferably less than 2 wt% of boron, more preferably less than 1 wt% of boron, and most preferably less than 0.1 wt% of boron. The preparation reaction may be conducted under the following conditions: The halogeno-alkylcarboxylic acid, the carrier molecule, and optionally catalyst and optionally solvent are combined and reacted at the given temperature for 1 -24 hours, more preferably 3-12 hours. The reaction mixture is then cooled, and the reaction product is subsequently separated from the reaction mixture optionally purified and dried. The solvent used may be an organic solvent such as toluene, xylene, mesitylene or any other suitable solvent. It can be used in combination with a Dean Stark apparatus to remove the formed water from the reaction mixture; in the latter, the reaction is performed under refluxing conditions. The reaction temperature usually ranges from 80-220 °C, more preferably between 100-180 °C.

As is common general knowledge for a skilled person, a suitable catalyst for this reaction may be an acid catalyst such as p-toluene sulfonic acid, triflic acid, any strong organic acid or strongly acidic (co-)polymer or resin (e.g. Nafion® PFSA, Amberlyst® 15 or Dowex® 50WX products). The separation of the product from the reaction mixture may be conducted by extraction such as extraction from a water/alkali mixture. The product may be further purified by washing with e.g. aqueous sodium bicarbonate, and dried. Preferably, the molecular weight (Mw, mass average molar mass, which is calculated by the formula wherein N, is the number of molecules of molecular mass M, as determined with NMR and when relevant with GPC, Gel Permeation Chromatography) of the resulting flame retardant is at most 7000, more preferably at most 5000. More particularly, in case the flame retardant merely contains one carrier molecule, the chemical structure, and thus the M _w , can be confirmed by NMR analysis. In case the flame retardant contains more units, GPC is used to determine the M^.

If the flame retardant becomes too bulky the compatibility with the polymer to be rendered flame retardant will be compromised too much. Further, the density of the halogen atoms within the compound may become too low and the flame-retardant properties will be detrimentally affected. Preferably, the resulting flame retardant comprises 1 to 4 carrier molecule residues per compound. This is to ensure compatibility with the polymer to be rendered flame retardant and to avoid that the flame retardant detrimentally affects the chemical and physical properties of the polymer to be rendered flame retardant. With the term "residue" is meant the unit in its form after reaction via its bonding sites.

It is important that the carrier molecule can bind a large amount of halogens via ester bonds. In principle any compound having multiple (i.e. in total at least three) hydroxy groups and/or unsaturated bonds (i.e. double carbon-carbon bonds) and not being sterically hindered to form ester bonds, may be suitable as carrier molecule as long as it does not have toxic groups. Preferred carrier molecules may comprise triallyl isocyanurate, glycerol, pentaerythritol, trimethylol propane, triallyl cyanurate (TAC), or trimethylol ethane (TME) according to the following structures I, II, III, IV, V and VI.

These types of carrier molecules provide numerous bonding sites for the halogeno alkyl carboxylic acids via either their unsaturated (carbon-carbon double) bonds or their OH-groups.

The halogeno alkyl carboxylic acid according to the present invention contains at least two halogen atoms. Suitable halogeno alkyl carboxylic acids comprise di- or tribromoacetic acid, di- or trichloroacetic acid, but also higher halogeno alkyl carboxylic acids such as halogeno propionic acid, halogenobutyric acid and halogeno valeric acid. It goes without saying that these higher alkyl carboxylic acids may be multiple halogenated, such as di-, tri-, tetra,- penta-, septa-, octa-, nona-, and decahalogenated. Preferred are the lower halogeno alkyl carboxylic acids such as di- or tribromoacetic acid or most preferably di- or trichloroacetic acid. It was found that compounds prepared with trichloroacetic acid provided the best flame retardancy, but the other compounds disclosed herein also provide good flame retardancy as well. The resulting flame-retardant preferably has at least 4 halogen atoms, preferably between 6-12 halogen atoms per molecule so as to ensure satisfactory flame retardancy. More particularly, the flame-retardant preferably has a halogen content of at least 25wt%, meaning that the total weight of the halogens in the flame-retardant is 25% of the total weight of the flame-retardant. More preferably, the flame-retardant has a halogen content of at least 30wt% and most preferably, of at least 40wt%, based on the total weight of the flame-retardant. Preferably, the flame-retardant has a halogen content which is lower than 90wt%, based on the total weight of the flame-retardant. The disclosure is further directed to a boron-free halogen-based flame retardant obtainable by the process according to the disclosure comprising an unsaturated and/or hydroxyl residue-containing carrier unit and a halogeno-alkylcarboxylate unit.

Said halogen-based flame retardant preferably comprises a carrier molecule unit that has been bound to at least three halogeno-alkylcarboxylic acid residues. Preferably the molecular weight, Mw, of the halogen-based flame retardant is at most 7000, more preferably at most 5000. Preferably, the M _w of the halogen-based flame retardant is at least 200, preferably at least 300, most preferably at least 400.

Said the unsaturated and/or hydroxyl-containing carrier unit optimally originates from one of the following structures I, II, III, IV, V and VI:

(I) (II) (ill) (iv)

The disclosure is further directed to the use of the halogen-based flame retardant according to the disclosure to render a polymer flame retardant and to a method for the preparation of a flame retardant polymer wherein the flame retardant according to the disclosure is blended into a polymer, optionally together with a synergist compound, (e.g. antimony(lll) oxide Sb ₂ 0 ₃ , Perkadox® BC or Perkadox® 30), typically in ranges of 1 -25 wt%. The disclosure is furthermore directed to a method for the preparation of a flame retardant polymer wherein the flame retardant is Suitable polymer classes include but are not limited to polyolefins (e.g. polypropylene, polyethylene), styrenics (e.g. expanded polystyrene, extruded polystyrene, high impact polystyrene), polyesters, polycarbonates, polyamides, polyacrylates, polyisocyanurate, epoxies/phenolics, natural and synthetic rubbers (eg EPDM), polyethylene terephthalate, polyvinyl chloride, starch and cellulose based polymers, and blends thereof. The most preferred polymers are polyolefins, styrenics and polyesters.

The flame retardant can be blended with the polymer in ranges of 1 -25 wt%, preferably 2-20 wt%, based on the total weight of the polymer. Optionally, also other conventional additives may be blended with the polymer such as other flame retardants, stabilizers, lubricants, UV absorbers, and clarifiers etcetera.

Biodegradability tests (carried out in according to the guidelines mentioned above) show that the flame retardants according to the present invention are classified as not being Very Persistent, Very Biocumulative (vPvB) and also not Persistent, bioaccumulative, Toxic (PBT). It is noted that various elements of the present disclosure, including but not limited to preferred ranges for the various parameters, can be combined unless they are mutually exclusive. The disclosure will be elucidated by the following examples without being limited thereto or thereby.

Examples

Example 1 (Preparation of flame retardant with pentaerythritol as carrier molecule: PTT-TCA)

A 250 mL three-neck round-bottom flask equipped with a thermometer, a magnetic stirrer and a Dean Stark apparatus (filled with toluene) connected to a water condenser, was loaded with 5.0 g (0.036 mol, 1 .0 equivalent) of pentaerythritol (PTT), 30.0 g (0.182 mol, 5.0 equivalents) of trichloroacetic acid, 200 mL of toluene and 3.0 g (0.016 mol, 0.4 equivalent) of para-toluenesulfonic acid then purged with nitrogen gas. The mixture was heated to reflux and stirred at this temperature for 6h then cooled down and allowed to stand overnight at room temperature. The mixture was poured into a mixture of 200 g of ice and 200 mL of aqueous sodium hydroxide 10% solution. The layers were separated; the organic layer was washed with three times 100 mL of aqueous sodium bicarbonate 5% solution then 200 mL of water, dried over anhydrous sodium sulphate. After filtration, toluene was removed on the rotating evaporator (15 mbar, 120°C, 20 min) to afford 16.8 g of white crystalline material (Yield 63%, assay >97%).

¹ H NMR (400 MHz, CDCI ₃ , 25°C) δ (ppm): 4.53 (singlet, 8H); ¹³ C NMR (100 MHz, CDCIs, 25°C) δ (ppm): 43.7, 65.3, 88.9, 161 .0.

Example 2 (Preparation of flame retardant with pentaerythritol as carrier molecule: PTT-MCA)

A 500 mL three necks round-bottom flask equipped with a Dean Stark apparatus (filled with extra xylene) with a water condenser, a thermometer and a magnetic stirrer was loaded with 24.0 g (0.176 mol, 1 .0 equiv.) of PTT, 100.0 g (1.058 mol, 6.0 equiv.) of MCA, 276.0 g of xylene and 3.4 g (0.018 mol, 0.1 equiv.) of PTSA then purged with nitrogen gas. The mixture was heated to reflux and stirred at this temperature for 14h then cooled down to RT. The xylene was removed under reduced pressure by means of a rotating evaporator (30 mbar, 60°C, 30 min) to afford a highly viscous yellow- brown oil. This oil was dissolved in 120 mL of dichloromethane; the obtained solution was washed with seven times 90 mL of aqueous sodium bicarbonate 5% solution then 200 mL of water, dried over anhydrous sodium sulphate. After filtration, the mixture was concentrated under reduced pressure by means of a rotating evaporator (30 mbar, 60°C, 30 min) to afford 38 g of pale yellow powder (Yield 49%, assay >98%).

¹ H NMR (400 MHz, CDCI ₃ , 25°C) δ (ppm): 4.09 (singlet, 8H), 4.28 (singlet, 8H); ¹³ C NMR (100 MHz, CDCI ₃ , 25°C) δ (ppm): 40.5, 42.6, 63.3, 166.7.

Example 3 (Preparation of flame retardant with triallyl isocyanurate as carrier molecule: TAIC-TCA)

A 100 mL three-neck round-bottom flask equipped with a water condenser, a thermometer and a magnetic stirrer was loaded with 7.5 g of TAIC (0.029 mol, 1.0 equivalent), 73.0 g (0.442 mol, 15.0 equivalents) of trichloroacetic acid and 1.0 g (0.005 mol, 0.18 equivalent) of para-toluenesulfonic acid then purged with nitrogen gas. The mixture was heated to 140 °C and stirred at this temperature for 6h then cooled down and allowed to stand overnight at room temperature. Some of the TCA was removed from the mixture by means of a rotating evaporator (10 mbar, 120 °C, 20 min) then the obtained material was dissolved in 100 mL of diethyl ether. The obtained solution was washed with five times 25 mL of aqueous sodium bicarbonate 5% solution, 10 mL of water, dried on anhydrous sodium sulphate then filtrated. The diethyl ether was removed on the rotating evaporator (15 mbar, 120°C, 20 min) to afford 13.7 g of solid brown material (Yield 64%, assay >97%).

¹ H NMR (600 MHz, CDCI ₃ , 25°C) δ (ppm): 1.45 (9H, multiplet), 3.95 (3H, multiplet), 4.38 (3H, multiplet), 5.34 (3H, multiplet); ¹³ C NMR (150 MHz, CDCI ₃ , 25°C) δ (ppm): 17.16, 17.20, 17.24, 46.32, 46.41 , 73.98, 74.10, 74.22, 89.7, 148.54, 148.60, 148.64, 161 .72, 161 .74, 161.75. Example 5 (Preparation of flame retardant with triallyl isocyanurate as carrier molecule: TAIC-MCA)

The TAIC-MCA flame retardant was prepared from triallyl isocyanurate and MCA (monochloroacetic acid) using the procedure given as Example 3 with extended reaction time (12h).

¹ H NMR (400 MHz, CDCI ₃ , 25°C) δ (ppm): 1.33 & 1 .34 (9H, 2 doublets), 3.91 (3H, multiplet), 3.99 (6H, multiplet), 4.23 (3H, multiplet), 5.36 (3H, multiplet).

Example 6 (Preparation of flame retardant with glycerol as carrier molecule: Glyc-TCA)

A 250 mL three-neck round-bottom flask equipped with a thermometer, a magnetic stirrer and a Dean Stark apparatus (filled with toluene) connected to a water condenser, was loaded with 6.0 g (0.064 mol, 1 .0 equivalent) of glycerol (Glyc), 54.0 g (0.327 mol, 5.1 equivalents) of trichloroacetic acid, 180 mL of toluene and 2.4 g (0.012 mol, 0.2 equivalent) of para-toluenesulfonic acid then purged with nitrogen gas. The mixture was heated to reflux and stirred at this temperature for 12h then cooled down and allowed to stand overnight at room temperature. The mixture was poured into a mixture of 180 g of ice and 200 g of aqueous sodium hydroxide 10% solution. The layers were separated; the organic layer was washed with twice 100 mL of aqueous sodium bicarbonate 5% solution then 200 mL of water, dried over anhydrous sodium sulphate. After filtration, toluene was removed on the rotating evaporator (15 mbar, 120°C, 20 min) to afford 30.1 g of slightly yellow oily liquid material (Yield 85%, assay >95%). ¹ H NMR (400 MHz, CDCI ₃ , 25°C) δ (ppm): 4.62 (doublet of doublet, 2H), 4.77 (doublet of doublet, 2H), 5.60 (multiplet, 1 H); ¹³ C NMR (100 MHz, CDCI ₃ , 25°C) δ (ppm): 65.0, 72.7, 88.9, 161.0, 161 .3.

Example 7 (Preparation of flame retardant with trimethylol propane as carrier molecule: TMP-TCA)

The molecule TMP-TCA was prepared from trimethylol propane and TCA, applying the procedure given in Example 1 .

¹ H NMR (400 MHz, CDCI ₃ , 25°C) δ (ppm): 1 .02 (triplet, 3H), 1 .60 (quadruplet, 2H), 4.39 (singlet, 6H); ¹³ C NMR (100 MHz, CDCI ₃ , 25°C) δ (ppm): 7.3, 23.0, 42.2, 67.5, 89.3, 161 .3. Example 8: Evaluation of flame retardancy in PolyPropylene (PP) formulations

The well-known UL 94 test was used for evaluating the flame retardancy. In this test a 12.7x1 .27 cm specimen (polypropylene comprising the flame retardant to be tested, along with a synergist compound; the thickness of the specimens can be varied between 0.7 and 3 mm; in this Example, the thickness of the specimen was 2 mm) was held with a clamp above a burner that was placed above a piece of cotton of specific height and diameter, at specific distance from the flame. If the specimen caught fire and glowing or flaming combustion reached the holding clam, it was categorized as non-retardant (NR). If it was not the case and if the cotton underneath the burner got ignited by drips from the specimen, the material was categorized as V2. If the specimen did not burn up to reaching the holding clamp, nor did it cause the cotton to ignite, then the flaming combustion time was measured. If it took up to 10 seconds for the specimen to stop burning, it was considered to fall in the V0 category, while a flaming combustion time up to 30 seconds the material was said to fall in the V1 category.

Burning tests were conducted of polypropylene comprising various amounts of flame retardants and synergist according to the disclosure prepared with methods similar to the method described in Example 3 and compared to commercial standard Armoquell® FR 930. To this end specimens of polypropylene (MFI 4) were made comprising either 4 wt% flame retardant and 2 wt% Timonox® White Star (i.e. antimony(lll) oxide, Sb ₂ 0 ₃ ), and 0.5 wt% of the thermal stabilizer Irganox® B225, 8 wt% flame retardant, 4 wt% Timonox® White Star and 0.5 wt% Irganox® B225, or 12 wt% flame retardant, 6 wt% Timonox® White Star and 0.5 wt% Irganox® B225. All results of the burning test are compiled in TABLE I:

TABLE I

(n.d.: not determined) It can be concluded that the flame retardants prepared by reaction of TCA (trichloroacetic acid) with TAIC (triallyl isocyanurate), PTT (pentaerythritol) and Glyc (glycerol) provided significant flame retardant properties to the corresponding formulated PP samples. On the contrary, the use of the compounds prepared by reaction of MCA with TAIC and PTT did not result in significantly improved flame retardancy properties of the formulated polypropylene. These experiments further show that flame retardants need to contain at least four halogen atoms per molecule to provide significant flame retardancy effect.

Example 9: Evaluation of flame retardancy in Polystyrene (PS) formulations

20 g of suspension consisting of PS (Crystal clear, BASF), flame retardant and Timonox® White Star (i.e. antimony (III) oxide, Sb ₂ 0 ₃ ) in dichloromethane were poured into a metallic Petri box (diameter 10 cm). The solvent was allowed to evaporate at room temperature in a fume cupboard for at least 2 days, affording (ca. 4 g) films. Then specimens (7.0 x 1.5 cm) were cut from the films and subjected to a modified UL94 test: the specimens were exposed to flame twice during 1 s (instead of twice 10 s in the original standard method). The flame retardancy of the specimen was rated by applying the same methodology described in Example 4 above, demonstrating a clear retardancy effect of the prepared esters. The results are shown in TABLE II. TABLE II

(n.d.: not determined)

Example 10: Biodegradability tests

Biodegradation of an organic chemical refers to the reduction in complexity of the chemical through metabolic activity of microorganisms. Under aerobic conditions, microorganisms convert organic substances in carbon dioxide, water and biomass. The theoretical oxygen demand (ThOD) of the tested compounds was calculated from their molecular formulae and molecular weights as follows:

16( 2( +0.5(// - CI - 3N) + 3S + 2.5P + 0.5 Nu - ())

ThOD _Nm (mg0 ₂ 1 mg)

MW

Provided that the oxygen concentrations in all bottles at the start of the test were equal, the amounts of oxygen consumed in the bottles were calculated as follows:

Oxygen consumption (mg/L) = Me — Mt _w j^

Mc = the mean oxygen level in the control bottle only containing inoculum, n-days after the start of the test.

Mt = the mean oxygen concentration in the bottles containing the test substance and inoculum, n-days after the start of the test.

The BOD (mg/mg) of the test substance was calculated by dividing the oxygen consumption by the concentration of the test substance in the closed bottle. The biodegradation was calculated as the ratio of the BOD to ThOD. The biodegradability of trimethylol propane (TMP), of a mixture prepared according to the procedure described by Hongsoo Park, Hyun-Sik Hahm and Eun-Kyung Park in Journ. of Appl. Polym. Science, Vol 61 , pp.421 -429 (1996), and of two flame retardants according to the present invention was evaluated using Closed Bottle tests. These tests were performed according to Test Guidelines (OECD 1992), as mentioned in the description. The bottles were incubated in the dark at temperatures ranging from 22 to 24 °C; the biodegradation was measured by following the course of the oxygen decrease in the bottles. The pH of the media was 7.3 (activated sludge) and 8.0 (river water) at the start of the test. The pH was 7.2±0.1 (activated sludge) and 7.9±0.1 (river water) at day 28.

TMP was found to be not persistent, with a biodegradation level of 66% after 21 days in both activated sludge and river water. A reaction mixture of stoichiometric amounts of TMP and 2,3-dibromopropionic acid, prepared according to the procedure described by Hongsoo Park, Hyun-Sik Hahm and Eun-Kyung Park in Journ. of Appl. Polym. Science, Vol 61 , 421 -429 (1996) on page 423 (section "Syntheses of 2,3-Dibromo Modified Polyesters - 2,3-DBP/TMP Intermediate"), was found to be persistent, as less than 40% biodegradation was measured in both activated sludge and river water after 84 days. The closed bottle test of the substance PTT-TCA, a flame-retardant according to the present invention, resulted in biodegradation of 65% within 6 weeks and 66% within 8 weeks with activated sludge and river water, respectively. The test substance should therefore be classified as not persistent. Pentaerytritol is inherently biodegradable and TCA was found to reach a biodegradation above 60% within 8 weeks in both activated sludge and river water. Biodegradation of the ester and both hydrolysis products within 60 days allows classification as not persistent. These results clearly demonstrate that the substance PTT-TCA is not a vPvB (Very Persistant, Very Bioaccumulative) nor a PBT (Persistent, Bioaccumulative, Toxic) compound.