Fire-retardant compositions

Abstract

Disclosed is a method for the use, as fire-retardant, of an aqueous composition including chitosan and at least one mineral filler, the inorganic filler being for example chosen from the group of mineral fillers in laminae, in particular chosen from the group consisting of talc, montmorillonite, saponite, sepiolite, bentonite, smectite, hectorite, kaolinite, halloysite and mica, and mixtures thereof.

Claims

1. A fire-retardant treatment method of a surface comprising the application on said surface of an aqueous composition consisting of water, chitosan, at least one mineral filler, said mineral filler being talc, and at least one acid selected from the group consisting of: acetic acid, hydrochloric acid, formic acid, L-ascorbic acid, L-glutamic acid, lactic acid, maleic acid, malic acid, and succinic acid, and mixtures thereof, and optionally at least one surfactant and/or at least one phosphorus additive.

2. The method according to claim 1, wherein the acid is acetic acid.

3. The method according to claim 1, wherein the content by weight of chitosan is between 20% and 99.99% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler, and the content by weight of mineral filler is between 0.01% and 80% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler.

4. The method according to claim 1, wherein the aqueous composition comprises from 0.01% to 25% by weight of dry extract of the mixture formed by chitosan and the mineral filler with respect to the total weight of the composition.

5. The method according to claim 1, wherein the aqueous composition has an acidic pH.

6. The method according to claim 1, wherein the number-average molar mass of chitosan is between 50,000 g/mol and 200,000 g/mol.

7. The method according to claim 1, wherein the degree of deacetylation of chitosan is between 75% and 95%.

8. The method according to claim 1, wherein the aqueous composition has a dynamic viscosity measured at 25° C. and a shear rate of 2 s.sup.−1 between 0.1 Pa.Math.s and 1000 Pa.Math.s.

9. The method according to claim 1, wherein the aqueous composition comprises from 75% to 99.99% by weight of water relative to the total weight of the composition.

10. The method of fire-retarding a surface of claim 1, the surface comprising wood, natural fibers or synthetic fibers.

11. The method of claim 10, wherein the aqueous composition is applied by dipping or spraying.

12. The method according to claim 2, wherein the content by weight of chitosan is between 20% and 99.99% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler, and the content by weight of mineral filler is between 0.01% and 80% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler.

13. The method according to claim 1, wherein the content by weight of chitosan is between 20% and 99.99% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler, and the content by weight of mineral filler is between 0.01% and 80% by weight relative to the weight of dry extract of the mixture formed by chitosan and the mineral filler.

14. The method according to claim 2, wherein the aqueous composition comprises from 0.01% to 25% by weight of dry extract of the mixture formed by chitosan and the mineral filler with respect to the total weight of the composition.

15. The method according to claim 1, wherein the aqueous composition comprises from 0.01% to 25% by weight of dry extract of the mixture formed by chitosan and the mineral filler with respect to the total weight of the composition.

16. The method according to claim 3, wherein the aqueous composition comprises from 0.01% to 25% by weight of dry extract of the mixture formed by chitosan and the mineral filler with respect to the total weight of the composition.

17. The method according to claim 1, wherein the aqueous composition has a pH between 3 and 6.

18. A method for the fire-retardant treatment in the mass of a material comprising wood, natural fibers or synthetic fibers, wherein the method comprises the addition during the manufacture of the material of an aqueous composition as defined in claim 1 to the constituent elements of the material before it is formed.

19. The method of mass fire-retardant treatment of a material according to claim 18, wherein the formation is a formation into a panel.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 relates to the monitoring of the flame height as a function of the time for ignitability tests on particle boards (the compositions correspond to those described in Table 1 below).

(2) The solid line curve with the black diamonds corresponds to the panel alone, the solid line curve with black squares corresponds to a Teknosafe paint-coated panel (1), while the solid line curve with black triangles corresponds to a panel coated with Teknosafe paint (2), the dotted line curve with stars and the dotted line curve with white squares correspond to panels coated with a formulation comprising 60% of chitosan and 40% of talc, the dotted line curve with white circles and dotted curve with white diamonds correspond to panels coated with a formulation comprising 60% chitosan and 40% Nanoclay montmorillonite, while the dotted line curve with white triangles and the dotted line curve with crosses correspond to panels coated with a formulation comprising 60% of chitosan and 40% of montmorillonite K10.

(3) FIG. 2 shows the results of the ignitability test on 160 kg/m.sup.3 (d3) density panels: measurements of the carbonized area according to the type of coating applied and the flame exposure time. The gray columns correspond to a 30-second exposure while the black columns correspond to a 1-minute exposure.

(4) FIG. 3 relates to heat transfer curves (HRR) obtained for the SBI type tests on wood fibreboard (density of 140 kg/m.sup.3) (d4) beveled at 45° untreated and treated by spraying with an aqueous formulation based on talc and chitosan (Aldrich source or GTC Bio). The black solid curve corresponds to the raw panels, the gray solid curve corresponds to the panels treated with an aqueous Aldrich-based talc and chitosan formulation (test 1), the gray dotted curve corresponds to the panels treated with an aqueous Aldrich-based talc and chitosan formulation (test 2), while the gray semi-dotted curve corresponds to the panels treated with an aqueous formulation based on talc and chitosan from GTC Bio source (IGC-5).

(5) FIG. 4 relates to the monitoring of flame height versus time for particle board ignitability tests (the compositions correspond to those described in Table 8 described later).

(6) The solid line curve with the black diamonds corresponds to the single panel, the curve in solid lines with white squares and the curve in solid line with black triangles correspond to panels treated with an aqueous formulation whose dry extract is composed of 100% of chitosan (tests 1 and 2), the curve in solid lines with white circles and the dashed curve with black rhombs correspond to panels treated with an aqueous formulation whose dry extract is composed of 20% of chitosan and 80% montmorillonite K10 (tests 1 and 2), while the curve in solid lines with stars and the dashed curve with crosses (x) correspond to panels treated with an aqueous formulation whose dry extract is composed of 60% chitosan and 40% montmorillonite K10 (tests 1 and 2).

DETAILED DESCRIPTION OF EMBODIMENTS

(7) Materials

(8) The first screening tests were carried out on wood particle boards. The d2 density fibreboard is PAVATEX type PAVABOARD panels with a density of 200 kg/m.sup.3 and 60 mm thickness, while d3 density fibreboard is STEICO top type with a density 160 kg/m.sup.3 and 100 mm thick. The d4 density fibreboard is STEICO special dry type panels with a density of 140 kg/m.sup.3 and 80 mm thickness.

(9) Chitosan was purchased from Aldrich (low molecular weight grade) (Mn=75,750 g/mol, degree of deacetylation measured (DD, %) of 87.3%).

(10) Two batches of industrial grade chitosan were supplied by GTC Bio, batch IGC-3 (Mn=65 090 g/mol with a degree of deacetylation indicated by the industrialist of 87.6%) and lot IGC-5 (Mn=147 900 g/mol with a degree of deacetylation indicated by the industrialist of 86.1%).

(11) The Luzenac 10M2 talcum powder was supplied by Imerys Talc.

(12) The Montmorillonite K10 was purchased from Aldrich.

(13) The glacial acetic acid grade Reagent Plus comes from Aldrich.

(14) The intumescent commercial formulation Teknosafe, used as a reference, was provided by Teknos.

(15) All products were used as is, without additional purification.

(16) Preparation of the Formulation

(17) Chitosan and mineral filler (60/40 weight ratio, for example) were placed in a container. The permutated water was then added to obtain a formulation with 3.75% (w/w) solids. The pH was then adjusted to between 4 and 4.5 by adding small volumes of 99% acetic acid.

Preparation Example

(18) 180 g of chitosan (low MW, Aldrich) and 120 g of mineral filler (Luzenac talc 10M2, Imerys talc) were placed in a 15 L bucket. A volume of 8 L of deionized water was then poured in and 54 g of 99% acetic acid were added per fraction to the solution with mechanical stirring to make the chitosan soluble.

(19) The formulation was then homogenized by means of a mechanical stirring blade for 48 hours, at room temperature. Following the gradual decantation of the mixture over time, the formulation was again homogenized just before application using a paint disperser equipped with a spiral mixing rod or a deflocculating turbine.

(20) Other formulations were prepared with chitosans other than those mentioned above and results similar to those described later have been obtained.

(21) Coating Application

(22) Liquid Application

(23) For samples to be tested for ignitability, 60 mL of formulation was poured onto the panel surface (25×9 cm) and spread with a brush. To limit the flow of the formulation on the edges of the panel, the sample was surrounded by parchment paper or aluminum foil held in a foam mold. The samples were dried at room temperature under a fume hood (0.4 m/s) for 24 hours and then in a vacuum oven (−30 in.Math.Hg) for 8 hours at 40° C.

(24) Spray Application

(25) The kinematic viscosity of the formulation was evaluated by measuring the flow time (1 min 50 s) in an AFNOR #4 flow viscosity cup. The formulation was sprayed using a 4 bar piston pump with a 33/1 pressure ratio and a Binks gun. The formulation was cross-applied to the surface of wood fibreboard placed horizontally to reach the targeted basic wet weight.

(26) For the d2 and d3 density panels, the product was applied in a single layer. The panels were dried in an oven for 24 hours at 40° C. (0.53 m/s, 80% fresh air). The panels were then cut into pieces of 100×100 mm depending on the test to be performed (cone calorimeter).

(27) For the d4 density panels, the product was applied in two layers: the first layer was cross-applied to a basic wet weight of approximately 600 g/m.sup.2, and then the panel was dried horizontally in an oven for 1 h30 at 60° C. (about 0.5 m/s, 100% fresh air as oven ajar), then the operation was repeated a second time to obtain a final wet weight of about 1200 g/m.sup.2.

(28) Climate Aging

(29) Aging was carried out according to the NF P92-512 standard “Fire safety—Building—Reaction to fire tests of materials—Determination of the durability of fire reaction classifications of materials—Tests”.

(30) The samples were submitted to 4 cycles of a duration of two weeks each.

(31) After stabilization at 50% relative humidity and a temperature of 23° C. until a constant mass at 2% is obtained, the samples are subjected alternately (at 23±3° C.): 4 days at 90% of relative humidity, three days at 15% relative humidity, three days at 90% relative humidity and four days at 15% relative humidity, so that the times passed in each condition are similar in every other cycle.

(32) After the last cycle, the samples are returned to atmosphere at 50% relative humidity and 23° C. temperature until stabilization, which is verified by a constant mass to 2%.

(33) Upon leaving the conditioned chamber, the samples are submitted to regulatory fire reaction tests.

(34) Fire Reaction Tests

(35) The samples were conditioned for at least 48 hours in a climatic chamber at 23° C. and 50% humidity before being tested for reaction to fire (in accordance with standard NF EN 13238 “Reaction to fire tests of construction products—Packaging procedures and general rules for the selection of substrates”).

(36) Ignitability

(37) The test was conducted according to the methodology described in standard NF EN ISO 11925-2: “Reaction to fire tests—Ignitability of building products subjected to the direct impact of the flame—Part 2: Test with the aid of ‘a single flame source’. This test makes it possible to evaluate the ignitability of a product exposed to a low thermal load simulated by a small flame.

(38) For the particle board tests, the samples were exposed to flame for 5 minutes. The flame height (in cm) as a function of time was evaluated visually according to the graduations in cm plotted on the sample.

(39) For the same sample of wood fiber board, one replica was exposed to the flame for 30 seconds while the second replica was exposed for 1 minute. After the test, the samples were photographed horizontally by means of a camera placed vertically. The charred surface was evaluated by image processing of these photographs using the NIVision software.

(40) Cone Calorimeter

(41) The test was conducted according to the methodology described in ISO 5660-1: 2002: “Reaction to fire tests—Heat release, smoke release rate and mass loss rate—Part 1: Heat transfer rate (cone calorimeter method) and smoke release rate (dynamic measurement)”. The test consists in evaluating the heat flow rate and the dynamic smoke release rate of horizontally oriented specimens exposed to irradiance levels controlled by an external source. The samples were exposed to a heat flux of 35 or 50 kW/m.sup.2. For each sample, 2 replicas were analyzed. A restraining frame was used, therefore the exposed area was 88.4 cm.sup.2. The heat rate was determined by measuring the oxygen consumption, as well as the flow rate in the flue of the combustion products. Ignition time (persistent flame) was also measured during this test. The following indices were obtained: tig: ignition delay PHRR: Peak of Heat Release Rate (“Peak of Heat Release Rate”) THR: total energy released after 600 s of test (Total Heat Release) Max HRR30 s: average heat output over 30 s maximum test (HRR30 s=RHR30 s)

(42) Single Burning Item (SBI)

(43) The test was inspired by the methodology described in the standard “NF EN 13823—Reaction to fire of building materials—Building materials” with the exception of floor coverings—exposed to thermal stress caused by an isolated object on fire. The specimen consisting of two vertical wings (1.50 m high and 0.50 m wide) forming a right angle was exposed to flames from a main burner placed at the bottom of the corner. The flames are obtained by burning propane gas injected through a bed of sand and producing a heat flow of 30.5 (+/−2) kW. The performance of the test piece was evaluated over a period of 21 minutes. The performance criteria are: heat generation and flame front propagation. The propagation of the flame front was estimated by visual observation. These quantities were automatically recorded and used for the calculation of the following indices: FIGRA (W/s): Fire Development Index THR600 s (MJ): Total energy released between 300 and 900 seconds

(44) Results

(45) I—Ignitability

(46) Screening on Particle Board

(47) The study was carried out on particle board coated with formulation (dry extract composed of 60% chitosan and 40% mineral filler) by liquid application. The dry weights obtained are listed in Table 1 below.

(48) 3 types of mineral fillers were tested: talc (non-exfoliable silicate), montmorillonite Nanoclay (classic, exfoliable, with Na.sup.+ counter-ions) and montmorillonite K10 (treated HCl, exfoliable, H.sup.+ counter-ions).

(49) The formulations were compared to a commercial intumescent paint for wood, Teknosafe.

(50) The samples were exposed to the flame for 5 min during which the flame height was read every 30 s. The results are shown in FIG. 1.

(51) TABLE-US-00001 TABLE 1 Quantity of residual formulation, after drying, on the surface of particle board samples for ignitability tests. Dry grammage Composition (g/m.sup.2) Teknosafe-1 175 Teknosafe-2 167 60% chitosan 40% talc 24 60% chitosan 40% talc 25 60% chitosan 40% Nanoclay 33 60% chitosan 40% Nanoclay 29 60% chitosan 40% MMT K10 94 60% chitosan 40% MMT K10 125

(52) Although the amount of formulation varies from one sample to another (see Table 1 above) because of the poorly controlled application method, the flame heights are significantly reduced compared to those measured for the raw panel, whatever the coating considered. The flame heights for the formulations according to the invention are of the order of magnitude of those obtained for the Teknosafe paint. The flame heights measured for the formulation containing MMT K10 are lower because of the higher basis weight.

(53) Screening on Fiberboard

(54) The study was carried out on panels of density d3 (160 Kg/m.sup.3) coated with formulation by liquid application. The results are shown in FIG. 2. Although the amount of formulation varies from one sample to another (see Table 2) because of the mode of application, the formulation containing 60% of chitosan and 40% of talc (percentage dry extract) shows the smallest charred surfaces after exposure to the flame. At equivalent grammage, the 60% chitosan 40% mineral filler formulations may be compared with each other, as well as the 80% chitosan and 20% mineral filler formulations. Both these comparisons show better results in the case of talc. Despite a higher grammage than that of the 80% chitosan 20% talc formulation, the 40% chitosan 60% talc formulation ignites after 1 min of exposure, unlike the 80% chitosan 20% talc formulation, which does not ignite in the analysis time.

(55) TABLE-US-00002 TABLE 2 Quantity of residual formulation, after drying, on the surface of small-flame-tested wood-fiber panel samples (ignitability test). composition dry grammage (g/m.sup.2) Teknosafe 329 349 40% chitosane 60% MMT K10 81 62 60% chitosan 40% MMT K10 172 206 80% chitosan 20% MMT K10 87 102 40% chitosan 60% talc 143 119 60% chitosan 40% talc 169 197 80% chitosan 20% talc 105 110

(56) II—Cone Calorimeter

(57) d2 Density Panels (200 kg/m.sup.3)

(58) The 60% chitosan 40% talc formulation was applied to the panels by spraying to a wet weight of about 500 or 800 g/m.sup.2. A part of the panels treated at 800 g/m.sup.2 was subjected to climatic aging for 8 weeks, then analyzed with a cone calorimeter.

(59) Before Climatic Aging

(60) The samples were subjected to a thermal flux of 50 kW/m.sup.2. The parameters calculated from the heat flow rate curves (HRR) obtained are summarized in Table 3 below.

(61) Although the coating barely increases the ignition time (tig), the peak heating rate (PHRR), representative of the surface effect, decreases markedly as the grammage increases. This decrease of the PHRR with the amount of surface product proves the effectiveness of the coating. The total energy released over 600 s (THR (600 s), area under the HRR curve over 600 s of test) is only slightly influenced by the coating because it represents more the effect of the mass on the reaction properties to fire.

(62) TABLE-US-00003 TABLE 3 Cone calorimeter data obtained for raw and treated d2 panels prior to weathering. tig is the ignition time, PHRR represents the heat flow peak, and THR (600 s) indicates the total energy released after 600 s of test. The samples were subjected to a thermal flux of 50 kW/m.sup.2. heat flux PHRR THR(600 s) sample (kW/m.sup.2) tig (s) (kW/m.sup.2) (MJ/m.sup.2) Before aging d2-raw 50 6 230 46 d2-567 g/m.sup.2 50 5 166 48 d2-793 g/m.sup.2 50 9 119 40

(63) After Climatic Aging

(64) The samples were subjected to a thermal flux of 35 kW/m.sup.2. The parameters calculated from the heat flow rate curves (HRR) obtained are summarized in Table 4 below. Even when aged, the coating helps to increase the ignition time (tig), from 13 to 19 s after treatment. The heat flow peak (PHRR), representative of the surface effect, is largely decreased with the coating. This drop in PHRR, coupled with the slight increase in tig after treatment, confirms the effectiveness of the coating, even after aging. The total energy released over 600 s (THR (600 s), area under the HRR curve over 600 s of test) is still only slightly influenced by the coating because it represents more the effect of the mass on the properties of reaction to fire.

(65) TABLE-US-00004 TABLE 4 Cone calorimeter data obtained for raw and treated d2 panels, after climatic aging. tig is the ignition time, PHRR represents the heat flow peak, and THR (600 s) indicates the total energy released after 600 s of test. The samples were subjected to a thermal flux of 35 kW/m.sup.2. THR heat flux PHRR (600 s) sample (kW/m.sup.2) tig (s) (kW/m.sup.2) (MJ/m.sup.2) After aging d2-gross aged 35 13 193 34 d2-793 g/m.sup.2 aged 35 19 94 32

(66) d3 Density Panels (160 kg/m.sup.3)

(67) The samples were subjected to a thermal flux of 50 kW/m.sup.2. The parameters calculated from the heat flow rate curves (HRR) obtained are summarized in Table 5 below. The coating contributes to increase the ignition time (tig), from 2 to 4 s after treatment. The heat flow peak (PHRR), representative of the surface effect, is halved with the coating. This drop in PHRR, coupled with the slight increase in tig after treatment, confirms the efficacy of the coating. The total energy released over 600 s (THR (600 s), area under the HRR curve over 600 s of test), is only slightly influenced by the coating because it represents more the effect of the mass on the reaction properties to fire.

(68) TABLE-US-00005 TABLE 5 Cone calorimeter data obtained for raw and treated d3 panels. tig is the ignition time, PHRR represents the heat flow peak, and THR (600 s) indicates the total energy released after 600 s of test. The samples were subjected to a thermal flux of 50 kW/m.sup.2. THR heat flux in PHRR (600 s) sample kW/m.sup.2 tig (s) (kW/m.sup.2) (MJ/m.sup.2) d3-raw 50 2 248 43 d3-800 g/m.sup.2 50 4 116 42

(69) III—Single Burning Item (SBI): Fire Reaction Test for Euroclass Classification (d4 Density Panels 140 kg/m.sup.3)

(70) Tests were carried out on panels treated with the formulation with talc and based on chitosan (Aldrich) and chitosan GTCBio (GTC Bio IGC-5, DD (%)=86.1% (supplier data, DD (%))=90.7%. Measured by 1H NMR) (Mn=147,900 g/mol), the panels having been cut beforehand at 45° to optimize the angle, and raw panels.

(71) Quantity of Formulation

(72) The models consist of 2 panels positioned at a 90° angle, each measuring 150×50 cm. The formulation was spray applied to the surface of d4 density panels.

(73) One of the fields in the length was cut at a 45° angle to optimize angle mounting. Only one face and the 45° field were treated. The amount of product actually sprayed on the surface of the panels is given in Table 6.

(74) TABLE-US-00006 TABLE 6 Quantity of formulation applied to the surface of density boards d4 (140 kg/m3) tested. The basic dry weight was estimated using the dry extract value of 3.6%. panel 1.sup.st layer 2.sup.nd layer Total wet Total dry N.sup.o (g/m.sup.2) (g/m.sup.2) (g/m.sup.2) (g/m.sup.2) Chitosan 1 597 620 1217 43.81 (Aldrich) 2 720 482 1202 43.27 3 560 660 1220 43.92 4 580 710 1290 46.44 Chitosan 5 628 582 1210 43.56 GTCBio IGC-5 6 640 586 1226 44.13

(75) SBI Test Results

(76) The SBI test measures the heat flow rate (HRRav) produced by the sample as it burns by the burner as a function of time (FIG. 3). From these data are calculated: the fire growth rate (FIGRA), characterized by the highest slope recorded on the HRRav curve as a function of time, the total energy released after 600 s of test (THRta), area under the HRR curve as a function of time from 300 s to 900 s), characteristic of fireproofing in the mass.

(77) The calculated parameters and the resulting Euroclass classification are given in Table 7.

(78) As indicated above, the panels were beveled at 45° and treated with two different types of chitosan: Aldrich chitosan (Mn=75,750 g/mol, Mw/Mn=2.3) and chitosan from GTCBio (Mn=147,900 g/mol and Mw/Mn=1.8). In view of the HRRav curves (FIG. 3), the angle design with the 45° bevels significantly reduced the THR values, since the opposite (untreated) face hardly ignited (no HRR jump).

(79) The heat transfer curves (HRR) obtained for the SBI tests on panels treated at 1200 g/m.sup.2 with Aldrich chitosan (45° beveled), the panels treated with chitosan GTCBio IGC-5 (beveled at 45°), and untreated raw panels (45° beveled panels) are shown in FIG. 3.

(80) TABLE-US-00007 TABLE 7 Results obtained for SBI type tests on panels treated with an Aldrich chitosan-based formulation and talc (tests with 45° beveled panels), panels treated with a GTCBio IGC chitosan formulation-5, and talc (with 45° beveled panels) and rough panels (45° beveled panels). Wet THRta Est. Est. grammage (t0_t0 + 600 s) Class FIGRA Class Retained Sample (g/m.sup.2) [MJ] THRta [W/s] FIGRA class Untreated raw panel-45° — 41.8 D 1 902.59   E E bevel chitosan Aldrich bevel 45°- 1 200 8.5 C 108.24 B C test 1 chitosan Aldrich bevel 45°- 1 200 9.2 C 127.87 C C test 2 chitosan GTCBio IGC-5 45°- 1 200 10.5 C 145.95 C C bevel

(81) The Euroclass ranking is based on the values of FIGRA and THR. Panels treated with formulations based on 2 sources of chitosan and talc are classified C with respect to the untreated raw panel which is classified E. An applied wet weight of 1200 g/m.sup.2 is sufficient to reach class C.

(82) IV—Particle Board Ignitability Tests

(83) The study was carried out on coated particle boards of aqueous formulation (whose dry extract is composed of 100% chitosan or 60%/40% acid-treated chitosan/montmorillonite (MMT K10) by liquid application).

(84) 60 mL of formulation was poured onto the panel surface (25×9 cm) and spread with a brush. To limit the flow of the formulation on the edges of the panel, the sample was surrounded by parchment paper or aluminum foil held in a foam mold. The samples were dried at room temperature under an extractor hood (0.4 m/s) for 24 hours and then in a vacuum oven. (−30 in.Math.Hg) for 8 h at 40° C. Dry grammages obtained after evaporation of water are listed in Table 8. The samples were exposed to the flame for 5 min during which the flame height was read every 30 s. The results are shown in FIG. 4.

(85) TABLE-US-00008 TABLE 8 Quantity of residual formulation, after drying, on the surface of particle board samples for ignitability tests. Sample Grammage dry (g/m.sup.2) 100% chitosan - test 1 114 100% chitosan - test 2 182  20% chitosan 80% MMT K10 - test 1 62  20% chitosan 80% MMT K10 - test 2 78  60% chitosan 40% MMT K10 - test 1 94  60% chitosan 40% MMT K10 - test 2 125

Comparative Example

(86) The same composition as that of the Preparation Example under Preparation of the formulation is carried out by replacing the acetic acid with the same amount of ricinoleic acid.

(87) Rheological and wettability tests are performed, as well as fire-retardant tests.

(88) These tests show results different from those obtained with a composition according to the invention as described above.