Ro

Ba

db

c

a

ARRAA

KRIOC

1

ibbcoithaaaatwtts

0h

Progress in Organic Coatings 76 (2013) 1633– 1641

Contents lists available at ScienceDirect

Progress in Organic Coatings

jou rn al hom ep age: www.elsev ier .com/ locate /porgcoat

esistance to fire of intumescent silicone based coating: The rolef organoclay

. Gardellea, S. Duquesnea, P. Vandereeckenb, S. Bellayera,c, S. Bourbigota,∗

ISP/UMET – UMR/CNRS 8207, Ecole Nationale Supérieure de Chimie de Lille (ENSCL), Avenue Dimitri Mendeleïev – Bât. C7a, BP 90108, 59652 Villeneuve’Ascq Cedex, FranceDow Corning S.A., High Performance Building solutions, Rue Jules Bordet 8, 7180 Seneffe, BelgiumEcole Nationale Supérieure de Chimie de Lille (ENSCL), Avenue Dimitri Mendeleïev – Bât. C7a, CS 90108, 59652 Villeneuve d’Ascq Cedex, France

r t i c l e i n f o

rticle history:eceived 27 February 2013eceived in revised form 24 June 2013ccepted 16 July 2013vailable online 6 August 2013

eywords:

a b s t r a c t

The fire performance of a curable-silicone based coatings containing expandable graphite (EG) and anorganoclay is evaluated in hydrocarbon fire scenario (standard UL1709) using a lab-scale furnace test.It is shown that the use of organoclay allows achieving better performance. The influence of the clay asadditional filler is investigated on the fire performance and on the mechanical properties of the char. Itis shown that the clay increases significantly the mechanical properties of the char and hence, the fireperformance of the silicone based coating. In a next part, the silicone/clay material was characterized

29

esistance to fire

ntumescent siliconerganoclayhar strength

by electron microscopy, wide-angle X-ray scattering and solid state Si nuclear magnetic resonance(NMR). It evidences the nanodispersion of the clay into the silicone matrix and two main interactions: (i)intercalation of some silicate layers and (ii) chemical reactions between the hydroxyl groups of the clayand the silicone matrix. Finally, X-ray fluorescence of the residue after fire testing shows the organoclayis present uniformly throughout the thickness of the char, due to the previous interaction, and henceincreasing the cohesion of the char.

. Introduction

The protection of metallic materials against fire has become anmportant issue in the offshore platform industry. Indeed, steelegins to lose its structural properties above 500 ◦C and it muste therefore protected against fire [1]. Prevention of the structuralollapse of building is paramount to ensure the safe evacuationf people from the building, and is a prime requirement of build-ng regulations in many countries. One of the most used systemso protect metallic structures is intumescent paint. These coatingsave the properties to swell when exposed to fire and thus make

thick insulative foam. Intumescent coatings are mostly based on combination of a char-forming material, a mineral acid catalyst, blowing agent and a binder resin [2,3]. However, these materialsre typically organic-based materials and exhibit some disadvan-ages. First, organic additives undergo exothermic decompositionhich reduces the thermal insulative value of the system. Second,

he resulting carbonaceous char in some cases has a low struc-

ural integrity and the coating cannot withstand the mechanicaltress induced by a fire and/or by other external constraints. And

∗ Corresponding author. Tel.: +33 03 20 43 48 88.E-mail address: serge.bourbigot@ensc-lille.fr (S. Bourbigot).

300-9440/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.porgcoat.2013.07.011

© 2013 Elsevier B.V. All rights reserved.

third, the coating releases organic gases (potentially toxic) whichare undesirable in a closed fire environment [4].

Currently, some alternative to organic intumescent coating havebeen studied [5–7] in our lab, we recently evaluated the fire per-formance of an intumescent silicone based coating [5]. In thisstudy, the fire performance of a phenyl silicone resin containingsilica-based modifier was evaluated in pure radiative and in con-vective/radiative heating conditions [5]. We reported the good heatbarrier properties of this intumescent silicone based coating inradiative/convective heating whereas fire performance of this coat-ing is rather limited in the case of pure radiative heating. Indeed, inpure radiative heating source, silicone based coatings cracks due tothe high vibration of Si O bond in infrared field and so, it exhibitslow fire performances.

On the other hand, expandable graphite (EG) is a “particular”intumescent additive known to impart fire retardancy to variousmaterials [8]. EG is a graphite intercalation compound in whichsulfuric acid and/or nitric acid is inserted between the carbon lay-ers of graphite. Upon heating, exfoliation of the graphite occurs, i.e.expansion along c-axis of the crystal structure by about hundredtimes. The material generates in that way is a puffed-up material

of low density with a “worm” like structure. In recent decades,more and more papers reported the use of expandable graphitein intumescent based coating. This intumescent additive increasesthe fire performance and anti-oxidant properties of traditional

1 ganic Coatings 76 (2013) 1633– 1641

itoot

ipracawtcebmmtcssp

eoTccro

2

2

hs(r

miopfiw1cfTc

2

tsilRtp

634 B. Gardelle et al. / Progress in Or

ntumescent based coating [2,9,10]. However, in the above men-ioned studies, expandable graphite is incorporated into complexrganic intumescent based formulations. It is noteworthy that inrganic based coating, EG decreases considerably the cohesion ofhe char [10].

We recently reported the used of silicone based coating contain-ng EG for the fire protection of steel [11]. We reported high fireerformance of silicone containing 25% of EG in curable siliconeesin cross linked with titanium-based catalyst. The good perform-nces were explained by the formation of an expanded insulativehar (3400% expansion) formed at a high expansion velocity (18%/s)nd exhibiting a low thermal conductivity (0.35 W/K m at 500 ◦C). Itas shown that the formation of a complex silicone/graphite struc-

ure at high temperature is responsible of the good cohesion of thehar in fire scenario. Mechanical strength of the char is a key param-ter to ensure its integrity in fire scenario. This factor is significantecause in fire scenario, char destruction can proceed not only byeans of ablation and heterogeneous surface burning but also byeans of an external influence such as wind, mechanical action of

he fire or convective air flows. To ensure the good cohesion of thehar, several additives are added to the intumescent formulationuch as mineral fibers [12], organo-clay [13,14], ceramic precur-or [15]. In this study, organoclay are used to increase mechanicalroperties of the char.

The purpose of this paper is to investigate the heat barrier prop-rties of curable silicone/expandable graphite based coating usingrganoclay as filler in hydrocarbon fire scenario (standard UL1709).he effect of the organoclay on the mechanical properties of thehar will be carefully investigated. To explain the influence of thelay on the fire performance and mechanical properties of theesulting char, silicone matrix containing organoclay and residuesbtained after furnace test will be characterized.

. Experimental

.1. Materials

The silicone resin called silicone 1 (S1), is composed of aydroxylated PDMS with a viscosity of 15,000 cS (viscosity is mea-ured using cone/plate rheometer CP-52), methyltrimethoxysilaneMTM) as crosslinking agent and a titanium catalyst. All the mate-ials were supplied by Dow Corning, Seneffe (Belgium).

25% of expandable graphite (ES350F5 from Graphitwerk Kropf-uehl (Germany)) with an average grain size of 300 �m was

ncorporated in the silicone matrix. 5% of Cloisite 30B ((C30B) –rganoclay mineral from Southern Clay) are incorporated and dis-ersed manually with a spatula in the formulations to increase there performances of the coating. Each formulation was appliedith a spatula on a 10 cm × 10 cm × 3 mm steel plate to obtain

.0 ± 0.15 mm and 1.5 ± 0.15 mm coatings. Steel plates were firstleaned before application with ethanol and a primer (Primer 1200rom Dow Corning) was applied to enhance the coating adhesion.o characterize the silicone/clay material, the ratio S1/clay was keptonstant and so, S1 + 6% clay material is characterized.

.2. Fire testing methods

The small scale furnace test was developed in our laboratoryo evaluate the fire performance of intumescent coatings in firecenario. This test was designed to mimic the UL1709 normal-zed temperature/time curve, referred to hydrocarbon fire. The

ab-made furnace exhibits an internal volume of 26 dm3 (Fig. 1).efractory fibers (stable up to 1300 ◦C) cover the different sides ofhe furnace. The furnace is equipped with two gas burners (20 kWropane burners). The gas pressure was fixed at 1.8 bars and the

Fig. 1. Furnace set up to mimic hydrocarbon fire scenario.

flow was regulated in order to mimic the UL1709 curve. A tem-perature probe inside the furnace regulates the temperature and aK-type thermocouple allows the furnace temperature profile to beregistered. The furnace is equipped with a quartz window allowingfollowing the intumescent process taking place during the test.

The temperature at the back side of the plate is measured asa function of time using a pyrometer (temperature measured inthe center of the plate). The plate is not insulated. The backsideof the plate is coated with a black paint (Jeltz) having a constantemissivity (0.92) and thermally resistant up to 800 ◦C in order toget reliable temperature measurements with the pyrometer. Thecritical temperature of steel defined as the temperature at whichonly 60% of the original strength remains, point at which failure isimminent under full design loads were obtained. The temperatureof 500 ◦C has been adopted as a standard for normally loaded struc-tural components whereas 400 ◦C is often used for heavily loadedstructure [16]. In this study, the time to reach the two temperaturesis taken into account to evaluate the heat barrier properties of theintumescent coatings.

2.3. Air jet test

To evaluate the mechanical properties of an intumescent mate-rial few experimental setup are reported in the literature [17,18].In the lab, we have developed an approach involving a rheometerin which the intumescent structure is developed at a given temper-ature (for example 500 ◦C) without any constrain. After the steadystate was obtained, the upper plate of the rheometer is broughtinto contact with the intumesced material and the gap betweenthe plates is reduced linearly (0.02 mm/s). The compression force isthus followed as a function of the gap [17]. However, this techniqueis limited for experimental reason to 500 ◦C and the silicone basedformulation are not degraded at this temperature and its expan-sion is not maximum, so this technique is not appropriate. Anothermethod was developed by Reshetnikov et al. [18] using a “Structur-ometer ST-1”, developed at the Moscow State Food Academy. Theymeasure the force required to destroy a char. The samples were firstpyrolyzed and then a destructive force was applied to the sample.Mechanical properties of the char are investigated at ambient tem-perature which may not be representative of its behavior at hightemperature.

Based on the above considerations, we have designed a new labscale test to evaluate the mechanical properties and the cohesion ofthe char obtained from intumescent coatings. Its set up is illustratedin Fig. 2.

The coating is directly exposed to an electrical cone heater pro-viding pure radiative heat (in this study, 35 kW/m2). When theexpansion of the coating is maximal (after 5 min testing), an airflow (25 m/s) impacts the char. The air flow was measured using

B. Gardelle et al. / Progress in Organic

Fo

ahtsafi

2

edt

ocsetwug

2

tfX

ig. 2. Schematic experimental set up used to determinate the cohesion propertiesf the char.

n anemometer Testo 425. By visual observation, it is possible toave an idea of the mechanical properties of the char. Indeed, ifhe char is fragile, it will be destroyed by the air impact whereas atrong char will not or only slightly be affected by the air jet. Thedvantage of this test is to evaluate the mechanical properties in are scenario.

.4. 29Si solid state NMR

29Si nuclear magnetic resonance (NMR) spectroscopy is a pow-rful tool for examining silicon surrounding. This technique canistinguish several kinds of structures including D, T and Q struc-ures which characterize silicone network (Fig. 3).

29Si NMR spectra were recorded on a Bruker Advance II 400perating at 9.4 T and using a 7 mm probe. Zirconia rotor andaps are used. NMR spectra were acquired with MAS (magic anglepinning) of 5 kHz. The reference used for 29Si NMR was tetram-thylsilane (TMS). For the pure clay sample, a delay of 30 s betweenhe pulses and a �/2 pulse length for 6 ms were used and 2560 scansere accumulated. For S1 and S1/clay samples a delay of 180 s wassed. Respectively 288 scans and 2267 scans were accumulated toet an acceptable signal to noise ratio.

.5. X-ray fluorescence (XRF)

Residue after fire testing were analyzed by X-ray fluorescenceo determine its chemical composition. Measurements were per-ormed on a Bruker S2-RANGER spectrometer equipped with a Pd-ray tube and a XFLASHTM Silicon Drift Detector (SDD).

Fig. 3. Schematics presentation of D

Coatings 76 (2013) 1633– 1641 1635

2.6. Wide-angle X-ray scattering (WAXS)

Measurements were performed by wide-angle X-ray scattering(WAXS) at room temperature. The X-ray source was generated bya 1.5 kW sealed tube with Cu target (� = 1.54 A), operated at 20 mAand 40 kV. The sample-to-detector distance was 20 cm. 2D-WAXSpatterns were recorded on a CCD camera from Photonic ScienceLtd.

2.7. Transmission electron microscopy (TEM)

Ultramicrotomy was used to prepare the sample. Bright-fieldTEM images of sample were obtained at 200 kV under low doseconditions with a FEI TECNAI 62 20 electron microscope, using aGatan CCD camera.

3. Results

3.1. Fire protection

Fig. 4 shows temperature profiles as a function of time onthe backside of the steel plates coated with the different for-mulations during hydrocarbon fire scenario. The influence of theclay on the heat barrier properties of silicone based coating isclearly shown: the paint prepared with 5% of clay exhibits a betterbehavior for 1 mm and particularly for 1.5 mm thick intumescentcoatings. The reproducibility of measurement is acceptable andan error of 10% about the time to reach critical temperature isobtained.

For 1 mm based coating, 500 ◦C is reached in 1300 ± 130 sand 1860 ± 180 s respectively for S1/EG and S1/EG/clay (Table 1)demonstrating the better fire performance of silicone based coat-ing when organoclay is used. Similar results are obtained at 400 ◦Csince the time to reach this temperature increases from 715 ± 70 sfor S1/EG to 900 ± 90 s for S1/EG/clay. For the two formulations,the expansion rate reaches 3400%. Both residue are presented inFig. 5(a) and (b) and they exhibits same morphology. The two charare covered by a white powder.

Fire tests carried out on 1.5 mm based coatings also evi-

dence the better cohesion of the char when clay is used asadditional filler. Indeed, when just 25% EG is added to siliconematrix there is a complete loss of adhesion of the char from theplate after 400 s (Fig. 5(c)) leading to the loss of the insulative

, Ti and Qi silicone structures.

1636 B. Gardelle et al. / Progress in Organic Coatings 76 (2013) 1633– 1641

Fig. 4. Temperature versus time curve of S1/EG and S1/EG/clay based coating in hydrocarbon fire scenario.

Table 1Time to reach critical temperature.

S1/EG – 1 mm S1/EG/clay – 1 mm S1/EG – 1.5 mm S1/EG/clay – 1.5 mm

90 s

180 s

p5i(

Fb

Time to reach 400 ◦C 715 ± 70 s 900 ±

Time to reach 500 ◦C 1300 ± 130 s 1860 ±

roperties. The temperature thus sharply increases to reach00 ◦C in 490 ± 40 s. However, when 5% of clay is incorporated

n the S1/EG formulation, there is good cohesion of the charFig. 5(d)) and so, the silicone based coating remains stuck on the

ig. 5. (a) S1/EG 1 mm based coating residue, (b) S1/EG/C30B 1 mm based coating residueased coating residue.

428 ± 40 s 1000 ± 100 s490 ± 40 s 2050 ± 200 s

plate providing high insulative properties. For the two formula-tions, critical temperature (500 ◦C) is reached in only 490 ± 40 sfor S1/EG while 2050 ± 200 s is needed for S1/EG/clay basedcoatings.

, (c) S1/EG 1.5 mm based coating during fire experiment and (d) S1/EG/C30B 1.5 mm

B. Gardelle et al. / Progress in Organic Coatings 76 (2013) 1633– 1641 1637

Fig. 6. (a) S1/EG, (b) S1/EG/C30B during air jet test before air impact, (c) S1/EG residue after air impact and (d) S1/EG/C30B after air impact.

icture

3

mScsicT

Fig. 7. TEM p

.2. Mechanical properties of the char

In order to investigate the mechanical properties of the intu-escent chars and the influence of the clay on it, the S1/EG and

1/EG/clay formulations were tested using the air jet test. Theoatings exhibit similar behavior (Fig. 6(a) and (b)) during the first

tep of the test (when the air does not impact the coating). Dur-ng the first minute, the coatings begin to swell; after 2 min, theoatings ignite and a white powder covers the expanded graphite.he residues of the two formulations after air jet testing are shown

s of S1/clay.

in Fig. 6(b) and (c). It could be observed that, when the air flowis switch on, a complete destruction of the char occurs for S1/EGbased coating while there is no complete char destruction whenthe clay is incorporated into the formulation.

This evidences the superior mechanical properties and theextremely good cohesion of the char when organoclay is used as

filler. We have thus demonstrated that the organoclay has a verysignificant influence on the cohesion and the mechanical proper-ties of the intumescent char and, hence, on the fire performance ofsilicone based coating. In the next step of this study, the mode of

1638 B. Gardelle et al. / Progress in Organic Coatings 76 (2013) 1633– 1641

Fig. 8. WAXS pattern of clay, silicone matrix and silicone/clay composite.

f (a) c

ao

3

cieo

mctpi

Fig. 9. 29Si NMR spectra o

ction of the organoclay is evaluated. First of all, the morphologyf the resin added with the clay is characterized.

.3. Silicone/organoclay characterization

In order to highlight possible interaction between the sili-one matrix and the clay, the system without expandable graphites studied. And, we can reasonably assume that the addition ofxpandable graphite should not affect significantly the dispersionf the clay in the material.

Fig. 7 shows the TEM pictures of S1/clay sample at differentagnitude. Some aggregates and tactoids of 3–10 layers in size

an be distinguished and individual platelets can be distinguish onhe pictures. Even if the dispersion of the organoclay is not com-letely, it is acceptable and we can conclude that a nanodispersion

s achieved.

lay, (b) S1 and (c) S1/clay.

WAXS patterns of the organoclay, silicone matrix and sili-cone/clay are presented in Fig. 8. For pure polymer, no crystallinitypeak in the diffraction pattern. Concerning, pure organoclay, asingle peak at 2� = 5.0◦ corresponding to a distance between thesilicate layer of about 18 A [19] is observed. This value is consistentwith the value found in the literature. When the clay is added tothe silicone matrix, there is an increase of the interlayer distanceof the silicate evidenced by a shift of the peak corresponding tothe clay from 5.0◦ to 3.0◦. The interlayer increases from 18 A to30 A.

TEM images indicate the nanodispersion of the organoclayand the WAXS analysis evidences a decrease of the peak in 2�

attributed to an intercalation into the galleries of the clay. Thenext step was to point out if potential interactions could occurbetween the clay and the silicone matrix. For this, 29Si NMR spec-troscopy experiments are carried out. Fig. 9 presents the NMR

B. Gardelle et al. / Progress in Organic Coatings 76 (2013) 1633– 1641 1639

of the

soptttcptl

trcacogsh

ati

Fig. 10. XRF spectrum of: (a) the top of the char, (b) 1.5 cm from the top

pectra of the clay, S1 and S1/clay materials. Three sites arebserved: D (−22 ppm), T2 (−57 ppm) and T3 (−66 ppm) for theure silicone. The T structures come from the reaction betweenhe hydroxyl function of the PDMS and the methoxy groups ofhe MTM as we previously reported [11]. For the organoclay spec-rum, three peaks appear at −89 ppm, −94 ppm and −108 ppmorresponding respectively to Q2, Q3 and Q4 structures [20]. Theresence of Q2 and Q3 structure indicates the presence of reac-ive hydroxyl groups at the surface or on the edges of the silicateayer.

Concerning S1/clay spectrum, the three peaks correspondingo the matrix are observed at −22 ppm, −57 ppm, −66 ppm cor-esponding respectively to D, T2 and T3 structure. The peaksorresponding to the clay (Q3 and Q4) at −94 ppm and −108 ppmlso show up. However, there is no peak (or with low intensity)orresponding to Q2 structure at −89 ppm. A possible explanationf this observation is the chemical reaction between the hydroxylroups of the organoclay and the silicone matrix. This reactionhould occur between the ethoxy groups from the MTM and theydroxyl groups of the clay [21].

This study evidences three main conclusions: (i) the organoclayre nanodispersed in the silicone matrix, (ii) there is an intercala-ion of silicone chains into the galleries of the clay and (iii) chemicalnteraction between organoclay and silicone could be suspected.

char, (c) 2.5 cm from the top of the char and (d) the bottom of the char.

3.4. X-ray fluorescence (XRF) of the residues

A good dispersion of the clay in the char should be necessaryto ensure its good cohesion during fire testing [22,23]. To examinethis, different parts of the residue of S1/EG/clay were analyzed byX-ray fluorescence. The final expansion of the char is about 3.4 cm.To have an idea of the dispersion of the clay in the char, four partswere analyzed by XRF: the top, 1.5 cm from the top, 2.5 cm fromthe top and the bottom of the char. Organoclay are alumino-silicateand thus if aluminum is detected in the sample it will evidence thepresence of the organoclay. Fig. 10 shows the XRF spectrum of thefour parts of the char.

For all samples the same peaks are observed. The main peakat 1.75 keV, 1.5 keV and 1.1 keV correspond to respectively silicon,aluminum and palladium. Palladium comes from the experimen-tal device (the source). These analyses evidence the fact that theorganoclay is detected in all parts of the char.

3.5. Discussion

The fire performance of silicone/expandable graphite basedcoating has been investigated in hydrocarbon fire scenario for twodifferent thicknesses: 1 mm and 1.5 mm. For thick intumescentbased coating, when 25% of EG is added to the silicone matrix

1640 B. Gardelle et al. / Progress in Organic Coatings 76 (2013) 1633– 1641

Fig. 11. Schematic intercalation of the organoclay in the silicone matrix.

clay

tTscrfiiaocitcimtnc

Fig. 12. Mode of action of the

here is a complete loss of adherence of the char from the plate.his is explained by a low cohesion of the char and by the impres-ive expansion of the system (around 3000%). In order to avoid theollapse of the structure for thick coating, organoclay was incorpo-ated in the formulation. In this case, S1/EG/clay exhibits excellentre performances at 1 mm and 1.5 mm thick coating. The mechan-

cal properties of the char have been investigated using air jet testnd evidence the better mechanical properties of the char whenrganoclay is added to the S1/EG. The characterization of the sili-one/organoclay composite shows the nanodispersion of the clayn individual platelets and intercalation of the silicone chains intohe galleries of the clay. Moreover, 29Si NMR lets us to proposehemical reaction between the clay and the silicone matrix lead-ng to a strong cohesion between the clay platelets and the silicone

atrix. This is illustrated in Fig. 11. During the cross-link step ofhe resin/clay material, there is formation of a three dimensionaletwork based on T structures and some Q structures due to thehemical reaction between the silicone and the clay.

in hydrocarbon fire scenario.

In hydrocarbon fire scenario, there is expansion of the coat-ing and the graphite platelets are embedded in complex siliconestructure [11].

The residue after furnace testing was analyzed by X-ray fluo-rescence and shows the presence of the clay in the whole. As aconsequence, it may be proposed that the good interaction betweenthe matrix and the clay platelets lead when the material degradeto the formation of a cohesive structure that will embedded theexpanded graphite platelets. The clay platelets will thus be presentuniformly throughout the thickness of the char and no decanta-tion or surface migration is observed. The char thus presents highmechanical properties and, therefore, high fire protective behavior.These are illustrated in Fig. 12.

4. Conclusion

In this paper we have investigated the fire performance ofsilicone/expandable graphite/clay in hydrocarbon fire scenario.

ganic

Wmiptsoh

A

ehs

R

[

[

[

[

[

[

[

[

[

[

[

[

[

B. Gardelle et al. / Progress in Or

e demonstrated that the organoclay increases significantly theechanical properties of the char and therefore its cohesion dur-

ng fire testing. This will result in high fire protective behavior. Thishenomenon was explained reasonably by the characterization ofhe material by TEM, WAXS and 29Si NMR evidencing a nanodisper-ion of the fillers and two main interactions: (i) the intercalationf some silicate platelets and (ii) a chemical reaction between theydroxyl groups from the clay with the silicone matrix.

cknowledgements

The authors gratefully acknowledge Mr. Revel for his skillfulxperimental assistance and his expertise in NMR and Dr. Stoclet foris experimental assistance and his expertise in wide-angle X-raycattering.

eferences

[1] L. Gardner, N.R. Baddoo, Fire testing and design of stainless steel structures,Journal of Constructional Steel Research 62 (2006) 532–543.

[2] E.D. Weil, Fire-protective and flame-retardant coatings – a state-of-the-artreview, Journal of Fire Sciences 29 (2011) 259–296.

[3] S. Duquesne, S. Magnet, C. Jama, R. Delobel, Intumescent paints: fire protec-tive coatings for metallic substrates, Surface and Coatings Technology 180/181(2004) 302–307.

[4] D.T. Nguyen, D.E. Veinot, J. Foster, Inorganic intumescent fire protectivecoatings, U.S. Patent, Canada (1989).

[5] B. Gardelle, S. Duquesne, V. Rerat, S. Bourbigot, Thermal degradation and fireperformance of intumescent silicone-based coatings, Polymers for AdvancedTechnologies 24 (2013) 62–69.

[6] B. Gardelle, S. Duquesne, P. Vandereecken, S. Bourbigot, Fire performancesof curable silicone based coatings, in: ACS book Fire and Polymers VI: NewAdvances in Flame Retardant Chemistry and Science, 2012.

[7] B. Gardelle, S. Duquesne, C. Vu, S. Bourbigot, Thermal degradation and fire

performance of polysilazane-based coatings, Thermochimica Acta 519 (2011)28–37.

[8] S. Duquesne, M. Le Bras, S. Bourbigot, R. Delobel, G. Camino, B. Eling, C. Lindsay,T. Roels, Thermal degradation of polyurethane and polyurethane/expandablegraphite coatings, Polymer Degradation and Stability 74 (2001) 493–499.

[

Coatings 76 (2013) 1633– 1641 1641

[9] G. Li, G. Liang, T. He, Q. Yang, X. Song, Effects of EG and MoSi2 on thermal degra-dation of intumescent coating, Polymer Degradation and Stability 92 (2007)569–579.

10] Z. Wang, E. Han, W. Ke, Influence of expandable graphite on fire resistanceand water resistance of flame-retardant coatings, Corrosion Science 49 (2007)2237–2253.

11] B. Gardelle, S. Duquesne, P. Vandereecken, S. Bourbigot, Characterization ofthe carbonization process of expandable graphite/silicone formulations in asimulated fire, Polymer Degradation and Stability 98 (5) (2013) 1052–1063.

12] R. Otáhal, D. Vesely, J. Násadová, V. Zíma, P. Nemec, P. Kalenda, Intumescentcoatings based on an organic-inorganic hybrid resin and the effect of mineralfibres on fire-resistant properties of intumescent coatings, Pigment and ResinTechnology 40 (2011) 247–253.

13] S. Bourbigot, F. Samyn, T. Turf, S. Duquesne, Nanomorphology and reaction tofire of polyurethane and polyamide nanocomposites containing flame retard-ants, Polymer Degradation and Stability 95 (2010) 320–326.

14] S. Ullah, F. Ahmad, P.S.M. Megat-Yusoff, Effect of boric acid with kaolin clay onthermal degradation of intumescent fire retardant coating, Journal of AppliedSciences 11 (2011) 3645–3649.

15] J.H. Koo, P.S. Ng, F.B. Cheung, Effect of high temperature additives in fire resis-tant materials, Journal of Fire Sciences 15 (1997) 488–504.

16] G. Gosselin, Stuctural fire protection predictive methods, in: Proceedings ofBuilding Science Insight, 1987.

17] M. Jimenez, S. Duquesne, S. Bourbigot, Characterization of the performance ofan intumescent fire protective coating, Surface and Coatings Technology 201(2006) 979–987.

18] I.S. Reshetnikov, M.Y. Yablokova, E.V. Potapova, N.A. Khalturinskij, V.Y. Chernyh,L.N. Mashlyakovskii, Mechanical stability of intumescent chars, Journal ofApplied Polymer Science 67 (1998) 1827–1830.

19] S. Filippi, M. Paci, G. Polacco, N.T. Dintcheva, P. Magagnini, On the interlayerspacing collapse of Cloisite® 30B organoclay, Polymer Degradation and Stability96 (2011) 823–832.

20] F. Babonneau, N. Baccile, G. Laurent, J. Maquet, T. Azaïs, C. Gervais, C. Bon-homme, Solid-state nuclear magnetic resonance: a valuable tool to exploreorganic-inorganic interfaces in silica-based hybrid materials, Comptes RendusChimie 13 (2010) 58–68.

21] P.A. Wheeler, J. Wang, J. Baker, L.J. Mathias, Synthesis, Characterization ofcovalently functionalized laponite clay, Chemistry of Materials 17 (2005)3012–3018.

22] S. Bourbigot, M.L. Bras, F. Dabrowski, J.W. Gilman, T. Kashiwagi, PA-6 clay

nanocomposite hybrid as char forming agent in intumescent formulations, Fireand Materials 24 (2000) 201–208.

23] S. Duquesne, C. Jama, M. Le Bras, R. Delobel, P. Recourt, J.M. Gloaguen, Elabo-ration of EVA-nanoclay systems – characterization, thermal behaviour and fireperformance, Composites Science and Technology 63 (2003) 1141–1148.