Fire-Retardant Performance of Intumescent Coatings Using Halloysites as a novel Fire-Retardant Additive
Corresponding author: Dr. Javier Sacristan Bermejo, Acciona Technological Center, ACCIONA Infraestructuras S.A, C/ Valportillo Segunda, Nº8, Polígono Industrial Alcobendas, 28108 Alcobendas (Madrid), Tel: (+34) 91.791.20.20
Fibre-reinforced polymer composites (FRP) are widely used in several industrial sectors, for example, building, electrical, and transportation, due to their excellent mechanical properties, easy processing, low density, and good corrosion resistance. However, the typical resins used to hold the fibre in place and transfer the load from fibre to fibre are highly flammable and produce large amount of smoke and toxic gases during burning. Therefore, to widen their commercial use it becomes mandatory the development of novel flame-retardant systems to reduce fire risks.
Intumescent fire resistant coatings which do not have any effect on the mechanical performance of the laminate as they are applied at the surface, would represent a good choice and will widen the existing range of applications in architecture and transportation, offering a considerable higher safety standard. Intumescence is the result of a sequential chemical reaction between several chemical products activated by the heat of the flames during the fire. The final result of the chemical reaction is a multi- cellular foam of low thermal conductivity. The outer surface comprises a charred layer acting as a physical barrier to the influx of radiant heat and oxygen, reducing the smoke
and the heat transmission. The chemical compounds of intumescent systems are classified into four categories: a carbonization agent, a carbon-rich polyhydric compound that influences the amount of char formed and the rate of char formation; an acid source, a foaming agent, which during their degradation release non-flammable gases such as CO2 and NH3 and the polymeric binder that holds everything together. Novel additives with slightly altered properties can be added to the formulation to optimize the fire-retardant performance. Some examples include urea which can replace melamine as foaming agent to improve water solubility. The ammonium salts of orthophosphoric acid have been used as additives for flame retardants (FR) for decades as acid source, currently it has been substituted by ammonium polyphosphate (APP), the ammonium salt of polyphosphoric acid. In the same context melamine, melamine cyanurate and other melamine salts are also currently used to decrease activation temperatures and enhance fire-retardant performance by a synergistic behavior between melamine and the corresponding salt anion .
Finally, incorporation of nanoparticles to polymers has attracted a lot of attention as a simple and cost effective method of enhancing polymer properties by the addition of a small amount of properly designed and dispersed nanometer fillers. This trend has also been followed in the development of novel coatings with added functionalities [4,5]. Several studies have shown that the mechanical and fire properties of intumescent coatings can be improved when nanoparticles are added owing to increased thermal stability of the char in the presence of nanoparticles [4,5]. Among the different nanoparticles used in flame retardancy, organo-clays such as montmorillonites, and double layer hydroxides have attracted much attention in the literature. These nano-coatings exhibit excellent properties that are different from those coatings in which the fillers are added to the formulation at micrometer level. As for other nanoparticles, their effect as flame-retardant has been linked to their dispersion state within the binder and their modifications [6,7]. The other types of naturally occurring clay materials are halloysite nanotubes which are unique and versatile nanomaterials composed of a double layer of aluminium, silicon, hydrogen, and oxygen. Halloysite nanotubes are ultra-tiny hollow tubes with diameters typically smaller than 100 nanometers and lengths typically ranging from about 500 nanometers to over 1.2 microns. Currently, the main areas of application of halloysite nanotubes are, as additives in polymers, in electronic components, in pharmacy as drug delivery vehicles, and in cosmetics. Halloysites are biocompatible and easily modified through cation exchange reactions. They are also highly promising flame retardants since their application perspectives are connected with a series of a unique combination of properties . The thermal stability and flame-retardant effects of hollow nanotubes (HNTs) are believed to result from the hollow tubular structures of HNTs, the barriers for heat and mass transport and the presence of
iron in the HNTs. HNTs are promising as non-halogen flame- retardant fillers in several thermoplastics such as, PP, PA6, and LDPE among others . Halloysite-coatings are very attractive in practical applications because of advantages such as permanent stability under UV exposure, high anti-corrosion, and chemical stability at high temperatures and enhancement of mechanical properties.
The present study reports the preparation and characterization of flame-retardant coatings with modified and non-modified halloysites in combination with ammonium polyphosphate, pentaerythritol, and melamine cyanurate for protecting carbon fibre-epoxy laminates. Details of the preparation of the halloysite-water-based coating, most relevant coating properties such as viscosity, adhesion, morphology, and drying times are studied as a function of halloysite concentration. Finally, fire-retardant performance of halloysite coatings applied on CRF epoxy laminates is evaluated by means of a cone calorimeter following UNE5660 standard.
Intumescent systems were formulated based on an in house intumescent recipe. The main components are shown in Table
1. APP phase II were obtained from Budenheim, melamine cyanurate was purchased from OMIA, and pentaerythritol was sourced from Perstop. Titanium dioxide and resin binder were supplied by Huntsman and Wacker, respectively. A copolymer of PVA-PVAc, Vinapas, was used as a binder due to their decomposition mechanisms yielding stable carbonaceous char structures which would aid the protection of the underlying material and their compatibility with many additives. Halloysite and modified halloysite were supplied by Applied mineral inc.
Table 1. Intumescent formulation.
|Resin binder||Water based (20% PVAac)|
|Acid source||Amonium poliphosphate phase II|
|Carbon source||Pentaeritriol (PER)|
|Blowing Agent||Melamine Cyanurate|
|FR Additive||Halloysite||Halloysite II|
Paints were mixed by high speed, high shear mixing instrument, Dispermat LC55, to ensure that all of the components are adequately incorporated. Each coating was subjected to preliminary testing; dispersion can be observed by spreading a thin layer of the coating onto a substrate. Poor stability is easily observed as the coating will clump, agglomerate, become gritty, and separate out leaving water and sediment
due to settling and incompatibility of components. Stability is of great importance as fire retardance is of limited use if the coating does not function primarily as a coating.
Carbon-fibre-reinforced laminates were prepared as follows, five-ply oven-cured carbon-fibre-reinforced epoxy laminates were manufactured by resin infusion. Laminates are composed of 0/90 carbon fibre plain weave from Toray and an epoxy resin from Huntsman, 1568LY with hardener Aradur© 3489. Samples were prepared in duplicate following the same experiemental procedure to ensure consistency and reproducibility. The fibre mass fraction is 58% and the average thickness 4.3mm.
2.1.1. Composite preparation
Carbon-fibre-reinforced laminates were manufactured by a vacuum infusion process (VIP). The fibrous reinforcement was laid onto an open-faced steel plate. Then, a release film was placed on top of the mold up to the vacuum tube, followed by the bleeder material that absorbs the resin exceeding during the infusion, assuring a good vacuum distribution. Then, a resin distribution medium was placed on the top and partially in contact with the heating plate, useful for helping the resin to impregnate all the plies. Finally, after placing a peel ply, a nylon vacuum bag was put on the layers to cover the entire plate and stitched to the plate by means of a high-temperature resistant adhesive. The curing schedule that followed consisted of 4h at 25ºC and 6h at 100ºC.
2.2. Testing equipment
Morphological and elemental composition analysis of the coatings were carried out by means of Scanning Electron Microscopy (SEM) coupled with X-Ray Energy Dispersive Spectroscopy (EDS). SEM images provide detailed information of the surface morphology, while EDS spectra gave qualitative and semi-quantitative information of the elements present. Two scanning electron microscopes were employed by LabCyP- UCA: thermionic emission FEI Quanta 200 with a Phoneix EDX analysis system and a Field Emission Gun FEI Nova NanoSEM 450.
The cone calorimeter experiments were carried out on the coated fibre-reinforced composite materials by using a Fire Testing Technology Ltd instrument according to UNE 5660 procedure.
Samples, with nominal dimensions of 100 x 100 x 5-7 mm,
were tested horizontally under an incident flux of 50 kW/m2. Generally, results from cone calorimeter are considered to be reproducible to ± 10% for experiments run in triplicates . However, in this study experiments were executed in duplicates due to additive’s availability.
The flame-retardant coatings were applied in TITANIA by spreading 50g of fire-retardant paint with a multiple clearance square applicator from BYK onto the CFR laminates sized 100×100 mm. Both the wet thickness and the dry thickness were measured employing a wet film thickness gauge and a micrometer, respectively.
Paint adhesion to carbon fibre laminates was evaluated following several European and USA standards. More specifically, the standards chosen were: spheres UNE-EN- ISO-91117-3: 2010, Tack free ASTM D1640-03, and Dry hard ASTM D1640-03.
Water-based paints generally show a higher intumescence than solvent based paints. The fire retardant performance of the intumescent paint depends on the choice of the ingredients and their appropriate combinations. In a previous work (not published yet), it was found that the use of melamine cyanurate instead of neat melamine produces thicker intumescence, with more uniform and closed cell structure, and no cracking. On the basis of our previous results two different types of inorganic halloysite nanotubes, were added to an intumescent water based paint at 5–15 wt% with respect to the pure resin, see Table 1. Paint morphology was investigated by means of SEM- EDS and the fire-retardant properties of the produced coatings applied onto carbon fibre laminates was investigated by cone calorimeter tests. The main objective of the intumescent coating is to block heat transferring to the substrate and protect it from fire. The expanding effect and char structure are important and are intimately related to the additive dispersion within the formulation. The SEM micrographs of coatings taken before testing are shown in Figures 1–4. Three filler concentrations, 5 wt%, 10 wt%, and 15 wt%, were studied. In terms of homogeneity, the dispersion of both the unmodified and modified halloysite was in general good for all the concentrations The absence of big particle aggregates was the main result.
A closer examination of SEM images allows to determine the effect of the addition of inorganic filler on the coating morphology. As can be observed in figure 1, coating M1 which
does not contain any filler has a granular structure and can visually identify the presence of particles of APP, TiO2, and melamine. By incorporating different amounts of halloysite, the coating morphology varies. The clearly distinguishable granular morphology changes toward a laminar/flaky structure which is directly related to the inorganic nanotubes embedded in the matrix. Addition of inorganic filler modifies the coating morphology. At higher magnifications employing FEG-SEM microscope, see figure 3, groups of nanotubes of approximately 50-150 nm of diameter and up to 500 nm of length can be observed. The nanotube structureof the halloysite remained after the high speed mixing process and the drying process. These results indicate that the nanotube structure did not collapse under the mixing conditions, which were carefully chosen to avoid their destruction. EDS elemental spectra also support the presence of areas richer in Al, Si, and O from the halloysite compared to the reference sample M1. The intensity of the EDS signals depend on the area studied. Figures 1, 2 and 4 contain representative spectra of the corresponding group of samples.
Figure 2. SEM Images and EDS spectra of coatings containing a) 5 wt%, b) 10 wt%, and c) and d) 15 wt% of non-modified halloysite, samplesM2, M3 and M4 respectively.
Coating drying times and adhesion
Figure 1. a) and b) SEM Images and c) EDS spectra of reference coating (no halloysite)(M1).
It is well known that coating drying time is of paramount importance for film formation. A too short or too long drying time can be the main reason for causing defects and adhesion failures to the substrate. The process of drying involves several physical and/or chemical changes, such as solvent evaporation, oxidation, and polymerization, all of which are time dependent. There exist several testing methods which give information regarding different drying stages. In this work, drying time was evaluated through three methods: spheres (UNE-EN- ISO-9117-3: 2010), tack free and dry Hard (ASTM D1640-03). The drying time calculated by means of the spheres method indicates the time during which any work that can generate particles or chips near the painted surface should be avoided, since they will be occluded if they eventually fell on the surface. The time of Tack Free reflects the interval during which the
surface must not be touched and small dust particles can still become attached to the surface. Finally, the Dry Hard is related to the time during which, although the surface is dry, it must not suffer mechanical solicitations, because the layer could be eventually removed.
Figure 3. FEG-SEM Images of coatings containing 15 wt% of non-mod- ified halloysite. (M4)
Thus, the spheres method gives the lowest drying time values. They are very similar among all coatings and the presence of halloysite does not seem to produce any effect. Tack free and Dry hard times are also very similar except in the formulation containing the highest concentration of modified halloysite in which drying time increases up to 20%. Apart from those small differences, drying times of the novel intumescent coating are found to be in the same range of those of commercially available coatings. So, there are no excepted issues regarding its application.
Figure 4. SEM Images and EDS spectra of coatings containing a) 5 wt%, b) 10 wt%, and c) and d) 15 wt% in weight of modified hal- loysite, samplesM5, M6 and M7 respectively.
On the other side, a successful fire protective surface coating relies on cohesion between the film-forming substances and adhesion between the coating and the substrate. A failure in adhesion results in blistering and delamination of the coating. There are several methods for adhesion testing used by product manufacturers to determine the level of protection required in individual cases. One of the most common methods is the cross cut test. This test is used to test the adhesion of cured coats of paint on their substrate by means of a series of cuts through the coating. A lattice pattern with eleven cuts in each direction is made in the film down to the substrate; adhesive tape is applied over the lattice and then peeled off. The result is evaluated by using a table chart. This method determines the suitability of the coating to be applied onto the existing substrate. Main results areshownin Table 2.
Table 2. Coating drying times and adhesion values of coatings deposited onto carbon fibre laminates.
|Sample name||Drying time(min)||Thickness (μm)||Adhesion
|M2 M3 M4 M5 M6
The adhesion was evaluated by comparison with descriptions and illustrations from the standard, rated in a scale from 0–5, 0 for the best adhesion (no observable detachments) and 5 for the worst adhesion (detachment higher than 65% of the area tested). The adhesion results in Table 2 stand for: 1C is the test method, x/x/x are the results in the scale from 0–5 obtained in three different areas. As can be observed, all coatings display similar results indicating that paint adhesion to the carbon fibre laminate does not depend on either halloysite presence or concentration.
The fire performance of CFR epoxy composites with or without a protective intumescent coating (with and without halloysite) was determined using cone calorimetry; Before testing, the composite panels have been cut to comply with the standard size (100 mm x 100 mm) of the ISO 5660 for the cone calorimeter tests and conditioned for 24h under vacuum at room temperature.
HRR curves and sample images taken at the end of the test are shown in Figures 5–7, a) and b), respectively, and cone calorimetry results depicted in Table 3.
TTI is not very strongly affected by the presence of additives but peak HRR, THR, and MARHE as well as total smoke production is different for all samples depending upon the halloysite content and grade. TTI decreases for the unprotected laminate in comparison to the coated ones. This is associated to a poor resin outer layer of the laminate; there is less material to burn out, especially during the first few minutes of the test. Coated samples ignite with the pilot spark at shorter times, but in most of the cases the 10-minute test finishes before the sample is fully carbonized. The burning rate of coated samples is slower and the heat released is lower than in both the uncoated samples and coated samples without halloysite.
The HRR of the unprotected composite, Figure 5b, increased rapidly after the initial 75s with a peak of 450 kW/m2, followed by a second higher peak of 730kW/m2 which occurred at 150s.
Table 3: Cone calorimetric results for coated and uncoated carbon fibre composite samples at 50 kW m-2 heat flux.
TTI Peak HRR (kW m-2)
HRR (kW m-2)
Total smoke production
Subsequently, the HRR decreased rapidly until it became negligible after 275s. At the end of the experiment no residual char was available, losing the composite all its structural integrity.
The shape of HRR curves depends on whether the CF laminate is coated or not. In the uncoated sample, the occurrence of two distinct peaks should be related to the initial burning of the composite surface and its propagation through the thickness that take place with different HRRs for the different distribution of the resin between the plies and within the plies. Coated samples also show a double peak HRR curve. M1 (without halloysite) shows a less pronounced first peak (290 kW/m2) followed by a slightly more intense second peak (355kW/m2) which may be due to the rapid formation of a char layer at the surface with a very limited insulation capacity. Halloysite- containing coatings (M2-M7) show a distinct behavior; the first peak (90-100kW/m2) is up to three times less intense than the second one and it is followed by a sharp drop in HRR. These results may be attributable to the formation of a thicker insulating char layer preventing a rapid increase in the core material temperature during the test. The second peak at 250- a370 kW/m2 is due to the rise of surface temperature caused by the failure of the thermal barrier efficiency offered by the expanded char. It leads to the deterioration of the composite structural integrity.
Figure 6. a) Pictures taken after the cone calorimeter test of CFRP laminates coated with the reference intumescent formulation (non-halloysite in the formulation) and b) HRR curves versus time.
In general all coated samples show a significant decrease of the value of PHRR with respect to the unprotected sample, M0. The improvement in fire resistance as measured by the reduction in PHRR when compared to an unprotected composite (PHRR = 747 kW/m2) follow the pattern; M2 > M4
>M5 > M6 > M3 >M7.However, there is not a clear relationship between PHRR and halloysite concentration. Moreover, samples protected by halloysite-intumescent coatings tend to result in longer total burning time than that of the reference sample M1. Halloysite-intumescent coatings are capable of providing a more effective insulative thermal barrier to last during the entire cone calorimeter test.Regarding fire safety, the maximum average rate of heat emission, MARHE, can be considered as a good measure of the propensity for fire development under real scale conditions. A high MARHE value indicates an increase of the hazard level of the material. Coated composites show a notable reduction of MARHE, up to a 60% with respect to the uncoated sample, while the incorporation of higher concentration of halloysite or modified halloysite did not show any further improvement.
Figure 5. a) Pictures of laminates without coating at the end of the test and b) corresponding HRR curves versus time.
Figure 7. Heat release rate (HRR) versus time and images taken at the end of the test for the composite protected with intumescent coatings containing halloysite I a) 5 wt%, b) 10 wt%, c) 15 wt%, samples M2, M3, and M4 respectively.
When these intumescent coatings begin to burn and decompose, the different components, ATH, APP, melamine, pentaerythritol, and halloysite go through a series of chemical reactions forming networks, or organoceramic layers which may act as a barrier to fuel transport. They slow down the release of fuel from the decomposing coating to the flame front. Decomposition products from both coating and CF laminate now have longer residence times in the condensed phase and, as such, are more likely to form char. The main consequence of a slower burning rate is that, large amounts of smoke evolve from these systems which is known to be hazardous in fire situations. Smoke production, Table 3, varies according to the burning behaviour of the sample. The amounts of smoke evolved were consistently lower for the samples containing halloysite than those corresponding to other fire-retardant additives. As can be seen, halloysite concentration did not significantly influence the amounts of CO evolved during combustion which generally follows the heat release curve. Longer burning time results in an increase in combustion products which is typical for incomplete combustion, with the corresponding increase in CO and smoke production.
Figure 8. Heat release rate (HRR) versus time and images taken at the end of the test for the composite protected with intumescent coatings containing modified halloysite at a) 5 wt%, b) 10 wt%, c) 15 wt% , corresponding to samples M5, M6, and M7 respectively.
Finally, pictures taken at the end of the test show relevant dif- ferences among protected and un-protected composites. The unprotected laminate, see figure 5, did not have any residual char; the resin was completely burned out and only the car- bon fibre was left, consequently the laminate lost its structural integrity. Meanwhile, in the laminates protected by intumes- cent coatings, Figures 6, 7 and 8, some black swelled residual char can be observed confirming the efficiency of the intumes- cent coating in foaming and expanding the composite surface. From these images of residual char, halloysite coatings present the largest residual char expansion having a more stable and porous char structure than M1 (without halloysite). The ob- served char characteristics are consistent with the higher fire retardant efficacy observed for these coatings.,
In order to widespread the use of CFRP composites, development of novel intumescent coatings for protecting them against fire is essential to avoid compromising fire safety.
a) Developed coatings with different concentration of halloysites have shown good adhesion to CFRP laminates and drying times in the same range than commercial paints. SEM-EDS images have shown the presence of the inorganic nanotubes within the dry paint. It confirms that the high shear mixing method used did not destroy the nanotube structure. The cone calorimeter results evidenced a more effective fire-retardant action of the halloysite-intumescent coatings by reducing the mass lost during burning, reducing the HRR, owing to an increase in the thermal stability of the char in the presence of inorganic nanotubes. A reduction of PHRR (21%), MARHE (10–20%), and total smoke (10–15%) for the protected CFRP composite were observed. These results suggest that halloysite is efficient as a fire-retardant additive due to the capability to form a good-quality char. In spite of these promising results, more experiments are necessary to better understand the role of halloysite as a fire-retardant additive in intumescent paints.
This work was supported by the project “Nuevas Tecnologías de refuerzo con materiales compuestos” (NUREMCO) funded by CDTI / FEDER INNTERCONECTA.
- Tavares, S. M. O., Camanho, P., and De Castro, P. M. S. T. (2013) Assessment of materials for fuselage panels considering fatigue behavior. 265-270.
- Bourbigot S, Le Bras M. Duquesne S and Rochery M. Recent advances for intumescent polymers. Macromolecular Materials and Engineering.2004, 289(6):499-511..
- Patrick Lima WK,Mariattia M,Chowa WS,Mar KT.Effect of intumescent ammonium polyphosphate (APP) and melamine cyanurate (MC) on the properties of epoxy/glass fiber composites. Composites Part B: Engineering .2012, 43(2):124– 128.
- Ye Xiaoxin, Tse Zion T. H., Tang Guoyi, Song Guolin. Effect of electroplastic rolling on deformability, mechanical property and microstructure evolution of Ti-6Al-4V alloy strip. Materials Characterization, 2014,98(0):147-161.
- Ye Xiaoxin, Tse Zion T. H., Tang Guoyi, Geng Yubo, Song Guolin. Influence of electropulsing globularization on the microstructure and mechanical properties of Ti-6Al-4V alloy strip with lamellar microstructure. Materials Science and Engineering: A, 015, 622 (0): 1-6.
- Ye Xiaoxin, Yang Yanyang, Tang Guoyi. Microhardness and corrosion behavior of surface gradient oxide coating on the titanium alloy strips under high energy electro-pulsing treatment. Surface and Coatings Technology. 2014,258(0):467- 484.
- Ye Xiaoxin, Wang Lingsheng, Tse Zion T. H., Tang Guoyi, Song Guolin. Effects of high-energy electro-pulsing treatment on microstructure, mechanical properties and corrosion behavior of Ti–6Al–4V alloy. Materials Science and Engineering: C.2015,49(0):851-860.
- Mingxian L., Zhixin J., Demin J., Changren Z. Progress in Polymer Science.Recent advance in research on halloysite nanotubes-polymer nanocomposite.2014,39(8):1498–1525,
- Marneya DCO, Russella LJ, Wua DY,Nguyena T,Crammb D et al.suitability of halloysite nanotubes as a fire retardant for nylon 6. Polymer Degradation and Stability Oct 2008, 93(10):1971–1978.
- Gilman JW, Kashiwagi T, Nyden M, Brown JET, Jackson CL et al. In Chemistry and technology of polymer additives. Eds. l-Maliaka S, Golovoy A, Wilkie CA. London Blackwell Scientific; 1999. 249- 265.