Chapter 24: Polymer–carbon nanotube composites for flame-retardant cable applications – Polymer-Carbon Nanotube Composites


Polymer–carbon nanotube composites for flame-retardant cable applications

G. Beyer,     Kabelwerk Eupen AG, Belgium


This chapter describes multi-walled carbon nanotubes (MWCNTs) as very efficient flame retardants at low filler contents in ethylene vinyl acetate (EVA). An optimized formulation for flame-retardant insulated wires was developed based on the filler-blend ‘multi-walled carbon nanotubes–organoclays–aluminium trihydrate’. The char was improved and strengthened by the long length/diameter ratio of the MWCNTs, resulting in better flameretardant performances of the wires.

Key words

aluminium trihydrate


carbon nanotubes

char formation



24.1 Introduction

Fire hazards are mainly the combination of different factors including ignitability, ease of extinction, flammability of volatiles generated, amount of heat released on burning, rate of heat release, flame spread, smoke obscuration, and smoke toxicity. The most important factors are the rate of heat release, rate of smoke production, and rate of toxic gas release.1 An early high rate of heat release causes a fast ignition and flame spread; furthermore, it controls the fire intensity and is much more important than ignitability, smoke toxicity or flame spread. The time for people to escape from a fire is controlled by the heat release rate.2 Once a fire starts in a room containing flammable materials, it will generate heat, which can heat up and ignite additional combustible materials. As a consequence, the rate at which the fire progresses increases, because more and more heat is released and a progressive increase of the room temperature is observed. The radiant heat and the temperature can rise to such an extent that all the materials within the room will be easily ignited, resulting in an extremely high rate of fire spread. This point in time, termed flash-over, leads to a fully developed fire. Escape from the room will then be nearly impossible, and spread of the fire to other rooms is very likely. When a fire goes to flash-over, all polymers in the fire will release roughly 20% of their weight as carbon monoxide, resulting in too much toxic smoke. Therefore, most people die in big fires, and 90% of fire deaths are the result of fires becoming too big, resulting in too much toxic smoke.3 Fire statistics report more than 12 million fires every year in the United States, Europe, Russia, and China, killing roughly 160,000 people; several hundreds of thousands of people are injured every year. Each year about 5000 people are killed by fires in Europe and more than 4000 people die in the USA. Direct property losses by fires are roughly 0.2% of the gross domestic product in Europe and the total costs of fires are around 1% of the gross domestic product.4 Therefore, it is important to develop well-designed flame-retardant materials to decrease both fire risks and fire hazards and save lives.

Polymers are being used in more and more fields of applications, and specific mechanical, thermal, and electrical properties are required. One further important property is the flame-retardant behavior of polymers, which can be fulfilled traditionally by using intrinsically flame-retardant polymers like poly(vinyl chloride) (PVC) or fluoropolymers, and by flame retardants like aluminium trihydrate (ATH), magnesium dihydroxide (MDH), organic brominated compounds, or intumescent systems based on nitrogen- or phosphorus-based compounds to prevent the burning of such polymers like polyethylene (PE), polypropylene (PP), poly(ethylene-co-vinyl acetate) (EVA), polyamide (PA) or other polymers. However, these flame retardants sometimes exhibit serious disadvantages. Use of aluminium trihydrate and magnesium hydroxide in flame- retardant cables requires a very high portion of these fillers for the applied polymers EVA, PE or PP; filling levels of more than 60 wt% are necessary to achieve a suitable flame retardancy. Clear disadvantages of these filling levels are the high density and the lack of flexibility of end products, the poor mechanical properties, and the problematic compounding and extrusion steps. In Europe, at least, there are reservations about the general use of brominated compounds as flame retardants. Intumescent systems are relatively expensive and electrical requirements can restrict the use of these products.

A new class of materials, called nanocomposites, avoids the disadvantages of the traditional flame-retardant systems. Generally the term nanocomposite describes a two-phase material with a suitable nanofiller (usually modified layered silicates like montmorillonites (organoclays) or carbon nanotubes) dispersed in the polymer matrix at the nanometer (10−9 m) scale. Compared with pure polymers, the corresponding nanocomposites show tremendous improvements; the content of nanofillers within the polymer matrix is usually between 2 wt% and 10 wt%. The most important properties improved by nanocomposites are mechanical properties such as tension, compression, bending, and fracture; barrier properties like permeability; and solvent resistance, translucence, and ionic conductivity. Other highly interesting properties exhibited by these polymer nanocomposites concern their increased thermal stability and flame retardancy at very low filler levels.57 The low filler content in nanocompos- ites for improved thermal stability is highly attractive to the industry because the end products can be made cheaper and easier to process. Beyer discusses these improvements by nanocomposites in detail.8

Beyer9 reported at an international flame retardancy conference in 2002 for the first time on the flame retardancy of carbon nanotubes. The first convincing explanation of the flame-retardant behavior of carbon nanotubes was proposed then by Kashiwagi et al.1012 Several of these papers described how single- and multi-walled carbon nanotubes enhance the thermal stability of polymers without using any organic treatment or additional additives. The carbon nanotubes were at least as effective flame retardants as organoclays. PP and PMMA were investigated and the dispersion of the nanotubes within the polymer matrix was the important key parameter for good flame retardancy. Kashiwagi reported that the ideal structure of a protective surface layer (consisting of clay particles and some char) was net-like and had sufficient physical strength not to be broken or disturbed by bubbling. The protective layer should remain intact throughout the entire burning period. The formation of a continuous, network-structured protective char was easiest with high aspect ratio nanoscale particles. Kashiwagi et al.13 pointed out that in general, a variety of highly extended carbon-based nanoparticles, such as single- and multi-walled carbon nanotubes as well as carbon nanofibres, will form this kind of a network if the nanofillers formed a network structure in the polymer matrix, so that the material as a whole behaves rheologically like a gel. Also Schartel et al.14 reported on the flame retardancy of MWCNTs in PA-6; again, the increased melt viscosity of the nanocomposites and the fibre-network character were the dominant factors influencing the fire performance.

This chapter will review in detail results of nanocomposite-based organoclays and carbon nanotubes and the synergistic effects of these fillers with micro-sized aluminium trihydrate as a traditional flame retardant for cable applications.

24.2 Carbon nanotube-based nanocomposites

24.2.1 General properties of carbon nanotubes

Carbon nanotubes (CNTs) are tubular derivatives of fullerenes. They were first observed in arc discharge experiments, and exhibit properties, which are quite different from those of the closed cage fullerenes such as C60, C70 and C76, etc. Special topologies are responsible for the unique and interesting properties of CNTs. As novel carbon materials, CNTs are of great interest in the field of material science research. Due to their high mechanical strength, capillary properties, and remarkable electronic structures, a wide range of potential uses is reported. Typical applications for CNTs are supports for metals in the field of heterogeneous catalysis, material for hydrogen storage, as composite materials in polymer science15, 16 and for immobilization of proteins and enzymes. Several techniques like arc discharge, laser ablation and catalytic methods and others have been developed for the production of CNTs.16 Many material science researchers and also companies (Nanocyl, Hyperion Catalysis International, Bayer Material Science, Arkema) are working on the development of methods for large-scale production of CNTs to realize their speculated applications. CNTs can consist of one (single-walled carbon nanotubes, SWCNTs) or more (multi-walled carbon nanotubes, MWCNTs) cylindrical shells of graphitic sheets. Each carbon is completely bonded to three neighboring carbon atoms through sp2 hybridization to form a seamless shell. CNTs can have a very high aspect ratio, above 1000. It is reported17 that polymer degradation can be retarded by CNTs as measured by thermogravimetric analysis. In PVOH, a loading level of 20 wt% MWCNTs shifts the polymer degradation to higher temperatures.

24.2.2 Synthesis and purification of CNTs

Crude MWCNTs and SWCNTs were produced by catalytic decomposition of acetylene on Co-Fe/Al(OH)3 catalysts.18 The CNTs contained catalysts and other by-products. The catalysts and support contents of the CNTs samples are shown in Table 24.1. Crude MWCNTs contained Co, Fe and alumina. Purified MWCNTs were synthesized from crude MWCNTs by dissolution of the catalyst support in concentrated NaOH, dissolution of the metal catalyst in concentrated HCl, drying at 120 °C in an air oven and additionally drying at 500 °C under vacuum. Crude SWCNTs contained Co and MgO. Purified SWCNTs were synthesized from crude SWCNTs by dissolution of the catalyst support in concentrated HCl, purification by air oxidation at 300 °C and then drying at 120 °C in an air oven.

Table 24.1

Properties of MWCNTs

24.2.3 Flammability of poly(ethylene-co-vinyl acetate) multi-walled carbon nanotube (EVA–MWCNT) compounds and EVA–MWCNT–organoclay compounds

It was possible to investigate for the first time worldwide the flame-retardant properties of carbon nanotube compounds by cone calorimeter.9, 18 All compounds were melt-blended in a Brabender mixing chamber. EVA was molten within 3 minutes; then MWCNTs were added and mixed for an additional 7 minutes. The compounds were pressed to plates at 180 °C, 200 bar for 5 minutes. It is evident from the results in Table 24.2 that all the filled polymers had improved flame-retardant properties. For EVA and the EVA-based nanocomposites containing 2.5 phr of filler, the PHRR decreased as follows: EVA > organoclays > purified MWCNTs. For EVA and EVA-based composites containing 5.0 phr filler, the peak heat release rate (PHRR) decreased as follows: EVA > organoclays > purified MWCNTs = crude MWCNTs. Crude MWCNTs were as effective in the reduction of PHRR as purified MWCNTs. Increasing the filler content from 2.5 to 5.0 phr caused an additional flame-retardant effect and this improvement was most significant when purified or crude MWCNTs were used. A synergistic effect for flame retardancy between MWCNTs and organoclays was observed for the nanocomposite containing 2.5 phr of purified MWCNTs and 2.5 phr of organoclays (Fig. 24.1). The latter sample was found to be the best flame-retardant compound. The variation of the screw velocity from 45 rpm (sample A) to 120 rpm (sample B) did not change the flame-retardant properties for the composites containing 5.0 phr of crude MWCNTs. There was also no reduction in time to ignition for the MWCNT-based EVA composite, in contrast to the only organoclay-based EVA composite, which undergoes an early thermal degradation of the quaternary ammonium compound within the galleries of the organoclay.19

Table 24.2

Peak of heat release rates at heat flux = 35 kW/m2 for various compounds with organoclays and multi-walled carbon nanotubes

Note: EVA: Escorene UL 00328 with 28 wt% vinyl acetate content.

Organoclay: Nanofil 15 (dimethyl ditallow exchanged montmorillonite).

aThe screw velocity was 45 rpm and the mass temperature was 136 °C.

bThe nanotubes and the organoclay were premixed before their addition.

cThe screw velocity was 120 rpm and the mass temperature was 142 °C.

24.1 Peak of heat release rates for various EVA-based compounds with organoclays and multi-walled carbon nanotubes. A, EVA + 5.0 phr organoclays. B, EVA + 5.0 phr pure MWCNTs. C, EVA + 2.5 phr organoclays + 2.5 phr pure MWCNTs. Organoclay: Nanofil 15 (dimethyl ditallow exchanged montmorillonite). Heat flux = 35 kW/m2. (From reference 18; John Wiley & Sons Limited. Reproduced with permission.)

24.2.4 Crack density and surface results of charred MWCNT compounds

For flame-retardant EVA-based composites containing 5 phr of fillers (Table 24.2), the crack density increased in the order of purified MWCNTs < organoclays. A very important synergistic effect reducing the crack density to a very low level was observed for the nanocomposite containing both 2.5 phr purified MWCNTs and 2.5 phr organoclays (Figs 24.2 (a)–(c)). The synergistic effect for improved flame retardancy by the filler combination MWCNTs and organoclays can be explained by the improved closed surface. The char acted as an insulating and non-burning material with reduction of emission of volatile products (fuel) into the flame area. The fewer cracks present, the better was the reduction of emission of fuel and therefore the reduction of PHRR. The fillers played an active role in the formation of this char, but obviously the MWCNTs, with their long aspect ratio, could strengthen it and made it more resistant to mechanical cracking.

24.2 (a) EVA with 5 phr organoclays after combustion. From reference 18; John Wiley & Sons Limited. Reproduced with permission. (b) EVA with 5 phr pure MWCNTs after combustion. From reference 18; John Wiley & Sons Limited. Reproduced with permission. (c) EVA with 2.5 phr pure MWCNTs and 2.5 phr organoclays after combustion. EVA: Escorene UL 00328 with 28 wt% vinyl acetate content. Organoclay: Nanofil 15 (dimethyl ditallow exchanged montmorillonite). From reference 18; John Wiley & Sons Limited. Reproduced with permission.

24.2.5 Flammability of low-density polyethylene (LDPE) carbon nanotube compounds

Compounds of SWCNTs and MWCNTs in LDPE (BPD 8063 from INEOS) were melt blended in a Brabender mixing chamber according to the formulations indicated in Table 24.3 and Table 24.4. The corresponding cone calorimeter measurements are shown in Figs 24.3 and 24.4. The results from the cone measurements of different carbon nanotubes in LDPE demonstrated that SWCNTs did not act as flame retardants in LDPE. MWCNTs acted as flame retardants in LDPE with no reductions on time to ignition (in contrast to organoclays), and crude MWCNTs showed similar reductions for the PHRR as purified MCNTs.

Table 24.3

PDE filled with SWCNTs

Table 24.4

LDPE filled with different MWCNTs

24.3 Heat release rates for SWCNTs in LDPE. Heat flux = 35 kW/m2. From reference 18; John Wiley & Sons Limited. Reproduced with permission.

24.4 Heat release rates for MWCNTs in LDPE. Heat flux = 35 kW/m2. From reference 18; John Wiley & Sons Limited. Reproduced with permission.

24.3 Cable with the multi-walled carbon nanotube (MWCNT)–organoclay–aluminium trihydrate (ATH) flame-retardant system

It was of interest to transform the results for the synergistic filler-based FR system ‘MWCNTs–organoclays’ into real products.20 To produce a flame-retardant insulated wire by a real cable production extruder, a minimum of approximately 60 kg of compound was needed to fill the extruder and to run a small insulated wire production. It was checked whether the filler blend system MWCNTs–organoclays could be transformed to a real cable compound without processing problems.20 A well-running only organoclay-based cable compound named compound 1 was used, and the weight ratio between these two fillers was changed stepwise. The sum of both fillers always remained constant (Table 24.5) within the cable compounds. Compounding was done on a twin-roll mill, and the reductions of PHRRs for the three compounds were measured (Fig. 24.5). The results clearly indicated that for the filler blend and the only MWCNTs-based compounds, the first PHRR was reduced maximally. The second PHRR was observed at the longest times for the two compounds 2A and 2B, indicating that the chars were less cracked (more stable in time). Therefore a 1:1 blend of MWCNTs and organoclays was used (compound 2A, see Table 24.5 for the cable compound used). This allowed production of the required quantity of the cable compound. Compounding of the formulation 2A was done on an 11 L/D BUSS Ko-kneader with a 46 mm screw diameter, and 60 kg were produced without any processing problems. Processing on the BUS S Ko-kneader improved the PHRRs compared to those with the corresponding twin-roll mill compounding (Fig. 24.6), may be due to a better dispersion of the fillers.

Table 24.5

Blends of MWCNTs and organoclays in cable compound formulations

Compound Composition
1 Technical cable compound (EVA/PE/ATH/organoclay/processing additives)
2A Same formulation as compound 1, but substitution of 50% organoclay by the same amount of MWCNTs
2B Same formulation as compound 1, but substitution of 100% organoclay by the same amount of MWCNTs

24.5 Heat release rates for compounds made by twin-roll mill with different filler blends MWCNTs–organoclays. Heat flux = 35 kW/m2. From reference 20; John Wiley & Sons Limited. Reproduced with permission.

24.6 Heat release rates for cable compound 2A with the filler blend MWCNTs–organoclays by twin-roll mill and BUSS Ko-kneader. Heat flux = 35 kW/m2. From reference 20; John Wiley & Sons Limited. Reproduced with permission.

Two insulated wires with identical geometric parameters were produced on a 20-L/D single-screw cable extruder with an 80 mm screw diameter. The diameter of the copper conductor was 1.78 mm, and the wall thickness for the insulation was 0.8 mm. For one wire, the insulation was compound 1 (filler combination of ATH–organoclay) and for the other wire, the optimized compound 2A (filler combination of ATH–organoclay–MWCNTs) was used as insulation. The MWCNT-based compound 2A showed a remarkable increased viscosity over that of the standard nanocomposite compound 1, as indicated by reduced revolutions per minute of the screw and increased power take-up by the extruder motor; a high pressure capillary viscosimeter showed a higher viscosity for the compound 2A for all shear rates up to 3000 s−1 obviously due to the high L/D ratio of the MWCNTs and their dispersion at a nanometer level. A small fire test according to IEC 60 332-1 (the insulation was exposed to a Bunsen burner ignition source) was very similar for both insulated wires. No dripping of burning polymer was noted, and the charred lengths were identical. But the char of the insulation made with the compound 2A was far less cracked compared to the char generated from the compound 1. This may be the result of the strengthening effect by the MWCNT, with its very great L/D ratio; proposal of such a mechanism for the function of carbon nanotubes was published recently.21 Heat release rates and times to ignition of the two insulated wires were measured using a cone calorimeter. The wires were cut in samples of 10 cm, and the standard cone sample holder was filled with the cable pieces. Twenty-six wires were mounted by building up a single layer of wires with no gap between them and the ends of the wires were not sealed; this mounting was called a ‘single layer design’. Also, four cut wires with no sealing of the ends, simulating an unjacketed cable22 were put together. An aramid fibre binder was used to maintain the integrity of the bundles; this mounting with 24 wires/layer and totally two layers was called a ‘bundle design’ (Fig. 24.7). Both mounting designs demonstrated quite different cone calorimeter results (Figs 24.8 and 24.9). For the single layer design, the cone tests were finished within 20 minutes. This is the flame application time for cables as defined in the new European proposal for flame tests of cables (prEN 50399). Wires insulated with the compound 1 had a higher PHRR within the first 5 minutes. For the bundle design, the duration of the cone calorimeter tests was very long, due to the insulation effect of the first charred layer on those located underneath in the second layer; therefore, the tests were stopped after 20 minutes. This mounting design is representative of many end-product applications. Wires insulated with the filler-blended compound 2A did not show any increase in the PHRR within the first 10 minutes compared to the wires insulated with the only organoclaybased compound 1.

24.7 Different insulated wire mounting designs for cone calorimeter tests From reference 20; John Wiley & Sons Limited. Reproduced with permission.

24.8 Heat release rates for the single layer design (FR wire, 2.5 mm2). Heat flux = 35 kW/m2. From reference 20; John Wiley & Sons Limited. Reproduced with permission.

24.9 Heat release rates for the bundle design (FR wire, 2.5 mm2). Heat flux = 35 kW/m2. From reference 20; John Wiley & Sons Limited. Reproduced with permission.

24.4 Conclusion

MWCNTs act as very efficient flame retardants at low filler contents in EVA. An optimized formulation for flame-retardant insulated wires was developed based on the filler-blend ‘multi-walled carbon nanotubes-organoclays’. Small fire tests according to IEC 60 332-1, with flames of a Bunsen burner attacking the insulation of the wire, showed that the char was strengthened by the long L/D ratio of the MWCNTs. The improved char results in better flame-retardant performances of the wires. Now there are many activities worldwide to investigate the combinations of nanostructurated fillers in combination with traditional non-halogenated or halogenated traditional flame retardants for technical applications to enhance the worldwide markets for flame-retardant polymers.

24.5 References

1. Hirschler, M.M., Fire performance of organic polymers, thermal, decomposition, and chemical decomposition. Polymeric Materials: Science and Engineering; 83. ACS Meeting, Washington, DC, 2000:79–80.

2. Babrauskas, V. The generation of CO in bench scale fire tests and the prediction for real scale fires. Fire and Materials. 1995; 19:205–213.

3. April Hirschler, M.M., Fire safety, smoke toxicity and halogenated materialsFlame Retardancy News. Norwalk, CT, USA: Business Communications Co, 2005.

4. Stevens, G., Countervailing risks and benefits in the use of flame retardants. Proceedings of the Flame Retardants 2000 Conference, London, UK, 8–9 February 2000. 2000:131–145.

5. Beyer, G. Nanocomposites offer new way forward for flame retardants. Plastics Additives & Compounding. 2005; 7:32–35. [September/October issue].

6. Gilman, J.W., Kashiwagi, T., Giannelis, E.P., Manias, E., Lomakin, S., Lichtenhan, J.D., Jones, P. Nanocomposites: radiative gasification and vinyl polymer flammability. In: Le Bras M., Camino G., Bourbigot S., Delobel R., eds. Fire Retardancy of Polymers: The Use of Intumescence. Cambridge: Royal Society of Chemistry; 1998:203–221.

7. Gilman, J.W., Kashiwagi, T., Lichtenhan, J.D. A revolutionary new flame retardant approach. SAMPE J. 33–40, 1997.

8. Beyer, G., Industry Guide to Polymer Nanocomposites. Plastics Information Direct. 2009. [ISBN: 978–1–906479–04–6].

9. 3–6 June Beyer, G., Improvements of the fire performance of nanocompositesProceedings of the 13th Annual BCC Conference on Flame Retardancy. CT, USA: Stamford, 2002.

10. Kashiwagi, T., Grulke, E., Hilding, J., Harris, R., Awad, W., Douglos, J. Thermal degradation and flammability properties of poly(propylene)/carbon nanotube composites. Macromolecular Rapid Communications. 2002; 23:761–765.

11. Kashiwagi, T., Grulke, E., Hilding, J., Groth, K., Harris, R., Butler, K., Shields, J., Kharchenko, S., Douglas, J. Thermal and flammability properties of polypropylene/carbon nanotube nanocomposites. Polymer. 2004; 45:4227–4239.

12. Kashiwagi, T., Du, F., Winey, K., Groth, K., Shields, J., Bellayer, S., Kim, H., Douglas, J. Flammability properties of polymer nanocomposites with single-walled carbon nanotubes: effects of nanotube dispersion and concentration. Polymer. 2005; 46:471–481.

13. Kashiwagi, T., Du, F., Douglas, J., Winey, K.I., Harris, R., Shields, J. Nanoparticle networks reduce the flammability of polymer nanocomposites. Nature Materials. 2005; 4:928–933.

14. Schartel, B., Pötschke, P., Knoll, U., Abdel-Goad, M. Fire behaviour of polyamide 6/multiwall carbon nanotube nanocomposites. European Polymer Journal. 2005; 41:1061–1070.

15. Breuer, O., Sundararaj, U. Big returns from small fibers: a review of polymer/carbon nanotube composites. Polymer Composites. 2004; 25:630–645.

16. Tang, B.Z., Xu, H. Preparation, alignment, and optical properties of soluble poly(phenylacetylene)-wrapped carbon nanotubes. Macromolecules. 1999; 32:2569–2576.

17. September Shaffer, M., Carbon nanotube modified polymers. paper presented to the conference Nanostructures in Polymer Matrices. Risley Hall, Derbyshire, UK, 2001:10–13.

18. Beyer, G. Carbon nanotubes as flame retardants for polymers. Fire and Materials. 2002; 26:291–293.

19. Gilman, J.W., Jackson, C.L., Morgan, A.B., Harris, R., Manias, E., Giannelis, E.P., Wuthenow, M., Hilton, D., Philipps, H. Flammability properties of polymer- layered-silicate nanocomposites, polypropylene and polystyrene nanocomposites. Chemistry of Materials. 2000; 12:1866–1873.

20. Beyer, G. Filler blend of carbon nanotubes and organoclays with improved char as a new flame retardant system for polymers and cable application. Fire and Materials. 2005; 29:61–69.

21. Beyer, G., Gao, F., Yuan, Q.A. A mechanistic study of fire retardancy of carbon nanotube/ethylene vinyl acetate copolymers and their clay composites. Polymer Degradation and Stability. 2005; 89:559–564.

22. Elliot, P.J., Whiteley, R.H. A cone calorimeter test for the measurement of flammability properties of insulated wire. Polymer Degradation and Stability. 1999; 64:577–584.