Polymer Degradation and Stability 91 (2006) 2275e2281www.elsevier.com/locate/polydegstab

Metal functionalized POSS as fire retardants in polypropylene

Alberto Fina a,*, Hendrikus C.L. Abbenhuis b, Daniela Tabuani a, Giovanni Camino a

a Centro di Cultura per l’Ingegneria delle Materie Plastiche e Politecnico di Torino, V.le T. Michel,

5 e 15100 Alessandria, Italy e and INSTM memberb Eindhoven University of Technology, Hybrid Catalysis BV, P.O. Box 513, 5600 MB Eindhoven, The Netherlands

Received 20 March 2006; received in revised form 18 April 2006; accepted 24 April 2006

Available online 6 June 2006

Abstract

This paper deals with the study of the combustion properties of dimeric and oligomeric Al-and Zn-isobutyl silsesquioxane (POSS)/polypro-pylene (PP) composites, in comparison with neat PP and PP/octaisobutyl POSS.

The presence of Al-POSS leads to a decrease in combustion rate with respect to PP, resulting in a decrease of Heat Release Rate (�43% at10 wt% POSS loading) as well as reduction in CO and CO2 production rates, whereas a negative effect on the above parameters is obtained withoctaisobutyl POSS. Zn-POSS does not significantly affect the PP combustion behaviour.

The effect of Al-POSS is most likely related to its chemical activity, which favours the formation of a moderate amount of char residue, bycatalysing secondary reactions in the polymer during combustion.� 2006 Elsevier Ltd. All rights reserved.

Keywords: POSS; Polysilsesquioxane; Flame retardancy; Metal-POSS; Cone calorimeter; Combustion

1. Introduction

In recent years, much effort was dedicated to the develop-ment of halogen free flame retardants for polymers, in order toavoid the use of traditional chlorinated or brominated com-pounds; in this field, attention was given to silicon containingflame retardants such as silanes, siloxanes and silsesquioxanes[1,2]. These compounds are recognised to be the precursors forthe formation of thermally stable ceramic materials so thatthey are also referred as preceramic compounds.

Different polysilsesquioxane or polycarbosilane resins wereshown to be effective fire retardants in thermoplastic polymerssuch as polypropylene (PP), styrene-butadiene-styrene blockcopolymers (SBS) and polyether-polyamide copolymers(PTME-PA), by reducing the Heat Release Rate during conecalorimeter tests [3,4]. In PP a 40% reduction in Heat ReleaseRate (HRR) peak was obtained with 20 wt% of a methyl/

* Corresponding author. Tel.: þ390131229316; fax: þ390131229331.

E-mail address: alberto.fina@polial.polito.it (A. Fina).

0141-3910/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.polymdegradstab.2006.04.014

phenyl polysilsesquioxane and even better results were ob-tained in PTME-PA, with HRR peak reduction of about 70%by incorporation of 10 wt% of the same polysilsesquioxane.The proposed mechanism of fire retardancy with preceramicsinvolves the formation of an insulating char, which shields theunderlying polymer from the radiated heat and limits polymerablation, thus reducing the amount of combustible volatileproducts available for burning in the gas phase [4].

Few scientific papers have been published so far on the useof Polyhedral Oligomeric Silsesquioxane (POSS) as flameretardants.

Ikeda reported the use of POSS, with different cage struc-tures and organic substituents, in polyphenylene ether (PPE),showing the formation of a tough foam layer on nanocompo-site surface during UL94 tests, leading to V0 classification,whereas neat PPE was classified as V1 [5].

Bourbigot et al. prepared polyurethane/POSS (TPUePOSS)nanocomposite coatings for textiles and tested knitted multifila-ment yarns on cone calorimeter. TPUePOSS coating on polyes-ter woven fabrics resulted in a remarkable reduction of the HeatRelease Rate peak (up to 50%), in particular with the use of

2276 A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

phenyl-T12 POSS and with poly(vinyl silsesquioxane) resin,whereas methyl-T8-POSS was found to be almost ineffective[6,7]. The same research group also prepared multifilamentyarn knitted fabric with PP/poly(vinyl silsesquioxane) resinnanocomposites, showing an increase of the Time to Ignitionwith no significant effects on the Heat Release Rate [7].

Recently, combustion behaviour of a PMMAetrisilanolphenyl POSS nanocomposites was studied by means of conecalorimeter tests, showing no improvements with respect toneat PMMA [8].

As regards thermoset polymers, interesting results were ob-tained on vinyl ester resins by the addition of vinyl POSS[9,10] in terms of reduced smoke release, lower HRR andhigher ignition time, with respect to neat resin.

Synergic formulations with phosphorous-based fire retar-dants were recently studied in vinyl ester thermoset nanocom-posites [11]. Significant reductions of the HRR peak wereobtained with vinyl POSS and an improvement was achievedby the addition of tricresylphosphate, also leading to importantreductions of the Total Heat Released.

The effect of metal nanoparticles on polymer flammabilitywas explored by Antonov et al. [12], showing that finely dis-persed metals at low concentration (�1 wt%) strongly enhancePP char yield, by catalysing dehydrogenation, despite a higherflammability, shown by lower LOI values.

Since the possibility to remove part of the organic polymerfrom combustion through carbonisation is one of the maingoals in polymer fire retardancy [13], the dehydrogenationeffect appears to be particularly interesting.

In this work the effect of metal-containing POSS on PPcombustion behaviour was investigated; such compoundswere chosen to couple the silsesquioxane preceramic effectwith the catalytic action of a metal site on POSS cage, takingadvantage from the possibility to disperse POSS cages ona submicron scale into polymer matrix [14]. These systemshave been previously investigated by the same authors, show-ing that thermoxidative degradation of composites is stronglyaffected by the presence of metal-containing POSS, resultingin an improved thermal stability, in terms of higher weightloss temperature [15].

2. Experimental

Octaisobutyl-T8-POSS, (i-C4H9)8Si8O12, (Fig. 1), referredto as T8-POSS in the following, was purchased from HybridPlastics Company and used as received.

Metal-POSS derivatives (Al-POSS and Zn-POSS, Figs. 2 and3, respectively) were prepared by deprotonation of incompletelycondensed POSS trisilanol, (i-C4H9)7Si7O9(OH)3 with eithertriethylaluminium or diethylzinc as previously reported [15].

PP was a Moplen HP501L, purchased from Basell.Composites were prepared by mixing PP and POSS in

a Brabender internal mixer (180 �C, 20 min, 60 rpm); theseblending conditions were chosen on the basis of what reportedby Fu et al. [16] and on our previous work [14]. POSS wasloaded into the polymeric matrix at 10 wt% ratio. Neat PPwas processed in the same conditions as a reference material.

Combustion tests were performed on a Fire Testing Tech-nology Cone Calorimeter, with 100� 100� 3 mm specimens,prepared by compression moulding on a hot plate laboratorypress at 190 �C and conditioned at 23 �C and 50% RH toequilibrium.

Tests were performed at 35 kW/m2 external heat flux, in or-der to evaluate the fire properties of the composites in condi-tions comparable to a mild fire scenario [17]; specimens werewrapped in an aluminium foil leaving the upper surface ex-posed to the radiator and placed on ceramic backing boardat a distance of 25 mm from cone base.

Three tests for each sample were performed and in the fol-lowing section the average values will be discussed; data arereported with their deviation with respect to the average.

Raman spectra using 633 nm excitation wavelength wererecorded on a Jobin Yvon Labram HR spectrophotometer,equipped with an Olympus microscope and using a CCD de-tector, using a 1800 lines/mm grating, determining a instru-mental resolution of 2.5 cm�1, with typical laser power setat 10 mW.

Scanning electron microscopy imaging was obtained bymeans of a LEO 1450 VP instrument on cryogenic fracturesurfaces.

3. Results and discussion

The main parameters obtained from cone calorimeter mea-surements are reported in Table 1 for each material and will be

SiO

Si

O

SiO

O

SiO

SiO

Si

O

SiO

O

Si

OO

O

R

R

R

R

RR

RR

R=i-C4H9

Fig. 1. Octaisobutyl POSS (T8-POSS).

O

Si

SiO

Si OSi

O

SiO

O

O

Si

O

OSi

O

O

O

R R

R

R

R

R

R

SiOSi

O

Si O

O

O

Si

O

OO

Si SiO

O

Si

O

O

RR

R

R

R

R

R

Al

Al

R=i-C4H9

Fig. 2. Al-POSS.

2277A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

discussed hereunder; they include Time to Ignition (TTI), HeatRelease Rate (HRR), Total Heat Released (THR), EffectiveHeat of Combustion (EHC), Maximum Average Rate ofHeat Emission (MAHRE), Total Smoke Released (TSR), COand CO2 yields, mass loss during combustion.

3.1. Ignition time

Time to Ignition for PP/POSS appears equal or lower with re-spect to neat PP, whereas a remarkable reduction is observed forPP/Al-POSS. From the visual observations during the tests, thislower ignition time is probably due to specimen warping underthe cone heater irradiance in the early stage of the test, resultingin a reduced distance of the sample from the cone heater. Whilethis kind of deformation is relaxed far before ignition for neat PP,deformation relaxation for composites is slower (in particularfor PP/Al-POSS) so that ignition occurs when the relaxation isstill not complete. In these conditions, the deformed part ofthe specimen is irradiated with a heat flux higher than the nom-inal [18], resulting in an anticipated ignition.

A retainer frame was not used in order to avoid extensivedripping during combustion. Trying to obtain reliable ignitiontimes, measurements were performed on non-standard50� 50� 3 mm specimens (three for each material): in thiscase limited deformation is observed and ignition times are56� 3 s for PP, 63� 7 s for PP/Al-POSS and 59� 4 s forPP/Zn-POSS. In conclusion, this parameter is not strongly af-fected by the presence of POSS.

O

Si

SiO

SiO

SiO

SiO

O

O

Si

O

OSi

O

O

O

R R

R

R

R

R

RZn

SiO

SiO

Si O

O

O

Si

O

OO

Si SiO

O

Si

O

O

RR

R

R

R

R

R

Zn

Zn

n

R = i-C4H9

Fig. 3. Zn-POSS.

3.2. Heat release

Heat Release Rate results are shown in Fig. 4a. As com-pared to neat PP, PP/T8-POSS shows a higher Heat ReleaseRate; this behaviour can be explained by partial evaporationof T8-POSS in the earlier stages of combustion, and subse-quent oxidation of POSS organic substituents in the flame,representing a further contribution in heat flux radiated fromthe flame.

This negative result gives even more prominence to whatobtained for PP/Al-POSS, which shows a significant reductionof Heat Release Rate with respect to PP (see Table 1): slowercombustion is evidenced by the lower HRR peak (pkHRR,�43%).

In contrast, PP/Zn-POSS does not show significant differ-ences on HRR curve as compared to PP.

The Average Rate of Heat Emission (AHRE) curve is re-ported in Fig. 4b. This parameter is defined as the cumulativeheat emission divided by time [19] and its peak value (Maxi-mum Average Rate of Heat Emission, MAHRE) has been re-cently proposed as a good measure of the propensity for firedevelopment under real scale conditions [20]. MAHRE forPP/Al-POSS (Table 1) shows a notable reduction (w20%)with respect to PP, whereas an increase is observed with T8-POSS and no effect is found for Zn-POSS.

By comparing the above radically different composite be-haviours, a strong effect due to the Al atom is suggested; infact, T8- and Al-POSS differs mainly for the presence of Alin the cage, being the groups linked to the Si atoms in anycase isobutyl radicals. This improved behaviour of Al-POSSas compared to T8-POSS can be ascribed to a catalytic effectinduced by the metal during the combustion process. Indeed,from direct observation during the tests, the mechanism lead-ing to combustion rate reduction seems to be chemical ratherthan physical as, in both cases, only a non-continuous thin ce-ramic layer from POSS degradation [15,21] is formed on thespecimen surface during combustion, which does not representan effective physical barrier for combustion processes.

The influence of Al- and Zn-POSS on PP thermo-oxidativedegradation was previously investigated [15] resulting in animproved thermal stability, explained by a chemical actionof the metals on PP degradation pathway.

Dealing with Total Heat Released (THR), values for PP/T8-POSS and PP/Zn-POSS are similar to that of PP, whereasTHR of PP/Al-POSS is 11% lower (Fig. 4c): given theamount of inorganic (non-combustible) fraction present in

Table 1

Main parameters from cone calorimeter measurements

TTI

(s)

PkHRR

(kW/m2)

MAHRE

(kW/m2)

THR

(MJ/m2)

EHC

(MJ/kg)

Residual

weight (%)

TSR Avg CO2

yield

(kg/kg)

Avg CO

yield

(g/kg)

PP 56� 5 1103� 50 509� 20 111� 2 47� 1 w0 1427� 30 3.39� 0.02 43� 5

PP/T8-POSS 50� 3 1325� 50 591� 20 112� 3 48� 1 0.2� 0.1 1564� 70 3.45� 0.04 42� 1

PP/Al-POSS 37� 1 624� 20 399� 10 98� 1 41� 1 2.5� 0.1 1850� 20 3.05� 0.03 35� 2

PP/Zn-POSS 54� 4 1069� 50 509� 20 108� 4 46� 1 3.3� 0.1 1525� 25 3.35� 0.08 42� 2

2278 A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

the PP/Al-POSS composite, a theoretical decrease of THRequal to 5% is expected. In fact, when adding 10 wt% ofAl-POSS, about the half of it (w5 wt%) is represented bynon-combustible SieOeAl fraction. The lower THR valuewith respect to prediction indicates that a part of the polymeris not completely combusted, probably undergoing a carbonisa-tion process. The 11% reduction falls at the limit of the gen-eral recognised reliability for cone calorimeter data [22e24]but data deviations seem to validate such difference.

Effective Heat of Combustion (EHC, Table 1) for PP/Al-POSS is lower than for reference PP, showing chemical actionof Al-POSS during PP combustion, whereas no significantchanges are observed with T8-POSS and Zn-POSS.

The black appearance of PP/Al-POSS residue (Fig. 8) at theend of the tests confirms the presence of a carbon char. Thisvisual observation confirms the chemical action of Al in thecombustion process. Since Lewis acid metals are known to

0

200

400

600

800

1000

1200

1400

0 50 100 150 200 250 300 350

0

20

40

60

80

100

120

0

100

200

300

400

500

600

c

b

PPPP 10% T8-POSSPP 10% Al-POSSPP 10% Zn-POSS

HR

R [k

W/m

2 ]

a

THR

[MJ/

m2 ]

Time [s]

AHR

E [k

W/m

2 ]

Fig. 4. Heat evolving during combustion (a) Heat Release Rate, (b) Average

Rate of Heat Emission, (c) Total Heat Released.

catalyse oxidative dehydrogenation reactions in the gas phase[25], Al probably acts as a well-dispersed catalyst in the mol-ten polymer, affecting the PP degradation and producing a lim-ited amount of carbonaceous phase.

0,00

0,05

0,10

0,15

0 50 100 150 200 250 300 350

0

500

1000

1500

2000b

SPR

[m2 /

s]

a

Time [s]TS

R

PPPP 10% T8-POSSPP 10% Al-POSSPP 10% Zn-POSS

Fig. 5. Smoke production during combustion (a) Smoke Production Rate,

(b) Total Smoke Released.

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

0 50 100 150 200 250 300 350

0

100

200

300

400

500

600

700 b

CO

2 [%

]

a

Time [s]

CO

[ppm

]

PPPP 10% T8-POSSPP 10% Al-POSSPP 10% Zn-POSS

Fig. 6. CO2 (a) and CO (b) production during combustion.

2279A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

3.3. Smoke

In Fig. 5 smoke production parameters are reported. Asregards smoke production rate (SPR), the results reflect thedifferences in HRR discussed previously, with higher SPRvalues for PP/T8-POSS and with a broader curve for PP/Al-POSS. Total Smoke Released (TSR) of PP/POSS showshigher values and the difference with respect to PP is partic-ularly relevant for PP/Al-POSS. The higher smoke produc-tion is to be ascribed to the dehydrogenation effect whichleads to aromatic volatiles in the flame, resulting in a sootincrease.

3.4. CO and CO2 release

In Fig. 6 CO and CO2 production is shown, mimicking theHRR curve shapes exactly. The ratio between CO and CO2

yield is not significantly changed; it is relevant that CO yieldfor composites (see Table 1) is not higher than for PP, so thatsmoke toxicity is not increased by POSS. Moreover, a signifi-cant decrease in CO yield is caused by the presence of Al-POSS.

3.5. Mass

In Fig. 7 mass curves of the composites during combustionare reported; the higher combustion rate for PP/T8-POSS re-sulted in an anticipated overall weight loss as compared toPP. A lower mass loss rate (MLR) is clearly observable forPP/Al-POSS. The residue amounts at the end of the analyses

0

20

40

60

80

100

0 5 100 150 200 250 300 350

0,00

0,05

0,10

0,15

0,20

0,25b

Mas

s [%

]

a

Time [s]

Mas

s lo

ss ra

te [g

/s]

PPPP 10% T8-POSSPP 10% Al-POSSPP 10% Zn-POSS

Fig. 7. Mass curves during combustion (a) mass (normalized on initial weight)

and (b) mass loss rate.

are reported in Table 1 and the residue appearance is shownin Fig. 8.

The weight loss curves for PP/metal-POSS obtained fromcone calorimeter test show differences with those obtainedfrom thermogravimetry in air: on heating in a thermobalanceat 10 �C/min in air, both metal compounds present a two-step

Fig. 8. Residues at the end of cone calorimeter tests (a) PP/T8-POSS, (b) PP/

Al-POSS, (c) PP/Zn-POSS. Neat PP gave no residue.

2280 A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

weight loss, corresponding to two different degradation mecha-nisms [15], whereas a single step is observed during the combus-tion test.

This is due to the radically different heating conditions; infact, during cone calorimeter test a high heating rate and hightemperatures are reached. In these conditions it is assumed thatonly the degradation process occurring at higher temperaturetakes place; in fact, weight losses are in agreement with theones observed during the thermogravimetric second degrada-tion stage, being the mass loss rate of PP/Al-POSS lowerthan that of PP/Zn-POSS.

The residue left by PP/T8-POSS is mainly formed of thinlight grey platelets probably due to silica network formationfrom POSS during combustion, as reported in previous works[21,26].

On the other hand, PP/Al-POSS residue appears almostcompletely black; Raman spectroscopy (Fig. 9) shows thepresence of two bands at ca. 1359 and 1607 cm�1 which aretypical of graphenic phases. The relative intensity and widthof the bands can be consistent with partially ordered carbon[27], thus confirming the above described PP charring processcatalysed by Al-POSS.

The appearance of PP/Zn-POSS residue is intermediate be-tween those of PP/T8-POSS and PP/Al-POSS; in fact, the res-idue at the end of combustion test is formed of a mixture ofgrey and black particles. Raman analyses confirmed the pres-ence of a graphenic phase in the darker particles.

The metal-POSS dispersion into the PP matrix is shown inFig. 10. Based on our previous experience on PP/isobutylPOSS systems [14], it is supposed that a fraction of the nano-filler is dispersed on a submicron scale, the remaining part be-ing still observable as residual micron-sized POSS aggregatesby SEM imaging. A complete dispersion to a submicron scalewould probably further enhance the combustion behaviour ofPP/metal-POSS, thanks to a uniform distribution of catalyticcentres, leading to a higher efficiency.

1000 1200 1400 1600 1800 2000

0

100

200

300

400

500

1607cm-1

Ram

an s

catte

ring

Wavenumber [cm-1]

PP/ Al-POSS

PP/ T8-POSS

1359cm-1

Fig. 9. Raman spectroscopy on PP/Al-POSS vs PP/T8-POSS.

4. Conclusions

From the analysis of the combustion parameters discussedin this paper, which give a clear overview on the materialburning behaviour, we can conclude that the presence of metalatoms in POSS structure can affect the polymer degradationpathway, resulting in improved fire retardancy properties.

Whereas the addition of octaisobutyl POSS shows a higherflame spread with respect to neat PP, Al isobutyl POSS ap-pears to be an effective fire retardant for PP, resulting in lowerHeat Release Rate and lower Effective Heat of Combustion.The PP/Al-POSS behaviour is probably due to a catalytic ef-fect of the Al moieties, promoting secondary reactions duringpolymer degradation and leading to partial PP charring, in-stead of complete volatilisation. The carbonaceous char ob-tained at the end of combustion test shows signals for theformation of a graphenic phase, evidencing an aromatisationeffect induced by the Al moieties.

On the other hand, PP/Zn-POSS does not show any fire re-tardancy improvements with respect to neat PP, even if a partialcharring during combustion is observed.

The possibility to subtract part of the organic polymer fromcombustion through carbonisation is one of the main goals inpolymer fire retardancy; in this sense metal-POSS compoundsare promising fire retardants and synergic formulation withother flame retardant additives should be studied.

Fig. 10. SEM micrographs for (a) PP/Al-POSS and (b) PP/Zn-POSS.

2281A. Fina et al. / Polymer Degradation and Stability 91 (2006) 2275e2281

Acknowledgements

This study was carried out in the frame of the STRP Euro-pean research program ‘‘NANOFIRE’’, No. 505637, in the 6thFramework Program.

The authors would like to gratefully thank Dr. Enrico Boc-caleri (University of Eastern Piedmont) for Raman character-ization and Andrea Castrovinci for useful discussions.

References

[1] Kashiwagi T, Gilman JW. In: Grand AF, Wilkie CA, editors. Fire retard-

ancy of polymeric materials. New York: Marcel Dekker Inc.; 2000.

p. 353e89.

[2] Lu SY, Hamerton I. Recent developments in the chemistry of halogen-

free flame retardant polymers. Prog Polym Sci 2002;27:1661e712.

[3] Gilman JW, Kashiwagi T, Harris RH, Lomakin S, Lichtenhan JD, Bolf A,

et al. In: Al-Malaika S, Golovoy A, Wilkie CA, editors. Chemistry and

technology of polymer additives. Malden: Blackwell Science Inc.;

1999. p. 135e50.

[4] Lichtenhan JD, Gilman JW. Preceramic additives as fire retardants for

plastics. US Patent 6,362,279; 2002.

[5] Ikeda M. Utilization of POSS in industrial applications. In: Proceedings

of POSS nanotechnology conference, Huntington Beach, CA; September

2002. p. 278e88.

[6] Devaux E, Rochery M, Bourbigot S. Polyurethane/clay and polyure-

thane/POSS nanocomposites as flame retarded coating for polyesters

and cotton fabrics. Fire Mater 2002;26:149e54.

[7] Bourbigot S, Le Bras M, Flambard X, Rochery M, Devaux E,

Lichtenhan JD. In: Le Bras M, Wilkie C, Bourbigot S, editors. Fire re-

tardancy of polymers: new applications of mineral fillers. Cambridge:

Royal Society of Chemistry; 2005. p. 189e201.

[8] Jash P, Wilkie CA. Effects of surfactants on the thermal and fire proper-

ties of poly(methyl methacrylate)/clay nanocomposites. Polym Degrad

Stab 2005;88:401e6.

[9] Gupta SK, Schwab JJ, Lee A, Fu BX, Hsiao BS. In: Rasmussen BM,

Pilato LA, Kliger HS, editors. Affordable materials technology e platform

to global value and performance. Long Beach, CA: SAMPE Pub.; 2002.

p. 1517e26.

[10] Lee A. POSS thermosettings polymers and composites. In: Proceedings

of POSS nanotechnology conference, Huntington Beach, CA; September

2002. p. 203e25.

[11] Chigwada G, Jash P, Jiang DD, Wilkie CA. Fire retardancy of vinyl ester

nanocomposites: synergy with phosphorus-based fire retardants. Polym

Degrad Stab 2005;89:85e100.

[12] Antonov A, Yablokova M, Costa L, Balavanovich A, Levchik G,

Levchick S. The effect of nanometals on the flammability and thermox-

idative degradation of polymer materials. Mol Cryst Liq Cryst Sci Tech-

nol Sect A 2000;353:203e10.

[13] Gilman JW, Kashiwagi T, Harris RH, Lomakin S, Lichtenhan JD, Bolf A,

et al. In: Al-Malaika S, Golovoy A, Wilkie CA, editors. Chemistry and

technology of polymer additives. Malden: Blackwell Science; 1999.

p. 135e50.

[14] Fina A, Tabuani D, Frache A, Camino G. Polypropylene-polyhedral olig-

omeric silsesquioxanes (POSS) nanocomposites. Polymer 2005;46:

7855e66.

[15] Fina A, Abbenhuis HCL, Tabuani D, Frache A, Camino G. Polypropyl-

ene metal functionalized POSS nanocomposites: a study by thermogravi-

metric analysis. Polym Degrad Stab 2006;91:1064e70.

[16] Fu BX, Yang L, Somani RH, Zong SX, Hsiao BS, Philips S, et al.

Crystallization studies of isotactic polypropylene containing nano-

structured polyhedral oligomeric silsesquioxane molecules under

quiescent and shear conditions. J Polym Sci B Polym Phys 2001;39:

2727e39.

[17] Babrauskas V. Specimen heat fluxes for bench-scale heat release rate

testing. Fire Mater 1995;19:243e52.

[18] Schartel B, Bartholmai M, Knoll U. Some comments on the use of cone

calorimeter data. Polym Degrad Stab 2005;88:540e7.

[19] European Standard under development prEN 45545e2:2004, under

approval.

[20] Duggan GJ, Grayson SJ, Kumar S. New fire classifications and fire test

methods for the European railway industry, Flame retardants 2004. Lon-

don: Interscience Communications; 2004.

[21] Fina A, Tabuani D, Frache A, Boccaleri E, Camino G. In: Le Bras M,

Wilkie C, Bourbigot S, editors. Fire retardancy of polymers: new appli-

cations of mineral fillers. Cambridge: Royal Society of Chemistry; 2005.

p. 202e20.

[22] European Standard ISO 5660; 2002.

[23] Babrauskas V. Development of the cone calorimetere a bench-scale heat

release apparatus based on oxygen consumption. Fire Mater 1984;8:

81e94.

[24] Gilman JW, Kashiwagi T, Nyden M, Brown JET, Jackson CL,

Lomakin S, et al. In: Al-Malaika S, Golovoy A, Wilkie CA, editors.

Chemistry and technology of polymer additives. London: Blackwell

Scientific; 1998. p. 249e65.

[25] Banares MA. Supported metal oxide and other catalysts for ethane con-

version: a review. Catal Today 1999;51:319e48.

[26] Fina A, Tabuani D, Carniato F, Frache A, Boccaleri E, Camino G.

Polyhedral oligomeric silsesquioxanes (POSS) thermal degradation.

Thermochim Acta 2006;440:36e42.

[27] Eklund PC, Holden JM, Jishi RA. Vibrational modes of carbon nano-

tubes: spectroscopy and theory. Carbon 1995;33:959e72.