European Polymer Journal 51 (2014) 112–119

Contents lists available at ScienceDirect

European Polymer Journal

journal homepage: www.elsevier .com/locate /europol j

Bulk or surface treatments of ethylene vinyl acetate copolymerswith DNA: Investigation on the flame retardant properties

0014-3057/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.eurpolymj.2013.12.009

⇑ Corresponding author. Tel.: +39 0131 229337; fax: +39 0131 229399.E-mail address: jenny.alongi@polito.it (J. Alongi).

Jenny Alongi ⇑, Alessandro Di Blasio, Fabio Cuttica, Federico Carosio, Giulio MalucelliDipartimento di Scienza Applicata e Tecnologia, Politecnico di Torino, sede di Alessandria, and Local INSTM Unit, Viale Teresa Michel 5, 15121 Alessandria, Italy

a r t i c l e i n f o a b s t r a c t

Article history:Received 23 October 2013Received in revised form 6 December 2013Accepted 11 December 2013Available online 19 December 2013

Keywords:Ethylene vinyl acetate copolymersDeoxyribonucleic acidFlame retardancyCombustionCone calorimeterBurning-through tests

Deoxyribose nucleic acid (DNA) has recently proven to be an efficient flame retardant forethylene vinyl acetate (EVA) copolymers, when added in bulk via melt-blending. Indeed,thanks to its char-former features, DNA was able to quite efficiently protect an EVAcopolymer (containing 18 wt.% of vinyl acetate) against an irradiative heat flux of35 kW/m2, strongly reducing the combustion kinetics and favouring a remarkable decreaseof CO and CO2 yields. In the present work, the evolution of the DNA flame retardant conceptis presented: in spite of bulk compounding, DNA has been confined as a coating on EVAsurface. Thus, a comparative study on the flame retardant properties of EVA loaded or coatedwith DNA has been thoroughly carried out. The collected results have shown that the DNAcoating blocks the ignition of the copolymer when tested by cone calorimeter under a heatflux of 35 kW/m2, increasing the time to ignition by 228s (+380%, with respect to pureEVA), while it greatly postpones (102s, +625% with respect to pure EVA) and reduces thecombustion kinetics under a heat flux of 50 kW/m2. Finally, unlike melt-compoundedDNA, the bio-macromolecule coating is able to protect the underlying material from abutane/propane torch applied three times consecutively to the specimen for 5s.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Generally speaking, a polymer combustion is fuelledby pyrolysis products escaping from its surface due tothe heat transferred from the flame to the polymersurface and also radiated from the flame itself, asschematised in Fig. 1. This process can be modelled atthe laboratory scale by cone calorimeter [1,2]. The oxygenrequired for sustaining the flaming combustion diffuses inand from the surrounding air environment. Solid particlesescape from the flame as smoke, which is accompanied bygaseous species, some of which can be toxic [3]. Asalready documented [1], the most significant polymerdegradation reactions usually occur in the condensedphase, as they take place mainly within 1 mm of theinterphase between the flame and polymer, where the

temperature raise is high enough. These reactions involvethe polymer and any additives (in particular flame retar-dants) included in the formulations or applied as surfacetreatments. Experimental studies of this region have beenpublished by Price and co-workers [4] and by Marosi andcoworkers [5,6]. The volatile species formed during com-bustion escape into the flame zone, whilst heavier speciesundergo further reactions and may eventually turn intochar: this multi-lamellar carbonaceous structure actingas a thermal insulator protects the surrounded polymer.It is common consensus that the flame retardants operatingin the condensed phase may be considered the unique andworthy alternative to halogen-based flame retardants (thatare currently under scrutiny by governments because ofenvironmental and human safety issues) [1], although theiraction mechanism is significantly different. Indeed, thesesystems are able to facilitate the char formation and toreduce the evolution of the flammable volatile species.In this scenario, DNA has proven to exhibit the same

Fig. 1. Schematic representation of a polymer combustion.

J. Alongi et al. / European Polymer Journal 51 (2014) 112–119 113

behaviour: indeed, due its chemical structure, it is anall-in-one intumescent bio-macromolecule able to pro-mote the char formation either when deposited on cottonfabrics through simple impregnation [7,8], Layer-by-Layerdeposition [9], or when added in bulk to ethylene vinylacetate (EVA) copolymers [10]. In particular, as far as theselatter are considered, compounds with different DNAamounts (namely, 10, 15 and 19 wt.%) have been recentlyprepared via melt-blending and the resulting thermal andfire properties have been thoroughly investigated. Thecollected results by cone calorimetry have shown thatDNA is able to promote a significant reduction of the heatrelease rate peak (�40%) as well as of CO and CO2 yields(approximately �50% and �40%, respectively). Pursuingthis research, further improvements can be foreseen whenDNA is deposited on EVA surface, instead of being added tothe bulk. Indeed, as depicted in Fig. 1, the surface plays akey role during the combustion of a polymer; as a conse-quence, a flame retardant loaded in the bulk has to diffusetoward the surface in order to perform its action. On theother hand, if confined on the surface, the same flameretardant can provide the full potential flame retardantfeature even from the beginning of combustion. Thus, inthe present manuscript, a comparative study on the flameretardant properties of EVA loaded or coated with DNA(being equal the bio-macromolecule content) has beenthoroughly performed. The resulting flame retardantproperties have been assessed by cone calorimeter andburning-through tests.

2. Experimental part

2.1. Materials

An EVA copolymer containing 18 wt.% vinyl acetate (El-vax�470 from DuPontTM; melt flow index: 0.7 g/10 min)was used. DNA from herring sperm was supplied as a highpurity grade reagent by Sigma Aldrich, Inc. and used asreceived.

2.2. Deposition of DNA on EVA

1 g of DNA was deposited on EVA square plates (6.7 g)and subsequently compressed, using a hot compressionmoulding press at 120 �C for 1.5 min (applied pressure:5 MPa). DNA:EVA weight ratio was set at 15 wt.%, that cor-responds to the bio-macromolecule content already addedto EVA in bulk [10]. Hereafter, we will refer to EVA addedwith DNA (EVA_DNA in bulk) and EVA coated by DNA(EVA_DNA on surface).

2.3. Characterization techniques

The surface morphology of EVA treated with DNA wasassessed by using an optical microscope (Nikon EclipseLV100D instrument) in transmission mode and aLEO-1450VP Scanning Electron Microscope (beam voltage:20 kV). An X-ray probe (INCA Energy Oxford, Cu Ka X-raysource, k = 1.540562 Å) was used to perform elementalanalysis. Fragments of the compounds obtained by a fragilefracture in liquid nitrogen (5 � 5 mm2) were fixed to con-ductive adhesive tapes and gold-metallized.

Cone calorimeter tests (Fire Testing Technology, FTT)were performed according to the ISO 5660 standard [2].The samples (50 � 50 � 3 mm3) were placed on a sampleholder and irradiated at a heat flux of 35 or 50 kW/m2 inhorizontal configuration. For each formulation, the testwas repeated three times and an experimental error of2% has been calculated as standard deviation for all themeasured parameters. The calculated value was signifi-cantly lower than that usually quoted for these tests(±10%); this finding can be ascribed to the controlled con-ditions adopted during these measurements: indeed, allthe samples were run on 1 day at 23 �C and 50% R.H.

Time To Ignition (TTI, s), Total Heat Release (THR,MJ/m2), and Heat Release Rate peak (PHRR, kW/m2). TotalSmoke Release (TSR, m2/m2), Rate of Smoke Release (RSR,1/s), peaks of carbon monoxide and dioxide ([CO] and[CO2], ppm and %, respectively) were evaluated, as well.

The temperature profile of square samples(50 � 50 � 3 mm3) as a function of time upon heatingwas monitored by using the cone calorimeter andfollowing the procedure described elsewhere [11,12]. Inparticular, the specimens were placed on a ceramic padholed in the centre and kept 25 mm far from the irradiatingsource. Two separated thermocouples (stainless steelsheathed K-type thermocouples with 0.5 and 1 mm outerdiameter for the thermocouple located at the top andbottom of the sample, respectively) connected with theheating source were used for monitoring the temperatureprofile as a function of time (heat flux: 35 kW/m2).

Burning-through tests were carried out placing thespecimen (50� 50� 3 mm3) in vertical configuration, andapplying a 50 mm butane–propane flame (65:35 vol./vol.)to its centre. The flame was applied for 5s at 100 mmdistance from the surface for three times. The temperatureon the back site of the specimen (the surface not exposedto the flame) was measured by using a thermocouple(stainless steel sheathed K-type thermocouples with0.5 mm diameter). The test was duplicated and theexperimental error on temperature peak was ±5 �C.

114 J. Alongi et al. / European Polymer Journal 51 (2014) 112–119

3. Results and discussion

As assessed by scanning electron microscopy (SEM), wehave already demonstrated that DNA can be homoge-neously and fine dispersed within EVA, without formingmicro-aggregates [10]. When the same amount of DNA isdeposited and hot-compressed on EVA surface, the forma-tion of a continuous and coherent coating was revealedthrough optical microscopy and SEM, as depicted inFig. 2a and b, respectively. Furthermore, it is noteworthyto observe the presence of a DNA-rich interphase zone be-tween coating and bulk, as clearly evident by C, O and Pdistribution in the elemental analyses reported in Fig. 2.

Fig. 2. Optical microscopy (A) and SEM (B) m

The thickness of the coating as well as of the interphasezone approaches a maximum of 150–160 lm.

3.1. Resistance to an irradiative heat flux

The resistance to an irradiative heat flux of EVA_DNA inbulk and EVA_DNA on surface has been compared withthat of neat EVA through cone calorimetry tests. Initially,these samples have been exposed to a heat flux of35 kW/m2. Fig. 3 plots HRR and THR curves as a functionof time (Fig. 3a and b, respectively). The presence of DNAin bulk causes a strong reduction of TTI EVA (28 vs. 62s[10]), but also a strong decrease of PHRR (�36%), as clearly

agnifications of EVA_DNA on surface.

Fig. 3. HRR (A) and THR (B) curves of EVA, EVA_DNA in bulk and EVA_DNA on surface as a function of time (heat flux: 35 kW/m2).

J. Alongi et al. / European Polymer Journal 51 (2014) 112–119 115

visible in Fig. 3a and reported in Table 1. When the sameDNA amount is placed on EVA surface, the specimen doesnot ignite for approximately 300s: meanwhile, EVA andEVA_DNA in bulk completely burn. On the contrary,EVA_DNA on surface ignites after 290s (TTI increase of380%) and starts to burn slowly with a very low PHRR(348 vs. 1588 kW/m2 for EVA_DNA on surface and EVA,respectively: this reduction roughly approaches 80%).Furthermore, comparing the THR values, it is possible toobserve an analogous trend (Fig. 3b): indeed DNA onsurface is more efficient than DNA in bulk to reduce thetotal heat release during combustion (namely, 67, 106and 108 MJ/m2 for EVA_DNA on surface, EVA_DNA in bulkand EVA, respectively). As far as smokes are concerned, TSRdoes not change in a remarkable way, regardless of thepresence of DNA (Fig. 4a); at variance, the rate of smokerelease (RSR) as well as CO and CO2 yields are stronglylowered when DNA is present on EVA surface (Fig. 4b–d).Once again, the reductions of CO and CO2 yields have beenestimated around 80%. At the end of tests, however, thefinal residue turned out to be around 5%, regardless ofthe presence of the bio-macromolecule in bulk or on thecopolymer surface. In both cases, DNA acts as char-former,exerting a thermal shield effect and protecting EVA fromcombustion. In order to confirm this hypothesis, the

Table 1Combustion data for neat EVA, EVA_DNA in bulk and EVA_DNA on surface by con

Sample TTI(s)

PHRR1(kW/m2)

PHRR2(kW/m2)

DPHRR1(%)

DPHRR2(%)

THR(MJ

Heat flux = 35 kW/m2

EVA 62 1588 0 0 0 108EVA_DNA in bulk 28 1013 846 �36 �47 106EVA_DNA on surface 290 348 – �77 – 67

Heat flux = 50 kW/m2

EVA 14 1980 – – – 106EVA_DNA in bulk 14 1311 – �34 – 106EVA_DNA on surface 116 375 – �81 – 96

⁄The experimental error was ±2%.

temperature profile during combustion has been moni-tored by using two thermocouples placed at the bottomand top of the specimen exposed to the cone calorimetersource, although the temperature values can be partiallyaffected by the intimate contact between polymer andthermocouple [13]. Fig. 5 shows the temperature profilesas a function of irradiation time. The neat copolymerignites at ca. 60s (as registered by the thermocouple placedat the top of specimen, Fig. 5a), reaching the temperatureof 500 �C that does not change in a remarkable way up to170s from the exposure to heat flux; then, it increases upto 850 �C (when PHRR is reached). On the other hand,EVA_DNA in bulk ignites sooner than neat EVA (ca. 30safter the test start) and its temperature grows up steadilyto 450–500 �C. After about 40s, the temperature starts todecrease as a non-completely coherent char, acting as ther-mal shield, is formed. This trend has been registered up to110s: subsequently, the char breaks and the samplereaches PHRR1 at 420 �C; such temperature is much lowerthan that reached by neat EVA. Then, the formation of amore coherent and consistent char that is thermally stableup to PHRR2 (at 160s) occurs. The presence of DNA on sur-face completely changes the temperature profile: indeed,upon exposure, after a short time of 50s, the temperaturereaches a plateau that is maintained constant during the

e calorimetry.⁄

/m2)DTHR(%)

TSR(m2/m2)

pkCO(ppm)

DpkCO(ppm)

pkCO2

(%)DpkCO2

(%)Residue(%)

– 1587 141 0 0.95 0 0– 1637 73 �48 0.54 �43 5.5�38 1520 32 �77 0.22 �77 5.0

– 1570 197 � 0.77 – 0– 1490 98 �50 0.58 �25 5.0�10 2020 31 �84 0.23 �70 4.5

Fig. 4. TSR (A), RSR (B), [CO] (C) and [CO2] (D) curves of EVA, EVA_DNA in bulk and EVA_DNA on surface as a function of time (heat flux: 35 kW/m2).

Fig. 5. Temperature profiles of EVA, EVA_DNA in bulk and EVA_DNA on surface registered by top (A) and bottom thermocouples (B), (heat flux: 35 kW/m2).

116 J. Alongi et al. / European Polymer Journal 51 (2014) 112–119

overall test. This finding confirms the formation of a ther-mal shield, more specifically a carbonaceous structure onEVA surface, able to protect the copolymer againstcombustion.

The temperature profile has been monitored also usinga thermocouple placed at the bottom of the specimen:Fig. 5b plots the corresponding profile. It is worth noticingthat, up to 150s from the test start, the thermocouple does

Fig. 6. HRR (A) and THR (B) curves of EVA, EVA_DNA in bulk and EVA_DNA on surface as a function of time (heat flux: 50 kW/m2).

Fig. 7. TSR (A), RSR (B), [CO] (C) and [CO2] (D) curves of EVA, EVA_DNA in bulk and EVA_DNA on surface as a function of time (heat flux: 50 kW/m2).

J. Alongi et al. / European Polymer Journal 51 (2014) 112–119 117

not register a remarkable change in the temperature pro-file, but only a gradual increase, hence the first ignitionof both EVA and its compounds is not detectable. Further-more, the temperature of DNA-treated samples is always

lower than that of neat EVA (in particular for EVA_DNAon surface) or at least comparable, taking into accountthe experimental error of the used instrumentation (forEVA_DNA in bulk). Thus, it is possible to conclude that

Fig. 8. Final residue of EVA_DNA on surface after cone calorimetry tests(heat flux: 50 kW/m2).

118 J. Alongi et al. / European Polymer Journal 51 (2014) 112–119

the thermal shielding effect exerted by DNA is mainly fo-cused on the surface of the specimen, which reaches a tem-perature high enough for promoting the degradationreactions in the condensed phase [4–6].

Upon exposure to a heat flux of 50 kW/m2, DNA on sur-face has proven to be more efficient than DNA in bulk(Figs. 6 and 7). Indeed, as exposed to this flux, EVA andEVA_DNA in bulk immediately start to vigorously burn(TTI value of EVA is not affected by the presence of DNAin bulk, as reported in Table 1), with the same combustionmechanism. On the other hand, DNA on EVA surface is ableto greatly postpone the ignition (116 vs. 14s for EVA_DNAon surface and EVA, respectively). This effect can be con-sidered of extreme importance as a strong delay in theignition of a polymer can avoid human casualties or prop-erty loss. PHRR is reduced by DNA in bulk (�34%), evenmore when DNA is on the surface (�81%) (Fig. 6a).However, THR values are not significantly affected by thepresence of the flame retardant (Fig. 6b). On the contrary,RSR as well as CO and CO2 yield values dramatically

Fig. 9. Temperature peak of EVA, EVA_DNA in bulk and EVA_DNA as a functionburning-through tests.

decrease, as clearly depicted in Fig. 7b–d. It is noteworthythat DNA on EVA surface favours the formation of a foamedcellular charred layer (shown in Fig. 8), which inhibits theheat and oxygen transfer, as confirmed by HRR curve. In-deed, EVA and EVA_DNA in bulk show a profile with a max-imum of heat release rate (namely, PHRR); on the contrary,the HRR curve for EVA_DNA on surface tends to a plateauthat remains constant during the exposure to the heat flux.This means that the protective barrier created by DNAforces EVA to undergo pyrolysis instead of burning; indoing so, the heat release rate remains almost constant.

3.2. Resistance to a flame application

Burning-through tests have been carried out in order tostudy the effect of DNA in bulk or on the surface on EVAheat penetration. To this aim, a butane/propane flamehas been applied to the samples under investigation(placed in vertical position), using three consecutive appli-cations of a butane/propane torch. During the test, thetemperature profile has been registered.

Fig. 9 plots the temperature peak as a function of flameapplication number. Referring to neat EVA, the tempera-ture increases by increasing the number of flame applica-tions, reaching a maximum of 346 �C during the thirdapplication (Table 2). At the end of test, neat EVA becamesoft and lost its original shape (Fig. 9). When DNA is addedto the bulk, the temperatures reached during each flameapplication are consistently lower than those of neat EVA,as a very thin film of char is formed during test. This carbo-naceous coating protects the copolymer from the subse-quent flame applications, as indicated by the final residuepresented in Fig. 9. On the other hand, when DNA is locatedon EVA surface, immediately after the first flame applica-tion, a swelled carbonaceous structure, acting as a thermalshield, is generated on the copolymer surface: indeed, thereached temperatures are much lower than those

of flame application number and corresponding final residue at the end of

Table 2Data collected for neat EVA, EVA_DNA in bulk and EVA_DNA on surface byburning-through tests.

Number offlame applications

Temperature peak (�C)

EVA EVA_DNA inbulk

EVA_DNA onsurface

1 50 40 292 167 118 403 346 188 56

J. Alongi et al. / European Polymer Journal 51 (2014) 112–119 119

registered for EVA and EVA_DNA in bulk, as well visible bycomparing the data listed in Table 2.

4. Conclusions

The flame retardant properties of EVA (18 wt.% of vinylacetate) loaded or coated with a constant amount of DNAhave been compared and thoroughly investigated by usingcone calorimetry and burning-through tests. The collectedresults have shown that the DNA coating is more effectiveas compared to its bulk addition, since the former is able toblock the ignition of the copolymer under a heat flux of35 kW/m2 (PHHR, [CO] and [CO2] reductions of 80%), aswell as to greatly postpone and reduce the combustionkinetics under a heat flux of 50 kW/m2 ([CO] and [CO2]reductions of 80% and 70%, respectively). Finally, unlikeDNA in bulk, it is worthy to note that the DNA on surfacecan exhibit exceptional thermal shielding properties as itis capable of protecting the underlying material from a bu-tane/propane torch during three consecutive applicationsfor 5s. However, in both the cases, DNA has been provento be an efficient char-former, exerting a thermal shieldingeffect on EVA copolymer.

Acknowledgement

The European COST Action ‘‘Sustainable flame retardan-cy for textiles and related materials based on nanoparticles

substituting conventional chemicals‘‘, FLARETEX (MP1105)is gratefully acknowledged.

References

[1] Horrocks AR, Alongi J. In: Alongi J, Carosio F, Horrocks AR, MalucelliG, editors. Update on flame retardant textiles: state of the art,environmental issues and innovative solutions. Fundamental aspectsof flame retardancy. Shawbury, Shrewsbury, Shropshire (UK):Smithers RAPRA Publishing; 2013, ISBN:978-1-90903-017-6[chapter 2].

[2] ISO 5660. Fire Test, Reaction to Fire, Rate of Heat Release (ConeCalorimeter Method). Geneva, Switzerland: InternationalOrganization for Standardization; 2002.

[3] Hull TR. In: Price D, Horrocks AR, editors. Advances in fire retardantmaterials. Cambridge (UK): Woodhead Publishing; 2008.

[4] Price D, Gao F, Milnes GJ, Eling B, Lindsay CI, McGrail TP. Laserpyrolysis/time-of-flight mass spectrometry studies pertinent to thebehaviour of flame-retarded polymers in real fire situations. PolymDegrad Stab 1999;64(3):403–10.

[5] Marosfoi B, Szabo A, Toldy A, Anna P, Marosi G, Tabuani D, et al. In:Lewin M, editor. Recent Advances in Flame RetardantPolymers. Norwalk (USA): BCC Inc.; 2006.

[6] Marosi G. In: Hull TR, Kandola BK, editors. Flame Retardancy ofPolymers: New Strategies and Mechanisms. Cambridge (UK): TheRoyal Society of Chemistry; 2009.

[7] Alongi J, Carletto RA, Di Blasio A, Carosio F, Bosco F, Malucelli G. DNA:a novel, green, natural flame retardant and suppressant for cotton. JMater Chem A 2013;1:4779–85.

[8] Alongi J, Carletto RA, Di Blasio A, Carosio F, Bosco F, Malucelli G.Intrinsic intumescent-like flame retardant properties of DNA-treatedcotton fabrics. Carbohydr Polym 2013;96:296–304.

[9] Carosio F, Di Blasio A, Alongi J, Malucelli G. Green DNA-based flameretardant coatings assembled through layer by layer. Polymer2013;54:5148–53.

[10] Alongi J, Di Blasio A, Cuttica F, Carosio F, Malucelli G. DNA: a novelgreen flame retardant for ethylene vinyl acetate copolymers,submitted to Industrial and Engineering Chemistry Research.(submitted for publication).

[11] Fina A, Camino G. Ignition mechanisms in polymers and polymernanocomposites. Polym Adv Technol 2011;22:1147–55.

[12] Fina A, Cuttica F, Camino G. Ignition of polypropylene/montmorillonite nanocomposites. Polym Degrad Stab2012;97:2619–26.

[13] Schartel B, Weiß A, Sturm H, Kleemeier M, Hartwig A, Vogt C, et al.Layered silicate epoxy nanocomposites: formation of the inorganic-carbonaceous fire protection layer. Polym Adv Technol2011;22:1581–92.