European Polymer Journal 48 (2012) 586–596

Contents lists available at SciVerse ScienceDirect

European Polymer Journal

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

Rheological and structural details of biocidal iPP-TiO2 nanocomposites

Cristina Serrano a, María L. Cerrada a,⇑, Marta Fernández-García a, Jorge Ressia b,c,Enrique M. Vallés b

a Instituto de Ciencia y Tecnología de Polímeros (ICTP-CSIC), Calle Juan de la Cierva 3, 28006 Madrid, Spainb Planta Piloto de Ingeniería Química – PLAPIQUI (UNS-CONICET), Camino La Carrindanga Km 7, 8000 Bahía Blanca, Argentinac Comisión de Investigaciones Científicas de la Provincia de Buenos Aires (CIC), La Plata, Argentina

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

Article history:Received 27 April 2011Received in revised form 13 December 2011Accepted 17 December 2011Available online 30 December 2011

Keywords:Nanocompositesa PolymorphRheological parametersCrystallinityInvariant

0014-3057/$ - see front matter � 2011 Elsevier Ltddoi:10.1016/j.eurpolymj.2011.12.012

⇑ Corresponding author. Tel.: +34 912587474; faxE-mail address: mlcerrada@ictp.csic.es (M.L. Cer

Nanocomposites obtained from incorporation of TiO2 nanoparticles in different amounts,ranging from 0.5 to 5 wt.%, into an isotactic polypropylene (iPP) matrix are achieved viaa straightforward and cost-effective melting process. These materials exhibit a powerfulgermicide capability over a wide variety of regular bacteria and other microorganismswidely present in the environment that cause infections and serious illness. The iPP-TiO2

nanocomposites show similar or improved structural characteristics than those of the pureiPP matrix and aspects as important as processability and final mechanical performanceseem to be not affected because of the incorporation of these TiO2 nanoparticles. Validationof time–temperature superposition of the molten polymers is observed within the temper-ature range analyzed. On the other hand, the a polymorph is the one primarily attained forthese specimens. Crystallinity and most probable crystallite size are slightly dependent onTiO2 content, both increasing as oxide composition is enlarged.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction offering an adequate balance between performance,

The use of photocatalytic semiconductor oxides hasemerged as a successful technology in the struggle againstbiological risks to guarantee the safety of products relatedto food/beverage packaging or containers for biomedical/pharmaceutical materials/devices. TiO2-Anatase is by farthe most widely used photocatalyst. Under UV illumina-tion TiO2 generates energy-rich electron–hole pairs ableto degrade cell components of microorganisms renderinginnocuous products. Moreover, no weakness with respectto the microorganism nature (bacteria, virus, fungus, etc.)is known [1,2]. Consequently, its incorporation as a constit-uent in polymeric multicomponent materials could be apromising alternative within the field of food and drugpackaging as well as medical devices.

The packaging industry can choose from an array of poly-meric materials, although polyolefins (polyethylene andpolypropylene) present a flexible and robust solution

. All rights reserved.

: +34 915644853.rada).

processability and cost to make them the industry’s materialof choice. The main factors that influence the final decisionfor a material selection are preservation, protection andsafety. Packaging must protect the contents from physicaldamage and from external contamination, i.e., microbiolog-ical contamination. It must also preserve the quality of thecontents, whether for a short shelf-life of some days, or foran extended shelf-life of several months. The intrinsicallyinert nature of polyolefins and, specifically, of polypropyl-ene helps to make this polymer a perfect candidate for safefood and drug packaging complying with relevant nationaland international regulations.

Therefore, isotactic polypropylene (iPP) has in recentyears been benefited versus other more customary materi-als, such as cellophane, metals and paper on account of itssuperior puncture resistance, low sealing threshold andcompetitive price. Although packages processed using con-ventional Ziegler–Natta iPP resins provide an attractiveexterior finish, they slightly obscure the clarity in appear-ance of the contents. The use of metallocene catalysts forits synthesis leads to iPPs with a significant improvement

C. Serrano et al. / European Polymer Journal 48 (2012) 586–596 587

in clarity, this feature making metallocene iPP an idealmaterial for packaging in general and for food and drugpackaging in particular.

The excellent biocidal capabilities exhibited by nano-composites based on a metallocene iPP and TiO2 have beenreported in previous works [3–5], varying either the nano-particle composition, this ranging from 0.5 to 5 wt.%, at agiven compatibilizer content, or the interfacial agent frac-tion at a specific TiO2 content of 2 wt.%. This strikingantimicrobial activity has been also found by using an eth-ylene–vinyl alcohol copolymer as polymeric matrix [6–8].The existence of energy/charge transfer(s) through thesepolymer/oxide interfaces and the subsequent developmentof new electronic states within the nanocomposite system[9] are behind this optimal germicide capability. At indus-trial level, the easiness of production as well as the lowestcost of new materials is a crucial aspect for their furthercommercialization. Very small concentrations of nanopar-ticles can significantly alter the phase behavior and theflow characteristics in polymeric materials.

Insight about how nanoparticles influence the associ-ated morphological structure [10–12] and system dynam-ics [13–15] of polymer–nanoparticle mixtures is onlybeginning to emerge, and the advancement of knowledgein these areas will be a key factor to develop design rulesto engineering materials with desired properties. The studyof the rheological behavior of these polymeric nanocom-posites with extremely good biocidal properties turns outto be necessary in order to learn how different their pro-cessability can be in comparison with that exhibited bythe neat iPP homopolymer. In addition to analyze the effectof different TiO2 contents on the rheological response ofthese nanocomposites, the influence of TiO2 nanoparticleson the polymeric crystalline morphology, the viscoelasticresponse in the solid state and the decomposition behavioris also evaluated. Several techniques are then required forthis investigation. They comprise rheological measure-ments, either in the molten or solid state, X-ray diffractionat wide (WAXS) and small angles (SAXS), differential scan-ning calorimetry (DSC) and thermogravimetry (TGA).

2. Experimental part

2.1. Nanocomposite preparation

The TiO2 component (characteristic primary particlesize below 10 nm) was prepared using microemulsion syn-thetic route by addition of titanium (IV) isopropoxide(Aldrich) to an inverse emulsion containing an aqueousphase dispersed in n-heptane (Panreac), using TritonX-100 (Aldrich) as surfactant and 1-hexanol (Aldrich) ascosurfactant. The mixture was stirred for 24 h, centrifuged,decanted, rinsed under stirring five consecutive times withmethanol (twice), water (twice) and acetone (one) to elim-inate any portion from the organic and surfactant media,dried at 110 �C for 24 h and calcined at 500 �C for 2 h.

A commercially available metallocene-catalyzed isotac-tic polypropylene, iPP (Basell Metocene X50081: melt flowindex of 60 g/10 min at 230 �C/2.16 kg, ASTM D1238),meeting FDA requirements for food contact (Federal Regu-lations, 21 CFR 177.1520), was used as polymeric matrix in

the preparation of these iPP-TiO2 nanocomposites. PP-g-MAH, polypropylene wax partially grafted with maleicanhydride, was used as interfacial agent (Licomont� AR504 fine grain from Clariant) in a composition of 80 wt.%with respect to the content in TiO2 nanoparticles. SeveralTiO2 compositions have been examined: 0.5, 1, 2, and5 wt.%. The resultant three-component nanocompositeswere labeled as iPPTix along the manuscript, being x theTiO2 content. These nanocomposites were preparedthrough a straightforward melt processing in an internalmixer with volumetric capacity of 3 cm3 at 160 �C and at60 rpm for 5 min. A previous TiO2 sonication was com-pleted in an ultrasonic device to minimize the aggregationof nanoparticles and maximize the performance of resul-tant nanocomposites.

Films of these TiO2 nanocomposites were obtained bycompression molding for 5 min in a Collin press betweenhot plates (175 �C) at a pressure of 1.5 MPa. After this,the films were rapidly quenched (at about 100 �C/min) byrefrigerating the plates of the press with cold water.

2.2. Sample characterization

The antimicrobial tests were performed using an Entero-coccus faecalis clinical isolate brs30 from human biliary.Cells were streaked from a glycerol stock onto an appropri-ate agar plate, grown overnight at 45 �C, and subsequentlyused for photochemical cell viability assays. To study theantimicrobial activity of the titania nanomaterials, theywere contacted with a solution containing microbial cells(ca.108–9 CFU mL�1) suspended in 1 mL broth solution.The system was placed in the UV spectrometer chamber(UVIKON 930) and irradiated with a UV light of 280 nmfor different periods of time. Loss of viability after eachexposure time was determined by the viable count proce-dure on Luria–Bertani agar plates after serial dilution (10�2

to 10�5). All plates were incubated for 24 h prior to enu-meration at the temperatures above mentioned, specificfor each microorganism. A minimum of three experimentalruns was carried out to determine the antimicrobialactivity.

The rheological characterization was carried out insmall-amplitude oscillatory shear mode using a dynamicrotational rheometer from Rheometrics Inc. (RheometricsDynamic Analyzer RDA-II). The tests were performed usingparallel plates of 25 mm in diameter, at a frequency rangebetween 1 and 500 rad s�1, and a temperature range of170–250 �C. All tests were carried out at small strains inorder to assure the linearity of the dynamic responses[16]. To verify this, the series of frequency sweeps were re-peated twice with the same sample at different strains.Excellent agreement between these results was found inall cases.

Transmission electron microscopy was performed atroom temperature in a 200 kV JEM-2100 JEOL microscopeto analyze material homogeneity. Samples were cut in thinsections (50–70 nm) by crioultramicrotomy (Leica EM UC6).

The X-ray synchrotron study was performed in thesoft-condensed matter beamline A2 at Hasylab (Hamburg,Germany), working at a wavelength of 0.150 nm. Theexperimental setup includes a specimen holder, a MARCCD

588 C. Serrano et al. / European Polymer Journal 48 (2012) 586–596

detector for acquiring two-dimensional SAXS patterns(sample-to-detector distance being 260 cm) and a lineardetector for 1D WAXS measurements (distance 21 cm). Asample of crystalline PET was used for WAXS calibrationand the different orders of the long spacing of rat-tail cor-nea (L = 65 nm) were utilized for the SAXS detector. The 2DX-ray diffractograms were processed using the A2tool pro-gram developed to support beamline A2 data processing.The profiles were normalized to the primary beam inten-sity and the background from an empty sample wassubtracted. All experiments comprise the heating of sam-ples from 26 up to 170 �C at 8 �C/min. The data acquisitionwas done in frames of 15 s. The WAXS degree of crystallin-ity, fc

WAXS, was determined from the X-ray diffractogramsafter subtraction of the amorphous profile [17].

Calorimetric analyses were carried out in a Perkin–Elmer DSC7 calorimeter connected to a cooling systemand calibrated with different standards. The sampleweights ranged from 6 to 9 mg and the heating rate usedwas 8 �C min�1. For crystallinity determinations, fc

DSC, avalue of 209 J g�1 has been taken as enthalpy of fusion ofa perfect crystalline material [18]. On the other hand, theglass transition temperature, Tg, was determined as thetemperature where the specific heat increment is half ofthe total one at the transition.

Dynamic mechanical relaxations were measured with aPolymer Laboratories MK II Dynamics Mechanical ThermalAnalyser, working in a tensile mode. The storage modulus,E0, loss modulus, E00, and the loss tangent, tand, of eachsample were obtained as function of temperature overthe range from �150 to 150 �C at fixed frequencies of 3,10, 30, 50 Hz, and at a heating rate of 1.5 �C min�1. To carryout these measurements, strips around 4.0 mm wide and18 mm length were cut from the molded sheets. The

0 5 10 15 20 25 300

2

4

6

8

10

0 5 10 15 20 25 300

2

4

6

8

10

iPPTi0

Log

CFU

ml-1

iPPTi2

Log

CFU

ml-1

Time (min)

Enterococcus faecalis

Fig. 1. Process come-up logarithmic reduction of microorganism population sirradiation time for iPP control and iPPTix samples. Values are obtained from th

apparent activation energy values were calculated on thebasis of tand according to an Arrhenius-type equation, con-sidering an accuracy of ±0.5 �C in the temperature assign-ment from the maxima.

The mass loss was estimated by thermogravimetryusing a TA Instruments TGA Q500 equipment working un-der inert and oxidant atmospheres. The equipment wascalibrated according to standard protocols. The sampleweights ranged from 4 to 6 mg, and the heating rate was10 �C min�1.

3. Results and discussion

Compared with other biocidal agents, an excellentpower for destruction of regular bacteria, as Bacillus stearo-thermophilus, Escherichia coli, Staplyococcus aureus, Pichia.jadini, and other microorganisms widely present in theenvironment that cause infections and serious illness, asPseudomonas aeruginosa and E. faecalis, has been found intitania nanoparticle-polypropylene nanocomposites [3–5].As an example, Fig. 1 shows this germicide activity againstthe E. faecalis Gram positive bacterium. The blank experi-ment in presence of the iPP matrix demonstrates the rela-tive innocuousness of UV radiation. Incorporation ofdifferent nanoparticled TiO2 contents has an important im-pact on the cell inactivation with respect to the blank testincreasing the initial velocity of elimination and furtherdecreasing the final log-reduction at the end of the exper-iments. Moreover, it is also observed that maximum per-formance is obtained using the iPPTi2 nanocomposite. Inthis case, an almost complete cell inactivation of the E. fae-calis is reached, accounting for a log-reduction of near ca. 8units. The germicide behavior exhibited by thesenanocomposites is dependent on the microorganisms eval-

0 5 10 15 20 25 300

2

4

6

8

10

0 5 10 15 20 25 300

2

4

6

8

10

iPPTi1

iPPTi5

Time (min)

uspended in LB medium. Survival curves of E. faecalis as a function ofree different measurements.

(a)

(b)

Fig. 2. (a) Reduced frequency dependence of storage modulus, G0(x) andloss modulus, G00(x) and (b) of dynamic viscosity ((bT/aT) g0) observed inthe metallocene iPP and its nanocomposites with TiO2.

C. Serrano et al. / European Polymer Journal 48 (2012) 586–596 589

uated [3,5] but, in all cases, a significant improvement ofthe biocidalactivity is detected with the employment of the iPPTi2nanocomposites.

Interactions of TiO2 with the polymer appear as a mustin order to boost the biocidal action [6,9]. This conclusionis based on the evidence that no detection of inorganicdomains across the film is noticed for the samples withconcentrations of TiO2 below 2 wt.%; whereas two typesof zones are observed in the iPPTi5 specimen: one lackingin oxide agglomerates and another with presence of thoseaggregates. Regions free of oxide agglomerates are, how-ever, dominant (70–90% of the volume).

The existence of this intimate contact between iPP andTiO2 will also affect the whole spectrum of propertiesshown by these nanocomposites, including their process-ability and, consequently, the capability of being producedand commercialized at an industrial level. Theoreticalworks suggest that the nature of the particle/polymerinteractions -attractive vs. repulsive- may play a role inthe changes exhibited by the viscoelastic properties ofthe molten polymeric matrix [19]. A faster polymerdynamics is observed for repulsive systems compared tothe one found in the pure melt and neutral systems andit is associated with a decrease in the polymer matrix den-sity. On the other hand, the slowing of polymer dynamicsdescribed in systems with significant attractive interac-tions is ascribed to a dramatic reduction in mobility ofadsorbed polymeric segments along the nanoparticlesurface.

Fig. 2 shows the master curves obtained for the storageand loss shear moduli as well as the dynamic viscosity ofthese nanocomposites. It is well known that bi-logarithmicplots of the isotherms of the G0(x), G00(x), and dynamic vis-cosity g0(x) can be superimposed for thermo-rheologicallysimple materials by horizontal shifts log(aT), along thelog(x) axis, and vertical shifts given by log(bT) such that[16]:

bT G0ðaTx; TÞ ¼ G0ðx; Tref Þ

bT G00ðaTx; TÞ ¼ G00ðx; Tref Þ

Fig. 3. TEM micrograph of the iPPTi2 nanocomposite.

ðbT=aTÞg0ðaTx; TÞ ¼ g0ðx; Tref Þ

All isotherms measured for the neat iPP and for the var-ious iPPTix nanocomposites are superimposed in this way,allowing attainment of time–temperature master curves.In the case of pristine polymeric samples, it is expectedthat they should exhibit characteristic homopolymer-liketerminal flow behavior, expressed by the power-laws closeto G0(x) �x2 and G00(x) �x at the temperatures and fre-quencies at which the rheological measurements are car-ried out. This terminal flow zone is clearly observed fromthe G0 and G00 representations (see Fig. 2a). This responseis distinct to that exhibited by other nanocomposites ofiPP and in situ generated TiO2 particles [20], with composi-tion ranging from 2.9 to 9.3 wt.%, where significantchanges in the rheological behavior in the terminal regionwas reported even at the lowest TiO2 content. The absenceof any terminal flow zone at low frequencies and the

appearance of a secondary plateau are shown in thosematerials. In our case, the terminal relaxation zone inthe nanocomposites is, however, observed probably be-cause of the excellent oxide/polymer contact existing atinterfaces and the non-presence of important TiO2 agglom-erates, as seen in Fig. 3 for the iPPTi2 nanocomposite. Thesefeatures seem to suggest the breakdown of particle–parti-cle interactions [21], allowing that conformations of mac-romolecular chains remain almost unaffected. Only aslight raise on the magnitude of G0 and G00 compared to thatfound in the pristine polypropylene is observed withincreasing concentration of the TiO2 content.

The reduced curve of dynamic viscosity is depicted inFig. 2b. Viscosity rises as inorganic content does in thenanocomposites and this increase is more significant at

590 C. Serrano et al. / European Polymer Journal 48 (2012) 586–596

low frequencies in the Newtonian plateau region. Conse-quently, g0 values obtained by extrapolating g0 to zero fre-quency are slightly dependent on TiO2 composition. Therise of g0 with increasing TiO2 content is shown in the insetof Fig. 2b. It should be mentioned that Y axis in Figure 2b isnot represented in the same scale than that in Figure 2a toallow a better visualization.

The fact that these nanocomposites show an excellentbiocidal activity with respect to the practically non-existing one in the neat iPP and that their viscoelastic re-sponses in the melt are similar to those seen in the pureiPP is quite interesting since they prove an attractive bio-cidal specificity with practically unchanged rheologicalproperties. This is very important from a practical pointof view since the processing conditions of these iPP-TiO2

nanocomposites should be analogous to those of the virginiPP.

It is also mandatory to know how different TiO2 con-tents affect the final crystalline structure of these nano-composites as well as their phase transitions, relaxationprocesses and thermal stability. Fig. 4a shows clearly thecrystalline character at room temperature of the differentspecimens recorded in a transmission mode. The most typ-ical and stable crystalline structure in iPPs is the mono-clinic a form, and was firstly characterized by Natta andCorradini [22]. The orthorhombic c modification, easilyidentified because of its characteristic (117) diffractionpeak at around 20� in 2h profiles [23], is specially favored

10 15 20 25 30

anatase

(a)

(150, 060) +

(101)

(117) γ polymorph

(131

, 041

)

(111

)

(130)(040)

(110)

iPPTi5

iPPTi2

iPPTi1

iPPTi05

iPPTi0

rela

tive

inte

nsity

(a.

u.)

2 θ (º)

10 15 20 25 30

(b)(117) γ polymorph

iPPTi5

iPPTi2

iPPTi1

iPPTi05

iPPTi0

rela

tive

inte

nsity

(a.

u.)

2 θ (º)

Fig. 4. WAXS diffraction patterns of pure iPP and the different nanocom-posites at room temperature: (a) quenched films as processed and (b)samples crystallized at 8 �C/min after their first heating.

in the case of iPP synthesized by metallocenic catalystsdue to the presence of errors homogeneously distributedamong the different polymer chains [24]. It can beobserved in this Fig. 4a that these profiles exhibit the maindiffractions characteristic of the a monoclinic iPP crystal-line modification [24,25]. On the other hand, it is notice-able that there is practically no evidence of the cmodification in the pristine iPP. The distinctive (117)reflection of this polymorph appears in the WAXS profilesof the different nanocomposites although in a small con-tent. These features can be ascribed to the relatively highcooling rate used during processing of these films sincethe c form is attained in other similar metallocenic iPPhomopolymers when crystallization is performed at veryslow rate [26]. In fact, the profiles at room temperature(depicted in the Fig. 4b) obtained after crystallizing at arate of 8 �C/min during real-time variable-temperatureWAXS experiments with synchrotron radiation show thedeep effect that cooling rate has on the development of thisc polymorph. In addition, it seems that the presence ofTiO2 favors this c form and, consequently, its content in-creases in the nanocomposites if compared with thatexhibited by iPPTi0 homopolymer. This larger develop-ment of c phase could be ascribed in the nanocompositesto the nucleant effect that the presence of TiO2 nanoparti-cles has on the crystallization of iPP macromolecules,which will be commented later on DSC results. The shiftof crystallization to higher temperatures favors the forma-tion of c crystals [27] up to a maximum value well-abovethe dynamic crystallization temperature here observed.

Crystallinity of the different specimens can be esti-mated from the WAXS profiles. The procedure used is thefollowing: firstly, the amount of amorphous phase hasbeen calculated from the comparison with the diffractionprofile corresponding to a totally amorphous, elastomeric,polypropylene sample, obtained in the same diffractome-ter and configuration [17]. In this way, the total X-raycrystallinity at room temperature is determined from thediffractograms representing the pure crystalline compo-nents of each specimen after subtracting the scaled amor-phous profile. The corresponding values obtained at roomtemperature are listed in Table 1. The degree of crystallin-ity is within the range of that found in analogous quenchedmetallocenic iPP specimens [26–30]. Under the quenchingconditions imposed during film preparation, crystallizationtakes place at conditions far from equilibrium and the fastcooling applied limits even more the development of crys-tallites and their perfection. Incorporation of TiO2 nanopar-ticles has a slight effect on the degree of crystallinityreached, its normalized value increasing as TiO2 contentdoes (see Table 1). Therefore, intensity of the differentreflections slightly rises in the nanocomposites (see Fig. 4).

The results obtained by the real-time variable-tempera-ture synchrotron radiation experiments at wide anglesobtained from the first heating after processing for thepure iPPTi0 homopolymer and all the iPPTix nanocompos-ites are represented in Fig. 5. This figure shows animprovement of crystalline structure as temperature is in-creased and, then, reflections become narrower in the dif-ferent specimens studied. Moreover, the (101) reflectioncharacteristic of TiO2 anatase polymorph is observed at

Table 1Normalized crystallinity (fc

WAXS), most probable long spacing (LSAXS) and most probable crystallite size (lc) estimated at room temperature; meltingtemperature determined from WAXS, SAXS and DSC measurements (Tm

WAXS, TmSAXS, and Tm

DSC, respectively); melting enthalpy (DHm), overall crystallinitydegree (fc

DSC) and crystallization temperature (Tc) estimated from DSC experiments for the different specimens.a.

Sample fcWAXS Tm

WAXS (�C) LSAXS (nm) TmSAXS (�C) lc (nm) DHm

DSC (J/g) TmDSC (�C) fc

DSC TcDSC (�C)

iPPTi0 0.54 143 10.2 142 5.5 98 139.0 0.47 115.5iPPTi05 0.55 143 11.3 142 6.2 102 141.0 0.49 117.0iPPTi1 0.56 144 11.3 143 6.3 105 139.5 0.50 117.5iPPTi2 0.56 143 11.2 143 6.3 105 139.5 0.50 117.5iPPTi5 0.57 143 11.4 144 6.5 105 140.5 0.50 116.5

a Estimated errors: temperatures ±0.5 �C; enthalpies ±4 J g�1.

12 18 24

2 θ (º)2 θ (º)2 θ (º)2 θ (º)2 θ (º)

rel.

int.

(a.u

.)

12 18 24

iPPTi5

12 18 24 12 18 24168 ºC

28 ºC

iPPTi0 iPPTi05 iPPTi1 iPPTi2

12 18 24

Fig. 5. Real-time variable-temperature WAXS profiles obtained with synchrotron radiation for all the quenched samples as processed in a meltingexperiment at 8 �C/min. Only one every three frames is plotted for clarity.

40 80 120 1600.4

0.5

0.6

0.615

0.620

0.625

(b)

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

wre

f110 (

º)

T (ºC)

(a)

d110 (

nm)

40 80 120 160

0.0

0.2

0.4

0.6

(d)

deri

vativ

e

T (ºC)

(c)

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

norm

aliz

ed f

cWA

XS

Fig. 6. Temperature dependence, on melting, for the (110) monoclinic reflection of: (a) peak spacing; (b) the width at half height of this diffraction; (c)normalized crystallinity determined from it; and (d) crystallinity derivative corresponding to the neat homopolymer and its nanocomposites.

C. Serrano et al. / European Polymer Journal 48 (2012) 586–596 591

around 25� in these 2h plots. Although this diffractionoverlapped with other small one from iPP, it is unambigu-ously observed that its intensity increases as TiO2 contentis raised in the nanocomposite.

A deeper analysis of the position, width and intensity ofthe (110) diffraction in those real-time variable-tempera-ture diffractograms allows us obtaining some importantquantitative information. Therefore, Fig. 6a represents thedependence on temperature of the position of this (110)reflection. The trend observed in the nanocompositesseems to point out a decrease of d110 spacing as TiO2 con-tent increases, the nanocomposite iPPTi2 exhibiting similar

values to those shown by iPPTi0 homopolymer. This fea-ture indicates that incorporation of very small amount ofTiO2 leads to a small distortion of the (110) plane withinthe crystalline lattice (iPPTi05 and iPPTi1 specimens). Thischange is not observed in the iPPTi2 whereas iPPTi5 pre-sents lower d110 values, pointing out to a more compactcrystalline morphology. The temperature variation of d110

is, however, very similar for the different samples, withan apparent temperature coefficient of (4.3 ± 0.3) �10�5 nm �C�1. The individual values (4.3 � 10�5 nm �C�1

for iPPTi0, 4.0 � 10�5 nm �C�1 for iPPTi05 and iPPTi1,4.1 � 10�5 nm �C�1 for iPPTi2 and 4.6 � 10�5 nm �C�1 for

0.04 0.08 0.12

iPPTi0

1/d (1/nm) 1/d (1/nm) 1/d (1/nm) 1/d (1/nm) 1/d (1/nm)

rel.

int.

(a.u

.)

0.04 0.08 0.12 0.04 0.08 0.12 0.04 0.08 0.12

iPPTi05 iPPTi1 iPPTi2 iPPTi5

0.04 0.08 0.12

Fig. 7. Real-time variable-temperature Lorentz-corrected SAXS profiles obtained with synchrotron radiation for all the quenched samples as processed in amelting experiment from 28 to 168 �C at 8 �C/min. Only one every three frames is plotted for clarity.

40 80 120 160

10

20

30

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

L (

nm)

T (ºC)

SAX

S in

vari

ant (

a.u.

)

(b)

(a)

inva

rian

t der

ivat

ive

(a.u

.)

Fig. 8. Temperature dependence, on melting, of (a) the SAXS relativeinvariant and its derivative and (b) most probable long spacing corre-sponding to the neat homopolymer and its nanocomposites.

592 C. Serrano et al. / European Polymer Journal 48 (2012) 586–596

iPPTi5) show the influence of TiO2 introduction on the finalcrystalline characteristics.

Other parameter that provides useful information is thewidth of the diffractions and its dependence on tempera-ture. Fig. 6b depicts the variation of this parameter forthe distinct nanocomposites. Width of monoclinic (110)diffraction remains rather constant in the iPPTi0 up toaround 60 �C becoming narrower at higher temperaturesas consequence of the improvement of crystallites withtemperature. A very small incorporation of nanoparticlesin iPPTi05 leads to a distortion of crystalline lattice, asaforementioned. The nanocomposites iPPTi1 and iPPTi2show a similar dependence than the homopolymerwhereas the lattice enhancement and, consequently, nar-rowness of the (110) reflection start at lower temperaturesin iPPTi5.

The analysis of intensity in this (110) diffraction allowsobtaining information on crystallinity changes with tem-perature (see Fig. 6c). These values correspond to crystal-linity normalized to the actual polymer content withineach nanocomposite. There are small differences withincreasing nanoparticle content, crystallinity increasingas TiO2 content does. The location of the melting of thesecrystalline entities can be estimated by derivation fromthis temperature dependence, as shown in Fig. 6d. Thenumerical values are listed in Table 1.

The real-time variable-temperature Lorentz-correctedSAXS profiles corresponding to the first melting processof the polymeric component in the different nanocompos-ites are shown in Fig. 7. It has to be commented that a peri-odicity peak is additionally observed after melting of theiPP polymeric matrix. This correlation peak is ascribed toa characteristic spacing between TiO2 nanoparticles [20].Looking at Fig. 7, a clear long spacing is seen for all of thespecimens after subtracting the TiO2 correlation peak. Animportant shift to lower 1/d values of the peak related tothe long spacing with temperature is observed in all ofthese specimens, independently of TiO2 content, associatedwith the improvement of the iPP crystalline morphologyby effect of temperature and, then, the thickening of itscrystallites. Further information can be obtained fromthese profiles by analysis of the relative SAXS invariant(see Fig. 8a). This is a magnitude used in the lamellar stackmodel theory of semicrystalline polymers which is directlyrelated to the electron density differences between the twophases (amorphous and crystalline), so that it has rather

important variations during crystallization, recrystalliza-tion or melting processes in semicrystalline polymers[31–33]. Data prove the existence of different regions: aninitial one, up to around 55 �C, where a very small increaseof the SAXS invariant is seen; a second zone, between 55 �Cto around 130 �C, with an important increase of invariantfollowed by a small diminishment at the end of this inter-val; and a final one characterized by a primary and sharpdecrease, related to the main melting process.

These distinct regions can also be observed from thevalues of the most probable long spacing as function oftemperature, deduced from the Lorentz-corrected SAXSprofiles, as observed in Fig. 8b. The first one, up to around55 �C, where the long spacing is almost constant; a secondregion, between 55 �C to around 130 �C, with a moderateincrease of L; and a final one, with a very important thick-ening, this being ascribed to a considerable melting-recrystallization phenomena.

It is also noticeable from Fig. 8b the development ofthicker crystallites in the nanocomposites with respect tothose existing in the iPP matrix probably due to the nucleanteffect performed by TiO2 nanoparticles. This assumption canbe confirmed once long spacing, LSAXS, is determined takingthe values of total WAXS crystallinity, fc

WAXS. The most

C. Serrano et al. / European Polymer Journal 48 (2012) 586–596 593

probable crystallite size in the direction normal to thelamellae, lc, can be estimated by assuming a simple two-phase model, i.e., lc = fc

WAXS � LSAXS. The results for lc are alsolisted in Table 1, a slight increase being observed as TiO2

content is raised in the nanocomposites. On the other hand,information on melting temperature is also obtained fromthe corresponding derivatives of the relative invariant, ascan be deduced from Fig. 8a. The agreement with thosedetermined from the WAXS profiles is really good, as canbe seen.

The crystalline nature of these nanocomposites can bealso figured out from calorimetric results represented inFig. 9a and b for the melting and crystallization processes,respectively. An endothermic peak ascribed to the meltingprocess of the a form is exhibited as seen in Fig. 9a. Weshould keep in mind that this a polymorph is practicallythe unique one developed in these as-processed Q speci-mens, as deduced from WAXS diffractograms. However,two melting processes seem to be merged in the homopol-ymer and the endotherm observed in the nanocompositesis rather asymmetric. These features are related to themelting-recrystallization phenomena of these a crystal-lites, which become so important at high temperatures.In addition to these melting-recrystallization phenomena,the existence of two overlapped endotherms could be alsodue to melting of other polymorph or to the melting of twodifferent crystallite populations. Nevertheless, thoseassumptions are not accomplished in these materials sinceWAXS profiles do not show development with temperatureof an important content of other polymorph different thanthe a form and SAXS peaks do not point out the presence oftwo distinct crystallite populations. On the contrary, bothWAXS and SAXS results confirm the improvement on tem-perature of crystalline structure through melting-recrys-tallization processes, as clearly seen in Figs. 6b and 8a.

Another aspect deduced from these melting curves isthe enthalpy of melting (see Table 1), and, consequently,the DSC crystallinity if the enthalpy of melting for the100% crystal is known. If the widely used value [18] of

40 80 120 160

iPPT

(a)

0.5 W/g

iPPT

iPPT

iPPT

iPPT

Hea

t Flo

w (

W/g

)

T (ºC)

Fig. 9. DSC curves corresponding to: (a) the first heating process and (b) the crystwith distinct TiO2 nanoparticle contents.

209 J g�1 is considered as enthalpy of fusion of a perfectcrystalline material, the DSC crystallinity variation is thatreported in Table 1. As previously observed from WAXS re-sults, the values for the nanocomposites are slightly higherthan that found in the iPPTi0 sample and there are not sig-nificant differences as TiO2 content increases in thespecimens.

Fig. 9b shows the crystallization process for the differ-ent samples. It is clear the nucleant effect that incorpora-tion of TiO2 has on the iPP matrix. Consequently,crystallization temperature, Tc, is moved to higher temper-ature (see Table 1). The nucleant influence that the highestcontent of TiO2 exerts (sample iPPTi5) is, however, less sig-nificant than that observed in the nanocomposites withlower TiO2 composition. This feature might be ascribed tothe presence of oxide agglomerates in iPPTi5, which arenot observed in the other nanocomposites with lower con-tents, fact that can minimize the nucleant effect of thenanoparticles.

Fig. 10 shows the results related to the thermal stabilityof iPPTi0 and the different nanocomposites under inert andoxidant conditions. Looking first at curves achieved undera nitrogen atmosphere (Fig. 10a), it can be said that degra-dability is not affected by incorporation of oxide nanopar-ticles. Degradation seems to take place in a unique stage(Fig. 10b) and the TiO2 content can be estimated fromresidual mass at the end of the experiment. A good agree-ment was found if compared with the theoretical values:0.6, 1.1, 1.9 and 4.9 for the iPPTi05, iPPTi1, iPPTi2 and iPP-Ti5, respectively. The picture significantly changes whenthermal stability is evaluated in an oxidant environment(Fig. 10c). In this case the degradation mechanism is morecomplex and takes place in several steps (Fig. 10d). Also,the presence of TiO2 nanoparticles favors degradabilityand the temperature of maximum degradation is shiftedtoward lower values than that corresponding to the homo-polymer. Nevertheless, the onset of degradation in thenanocomposites is moved to higher temperatures about20 �C, this means that the onset of the degradation process

100 110 120 130

i0

i5

i2

i1

i05

1 W/g(b)

endo

T (ºC)

allization corresponding to the neat homopolymer and its nanocomposites

0

20

40

60

80

100

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

Mas

s L

oss

(%)

100 200 300 400 500 600

inert atmosphere

-dw

/dt (

w%

ºC

-1)

T (ºC)

oxidant atmosphere

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

(c)

200 300 400 500 600

iPPTi5

iPPTi2

iPPTi1

iPPTi05

iPPTi0(d)(b)

(a)

T (ºC)

Fig. 10. Left plot: (a) TG and (b) DTG curves obtained under an inert atmosphere; Right plot: (c) TG and (d) DTG curves obtained under an oxidantatmosphere of neat homopolymer and the different nanocomposites analyzed. DTG curves have been vertically shifted for clarity.

100

1000

10000

0.00

0.04

0.08

-150 -100 -50 0 50 100 15010

100

iPPTi0 iPPTi05 iPPTi1 iPPTi2 iPPTi5

E' (

MPa

)ta

n δ

E'' (

MPa

)

T (ºC)

Fig. 11. Temperature dependence of the two components to the complexmodulus (E0 and E00) and of the loss tangent (tand) for the neathomopolymer and its nanocomposites with distinct TiO2 nanoparticlecontents.

594 C. Serrano et al. / European Polymer Journal 48 (2012) 586–596

is postponed with the incorporation of TiO2. These featuresderived from the TiO2 catalytic effect could have a highinterest from an industrial and commercial standpoint.

After assessing their excellent biocidal properties, theconstancy in processability conditions, the crystallite char-acteristics developed during film preparation as well as theinfluence of nanoparticles in the thermal degradability forthese nanocomposites with different TiO2 contents, thestudy of their viscoelastic behavior in solid state is essen-tial to learn the final mechanical response. Therefore,Fig. 11 shows the viscoelastic mechanisms that occur in abroad temperature interval, from �150 to 150 �C. Threeprocesses are observed in the plots of loss magnitudes,tand and E00, labeled as a, b and c in order of decreasingtemperatures. The a relaxation is related to motions withinthe polymer crystalline phase, especially to defect diffusion[34]. The mechanism that takes place at around 0 �C (b pro-cess) is ascribed to generalized motions of long chainsegments that occur along the glass transition. The cooper-ative nature of this movement explains the importantdecrease in E0 found in this temperature range. The otherrelaxation process observed at temperatures lower thanthat related to cooperative motions is labeled as c and isassociated with rotational motions of methyl groups frompolypropylene. It does actually appear as a shoulder andnot as a well-defined peak even in the iPP homopolymer(see Fig. 11).

Due to the partial overlap of the b and c relaxations, theseparation of the loss response into the different processcontributing to the overall viscoelastic spectrum is ratherconvenient to estimate more accurately the location ofthe different relaxations. With this purpose, the deconvo-lution of tand magnitude into three distinct relaxationswas performed. The tand response has been consideredas composed by three distinct Gaussian curves, one foreach observed relaxation process (also loss modulus curves

can be fitted to this type of mathematical functions). Sucha deconvolution does not have a theoretical basis that canexplain satisfactorily the shape of the dependence of tand(or loss modulus) on temperature though some factors thatcan influence it are known. A method of curve deconvolu-tion [35] has been proposed to analyze the dynamicmechanical loss curves in the region of the glass transitionof several polymers, confirming the validity of this empir-ical approximation. In addition, it was shown that a

Table 2Relaxation temperatures (tand basis, at 3 Hz) for the different processes and storage modulus values at different temperatures found in theiPPTi0 homopolymer and its nanocomposites with different TiO2.nanoparticle contents.

Sample Tc (�C) Tb (�C) Ta (�C) E0�100 �C (MPa) E025 �C (MPa) E0100 �C (MPa)

iPPTi0 �72.0 0.0 67.0 4900 1800 315iPPTi05 �70.0 �2.5 69.0 4700 1650 310iPPTi1 �70.0 1.5 69.5 4700 1800 310iPPTi2 �70.0 2.0 68.5 5000 1800 320iPPTi5 �70.0 4.0 68.5 4700 1650 310

C. Serrano et al. / European Polymer Journal 48 (2012) 586–596 595

Gaussian function provided the best fitting. This methodconstitutes a useful tool to determine the peak positionswhether a good overall fitting is attained over the wholeexperimental temperature range measured.

The results obtained from this procedure are listed inTable 2. In view of these results we may conclude thatthe incorporation of TiO2 nanoparticles does not have aremarkable effect on the location and intensity of the cand a secondary relaxations. However, a shift to highertemperature with oxide content is seen for the primary bprocess. Regarding the elastic component of Young’s mod-ulus, the values for the different specimens are also similarat different temperatures. Stiffness in the solid state is notsignificantly affected by introduction of small contents innanoparticles. This means that the nanocomposites main-tain the good mechanical performance exhibited by theirpolymeric iPP matrix.

4. Conclusions

Several iPP/TiO2 nanocomposites exhibiting an extraor-dinary power for destruction of numerous microorganismswere prepared via a straightforward and cost-effective ap-proach. These polymers show similar or improved struc-tural characteristics than those of the pure iPP matrixand aspects as important as processability and finalmechanical performance are not affected because of theincorporation of these TiO2 nanoparticles. Time–tempera-ture superposition principle is validated for all of the nano-composites and Newtonian viscosity slightly increases asTiO2 content is enlarged in the final material. Thequenched films primarily develop the a polymorph,although the presence of nanoparticles seems to favorsomehow the minor appearance of the orthorhombic c lat-tice, probably because of their nucleant effect and, conse-quently, the shift of crystallization to slightly highertemperatures. On the other hand, crystallinity and mostprobable crystallite size tend to rise as TiO2 content in-creases. The presence of these nanoparticles does not affectthe degradability of the nanocomposites performed in aninert environment, although it has a favorable effect underoxidant conditions since the beginning of degradation ispostponed and the temperature of maximum degradationis slightly decreased. Accordingly, new functional poly-meric nanocomposites with tailored and precise propertiescan be prepared without increasing their production costs.

Acknowledgements

The authors are grateful for the financial support ofMinisterio de Ciencia e Innovación (project MAT2010-

19883). J. Ressia and E. Vallés acknowledge the financialsupport from CONICET, FONCyT and UNS. Ms. C. Serranois also grateful to Ministerio de Ciencia e Innovación forher FPU predoctoral grant. The synchrotron work leadingto these results has received funding from the EuropeanCommunity’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 226716. We thank thecollaboration of the Hasylab personnel in the soft-con-densed matter beamline A2, especially Dr. S. S. Funari.

References

[1] Matsunaga T, Tomoda R, Nakajima T, Wake H. FEMS Microbiol Lett1985;29:211–4.

[2] Joo J, Kwon SG, Yu T, Cho M, Lee J, Yoon J, et al. J Phys Chem B2005;109:15297–302.

[3] Kubacka A, Cerrada ML, Serrano C, Fernandez-Garcia M, Ferrer M,Fernandez-Garcia M. J Nanosci Nanotechnol 2008;8:3241–6.

[4] Cerrada ML, Serrano C, Sánchez-Chaves M, Fernández-García M, deAndrés A, Riobóo RJ, et al. Environ Sci Technol 2009;43:1630–4.

[5] Kubacka A, Ferrer M, Cerrada ML, Serrano C, Sánchez-Chaves M,Fernández-García M, et al. Appl Catal B 2009;89:441–7.

[6] Kubacka A, Serrano C, Ferrer M, Lunsdorf H, Bielecki P, Cerrada ML,et al. Nano Lett 2007;7:2529–34.

[7] Cerrada ML, Serrano C, Sánchez-Chaves M, Fernández-García M,Fernández-Martín F, de Andrés A, et al. Adv Funct Mater2008;18:1949–60.

[8] Kubacka A, Cerrada ML, Serrano C, Fernández-García M, Ferrer M,Fernández-Garcia M. J Phys Chem C 2009;113:9182–90.

[9] Jiménez Riobóo RJ, Serrano-Selva C, Fernández-García M, CerradaML, Kubacka A, de Andrés A. J Lumin 2008;128:851–4.

[10] Thompson RB, Ginzburg VV, Matsen MW, Balazs AC. Science2001;292:2469–72.

[11] Mackay ME, Tuteja A, Duxbury PM, Hawker CJ, Van Horn B, Guan Z,et al. Science 2006;311:1740–3.

[12] Warren SC, DiSalvo FJ, Wiesner U. Nat Mater 2007;6:156–61.[13] Mackay ME, Dao TT, Tuteja A, Ho DL, Van Horn B, Kim HC, et al. Nat

Mater 2003;2:762–6.[14] Bansal A, Yang H, Li C, Cho K, Benicewicz BC, Kumar SK, et al. Nat

Mater 2005;4:693–8.[15] Narayanan RA, Thiyagarajan P, Lewis S, Bansal A, Schadler LS, Lurio

LB. Phys Rev Lett 2006;97:075505/1–5/4.[16] Ferry JD. Viscoelastic properties of polymers. New York: John Wiley

and Sons; 1980.[17] Mansel S, Pérez E, Benavente R, Pereña JM, Bello A, Röll W, et al.

Macromol Chem Phys 1999;200:1292–7.[18] Wunderlich B. Macromolecular Physics, vol. 3. New York: Academic

Press; 1980.[19] Smith GD, Bedrov D, Li L, Byutner O. J Chem Phys

2002;117:9478–89.[20] Bahloul W, Bounor-Legaré V, David L, Cassagnau P. J Polym Sci Part

B: Polym Phys 2010;48:1213–22.[21] Bartholome C, Beyou E, Bourgeat-Lami E, Cassagnau P, Chaumont P,

David L, et al. Polymer 2005;46:9965–73.[22] Natta G, Corradini P. Il Nuovo Cimento 1960;15:40–51.[23] Krache R, Benavente R, López-Majada JM, Pereña JM, Cerrada ML,

Pérez E. Macromolecules 2007;40:6871–8.[24] Turner-Jones A. Polymer 1971;12:487–508.[25] Arranz-Andrés J, Benavente R, Pérez E, Cerrada ML. Polym J

2003;35:766–77.[26] Cerrada ML, Pérez E, Benavente R, Ressia J, Sarmoria C, Vallés EM.

Degrad Stab 2010;95:462–9.

596 C. Serrano et al. / European Polymer Journal 48 (2012) 586–596

[27] Alamo RG, Kim M-H, Galante MJ, Isasi JR, Mandelkern L.Macromolecules 1999;32:4050–64.

[28] Palza H, López-Majada JM, Quijada R, Benavente R, Pérez E, CerradaML. Macromol Chem Phys 2005;206:1221–30.

[29] López-Majada JM, Palza H, Guevara JL, Quijada R, Martínez MC,Benavente R, et al. J Polym Sci. Part B: Polym Phys 2006;44:1253–67.

[30] Arranz-Andrés J, Peña B, Benavente R, Pérez E, Cerrada ML. EurPolym J 2007;43:2357–70.

[31] Baltá-Calleja FJ, Vonk CG. X-Ray Scattering of SyntheticPolymers. Amsterdam: Elservier; 1989.

[32] Ryan AJ, Stanford JL, Bras W, Nye TMW. Polymer 1997;38:759–68.[33] Crist B. J Polym Sci Part B: Polym Phys 2001;39:2454–60.[34] Jourdan C, Cavaille JY, Perez J. J Polym Sci Part B: Polym Phys

1989;27:2361–84.[35] Rotter G, Ishida H. Macromolecules 1992;25:2170–6.