lable at ScienceDirect

Polymer 55 (2014) 5584e5595

Contents lists avai

Polymer

journal homepage: www.elsevier .com/locate/polymer

Photo-vulcanization using thiol-ene chemistry: Film formation,morphology and network characteristics of UV crosslinked rubberlatices

Sandra Schl€ogl a, *, Marie-Luise Trutschel b, Walter Chass�e b, Ilse Letofsky-Papst c,Raimund Schaller d, Armin Holzner d, Gisbert Riess e, Wolfgang Kern e, Kay Saalw€achter b

a Polymer Competence Center Leoben GmbH, Roseggerstraße 12, A-8700 Leoben, Austriab Institut für Physik e NMR, Martin-Luther-Universit€at Halle-Wittenberg, Betty-Heimann-Straße 7, D-06120 Halle, Germanyc Institute for Electron Microscopy, Graz University of Technology, Steyrergasse 17, A-8010 Graz, Austriad Semperit Technische Produkte GmbH, Triester Bundesstraße 26, A-2632 Wimpassing, Austriae Chair of Chemistry of Polymeric Materials, University of Leoben, Otto Gl€ockel-Straße 2, A-8700 Leoben, Austria

a r t i c l e i n f o

Article history:Received 8 April 2014Received in revised form1 June 2014Accepted 3 June 2014Available online 10 June 2014

Keywords:Diene-rubber latexFilm formationThiol-ene chemistry

* Corresponding author. Tel.: þ43 3842 402 2354;E-mail address: sandra.schloegl@pccl.at (S. Schl€og

http://dx.doi.org/10.1016/j.polymer.2014.06.0070032-3861/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

The photo-vulcanization with versatile thiol-ene chemistry represents an innovative approach tocrosslink diene-rubber materials both in latex and in solid film state. In this work, the structure ofelastomer-based thiol-ene networks and the morphology after film formation are studied in detail usingelectron microscopic techniques, atomic force microscopy and multiple-quantum solid-state NMRspectroscopy. Additionally, film formation properties and corresponding macroscopic properties ofphoto-vulcanized natural rubber (NR) latex and its synthetic counterpart, isoprene rubber (IR) latex, aredetermined in dependence on the curing procedure (pre- and post-vulcanization). The results reveal thatthiol-ene cured elastomers comprise homogenously distributed crosslinks with a low amount of shortchain defects. Whilst photochemically pre-cured NR latex particles provide coherent films, the filmformation and mechanical properties of IR are strongly governed by the crosslink density of the latexparticles. In film state, photo-vulcanization promotes narrow crosslink distributions and excellent tensileproperties of both NR and IR.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Although earliest reports on thiol-ene chemistry date back to1905, the publications of Morgan and Ketley on photo-polymerization of thiol-ene monomers have become the startingpoint for extensive studies and widespread applications of thiol-enes in the 1970s [1,2]. The photoinduced addition of thiolsacross unsaturated C]C double bonds follows a free radicalmechanism where in the first step thiol radicals are formed by ahydrogen abstraction from photoinitiator radicals. The subsequentaddition of the thiol radical to the alkene double bond creates acarbon centered radical. This radical can abstract a hydrogen fromanother thiol group leading to a perpetuation of the chain processin photopolymerization reactions [3,4]. Besides photo-polymerization, UV induced thiol-ene chemistry is widely used in

fax: þ43 3842 402 2352.l).

surface functionalization techniques, fabrication of polymerbrushes, photo-lithography and modification of polymer materials[5e7].

Apart from the photoinitiated free-radical addition of thiols toelectron-rich and electron-deficient alkenes, radical-mediatedthiol-ene reactions can also be initiated by thermal initiators suchas peroxides or 2,20-azobis(isobutyronitrile). Thermal initiators areoften used for large-scale synthesis as an uniform generation ofradicals in large volumes can be easily achieved at elevated tem-perature [4]. The employment of thermally sensitive radical initi-ators has also recently gained increased attention in themodification of biological and polymer-based materials [8,9]. Inparticular, Schlaad and co-workers extensively studied the func-tionalization of 1,2-poly(butadiene) and related copolymers viaboth thermally and photoinitiated thiol-ene chemistry. They havedemonstrated that the selected attachment of functional thiolsmakes the preparation of stimuli-responsive copolymers withtailored physical and chemical properties feasible [10e12].

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5585

Account has to be taken into the fact, that not only radical-mediated thiol-ene reactions but also catalyzed Michael additionreactions carry salient features of click reactions which include highefficiency, rapid reaction rates, mild process conditions and highyields. The thiol-Michael addition is a catalyzed reaction betweenthiols and alkenes comprising electronwithdrawing groups amongthem (meth)acrylates, acrylonitrile, maleimides or a,b-unsaturatedketones. Numerous catalysts ranging from strong bases and Lewisacids to nucleophilic amines and alkyl phosphines can be used toinitiate the thiol-Michael reaction [4,13]. Due to its short reactionstimes under ambient conditions and its insensitivity to the pres-ence of water and oxygen, the thiol-Michael reaction is particularlyapplied in the preparation of hydrogels and biomaterials [14,15].

In our work we have recently highlighted that versatile andunique UV induced thiol-ene chemistry can be employed to photo-vulcanize diene-rubber materials in liquid and solid state (seeScheme 1). Although side chain enes on the investigated diene-rubbers provide a higher reactivity than the C]C groups of themain chain, cis-1,4-poly(isoprene) rubber materials can be cross-linked successfully with crosslink yields comparable to sulfur curedmaterials [16]. Whilst the photo-vulcanization in solid state can beaccomplished easily by a UV irradiation of the rubber films, thephoto-vulcanization in latex state requires a homogenous illumi-nation of thin liquid films (layer thickness below 1 mm) since latexemulsions exhibit a low light transmissivity. In a new approach wehave developed a continuous pre-vulcanization process using afalling film photoreactor to ensure a homogenous UV illuminationof liquid latex formulations in thin films [17,18].

It should be noted that thermally initiated thiol-ene reactionsallow a more flexible reactor design since they are not limited bylight attenuation. However, our study focuses on UV induced thiol-ene chemistry as photochemical reactions provide desirable fea-tures when it comes to the large-scale vulcanization of diene-rubber latices. The energy consumption of the pre-vulcanization

Scheme 1. Photo-vulcanization of diene-rubbers in l

process can be reduced significantly because the UV inducedcrosslinking proceeds at room temperature. In addition, the designof the falling film photoreactor establishes a new way for contin-uous pre-vulcanization processes whilst thermally induced pre-vulcanization techniques are typically carried out in discontin-uous processes [17,19].

The pre-vulcanization of rubber latices plays an important partin the enhancement of tensile properties and aging stabilities ofthin film dipped latex goods [19]. On industrial scale processes thepre-vulcanization is usually carried out in the presence of sulfur orsulfur-donor agents. In addition, accelerators and activators arerequired that reduce the reaction time and reaction temperature[20]. Other thermal crosslinking processes involve the curing withperoxides or hydroperoxides whilst the curing with high energyradiation such as e-beam or gamma rays has become a well-established technique to produce pre-cured NR latex in pilotplant quantities [21e23].

Besides the mechanical properties, the film formation proper-ties are directly affected by the employed pre-curing system. Recentwork has revealed that sulfur pre-vulcanization as well as gamma-ray induced pre-curing of natural rubber (NR) latex leads to uni-formly crosslinked rubber particles [24e26]. With respect toperoxide curing, only small NR particles exhibit a homogenouscrosslink structure whereas a coreeshell structure has beenobserved for larger particles [26,27]. The film formation of bothnatural and synthetic latices has been studied in detail and researchhas demonstrated that the film formation of polymer emulsionstakes place in three major stages [28e30]. The first step involvesthe drying of the latex due to the evaporation of water from thelatex surface. The evaporation rate is affected by various parame-ters such as the particle size, the chemical composition and stabi-lization of the latex or the glass transition temperature [31]. Aconcentration of the rubber particles takes place leading to thesecond step in which the latex particles come in contact and stick

iquid and solid state using thiol-ene chemistry.

Fig. 1. Schematic representation of a falling film photo-reactor used in the photo-vulcanization of diene-rubber materials in latex state (pre-vulcanization).

S. Schl€ogl et al. / Polymer 55 (2014) 5584e55955586

together irreversibly. In dependence on the hardness of the parti-cles, deformation and coalescence are observed. Various theoriesinvolvingwet-sintering and dry-sintering processes or interparticlecohesion promoted by surface forces have been proposed to explainthe formation of ordered particle structures during this step[32e34]. In the last stage, the cores of the latex particles inter-penetrate and the soft latex film obtains its mechanical properties(wet gel strength). In terms of highly pre-vulcanized latex particlesor particles with a highly crosslinked shell structure, the polymerdiffusion across particle boundaries is reduced or hindered due tothe restricted mobility of the macromolecules chains [35]. As themacroscopic properties of the dried latex film are directly related toa coherent and efficient film formation it has to be considered that abalance has to be found between pre-vulcanization, film formationand mechanical properties of the final latex articles [29].

The present study focuses on the crosslink structure ofelastomer-based thiol-ene networks and its influence on film for-mation properties, film morphologies and corresponding tensileproperties of diene-rubber materials. In particular, this work ad-dresses the network characteristics of thiol-ene natural rubber (NR)latex and its synthetic counterpart isoprene rubber (IR) latex. Thephoto-vulcanization is carried out in liquid state (pre-vulcanizationof the emulsion) as well as solid state (post-vulcanization of solidfilms) and the effect of both procedures on the microscopic andmacroscopic properties is determined in detail. Real time FT-IRspectroscopy is employed to investigate the reaction kineticswhilst the morphology of the photo-vulcanized elastomer films ischaracterized by scanning electronmicroscopy (SEM), transmissionelectron microscopy (TEM) and atomic force microscopy (AFM).Both multiple-quantum (MQ) solid-state NMR spectroscopy andequilibrium swelling measurements are used to obtain quantitativeinformation on the network structure after film formation, asstudied previously with a focus on entanglement effects [36,37]. Inaddition the corresponding mechanical properties, in particular thetensile strength, are measured by tensile tests. The correlation be-tween crosslink structure, film morphology and macroscopicproperties gives new insights into the characteristics of elastomer-based thiol-ene networks. Furthermore, crucial process parametersin photochemical pre- and postvulcanization procedures arehighlighted that considerably influence film formation and corre-sponding tensile properties.

2. Experimental part

2.1. Materials and chemicals

Natural concentrated rubber latex (high ammonia, 60 wt.-% dryrubber content)was purchased fromaMalaysian supplier. Syntheticisoprene rubber latex (Kraton® IR0401 BU) with a dry rubber con-tent of 63 wt.-% and a molecular weight ranging from 1,500,000 to2,500,000 g/mol was obtained from Kraton Polymers (Amsterdam,The Netherlands). Ethyl 2,4,6-trimethylbenzoylphenylphosphinate(Lucirin TPO L) was provided by BASF (Ludwigshafen, Germany)and trimethylolpropane tris(3-mercaptopropionate) (TriThiol) wassupplied by Bruno Bock Thiochemicals (Marschacht, Germany). Allother reagents were from SigmaeAldrich (St. Louis, United States)and were used without further purification.

2.2. Photo-vulcanization of NR and IR in latex state

In the first step of the UV pre-curing, 1.0 phr (parts per hundredparts of rubber) Lucirin TPO L and 1.0 phr TriThiol were emulsifiedin 2.0 phr deionized water. The emulsion of the photoinitiator andthe thiol was added to 10.0 kg IR and NR latex, respectively, with adry rubber content of 40 wt.-%. This latex mixture was stirred at

room temperature bymeans of a magnetic agitator for 2 h. The pre-vulcanization of NR and IR latex mixtures was carried out in afalling film photoreactor using the same process parameters.

The scheme of the falling film reactor which was tailor made isdepicted in Fig. 1. The reactor tube with an inner diameter of135 mm was made of glass and contained an inlet for the latexmixture at the top and an outlet at the bottom. In the reactor tube amedium pressure Hg lamp (Heraeus) equipped with a quartzcooling tube was positioned centrically. The UV lamp comprised anarc length of 25 cm and was flooded with inert gas (nitrogen) toextend the longevity of the UV system. The latex mixtures werepumped continuously from a storage vessel to the top of the fallingfilm reactor using an eccentric screw pumpwith a flow rate of 1.3 L/min. With these process parameters continuous and stable liquidfalling films could be achieved with a film thickness in the range of0.5 and 1.0 mm. The UV assisted pre-vulcanization was performedwith a lamp power of 800 and 3000 W corresponding to a lightintensity of 0.4 and 0.9W/cm2 (wavelength range between 240 and460 nm), respectively. After each illumination cycle samples weretaken and 0.5 phr Ionol LC as antioxidant was added to the pre-cured latices.

2.3. Preparation of solid latex films

A conventional coagulant dipping process was employed toprepare solid films from either non crosslinked or pre-cured diene-rubber latices. In the first step porcelain formers were cleaned withacid and alkaline solutions and placed in a coagulant solutioncontaining calcium salts (coagulant), calcium carbonate (releaseagent) and surfactants. The formers were dried for 60 s at 120 �C,immersed in the various latex formulations for 30 s and dried againat 120 �C for 20 min. The thickness of the dried latex films wasbetween 200 and 300 mm.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5587

2.4. Photo-vulcanization of NR and IR in film state

An emulsion with 1.0 phr Lucirin TPO L and 1.0 phr TriThiol in2.0 phr deionized water was added to NR and IR latex with a dryrubber content of 40 wt.-%. After adding 0.5 phr Ionol LC, the latexmixture was stirred at room temperature for 2 h and then solidfilms were prepared by the dipping process. These films were UVilluminated with a medium pressure Hg lamp (Hereaus) employingan exposure dose between 1.0 and 7.5 J/cm2 (wavelength rangefrom 240 to 460 nm). The UV irradiationwas carried out under inertconditions to avoid any photo-oxidation effects on the elastomersurface.

2.5. Characterization of the crosslink kinetics

Real time FT-IR measurements were carried out with a largeangle reflectance infrared (IR) cell (Perkin Elmer). The rubbersamples were irradiated with a UV spot curing unit (Efos Novacurehigh pressure Hg emitter with a wave length range between 290and 410 nm). IR spectra were taken with an FT-IR spectrometer(Spectrum One, Perkin Elmer) and the absorption peak areas werecalculated with Spectrum 3.02 and Spectrum TimeBase 1.1 soft-ware. Sample preparation: 2 wt.-% rubber was dissolved in tolueneand Lucirin TPO L and TriThiol were added to the solution. Theconcentration of the photoinitiator and thiol is stated in the dis-cussion section. The solution was stirred for 15 min at room tem-perature by means of a magnetic agitator and then spin cast(4000 rpm, 20 s) on gold substrates. Each sample was fixed on aspecimen holder in the large angle reflectance IR cell and the cellwas flooded with inert gas for 10 min to avoid photo-oxidationduring UV irradiation. The crosslinking via thiol-ene reaction wasmonitored during UV irradiation (1.1 W/cm2 light intensity) bytaking FT-IR spectra after various irradiation times. The light in-tensity into the sample plane was determined with a spectroradi-ometer (Solatell, Solascope 2000™).

2.6. Characterization of the film morphology

Atomic force microscopy (AFM): The elastomer films wereplaced between two poly(methyl methacrylate) plates and speci-mens were prepared by cutting ultra-thin sections cryogenically(�120 �C) with a Leica FCS microtome fitted with a histo cryodiamond knife (Diatome). AFM operating in tapping mode wascarried out in air under ambient conditions with a Dimension 3100connected to a NanoScope IV controller (Veeco Instruments, SantaBarbara, USA). The silicon cantilever comprised a nominal springconstant of 40 N/m and a resonance frequency of 360 kHz.

Scanning electron microscopy (SEM): SEM images were takenwith a JEOL JSM-5410 scanningmicroscope using aworking voltageof 5 keV.

Transmission electron microscopy (TEM-EDX): The latex filmswere placed between two poly(methyl methacrylate) plates andspecimens were prepared by cutting ultra-thin sections cryogeni-cally (�120 �C) with a Leica FCS microtome fitted with a histo cryodiamond knife (Diatome). The sections were collected on nickelgrids and stained with osmium tetra oxide vapor for 2 h prior toexamination. The cross-sections of the films were examined with aPhilips CM 20 transmission electron microscope operating at10 keV. EDXS spectra were taken with a Noran high purityGermanium EDX detector (TN-5500 Series II).

2.7. Characterization of the rubber network

An in-depth assessment of the rubber network microstructurewas the topic of two previous publications [36,37], which we refer

to for details. We have studied the given sample series with a focuson entanglement effects, andwe here summarize themost relevantbasic principles, as we discuss below some of the previously pub-lished results under a different viewpoint.

Equilibrium swelling measurements: To determine the crosslinkdensity of the photo-vulcanized NR and IR latex films equilibriummeasurements were carried out in accordance with the theory ofFlory and Rehner [38,39]. From each solid film, 10 samples with adimension of 2 � 2 cm were cut and weighed on an analyticalbalance with an accuracy of 0.1 mg. The samples were placed invials and equilibrated with toluene (20 mL) for 48 h at 21 �C. Theswollen films were removed from the solvent, blotted onto tissuepaper and wereweighed immediately. The films were then dried toconstant weight under vacuum at 25 �C and re-weighed. Theeffective crosslink density (1/Mc) in mol/cm3 was calculated anddetermined in agreement with recently published work and wasaveraged over 10 specimens [40].

1H multiple-quantum (MQ) solid-state NMR spectroscopy:Proton NMR was used to determine the degree of crosslinking on alocal (molecular) level based upon the quantitative measurementof the segmental residual (time-averaged) dipoleedipole couplingconstant, dres ¼ Dres/2p in units of Hz [36,41,42]. As compared toisotropic liquids, this quantity is non-zero due to the restrictions onthe network chain ends posed by crosslinks and also entangle-ments. It is related to the effective crosslink density via 1/Mc,NMR¼ dres/dref*f/(f� 2), where f is the crosslink functionality. Thereference coupling dref is model dependent and has been deter-mined to 617 Hz kg/mol and 811 Hz kg/mol for NR [42] and IR [37],respectively. Themethod not only yields an averageDres, but also itsdistribution function in case the network is inhomogeneouslycrosslinked. The Dres distribution was measured and analyzed ac-cording to previously published procedures [36,41,42], and char-acterized by its average (Dav) and width in terms of the standarddeviationys. Finally, the method also provides the fraction of de-fects, i.e., isotropically mobile chains that do not carry load, such asloops and dangling chains.

Additional NMR experiments have been carried out on samplesswollen in deuterated toluene to different degrees up to swellingequilibrium [36]. First, this leads to a higher and more realisticmeasure of the defect content, as long dangling chain may relaxslowly on contribute to transient elasticity in the bulk, but are fullyrelaxed in the swollen state. Second, a back-extrapolation of Davfrom differently swollen samples to its apparent value in the bulkprovides a measure of the true (chemical) crosslink density of thenetwork, referred to as phantom reference state (PRN), see Refs.[36,37]. The difference between this value and the Dav measured inbulk provides a measure of the entanglement contribution to theelasticity.

The measurements were performed with a Bruker minispecmq20 spectrometer operating at 0.5 T with 90� pulses of 1.7 mslength and a dead time of 12 ms. The experiments were carried outat 60 �C.

2.8. Characterization of the tensile properties

Tensile testing was performed with a ZWICK Z005 tensile testerin compliance with ASTM Standard D412-98a.

3. Results and discussion

The photo-vulcanization of diene-rubbers by thiol-ene additionreaction is characterized by short reaction times, low energy con-sumption and allows themanufacture of latex goods with good skincompatibility since any employment of skin hazardous processchemicals can be avoided [43]. These tremendous advantages

S. Schl€ogl et al. / Polymer 55 (2014) 5584e55955588

together with additional considerations such as ease of operationand operational safety concerns promoted the industrial imple-mentation of the UV assisted pre-vulcanization for the commercialproduction of surgical gloves [44]. As the reliability of dipped latexarticles is not only governed by the type of latex but also by theemployed curing procedure it is crucial to assess the influence ofthe photo-vulcanization on network characteristics, film formationproperties and corresponding tensile properties.

3.1. Determination of the crosslink kinetics

To evaluate the reactivity of the diene-rubber materialsemployed in the photo-vulcanization, the UV initiated addition ofthe thiol crosslinker to the C]C double bonds was monitored byreal time FT-IR spectroscopy. Fig. 2 depicts the consumption of theC]CH wagging band at 836 cm�1 versus the UV irradiation time.Upon 15 min of UV exposure and with 1.0 phr photoinitiator andthiol, respectively, IR films exhibit a significant depletion of therelative peak area up to 11% whilst the decrease of the C]C doublebonds in NR films amounts to 8%. At higher concentrations of theprocess chemicals (5 wt.-%) a pronounced C]C consumption in therange of 28% is observed in IR films. In contrast, the decrease of theC]C peak area is slightly lower in NR films (24%). Account has to betaken into the fact that the infrared measurements of solid filmsmay not fully reflect the reaction conditions in the liquid latexwhere other process parameters such as the polarity of the processchemicals or the diffusion of the process chemicals in the latexparticles play a significant role [18]. However, the results give evi-dence that in IR films the depletion of C]C double bonds proceedsat a higher rate compared to NR. Since the UV assisted pre-vulcanization via the thiol-ene reaction is a free radical inducedprocess the lower number of crosslinks observed in NR may beattributed to the presence of radical scavenging proteins. NR isderived from the Hevea brasiliensis tree and consists of a consid-erable amount of organic components including proteins (2e3 wt%), fatty acids (1e1.5 wt%) and lipids (1e1.5 wt.-%) [45e47]. Previ-ous work has shown that not only lignin but also proteins can act asnatural antioxidants [48,49]. Regarding NR latex, it was repeatedlydemonstrated that the proteins present in the latex sap terminatefree radical reactions such as the grafting of methacrylic monomersamong them methyl methacrylate or 1,9-nonandioldimethacrylate[50e52]. In terms of thiol-ene crosslinking, these data clearly

Fig. 2. Monitoring the thiol-ene reaction by means of real time FT-IR spectroscopy:Depletion of the C]CH wagging band at 836 cm�1 in NR and IR films upon UVillumination.

support the finding of the decreased reactivity of NR filmscompared to IR.

3.2. Determination of the network structure

Both crosslink density and spatial inhomogeneities of thiol-enecrosslinked diene-rubber films were determined by 1H multiple-quantum (MQ) solid-state NMR spectroscopy. It should be notedthat a detailed NMR study of thiol-ene crosslinked NR and IRsamples has already been published in two papers aimed at afundamental microscopic understanding of the swelling processand the influence of entanglements [36,37]; selected data are herediscussed in some more detail with regards to the materials'characteristics. The normalized distribution widths of the residualdipolaredipolar couplings which reflect the homogeneity of thecrosslinked rubber matrix are depicted in Fig. 3a. The results clearlyshow that the crosslink distribution is strongly governed by thevulcanization procedure. In film state photo-vulcanization pro-motes highly homogenous elastomer-networks which are compa-rable to sulfur cured systems (s/Dav < 0.15) whilst somewhatbroader spatial crosslink distributions (0.15 < s/Dav < 0.3) are ob-tained by the UV assisted pre-curing in latex state [42]. In contrastto thermally peroxide-cured NR samples (s/Dav > 0.3) photo-vulcanized networks do not exhibit higher crosslink densities assecond component that can be attributed to radical chain scissionreactions [42]. Due to the absence of any bimodal distribution of thecrosslinks, the experimental data give an indication that UV pre-cured latex particles exhibit no coreeshell structure with a highlycrosslinked shell and a slightly crosslinked core [26].

In order to address topologically trapped entanglements, thesamples were not only measured in dried state but also at differentdegrees of swelling using deuterated toluene as solvent. In Fig. 3bthe apparent defect content in dried and swollen photo-vulcanizedNR and IR samples is compared. A low defect fraction (~5%), com-parable to hydroperoxide-cured NR rubbers can be observed indried state which is less dependent on either vulcanization pro-cedure or type of latex [26,42]. The bulk measurements in driedstate reflect quickly relaxing low molecular weight defectsincluding short loops and dangling ends. With respect to peroxide-cured NR materials, the defect fraction is considerably higher(>20%) related to chain scission reactions at higher temperatures[42].

However, when removing the entanglement constraints duringswelling in toluene a completely different performance can beobserved. Whilst low crosslinked samples comprise up to 90% non-elastic chains, a distinctive decrease below 30% is obtained withincreasing crosslink density. This behavior gives an indication thatthe defect fraction is governed by long precursor chains which arehighly entangled in the elastomer bulk and may comprise longrelaxation times. Whilst the defect content in peroxide-cured NRsamples does not increase upon swelling due to the contributionsof short chain fragments, this trend is also observed in sulfur curedNR samples or poly(dimethylsiloxane) based elastomers [40,42,53].

The comparison of the NMR results to equilibrium swellingmeasurements is provided in Fig. 4. A very good linear correlationof the results from the two methods is observed, with deviationsfrom a unity slope (dashed line) that are well within the un-certainties related to the different model assumptions. Theentanglement-free phantom reference network (PRN) result ob-tained by back extrapolation of measurements on swollen samplescorrelates with the swelling result with zero offset, providing aconvincing proof of principle, and allows for an estimation of thenear-constant entanglement contribution in the bulk. It is inter-esting to note that NR samples exhibit higher crosslink densitiesthan IR films at a given illumination dose which does not reflect the

Fig. 3. (a) Spatial crosslink distributions and (b) apparent, non-elastic defect fractions of thiol-ene crosslinked NR and IR after film formation. A part of these data is replotted fromRef. [37].

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5589

results of the infrared spectroscopy where a higher reactivity of IRsamples was observed. The higher crosslink density of NR latexfilms might be a result of the considerable gel content of highammonia NR latex. Recent work has shown that during transportand storage of NR latex, a loose network is formed due to crosslinkreactions induced by non-rubber contents, in particular proteinsand phospholipids, that increases the gel content and the apparentcrosslink density. In contrast, IR as synthetic counterpart does notcontain any proteins or fatty acids and no significant gel content isobserved [19].

3.3. Characterization of the film morphology

To address the influence of the photo-vulcanization on both filmformation properties and morphology of solid latex films SEM, TEMand AFM measurements were conducted. In Fig. 5aed the cross-section of NR as well as IR latex films pre-cured with an irradia-tion intensity of 0.9 W/cm2 upon two illumination passes arecompared. Although both samples comprise comparable crosslinkdensities of 0.085 (NR) and 0.076 (IR) mmol/cm3, respectively, astriking difference between the corresponding filmmorphologies isachieved. For the pre-vulcanized NR sample, a smooth andcontinuous cross-section without defects and particle contours isobserved. The coherent film morphology suggests that an efficientinterparticle diffusion has occurred during the coalescence step of

Fig. 4. Network chain densities obtained from NMR experiments versus the results ofequilibrium swelling measurements. Data replotted from Ref. [37].

the film formation. In contrast, SEM micrographs of photochemi-cally pre-cured IR samples exhibit single and partly coalesced latexparticles being very polydisperse in size (~1e10 mm) and predom-inately spherical in shape. The inhomogeneous film morphology ofIR latex films reflected by the separated particles gives a goodindication that the polymer diffusion across the particle interfaceshas not occurred sufficiently during the last step of the film for-mation process.

Both IR and NR latex comprise negatively charged particles thatare held apart by electrostatic forces. Whilst the IR latex particlesare stabilized by adsorbed surfactants the NR latex particles aresurrounded by negatively charged protein-phospholipide layersthat prevent particle coalescence. In addition, ammonia is usuallyadded to commercial NR latex to further improve the colloidalstability. During the dipping process the electrostatic forces areovercome upon neutralization with calcium cations initiating thefilm formation. A subsequent drying step at elevated temperatureremoves the residual water and leads to further particle coales-cence and to polymer chain diffusion across the particle boundaries[54]. As evidenced by NMR experiments, UV pre-cured IR and NRsamples exhibit homogenous crosslink distributions after filmformation which gives an indication that both pre-cured latices donot comprise a coreeshell structure. As both network structure andglass transition temperature are comparable, the difference be-tween the film formation properties of NR and IR materials mightbe mainly attributed to the protein and phospholipid layers of NRlatex that promote particle coalescence during film formation [19].Literature data would suggest that the film formation properties ofUV pre-cured IR latex are strongly affected by the overall crosslinkyields which increase the internal viscosity and the hardness of theparticles hindering their coalescence. Zosel and Lay have shown intheir work that the film formation properties of synthetic laticessuch as poly(butyl methacrylate) (PBMA) are substantially influ-enced by their crosslink densities. They observed that highlycrosslinked PBMA particles do not undergo particle coalescencegiving rise to brittle latex films with poor mechanical propertieswhereas non crosslinked PBMA particles formed homogeneous andcoherent films [35].

To confirm the effect of the crosslink density on the film for-mation properties of thiol-ene pre-vulcanized IR latex, the pre-vulcanization was carried out with a lower light intensity of0.4 W/cm2 upon two illumination passes corresponding to acrosslink density of 0.046 mmol/cm3 Fig. 6a and b depict thecharacteristic morphology of the IR film formed from lightlycrosslinked latex particles. The SEMmicrographs clearly reveal thatslightly crosslinked IR latex particles form latex films with smoothand almost structureless morphologies. Singular particles with

Fig. 5. SEM micrographs of dipped (a,b) NR and (c,d) IR latex films photochemically pre-cured with an irradiation intensity of 0.9 W/cm2.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e55955590

distinctive contours are notably absent indicating an improvementin the particle coalescence.

The morphology of UV pre-vulcanized IR latex films in depen-dence on the crosslink density was further characterized by TEMand AFM measurements. Fig. 7a and b provide bright field TEMmicrographs of IR latex films pre-cured with 0.4 and 0.9 W/cm2,respectively. In order to increase the material contrast, the speci-mens were stained with osmium tetroxide that reacts with the C]C double bonds of the poly(isoprene). With respect to highlycrosslinked films (see Fig. 7b), the TEM images reveal sphericallyshaped latex particles with sharp contours. The TEM micrographssupport the results of the SEM measurements and confirm that fullcoalescence of the latex particles was retarded resulting in anincoherent film formation. When it comes to slightly crosslinked IRfilms (see Fig. 7a), the contours of the individual particles havedisappeared. In addition, bright domains are observed at the borderof the latex shell. EDX spectra were taken from the bright domainsas well as the latex particles illustrated in Fig. 8a and b. Besides the

Fig. 6. SEM micrographs of dipped IR latex films photochemic

characteristic Os signals (1.9, 8.9 and 13 keV) attributed to theprevious staining and the Ni signals (0.9, 7.5 and 8.2 keV) from theTEM grid, the EDX spectrum of the latex particles reveal the C Kapeak at 0.2 keV and the O Ka peak at 0.5 keV. In the rubber matrix,the photoinitiator as well as the cross-link agent levels are belowthe detection limit. Hence, sulfur or phosphor signals are notdetected with EDX. Regarding the bright domains observed in theTEM images, Ca and Cl signals at 2.8 and 3.7 keV are observed. Theresults indicate that the process chemicals of the coagulant bathincluding calcium carbonate and calcium salts are incorporated inthe bulk material during the film formation forced by the coagu-lation dipping process.

From the cross-sections of the IR films pre-cured with 0.4 and0.9 W/cm2, also AFM phase-contrast images were taken to char-acterize the viscoelastic properties. In these measurements ob-tained in tappingmode the dark contrast regions correspond to softdomains, whereas regions with light contrast correspond to harddomains. As illustrated in Fig. 9a and b, it is interesting to note that

ally pre-cured with an irradiation intensity of 0.4 W/cm2.

Fig. 7. TEM micrographs (bright field) of dipped IR films photochemically pre-cured with an irradiation intensity of (a) 0.4 and (b) 0.9 W/cm2.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5591

the boundaries of slightly crosslinked IR particles are partly dis-rupted suggesting that the particles have been fused together. Inaddition, similar to the TEM micrographs light domains with asharp shape are observed at the boundaries of the partly fused latexparticles. EDX spectra of these domains show a relatively highamount of Ca whereas no Cl signal can be detected. The results givea good indication that the light domains represent calcium car-bonate particles from the coagulant bath.

Fig. 8. EDX spectra of dipped IR films photochemically pre-cured

With respect to highly crosslinked IR films (see Fig. 9c and d) theAFM images display sharp polyhedral contours and a regularpacking. The majority of the observed particle boundaries are notdisrupted confirming that the particle coalescence was hinderedduring the film formation process. The darker layer surrounding theparticles may be attributed to surfactants such as sodium dodecylsulfate that are widely used in the stabilization of the raw material.A similar morphology is observed in phase contrast images of NR

with an irradiation intensity of (a) 0.4 and (b) 0.9 W/cm2.

Fig. 9. Phase contrast images (tapping mode) of dipped IR films photochemically pre-cured with an irradiation intensity of (a,b) 0.4 and (c,d) 0.9 W/cm2. Zone A and B highlight thesample areas which are magnified out of in Fig. 9b and d, respectively.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e55955592

latex particles where the double layers of proteins and phospho-lipids are depicted as dark areas at the surface of the particles [46].From the AFM measurements it can be further obtained, that thepre-cured rubber particles are inhomogeneously covered with darkstreaked layers. In Fig. 10a and b the cross-section of highly cross-linked IR latex films are provided that has been imaged under“hard-tapping” and “soft-tapping” conditions. Imaging underdifferent tapping forces leads to different extents of deformation insofter and more rigid regions that makes the characterization ofliquids on both organic and inorganic surfaces feasible and repre-sents an applicable tool to characterize material surfaces withoutadsorbed soft molecular layers [55]. It can be clearly shown that thedark structures observed under a small force vanish when imagingis performed with higher forces suggesting that the dark areascomprise liquid layers related to surfactants used in the stabiliza-tion of IR latex during storage.

Besides the illumination intensity or the concentration of thecrosslink chemicals, the crosslink density of UV pre-cured latexparticles is also governed by the distribution and diffusion of thephotoinitiator and the thiol in the rubber particles [18]. In terms ofsulfur pre-vulcanized NR latex particles, Ho and Khew have pro-posed that the diffusion rate of the process chemicals and thecrosslink rate determine the distribution of the crosslinks in the

rubber particles. Hence, a highly crosslinked shell together with asoft non crosslinked vulcanized core is achieved if the diffusion rateof the crosslink chemicals is considerably lower than the reactionrate of the crosslinking. Previous work of Che at al. has confirmedthat peroxide-cured NR latex particles provide coreeshell struc-tures since the crosslinking reaction proceeds too fast to ensure asufficient diffusion of the peroxide crosslinker inside the particle. Incontrast, sulfur cured NR latex particles, using dipentamethylene-thiuram tetrasulfide as sulfur source, are characterized by narrowspatial crosslink distributions which Che at al. mainly relate to theslow crosslink kinetics [26].

In terms of photo-vulcanization, the diffusion of the photo-initiator and the thiol across the water phase into the latex particlesoccurs during the 2 h stirring of the reaction mixture at room tem-perature. Hence, themajority of the required crosslinking chemicalsis already present in the latex particles during the photo-vulcanization which proceeds within minutes whereas thermalprocesses such as the accelerated sulfur vulcanization require re-action times in the range of hours [20]. As a consequence homoge-nous networks are obtained although the reaction time of the thiol-ene photo-vulcanization ismuch faster than the diffusion rate of thechemicals. However, our previous work has demonstrated that thediffusion rateof the chemicals playa key roleduring the stirring time

Fig. 10. Phase contrast images (tapping mode) of dipped IR films photochemically pre-cured with an irradiation intensity of 0.9 W/cm2. (a) soft tapping; (b) hard tapping.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5593

where a lowpolarity of the photoinitiator and the thiol crosslinker iseminent to obtain high crosslink yields and good mechanicalproperties of thiol-ene pre-cured NR films [18].

3.4. Determination of the mechanical properties

Fig. 11a provides the tensile properties of both NR and IR latexfilms pre-cured with a radiation intensity of 0.4 and 0.9 W/cm2,respectively. With respect to pre-cured NR latex films the tensilestrength rises with increasing exposure dose and amounts up to30 ± 2 MPa. In terms of IR latex, no values could be obtained fromfilms pre-cured with 0.9 W/cm2 upon two and three illuminationcycles since the incoherent film formation leads to brittle films withextensive macroscopic cracks. Although an improved film forma-tion is obtained at lower exposure doses corresponding to crosslinkdensities <0.06 mmol/cm3, the tensile strength does not exceed5 MPa. It has to be considered that compared to NR, IR comprisespoor wet gel strength, low amount of branching, reduced gel con-tent and low Mooney viscosity which can be associated with thedifferent microstructure and stabilization of IR and results inconsiderably lower tensile properties of the uncured raw material[19,56].

To provide coherent IR films with higher crosslink densities thephoto-vulcanization was carried out in solid state after the film

Fig. 11. Tensile properties of (a) photochemically pre-cured and (b) photochemically post-curespectively.

formation. Fig. 11b illustrates the tensile strength of both NR and IRlatex films versus the exposure dose employed in the UV illumi-nation of the solid films. It can be clearly shown that both rubbermaterials provide comparable tensile properties in the range of26 ± 2 MPa upon an exposure dose >2 J/cm2. From the results it canbe concluded that NR as well as IR latex films can be successfullycrosslinked by the thiol-ene reaction and provide excellent me-chanical properties when the photo-vulcanization is carried out insolid films. It should be noted that a detailed correlation of themacroscopic (obtained from the Mooney-Rivlin analysis of me-chanical data) and the microscopic properties (characterized by DQNMR) of thiol-ene cured elastomers is the subject of a separatepublication [37].

4. Conclusions

In the present work thiol-ene chemistry has been employed tophoto-vulcanize diene-rubbers, particularly NR and IR, in latexstate as well as in solid film state. By following the crosslink kineticsof NR and it synthetic counterpart with real time infrared spec-troscopy a higher crosslinking reactivity is determined for IR filmsthat can be attributed to the absence of natural radical scavengers(e.g. proteins). After film formation the main characteristics ofthese elastomer-based thiol-ene networks have been determined

red NR and IR latex films versus the number of illumination passes and exposure dose,

S. Schl€ogl et al. / Polymer 55 (2014) 5584e55955594

and their influence on film formation properties, film morphologyand corresponding tensile strength has been highlighted. 1Hmultiple-quantum (MQ) solid-state NMR experiments revealedthat photo-vulcanization in film state promotes homogenous net-works with narrow spatial crosslink distributions which somewhatget broader if the photo-vulcanizationwas carried out in latex state.In both cases the distributions did not show a second componentwhich gives a strong indication that UV pre-cured latex particles donot comprise coreeshell structures with highly crosslinked shellsand lightly crosslinked cores. Additional NMR experiments oftoluene-swollen samples further demonstrated that elastomer-based thiol-ene networks are characterized by a low amount ofshort chain defects (<5%). In terms of film formation, SEM, TEM andAFM measurements revealed that highly crosslinked IR latex par-ticles suffer from inhomogeneous film morphologies with onlypartial coalescence of the latex particles. In contrast, NR filmsexhibit smooth and uniform film morphologies which are lessdependent on the crosslink density of the latex particles. As themacroscopic properties such as the tensile strength are directlyrelated to the film formation properties, UV pre-cured NR filmscomprise good tensile properties whilst highly crosslinked IR filmsare brittle and suffer from substantial cracking. However, in filmstate, photo-vulcanization leads to excellent tensile strengths(26 ± 2 MPa) of both NR and IR materials.

Acknowledgments

This study was performed at the Polymer Competence CenterLeoben GmbH (PCCL, Austria) within the framework of the CometK1 program and the “Produktion der Zukunft” program of theAustrian Federal Ministry for Transport, Innovation and Technologyand of the Austrian Federal Ministry for Economy, Family and Youthwith contributions of University of Leoben and Semperit Techni-sche Produkte GmbH. PCCL is funded by the Austrian Governmentand the State Governments of Styria, Lower Austria and UpperAustria.

References

[1] Posner T. Beitr€age zur Kenntniss der unges€attigten Verbindungen. II. Ueber dieAddition von Mercaptanen an unges€attigte Kohlenwasserstoffe. Ber DtschChem Ges 1905;38(1):646e57.

[2] Morgan CR, Magnotta F, Ketley AD. Thiol/ene photocurable polymers. J PolymSci Polym Chem Ed 1977;15(3):627e45.

[3] Hoyle CE, Lee TY, Roper T. Thiol-enes: chemistry of the past with promise forthe future. J Polym Sci Part A Polym Chem 2004;42(21):5301e38.

[4] Hoyle CE, Bowman CN. Thiol-en-klickchemie. Angew Chem 2010;122(9):1584e617.

[5] Jonkheijm P, Weinrich D, K€ohn M, Engelkamp H, Christianen Peter CM,Kuhlmann J, et al. Photochemical surface patterning by the thiol-ene reaction.Angew Chem Int Ed 2008;47(23):4421e4.

[6] Khire VS, Yi Y, Clark NA, Bowman CN. Formation and surface modification ofnanopatterned thiol-ene substrates using step and Flash imprint lithography.Adv Mater 2008;20(17):3308e13.

[7] Hagberg EC, Malkoch M, Ling Y, Hawker CJ, Carter KR. Effects of modulus andsurface chemistry of thiol-ene photopolymers in nanoimprinting. Nano Lett2007;7(2):233e7.

[8] DeForest CA, Polizotti BD, Anseth KS. Sequential click reactions for synthe-sizing and patterning three-dimensional cell microenvironments. Nat Mater2009;8:659e64.

[9] Dondoni A. The emergence of thiol-ene coupling as a click process for mate-rials and bioorganic chemistry. Angew Chem Int Ed 2008;47(27):8995e7.

[10] Lutz J, Schlaad H. Modular chemical tools for advanced macromolecular en-gineering. Polymer 2008;49(4):817e24.

[11] Hordyjewicz-Baran Z, You L, Smarsly B, Sigel R, Schlaad H. Bioinspired poly-mer vesicles based on hydrophilically modified polybutadienes. Macromole-cules 2007;40(11):3901e3.

[12] Justynska J, Hordyjewicz Z, Schlaad H. Toward a toolbox of functional blockcopolymers via free-radical addition of mercaptans. Polymer 2005;46(26):12057e64.

[13] Mather BD, Viswanathan K, Miller KM, Long TE. Michael addition reactions inmacromolecular design for emerging technologies. Prog Polym Sci 2006;31:487e531.

[14] Metters A, Hubbell J. Network formation and degradation behavior ofhydrogels formed by michael-type addition reactions. Biomacromolecules2005;6(1):290e301.

[15] Nimmo CM, Shoichet MS. Regenerative biomaterials that “Click”: simple,aqueous-based protocols for hydrogel synthesis, surface immobilization and3D patterning. Bioconjug Chem 2011;22(11):2199e209.

[16] Lenko D, Schl€ogl S, Temel A, Schaller R, Holzner A, Kern W. Dual crosslinkingof carboxylated nitrile butadiene rubber latex employing the thiol-enephotoreaction. J Appl Polym Sci 2013;129(5):2735e43.

[17] Schl€ogl S, Temel A, Schaller R, Holzner A, Kern W. Prevulcanization of naturalrubber latex by UV techniques: a process towards reducing type IV chemicalsensitivity of latex articles. Rubber Chem Technol 2010;83(2):133e48.

[18] Schl€ogl S, Temel A, Schaller R, Holzner A, Kern W. Characteristics of thephotochemical prevulcanization in a falling film photoreactor. J Appl PolymSci 2012;124(4):3478e86.

[19] R€othemeyer F, Sommer F. Kautschuk-Technologie: Werkstoffe e Verarbeitunge Produkte. 2nd ed. München [u.a.]: Hanser; 2006.

[20] Akiba M. Vulcanization and crosslinking in elastomers. Prog Polym Sci1997;22(3):475e521.

[21] Chen M, Ao N, Zhang B, Den C, Qian H, Zhou H. Comparison and evaluation ofthe thermooxidative stability of medical natural rubber latex products pre-pared with a sulfur vulcanization system and a peroxide vulcanization system.J Appl Polym Sci 2005;98(2):591e7.

[22] Haque ME, Makuuchi K, Mitomo H, Yoshii F, Ikeda K. A new trend in radiationvulcanization of natural rubber latex with a low Energy electron beam. PolymJ 2005;37(5):333e9.

[23] Cheong IW, Fellows CM, Gilbert RG. Synthesis and cross-linking of poly-isoprene latexes. Polymer 2004;45(3):769e81.

[24] Tangboriboonrat P, Lerthititrakul C. Morphology of natural rubber latex par-ticles prevulcanised by sulphur and peroxide systems. Colloid Polym Sci2002;280(12):1097e103.

[25] Tangboriboonrat P, Tiyapiboonchaiya C. Novel method for toughening ofpolystyrene based on natural rubber latex. J Appl Polym Sci 1999;71(8):1333e45.

[26] Che J, Toki S, Valentin JL, Brasero J, Nimpaiboon A, Rong L, et al. Chain dy-namics and strain-induced crystallization of pre- and postvulcanized naturalrubber latex using proton multiple quantum NMR and uniaxial deformationby in situ synchrotron x-ray diffraction. Macromolecules 2012;45(16):6491e503.

[27] Tangboriboonrat P, Polpanich D, Suteewong T, Sanguansap K, Paiphansiri U,Lerthititrakul C. Morphology of peroxide-prevulcanised natural rubber latex:effect of reaction time and deproteinisation. Colloid Polym Sci 2003;282(2):177e81.

[28] Steward P, Hearn J, Wilkinson M. An overview of polymer latex film formationand properties. Adv Colloid Interface Sci 2000;86(3):195e267.

[29] Ho CC, Khew MC. Surface morphology of prevulcanized natural rubber latexfilms by atomic force microscopy: new insight into the prevulcanizationmechanism. Langmuir 1999;15(19):6208e19.

[30] Rippel MM, Lee L, Leite CA, Galembeck F. Skim and cream natural rubberparticles: colloidal properties, coalescence and film formation. J ColloidInterface Sci 2003;268(2):330e40.

[31] Bouyer D, Philippe K, Wisunthorn S, Pochat-Bohatier C, Dupuy C. Experi-mental and numerical study on the drying process of natural rubber latexfilms. Dry Technol 2009;27(1):59e70.

[32] Dillon RE, Matheson LA, Bradford EB. Sintering of synthetic latex particles.J Colloid Sci 1951;6(2):108e17.

[33] Vanderhoff JW, Tarkowski HL, Jenkins MC, Bradford EB. Theoretical consid-eration of interfacial forces involved in coalescence of latex particles. RubberChem Technol 1967;40(4):1246e69.

[34] Kendall K, Padget JC. Latex coalescence. Int J Adhesion Adhesives 1982;2(3):149e54.

[35] Zosel A, Ley G. Influence of crosslinking on structure, mechanical properties,and strength of latex films. Macromolecules 1993;26(9):2222e7.

[36] Chass�e W, Schl€ogl S, Riess G, Saalw€achter K. Inhomogeneities and local chainstretching in partially swollen networks. Soft Matter 2013;9(29):6943.

[37] Schl€ogl S, Trutschel M-L, Chass�e W, Riess G, Saalw€achter K. Entanglementeffects in elastomers: macroscopic vs. Microscopic properties. Macromole-cules 2014;47(9):2759e73.

[38] Flory PJ. Principles of polymer chemistry. Ithaca: Cornell University Press;1953.

[39] Brandrup J, Immergut EH, Grulke EA. Polymer handbook. 4th ed. New York:Wiley-Interscience; 1999.

[40] Valentín JL, Carretero-Gonz�alez J, Mora-Barrantes I, Chass�e W, Saalw€achter K.Uncertainties in the determination of cross-link density by equilibriumswelling experiments in natural rubber. Macromolecules 2008;41(13):4717e29.

[41] Saalw€achter K. Proton multiple-quantum NMR for the study of chain dy-namics and structural constraints in polymeric soft materials. Prog Nucl MagReson Spectrosc 2007;51(1):1e35.

[42] Valentín JL, Posadas P, Fern�andez-Torres A, Malmierca MA, Gonz�alez L,Chass�e W, et al. Inhomogeneities and chain dynamics in diene rubbers vul-canized with different cure systems. Macromolecules 2010;43(9):4210e22.

[43] Schl€ogl S, Aust N, Schaller R, Holzner A, Kern W. Survey of chemical residuesand biological evaluation of photochemically pre-vulcanized surgical gloves.Monatsh Chem 2010;141(12):1365e72.

S. Schl€ogl et al. / Polymer 55 (2014) 5584e5595 5595

[44] Holzner A, Kern W, Schaller R, Schl€ogl S. Method for producing a cross-linkedelastomer (WO 2010/105283); 2010.

[45] Sussman G, Beezhold DH, Kurup VP. Allergens and natural rubber proteins.J Allergy Clin Immonol 2002;110(2):S33.

[46] Nawamawat K, Sakdapipanich JT, Ho CC, Ma Y, Song J, Vancso JG. Surfacenanostructure of Hevea brasiliensis natural rubber latex particles. ColloidSurface Physicochem Eng Aspect 2011;390(1e3):157e66.

[47] Sansatsadeekul J, Sakdapipanich J, Rojruthai P. Characterization of associatedproteins and phospholipids in natural rubber latex. J Biosci Bioeng2011;111(6):628e34.

[48] Dizhbite T. Characterization of the radical scavenging activity of lignins asnatural antioxidants. Bioresour Technol 2004;95(3):309e17.

[49] Bateman L. Natural rubber Producers' Research Association. The chemistryand physics of rubber-like substances: studies of the natural rubber Pro-ducers' Research Association. London: Maclaren; 1963.

[50] Pukkate N, Horimai T, Wakisaka O, Yamamoto Y, Kawahara S. Photoreactiveparticle prepared from natural rubber and 1,9-nonandioldimethacrylate.J Polym Sci A Polym Chem 2009;47(16):4111e8.

[51] Nakason C, Kaesaman A, Yimwan N, Ketsarin K. Preliminary study on thepreparation of latex foam rubber from graft copolymer of deproteinisednatural rubber and methyl methacrylate. J Rubber Res (Kuala Lumpur,Malays.) 2001;4(3):141e52.

[52] Abad LV, Relleve LS, Aranilla CT, Aliganga AK, San Diego CM, dela Rosa AM.Natural antioxidants for radiation vulcanization of natural rubber latex. PolymDegrad Stab 2002;76(2):275e9.

[53] Chass�e W, Lang M, Sommer J, Saalw€achter K. Cross-link density estimation ofPDMS networks with precise consideration of networks defects. Macromol-ecules 2012;45(2):899e912.

[54] Blackley DC. Polymer latices. Science and technology: applications of latices.2nd ed. London: Chapman & Hall; 1997.

[55] M�endez-Vilas A, J�odar-Reyes AB, Gonz�alez-Martín ML. Ultrasmall liquiddroplets on solid surfaces: production, imaging, and relevance for currentwetting research. Small 2009;5(12):1366e90.

[56] Toki S, Sics I, Ran S, Liu L, Hsiao BS. Molecular orientation and structuraldevelopment in vulcanized polyisoprene rubbers during uniaxial deformationby in situ synchrotron X-ray diffraction. Polymer 2003;44(19):6003e11.