α,ω-Alkanedithiol cross-linking of high vinyl 3,4-polyisoprene

Abstract

This paper evaluates the ability of α,ω-alkanedithiols to cross-link high vinyl 3,4-polyisoprene rubber (represented by Isogrip) through the thiol-ene addition reaction and provides microstructural insights into where cross-linking takes place. The thiol-ene reactions are the hydrothiolation of a C = C bond, which can be initiated in a number of ways. It has been postulated that organic peroxides are very effective in initiating such cross-links and that α,ω-alkanedithiol-ene cross-linking of high vinyl 3,4-polyisoprene primarily takes place on the α-carbon of the unsaturated sites of the rubber chains.

Keywords

Isogrip 3 4-Polyisoprene Alkanedithiols Cross-linking High vinyl

Introduction

High vinyl 3,4-polyisoprene refers to polyisoprene with 50–70% 3,4-vinyl group content. In this study, Isogrip was used to represent this family. Isogrip is the commercial name for a dry synthetic polyisoprene with a special microstructure of a mixture of predominantly 3,4-vinyl isoprene (∼60%) units and the rest being 1,4-isoprene (∼30%) and 1,2-isoprene (∼10%) units (Fig. 1).¹

Possible polyisoprene structural configurations

The general reactions of thiols with dienes (thiol-ene reactions) undergo radical addition of thiols to unactivated carbon–carbon double bonds as well as Michael type addition reactions.^{2, 3} The thiol-ene addition process begins by the formation of thiyl radicals. These radicals then add to α,β unsaturation to yield a carbon centred radical. Subsequent abstraction of hydrogen by this carbon centred radical then follows.⁴ The general mechanism is shown in Fig. 2.

General mechanism for peroxide initiated thiol-ene polymerisation process⁴

From the mechanism, it can be deduced that cross-linking can be achieved by combination of a polyene with a dithiol. Thiol-ene cross-linking can be initiated by standard cleavage initiators that produce radicals.^5–7 These include photoinitiators and hydrogen abstraction type initiators. The detailed mechanism and modelling of thiol-ene photopolymerisation is described by Cramer et al.⁸

Sasin et al.⁹ reported that the addition of alkanedithiols to olefins takes place to form compounds of anti-Markownikov configuration. Boileau et al.¹⁰ investigated the functionalisation of vinylidene terminated oligoisobutenes with several thiols, with free radicals being generated from peroxydicarbonates or ultraviolet radiation. They reported that the reaction is of an anti-Markovnikov type with the RS function adding to the vinylidene = CH₂ group selectively in the presence of more substituted olefin groups in polyisobutene.

Thiol-ene curing has been used in a number of specific applications, mainly on surface coatings and adhesives.^11–16 The pursuit of additional applications has been a driving force behind academic and industrial research for the continued development in thiol-ene cross-linking. The variety of enes one may select as coreagents with the thiols makes thiol-enes very versatile systems with a large room for further exploration. Therefore, this study explores the likelihood of α and ω –SH functional groups of homobifunctional alkanedithiols forming thiol-ene cross-links on high vinyl 3,4-polyisprene.

Experimental

Table 1 shows a list of various industrial grade materials and chemical agents used in this study.

Table 1

Materials and chemical agents used in this study

Materials/chemical agents	Supplier
3,4-Polyisoprene (Isogrip)	Karbochem, Newcastle (South Africa)
1,3-Propanedithiol, assay≥97%	Merck, Hohenbrunn (Germany)
1,4-Butanedithiol, assay≥97%	Merck, Hohenbrunn (Germany)
1,5-Pentanedithiol, assay≥95%	Sigma-Aldrich, St Louis (USA)
1,6-Hexanedithiol, assay≥96%	Sigma-Aldrich, St Louis (USA)
1,8-Octanedithiol, assay≥97%	Sigma-Aldrich, St Louis (USA)
1,9-Nonanedithiol, assay≥95%	Sigma-Aldrich, St Louis (USA)
Dicumyl peroxide, assay≥97%	Sigma-Aldrich, St Louis (USA)
Chloroform, assay≥99%	Merck, Wadeville (South Africa)

*Chloroform was preferred for swelling experiments over other organic solvents because of its superior degree of swelling for hydrocarbon rubbers.^{17, 18}

Preparation of samples

The compounds were prepared in a Brabender Plasticorder internal mixer (model PL 2200), with a mixing head volume of 85 mL, by first adding three-fourths of the rubber to the mixing head at a rotor speed of 60 rev min⁻¹. The rotor speed was dropped to 30 rev min⁻¹, and the curatives were added over 1 min and the remainder of the rubber was then added. The rotor speed was increased again to 60 rev min⁻¹, and the compounds were mixed for 3 min. They were then removed, milled on a Schwabenthan two-roll mill with a nip size of 4 mm and remixed in the mixing head for a further 3 min at a rotor speed of 60 rev min⁻¹. The internal temperature of the mixing head was maintained at 30–40°C throughout the procedure by water circulation.

Rubber vulcanizates of ∼2 mm thickness were prepared in a preheated hydraulic press at 150°C for 1 h (to accommodate curing times of all samples for easier comparison). At the end of the run, samples were quickly removed from the press and put in cold water to quench any reactions.

Cross-link density

As described in previous studies,^{17, 19, 20} swelling measurements were carried out using chloroform as solvent (chloroform: χ = 0·37, ρ = 1·47 g mL⁻¹).²¹ Appropriate samples were clipped from the prepared rubber sheets and immersed in chloroform for 48 h in the dark, replacing the solvent every 24 h.

The swollen samples were then taken out, and the excess liquid on the surface of the specimen was removed by dabbing carefully with a paper towel. The swollen weight was immediately measured. With regard to the deswelling, samples were placed to dry in the cupboard for 2 days and then under vacuum for a further 2 days. The mass of the dry pellets and that of swollen pellets were used to calculate the volume fraction of rubber in the swollen gel V_r using equation (1) (1) where M_r is the residual mass (g) of rubber after extraction, M_s is the solvent mass (g) at equilibrium swelling, ρ_r is the density of IR (0·908 g mL⁻¹) and ρ_s is the density of solvent.

The number of cross-links per unit mass of rubber is a more useful parameter. This is determined by acknowledging the fact that each chain has two ends, and therefore, each cross-link unites four ends.²² This leads to the following equation (2) where 1/2M_c is the cross-link density in mol g⁻¹ (2) where M_c is the average molecular mass of network chains between adjacent cross-links, ρ is the density of solvent, V_s is the molar volume of solvent, χ is the Flory–Huggins interaction parameter (a constant for a polymer/solvent pair).

Cross-link structure

The attenuated total reflection (ATR) spectra were taken with a Platinum ATR spectrometer, model TENSOR 27 (Bruker). The samples were measured directly on a Diamond IRE with a single bounce mode. The spectra were collected over 4000–400 cm⁻¹, with a resolution of 4 cm⁻¹ and 32 scans per sample. The force applied to the samples during spectral measurements was ∼80 N. Spectra were analysed using OPUS spectroscopy software.

The ¹³C solid state nuclear magnetic resonance (NMR) experiments were conducted at ambient probe temperature using a Varian VMRS Wide Bore 500 MHz spectrometer operating at 75·4 MHz. The spectra were collected using a Varian 4 mm HX probe and the samples spun at 10–15 kHz. Cross-polarisation magic angle spinning (CP/MAS), with dipolar dephasing, experiment was carried out with a dephasing time of 40 μs. The ¹³C chemical shifts were calculated using Spin Works 3·0 processing software.

Results and discussion

The swelling characteristics of peroxide initiated α,ω-alkanedithiol cross-linked Isogrip as a function of α,ω-alkanedithiol chain length at different concentrations of α,ω-alkanedithiols are shown in Fig. 3.

Effect of size of alkanedithiol chains on peroxide initiated Isogrip/α,ω-alkanedithiols systems (1 phr DCP)

According to the theory of swelling of cross-linked polymers,²³ the extent of swelling depends on the degree to which the network is cross-linked. The equilibrium swelling of a rubber network immersed in a solvent is influenced by the resisting force of the rubber network, driven by its entropic elasticity. Therefore, the high cross-link density values (Fig. 3) can be attributed to the ability of dicumyl peroxide to initiate the high vinyl thiol-ene cross-linking process. Moreover, it is clear from Fig. 3 that an increase in α,ω-alkanedithiols concentration causes an increase in cross-link density. Although cross-link density does not seem to be affected much by the chain size, it is worth noting that the shortest chain, 1,3-propanedithiol, displays relatively higher values of cross-link density. This may be attributed to the reactivity of the 1,3-propyldithyil radicals.²⁴ This behaviour is possibly analogous to peptisation during mastication of raw rubber to reduce rubber viscosity.^25–27 Peptising agents are known to commonly decrease rubber viscosity and therefore improve processability,^26–28 better distributing curatives and enhancing the formation of cross-links. During mixing, shear generated polymer radicals are prevented from recombination by the action of thiol radicals specifically added for this purpose. In the case of a dithiol, the dithiol can be a di-radical and so does not terminate recombination. Such reactions likely only make minor contributions to cross-linking.

One may argue that the cross-links may be as a result of the weaker C–S bond strength (∼272 kJ mol⁻¹)²⁹ in the α,ω-alkanedithiols, which possibly ruptures when exposed to elevated temperatures and not necessarily via the thiol-ene reaction as anticipated. This possibility could not be overlooked. Figure 4 presents cross-link density as a function of α,ω-alkanedithiol chain size for uninitiated Isogrip/α,ω-alkanedithiols systems. It can be seen that significant cross-link density is obtained only in the cases where 1,3-propanedithiol was used. As discussed above, this could be due to the relatively enhanced reactivity of 1,3-propyldithiol chain and/or the associated breaking down of Isogrip when treated with 1,3-propyldithiol. With regard to other α,ω-alkanedithiols, the cross-link density values are very low, indicating very low likelihood of the C–S bond scission upon heating. Furthermore, the distinctive smell of these thiols was still present after curing. This is indicative of unaffected –SH groups in the resultant samples. The little amount of cross-links obtained could in fact be as a result of radicals formed by thermal scission and shear stress during processing.³⁰

Effect of size of alkanedithiol chains on uninitiated Isogrip/α,ω-alkanedithiols systems

Organic peroxides can also initiate rubber cross-linking on their own.^31–33 When Isogrip is cross-linked solely with 1 phr dicumyl peroxide (DCP), the cross-link density is ∼36×10⁻⁵ mol g⁻¹, which is generally low compared to when α,ω-alkanedithiols are present. This also rules out the possibility of peroxide dominating the cross-linking process.

Further evidence for the presence of cross-links may be found in the ATR spectra. Figures 5–7 show different regions of the ATR spectrum of raw Isogrip. The main peaks in the spectra are identified as follows: the relatively weak absorption at 3069 cm⁻¹ is due to the C–H stretching vibrations in = CH₂ groups, which indicate the existence of vinyl groups.^{34, 35} Cornell and Koenig³⁶ indicate that this vibration is present in both the 1,2- and 3,4-structural units. The strong asymmetrical and symmetrical –CH₂ stretching of methylene group vibrations present are observed near 2922 and 2854 cm⁻¹ respectively with the band near 2964 cm⁻¹ corresponding to methyl stretching.^{37, 38} The closeness of these bands for different isomers does not allow them to be resolved. The shoulder at 2980 is characteristic of a = CH₂ stretch only present in 3,4-units. The minor band at 2727 cm⁻¹ is the result of another methylene group stretch in 1,4-units. These methyl and methylene stretches can be seen more clearly in Fig. 5.

Attenuated total reflection spectra of peroxide initiated Isogrip/α,ω-alkanedithiols systems (1·5 phr α,ω-alkanedithiols, 1 phr DCP)

The vibration at 1664 cm⁻¹ is the C = C stretching vibration of 1,4-units, while the 1643 cm⁻¹ absorption is due to the C = C stretching vibrations at vinyl end groups.^{32, 33, 36} The relative strength of the 1643 cm⁻¹ is indicative of the preponderance of 3,4-structures in Isogrip. The band at 1591 cm⁻¹ is characteristic of conjugated C = C stretches because conjugation lowers the wavenumber associated with the stretch.

Absorption at 1432 cm⁻¹ is due to the C–H bending vibrations in methylene groups, while the nearby band at 1450 cm⁻¹ is due to methyl bending vibrations. These bands cannot be resolved between isomers. The band at 1374 cm⁻¹ corresponds to the CH₂ wagging.³⁸ The absorptions between 1350 and 1200 cm⁻¹ are C = C–H bending vibrations. Figure 6 provides a close-up of this region in raw Isogrip.

Attenuated total reflection spectra of peroxide initiated Isogrip/α,ω-alkanedithiols systems (6 phr α,ω-alkanedithiols, 1 phr DCP)

The strong absorbing peak at ∼885 cm⁻¹ is due to the = C–H out of plane deformation vibrations in allyl groups.^{34, 35} The shoulders on this peak are due to out of plane vibrations of 1,4-units. The peaks at ∼725 and 757 cm⁻¹ are attributed to the typical (CH₂)_n≥5 for rocking vibrations.^38–40 Cornell and Koenig suggest that the band at 1005 cm⁻¹ is associated with methyl vibrations.³⁶ This region is expanded in Fig. 7.

Attenuated total reflection spectra of peroxide cross-linked Isogrip (1 and 3 phr DCP)

Since the degree of curing and conversion may be assessed by the decrease in the intensity of vinyl absorption bands,³⁵ the relative intensities of different bands were measured, and the data are presented in Table 2. The data for each band are normalised relative to an internal reference which is the –CH₂- stretch at 2922 cm⁻¹ which is almost unaffected by vulcanisation. The value of each for Isogrip is set equal to 100 and all others are reported relative to this value. This treatment corrects for an uncertainty by differences in IR path length, which could occur if the sample is not placed in the exact same position on the ATR.

Table 2

Effect of α,ω-alkanedithiols on intensities of various 1 phr DCP initiated α,ω-alkanedithiols cured systems

Band/cm⁻¹	α,ω-alkanedithiols
Band/cm⁻¹	1,3-propyldithiol/%	1,6-hexyldithiol	1,9-nonyldithiol
3069	72·3	71·4	71·2
1664	48·6	40·6	47·0
1643	82·2	80·5	80·1
1002	27·2	29·2	25·1
885	93·1	90·8	90·7
725	33·5	38·6	35·7

*Values reported relative to the intensity of the band in uncured Isogrip.

Similar behaviour was observed with the other α,ω-alkanedithiols. No significant differences were seen between the thiols, including 1,3-propyldithiol despite the higher cross-link density observed with a curing system based on this dithiol.

This might be because the additional cross-linking observed with 1,3-propyldithiol may involve a different mechanism or more likely because Fourier transform infrared (FTIR)-ATR cannot differentiate between a pendent group or a cross-link. The reason for this behaviour is unclear, but since other α,ω-alkanedithiols did not show such a behaviour, it is therefore likely that cross-linking of Isogrip in the presence of 1,3-propyldithiol follows a different mechanism. Possible reactions may include dithiol-ene linking between vinyl groups of the same chain and cyclisation occurring on the same double bond.

Figure 8 illustrates the changes to the methyl and methylene stretching vibrations. It can clearly be seen that upon addition of dithiols, the band at 3069 cm⁻¹ decreases in size. This indicates reaction at the = CH₂ group of 1,2- and 3,4-units indicating that thiol-ene reactions are occurring.

Methyl and methylene stretching vibration region of FTIR spectrum of Isogrip upon curing with DCP accelerated α,ω-alkanedithiols

Figure 9 illustrates changes in the C = C region. Both the absorbances at 1643 and 1664 cm⁻¹ decrease. Surprisingly the stretch at 1664 cm⁻¹ is affected more. This is likely the effect of DCP alone cross-linking. The postulated conjugated stretches at 1591 cm⁻¹ have all but disappeared.

C = C stretching vibrations of raw Isogrip of Isogrip upon curing with DCP accelerated α,ω-alkanedithiols

In the region where out of plane bending vibrations occur, there is a significant reduction in the peak at 1002 cm⁻¹ as well as at 725 cm⁻¹. This is illustrated in Fig. 10.

Out of plane and rocking vibrations in FTIR spectrum of raw Isogrip upon curing with DCP accelerated α,ω-alkanedithiols

High resolution solid state ¹³C NMR response can also give good information on cross-linked samples. Figure 11 exhibits the ¹³C CP/MAS NMR spectrum of untreated Isogrip. The assignment of sequence structure of respective isomeric units is based on the pioneering ¹³C NMR studies of 3,4-/cis-1,4-polyisoprene by Gronski et al., ⁴⁰ Beebe⁴¹ and the recent study by Pilichowski et al.⁴² The latter authors allotted peaks to the respective isomeric units as shown in Fig. 12.

¹³C NMR spectrum of untreated Isogrip

¹³C NMR shift attribution of polyisoprene⁴²

In the aliphatic region of the spectrum (0–50 ppm), the resonance of the methyl carbon in the vinyl side group of 3,4-units (Fig. 13, (i)) is depicted by a signal at ∼18 ppm, whereas the methyl carbon of 1,2-units is represented by a signal at ∼16 ppm. The peak with the signal at 23 ppm is assigned to the methyl carbon of 1,4-units (Fig. 13, (ii)). The region between 25 and 50 ppm represents methylene and methyl carbon atoms of different isomeric units.

Structures of different isomeric units

In the olefinic region of the spectrum (100–170 ppm), the peak with the signal at ∼111 ppm represents the CH₂ for the vinyl sequence (Fig. 13, (iii)). The α and β carbon atoms on the 1,4-units are assigned peaks with the signals at ∼135 and 125 ppm respectively (Fig. 13, (iv and v)). α carbon atoms of the 3,4-units (Fig. 13, (vi)) are observed at ∼148 ppm.

Comparison of the treated samples to the untreated one in Fig. 14 reveals that upon addition of α,ω-alkanedithiols, in the presence of the initiator, the intensities of peaks in the olefinic region (except the peak at ∼111 ppm) are greatly reduced. Since these peaks represent unsaturation in all isomeric units of Isogrip, it is evident that cross-linking predominantly affects the in-chain double bonds most probably due to the associated addition reactions. It is generally known that vinyl units submit to addition reactions more readily than the in-chain double bonds.³⁷ However, the peak assigned to the = CH₂ of the vinyl sequence at ∼111 ppm appears to be minimally affected when compared to other peaks. This suggests that, in this part of the microstructure, cross-linking occurs via the abstraction of a hydrogen atom from the carbon alpha to the vinyl double bond. This is confirmed by the decrease in the intensity of the methyl carbon in the vinyl side group of 3,4-units at ∼18 ppm.

1: untreated Isogrip; 2: iso/6(1,9-nonanedithiol)/1DCP; 3: iso/6(1,3-propanedithiol); 4: iso/6(1,9-nonanedithiol); 5: iso/1,5(1,9-nonanedithiol)/1DCP; 6: iso/1,5(1,3-propanedithiol)/1DCP

With regard to the aliphatic region, it can clearly be seen in Fig. 14 that there is a significant decrease in the intensity of the signal at ∼18 ppm. This resonance is assigned to the aliphatic methyl carbon in the vinyl group of 3,4-units, which confirms that cross-linking occurs by the abstraction of the hydrogen from this methyl carbon on the vinyl site. The general enhancement of the aliphatic methylene peaks between 25 and 50 ppm, which are generally assigned to the methylene and methyl carbon atoms of different isomeric units, suggests that the cross-linking process might involve double bond migration which then result in isomeric structural changes.

Since hydrocarbon signals are highly dominant in the samples spectra, other chemical components are therefore susceptible to low sensitivity for detection. As a result, it is rather challenging to confirm the structural information of cross-links. However, it can be seen that when high concentrations of initiated α,ω-alkanedithiols are used [i.e. iso/6(1,9-dithiol)/1DCP], the intensity of the peak at ∼30 ppm increases and a shoulder peak develops at ∼45 ppm. According to the findings in the solid state HRMAS study on cross-link structures in high cis-butadiene rubber carried out by Hulst et al., ⁴³ these peaks could be attributed to the C–S–C link. These findings substantiate the peroxide initiated α,ω-alkanedithiols cross-linking process, in agreement with the cross-link density data and the ATR findings.

Conclusions

Cross-link densities increased as the concentration of α,ω-alkanedithiols increased when DCP was used as an initiator. In all cases, the extent of cross-linking exceeded that of DCP alone, indicating that the α,ω-alkanedithiols were able to produce cross-links. The greatest cross-link density was achieved with 1,3-propyldithiol. In the absence of DCP, cross-linking was much reduced. This indicates that, for dithiols to be used as vulcanising agents, the systems need to be activated with a radical producing initiator. Although not tested here, ultraviolet light could also fulfil this role when producing transparent or translucent vulcanizates. The extent of cross-linking was independent of the dithiol chain length with the exception of 1,3-propyldithiol.

1,3-Propyldithiol was able to produce cross-links on its own, suggesting that it is able to act as a self-initiator. This may also explain why an increased level of cross-linking was achieved when 1,3-propyldithiol was used in conjunction with DCP compared with longer α,ω-alkanedithiols.

Nuclear magnetic resonance studies also confirmed that cross-linking takes place at 3,4-units to a greater extent than 1,2-units. Spectra (FTIR-ATR) suggested however that dithiols appear capable of reacting with all types of double bonds given the decrease in absorbance for the vibrations associated with each type of polyisoprene unit but cannot distinguish between pendent groups and cross-links. Nuclear magnetic resonance also revealed that cross-linking also took place within 1,4-units, but this is likely the result of DCP cross-linking.

Footnotes

Acknowledgements

The authors are grateful to NRF (Thuthuka grant) and Nelson Mandela Metropolitan University for financial support of this work.

References

Nordsiek

and Wolpers

: ‘High Tg polyisoprenes for superior wet grip of tyre treads’, Kaut. Gummi. Kunst. , 1990, 43, 755–760.

Ikeda

, Yasuda

, Makino

, Yamamoto

, Tosaka

, Senoo

and Kohjiya

: ‘Strain-induced crystallization of peroxide-crosslinked natural rubber’, Polymer , 2007, 48, 1171–1175.

Mather

, Viswanathan

, Miller

and Long

: ‘Michael addition reactions in macromolecular design for emerging technologies’, Prog. Polym. Sci. , 2006, 31, 487–531.

Hayden

: ‘The cross-linking and chain fracture of rubber by high-energy radiation’, Int. J. Appl. Radiat. Isot. , 1960, 8, 65–71.

Shin

, Nazarenko

, Philips

and Hoyles

: ‘Physical and chemical modifications of thiol-ene networks to control activation energy of enthalpy relaxation’, Polymer , 2009, 50, 6281–6286.

Carioscia

, Stansbury

and Bowman

: ‘Evaluation and control of thiol-ene/thiol-epoxy hybrid networks’, Polymer , 2007, 48, 1526–1532.

Jia

, Liu

, Davis

and Bulmus

: ‘RAFT polymerization and thiol-ene modification of 2-vinyloxyethyl methacrylate: towards functional branched polymers’, Polymer , 2009, 50, 5928–5932.

Cramer

, Davies

, O'Brien

and Bowman

: ‘Mechanism and modelling of a thiol-ene photopolymerization’, Macromolecules , 2003, 36, 4631–4636.

Sasin

, Doyle

, Campbell

, Caruso

and Sasin

: ‘Addition of alkanedithiols to olefins’, J. Chem. Eng. Data , 1963, 8, 102–103.

10.

Boileau

, Mazeaud-Henri

and Blackborow

: ‘Reaction of functionalised thiols with oligoisobutenes via free-radical addition. Some new routes to thermoplastic crosslinkable polymers’, Eur. Polym. J. , 2003, 39, 1395–1404.

11.

, Zhou

and Hoyle

: ‘The effect of thiol and ene structures on thiol-ene networks: photopolymerization, physical, mechanical and optical properties’, Polymer , 2009, 50, 2237–2245.

12.

Decker

and Nguyen Thi Viet

: ‘Photocrosslinking of functionalized rubbers IX. Thiol-ene polymerization of styrene-butadiene-block-copolymers’, Polymer , 2000, 41, 3905–3912.

13.

Roper

, Kwee

, Lee

, Guymon

and Hoyle

: ‘Photopolymerization of pigmented thiol-ene systems’, Polymer , 2004, 45, 2921–2929.

14.

Yang

, Wicks

, Yuan

, Pu

and Liu

: ‘Newly UV-curable polyurethane coatings prepared by multifunctional thiol- and ene-terminated polyurethane aqueous dispersion: photopolymerization properties’, Polymer , 2010, 51, 1572–1577.

15.

Black

and Rawlins

: ‘Thiol-ene UV-curable coatings using vegetable oil macromonomers’, Eur. Polym. J , 2009, 45, 1433–1441.

16.

Guenthner

, Hess

and Cash

: ‘Morphology development in photopolymerization-induced phase separated mixtures of UV-curable thiol-ene adhesive and low molecular weight solvents’, Polymer , 2008, 49, 5533–5540.

17.

Sohby

, Mahdy

MMM

, El-Fayoumi

MAK

and Abdel-Bary

: ‘Effect of waste rubbers powder in SBR formulations on the swelling of different organic solvents’, Polym. Test. , 1997, 16, 349–362.

18.

Alli

, Hazer

and Baysal

: ‘Determination of solubility parameters of cross-linked macromonomeric initiators based on polypropylene glycol’, Eur. Polym. J. , 2006, 42, 3024–3031.

19.

Schlick

, Gao

, Matsukawa

, Ando

, Fead

and Rossi

: ‘Swelling and transport in polyisoprene networks exposed to benzene-cyclohexane mixtures: a case study in multicomponent diffusion’, Macromolecules , 1998, 31, 8124–8133.

20.

Mousa

and Karger-Kocsis

: ‘Rheological and thermodynamical behaviour of styrene/butadiene rubber-organoclay nanocomposites’, Macromol. Mater. Eng. , 2001, 286, 260–266.

21.

Wolf

: ‘Polymer-solvent interaction parameters’, in ‘Polymer handbook’, (ed. Brandrup

and Immergut

J H

), 2nd edn, IV-131-134; 1975, New York, Wiley-Interscience Publishers.

22.

Barton

: ‘Vulcanization studies’, J. Polym. Sci. , 1955, 18, 559–562.

23.

Sombatsompop

: ‘Swelling characteristics of some natural rubber based elastomers’, Prog. Rubber Plast. Technol. , 1998, 14, 208–225.

24.

Boileau

, Mazeaud-Henri

and Blackborow

: ‘Reaction of functionalised thiols with oligoisobutenes via free-radical addition. Some new routes to thermoplastic crosslinkable polymers’, Eur. Polym. J. , 2003, 39, 1395–1404.

25.

Bateman

, Moore

, Porter

and Saville

: ‘The chemistry of vulcanization’, in ‘The chemistry and physics of rubber-like substances’, (ed. Bateman

), Chap. 15; 1963, London, Maclaren.

26.

Liebman

, Behringer

and Derrington

: ‘Masticating polybutadiene’, US patent no. 3,324,100, 6 June 1967. 419–462.

27.

Kaplan

and Pinto

: ‘Peptizing agent for natural rubber and synthetic butadiene-styrene rubber’, US patent no. 4,129,538, 12 December 1978.

28.

M. Hensel, K.H. Menting, T. Mergenhagen and H. Umland: ‘New insights into the mastification process’, Kaut. Gummi. Kunst., 2004, 57, 95–98.

29.

Brydson

: ‘Rubber chemistry’, Chap. 4, 66–105; 1978, London, Applied Science Publishers.

30.

Pinheiro

, Chinelatto

and Canevarolo

: ‘The role of chain scission and chain branching in high density polyethylene during thermo-mechanical degradation’, Polym. Degrad. Stab. , 2004, 86, 445–453.

31.

Thitithammawong

, Nakason

, Sahakaro

and Noordermeer

: ‘Effect of different types of peroxides on rheological, mechanical, and morphological properties of thermoplastic vulcanizates based on natural rubber/polypropylene blends’, Polym. Test. , 2007, 26, 537–546.

32.

Barroso-Bujans

, Verdejo

, Perez-Cabero

, Agouram

, Rodriguez-Ramos

, Guerrero-Ruiz

and Lopez-Manchado

: ‘Effects of functionalized carbon nanotubes in peroxide crosslinking of diene elastomers’, Eur. Polym. J. , 2009, 45, 1017–1023.

33.

and Koenig

: ‘Proton spin–spin relaxation and NMR imaging of peroxide/coagent crosslinked cis-1,4-polyisoprene’, Polymer , 1999, 40, 4703–4708.

34.

Martinez-Barrera

, Lopez

, Castano

and Rodriguez

: ‘Studies on the rubber phase stability in gamma irradiated polystyrene-SBR blends by using FT-IR and Raman spectroscopy’, Radiat. Phys. Chem. , 2004, 69, 155–162.

35.

Gunasekaran

, Natarajan

and Kala

: ‘FTIR spectra and mechanical strength analysis of some selected rubber derivatives’, Spectrochim. Acta A , 2007, 68A, 323–330.

36.

Cornell

and Koenig

: ‘Raman spectra of polyisoprene rubbers’, Rubber Chem. Technol. , 1970, 43, 313–321.

37.

Everall

, Chalmers

and Griffiths

: ‘Vibrational spectroscopy of polymers: principles and practice’; 2007, Chichester, John Wiley & Sons Ltd.

38.

Coates

: ‘Interpretation of infrared spectra, a practical approach’, in ‘Encyclopedia of analytical chemistry’, (ed. Meyers

R A

); 2000, Chichester, John Wiley & Sons Ltd. 10815–10837.

39.

Ivan

, Meghea

, Giurginca

and Iftimie

: ‘Thermo-oxidative degradation of some elastomers with a high content of 3,4 isoprene units’, Polym. Degrad. Stab. , 2003, 80, 397–402.

40.

Gronski

, Murayama

and Cantow

: ‘On the structure of 3,4-/cis-1,4-polyisoprene by ¹³C NMR’, Polymer , 1976, 17, 358–360.

41.

Beebe

: ‘Structure of 3,4-(cis-1,4-)trans-1,4-polyisoprene by ¹³C NMR’, Polymer , 1978, 19, 231–233.

42.

Pilichowski

, Morel

, Tamboura

, Chmela

, Baba

and Lacoste

: ‘Crosslinking and ageing of ¹³C labelled polyisoprene part 1: synthesis and polymerisation of 4-¹³C-isoprene’, Polym. Degrad. Stab. , 2010, 95, 1575–1580.

43.

Hulst

, Seyger

, van Duynhoven

JPM

, van der Does

, Noordermeer

JWM

and Bantjies

: ‘Vulcanization of butadiene rubber by means of cyclic disulfides. 2. A 2D solid state HRMAS NMR study on cross-link structures in BR vulcanizates’, Macromolecules , 1999, 32, 7509–7520.