Comparative analysis of delayed-onset peroxide crosslinking formulations.
Peroxide-initiated crosslinking is practiced widely to improve the mechanical and thermal properties of ethylene-rich polyolefins [1, 2], When performed on saturated polymers in the absence of co-agents, these reactions are stoichiometric processes, where crosslink yield is a fraction of the amount of initiator charged to the formulation . This is a consequence of the reaction mechanism (Scheme 1), in which hydrogen atom abstraction from the polymer yields alkyl macro-radicals whose termination by combination generates the desired crosslink functionality. Inefficiencies result from the [beta]-scission of alkoxy radicals to ketone + methyl radicals [4, 5], the latter being much less reactive in hydrogen abstraction than the cumyloxyl intermediates derived directly from dicumyl peroxide (DCP) thermolysis , Further inefficiency arises from macro-radical termination by disproportionation to alkane + olefin, which has no immediate effect on the polymer's architecture.
This report is concerned with strategies to delay the onset of crosslinking in peroxide cure formulations. Typical commercial processes are non-isothermal batch reactions where the polymer is compounded with initiator and other additives prior to forming into the desired shape, and then cured to generate a thermoset material. To avoid crosslinking before the compound is shaped, a problem known as scorch, the onset of cure must be timed appropriately , This can be challenging, since peroxide thermolysis is a relatively slow, first-order decomposition that controls the rate of subsequent radical reactions , Cure rates are, therefore, fastest in the early stages, and methods of suppressing radical concentrations during this initial phase can be an important process design consideration.
We recently described the use of acrylated nitrxoxyls such as 4-acryloyloxy-2,2,6,6-tetramcthylpiperidine-N-oxyl (AOTEMPO) (Scheme 2) for this purpose, and demonstrated their ability to delay crosslinking precisely and without compromising crosslink density [9, 10]. However, a wide range of scorch-protecting additives are described in the open and patent literature, some of which are claimed to provide both induction delays and improvements to crosslink yields. This report examines six additives, butylated hydroxyl toluene (BHT), [alpha]-methylstyrene dimer (MSD), 1,1-diphenyl ethylene (DPE), 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO), bis(l-oxyl-2,2,6,6-tetramethylpiperidine-4-yl)sebacate (bis-TEMPO), and AOTEMPO, with the objective of measuring their performance under standardized conditions. Direct comparisons of their scorch protection and cure suppression effects are discussed in the context of what is known of their radical chemistry. Extensive data on linear low-density polyethylene (LLDPE) crosslinking are augmented with results on m-polyisoprcne (PIP) cure formulations.
Linear LLDPE (5 mole % octene copolymer, Dow Chemical Company) and PIP (97%, Scientific Polymer Products) were purified by dissolution/precipitation (hexanes/acetone) and dried under vacuum prior to use. DCP (98%), hydroxy-TEMPO (4hydroxyl-2,2,6,6-tetramethylpiperidin-l-oxyl, 97%), TEMPO (2,2,6,6-tetramethyl-l-piperidinyloxy, 98%), triethylamine ([greater than or equal to]99.5%), sebacoyl chloride (95%), 1,1-diphenylethylene (DPE, 97%), 2,4diphenyl-4-methyl pentene (MSD, 97%) were used as received from Sigma Aldrich (Oakville, ON). 4-Acryloyloxy-2,2,6,6,-tetramethylpiperidine-N-oxyl (AOTEMPO) was prepared as described previously .
Bis(2,2,6,6-tetramethyl-4-piperidinyl) sebacate was prepared by reaction of a solution of TEMPOH (2.0 g, 12.8 mmol), triethylamine (1.042 g, 1.436 mL, 10.6 mmol), and sebacoyl chloride (2.534 g, 2.230 mL, 10.6 mmol) in toluene were maintained at room temperature for 12 h. A white precipitate was compacted by centrifugation prior to decanting the liquid and isolating a viscous, tan liquid under reduced pressure. [sup.1]H-NMR: (CD[Cl.sub.3]): [delta] (ppm) = 5.05 (2 H, tt, HC-O-C), 4.01 (4 H, tt, COOC-H), 1.75 (4 H, dd, piperidinyl CH), 1.60 (4 H, t, piperidinyl CH), 1.215 (3 H, s, piperidinyl C[H.sub.3]), 1.17 (3 H, s, piperidinyl C[H.sub.3]). High-res. M.S. [C.sub.28][H.sub.50][N.sub.2][0.sub.6] calculated m/z 510.3669; found m/z 510.3665.
LLDPE powder (5 g) was coated with an acetone solution containing the required amounts of DCP and scorch additive, hand mixed, and allowed to dry thoroughly. PIP was sheeted on a two-roll mill, treated with initiator + additive solution and remilled post-drying. Excessive drying of solutions containing TEMPO was avoided. The resulting polymer mixtures were cured in a controlled strain rheometer (Advanced Polymer Analyzer2000, Alpha Technologies) equipped with biconical disks and operating at 1 Hz and 3[degrees] arc, unless stated otherwise.
[FORMULA NOT REPRODUCIBLE IN ASCII]
Physical Property Analysis
Differential scanning calorimetry (DSC) measurements were acquired with a DSCQ100 calorimeter from TA Instruments using heating and cooling rates of 5[degrees]C/min. Crystallization temperatures were measured following an initial heating to 200[degrees]C, while melting data were generated from the second sample heating.
LLDPE (40 g) pellets were coated with an acetone solution containing the required amounts of DCP and AOTEMPO, and allowed to dry. The resulting mixture was compression molded at 160[degrees]C and 20 MPa for 60 min to yield macrosheets of 2.00 [+ or -] 0.05 mm thickness, from which dog bones were cut according to ASTM D4482 . Tensile strength data were acquired using an INSTRON Series 3360 universal testing instrument, operating at a crosshead speed of 500 mm/min at 23 [+ or -] 1[degrees]C.
[FORMULA NOT REPRODUCIBLE IN ASCII]
RESULTS AND DISCUSSION
Scorch Protection of LLDPE Formulations
A standard method of monitoring the progress of a polyolefin crosslinking process is to measure dynamic storage modulus (G') at a fixed temperature, frequency, and shear strain amplitude . Efficient stress relaxation within uncrosslinked polymer melts produces a relatively inelastic response to an oscillating shear deformation. The covalent polymer network created by macro-radical combination restricts polymer segment mobility, thereby raising the compound's G' in a manner that correlates with crosslink network density. Figure 1 presents a plot of G' versus time for LLDPE containing 18.5 [micro]mol DCP per gram of polymer. The sample started to crosslink immediately, and progressed with kinetics that closely resemble a first-order process. Comparing the evolution of G' with the first-order conversion of the initiator (predicted on the basis of the reported 5.5 min half-life at 160[degrees]C) , demonstrates the controlling influence of peroxide thermolysis on crosslinking dynamics. Since initiator decomposition dictates the instantaneous concentration of macro-radicals, additives are needed to quench radical activity in the early stage of the curing process.
The compounds illustrated in Scheme 2 were assessed as scorch-protecting additives by monitoring their effect on LLDPE + DCP formulations at 140, 160, and 180[degrees]C. The molar ratio of additive to cumyloxyl radicals, hereafter referred to as the trapping ratio, was maintained at 0.25. If a compound is an effective scorch-protectant, this loading should be sufficient to quench crosslinking activity in the reaction's early stages. Moreover, if an additive affects the final cure yield (either positively or negatively), this concentration should produce a measureable response.
Figure 2 shows the effect of BHT on the dynamics and yields of an LLDPE cure, with data plotted on a double log scale to highlight changes in G' in the earliest stages of the process. The response to BHT is typical of a relatively efficient anti-oxidant , in that the initial cure rate is retarded, but not quenched entirely. This suggests that the rate of hydrogen atom donation to cumyloxyl, methyl, and macro-radical intermediates is not sufficient to eliminate all macro-radical termination events, resulting in a small degree of crosslinking from the onset of peroxide decomposition. Whereas a nitroxyl-based system such as TEMPO (Fig. 2) provides an induction period during which time the modulus is unchanged, kinetically less reactive traps such as BHT only slow the initial rate. As such, these additives are usually characterized by a scorch time, [t.sub.10%], defined as the time needed for a cure formulation to reach 10% of its ultimate storage modulus change. For example, the BHT formulation reacted at 160[degrees]C reached a final modulus that was 115 kPa above its initial value ([DELTA]G' = [G'.sub.max] - [G'.sub.min] =115 kPa), and reached 10% of this value (11.5 kPa) in 1.9 min, giving [t.sub.10%]= 1.9 min.
The data illustrated in Fig. 3a shows that the efficacy of BHT as a radical trap is highest at 140[degrees]C, both in terms of delaying crosslinking and affecting the final storage modulus. That only 0.25 equivalents of BHT relative to DCP-derived initiator radicals can have such a great effect on ultimate crosslink density is attributed to the multiple quenching reactions provided by this antioxidant. It is well known that hydrogen atom donation by BHT yields a resonance stabilized phenoxy radical whose relative stability precludes subsequent hydrogen atom abstraction from polymer . Therefore, termination pathways are preferred, yielding non-radical products through combination and/or disproportionation (Scheme 3) . This sequence can quench two radicals per molecule of BHT, but the trapping efficiency can be much greater, given that termination products are capable of further radical reactivity that can affect macro-radical concentrations.
The macro-radical trapping efficiency of BHT, as indicated by the delayed onset character of its LLDPE cure formulations, is sufficient at 180[degrees]C to establish it as a scorch protecting additive. However, the depression of ultimate crosslink density is problematic. In order to reach a given modulus target, a BHT formulation will require more initiator than a BHT-free formulation. More initiator increases initial cure rates, thereby necessitating the use of more scorch protectant. Continued increasing of peroxide and anti-oxidant is not only undesirable from a cost perspective; it leads to higher levels of reaction by-products in the thermoset. Because these by-products are not polymerbound, they may be extracted or leached from the article.
Nitroxyl-based additives such as TEMPO provide several performance advantages over hydrogen donor antioxidants such as BHT. In the first place, they combine with alkyl radicals at the diffusion limit of reaction velocities . As demonstrated in Fig. 3b, TEMPO provides induction times during which no crosslinking activity whatsoever is observed. This condition holds as well at 140[degrees]C as it does at 180[degrees]C, thereby providing a robust method of overcoming scorch problems. Alkyl radical quenching yields alkoxyamines whose stability to thermolysis and disproportionation depends on the alkyl substituent (Scheme 4; [R.sup.=] = olefin) . Reversibility is valued by those practicing controlled radical polymerization of acrylic and styrenic monomers , but not in scorch-retardant applications. Since an ideal delayed-onset formulation does not generate unbound byproducts, it is desirable to have stable alkoxyamines derived from macroradical trapping.
The limited data available on alkoxyamine stability indicates that secondary alkoxyamines do not undergo nitroxyl exchange readily at 160[degrees]C, meaning that their formation under the present reaction conditions is effectively irreversible . Moreover, disproportionation to olefin + hydroxylamine requires on the order of hours to reach measurable conversion, indicating that alkoxyamine stability is sufficient for the purposes of polyolefin crosslinking. Note that the disproportionation mechanism is not well understood, and the reaction may occur through radical or non-radical pathways. In either case, the resulting hydroxylamine would be readily oxidized back to TEMPO, thereby trapping another initiator-derived radical intermediate .
One of the potential benefits of a nitroxyl-based cure formulation is the binding of the scorch-protecting reagent to the polymer, thereby rendering it unextractable from the cured product. Unlike phenolic anti-oxidants, DPE and MSD, nitroxyls do not trap oxygen-centered radicals, and cannot, therefore, quench cumyloxy radicals derived from DCP. Rather, they trap only polymer macro-radicals and the methyl radicals originating from cumyloxyl fragmentation. The latter are particularly troublesome from the perspective of by-product leaching, since 1-methoxy2,2,6,6-tetramethyl-l-piperidine (methyl-TEMPO) is a volatile organic compound that is not covalently bonded to the polyolefin. The potential of the AOTEMPO system to convert methylalkoxyamine to polymer-bound functionality will be considered, following an examination of the bis-TEMPO, MSD, and DPE systems.
Figure 3c illustrates the effect of bis-TEMPO on peroxideinitiated LLDPE cures. This approach is reported to retard initial cure rates at 140[degrees]C, and increase cure extents at 180[degrees]C, making this the first of the so-called "cure-boosting" scorch retardants examined in this study . This non-isothermal experimental approach is widely practiced, since most commercial crosslinking processes activate the peroxide while heating continuously to a temperature that provides the appropriate melt viscosity. The data presented in Fig. 3c show that bis-TEMPO affected LLDPE crosslinking in a manner similar to that of BHT. Whereas TEMPO provided a definitive induction time, bis-TEMPO only retarded the initial rate while reducing ultimate crosslink densities.
Scheme 5 provides some insight into the limited scorch protection performance of the bis-TEMPO system. Given that nitroxyl functionality will trap methyl radicals and polymer macro-radicals at a rate that is limited only by molecular diffusion, the rate of alkoxyamine formation is controlled strictly by the rate of peroxide thermolysis. This is true of peroxide-only formulations as well, in which macro-radical termination rates are limited to the rate of peroxide breakdown. Note that initial trapping by bis-TEMPO does not quench a macro-radical, but merely converts it to a pendant nitroxyl intermediate (Scheme 5). This new macro-radical can combine with another macroradical to yield a crosslink, or with a methyl radical to give a pendant alkoxyamine. As such, bis-TEMPO cannot completely inhibit polyolefin crosslinking in the manner of a monofunctional nitroxyl such as TEMPO. The observed reduction in the state of cure may arise from the trapping of methyl radicals, preventing their abstraction of hydrogen from the polymer to yield the macro-radicals needed for crosslink formation.
[FORMULA NOT REPRODUCIBLE IN ASCII]
The MSD system is a rational approach to controlling crosslinking rates and yields , It is based on the trapping of macro-radicals by C=C addition to give a benzylic radical that is susceptible to fragmentation (Scheme 6) [24, 25]. The resulting cumyl radical is resonance-stabilized, and has a limited capacity to abstract hydrogen from the polyolefin. Hence, when R- is cumyloxyl, methyl or a macro-radical, addition to MSD quenches its reactivity, leading to cure-retardation . Because the rate of radical addition to a styrenic monomer is much slower than the rate of trapping by nitroxyl, MSD cannot provide the induction behavior generated by TEMPO (Fig. 3d). However, macro-radical trapping by MSD can yield terminal styrenic functionality, thereby converting the polyolefin into a macro-monomer. If this functionality is oligomerized later in the cure process, the potential exists for recovering crosslink density that was lost to macroradical trapping , Unlike the stoichiometric macro-radical formation + macro-radical termination sequence that comprises a peroxide-only cure, a macro-monomer oligomerization process can have kinetic chain length, producing many crosslinks from a single initiator-derived radical.
While some reports indicate that MSD can boost cure yields , the data presented in Fig. 3d show no such improvements for a LLDPE + DCP formulation. Clear evidence of delayed crosslinking was observed, but irrespective of the temperature employed, significant losses in crosslink density were incurred. This may be due to excessive trapping of cumyloxyl by MSD, which could suppress crosslinking while producing the corresponding vinylether. Ideally, MSD would intervene in the crosslinking process by trapping macro-radicals to give pendant styrenic groups in high yield. If the yield of macro-monomer groups is too low, then the potential for generating crosslinks by oligomerization will be stunted, as observed in the present study.
[FORMULA NOT REPRODUCIBLE IN ASCII]
[FORMULA NOT REPRODUCIBLE IN ASCII]
While literature on the radical chemistry of DPE is largely focused on controlled radical polymerization , its efficacy as a scorch protectant has also been claimed . The cure data plotted in Fig. 3e demonstrates that 0.25 equivalents of DPE relative to cumyloxyl depressed both the rate and the extent of polyolefin crosslinking. The mechanism of action is likely to be similar to that of MSD, in that radical trapping by C=C addition yields a resonance-stabilized benzylic radical (Scheme 7) . In this case, radical fragmentation is not as favorable, and the likely outcome for the persistent radical intermediate is termination. Without a mechanism for regaining crosslink density, the use of DPE provides scorch protection at the expense of cure yield.
AOTEMPO technology is a hybrid of the TEMPO and MSD scorch protection strategies . It is based upon differences in the kinetic reactivity of nitroxyl and C=C functionality toward radical trapping , Recall that the rate constants the rate constants for alkyl radical combination with nitroxyls are generally of the order of [10.sup.8] - [10.sup.9] [M.sup.-1]s[l] , while those for alkyl radical addition to acrylates are of the order of [10.sup.3]-[10.sup.5] [M.sup.-1]s[l] . Therefore, AOTEMPO can trap macro-radicals by combination, as opposed to acrylate addition, thereby quenching radical activity while transforming the polyolefin into a macromonomer-bearing pendant acrylate functional groups. This approach is analogous to MSD, which has the potential to introduce styrenic functionality to the polymer (Scheme 6). However, the kinetic reactivity of AOTEMPO is far superior to that of MSD, as is its selectivity for carbon-centered radicals over oxygen-centered radicals. The latter is particularly important, since AOTEMPO allows cumyloxyl to abstract hydrogen from the polymer to produce the requisite macro-radical intermediates. The result is a superior yield of polymerizable functionality in the macromonomer .
[FORMULA NOT REPRODUCIBLE IN ASCII]
The cure rheology data presented in Fig. 3f demonstrate the three phases of an AOTEMPO cure, an induction period, an accelerated crosslinking phase, and continuing crosslinking phase involving peroxide-only chemistry. An idealized reaction mechanism is illustrated in Scheme 8. During the induction phase, macro-radicals and methyl radicals are trapped by combination with AOTEMPO, quenching radical activity to produce alkoxyamine intermediates. This eliminates crosslinking until nitroxyl concentrations fall to the point where radical addition to acrylate functionality is kinetically competitive. Once all AOTEMPO is consumed, oligomerization of pendant acrylate groups generates crosslinks by a kinetic chain process that requires relatively few initiating radical species, owing to the efficiency of acrylate polymerizations. Once all macromonomer functionality is converted, curing proceeds by a stoichiometric process that supports conventional polyolefin crosslinking.
[FORMULA NOT REPRODUCIBLE IN ASCII]
The performance of AOTEMPO is remarkable, in that it provides an induction period of no crosslinking, followed by a complete recovery of crosslink density during the acrylate oligomerization phase of the process. This unique combination is not provided by any other additive, as can be seen from the scorch time and cure yield data summarized in Table 1. Moreover, the induction time ([t.sub.ind]) provided by AOTEMPO abides by a simple function of the peroxide half-life ([t.sub.1/2]) and the trapping ratio [10, 35].
[t.sub.ind] = [[t.sub.1/2]/ln(1/2)] ln [1 - [AOTEMPO]/2[DCP]]
As a result, the onset of an AOTEMPO-mediated cure can be predicted reliably, whereas the scorch time provided by kinetically less reactive traps such as BHT, DPE and MSD must be established by trial and error experimentation.
This chemistry is also remarkable in terms of its potential use of methyl radical byproducts. In all of the systems examined to date, scorch protection was achieved at the expense of volatile organic compound production. This has consequences in terms of thermoset properties such as color, odor, and potential toxicity. AOTEMPO quenches methyl radicals to yield the corresponding acrylate monomer, whose oligomerization is expected to render it polymer-bound, or as a high molecular weight acrylate oligomer. As such, this technology has the potential to be more environmentally friendly than other scorch protection strategies.
Scorch Protection of PIP Formulations
Up to this point in the study, the efficacy of scorch retardants was examined only for LLDPE; a saturated polyolefin that requires quenching of alkyl macro-radicals. However, good performance on these materials does not necessarily translate to unsaturated polymers such as PIP. These unsaturated materials are better hydrogen atom donors, owing to the low-bond dissociation energy of allylic C-H bonds. As a result, hydrogen atom abstraction from these materials by cumyloxyl is faster, giving a greater proportion of allylic macro-radicals than methyl radicals, when compared to a polymer such as LLDPE , Moreover, the stability of quenched intermediates can differ substantially between alkyl and allylic radical systems. For example, cleavage of the benzylic radicals illustrated in Schemes 6 and 7 back to their starting materials may be more important for PIP cures than for a LLDPE system.
Figure 4 provides storage modulus data for DCP-initiated cure formulations of PIP operating at 160[degrees]C. Note that the PIP used in this work had a higher initial modulus than LLDPE, starting at 32 kPa compared to just 8 kPa for LLDPE. Nonetheless, 18.5 [micro]mol DCP per gram of polymer cured the diene-based elastomer efficiently, yielding an ultimate modulus of 234 kPa after 60 min reaction time. The inability of BHT, MSD, and DPE to affect the cure process is noteworthy, since these additives retarded cure rates and yields in LLDPE formulations. In the case of PIP, 0.25 equivalents of BHT and DPE relative to cumyloxyl had no measurable effect, implying that neither compound affected macro-radical concentrations at any point in the curing process.
[FORMULA NOT REPRODUCIBLE IN ASCII]
The nitroxyl-based reagents provided ample scorch protection by retarding crosslinking rates substantially, but not completely. The early stages of the TEMPO and AOTEMPO cures were indistinguishable, with G' increasing marginally, as opposed to the LLDPE system where the modulus remained constant throughout the induction period. Previous research on the stability of l-(l-ethylpent-2-enyloxy)-2,2,6,6-tetramethylpiperidine (TEMPO-heptene) to nitroxyl exchange and disproportionation showed these allylic alkoxyamines to be susceptible to thermolysis at 160[degrees]C , That is, the trapping of allylic radicals by TEMPO at these temperatures is reversible, with a dynamic equilibrium established between radical intermediates and the spin-paired alkoxyamine (Scheme 9). In the present context, this instability would support a small population of allylic macro radicals, whose termination by combination gives rise to the observed storage modulus increases. Fortunately, the allylic alkoxyamines derived from PIP appear to be sufficiently stable to provide useful delayed-action cure performance.
Whereas the TEMPO formulation provided scorch protection at the expense of crosslink density, AOTEMPO proved capable of restoring the crosslink density lost to allyl radical quenching. The final modulus of the AOTEMPO formulation was equivalent to that provided by DCP alone, and the thermoset showed no evidence of cure reversion.
Physical Properties of LLDPE Thermosets
We conclude our studies of AOTEMPO-mediated polymer crosslinking with a brief examination of melt-state and solid-state properties. Given that DCP alone cures the polymer by radical combination, crosslinks generally have an "H-type" structure, whereas oligomerization of the acrylate functionality introduced during an AOTEMPO cure can yield polyfunctional branch points wherein more than two polymer chains are bound at a single locus , The purpose of the present work is to determine whether these formulations produce significantly different rheological, crystallization, and tensile properties.
Figure 3e shows that AOTEMPO formulations with a trapping ratio of 0.25 produce a final storage modulus slightly greater than that generated by peroxide alone. Since this nitroxyl loading quenches only 25% of the cumyloxy radicals produced by the initiator, the cure yield should be dominated by macroradical coupling, as opposed to acrylic macro-monomer oligomerization. That is, the network generated by DCP + AOTEMPO at low trapping ratios should be very similar to that generated by DCP alone, since 75% of macro-radicals can terminate without interference from nitroxyl, and the amount of pendant acrylate functionality available for crosslinking is relatively small. The rheological data presented in Fig. 5 is consistent with this distribution of network structures. Storage modulus values for DCP and DCP + AOTEMPO formulations responded to frequency and strain amplitude in a consistent manner, indicating that the covalent networks established within each thermoset are comparable.
Similarities between DCP-only and AOTEMPO-mediated thermoset properties were also observed in DSC and tensile analyses (Table 2). Reductions in crystallization temperature ([T.sub.c]), melting temperature ([T.sub.m]), and degree of crystallinity that result from increases in polyethylene branch content  and crosslink density  are well established. In the present case, significant changes in [T.sub.c] were accompanied by small shifts in [T.sub.m] and melting enthalpy. Static tensile analysis of compression molded specimens demonstrated reductions in elongation at break values that are widely reported for LLDPE crosslinking [40, 41]. More importantly, the properties of each crosslinked LLDPE derivative were statistically equivalent, thereby providing further assurance that AOTEMPO can provide precise control over peroxide cure dynamics without adversely affecting thermoset properties.
The extent of LLDPE crosslinking in the early stages of a DCP-initiated cure is reduced by a range of radical traps, including hydrogen donors (BHT), fragmenting styrenic monomers (MSD, DPE), and nitroxyl-based additives (TEMPO, bisTEMPO, AOTEMPO). Of these scorch protectants, AOTEMPO provided predictable induction periods without affecting the melt-state rheological or solid-state tensile properties of an LLDPE thermoset. Quenching of the allylic radical population that supports PIP crosslinking could only be accomplished by mono-nitroxyl reagents, which provided effective scorch protection, if not complete crosslinking suppression.
[1.] I. Chodak, Prog. Polym Sci., 20, 1165 (1995).
[2.] A. Msakni, P. Chaumont, and P. Cassagnau, Polym. Eng. Sci., 46, 1530 (2006).
[3.] D. Maximilian, Adv. Polym.Tech., 5, 87 (1985).
[4.] J.R.L. Smith, E. Nagatomi, A. Stead, D.J. Waddington, and S.D. Beviere, J. Chem. Soc. Perkin Trans., 2, 1193 (2000).
[5.] A. Baignee, J.A. Howard, J.C. Sciano, and L.C. Stewart, J. Am. Chem. Soc., 105, 6120 (1983).
[6.] T. Zytowski and H. Fischer, J. Am. Chem. Soc., 119, 12869 (1997).
[7.] S. Mani, P. Cassagnau, M. Bousmina, and P. Chaumont, Macromolecules, 42, 8460 (2009).
[8.] J. Liu, W. Yu, C. Zhao, and C. Zhou, Polymer 48, 2882 (2007).
[9.] D.K. Hyslop and J.S. Parent, Macromolecules, 45, 8147 (2012).
[10.] D.K. Hyslop and J.S. Parent, Polymer, 54, 84 (2013).
[11.] ASTM Standard D4482-11, "Standard Test Method for Rubber Property: Extension Cycling Fatigue," in ASTM International, West Conshohocken, PA, 2007, doi: 10.1520/D4482-11.
[12.] F. Romani, R. Corrieri, V. Braga, and F. Ciardelli, Polymer, 43, 1115 (2002).
[13.] Technical Bulletin, Atofina Chemicals (2001).
[14.] N. Grassie and G. Scott, in Polymer Degradation and Stabilization. Cambridge University Press, New York, 119 (1985).
[15.] J. Pospisil, Polym. Degrad. Stahil. 20, 181 (1988).
[16.] T. Yamazaki and T. Seguchi, J. Polym. Sci.: Part A: Polym. Chem., 35, 2431 (1997).
[17.] E. Rizzardoand D.H. Solomon, Polym. Bull. (Berlin) 1,529 (1979).
[18.] F. Cicogna, S. Coiai, E. Passaglia, I. Tucci, L. Ricci, F. Ciardelli, and A. Batistini, J. Polym. Sci. Part A: Polym. Chem., 49, 781 (2011).
[19.] C.J. Hawker, A.W. Bosnian, and E. Harth, Chem. Rev., 101, 3661 (2001).
[20.] M.E. Scott, J.S. Parent, J. Dupont, and R.A. Whitney, Ind. Eng. Chem. Res., 42, 3662 (2003).
[21.] G.S. Ananchenko and H. Fischer, J. Polym. Sci.: Part A: Polym. Chem., 39, 3604 (2001).
[22.] B.I. Chaudhary, L. Chopin, and J. Klier, Pohm. Eng. Sci., 47, 50 (2006).
[23.] S. Suyama, H. Ishigaki, Y. Watanabe, and T. Nakamura, Polymer J., 27(4), 371 (1995).
[24.] Y. Watanabe, H. Ishigaki, H. Okada, and S. Suyama, Polymer J., 29(11), 940 (1997).
[25.] A.A. Gridnev, J. Polym. Sci. Part A: Polym. Chem., 40, 1366 (2002).
[26.] Y. Watanabe, H. Ishigaki, H. Okada, and S. Suyama, Polymer J., 29(4), 366 (1997).
[27.] L. Rossiello, A. Buttafava, L. Martinotto, F. Peruzzotti, G.M. Ghisoni, and A. Faucitano, Res. Chem. Interned., 28, 89 (2002).
[28.] H. Ishigaki and S. Suyama, U.S. Patent 5,298,564 (Assigned to Nippon Oil and Fats Co.) (1994).
[29.] P.C. Wieland, B. Raether, and O. Nuyken, Macromol. Rapid Commun., 22, 700 (2001).
[30.] F.F. Nelson, T.T. Nguyen, A. Biswas, P.J. Caronia, M.J. Keogh, J.S. Furno, and L.H. Gross, Patent EP1036804A1, 14 03 (2000).
[31.] S. Viala, M. Antonietti, K. Tauer, and W. Bremser, Polymer, 44, 1339 (2003).
[32.] F. Debaud, A. Defrancisci, O. Guerret, and J. Kervennal (Atofina Co.) PCT application WO 02/028950, (2002).
[33.] J. Chateauneuf, J. Lusztyk, and K.U. Ingold, J. Org. Chem., 53, 1629 (1987).
[34.] K.E. Russell, Prog. Polym. Sci., 27, 1007 (2002).
[35.] S. Mani, P. Cassagnau, M. Bousmina, P. Chaumont, Polymer, 51, 3918 (2010).
[36.] J. Lai, J.E. McGrath, and R.D. Board, J. Polym. Sci. Part A-I:Polym. Chem., 6, 821 (1968).
[37.] M. Lazar, Adv. Polym. Sci., 5, 149 (1989).
[38.] M.A. Kennedy, A.J. Peacock, M.D. Failla, J.C. Lucas, and L. Mandelkem, Macromolecules, 28. 1407 (1995).
[39.] H.A. Khonakdar, J. Morshedian, M. Mehrabzadeh, U. Wagenknecht, and S.H. Jafari, Eur. Polym. J., 39 1729 (2003).
[40.] C.J. Perez, G.A. Cassano, E.M. Valles, L.M. Quinzani, and M. D. Failla, Polym. Eng. Sci., 43, 1624 (2003).
[41.] H.A. Khonakdara, J. Morshedian, U. Wagenknecht, and S.H. Jafari, Polymer, 44, 4301 (2003).
Brian M. Molloy, David K. Hyslop, J. Scott Parent
Department of Chemical Engineering, Queen's University, Kingston, Ontario, Canada
Correspondence to: J. Scott Parent; e-mail: email@example.com
Published online in Wiley Online Library (wileyonlinelibrary.com).
TABLE 1. Summary of scorch protecting reagent performance. [DELTA] [G'.sub.DCP]- Temp [t.sub.10%] [DELTA]G' [DELTA] Additive ([degrees] (min) (kPa) [G'.sub.Additive] C) (kPa) None 140 10.9 308 -- BHT 140 10.9 129 179 DPE 140 21.7 207 101 MSD 140 14.5 198 110 Bis-TEMPO 140 18.1 207 99 AOTEMPO 140 25.3 302 6 TEMPO 140 43.3 142 165 None 160 2.2 271 -- BHT 160 1.9 115 156 DPE 160 2.8 155 116 MSD 160 2.5 147 124 Bis-TEMPO 160 2.5 204 67 AOTEMPO 160 3.1 299 -28 TEMPO 160 4.3 187 84 None 180 0.6 197 -- BHT 180 0.6 99 98 DPE 180 0.7 158 39 MSD 180 0.7 144 54 Bis-TEMPO 180 0.7 141 56 AOTEMPO 180 0.9 232 -34 TEMPO 180 1.1 153 44 DCP = 18.5 [micro]mol/g; BHT = TEMPO= MSD= AOTEMPO = 9.2 [micro]mol/g; Bis-TEMPO = 4.6 [micro]mol/g TABLE 2. DSC and tensile analyses of LLDPE and its thermoset derivatives. LLDPE + LLDPE DCP Crystallization 84 78 temperature ([degrees]C) Melting 102 100 temperature ([degrees]C) Enthalpy of 33 32 melting (J/g) Young's 198 [+ or -] 10 181 [+ or -] 1 modulus (kPa) Stress at break 49.3 [+ or -] 1.0 46.6 [+ or -] 1.4 (MPa) Elongation at 441 [+ or -] 6 285 [+ or -] 10 break (mm) LLDPE + DCP + AOTEMPO Crystallization 77 temperature ([degrees]C) Melting 99 temperature ([degrees]C) Enthalpy of 32 melting (J/g) Young's 186 [+ or -] 4 modulus (kPa) Stress at break 43.4 [+ or -] 1.9 (MPa) Elongation at 275 [+ or -] 19 break (mm) DCP= 18.5 [micro]mol/g; AOTEMPO = 9.2 [micro]mol/g; 160[degrees]C cure temperature
|Printer friendly Cite/link Email Feedback|
|Author:||Molloy, Brian M.; Hyslop, David K.; Parent, J. Scott|
|Publication:||Polymer Engineering and Science|
|Date:||Nov 1, 2014|
|Previous Article:||Synthesis and thermal properties of poly(ethylene 2,6-naphthalate) copolyesters containing modified diacid monomer.|
|Next Article:||Effect of water absorption on the mechanical properties of poly(vinylidene fluoride-trifluoroethylene) copolymer films.|