The role of functional polymers in improving tire performance.
Anionic polymerizations of 1,3-dienes, and their copolymerizations with styrene monomers, are typically initiated using alkyllithium reagents such as sec-butyl- or n-butyl lithium. Under proper conditions, i.e., anhydrous and inert atmosphere, the polymerization is living, indicating the absence of chain transfer and chain termination. Well-defined polymer architectures may be synthesized, e.g., branched, block and chain-end functional polymers. Chain-end functionalization of polymers synthesized from compounds such as tin tetrachloride (refs. 1 and 2), trialkyltin chlorides (refs. 1 and 2), dimethylilnidazolidinone (refs. 3 and 4) or alkylthiothiazoline (ref. 5) in carbon black filled rubber compounds has given reduced hysteresis. Alternatively, chain-end functional polymers may be synthesized by initiating anionic polymerization with functionalized initiators such as trialkyltin lithium (ref. 6), secondary amino lithium (refs. 7 and 8), protected hydroxyl initiators (refs. 9-13), 2-1ithio-l,3-dithianes (refs. 14 and 15) and protected amine initiators (refs. 9, 10 and 16-19).Recently, the use of silica as a filler in rubber compounds has become more common, which has lead to the need for a functional polymer that will interact with silica. The alkoxysilane functional group has long been used in coupling agents for silica compounds, e.g., bis-(3-triethoxysilylpropyl)tetrasulfane (ref. 20). Prior work has shown that alkoxysilanes are able to react with polyisoprenyl lithium chains (refs. 21 and 22) and polystyryl lithium chains (refs. 23 and 24).Termination of anionic styrene-butadiene copolymerization with tetraethylorthosilicate (TEOS) yields an alkoxysilane functionalized polymer that has been shown to interact with silica in rubber stocks on a laboratory scale (ref. 25).
This article will elucidate the challenges and benefits of using alkoxysilane functionalized polymers. Although prior work utilized [sup.29]Si NMR to characterize the chain-end functionality on alkoxysilane polymers (refs. 21-23), the molecular weight was much lower than that used in tire compounds. At molecular weights greater than 100 kg/mol, the use of [sup.29]Si NMR to characterize chain-end functionality is not currently possible. Thus, a new method of characterizing the coupling will be presented using tetrabutylammonium fluoride to predominantly decouple siloxane linkages. Differentiation of coupling resulting from condensation of alkoxysilanes vs. either thermal coupling or coupling resulting from two polymer chains reacting with one alkoxysilane molecule is now possible. Alkoxysilane compounds are inherently hydrolytically unstable and condense to form siloxane linkages that will increase viscosity during storage. Methods to prevent this increase will be shown. Finally, results of an evaluation of tire properties will be presented comparing a non-functional polymer to a TEOS terminated polymer.
Experimental
Materials
All solvents and monomers were desiccated through columns, and all vessels and reactors were dried carefully before use. Dried butadiene in hexane, dried styrene in hexane, dried hexane, dried cyclohexane, n-butyllithium (1.68 M in hexane), 2,2-bis(tetrahydrofuranyl) propane (BTP) in hexane (1.6 M solution in hexane, stored over calcium hydride), and di-tbutyl-p-cresol (DBPC) solution in hexane were used. Hexamethyleneimine (Sigma-Aldrich Chemical) was distilled from calcium hydride and used as a solution in cyclohexane. Commercially available reagents and starting materials (SigmaAldrich Chemical and Fisher Scientific) include the following: Tetraethylorthosilicate (TEOS), tin (IV) chloride, silicon (IV) chloride, sorbitan monooleate, 2-ethylhexanoic acid, calcium 2-ethylhexanoate and 1 M tetrabutylammonium fluoride in tetrahydrofuran were used as purchased without further purification.
Polymerizations
The polymerizations were carried out by placing monomers and solvent (mixture of hexanes) in a thoroughly dried reactor, then charging the initiator reagent, followed by BTP. Either bottles or stainless steel autoclaves were used as reactors. The polymerization bottles were baked dry, and had crown caps fitted with cyclohexane-extracted rubber liners through which reagents were charged by syringe (ref. 26). All polymerizations and operations involving lithium reagents were carried out under nitrogen, using standard air-free techniques. The polymerizations were conducted in reactors heated to about 50-80[degrees]C, while agitating for a minimum of about 30 minutes. The active polymerization cements were quenched by introducing a proton source such as nitrogen-sparged 2-propanol, or by treatment with terminating reagents. The polymers were stabilized with DBPC, coagulated from 2-propanol and steam desolventized or passed through a drum-dryer to remove residual solvent.
Coupling analysis
First, the coupling of the alkoxysilane functionalized polymer was measured by gel permeation chromatography. Then, a solution of alkoxysilane functionalized polymer in tetrahydrofuran was treated with excess 1 M tetrabutylammonium fluoride. After 30 minutes, the reaction was terminated through addition of excess calcium 2-ethylhexanoate. Coupling was again measured on the product. The difference of the starting and final coupling was the siloxane linkages. The remaining coupling was assigned to permanent coupling.
Rubber compound preparation
The recipes used for compounding are shown in table 1. For lab scale studies, the masterbatches were prepared by mixing the initial compounds in a 300 g internal mixer operating at 60 rpm and 133[degrees]C. First, the polymer was placed in the mixer, and after 0.5 minutes, the remaining ingredients except the stearic acid were added. The stearic acid was then added after three minutes. The initial components were mixed for five to six minutes. At the end of mixing the temperature was approximately 165[degrees]C. Each sample was transferred to a mill operating at a temperature of 60[degrees]C, where it was sheeted and subsequently cooled to room temperature. The mixtures were re-milled under milder conditions than those of the masterbatch stage. The final components were mixed by adding the masterbatch and the curative materials to the mixer simultaneously. The initial mixer temperature was 65[degrees]C, while operating at 40 rpm. The final material was removed from the mixer after 2.25 minutes when the material temperature was between 100[degrees]C and 105[degrees]C. The finals were sheeted and molded into cylindrical buttons and 15 x 15 x 0.19 cm sheets. The samples were cured at 171[degrees]C for 15 minutes in standard molds placed in a hot press.
[FORMULA NOT REPRODUCIBLE IN ASCII]
Physical testing
The dynamic mechanical properties were measured using two techniques. A TA Instruments ARES rheometer in the parallel plate mode was used with cured 15 mm high by 9.27 mm diameter buttons that were glued to the plate surface. The loss modulus, G", storage modulus, G', and tan [delta] were measured over deformation strain amplitude of 0.25-14.5% at 10 Hz and 60[degrees]C. The Payne Effect was estimated by calculating the difference of G' (0.25% strain)-G' (14.0% strain). A TA Instruments ARES rheometer in the torsion mode was used to measure the temperature dependence. Rectangular samples were used having the dimensions 31.7 mm x 12.7 mm x 2.0 mm. The temperature was increased at a rate of 2[degrees]C [min..sup.-1] from -80[degrees]C to -10[degrees]C and 5[degrees]C [min..sup.-1] from - 10[degrees]C to 100[degrees]C. The moduli (G' and G") were obtained using a frequency of 10 Hz and a deformation of 0.25% from -80[degrees]C to -10[degrees]C and 2% from -10[degrees]C to 100[degrees]C.
Mooney viscosity measurements were conducted according to ASTM-D 1646-89. The test was performed using a large rotor at 130[degrees]C for the filled rubbers and 100[degrees]C for unfilled polymer. The sample was preheated at the test temperature for one minute before the rotor started, and then the Mooney viscosity ([ML.sub.4]) was calculated from the recorded torque after the rotor had rotated for four minutes at 2 rpm (average shear rate about 1.6 [sec.sup.-1]).
Results and discussion
Evaluation of the coupling of tetraethyorthosilicate terminated polymers
The synthesis of alkoxysilane functionalized polymers from an anionic polymerization terminated with tetraethylorthosilicate at equimolar stoichiometry to polymeric lithium led to a polymer with a high coupling. The coupling may result from the hydrolysis and condensation of the alkoxysilane endgroup during workup, reaction of two or more living polymer chains with a single TEOS molecule, or thermal coupling of the polymeric lithium at high temperature (scheme 1). While [sup.29]Si NMR is useful to examine endgroups with low molecular weight polymers (<25 kg/mol), the signal to noise is too low to evaluate high molecular weight polymers (>100 kg/mol) used in the tire industry (refs. 21-23).
Tetrabutylammonium fluoride (TBAF) is a known reagent used to deprotect alcohols with trialkylsilyl protecting groups (ref. 27). Thus, it was expected that this reagent would be active for decoupling of Si-O-Si coupled polymer. Control experiments indicated that TBAF did not influence coupling of non-modified polymer. However, TBAF slowly reacted with Si[Cl.sub.4] coupled SBR. The reaction was found to be first order in silicon-carbon bonds. After 30 minutes, approximately 6.5% of the silicon-carbon bonds reacted at room temperature (figure 1). Since calcium salts are known to react quickly with HF, the addition of calcium 2-ethylhexanoate to the reaction stopped further decoupling. Silicon oxygen bonds were found to react much faster. Thus, TEOS terminated polymer (1 mol TEOS/mol BuLi charge) with a 67% coupled portion was treated with TBAF for 30 minutes leaving 2.0% remaining coupling. The remaining coupling was presumably due to either two polymeric lithium chains reacting with one TEOS molecule or thermal coupling. This facile method is able to distinguish coupling with relatively inexpensive analytical equipment and reagents.
[FIGURE 1 OMITTED]
Stabilizing viscosity of alkoxysilane terminated polymers
The alkoxysilane terminated polymer is able to hydrolyze and condense to form coupled products. Unfortunately, this also occurs during storage and so the viscosity can increase to such an extent that it becomes difficult to mix during compounding. This is more evident when the polymer also contains functionality that is known as a condensation catalyst (refs. 28-30) for alkoxysilanes. Therefore, a method to minimize the viscosity increase was needed. The prevention of condensation of these polymers was attempted through two methods: Neutralization of polymer solution prior to desolventizing and addition of sterically hindered alcohols.
The influence of bases and acids on condensation of alkoxysilanes is cited by Osterholtz et. al. (ref. 31). Since both acids and bases catalyze condensation, neutralization of the basic residue, lithium ethoxide, after termination can greatly slow the rate of viscosity increase in an alkoxysilane terminated polymer. The hydrolysis and coupling is most evident in steam desolventization where the polymer is in contact with excess water at high temperatures. Neutralization of the polymer solution with 2-ethylhexanoic acid yields a polymer with a Mooney viscosity of 23.6 after steam desolventization vs. the same polymer solution untreated at 49.5 Mooney viscosity. Upon aging the polymer also increases in viscosity (table 2).
Scheme 2--mechanism of equilibration of alcohols with alkoxysilanes--P designates a polymer chain
PSi[(OEt).sub.3] + HOR [right arrow] PSi(OR)[(OEt).sub.2] + HOEt
For example, an untreated TEOS terminated polymer with a starting Mooney viscosity of 52.4 increased to 66.4 in nine days aging at 55[degrees]C in 85% relative humidity. In contrast, by neutralizing the polymer cement with 2-ethylhexanoic acid, a polymer with a starting Mooney viscosity of 32.3 increased to only 37.3 Mooney viscosity after ten days aging at 55[degrees]C in 85% relative humidity.
An alternative method of stabilizing the alkoxysilane terminated polymer is to exchange the ethoxy group with a sterically hindered alcohol. Alcohols and alkoxysilanes are known to be in equilibrium (scheme 2). Azeotropic removal of ethanol during steam desolventization drives the reaction to form the sterically hindered polymeric alkoxysilane. The conversion may be measured by gas chromatography after hydrolysis of the alkoxysilane polymer. Thus, a TEOS terminated polymer was treated with two equivalents of 2-ethylhexylalcohol per ethoxy group and steam desolventized. The resulting polymer showed 73.7% conversion of the ethoxy groups measured by gas chromatography after treatment of the polymer with excess p-toluene sulfonic acid, as previously reported by Lin, et. al. (ref. 32).
The aged polymer viscosity was greatly reduced using this procedure. With two equivalents of 2-ethylhexanol per ethoxy group on the polymer, the Mooney viscosity increased from 35.2 to 39.9 in nine days at 55[degrees]C and 85% relative humidity (table 2). Concentration of the sterically hindered alcohol influenced the stability of the resultant polymer. A polymer with 3.7 equivalents of sorbitan monooleate per ethoxy group showed an increase of only 2.7 Mooney viscosity units after nine days at 55[degrees]C and 85% relative humidity. For comparison, a polymer with 0.75 equivalents of sorbitan monooleate per ethoxy group showed an increase of seven Mooney viscosity units after nine days at 55[degrees]C and 85% relative humidity. Since sorbitan monooleate is more sterically hindered than 2-ethylhexanol, at equivalent concentrations it is more effective at preventing viscosity growth.
The two methods may be used individually or in combination to control viscosity of alkoxysilane terminated polymers. Both acid neutralization of basic residues from anionic polymerization and addition of sterically hindered alcohols are able to control viscosity growth.
Comparison of alkoxysilane functional polymer to non-functional polymer in rubber compounds
While alkoxysilane terminated polymers have been examined on a laboratory scale in the past (ref. 25), the results from tire testing have not been reported. Thus, a high silica tread formulation (table 1) was examined in a P215/55R17 passenger tire with either 75 phr of non-functional poly(styreneco-butadiene) or 75 phr of TEOS terminated poly(styrene-cobutadiene). A laboratory indicator of improved rolling resistance is lower tan [delta] at 60[degrees]C. The tan [delta], the normalized energy losses measured at different temperatures, are usually used to predict the tire performances. For example, tire rolling resistance and wet traction are both dictated by the energy loss from the tire service, but encompass different deformation magnitudes and frequencies (refs. 33-35)
Use of tan [delta] at 0[degrees]C as a predictor of tire wet traction, and tan [delta] at 60[degrees]C as a predictor of rolling resistance are widely practiced in the tire industry. The functional polymer containing stock had a 9.5% lower tan [delta] at 60[degrees]C than the non-functional polymer containing stock (table 3). When coastdown rolling resistance, SAE J2452, was measured on the tires, the improvement was 7%. The rolling resistance predictor tan [delta] at 60[degrees]C agrees with the tire test results. The improvement in rolling resistance can be attributed to the results of improved polymerfiller interaction and a less developed filler network formed in the compound (ref. 25). The degree of filler networking is typically measured by the Payne Effect (ref. 36) (G' 0.25%E--G' 14.5%E) and is reduced as illustrated from strain sweep data (figure 2 and table 3). In addition, stronger temperature dependence on hysteresis, shown in figure 3, was found for the alkoxysilane functionalized polymer. It is suggested that improved polymer-filler interaction is obtained to give such a dependence (ref. 37). Thus, at low temperatures (0[degrees]C), the hysteresis for the functional polymer stock is equal to or greater than that of the non-functional stock. However, at higher temperatures (50-70[degrees]C), the hysteresis is lower. A tread compound with these properties may benefit the tire with improved rolling resistance, comparable wet traction and more reinforcement. Tire test data listed in table 3 confirm these predictions.
An additional benefit of using functionalized polymer for tire tread is the improved wear. For example, the tire with alkoxysilane terminated polymer in the tread compound showed 20% better wear over the tire with non-functional polymer. Wear was measured after 30,000 miles driving consisting of 94% highway driving and 6% city driving with uniform applied load and the tires rotated every 5,000 miles. These benefits were obtained while maintaining the same dry and wet traction (table 3).
Conclusions
Use of alkoxysilane functionalized polymers in tire tread formulations yielded reduced rolling resistance and improved wear. With these benefits came the challenges of developing analytical procedures to characterize the new polymers and control of the viscosity during aging. A method for evaluating the coupled portion of the polymer was developed using tetrabutylammonium fluoride. This showed that under proper termination conditions, the coupled portion came predominantly from condensation of alkoxysilane groups. Untreated polymers increased in Mooney viscosity during humidity aging. This was controlled through the use of neutralization or addition of a sterically hindered alcohol.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
References
(1.) F. Tsutsumi, M. Sakakibara and N. Oshima, Rubber Chem. Technol., 63, 8 (1990).
(2.) T. Fujimaki, M. Ogawa, S. Yamaguchi, S. Tomita and M. Okuyama, Proc. Int. Rubber Conf., Kyoto, Japan, 184 (1985).
(3.) S. Akita, F. Suzuki, Y. Kikuchi and H. Oishi, U.S. Patent 4,929,679 (to Nippon Zeon Co. and Yokohama Rubber Co.), May 29, 1990.
(4.) M. Aonuma, H. Watanabe, H. Kawada, K. Noguchi, A. Ueda, S. Akita, T. Ohyama and T. Kase, U.S. Patent 4,647,625 (to Nippon Zeon Co.), March 3, 1987.
(5.) T.E. Hogan, C.M. Rademacher, W.L. Hergenrother and D.F. Lawson, World Patent Application WO2004/020475A1 (to Bridgestone Corp.), March 11, 2004.
(6.) T.W. Bethea, W.L. Hergenrother, F.J. Clark and S.B. Sarkar, Rubber and Plast. News, 24(2), 17 (1994); Rubber and Plast. News, 1994 Technical Notebook, Crain Communications, Inc. Akron, 1995, pp. 73-76.
(7.) D.F. Lawson, D.R. Brumbaugh, M.L. Stayer, J.R. Schreffler, T.A. Antkowiak, D. Saffles, K. Morita and S. Nakayama, Polymer Preprints, 37 (2), 728 (1996).
(8.) D.F. Lawson, et. al., ACS Rubber Div. Mtg., Presented at Rubber Div., ACS, Indianapolis, Indiana, May 5 (1998).
(9.) R.P. Quirk, S.H. Jang and J. Kim, Rubber Chem. Technol., 69, 444 (1996).
(10.) R.P. Quirk, S.H. Jang, H. Yang and Y. Lee, Macromol. Syrup., 132, 281 (1998).
(11.) R.P. Quirk, S.H. Jang, K. Han, H. Yang, B. Rix and Y. Lee, Functional Polymers, 71 (1998).
(12.) D.N. Schulz, A.F. Halasa, and A.E. Oberster, J. Polym. Sci. Polymer Chemistry Edition, 12, 153 (1973).
(13.) R.P. Quirk, S.H. Jang, H. Yang and Y. Lee, Macromol. Sympos., 161, 37-44 (2000).
(14.) T.E. Hogan, Y.Y. Yan, W.L. Hergenrother and D.F. Lawson, Rubber Chem. Technol., 80, 194 (2007).
(15.) Y-Y. Yan, T.E. Hogan, W.L. Hergenrother and D.F. Lawson, Polymer Preprints, 47 (2), 467 (2006).
(16.) D.E. Sutton and J.A. Schwindeman in "Modern synthetic methods and novel structure," A.O. Paitil, D.N. Schulz and B.M Novak, Eds., ACS Symposium Series No. 174, 58 (1997).
(17.) D.N. Schulz and A.F. Halasa, J. Polym. Sci. Polymer Chemistry Edition, 15, 2,401-2,410 (1977).
(18.) Y. Miura, K. Hirota, H. Moto and B. Yamada, Macromolecules, 32, 8,356-8,362 (1999).
(19.) N. Hadjichristidis, S. Pispas and M. Pitsikalis, Prog. Polym. Sci., 24, 875-915 (1999).
(20.) R. Rauline (to Compagnie Generale des Etabissements Michelin-Michelin & Cie), U.S. 5,227,425 (July 13, 1993).
(21.) D. Derouet, S. Forgeard and J. C. Brosse, Macromolecular Chemistry and Physics, 199, 1,835 (1998).
(22.) D. Derouet, S. Forgeard and J. C. Brosse, Macromolecular Chemistry and Physics, 200, 10 (1999).
(23.) T.E. Long, L.W. Kelts, S.R. Turner, J.A. Wesson and T.H. Mourey, Macromolecules, 24, 1,431 (1991).
(24.) M. Ohata, M. Yamamoto, A. Takano and Y. Isono, J. Appl. Polymer Science, 59, 399 (1996).
(25.) C.C. Lin, T.E. Hogan and W.L. Hergenrother, Rubber Chem. Technol., 77, 90 (2004).
(26.) T.A. Antkowiak, D.F. Lawson, R. W. Koch and M.L. Stayer, Jr., U.S. Patent 5,153,159 (to Bridgestone/Firestone, Inc.), Oct. 6, 1992.
(27.) T.W Greene and P.G.M. Wuts, Protective Groups in Organic Synthesis, 3rd ed, John Wiley & Sons, New York, 1999.
(28.) K. Deka, R.J. Sarma and J.B. Baruah, Inorganic Chemistry Communications, 8, 1,082 (2005).
(29.) M Luechinger, R. Prins and G.D. Pirngruber, Microporous and Mesoporous Materials, 85, 111 (2005).
(30.) S.A. Torry, A. Campbell, A. Cunliffe and D.A. Tod, International Journal of Adhesion and Adhesives, 26, 40 (2006).
(31.) F.D. Osterholtz and E.R. Pohl, Journal of Adhesion Science and Technology, 6, 127 (1992).
(32.) C.J. Lin, W.L. Hergenrother and A.S. Hilton, Rubber Chem. Technol., 75, 215 (2002).
(33.) S. Futamura, Tire Science and Technology 18, 3 (1990).
(34.) S. Futamura, Rubber Chem. Technol., 64, 57 (1991).
(35.) S. Futamura, Rubber Chem. Technol., 69, 648-653 (1996).
(36.) A.R. Payne, J. Polym. Sci., 6, 57 (1962).
(37.) A. Scurati and C.J. Lin, Rubber Chem. Technol., 79, 170197 (2006).
by Terrence E. Hogan, Amy Randall, William L. Hergenrother and Chenchy J. Lin, Bridgestone Americas
Table 1--tread compound formulation Masterbatch Experimental polymer 75 (phr) Natural rubber 25 Silica 44.8 Carbon black 11.2 Oil 13.25 Wax, stearic acid, AO, silane 10.59 Total (phr) 179.84 Remill Masterbatch 179.84 Total (phr) 179.84 Final mix Remill 179.84 Sulfur 2.10 Curatives 3.90 Total (phr) 185.84 Table 2--Mooney viscosity change during accelerated polymer aging study (a) Time Control [SBR-Si(OEt).sub.3] [SBR-Si(OEt).sub.3] (days) SBR- treated with 2 eq. treated with [Si(OEt).sub.3] 2-ethylhexanol 0.75 eq. sorbitan monooleate 0 52.4 35.2 37.7 1 56.4 37.4 38.6 2 58.5 36.3 39.6 5 59.1 37.7 42.2 9 66.4 39.9 44.7 Time [SBR-Si(OEt).sub.3] (days) treated with 3.7 eq. sorbitan monooleate 0 30.5 1 29.6 2 30.4 5 31.0 9 33.2 (a.) Samples were aged at 55[degrees]C and 85% relative humidity. Table 3--comparison of tire properties for tires containing treads with and without alkoxysilane polymer Property Non- TEOS Polymer in tread functional terminated [ML.sub.1+4] at 130[degrees]C 26.6 56.7 tan [delta] 2% E, 0[degrees]C, 0.348 0.341 31.4 rad./sec. tan [delta] 7% E, 60[degrees]C, 0.138 0.122 31.4 rad./sec. [DELTA] G' (60[degrees]C) (MPa) 2.40 1.68 Tire testing data Tire rolling resistance Index (a) 100 93 Wet skid peak force index (b) 100 99 Dry skid peak force index (c) 100 102 Wear Index (average of main groves) 100 123 (a.) Lower values in tire rolling resistance index indicate improved fuel economy. (b.) Higher values in wet skid peak force index indicate better wet traction. Value was taken at 20 mph. (c.) Higher values in dry skid peak force index indicate better dry traction. Value was taken at 40 mph.
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Author: | Hogan, Terrence E.; Randall, Amy; Hergenrother, William L.; Lin, Chenchy J. |
|---|---|
| Publication: | Rubber World |
| Date: | Sep 1, 2010 |
| Words: | 3759 |
| Previous Article: | Process development of iron catalyzed 3,4 polyisoprene. |
| Next Article: | Milwaukee hosts Rubber Division. |
| Topics: | |