268
Macromolzcules 1988,21, 268-270
trochem. SOC. 1986,133, 117. (h) For a description of catalytic properties of n-PA, see: Soga, K; Ikeda, S. In Handbook of Conducting Polymers; Skotheim, T.,Ed.; Marcel Dekker: New York, 1986; Chapter 18, pp 661-672. (8) Such a distinction has been criticized by Pross who encourages the idea that, for example, a so-called (single) electron transfer and a (two electron) nucleophilic attack both actually involve single electron shifts (see: Pross, A. Acc. Chem. Res. 1985, 18, 212). For reasons of consistency with earlier literature,'OJ' we will continue to employ this distinction. (9) The charge carriers are believed to be delocalized carbanions (pairs of which are referred to as bipolarons) formed by intramolecular radical recombination of radical anions. See: Bredas, J. L.; Street, G. B. Acc. Chem. Res. 1985, 18, 309. (10) Garst, J. F. Acc. Chem. Res. 1971, 4 , 400. (11) (a) Garst, J. F.; Roberts, R. D.; Pacifici, J. A. J . Am. Chem. SOC. 1977. 99. 3528. (b) Svmons. M. C. R. J . Chem. Res.. Synop. 1978, 360. (12) The n-PA was DreDared bv immersion of PA films13ain 0.5 M sodium naphthalide in T H F for 24-48 h followed by washing with dry T H F until the washings were colorless. All operations were carried out in a glass apparatus13bwhich was previously flame dried on a vacuum line. The compositions of the "doped" films were determined by gravimetric analysis and were occasionally cross-checked by treatment with MeOH followed by titration of the supernatant with standardized HCl. (13) (a) The procedure for the synthesis of PA films can be found in: Chien, J. C. W. Polyacetylene: Chemistry, Physics and Material Science; Academic: New York, 1984; p 25-34. Occasionally, CzDz (from LizCz and DzO) was polymerized and the resulting deuteriated PA was used in our experiments in order to confirm the extent of alkyl (nondeuteriated) incorp ~ r a t i o n 'by ~ IR spectroscopy. (b) Reference 13a, p 327. (14) All reactions were carried out in a drybox. As an example, ca. 100 mg of n-PA film having a composition [CH(Na)o,no],(1.1 mmol charge carriers) in 5 mL of dry T H F in a screw-cap vial was treated with 0.165 mL (1.3 mmol) of 1-bromooentanefor 24 h. Both the resulting swellable film and soluble fraction contain n-pentyl groups as indicated by IR spectra of samples prepared from deuteriated PA. The amounts of substrate consumed and reduction products (Scheme I) formed were determined by GLC. The extents of alkylation were determined by elemental analysis (Schwartzkopf Laboratories, Woodside, NY). The intermediacy of radicals is inferred from the experimental data (see text). An experiment employing 6-bromo-1-hexene afforded PA containing terminal olefins (IR spectroscopy) and, as reduction products, 1-hexene and methylcyclopentane in a 5:l ratio. The formation of methylc clopentane suggests the intermediacy of 1-hexenyl radicals.' B The greater amount of 1-hexene is likely due to the high local concentration of reducing species in n-PA films in these heterogeneous reactions, and thus reduction of 1-hexenyl radicals competes with cyclization. In fact, Garst et a1.I6 found that 1-hexene is the major reduction product when 6-bromo-1-hexene is treated with excess sodium naphthalide radical anion. Garst, J. F.; Ayers, W. P.; Lamb, R. C. J. Am. Chem. SOC. 1966, 88, 4260. Group I alkyls (e.g., n-BuLi) are able to dope polyacetylene to n-PA. See: (a) Francois, B.; Mathis, C.; Nuffer, R.; Rudatikira, A. Mol. Cryst. Liq. Cryst. 1985,117, 113. (b) Elsenbaumer, R. L.; Miller, G. G.; Toth, J. E. U. S. Patent 4526708, July 2, 1985. This is not inconsistent with our contention that alkyl carbanions can be produced from n-PA by reduction of alkyl radicals since in our experiments the alkyl carbanion concentration is small relative to the concentration of n-PA carbanions. (The opposite situation exists in d ~ p i n g . " ~ , ~ ) (a) Lambert, F. L. J. Org. Chem. 1966,31,4184. (b) Lambert, F. L.; Ingall, G . B. Tetrahedron Lett. 1974,36, 3231. (c) Alkyl fluorides are expected to be even more difficult to reduce and in fact no reaction of n-PA with 1-fluoropentane occurred over a period of 3 days as indicated by the observation that the electrical conductivity of the n-PA decreased by only a factor of 2. On the contrary, tert-butyl chloride is easier to reducelab and consequently the extent of reaction with n-PA (Table I) is greater than that with 1-chloropentane. The reduction potential of n-PA varies from -2.8 to -2.05 V versus SCE for dopant compositions from ca. 10% to ca. 0.01 %, respectively. See: MacDiarmid, A. G.; Kaner, R. B. In Handbook of Conducting Polymers, Skotheim, T., Ed.; Marcel Dekker: New York, 1986; Chapter 20, p 696. (a) Chvalovsky, V.; Bellama, J. M. Carbon-Functional Organosilicon Compounds; Plenum: New York, 1984; pp 76-77. (b) The material balance in Table I is based on the amount of ,
I
I
tetramethylsilane, 1,2-bis(trimethylsilyl)ethane,and bis(trimethylsily1)methane detected by NMR. While these are also the major products in the reaction of CMTMS with sodium, lesser amounts of other products have been reported. See: Connolly, J. W.; Urry, G. J . Org. Chem. 1964, 29, 619. (c) Boak, D. S.; Gowenlock, B. G. J. Organomet. Chem. 1971,29, 385. (21) It is interesting to compare the reactions of n-PA and another heterogeneous reactant, C8K (K-intercalated graphite or n-type graphite); the amount of alkylation in the latter is not greater than 5%. See: Bergbreiter, D. E.; Killough, J. M. J . Am. Chem. SOC. 1978,100, 2126. (22) The compositions suggest that, on average, there are four C=C bonds in conjugation between the alkylated (presumably sp3) carbons. However, the alkylated films have a similar appearance to PA and the ,A, of an alkylated film (grown as a thin film on NaCl and then reduced and alkylated) is ca. 620 nm, a value quite inconsistent with tetraenes. These materials are presently being characterized in detail.
Amorphous Phase Segregation in Polyethylene-Based Ionomers Y. P. KHANNA* and W. M. W E N N E R Corporate Technology, Allied-Signal, Inc., Morristown, New Jersey 07960
L. KRUTZEL Engineered Materials Sector, A-C Polyethylene, Allied-Signal, Inc., Morristown, New Jersey 07960. Received March 31, 1987
Introduction "Ionomer" is a generic term for copolymers of an olefin (ethylene, butadiene, styrene, etc.) with a carboxylic monomer (acrylic acid, methacrylic acid, etc.) which have been neutralized, the H+ions being replaced by other cations (Na', Zn2+,etc.). Ionomer materials have been the subject of extensive studies over the last 2 decades. The morphology of ionomers is more or less understood a t present.1-3 Accordingly, the general understanding is that in a medium of low dielectric constant (e.g., hydrocarbon chains) the ions form aggregates. At lower ion contents, the ions aggregate to form multiplets, i.e., small aggregates (several ion pairs). At higher ion concentrations, many ions along with some nonionic material give rise to sizeable clusters which act not only as cross-links but more like microcrystallites. The literature4* reveals that the Tgof ionomers increases with an increase in cation content, presumably due to a fairly uniform distribution of aggregates or clusters. Also documented is the existence of a double Tg;one corresponding to the predominantly nonionic regions (i.e., unneutralized copolymer) and the other a t higher temperature representing the predominantly ionic microphase.' Studies on ethylene (E)/acrylic acid (AA) copolymer salts by Otocka and Kwei reveal a regular increase in the T,as a function of increasing ion content.' Contrary to the generally accepted increase in Tgwith ion content, a single reference has come to our attention in which MacKnight et a1.8 suggest that the &relaxation (or T,)of an ethylene/methacrylic acid copolymer decreases upon neutralization while producing, simultaneously, a higher temperature relaxation. In this note we present our independent results on the relaxation phenomenon in low molecular weight E/AA ionomers and support the observation of MacKnight et aLS that in ethylene ionomers the main Tgdecreases upon neutralization. 0 1988 American Chemical Society
Notes 269
Macromolecules, Vol. 21, No. 1, 1988 -5°C O"'0
-100
zn
O'C
-50
0
Temperature "C
--
50
Figure 1. Effect of percent neutralization(Zn2+)on the Tgregion of ethylene/acrylic acid copolymer revealed by dynamic storage modulus versus temperature data.
I
-50
I
0 Temperature 'C
-
I
50
Figure 2. Effect of percent neutralization (Zn2+)on the Tgregion of ethylene/acrylic acid copolymer revealed by dynamic loss modulus versus temperature data.
Experimental Section Materials. A low molecular weight (waxy) copolymer of ethylene and acrylic acid (93/7 mol %) was the base material selected for this study. This copolymer was partially to fully neutralized with cations according to the followingprocedure. The starting material was melted in the 90-100 "C region and mechanically agitated. A calculated amount of the cation oxide, e.g., ZnO, was added to the molten material so as to achieve the desired degree of neutralization. The temperature was raised to 150-220 "C and the reaction stopped in about 30-60 min. Dynamic Mechanical Analysis (DMA). The resulting samples were compression molded into plaques of about 0.05-in. thickness. Thin strips were cut for dynamic mechanical analysis. A Polymer Laboratories DMTA unit was used in obtaining the mechanical spectra. Samples were analyzed in the bending mode of deformation at a heating rate of 3 "C/min from -135 to about +50 OC under an argon atmosphere. A constant frequency of 1 Hz and a constant strain level were used. Results and Discussion Our recent studies on the dynamic mechanical properties of polyethyleneg and polyethylene waxeslO have indicated that the p-relaxation at about -30 "C is the Tgof polyethylene. [Note: Although this is in agreement with the claims by Boyd1' and Wunderlich,12 we do not wish to imply that the subject of polyethylene Tgis not controversial. Some groups of workers13J4 believe that the Tgof polyethylene is about -125 "C. While the solution to this controversy is elusive, our own resultsgJOpoint out that the 0-relaxation at about -30 "C has its origin in the amorphous phase and we will use this basis in this paper.] Dynamic storage modulus (E? versus temperature data reveal a Tgof about 0 "C for the E/AA copolymer used in this study, the higher Tgof the copolymer being attributed to the acrylic acid. [Note: The samples had no significant moisture content, e.g., 5 0.1%, as shown by TGA.] Figures 1-3 shows the effect of percent neutralization (Zn2+)on the Tgregion of an E/AA copolymer wax. It is clearly shown that the Tgof the starting material begins to split into a doublet upon neutralization; the lower Tg progressively decreases while the upper one progressively increases as the E/AA copolymer is neutralized. Inter-
I
-1 00
._ 50
0
100
% Neutralization (Zn")
Figure 3. Splitting of a single Tgof ethylene/acrylic acid copolymer into a doublet upon neutralization. Table I Effect of Percent Neutralization on the T,of Ethylene/Acrylic Acid Copolymer % cation concn p-relaxation T,,"C in ionomer E' basis E" basis Zn
0 25 50 100
Ca
9.5 25 50 100
-0 -16, 35 -15, 45 -25, 43 -30, 48 -32,49 -10, 25 -16, 44 -24, 42 -32, 44
-5 -16, 33 -13, 46 -18,40 -21, 50 -24, 50 -3, 13 -9, 38 -15, 47 -22,60
estingly, the lower Tg of the 100% neutralized copolymer falls in the range for polyethylene itself. Similar observations are made with other cations, e.g., Ca2+,Mg2+,and Na+ (Tables I and 11). We explain these observations as follows. The starting E/AA copolymer contains a crystalline phase of polyethylene (X-ray crystallinity = 19%, DSC, T, = 85 "C)
270
Macromolecules 1988, 21, 270-273 Table I1 Effect of Cation on the T,of Partially Neutralized Ethylene/Acrylic Acid CoDolsmer 70 cation "C concn in 6-relaxation Tg, ionomer E' basis E " basis 0 -0 -5 Ca -24, 42 -15,47 Mg -30, 45 -20, 27 50 -15, 45 -13, 46 Zn -25, 43 -18, 40 Na -10,39 -31, 35
and an amorphous phase of randomly mixed E and AA segments (DMA, Tg= 0 "C). Upon partial neutralization, part of the AA moieties are converted into ionic salts and these form aggregates. Depletion of free AA segments from the initial amorphous phase produces a lower Tgwhile simultaneously giving rise to a higher Tgdue to the phase containing ionic aggregates. In the fully neutralized E/AA copolymer, the amorphous phase has no free AA units, and thus, a Tgresults corresponding to the ethylene amorphous phase (Tg -30 "C). Simultaneously, however, another phase develops which is rich in ionic clusters, and this accounts for the higher Tg. Finally, we wish to comment that we have only dealt with a low molecular weight (waxy) material in this work and the relaxation patterns may be quite different in a similar copolymer of high molecular weight.lS
References and Notes Eisenberg, A. Pure Appl. Chem. 1976, 46, 171. Marx, C. L.; Caulfield, D. F.; Cooper, S. L. Polym. Prep. (Am. Chem. SOC.,Diu. Polym. Chem.) 1973, 24 (2), 890. Mattera, V. D.; Risen, W. M. J . Polym. Sci., Polym. Phys. Ed. 1986, 24, 753. Williams, M. W. Polym. Prep. (Am. Chem. SOC., Div. Polym. Chem.) 1973, 14 (2). 896. Matsuura, H.; Eisenberg, A. J . Polym. Sci., Polym. Phys. Ed. 1976, 14, 1201. Rahrig, D.; MacKnight, W. J. A C S Symp. Ser. 1980, 187, 77. MacKnight, W. J.; Earnest, T. R. J . Polym. Sci., Macromol. Rev. 1981, 16, 41. MacKnight, W. J.; McKenna, L. W.; Read, B. E. J . Appl. Phys. 1967,38,4208. Khanna, Y. P.; Turi, E. A.; Taylor, T. J.; Vickroy, V. V.; Abbott, R. F. Macrmolecules 1985, 28 (6), 1302. Khanna, Y. P.; Wenner, W. M.; Krutzel, L., submitted for publication in Macromolecules. Boyd, R. H. Macromolecules 1984,17, 903. Wunderlich, B. In Thermal Characterization of Polymeric Materials; Turi, E., Ed.; Academic: New York, 1981; Chapter 2, p 135. Beatty, C. L.; Karasz, F. E. J. Macromol. Si-Reu. Macromol. Chem. 1979. C27 (1).37. Axelson, D. 'E.; Mandelkern, L. J . Polym. Sci., Polym. Phys. Ed. 1978, 16, 1135. Eisenberg, A,; Navratil, M. Macromolecules 1973, 6 (4), 604.
Spectroscopic Analysis of Phase-Separation Kinetics in Model Polyurethanes HAN SUP LEE, YING KANG WANG,' WILLIAM J O H N MACKNIGHT, and SHAW LING HSU* Polymer Science a n d Engineering Department, The University of Massachusetts, Amherst, Massachusetts 01003. Received May 26, 1987, Revised Manuscript Received August 28, 1987
Introduction Because of their incompatible structural components, segmented polyurethanes mainly exist as phase-separated
* To whom
correspondence should be sent. 'Permanent address: Department of Chemistry, Beijing University, Peiking, People's Republic of China.
0024-9297/88/2221-0270$01.50/0
systems. The degree of phase separation as reflected in the size and perfection of the domains is the most important criterion for the determination of overall sample mechanical proper tie^.'-^ Although phase-separation kinetics of polymer mixtures have been active areas of few investigations have been carried out to characterize the phase-separation kinetics of polyurethane block copolymer^.^^^^^ Not only is the characterization of the phase-separated structure present in segmented polyurethanes important from a practical standpoint, but also the mechanism associated with the phase separation is extremely interesting from a fundamental viewpoint as well. The connectivity between the components, the effects of chain stiffness, and the polydispersity of the hard or soft segments are all important parameters and need to be examined in greater detail in order to understand more fully the phase-separation mechanism. In this study we have characterized the phase-separation behavior of well-defined polyurethanes consisting of the reaction product of methylenebisb-phenyl isocyanate) (MDI) and butanediol (BD) as the hard segments and poly(propy1ene oxide) (PPO) as the soft segment. The success in utilizing vibrational spectroscopy for the study of phase separation depends on the existence of bands sensitive to mixed and phase-separated states. In a previous study,g we assigned spectroscopic features characteristic of urethane linkages dispersed in the soft segments as compared to interurethane hydrogen bonding confined to hard-segment domains. With a specially designed sample cell, we were able to trap a phase-mixed structure at a temperature 60 "C below the glass transition temperature of the soft segments. When this quenched sample was brought up to a higher temperature, the increase of one spectroscopic component and the corresponding decrease of the other provided a direct measurement of phase-separation kinetics. In this paper, we report the initial findings associated with a well-defined model polyurethane. The phase separation behavior has been interpreted to proceed by a nucleation and growth mechanism with a two-dimensional growth initially, followed by a diffusion-dominated mechanism. Additional studies for the entire family of model polyurethanes will be published separately.
Experimental Section The model polyurethanes containing monodisperse hard and soft segments used in this study have been characterized extensively.*O The sample used in this study contains three MDI units per hard segment. All the infrared spectra were obtained with an IBM Model 98 vacuum Fourier transform infrared spectrometer. We built a "purge box" to allow ease in sample access by keeping the sampling area a t atmospheric pressure and constantly purged it with boiled-off nitrogen gas. Polyurethane samples used for this infrared study were dissolved in THF (2% w/v) and then casted onto AgCl windows. After drying overnight at atmospheric pre-sure, the casted films were then dried in a vacuum oven for 24 h a t room temperature. Finally the films were annealed a t 60 OC for an additional 24 h. As described earlier,g one of the most critical components in this experiment is our ability to cool the sample very rapidly. The sample in the beam area can be cooled from the molten phase (-130 "C) to -100 OC in less than 30 s. As before, we have assumed that the sample quenched from the melt is essentially phase mixed.g To ensure as much phase mixing as possible, the polyurethane sample was kept a t 130 OC for approximately 10 min. When the sample temperature is raised above the soft the basically incompatible hard and soft segments segment Tg, phase ~ e p a r a t e .Spectra ~ in the 1700-cm-' region, obtained a t various annealing times at 10 OC,are shown in Figure 1. Spectral resolution is maintained a t 2 cm-' for all studies. Although there are multiple bands of interest in this region, we have not carried
0 1988 American Chemical Society
