Crosslinking structure and properties of poly(urethane-urea) elastomers.
Poly(tetra-oxymethylene glycol)di-isocyanates were used as prepolymers. Both polymers have isocyanate groups at the telechelic position of the poly(tetra-oxymethylene glycol) chains. A methylene-ortho- chloro-aniline (hereafter MOCA, having two amine groups) was employed as a curing agent for these prepolymers.
The prepolymer was degassed in a flask, stirred at 80[degrees]C under a reduced pressure for ten minutes and precisely weighed MOCA was added. The two materials were thoroughly mixed and degassed again. This mixture was then poured into a mold to form a sheet (2 mm thick, 150 mm by 150mm square) and covered with a glass plate. The mold was cured in an oven at 100[degrees]C for three hours to complete the reaction between the isocyanate and the amine groups. The ratio of amine to isocyanate (hereafter A/I ratio) varied from 0.82 to 1.1. It should be noted that the chemical bonds exist below a A/I ratio of 1.0 when the completion of reaction is assumed. After curing, the sheet was stored in a desiccator for a week before testing.
Characterization of crosslink structure
A 2mm cubic sample was cut from the sheet and its compressive strain and stress properties were measured in two solvents, chloroform and tetrahydrofuran. The instrument employed in this measurement is illustrated in figure 2. The solubility parameters of chloroform and tetrahydrofuran are 9.3 and 9.35, respectively (ref. 7). Though the difference in solubility parameter for both the solvents is small, chloroform dissolves only the soft segment of a urethane elastomer, poly(tetra-oxymethylene glycol) chains, while tetrahydrofuran dissolves both the soft and the hard segments. In other words, chloroform remains both chemical and physical bondings while tetrahydrofuran leaves chemical bonds. This is attributed to the difference in polarity of the two solvents.
Rubber test samples were evaluated after reaching an equilibrium in solvent, step by step, adding weight to the weight tray of the test instrument (figure 2). The resulting compressive strain of the swollen rubber was measured using a linear transducer The network chain density of the rubber was calculated using equation 1 (ref. 6).
(1) v=[Vr.sup.1/3] F/RT(a-[a.sup.2])Ao where:
v: network chain density Vr: volume fraction of rubber in a swollen gel
F: applied weight
R: gas constant
T: absolute temperature
[alpha]: compression ratio
Ao: cross sectional area of a sample before swelling
When using chloroform as a solvent, the network density obtained by the equation 1 contains both the physical crosslinks and the chemical crosslinks (hereafter defined as [v.sub.c]). On the other hand, only the chemical network density(hereafter defined as [v.sub.t]) is obtained by using tetrahydrofuran as a solvent. Accordingly, the physical crosslinks are obtained as the difference between the [v.sub.c] and the [v.sub.t] .
The stress strain properties and tear strength were measured using a tensile tester at a cross head speed of 50 mm/mint A test piece of 1 cm by 10 cm was cut from the sheet and a 2 mm cut was made at the edge in the center of the test piece for the tear strength measurement. JIS 3 dumbbell type test pieces were used for the tensile measurement. A creep test was carried out at 80[degrees]C. Test pieces of 1 cm by 10 cm were used in this measurement. A dynamic fatigue tester was used at 500 cycles per minute with a strain of 10% to evaluate the fatigue life of the elastomer. And another dynamic fatigue test using urethane tubings was done according to JIS L 1017-1983, originally designed for fiber testing. Urethane tubings of o24 cm long, 2.7 cm outer diameter and 1.9 cm inner diameter respectively, were prepared. Then the inside of the tubings was inflated with air at a pressure of 0.34 MPa and rotated the samples at 1,000 rpm. The tubings were bent at an angle of 40[degrees] to give both compressive and tensile strains reciprocally to the samples.
The effect of heat treatment on the urethane was evaluated. For this test, urethanes with 0.82 A/I ratio were employed. The samples were treated in an oven at 100-150[degrees]C for up to 50 hours before testing.
Results and discussion
Network chain density
Tetrahydrofuran and chloroform have almost the same solubility parameter, yet the swelling characteristics of urethane to each solvent is quite different (figure 3). The relationship between the A/I ratio and [v.sub.c] and [v.sub.t] is shown in figure 4. From this figure, it is seen that [v.sub.t] goes to almost zero at an A/I ratio of 1.0 suggesting that the [v.sub.t] represents the network chain density based on only the chemical bonds. When one subtracts the [v.sub.t] from the [v.sub.c], it is observed that the difference which denotes physical bondings (hereafter [v.sub.p]) is almost independent of the A/I ratio up to 1.0. At a ratio of 1.1, the [v.sub.c] (=[v.sub.p]) slightly reduced. This is possibly attributed to a kind of plasticizer effect of the excess amine which affects the hard segments of the urethane elastomer.
Table 1 shows the relationship between the A/I ratio and the molecular weight of the chain between crosslink points (hereafter Mc). It should be noted that the Mc, calculated from the [v.sub.t], must theoretically be infinity at the A/I ratio of 1.0, however, the Mc observed vas 5,100 at that point. This may be attributed to the technical difficulty of precise determination of the number of isocyanate groups of a prepolymer, incompleteness of the reaction or an error in measuring the raw materials prior to reaction.
Mechanical properties and crosslink structure
The mechanical properties of urethane elastomers depend on the complex crosslink structure and a function of the A/I ratio. Figure 5 shows the stress strain properties of urethane elastomers in relation to their A/I ratio. This shows a strong relation between the strain at failure of an elastomer and an increase in the A/I ratio. Similarly, the energy of tearing, the elasticity at 10% strain, the fatigue resistance, and the creep of the urethane elastomers as tabulated in table 2 show a strong dependence on the A/I ratios. The tearing energy, fatigue resistance and creep deformation increase with the increase of the A/I ratio. It should be pointed out that the total crosslink density ([v.sub.t] +[v.sub.p] = [v.sub.c] ) decreases as A/I ratio increases (figure 4). It is well known that the similar phenomena were observed for sulfur cured elastomers (ref. 11) and almost the same elucidation may be applied to the relationship between the above properties and the crosslink chain density. It was found that a plot of the elongation (strain at failure divided by the original length) defined as ab, the plots of ab versus [v.sub.c], is a linear relationship (figure 6) with a slope of one-half This result can be explained by a theory proposed by Smith (ref. 8). Smith found that when folded polymer chains between crosslink points extend to the limit of their length at failure, the following equation results.
[[alpha].sub.b] =k [[v.sub.c].sup.1/2] where k is a constant.
Figure 7 shows the change of the [v.sub.p] and the [v.sub.t] with strain. The [v.sub.p] decreases as strain increases while [v.sub.t] remains constant. This result implies that the hard segments which form the physical crosslink points are gradually destroyed when the polymer chains are extended. However, the chemical bonds are not destroyed until the polymer chains are broken.
To know the change in crosslink structure of urethane by mechanical fatigue, the tube test, JIS L1017-1983 was carried out as described earlier. The maximum temperature observed during the test was about 110[degrees]C for all samples The change of crosslink structure before and after the test is shown in figure 8. Decrease of crosslink density after the fatigue test is observed for all samples. However, this is more remarkable for the samples at lower A/I ratio and the decrease of the [v.sub.t] contributes a great deal whereas the [v.sub.p], physical crosslink density, shows small change.
The change of the [v.sub.t] versus time at 100-150[degrees]C is shown in figure 9. It is found that the change of the [v.sub.t] in this figure is expressed by a first order reaction. When an Ahrenius type reaction is assumed, the rate constant k is expressed by the following equation.
(3) k = Aexp(-E/RT) where:
A: frequency factor
E: activation energy
R: gas constant
T: absolute temperature
From the k versus 1/T plot, E was found to be 28.5 kcal/mole. This E value coincides with the value obtained by Tobolsky for biuret bonding (ref. 8).
Figure 10 presents the mutual relationships among temperature. length of heat treatment time, and the v/v0. When two parameters out of the three are known, the remaining parameter can be determined from the figure. Using this figure, one can estimate the temperature or time when a urethane sample is treated.
Application to a practical problem
* Service life of urethane sieve in use. Polyurethane elastomers have excellent abrasion resistance and are used in a quarry as sieves to classify macadam. Occasionally, the service life of a sieve was much shorter than expected, for unknown reason. Figure 11 shows the crosslink structure of urethane sieves used for one year under a normal condition. Small difference can be seen between the original crosslink structure and the sieve. On the other hand, a sample which cracked in use shows no chemical crosslinks (also shown in figure 11). Since the decrease of chemical crosslinks in a urethane elastomer is caused by heat and moisture, it is strongly suggested that this sample was used excessively in hot and humid conditions.
* Urethane sieve stored. Urethane sieves stored in a warehouse (sample A) for five years showed cracks. Whereas others stocked in another warehouse (sample B) did not. The warehouse for sample A (hereafter house C) is a simple house which allowed moisture and air penetration. Contrary, the warehouse for sample B was air conditioned. Change in crosslink density for both sample A and B was analyzed. Test pieces were cut from the surface of samples, 100p thick, step by step, to a depth of 1 mm. Since the thickness of each sample is too thin to apply the compressive swollen method, the Flory-Huggins equation (ref. 10) was used to obtain crosslink densities. The interaction parameter between solvent (tetrahydrofuran) and urethane was determined by using the compressive swollen method as 0.365 at 20[degrees]C, beforehand. Figure 12 shows a depth profile of crosslink density from the surface to 1 mm. Though both sample A and B show decrease in [v.sub.t], comparing with that of the original product, the decrease is dominant for sample A, especially in the surface region. This suggests that the cracks in sample A were caused by moisture and hot air in house C in summer time.
The crosslink structure of poly(urethane-urea) elastomers was studied by analyzing the compressive stress-strain properties of the elastomers under an equilibrium swollen state in two solvents. The crosslink chain density measured in chloroform; [v.sub.c] represents both the chemical and physical crosslinks. On the other hand, the crosslink chain density measured in tetrahydrofuran; [v.sub.t] represents only the chemical crosslinks. when the elongation (strain at failure divided by the original length of a sample) defined as [[alpha].sub.b], was plotted versus [v.sub.c], a linear relationship with a slope of one-half resulted This result can be explained by the theory proposed by Smith (ref. 8). It was also observed that [v.sub.p] decreases as strain increases while [v.sub.t] remains constant. This result implies that the hard segments which form the crosslink points were gradually destroyed when the polymer chains were extended. However. the chemical bonds are not destroyed on stretching until the polymer chains are broken. It is found that the change of [v.sub.t] in heat treatment can be expressed as a first order reaction. The activation energy E was found to be 28.5 Kcal/mole and this value coincides with the value obtained by Tobolsky for biuret bonding (ref. 9). The mutual relationships among temperature, time of heat treatment, and the network chain density before and after heat treatment; v/[v.sub.o] were developed. When two parameters out of the three are known, the third parameter can be determined. These relationships are useful to estimate the time or temperature history of a sample. The analysis of crosslink structure of poly(urethaneurea) elastomers was successfully applied to solve practical industrial problems.
[Figures 1 to 12 ILLUSTRATION OMITTED]
(1.) R. Bonart, L. Morbitzer and G. Hentze, J. Macromol. Sci., B3 (2) 337 (1969). (2.) I. Kimura, H. Ishihara, H. Ono, N. Yoshihara, S. Nomura and H. Kawai, Macromolecules 7, 335 (1974). (3.) G.M. Estes, R.W. Seymour and S.L. Cooper: ibid, 4,452 (1971). (4.) L.M. Leung and J. T. Koberstein, J. of Polymer Sci., Polymer Phys. Ed., 23, 1883 (1985). (5.) Polyurethane Handbook, K. Iwata Ed., Ch. 2, Nikkan Kogyo Press (1987) (6.) P.A. Small, J. Appl. Chem., 3, 71 (1953). (7.) M. Yamamoto: Properties and processing of rubbers, Chijin Shokan, Tokyo (1965). (8.) T.L. Smith and A.B. Yagnussun: Rubber Chemistry and TechnoL, 35, 753 (1962). (9.) P.C. Colodmy and A.V. Tobolsky: J. Am. Chem. Soc., 71, 4320 (1957). (10.) P. J. Flory: Principle of polymer chemistry, Cornell Univ. Press, New York (1954). (11.) A.Y. Coran: Science and technology of rubber, Chapter 7, Academic Press, (1978).
|Printer friendly Cite/link Email Feedback|
|Date:||Apr 1, 1996|
|Previous Article:||A new era for MDI: flexible PU slabstock foam.|
|Next Article:||Rubber-to-metal bonding agents.|