Unfilled Vulcanizates Research Articles

PMR study of the effect of active fillers and vulcanization on macromolecular mobility of isoprene rubber

Studied were made of the temperature dependences of line width, second moment and the form of PMR line in the temperature range of −190+60° using isoprene rubber, crude and vulcanized systems (unfilled and containing carbon black, an active technical carbon). It was shown that vulcanization displaces transition from the glassy state to the high elastic state by 10° in the direction of higher temperatures. Crude carbon produces oriented ranges, in which macromolecular mobility is higher than in inhibited regions of unfilled vulcanizates. A narrow component is observed in spectra of samples containing untreated carbon, which is due to groups containing proton on the surface of untreated carbon particles.

The viscoelastic properties and the scatter of elasticity constants of vulcanizates

Determination of the viscoelastic properties on unfilled vulcanizates with known molecular network parameters has shown a well defined correlation between the scatter of elasticity constants and the variance of relaxation period spectra. These correlations indicate that it is possible to assess the scatter of the viscoelastic properties on the basis of existing combinations of ordered and chaotic zones in the cured elastomers.

Stress Relaxation Mechanisms in Rubbers Reinforced with Carbon Blacks

Abstract 1. In vulcanized rubbers containing blacks a multi-stage mechanism for stress relaxation was observed. It was discovered that the stress relaxation process consists of five fundamental processes: the first three relaxation processes, related to the slow stages of physical relaxation within the bulk of the rubber, have no connection with the fillers (“soft” domains); the fourth process has to do with the relaxation in the black-rubber domain; the fifth process involves the chemical relaxation of vulcanizates. 2. The fundamental mechanisms of the first 3 relaxation processes in the soft domains have the same activation energy values and the same segmental mechanism as the rearranged domains found in supermolecular weight structures, which are also present in unfilled vulcanizates. 3. In the investigated stress range of up to 200% elongation, the activation energy for the first 3 relaxation processes in the soft domains of filled vulcanizates is not a function of the deformation strain, whereas the activation energy of the fourth relaxation process in the black-rubber domains of filled rubbers is a function of the deformation and of the filler content. For these reasons, rubber loaded with carbon blacks, in contrast to unfilled rubbers, possess the typical nonlinearity of viscoelastic materials. 4. The activation energies of the relaxation processes in the black-rubber domains decrease in a linear fashion with the value for the initial tensile stress in filled vulcanizates, and decrease in like manner for vulcanizates containing different proportions of fillers. The kinetic units, determined from the activation energies of these processes, appeared to be segments of chains with activation energies of up to 40% more than the activation energies of the physical relaxation processes in the soft domains. The other kinetic units of the processes proved to be black particles, the dimensions of which were calculated from the values for the coefficients in the formula for relaxation time.

Creep and Stress Relaxation of Natural Rubber Vulcanizates. Part I. Effect of Crosslink Density on the Rate of Creep in Different Vulcanizing Systems

Abstract The physical creep of unfilled natural rubber vulcanizates, prepared with different vulcanizing systems, has been studied. For each of the three vulcanizing systems chosen there is a strong dependence of creep rate on crosslink density, but the rates for accelerated sulfur vulcanizates are two or three times higher than those of peroxide vulcanizates of similar crosslink density. Supplementary experiments, in which the crosslink structure of sulfur vulcanizates is modified either by chemical treatment or by variations in the vulcanizing conditions, show that the nature of the crosslink itself is not a determining factor in the type of vulcanizate. Other features, such as the type and quantity of extranetwork material arising from the vulcanizing process, contribute significantly to the viscoelastic behavior of accelerated sulfur vulcanizates.

Low Strain Dynamic Properties of Filled Rubbers

Abstract Carbon black does not exist as single spherical particles but forms itself into a rodlike primary structure. These rodlike structures then form into an aggregated secondary network. This secondary network is believed to be held together by Van der Waals-London attraction forces. The decrease in shear modulus of filled rubber vulcanizates with strain is due almost certainly to these secondary forces. Special mixing techniques such as attrition of the carbon black, increased time of mixing, or the addition of chemical promoters which aim at dispersing the carbon black within the mix better are shown to decrease the value of G′0−G′∞. The absence of any modulus change with strain for unfilled vulcanizates and secondly the little change observed in values of G′0−G′∞ with increasing vulcanization of the rubber when containing the same amount of carbon black confirms that the decrease in modulus with strain amplitude is in no way associated with the gum phase of the filled vulcanizate. The similarity in behavior of carbon black filled rubbers with clay/water and clay/rubber systems indicates that the decrease in modulus with amplitude is due to the breakdown of the three dimensional filler aggregates. A number of rheological studies on clay systems has confirmed that clay particles form into rigid three dimensional structures when dispersed in a medium. Evidence for the aggregated filler structure to be held together by Van der Waals-London attraction forces comes from the reasonable agreement between the experimental values for the forces required to breakdown the carbon black aggregates in paraffin oil and the forces calculated from Van den Tempel's model for flocculated solid particles in a liquid. The successful application of a domain model to the hysteretical behavior exhibited by carbon black filled vulcanizates at low strains indicates that the carbon black structure breaks down under stress but reforms to the original state when the stress is removed. This conclusion is also supported by the similarity in behavior between filled rubbers and a dendritic crystal structure of PBNA in rubber. Under the optical microscope the PBNA is seen to break down and reform under a stress-strain cycle. The breakdown and reformation of this secondary aggregated carbon black structure increases the hysteresis in filled rubber vulcanizates. Other sources of hysteresis include viscoelasticity of the polymer, crystallization, stress-softening, and changes in network structure (e.g., breakage of weak crosslinks). These mechanisms have been discussed in depth in previous publications. Recent work has shown, however, that the strength of a rubber is dependent on the combined effect of the different hysteretical mechanisms. The breakdown and reformation of the carbon black structure at low strains in filler reinforced rubbers therefore not only affects the heat build up, transmissibility, and fatigue behavior but also influences the failure properties of the filled vulcanizate.

Effect of Finite Extensibility on the Viscoelastic Properties of a Styrene-Butadiene Rubber Vulcanizate in Simple Tensile Deformations up to Rupture

Abstract Stress-strain and rupture data were determined on an unfilled styrene-butadiene vulcanizate at temperatures from −45 to 35° C and at extension rates from 0.0096 to 9.6 min−1. The data were represented by four functions: (1) the well-known temperature function (shift factor) aT; (2) the constant-strain-rate modulus, F (t, T) reduced to temperature T0 and time t/aT, i.e., T0F (t/aT)/T (3) the time-dependent maximum extensibility λm (t/aT); and (4) a function Ω(χ) where χ=(λ−1)λm0/λm, in which λ is the extension ratio and λm0 is the maximum extensibility under equilibrium conditions. The constant-strain-rate modulus characterizes the stress-time response to a constant extension rate at small strains, within the range of linear response; λm is a material parameter needed to represent the response at large λ; and Ω(χ) represents the stress-strain curve of the material in a reference state of unit modulus and λm=λm0. The shift factor aT was found to be sensibly independent of extension. At all values of t/aT for which the maximum extensibility is time-independent, the relaxation rate was also found to be independent of λ. These observations indicate that the monomeric friction coefficient is strain-independent over the ranges of T and λ covered in the present study. It was found that λm0=8.6 and that the largest extension ratio at break (λb)max is 7.3. Thus, rupture always occurs before the network is fully extended.

Developments in self‐reinforced elastomers

AbstractUntil recently it has been believed that either stress‐crystallization or the presence of reinforcing filler is necessary for the development of high strength in elastomers. Now it has been observed that gum vulcanizates of isoprene/acrylonitrile copolymers have tensile strengths in the order of 3000–4000 psi in the absence of fillers. In this case crystallization does not occur and the strength is associated with the ability of the polymer molecules to become highly oriented on stretching. Similar observations have been made with ABA block copolymers polystyrene‐polybutadiene‐polystyrene which are highly elastomeric in the absence of chemical crosslinks. The elastomeric properties and high strength of the block copolymers are attributed to the mutual incompatibility of the A and B blocks which leads to phase separation on a microscopic scale. The effective network structure can be temporarily destroyed by heating the polymer above the softening temperature of the resinous blocks which makes the material readily processable. The initial decrease in retractive force with increasing temperature up to 60–70°C is attributed to an increase of mobility of the resinous domains in the rubbery matrix. At higher temperatures the crosslinking effectiveness decreases markedly due to softening of the resinous regions. The process of yielding, hysteresis effects, and delayed elastic creep arise because of viscous forces which oppose the motions of the colloidal polystyrene domains. The strength and effectiveness of physical crosslinks in block copolymers shows a similar temperature dependence as has been observed for other filled and unfilled vulcanizates. The process of self‐reinforcement in isoprene‐acrylonitrile copolymers and polystyrene‐polybutadiene‐polystyrene block copolymers is attributed to viscous effects which make possible the alignment of a large number of polymer chains in the stretching direction.

Large Deformation Tensile Properties of Elastomers. I. Temperature Dependence of C1 and C2 in the Mooney-Rivlin Equation

Abstract Tensile stress strain data were determined at various extension rates and temperatures on unfilled butyl and silicone vulcanizates and on six hydrofluorocarbon (Viton A-HV) vulcanizates that contain between 0.90 and 12.2×10−5 mole of effective network chains per unit volume of gel rubber. From one min isochronal data, values of C1 and C2 in the Mooney-Rivlin equation were derived. For the butyl and silicone vulcanizates from, respectively, −20 to 150° C and —45 to 200° C, C1 increases with temperature at a rate which is in reasonable agreement with published data from force-temperature measurements. Because C2 is sensibly temperature independent over these extended temperature ranges, it is concluded that C2 is a finite quantity under equilibrium conditions. Time dependent behavior was observed on the Viton A-HV vulcanizates between —5 and 230° C. For each vulcanizate between about 25 and 230° C, C1273/T is temperature independent, but it increases with decreasing temperature below about 25° C. Except at the lowest temperatures, C2273/T decreases with increasing-temperature, the rate of decrease becoming progressively less with an increase in crosslink density. Above 25° C the ratio C2/C1, which ranges from about 90 to 0.6, decreases with an increase in either temperature or crosslink density. The time-dependent behavior of C2273/T appears to arise primarily from molecular processes whose relaxation times are considerably longer than the terminal relaxation time of individual chains.

Anomalous freezing point depression of swollen gels

AbstractThe anomalous freezing point depression, ΔT, of benzene‐swollen vulcanizates has previously been attributed to the limitation of (benzene) crystallite size by the polymer network. This study was initiated to determine the benzene crystallite size in a number of rubber and benzene systems. A special low‐temperature specimen holder was designed and constructed in the Cambridge Laboratories for running diffraction patterns at temperatures near −30°C. X‐ray line broadening techniques were used to study a series of filled and unfilled vulcanizates of varying crosslink density. The results indicate that crystallite size is not depressed to the degree predicted by freezing‐point measurements. Benzene crystallite sizes were similar in all rubber benzene systems, regardless of degree of crosslinking or benzene fraction, although carbon black loading appears to increase crystallite size. This effect may be attributed to lesser depth of penetration of the x‐rays due to greater density as carbon black loading increases. Additional studies measuring the ΔT for solutions and similar vulcanizates of NR and SBR over a wide range of rubber concentrations showed that at the same rubber in benzene fraction, crosslinking increases ΔT but the addition of carbon black reduces ΔT. An explanation for the observed phenomena is advanced.

Dynamic behavior of natural rubber during large extensions

AbstractThe stresses and energy losses during simple extension cycles up to a maximum elongation of 530% have been determined for an unfilled vulcanizate of natural rubber as a function of the temperature and extension rate. At sufficiently short elongation times and low temperatures, the rate and temperature dependence of the ascending stresses are connected by the Ferry transform, and the superposition principle can be applied to them. Outside this experimental range, the stresses are increased by crystallization. The validity of the Ferry transform for the energy losses and the energy loss ratio is more restricted than for the stresses, and the losses are always higher than can be expected from a purely viscoelastic mechanism. The additional losses are tentatively ascribed to incipient crystallization and stress‐softening effects. At short elongation times and low temperatures, the losses approach the values predicted by viscoelasticity, and the loss ratio becomes independent of the maximum extension of the strain cycle.

Structural Characterization of Filled Vulcanizates Part 1. Determination of the Concentration of Chemical Crosslinks in Natural Rubber Vulcanizates Containing High Abrasion Furnace Black

Abstract The degree to which HAF black restricts the swelling of natural rubber vulcanizates in n-decane has been determined using a vulcanizing system in which the stoichiometry of crosslinking is unaffected by the carbon black. The dependence of the degree of restriction, as measured by the ratio of the volume fractions of rubber in the filled and unfilled vulcanizates swollen to equilibrium, on the concentration of carbon black follows an exponential relationship previously proposed by Lorenz and Parks. This is found to be equivalent to a simple linear relationship between the apparent and actual crosslink concentrations: napparent/nactual=1+Kϕ, where K is a constant characteristic of the filler and φ is its volume fraction in the vulcanizate. The relation has been used to determine actual crosslink concentrations in filled natural rubber vulcanizates. HAF black is found to cause increases of up to 25 per cent in the yield of polymer to polymer crosslinks in conventional sulfur vulcanizing systems, accompanied by changes in rate of cure and of crosslink reversion. All these are small compared with the effect of the filler on many physical properties.

Mixing of carbon black and polymer: Interaction and reinforcement

AbstractInteraction between carbon black and polymer starts during the mixing process; a primary agglomerate is formed, the composition of which is dependent upon the structure. The important carbon black properties are surface area, specific activity, structure (void, volume, anisometry), and porosity of the particles. On heat treatment of black at 3000°C., it loses its sites of high specific activity. Structure and specific activity determine incorporation time and further dispersion. During mixing, bound rubber is formed which is used as a measure of specific surface activity. In the final vulcanizate, the filler–polymer interaction is evident in reduced swelling in solvents (benzene, chloroform, cyclohexane,) etc. Below a certain critical degree of swelling, the percentage swelling is no longer dependent upon the amount of filler in the vulcanizate. Graphitized black vulcanizates exhibit in all solvents the same degree of swelling as the unfilled vulcanizate. This phenomenon is explained by the assumption of mobile adsorption of rubber chains on the carbon black surface. In untreated blacks the mobility on the surface is limited by sites of high energy of adsorption. In graphitized blacks such sites are no longer found, and swelling is unhindered by the presence of black. Reinforcement is explained by the more homegeneous distribution of tension between molecular chains due to slippage on the carbon surface.

Dynamic Behavior of Rubber During Moderate Extensions

Abstract The stresses and energy losses during simple extension cycles to about 130 per cent maximum elongation have been determined for unfilled vulcanizates of natural rubber and acrylontrile-butadiene rubber at various rates of extension and temperatures. The results obey the Ferry transform, with the exception of natural rubber at the larger strains near the high temperature end of the experimental range. The stress-strain curves at increasing strain cannot be described as the product of one strain function and one time function, but the two-term Mooney equation with time dependent coefficients represents the results over a large strain range. Maxima in the strain-increasing stress curves have been observed which are not connected with necking. A suggestion is made as to why the negative slopes in these curves need not necessarily lead to instability.

Complexities of Crosslink Density in Filled Elastomers

Abstract This work was undertaken to learn more about the complicated role of fillers in reinforcing siloxane polymers and attempt to clarify why silica fillers cause the crosslink density to increase significantly more in siloxane polymers than in hydrocarbon polymers. To carry out this study it was necessary to decide upon a method for determining the effective number of crosslinks. A simple mathematical expression applicable to both filled and unfilled elastomers was developed for use with a compression modulus technique for swollen vulcanizates. The fillers studied with polydimethylsiloxane (containing a small amount of methylvinylsiloxane) were a fume silica, wet process silica, diatomaceous earth, and two ground quartz silicas. The effect of concentration of network chains on tensile strength, modulus, and dynamic mechanical properties was compared for fume silica and diatomaceous earth filled stocks. All of the types of silica studied caused the crosslink density to increase several times over that of the unfilled vulcanizate. This may be contrasted with a factor of less than 2 for a filled (carbon black or silica filler) hydrocarbon polymer. The difference in behavior of the two polymer systems was attributed to a pronounced interaction between the siloxane polymer and the silica surface. The data indicated that filler structure has a more pronounced effect on crosslink density than surface area. It was found that the log of concentration of network chains gives a straight line when plotted as a function of volume fraction filler.

Dynamic behaviour of rubber during moderate extensions

The stresses and energy losses during simple extension cycles to about 130% maximum elongation have been determined for unfilled vulcanizates of natural rubber and acrylonitrile butadiene rubber at various rates of extension and temperatures. The results obey the Ferry transform, with the exception of natural rubber at the larger strains near the high temperature end of the experimental range. The stress-strain curves at increasing strain cannot be described as the product of one strain function and one time function, but the two-term Mooney equation with time dependent coefficients represents the results over a large strain range. Maxima in the strain-increasing stress curves have been observed which are not connected with necking. A suggestion is made as to why the negative slopes in these curves need not necessarily lead to instability.

Relations between ultimate tensile properties of elastomers and their structure

Various factors which affect the tensile stress-at-break ( σ b ) and the corresponding ultimate extension ratio ( λ b ) of unfilled amorphous elastomers are considered. Such ultimate properties can be characterized by a failure envelope which is independent of the test temperatures and apparently of the stress-strain-time history prior to rupture. The envelope results from a plot of log σ b T 0 /T against log ( λ b — 1), where T and T 0 are, respectively, the test temperature and an arbitrary reference temperature, both expressed in degrees absolute. Characteristics of failure envelopes are illustrated in terms of data obtained at various rates of extension over extended temperature ranges on unfilled vulcanizates of silicone, butyl, styrene-butadiene, Viton (hydrofluorocarbon), and natural rubber. It is shown that the high-temperature segment of a failure envelope is identical to the near-equilibrium stressstrain curve and that the envelopes for vulcanizates having different cross-link densities yield a master envelope on a plot of log λ b σ b T 0 / T against log E e ( λ b — 1 ) T 0 / T , where is the true λ b σ b stress-at-break and E e is the equilibrium modulus. On this type of plot, envelopes for all vulcanizates studied, except those of silicone and natural rubber, are essentially identical. Effects resulting from crystallization prior to rupture are discussed briefly in terms of data obtained on the natural rubber vulcanizate at temperatures below 90°C.

Ultimate Tensile Properties of Elastomers. II. Comparison of Failure Envelopes for Unfilled Vulcanizates

Abstract The tensile stress-at-break σb (based on the initial cross-sectional area) and the corresponding ultimate extension ratio λb of unfilled vulcanizates of silicone, hydrofluorocarbon (Viton B), butyl (both sulfur-cured and resin-cured), and natural rubber were determined at many strain rates and temperatures; the latter ranged from slightly above the glass transition temperature Tg, up to a temperature somewhat below that at which chemical degradation affected the results. For each vulcanizate except natural rubber, data obtained over an extended temperature range superposed to give a time- and temperature-independent failure envelope on a plot of log (σb273/T) vs log (λb−1), where T is the test temperature in °K; for natural rubber, data obtained between 90° and 120° C superposed, but those at lower temperatures did not because of strain-induced crystallization. For each vulcanizate, data at elevated temperatures gave, or tended toward, a line of unit slope on a plot of log (λbσb273/T) vs log (λb−1), where λbσb is the breaking stress based on the cross-sectional area at the moment of rupture. The position of each line corresponded to the equilibrium modulus Ee derived from stress-strain curves. Failure envelopes previously obtained for two styrene—butadiene vulcanizates, which had different crosslink densities, superposed to give a master failure envelope on a plot of log (λbσb273/T) vs logEe(λb−1). On this type of plot, failure envelopes for all the vulcanizates except silicone and natural rubber were found to be essentially identical. At a given value of λbσb, silicone had a smaller λbλb and natural rubber a somewhat larger λbλb than the vulcanizates of the three other rubbery polymers.

Swelling of Filler-Reinforced Vulcanizates

A theory is developed to account for the restricted swelling in solvents of crosslinked elastomers containing reinforcing fillers. Assuming swelling to be completely restricted at the filler-rubber interface due to adhesion, the following relation is obtained: where vr is the volume fraction of rubber in the swollen rubber phase, Vro is the same quantity referred to on otherwise analogous, unfilled vulcanizate, Φ is the volume fraction of filler, and c a parameter depending on the filler, but independent of Φ and Vro. This equation is shown to hold for a large volume of experimental data on carbon blacks, involving four rubbers, several sulfur vulcanizing systems, five solvents, and a wide range of crosslinking. Conformance with the theory indicates that carbon blacks are firmly bonded to the rubber and that, in the sulfur crosslinking systems investigated, they have no significant effect on the stoichiometry of vulcanization (although they may affect the rate of vulcanization). Illustrative examples of applications of the theory to problems in filler reinforcement and vulcanization are shown.

Ultimate Tensile Properties of Elastomers. II. Comparison of Failure Envelopes for Unfilled Vulcanizates

The tensile stress-at-break σb (based on the initial cross-sectional area) and the corresponding ultimate extension ratio λb of unfilled vulcanizates of silicone, hydrofluorocarbon (Viton B), butyl (both sulfur-cured and resin-cured), and natural rubber were determined at many strain rates and temperatures; the latter ranged from slightly above the glass transition temperature Tg, up to a temperature somewhat below that at which chemical degradation affected the results. For each vulcanizate except natural rubber, data obtained over an extended temperature range superposed to give a time- and temperature-independent failure envelope on a plot of log(σb273/T) vs log(λb−1), where T is the test temperature in °K; for natural rubber, data obtained between 90° and 120°C superposed, but those at lower temperatures did not because of strain-induced crystallization. For each vulcanizate, data at elevated temperatures gave, or tended toward, a line of unit slope on a plot of log (λbσb273/T) vs log(λb−1), where λbσb is the breaking stress based on the cross-sectional area at the moment of rupture. The position of each line corresponded to the equilibrium modulus Ee derived from stress—strain curves. Failure envelopes previously obtained for two styrene—butadiene vulcanizates, which had different crosslink densities, superposed to give a master failure envelope on a plot of log(λbσb273/T) vs logEe(λb−1). On this type of plot, failure envelopes for all the vulcanizates except silicone and natural rubber were found to be essentially identical. At a given value of λbσb, silicone had a smaller λb and natural rubber a somewhat larger λb than the vulcanizates of the three other rubbery polymers.

Relaxation Processes in Vulcanized Rubber. III. Relaxation at Large Strains and the Effect of Fillers

Abstract Some experimental measurements are described of stress relaxation and creep at room temperatures in vulcanizates of natural rubber, butyl, and SBR. In an unfilled natural rubber vulcanizate the rate of stress relaxation is found to rise sharply for extensions of more than about 200%. Reasons are given for attributing this to the growth of a crystalline phase. Similar rates are observed at all extensions for a carbon black filled natural rubber vulcanizate. This is shown to be in satisfactory accord with the Mullins-Tobin model structure for filled vulcanizates, when the whole of the observed relaxation occurs in “softened” regions at rates appropriate to the high local deformations. The failure of rubber-carbon black associations with time does not appear to constitute a major relaxation process. In noncrystallizing unfilled vulcanizates the rate of relaxation is found to decrease somewhat with extension, possibly due to finite-extensibility effects. Preliminary measurements on a filled SBR vulcanizate suggest that a significant contribution to the observed relaxation arises from progressive failure of rubber-filler associations in this case. The relation derived previously between the rates of creep and stress relaxation at equivalent deformations is confirmed in all cases, within experimental error. Its validity in highly-irreversible systems is thus established experimentally.