Blends of Bitumen With Polymers Having a Styrene Component.
T. McNALLY [**]
The properties of a 100 penetration grade bitumen are modified considerably, and in a number of ways by the addition of 10 to 40 parts per hundred (pph) of a homopolystyrene and graft, block and random copolymers of styrene with butadiene and acrylonitrile. At low temperatures some blends have a similar stiffness to or even lower stiffness than the bitumen, but generally the blends are more than one order of magnitude stiffer, even when a rubber is added. The contrasting behavior is displayed by a polystyrene and a high impact polystyrene, [sim]3% to 4% of grafted rubber on the latter being sufficient to cause the enhancement, even at the 10 pph level, by two different random styrene-butadiene copolymers. and also by blends consisting of different amounts of SBS block copolymer. Some polymers apparently trigger a Hartley inversion of the micellar structure of the asphaltene micelles. High low temperature stiffness correlates roughly with a lower [T.sub.g], as measured by the peak maximum in the E" plots of th e dynamic mechanical thermal analysis (DMTA) and by the steps in the differential scanning calorimetry (DSC) curves at temperatures below 0[degrees]C. Tan [sigma] maxima and DSC traces detected the glass transition in the continuous phase and in the dispersed phases, but none of these amorphous polymers formed a crystalline phase. though the DSC traces of the polystyrene and the SBS blends suggested that the polymer-rich phases underwent an aging/ordering process on cooling. Our SBS blends differ in phase inversion behavior and the pattern of loss processes from others that had a smaller asphaltene component.
This work and related studies are designed to establish the mechanical properties of blends of a range of polymers with a commercial bitumen, and seeks to explain them in terms of the behavior of the polymers within a complicated matrix. A number of such blends, with atactic polypropylene (APP) and some rubbers, are already of industrial use [1-3]. The finite nature of APP resources has promoted the provision of certain polymers, whose first intention is to be used in this manner, as well as our exercise in examining others that might be plentifully available as scrap, as candidates for bitumen modification. We have noted one commercial use of polymer-bitumen blends  that has been described since a preliminary communication  mentioned the surprising stiffness of a certain isotactic polypropylene blends. Our studies may lead to other recycling opportunities, as well as promote the disposal of waste plastics in road pavements.
Bitumen itself is a thermoplastic that remains after volatile components have been removed from crude oils [2, 6], and consists of a colloidal system of at least four readily recognizable components: paraffinic saturates, aromatics. resins and graphitic asphaltenes  whose molecular weights are probably at most a few thousand . It is often considered that the asphaltenes are dispersed as a sol within the first two components by the influence of the resins, an idea supported by the following cohesive energy density [8, 9], or solubility parameters, in respective order 14.3, 17.8, 19.7 and 23 [(J/[cm.sup.3]).sup.0.5]. Should the proportion of the asphaltenes and paraffins rise above 50%, a gel may be obtained [3, 10], but polar entities such as water and salts may also induce order in the asphaltenes . It has been suggested  that Hartley micelles, with the polar groups extended, may be obtained together with the reversed form, and that these themselves may be ordered within larger structures of simi lar spherical shape or as multi-lamellar.. When a single polymer is introduced there then arises as the temperature is lowered the possibility of bitumen-rich and polymer-rich phases [9, 10, 12, 13] if the polymer is glassy or rubbery, depending on its intermolecular forces, as well as the formation of a crystalline phase of the polymer. The latter has been encountered after cooling several blends with bitumen of polyolefins and their copolymers [1, 3, 5.14, 15], and has recognizable influence on the mechanical properties through the formation of crystallite crosslinks, but is unlikely to be found in this work, which focuses on an atactic styrene homopolymer and several styrene copolymers with extra polar or rubber residues, either present within the backbone or grafted on to it. Included here also are an acrylonitrile-butadiene-styrene (ABS) graft copolymer and a styrene-butadiene-styrene (SBS) block copolymer system, the microphases of which provide a further elaboration , with beneficial effects on rhe ological properties for many uses. A specific interaction was found when an ionomeric polyethylene was added to the bitumen we employ, which seems to have caused a particularly high modulus : here strongly polar residues ([micro] = 4.5D) are present in the styreneacrylonitrile (SAN) and ABS systems, and may cause a similar effect.
In general we have used proportions of polymer that rise to 40 pph in four steps (or to 29% by weight), in order to locate any phase inversion-the development of a polymer-rich phase-a change that was encountered just below 40 pph with APP systems , but at only 8% with an SBS rubber [1, 9] in one bitumen. If our blend compositions differ in the type of polymer used and in the proportion employed from certain industrial practices, they may yet be of service in assisting with understanding the behavior, and with the development of new systems. The type of bitumen we use, from a Venezuelan oil field, is typically used in commercial blends with APP [1, 14, 15]. We use two investigative methods, fluorescence microscopy and DSC, to guide our interpretation of the DMTA measurements and their changes with composition. For the behavior of the final system, an SBS block copolymer. literature studies are also available for comparison [8, 9,15,16. 18-21], for this is widely used.
All blends were prepared at about 18000 in a Z-blade masticator mixer at concentrations of 10, 20, 30 and 40 parts of polymer per hundred parts of bitumen (pph), as described in detail . This range includes the concentrations of several useful polymer-bitumen blends [6, 22-24]. The polymers used in this study are shown in Table 1. They were examined by fluorescence optical microscopy using an Olympus BHS-RFL microscope with a mercury lamp as a light source, by a Perkin-Elmer differential scanning calorimeter and by a Polymer Laboratories Mark II dynamic mechanical thermal analyzer at 1 Hz, to measure the variation in modulus with temperature. For this purpose, melted blends were cast in the form of 2mm thick slabs, from which 5 mm X 20 mm samples were cut . The measurements on the bitumen alone were made in the shear mode (G') using 6 mm diameter disks, so the values were one third the value that would have been obtained in the bending mode (E') for the blends were obtained in the bending mode  w ith a single cantilever configuration. The bitumen itself was a Nynas product (100 penetration grade), and, by iatroscope anaylsis [14, 17] contained aromatics, resins, asphaltenes and paraffins in the proportion 47:31:16:6, as Fig. 1 shows.
RESULTS AND DISCUSSION
Blends of Bitumen With a Linear Homo Polystyrene
The blends prepared over two hours with the homopolystyrene were not very viscous at elevated temperatures, but were tacky on a glass surface. The fluorescence photomicrographs of this polymer's blends are shown in Fig. 2a and 2b, where two phases are observed. It is probable that the polystyrene was partitioned as the blends cooled from the melt. The first, lightly colored, was a polystyrene-rich macro phase, present as spherical droplet-type structures (varying diameters from 2 to 20 [micro]m). which are dispersed within a matrix of the dark second phase, which is bitumen-rich. These large droplets became more frequent and more elaborate in shape as the concentration rose. They differ considerably in size from microdomains [16, 26], whose dimensions are limited by molecular weight.
The presence of two phases in these blends was also evident from the DSC results (see Fig. 3). The polystyrene by itself exhibited one process, a glass transition at 81[degrees]C. In the curves from the blends, two processes are observed, a low temperature glass transition with a step in the specific heat capacity at about -23[degrees]C, which may be assigned to the bitumenrich phase, as with other systems we have examined [1, 5, 14. 15, 17], and a higher temperature 'distorted' glass transition at about 79[degrees]C, which may be attributed to the polystyrene-rich phase. As the weight fraction of the polymer in the blend was increased, the step size of the glass transition at 80[degrees]C rose, and the feature came to resemble less a pure endotherm, and that of the unambiguous glass transition near -25[degrees]C decreased. These trends are clearly established from the change in specific heat values obtained for both processes (see Table 2). The two phases are standard thermodynamic consequence of cooling a polymer in a poor solvent (12, 13), and here, at low temperatures, each is below a glass transition. A 'distortion' to the glass transition process attributed to a polystyrene-rich phase within the blends might be caused by the heating rate being faster than the previous cooling rate (27), but for these studies, the value of each was the same, 10[degrees]C [min.sup.-1]. The feature may perhaps result from the presence of a thermo-reversible physical gel formed from cooperative bonding between chain segments. This phenomenon has been studied by several workers for dilute and semidilute solutions of polystyrene [28-32] in poor solvents such as [CS.sub.2] and decalin, but in these cases the gels form at temperatures much lower than here. The present features, rather, have the form consistent with the promotion of physical aging in a polystyrene-rich phase by a component of the bitumen whose influence is less effective the greater the proportion of the polymer or the greater the size of the droplets: elsewhere it has been shown that PMMA and PVC have prominent endothermic features when a liquid nitrogen chill and annealing was used to accelerate aging .
The variation in storage modulus and tan [delta], and loss modulus of the blends with temperature are shown in Figs. 4 and 5. Except for the 10 pph blend, the modulus begins to level off to a large extent as a rubbery plateau is reached  just after the DSC-determined [T.sub.g] of the polystyrene-rich phase has been passed. The DSC transition obtained is not associated with a fall in the modulus, as when the crystallites that crosslink a gel melt [14, 15], but rather with this plateau, a consequence of chain entanglements, until, at about 150[degrees]C, the blends ceased to be measurable at the instrument setting. For the 40 pph there was a clear rise in tan [delta] above 130[degrees]C as that more viscous blend began to flow. The more dilute 10 pph blend suffered less from entanglements, and became soft within the measurement system at just 110[degrees]C. The E' curves for the 30 pph and 40 pph blends cross at about 85[degrees]C, just above the temperature at which the polystyrene-rich phase has its glas s transition, as do the E" curves. The tan [delta] plots display two processes for the blends, one centered at around 40[degrees]C and which falls in amplitude but starts at lower temperatures once 20 pph of the polymer has been used and suggests that the styrene-rich phase has sequestered a bitumen-stabilizing factor: the other process at about 80[degrees]C for the 10 pph blend, becomes less intense and broadens and shifts upwards to 110[degrees]C/115[degrees]C for the 30 and 40 pph blends.
Interestingly, these blends are less stiff than bitumen itself below about 25[degrees]C, which in itself is unusual [1, 5, 14, 15, 17, 19, 34, 35]. However, above 90[degrees]C, the transition temperature of the dispersed phase, the stiffness rises with polymer concentration, presumably for standard rheological reasons. The glass transition process as displayed in the loss modulus plots, first moves to a slightly higher temperature and then broadens and falls to lower temperatures with increasing polymer content in the blends.
Blends of Bitumen With High Impact Polystyrene (HIPS)
This high impact material (see Table 1) was a polystyrene grafted onto a poly-cis-butadiene rubber at a 3% to 4% proportion during a free-radical styrene polymerisation. The rubber is dispersed in the solid polystyrene matrix in the form of discrete micro-phases. HIPS was blended with bitumen for two hours at 180[degrees]C. All the blends we prepared wet glass at elevated temperatures and had a shiny surface when spread thinly.
A fluorescence photomicrograph of one of these blends is shown in Fig. 2c. As was observed for the blends with polystyrene, two phases are present, a light colored polystyrene-rich phase which forms droplets, and a dark colored bitumen-rich phase. The grafted rubber component of HIPS may be dispersed within the bitumen matrix at the interface. However, the average diameter of the polystyrene-rich droplets appears to increase with increasing polymer content in the blend from about 8 [micro]m in the 10 pph to 12-16 [micro]m in the 40 pph blend, which is a contra-indication to polystyrene-rich microphases being stabilized.
HIPS itself and its blends with bitumen were examined by DSC (34). High impact polystyrene exhibited two glass transition processes, one at about -29[degrees]C, presumably derived from the rubber component in HIPS, and a second at 84[degrees]C, undoubtedly derived from the polystyrene phase. When we blended this HIPS with 100 pen. bitumen, three features were observed in the DSC traces: the first a low temperature glass transition between -24[degrees]C and -28[degrees]C, which may be attributed to both the bitumen and the rubber component in HIPS either in one or two phases. The second feature, somewhat evident in the 10 pph blend and clearly present in the other blends, as Fig. 3f shows, was a glass transition process near 80[degrees]C that may be assigned as a polystyrene-rich phase, for it increased in step height with increased HIPS content. The presence of the rubber graft has removed the anomalous behavior seen beneath in Fig. 3 for the blends of polystyrene itself. A third feature is just about discer nible at around 111[degrees]C, and perhaps may be a small endotherm derived from a 'waxy' paraffinic component crystallize from bitumen, though the enthalpy value decreased with increased polymer content in the blend.
It may be seen from the variation in loss modulus, storage modulus and tan with temperature from Figs. 5 and 6, addition of the HIPS enhances the modulus of the bitumen, the ultimate effect being just more than one and a half orders of magnitude, an increment that we attribute to the graft nature of the copolymer, for two reasons: the rubber by itself lacks that effect (35), and, as we have just seen, polystyrene by itself causes E' to fall. What is remarkable is that this effect is fully developed for the 10 pph blend, and appears to derive from only 0.4 pph of the grafted rubber. That much rubber counteracts the modulus-lowering effect of 25 times as much as polystyrene, and has as dramatic an effect as had a few percent of ionic groups in an ethylene-methacrylic acid copolymer , yet the cohesive energy densities of the butadiene residues (17.4 [+ or -] 0.5) are a little less than that of styrene residues (18.7 [+ or -] 0.5). All the blends were stiffer than the bitumen itself at any temperature, the H IPS blends being stiffer than the corresponding polystyrene blends by about two orders of magnitude over the measured temperature ranges: the stiffening effects of the 0-4 pph rubber component extends well above the [T.sub.g] of the bitumen-rich phase.
Surprisingly, as the amount of high impact polystyrene in the blends was increased above 30 pph, the temperature at which the blends began to soften and flow fell from 120[degrees]C to about 80[degrees]C. This behavior must relate to the increasing rubber contribution from the HIPS in the blends, for it is the opposite to that seen above for the blends prepared with polystyrene despite the HIPS having the lower MFI. It seems to be linked to the minor feature at 40[degrees]C in the E" plot, for it grew as the loading rose and at 70[degrees]C the loss rose by a factor of 10. It was also observed in the loss modulus plots of Fig. 5 that as the concentration of HIPS (and hence rubber content too) in the blends was increased, the main peak in E" plots broadens, and moved until finally the peak maximum was just some 5[degrees]C below that of the bitumen. At low loading the homo polystyrene had the opposite effect.
Blends of Bitumen With Two Butadiene-Styrene Copolymers (EDS and SBR)
These linear and random copolymers contained 70% and 23.5% of styrene residues respectively, and so provide points on a composition transition from polystyrene towards a homo polybutadiene, a blend of which in this bitumen has been discussed elsewhere . The blends prepared with the first material (KK38 resin) and the 100 pen. bitumen wet glass and were tacky at elevated temperatures; however, the tackiness appeared to decrease as the concentration of BDS in the blends was increased. A fluorescence photomicrograph of one of this polymer's blends with bitumen is shown in Fig. 2d, and close examination of the prints reveals the presence of a lightly colored polymer-rich phase extended throughout the bitumen matrix. Very small spherical droplets of a presumably polymer-rich phase are present (with diameters typically of about 0.8 [micro]m in the blend consisting of 40 pph BDS), but to a very limited extent, this behavior perhaps dictated by the chemical configuration of the copolymer. The BDS and its blends with bitumen have been examined by DSC , and both displayed two clearly resolved glass transition processes. The first feature at low temperatures was at -33[degrees]C for BDS itself, at -25[degrees]C for the 10 pph blend, and fell to -31[degrees]C on increasing the polymer content in the blend to 40 pph. A second glass transition process was at l03[degrees]C for the copolymer, but what appeared to be an endotherm was present some 10[degrees]C above this at 113[degrees]C in the blends. There was also some evidence for a third process at about 80[degrees]C for the blends. This feature has been seen at this temperature in the blends consisting of bitumen with polystyrene and HIPS.
The variation in loss modulus, storage modulus and tan [delta] for BDS, SBR and their blends with bitumen are shown in Figs. 5 and 7. We take the DMTA plots of the SBR first, for the single blend we have studies, at 20 pph. The curves obtained resemble the curves reported above for polystyrene and elsewhere for a polybutadienez . In particular, the modulus at low temperatures is similar to that of the bitumen, and at temperatures as high as 190[degrees]C a decent stiffness remains. The extra rubber residues seem to extend the chains beyond those of the polystyrene, whose blend at 150[degrees]C is one order of magnitude less stiff than the blend of the copolymer; the copolymer blend at that temperature was also stiffer than the corresponding blend of the poly(cis-butadiene) . In this case (23.5% styrene) the main loss process was fairly sharp and centred at 35[degrees]C, and a very minor feature appeared at 110[degrees]C; for the rubber homo polymer blend the loss extended from about 0[degrees]C to 150 [degrees]C, and, as we have seen, the polystyrene blend has features at 40[degrees]C and 90[degrees]C.
The modulus of each of the BDS blends, in contrast to those of the homo polystyrene, the homo poly(butadiene) and the previous copolymer, is more than one order of magnitude greater than that of the bitumen below about -25[degrees]C. The blends above this temperature then began to soften before the bitumen alone, but were always stiffer than the bitumen. As with the polystyrene blends, there were two loss processes evident, but here, presumably on account of the rubber residues in each phase, they came at lower temperatures. At 0[degrees]C a loss process arose as more polymer was added, and became recognizably distinct in the tan [delta] curve for the 40 pph blend, a feature we have found elsewhere when the polymer-rich phase became extensive [14, 17]. Though the 0[degrees]C loss process was not resolved in the tan [delta] curve for the 30 pph blend, the qualitative difference in the tan [delta] and E' curves appeared to develop between 20 and 30 pph loading. The peak maxima in the loss modulus plots was broa der and some 30[degrees]C below that of bitumen itself. A feature at 55[degrees]C in the tan [delta] plot became less prominent as the polymer loading was increased, unlil, at 40 pph it was barely discernible. Once above 50[degrees] the moduli rose with polymer content, their shape suggesting the approach of a rubbery plateau at about 100[degrees]C, but the flow commenced in the extensive polymer-rich phases of the 30 and 40 pph blends, and the tan [delta] curves rose rapidly. The blends persisted within the measurement system up to between 102[degrees]C and 112[degrees]C. some 50[degrees]C to 60[degrees]C above that of bitumen itself, but about 80[degrees]C below that of the SBR blend. The BDS blends then flowed abruptly after the main glass transition of the copolymer at 103[degrees]C (determined from DSC studies) had been passed, so there may be some glassy phase that also stabilized the rubbery plateau.
Blends of Bitumen With Styrene-acrylonitrile Copolymer (SAN)
The blends prepared over two hours containing SAN, a material with a higher [T.sub.g] than homo polystyrene-from the presence of the polar groups, appeared to be very tough. They wet glass at elevated temperatures and on cooling too. The 10 pph blend was tacky at temperatures above 120[degrees]C. The fluorescence photomicrograph of a typical blend with bitumen is shown in Fig. 2e. There, in the 30 pph blend (Fig. 2e), and to a lesser extent, the 20 pph blend , striations of a light colored phase associated with a copolymer-rich species were observed, the shape of which probably reflected strain from application to the glass slide. A small number of brightly colored spheres (from 5 to 25 [micro]m) were also seen in each blend.
Blends of bitumen with this polymer were examined by DSC . The copolymer itself showed one clearly defined process, a glass transition at 109[degrees]C. When we blended the SAN with bitumen, four features were observed in the DSC curves of the blends. A low temperature glass transition, at about -27[degrees]C for the 10 pph blend, that rose to about -18[degrees]C for the remaining blends. The second feature at about 101[degrees]C was also a simple glass transition, as Fig. 3g shows, and may correspond to that of the SAN copolymer shifted down-wards by about 7[degrees]C or 8[degrees]C by some plasticizing component of the bitumen attracted by the acrylonitile groups. The third process was found at 73[degrees]C to 75[degrees]C and may be either a distorted glass transition or an endothermic-like process. The fourth and final feature at 120[degrees]C present in all the blends would appear to be an endothermic process, similar to those seen for blends of certain rubbers with bitumen , and may be derived from a paraffinic component of the bitumen, which has perhaps been expelled by the migration of another--perhaps more polar--component to the polymer-rich phase.
Addition of SAN at all concentrations enhanced the storage modulus of bitumen by more than one order of magnitude below -15[degrees]C, as may be seen from the variation in loss modulus, and storage modulus and tan [delta] plots with temperature (see Figs. 5 and 8). In comparison with the homopolystyrene-bitumen blends, all of these are about 100 times as stiff at low temperatures, so since they have similar cohesive energy densities, it may be that the polar acrylonitrile residues of some polymers within the bitumen-rich phase bind directly to the stiff components, the asphaltenes. The blends began to soften before the bitumen did, according to the tan [delta] and loss modulus curves, and had glass transitions some 5[degrees]C to 10[degrees]C below that of the bitumen, but they were always stiffer than bitumen. It may be seen from the tan [delta] and storage modulus plots that the persistence of the blends within the measurement system merely rose from 50[degrees]C to 80[degrees]C on increasing the concentrat ion of SAN in the blend from 10 pph to 40 pph. The rapid rise in the tan [delta] curves indicate a propensity to flow, and there is a suggestion of a loss process at about 12[degrees]C only in the most loaded system.
Since 80[degrees]C is some 29[degrees]C below the glass transition temperature of the SAN copolymer (as determined from DSC measurements), it would appear that the continuous phase is bitumen-rich for each of these blends, the common swelling of the polymer-rich phase caused by take-up of bitumen components being inhibited by the acrylonitrile residues, which are more polar than the paraffin or aromatics. Thus, when the glass transition temperature of bitumen (i.e. 9[degrees]C) in the blends is passed, the properties of bitumen predominate. The SAN chains are not extensively dissolving in the bitumen, perhaps because they remain below a [theta]-temperature. The temperature range and intensity of the peaks in the loss modulus plots of the blends are broader and greater than those of bitumen itself, so the SAN is dispersing and enhancing E'. but at high temperatures is not producing a rubbery plateau, as did certain other polar rubbers [17, 35].
Blends of Bitumen With Acrylonitrile-Butadiene-Styrene (ABS)
ABS polymers, like rubber modified polystyrene, are two-phase systems, and in this case too, the glassy styrene-acrylonitrile copolymers with strongly polar acrylonitrile groups have been grafted onto the elastomer component. Each of the blends prepared during two hours with ABS and bitumen were quite tough; however, large lumps of what was presumed to be polymer were obtained in the final 40 pph blend, perhaps following an unwanted free radical crosslinking reaction.
The fluorescence photomicrograph of one blend of bitumen with ABS is shown in Fig. 2f Again, as for blends with a polystyrene, a light colored, perhaps polymer-rich phase was evident, the shapes obtained ranging from discrete spherical droplets to particles, which in the 40 pph photomicrograph were continuous. However, the number of droplets seen would appear to decrease with increasing polymer content in the blends, when examined at this magnification.
Blends of ABS with this bitumen were examined by DSC . ABS itself displayed one process in the temperature range examined, a glass transition at about 100[degrees]C. When we blended ABS with bitumen, two features were observed in the DSC curves, a glass transition which fell from -23[degrees]C to 28[degrees]C when the concentration of ABS in the blend was increased from 10 pph to 40 pph. This low temperature process may be associated with a bitumen-rich phase, but a butadiene-rich microphase may also be present. The second feature, either a glass transition or an endotherm, was absent from the 10 pph curve, but was present and rose from 88[degrees]C to 99[degrees]C when the concentration of ABS in the blend was increased from 20 pph to 40 pph. This feature, shown in Fig. 3h, may be attributed to a poly(styrene-acrylonitrile)-rich phase, and develops in the form of an endotherm just where the pure polymer had its--slightly distorted--[T.sub.g]. There is also some evidence, certainly from the 30 pph and 40 pph blends, of a third small process at about 60[degrees]C, and possibly in ABS slightly above 60[degrees]C, perhaps derived from a butadiene-rich phase. A similar feature may also be seen in the DSC traces of blends of bitumen with SBS at about 50[degrees]C (see next section).
These blends were also examined by DMTA (see Figs. 5 and 9), the considerable but nonprogressive enhancement of the modulus at low temperatures by the polymer, indicating that it had been dispersed in the bitumen-rich phase. The acrylonitrile and rubber residues may both induce this effect, the former with the asphaltenes, and the latter with the aromatics and paraffins. The presence of a broad peak in the tan 8 plots is detected with a maximum at about 45[degrees]C measured at 1 Hz, but the polar acrylonitrile component has prevented the emergence from the broad curve of the feature seen at 0[degrees]C in the study on the BDS copolymer. The temperature range of the peaks in the loss modulus plots of the blends, possibly derived from a bitumen-butadiene-rich phase, is much broader that that of bitumen itself, and is some 15[degrees] to 25[degrees]C lower, a lowering of the [T.sub.g] which we attribute to the rubber component. The blends persist within the measurement system up to about 115[degrees]C (that is, until the glass transition of the styrene-acrylonitrile-rich phase has been passed). Above a maximum in the tan [delta] plots, the storage modulus curves of the 20, 30 and 40 pph blends leveled off to an extent, from 60[degrees]C to 90[degrees]C, and approach a rubbery-plateau, being most clear in the 30 pph blend. Further, above this temperature, a sharp drop in the modulus and sudden increase in tan [delta] was observed with very sharp peaks for all but the 40 pph blend. This sharp feature, not seen by us in any other bitumen blend , has a width of less than 20[degrees]C for the 30 pph blend, resembles the [alpha]-process of some entirely glassy polymers , and so may derive from an AS microphase.
Blends of Bitumen With Styrene-Butadiene-Styrene (SBS) Block Copolymer
SBS is a thermoplastic elastomer, a triblock system of styrene and butadiene, and is readily used in bitumen modification, normally at a level of about 15% for roofing systems [1, 9], but just a half or a third of that when used in roads . The blends prepared over two hours with bitumen and SBS were very viscous at all compositions at elevated temperatures, displayed elastic properties and appeared to become less tacky with increasing concentration of SBS in the blends. Our fluorescence photomicrographs of this polymers blends with bitumen showed that they were well dispersed in all proportions , though contrast between the polymer-rich and bitumen-rich phases was rather poor for those blends containing a higher concentration of polymer. Close examination of the photomicrographs revealed spherical lightly colored droplets extended throughout the dark colored bitumen matrix. The droplets are too large to be microphases of agglomerating polystyrene blocks of the SBS but may reflect the presence of the 'h oneycombus' observed by others with an electron microscope . In these, polystyrene microphases form physical crosslinks between the polybutadiene mid-blocks to give an elastomeric network, which may extend throughout the bitumen matrix if enough rubber is added.
The DSC characteristics for SBS and its blends are shown in Tables 3 and 4. This polymers blends displayed three similar features, a glass transition process at about -30[degrees]C presumably derived from a bitumenrich phase within which the rubber may be dispersed, the second feature at 50[degrees]C may either be a glass transition or distorted endotherm, which may now be attributed to a styrene-rich microphase. The third feature was a small endotherm at about 120[degrees]C, perhaps reflecting some paraffinic component of the bitumen, as the transition was absent from the broad amorphous DSC curve of SBS itself.
Both the 5 pph and 20 pph blends were annealed at 115[degrees]C for 2 and 24 hours (see Table 4). The glass transition feature at about 50[degrees]C in the blends after annealing resembled more an endotherm than a glass transition process, as was found for the polystyrene blends of Fig. 3, the micelles might also have enjoyed a physical aging process. The high temperature endotherm observed became sharper upon annealing, but the enthalpy value for this process remained constant; the time allowed for annealing presumably facilitating the ordered (or crystalline) component to achieve a more regularly sized state.
The variation in loss modulus, and, storage modulus and tan [delta], for some the polymer blends are shown in Figs. 5 and 10, where it can be seen that the blends consisting of only 5 pph SBS were considerably stiffer than bitumen itself below -20[degrees]C. This enhancement of the modulus of a brittle material at room temperature maybe consistent with other studies with another bitumen-SBS system at the 7% level , and is similar to the results obtained above with HIPS, that contained 3% or 4% poly-(cis-butadiene) grafts but is in sharp contrast to the first system we examined, the homo polystyrene, which, if anything, lowered the low temperature modulus. These are very subtle effects not readily explicable in terms of preferential binding or intermolecular forces, for the cohesive energy densities are respectively, for the rubber and styrene blocks, about 17 and 18 [MPa.sup.0.5] [([J/cm.sup.3]).sup.0.5] , and so neither might bind strongly to the polar asphaltenes. The curiosity deepens, for when the polymer concentration was raised to 15 pph (13% by weight), the low temperature modulus fell by a factor of 100 or more, in keeping with the observations with an electron microscope of a phase inversion near 7% in another system, a phase inversion there supported by tensile studies and interpreted with the Ashby-Gibson model of the dispersion . Such phase inversion was not seen, however, when a 20% blend of another SBS rubber was examined: to demonstrate these contrasting behaviors we have drawn up Table 5. The tan [delta] plots (Fig. 10) show how loss processes evolve and develop as the block copolymer was added to the bitumen. The process at -15[degrees]C was better resolved than would be expected from the trends in the previous study with a lower asphaltene content , though here we find an extra feature at 30[degrees]C. This we presume reflects the higher proportion of the asphaltenes and the lower proportion of aromatics in the present bitumen, and anticipates the 2000 feature in our tan [delta] curves.
All the blends were stiffer than bitumen above 30[degrees]C, the modulus of the blends consisting of 15 pph and 40 pph SBS falling to a plateau at about 90[degrees]C. These blends persisted within the measurement system up to about 150[degrees]C, perhaps the temperature at which the styrene droplets disintegrate. There was an interesting rise in the E' as the temperature of the 40 pph blend rose from 50[degrees]C to 70[degrees]C, which may reflect the occurrence of a change in microphase morphology, say from spheres to cylinders. Such morphologies are well known for SBS block copolymers [15, 38, 39], and in aqueous solutions, ethylene oxide-propylene oxide pluronics have been known to show dramatic changes of modulus with temperature at microphase transitions [10, 33, 40].
With this series of a polystyrene and its graft, random and block copolymers with butadiene and two systems with acrylonitrile residues, the properties of a Venezuelan 100 penetration grade bitumen has been modified considerably. The addition of the polystyrene in all the proportions examined [10, 20, 30 and 40 pph] and certain proportions of the SBS rubber [5, 15 and 40 pph] have been found to cause a small fall in the low temperature modulus, as also, did a random copolymer with 23.5% cis-butadiene and a homo poly(cis-butadiene) . Other polymers had the more usual effect of enhancing the modulus from about 2 X [10.sup.7] Pa to more than [10.sup.9] Pa, perhaps the most dramatic rise being caused by the HIPS, that differed from the polystyrene in having only 3% to 4%, of a poly(cis-butadiene) grafted into the backbone. Unfortunately this enhancement, induced in the 10 pph blend by 0.3% of the rubber and contrary to what is found for such polymers by themselves , was not continued when extra quantities of HIPS was added to the bitumen: a similar enhancement to 4 X [10.sup.9] Pa was found with an ionomer in this bitumen, caused by a similar proportion of the polar residues , but the reference material, a polyethylene blend then had a modulus of [10.sup.9] Pa , so the specific effect was in fact smaller. Our work with polymers has found a modulus of [10.sup.9] to 2 X [10.sup.9] Pa with so many that it is simpler to record that the only other systems, besides those found here, that do not promote a rise in E' to above [10.sup.9] Pa when added at the 10 pph level were APP (15), a chlorinated polyethylene, Lycra and a homopoly(cis-butadiene) . Because the enhancement is so consistent, takes place before any polymer-rich phase can become extensive, always reaches a value that is less than that of the semicrystalline or glassy polymers themselves, is caused by polymers of quite different polarities or cohesive energy densities, and is found whether the additive is above or below its own [T.sub.g], the low temperature modulus of the blends must reflect a reorganization of the bitumen prompted by the entry of some polymer into the bitumen-rich phase. Some workers have noted that strongly polar substances such as water and salts can create order in asphaltenes . Here we conclude that soft materials such as rubbers as well as glassy polymers such as SAN and PMMA  do so too. Stiffening road pavement asphalts by incorporating scrap plastics might be one application of this effect.
While semicystalline polymers such as LDPE may be acting, in the Nellenstein manner, as paraffinic components, which provide a gel if they with the asphaltenes exceed 50% of the bitumen, here a more subtle effect in indicated, for while the modulus is raised, the glass transition is generally lowered. Indeed, as an examination of Fig. 5 reveals, blend stiffness at lower temperatures correlates with a low [T.sub.g], The contrast is best found between the polystyrene and HIPS systems, between the two random styrene-butadiene copolymers and within the SBS blends. Of the random copolymers, the BDS blends have perhaps the lowest temperature for the E" maxima, but the SAN system is almost equally effective at providing a low [T.sub.g].
The DSC measurements invariably show that the bitumen-rich phase has a glass transition between -20[degrees]C and -30[degrees]C, the magnitude of the step usually falling as more polymer was added, at a rate, as for polystyrene itself (Fig. 3). which suggests that the phase loses material to the polymer-rich phase. The glass transition of the homopolystyrene blends near 80[degrees]C and possibly of the AS-rich phase of the ABS system resembles more an endotherm, as if the glassy phase achieved a high degree of aging. The micro-spheres of the SBS system may enjoy the same accelerated aging, to provide an endotherm near 50[degrees]C, but the SAN and the HIPS blends have standard glass transitions at about 102[degrees]C and 77[degrees]C, only 7[degrees]C below those of the pure polymer. The SAN polymers' blends flow most readily, as the temperature is raised, presumably because the polar residues do not extend the coils in the non-polar environment, and it is the polystyrene and then the SBS and finally the SBR copolymer that postpone flow to the highest temperature. It seems that entanglements (or the presence of a glassy phase) create a clear rubbery plateau in the modulus of only the 15 pph SBS system, in the SBR blend, but entanglements do enhance the moduli of other systems.
A phase change in the polymer-rich phase in the 40 pph SBS system causes a rise in the modulus at about 70[degrees]C by about a factor of two. While the glass transitions are clearly found in the E" plots of Fig. 5, only in the 15 pph SBS blend and in the 40 pph BDS blend is a tan [delta] process clearly resolved in the manner that has led us to recognize a pliable extensive polymer-rich phase elsewhere [5, 14, 15, 17]. Generally the first tan [delta] maximum was found between 40[degrees]C and 80[degrees]C, and observed to diminish in intensity and shift to higher temperatures as plasticizing components were diluted with more polymer. Whereas the 5 pph and 20 pph SBS blends had three tan [delta] processes, the 15 pph blend had only one with this bitumen, a behavior that differed from experiments by others with lower molecular weight polymers and with bitumen's having fewer asphaltenes .
The authors gratefully acknowledge the financial and technical support of Dussek-Campbell APP Polymers Division of the Burmah-Castrol Group, and the encouragement of Mr. J. Turford and Mr. R. Bailey. We thank Dr. David George (Rubberoid Ltd.) for the Iatroscope analysis of the 100 penetration grade Nynas bitumen, and the manufacturers listed for supplying the polymers used in this study.
(*.) Corresponding author.
(**.) Present Address: MVC Technology Research Group, Polymer Processing Research Centre. The Queen's University of Belfast, Belfast. BT9 5AH, U.K.
(1.) A. H. Fawcett, T. McNally, and G. M. McNally, J. Elastomers & Plastics, 31, 334 (1999).
(2.) British Standards 3690: Part 1: 1989: Bitumens for Building and Civil Engineering. Part 1. Specification for Bitumens for Road Purposes.
(3.) D. Whiteoak, The Shell Bitumen Handbook, Shell Bitumen U.K., Surrey (1990).
(4.) D. R. Vermillion and J. A. Marzari, SPE ANTEC Tech. Papers, 46, 2956 (2000).
(5.) A. H. Fawcett, T. McNally, and G. M. McNally, ACS Polym. Preprints, 40.1, 216 (1999).
(6.) R. N. Traxler, Asphalt, Its Composition. Properties and Uses, New York, Reinhold (1961).
(7.) B. Boutevin, Y. Pietrasanta, and J-J. Robin, Progress in Organic Coatings, 17, 221 (1989).
(8.) T. F. Yen, Asphaltic Materials, Encyclopedia of Polymer Science & Engineering, H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, eds., Wiley, Chichester, Index Vol., p. 1 (1990).
(9.) P. Wloczysiak, A. Vidal, and E. Papirer, J. Appl. Polym. Sci., 65. 1609 (1997).
(10.) F. Nellenstein, J. Int. Pet, 10, 311 (1924).
(11.) T. F. Yen, The Future of Heavy Crude and Tar Sands, p. 174, McGraw-Hill, New York (1981).
(12.) P. J. Flory, Principles of Polymer Chemistry, Cornell Univ. Press, Ithaca, New York (1953).
(13.) N. te Nijenhuis, Adv. Polym. Sci., 1, 130 (1997).
(14.) A. H. Fawcett, T. McNally, G. M. McNally, F. Andrews. and J. Clarke, Polymer, 40, 6337 (1999).
(15.) A. H. Fawcett and T. McNally, Polymer, 41, 5315 (2000).
(16.) L. S. Flosenzier and J. M. Torkelson, Macromolecules, 25, 735 (1992).
(17.) A. H. Fawcett and T. McNally, Macromol. Mater. & Eng., 286, 126 (2001).
(18.) P. Wloczysiak, A. Vidal, E. Papirer, and P. Gauvin, J. Appl. Polym. Sci., 65, 1595 (1997).
(19.) R. Blanco, R Rodriguez, M. Garcia-Garduno, and V. M. Castano, J. Appl. Polym. Sci., 56, 57 (1995).
(20.) E.J. Van Beem and P. Brasser, J. Int. Pet., 59, 91 (1973).
(21.) J. S. Bahl, N. Atheyra, and H. Singh, Erdol und Kohle-Erdgas-Petrochemie vereinigt mit Brennstoff-Chemie, 46, 23 (1993).
(22.) A. Ait-Kadi, B. Brahmi, and M. Bousmina, Polym. Eng. Sci., 36, 1724 (1996).
(23.) J. Read, Chem. Br., 34, 46 (1998).
(24.) E. N. Kresge, R H. Schatz, and H-C Wang, Isobutylene Polymers, Encyclopedia of Polymer Science & Engineering, H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, eds., Wiley, Chichester, Vol. 8, p. 423 (1987).
(25.) H-G. Elias, Macromolecules, Vol. 1, p. 425, Wiley, London (1977).
(26.) S. Krause and Z-H. Lu, J. Polym. Sci., Polym. Phys. Ed., 19, 1925 (1981).
(27.) R. Ryong-Joon, Glass Transition, Encyclopedia of Polymer Science & Engineering, H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, eds., Wiley, Chichester, 2nd Ed., Vol. 7 (1987).
(28.) N. Lehasaini, R. Muller, G. Weill, and J. Francois, Polymer, 35, 2180 (1994).
(29.) C. Daniel, C. Dammer, and J-M. Guenet, Polymer, 35, 4243 (1994).
(30.) J-M. Guenet, Macromolecules, 19, 1961 (1986).
(31.) R. F. Boyer, E. Baer, and A. Hiltner, Macromolecules, 18, 427 (1985).
(32.) H-M. Tan, A. Moet, A. Hiltner, and E. Baer, Macromolecules, 16, 28 (1983).
(33.) F. R. Schwarzl, Viscoelasticity, Encyclopedia of Polymer Science & Technology, H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, eds., 2nd Ed., Vol. 17, p. 587 (1989).
(34.) T. McNally, PhD thesis, The Queen's University of Belfast (1996).
(35.) A. H. Fawcett and T. McNally, J. Appl. Polym. Sci., 76, 586 (2000).
(36.) D. W. Hadely and I. M. Ward, Mechanical Properties, Encyclopediea of Polymer Science & Technology, H. F. Mark, N. M. Bikales, C. G. Overberger, and G. Menges, eds., 2nd Ed., Vol. 9, p. 378 (1987).
(37.) Polymer Handbook, J. Brandkup, E. H. Immergut, and E. A. Grulke, eds., Wiley, Chichester (1999).
(38.) E. Helfand, Macromolecules, 8, 552 (1975).
(39.) E. Helfand and E. Wauserman, Macromolecules, 11, 960 (1978).
(40.) E. B. Jorgensen, J. H. Jensen, and S. Huidt, J. Non-Cryst Solids, 172-4, 972 (1994).
(41.) S. N. Goyanes, J. Appl. Polym. Sci., 75, 865 (2000).
Table 1. The Polymers Used in This Study. Polymer Manufacturer/Grade Polystyrene BP Chemicals HF555 High Impact Polystyrene (HIPS) BP Chemicals 2220 Butadiene-styrene copolymer (BDS) Philips K-resin KK38 Styrene-butadiene rubber (SBR) Buna EM 1507 Styrene-acrylonitrile (SAN) BASF Luran 378P Acrylonitrile-butadiene-styrene (ABS) Bayer Novodur P2L-AT1792 Styrene-butadiene-styrene Shell Chemicals Cariflex TR 1184 block copolymer (SBS) Polymer Composition Polystyrene 100% Homopolymer High Impact Polystyrene (HIPS) 3 to 4% poly(cis-butadiene) graft Butadiene-styrene copolymer (BDS) 70% styrene Styrene-butadiene rubber (SBR) 23.5% styrene Styrene-acrylonitrile (SAN) Ca. 75% styrene Acrylonitrile-butadiene-styrene (ABS) 60% styrene Styrene-butadiene-styrene 30% styrene block copolymer (SBS) Polymer MFI (g/lO min.) * Polystyrene 20 High Impact Polystyrene (HIPS) 14 Butadiene-styrene copolymer (BDS) 8 Styrene-butadiene rubber (SBR) -- Styrene-acrylonitrile (SAN) 22 Acrylonitrile-butadiene-styrene (ABS) 45 Styrene-butadiene-styrene -- block copolymer (SBS) (*)MFI: Melt Flow Index. Table 2. DSC Characteristics of Blends of Bitumen with Polystyrene. pph Polymer 10 20 30 % polymer 9.1 16.7 23.1 [T.sub.g] bitumen-rich phase -22 -25 -24 ([degrees]C) [delta][C.sub.p] bitumen-rich phase 0.14 0.08 0.07 (J/g/[degrees]C) [T.sub.g] polystyrene-rich phase 78 79 79 ([degrees]C) [delta][C.sub.p] polystyrene-rich phase 0.87 1.34 1.66 (J/g/[degrees]C) pph Polymer 40 Polystyrene % polymer 28.6 100 [T.sub.g] bitumen-rich phase -23 -- ([degrees]C) [delta][C.sub.p] bitumen-rich phase 0.03 -- (J/g/[degrees]C) [T.sub.g] polystyrene-rich phase 80 81 ([degrees]C) [delta][C.sub.p] polystyrene-rich phase 2.10 2.9 (J/g/[degrees]C) Table 3. DSC Characteristics of Blends of Bitumen With SBS. pph Polymer 5 10 15 20 40 % polymer 4.8 9.1 13 16.6 28.6 [T.sub.g1] ([degrees]C) -25 -25 -25 -20 -- [delta][C.sub.p][T.sub.g1] 0.7 0.9 -- -- -- (J/g/[degrees]C) [T.sub.g2] ([degrees]C) 52 48 50 50 53 [delta][C.sub.p] [T.sub.g2] 1.05 2.39 -- -- -- (J/g/[degrees]C) [T.sub.m] ([degrees]C) 121 121 121 121 121 [delta]H [T.sub.m] (J/g) 0.71 0.91 0.92 0.80 0.86 Table 4. DSC Characteristics of Annealed Blends of Bitumen With SBS. 5 pph SBS Blend & annealed annealin 5 pph SBS 2 hours conditions unannealed @ 115[degrees]C [T.sub.g1] ([degrees]C) -25 -25 [delta][C.sub.p] [T.sub.g1] 0.70 0.07 (J/g/[degrees]C) [T.sub.g2]([degrees]C) 52 50 [delta][C.sub.p] [T.sub.g2] 1.05 2.85 (J/g/[degrees]C) [T.sub.m] ([degrees]C) 121 121 [delta]H [T.sub.m] (J/g) 0.71 0.95 5 pph SBS Blend & annealed annealin 24 hours 20 pph SBS conditions @ 115[degrees]C unannealed [T.sub.g1] ([degrees]C) -25 -20 [delta][C.sub.p] [T.sub.g1] 0.09 -- (J/g/[degrees]C) [T.sub.g2]([degrees]C) 50 50 [delta][C.sub.p] [T.sub.g2] 2.24 -- (J/g/[degrees]C) [T.sub.m] ([degrees]C) 120 121 [delta]H [T.sub.m] (J/g) 0.69 0.92 20 pph SBS 20 pph SBS Blend & annealed annealed annealin 2 hours 24 hours conditions @ 115[degrees]C @ 115[degrees]C [T.sub.g1] ([degrees]C) 16 -29 [delta][C.sub.p] [T.sub.g1] -- -- (J/g/[degrees]C) [T.sub.g2]([degrees]C) 53 50 [delta][C.sub.p] [T.sub.g2] 0.60 0.47 (J/g/[degrees]C) [T.sub.m] ([degrees]C) 120 120 [delta]H [T.sub.m] (J/g) 0.96 0.96
|Printer friendly Cite/link Email Feedback|
|Author:||FAWCETT, A.H.; McNALLY, T.|
|Publication:||Polymer Engineering and Science|
|Date:||Jul 1, 2001|
|Previous Article:||Styrene Grafting Onto a Polyolefin in an Internal Mixer and a Twin-Screw Extruder: Experiment and Kinetic Model.|
|Next Article:||The Design of Conformal Cooling Channels in Injection Molding Tooling.|