Polyethylene Compounds Containing Mineral Fillers Modified by Acid Coatings. 2: Factors Influencing Mechanical Properties.
I. SUTHERLAND [**]
High molar mass medium density polyethylene (MDPE) compounds containing magnesium hydroxide filler, surface-treated by a range of fatty acid coatings of variable aliphatic chain length were injection molded and subjected to mechanical property evaluation. Surface treatment modifies yield stress and modulus, with property maxima observed close to the monolayer coverage. Acid-group terminated polyethylene (ATPE) coatings produced the highest yield stress, as a result of physical interaction with the matrix polymer. Retention of high-energy, ductile mode impact failure of unfilled MDPE was obtained when using short-chain decanoic acid coating, as a result of enhanced dispersion and reduced particle-matrix interactions promoting microscopic matrix yielding. A mechanism for enhanced Mg[(OH).sub.2] particle dispersion based upon surface energy and molecular adsorption has been proposed, which is consistent with the properties data derived. Fractographic analysis (SEM) has confirmed the predominance of matrix yieldi ng in compounds exhibiting enhanced impact resistance, while X-ray photoelectron spectroscopy (XPS) has also highlighted a different crack growth mechanism in MDPE compounds containing coated fillers, which are less prone to agglomeration. In addition, thermomechanical history during processing also modifies physical properties of MDPE/Mg[(OH).sub.2] composites to some extent. Anisotropic effects include molecular orientation and filler particle alignment induced by shear stress during injection mold filling, which predominate in compounds containing coated fillers. Overall, the application of organo-acid coatings reduces polymer-particle surface interaction and thermodynamic work of adhesion, leading to improved dispersion and enhanced properties. Appropriate property balances can be tailored by judicious selection of aliphatic chain length and addition level of fatty acid filler coatings.
In the first part of this communication (1), a review of the characterization of filler coating and subsequent processing of magnesium hydroxide/MDPE composites was presented, in order to determine factors responsible for mechanical property enhancement. Particulate additives increase the tensile modulus of a polymer to an extent determined by filler volume fraction (2-8), whereas compound ductility usually decreases with increased filler loading as a result of lower energy absorption (5, 9-11). The effect on yield stress is not so easily generalized. Most empirical models predict decreases in yield stress with increasing filler content (5, 6, 12), yet some publications report a yield stress maximum, followed by a decrease to below the datum yield stress of the unfilled polymer (10, 13). This type of response has been attributed to secondary effects such as heterogeneous nucleation, as observed in the case of talc additives in polypropylene (PP) (13). Particle morphology is also an important parameter in this respect, as observed previously for Mg[(OH).sub.2] fillers in PP by Cook and Harper (14, 15).
It is recognized that the surface character of mineral fillers and other reinforcing additives determines the extent to which physical properties of thermoplastics are modified (16). For example, stearic acid has been shown to render inorganic fiber surfaces hydrophobic, reducing fiber-fiber interaction and improving additive dispersion in HDPE (17). In contrast, maleated PE improves the adhesion between filler and matrix. The amount of surface treatment is also significant, since this is a function of the specific surface area, hence particle size and morphology of the fillers. Property maxima have been observed, suggesting that at high addition levels, unreacted treatments were having a plasticizing effect on the HDPE. Interfacial interactions in PP/Ca[CO.sup.3] systems were also investigated by Pukanszky et al. (18), using stearic acid and maleated polypropylene (mPP) treatments. Carboxyl groups in the mpp react with the hydroxyl groups on the filler surface and also, the mPP chain molecules physically ent angle with the PP matrix. Hence, stress can be transferred from matrix to filler very effectively, by a combination of chemical and physical interactions. With stearic acid as a filler coating, the carboxyl groups in the organic acid also react with the hydroxyl groups on the filler surface but the chains are not long enough to provide chain entanglement. It was concluded that for a high molecular weight surface treatment to be beneficial, it must be compatible and interactive with the matrix polymer.
According to Jancar (19), the tensile yield stress of particle-modified polymers is heavily dependent on the interphase and adhesion of the filler to the matrix. Theoretical 'zero adhesion' was achieved by the use of stearic acid, which reduces the thermodynamic work of adhesion between particles and polymer matrix. For Ca[CO.sup.3] and Mg[(OH).sub.2] fillers in PP, the stearic acid coating reduced the tensile yield stress, for all compositions. In the case of the Mg[(OH).sub.2] particles it was also shown that a critical filler volume fraction exists, after which the anisotropic, plate-like particles act as stress concentrators resulting in a dramatic reduction in yield stress. When stearic acid coatings are applied to the filler platelets, the position of the corresponding property reduction is shifted to a higher critical volume fraction, since the stearic acid promotes dis-agglomeration of the particulate filler. It was concluded that the uncoated filler produces an immobilized PP interlayer on the surfa ce of the filler and that this interphase is reduced by the addition of stearic acid. Enhanced interfacial adhesion was then investigated by using maleated polypropylene (mPP) to provide a chemical bond between filler and PP (20). It was shown how the yield stress increases rapidly with acid group content for all fillers studied (both uncoated and stearic acid coated) up to a critical concentration, then increases more gradually. It was suggested that this be attributed to the filler surface being totally covered by mPP molecules and/or that the adhesive strength of the interface is greater than the yield stress of the PP.
The nature of the polymer will determine its ability to undergo yielding, although in filled polymers, particulate additives are also responsible for modification of the localized stress concentration. Crack propagation was discussed by Fu and Wang (10), who found that the addition of uncoated Ca[CO.sub.3] to HDPE decreased the notched Izod impact strength. However, the addition of phosphate-treated Ca[CO.sub.3] increased the impact strength with filler addition level, an effect attributed to finer dispersion of filler particles. With agglomerated particles, the stress field in the polymer is concentrated around any isolated aggregate so that large cracks can propagate between the agglomerates easily and rapidly, causing premature, low-energy failure. The improved dispersion of the treated Ca[CO.sub.3] leads to a greater number of small localized crazes around the filler particles, which are able to interact if sufficiently close together. The cavitation around disper sed filler particles and matrix yielding thereby contribute to impact energy absorption (22); it has also been suggested that impact strength is independent of dispersion when using notched impact tests since, unless the agglomerates are very large, the notch creates a much greater degree of stress concentration. Falling weight impact tests (on un-notched specimens) are sufficiently sensitive to the state of dispersion to demonstrate the transition of brittle to ductile failure with improved dispersion. It is therefore beneficial to treat the filler surface with a dispersion aid rather than a coupling agent, in order to promote this mechanism.
The influence of fillers on the toughness and energy absorbing properties of polymers is of strategic importance to commercial developments but owing to complex stress distributions, the notched impact resistance of filled polymers is far more difficult to predict than modulus or yield stress. In general, filled polyner composites are often assumed to be more brittle than the unfilled polymer, interpreted by reduced crack propagation resistance as a result of inadequate filler dispersion and the formation of low cohesive strength agglomerates. Impact tests are often used to measure energy absorption, since matrix yield stress increases at high strain rates, thereby promoting brittle-mode failure in many circumstances. However, if a polymer can absorb most of the impact energy by small-scale yielding (for example by craze formation at the crack tip), then crack propagation may be retarded resulting in a transition to ductile-mode failure. Reference should be made to more specialized texts for an in-depth appr eciation of these complex phenomena (21).
Inter-particle spacing has been discussed by Fu and Wang (23). It is postulated that when the matrix ligament thickness is reduced to a critical level, yielding zone interaction becomes more feasible and a brittle-ductile transition occurs. Inevitably when imperfect distribution of particles occurs, there is a distribution of ligament thicknesses. If short-range ligaments predominate, yielding can occur over the area of deformation. Enhancements in impact resistance have been determined for ultrafine calcium carbonate
particles in thermoplastics (24), although it is recognized that effective additive dispersion is essential (24, 25). Heterogeneous nucleation effects may offset the measured improvements in impact resistance (25, 26), although it appears that high aspect ratio additives may be more influential, in this respect (14, 15, 25). Small particle size Mg[(OH).sub.2] fillers have also been found to generate optimum impact toughness in PP (14) but once again, impact resistance reduces dramatically if mu ch higher filler loadings are incorporated .
Overall, it is clear that the mechanical properties of artifacts molded from particle-filled polymers are influenced by a number of complex and interactive variables. These relate not only to the type and concentration of selected polymer(s) and additives but also to the stages of processing from filler preparation, coating and subsequent additive dispersion induced at the compounding stage. Dispersive mixing enhancement created by judicious selection of filler coatings is the key to achieving a good balance of rheological properties in MDPE/[Mg(OH).sub.2] composites (1). The main objective in Part 2 of this communication is to progress the chain of knowledge by investigating the influence of filler coatings (aliphatic chain length and addition level) on selected mechanical properties of injection molded magnesium hydroxide filled MDPE composites, and to identify the factors primarily responsible for physical property enhancement.
A high molecular weight grade of MDPE (supplied by BP Chemicals, Grangemouth, U.K.) was used throughout the study. The magnesium hydroxide filler was supplied by Premier Periclase Ltd. (Ireland); grade DP393 has a high aspect ratio, planar 'platey' character with a primary particle size of around 0.8 [micro]m, corresponding to a specific surface area of 13 [m.sup.2]/g. Different concentrations of fatty-acid surface coatings have been added to the [Mg(OH).sub.2] filler; these vary in terms of the aliphatic chain length and include decanoic acid [C-10], stearic acid [C-18], behenic acid (C-22) and an acid-group terminated polyethylene (ATPE) of molecular weight 5000 g/mol. supplied by AlliedSignal Corporation (Belgium). Full details of individual material specifications (including a micrograph to show the [Mg[OH].sub.2] particle morphology) and related experimental procedures have been provided in the first part of this communication .
For the results presented here, the filler concentration was held constant at a nominal level of 30% (by weight). Following the twin-screw compounding stage (1), MDPE-[Mg(OH).sub.2] composites were injection molded using a Negri Bossi NB55 machine (55 tonnes maximum locking force; 32mm diameter screw) to produce both dumbbell specimens for uniaxial tensile tests and in a separate mold, single edge notched (SEN) Izod impact test bars. Unfilled MDPE was also molded to produce control specimens. In each case, maximum melt temperature was 22000 and the mold temperature surface was controlled by circulating water at 1500. Many individual settings (pressures, speeds, phasetimings) were controlled by the Dimicard 200 programming system; full details of all parameters are given elsewhere .
All molded specimens were tested according to BS2782: Part 3, Method 320A (1976) using a Lloyd Type L2000R tensile testing unit. A constant crosshead velocity of 5mm/minute was selected, and force measurements were taken via a 2.5kN load cell. A laser extensometer was used to measure extension at yield and at failure, while a mechanical, 'clip-on' type extensometer was utilized in low-strain tests to obtain secant modulus values. The force limits over which the modulus was measured were set between 400 and 700 N, for all specimens. All parameters studied and associated data, including low strain modulus and stress, strain and energy absorption data at the yield point and at ultimate failure, were computed using Dapmat software.
Instrumented Impact Testing
A Rosand IFW5 instrumented falling-weight impact (IFWI) tester equipped with a microprocessor-based data capture system was used to measure the impact toughness of MDPE-[Mg[OH].sub.2] composites at ambient temperature. Injection molded, single edge-notched (SEN) Izod-type test bars (molded-in notches) were used for all impact tests performed in three point bending mode. The drop height of the impactor determines the impact speed, and this was set at 460 mm to produce an impact velocity of 3 m/s. The impact head has a mass of 26 kg and the span of the sample holder was set to 40 mm. Force data are taken throughout impact (5 ms, approx.) and if excess energy conditions are assumed (zero deceleration of impact head), calculations of force can be translated as a function of deflection, allowing energy absorption to be calculated by integration. For all tests, eight specimens were tested and the results were averaged.
Fracture Surface Characterization
The morphologies of selected specimen surfaces were studied using a Cambridge Stereoscan 360 scanning electron microscope (SEM). Low magnification conditions were used initially to obtain an overall assessment of the fracture surfaces, and then much higher magnifications were adopted to view topographical features, composite microstructures and damage zones.
In addition, semi-quantitative elemental analysis of failure surfaces was obtained using X-ray photoelectron spectroscopy (XPS). The instrument used is a VG Escalab Mark 1 and the samples were irradiated with X-rays from an Al k[alpha] source with energy 1486 eV. Only the parts of the specimen across which failure had occurred were considered and to obtain statistically meaningful information, 15 samples of each of two compounds were analyzed. Use of this technique to characterize the fracture surfaces and particle-polymer interactions in filled polymers is a relatively new development and, to our knowledge, has not been published previously.
Differential Thermal Analysis
Crystallinity measurements were made on small samples of all compounds machined from the center portions of the tensile bars, using a DuPont Thermal Analyzer (Model 2000) fitted with a differential scanning calorimeter (DSC) cell. The samples were scanned at temperatures between 20[degrees]C and 200[degrees]C at a heating rate of 10[degrees]C/minute, held in the melt state in a 10-minute isothermal conditioning phase, then were cooled at the same rate to ambient temperature. All data from mineral-filled compounds were normalized to account for inorganic content and PE crystallinity values were determined from heat of fusion data, assuming the heat of fusion of a 100% crystalline polyethylene to be 286.74 J/g (28).
In order to compare and interpret mechanical testing results more exactly and to attribute changes observed to the presence of fillers and/or coatings, a set of unfilled MDPE samples were produced that had been subjected to the same process history (compounding and molding) as all filled polymer compounds. A selection of filled and 'control' specimens were then analyzed by a number of techniques, to measure properties that are sensitive to molecular weight variation. These included gel content determination to measure crosslinking (using xylene at 138[degrees]C, according to ASTM D2765), high temperature gel permeation chromatography (GPO) (Millipore-Waters 150CV instrument) and indirect shear flow techniques (melt flow index at 190[degrees]C and capillary rheometry at 210[degrees]C).
Wide-angle X-ray diffraction (XRD) was carried out on a small selection of samples in order to assess any preferential alignment of filler particles in the MDPE compounds. Diffraction patterns were obtained using an X-ray generator operating at 40 kV, with a Phillips goniometer. CuK[alpha] radiation was used (wavelength, [lambda] = 0.1542 nm) and the diffraction intensity was measured over a Bragg angle (2[theta]) range of 12[degrees]-75[degrees].
RESULTS -- MECHANICAL PROPERTIES
In this section, mechanical properties data are presented as a function of composition, with systematic changes in fatty acid coating type and addition level. In all these results, comparison can be made with an unfilled 'control' sample. Because of the changes in molecular structure induced at the twin-screw compounding stage (a point addressed later in the paper, interpreted with rheological data), the control (unfilled MDPE) specimens were also prepared with an equivalent process history, in order to ensure comparative validity.
Properties in Tension
Results obtained from ambient temperature tensile testing are shown in Figs. 1-3, expressed as a function of filler coating type and addition level. Addition of 30% uncoated filler enhances the low-strain modulus of MDPE (Fig. 1) significantly. When coated fillers are incorporated into MDPE compounds, modulus generally increases further. Stearic acid increases modulus marginally, but different levels of coating show no further effect. Behenic acid produces MDPE compounds with very high moduli, particularly at higher addition levels. High molar mass ATPE coating produces a compound with the highest recorded modulus, at the lowest coating levels (5%). The degree of scatter in the experimental data is relatively low; as shown by the error bars in Fig. 1; variation appears to increase with the level of modulus enhancement observed. Overall, the optimum modulus values consistently occur at the lowest coating level of those studied, for this grade of magnesium hydroxide filler in MDPE.
Analogous data for yield stress are shown in Fig. 2, and once more, clear trends have emerged with respect to the chosen type of fatty acid coating. All MDPE compounds failed by ductile yielding and the yield stress increases with increasing aliphatic chain length of coating. Relative to the unfilled MDPE, the yield stress decreases marginally on the addition of uncoated filler. Low coating levels (5%-6%) initially enhance yield stress, but in general, further increases in coating concentration beyond the monolayer range decrease the yield stress, for all coating types. Only with the use of behenic acid and the polymeric ATPE coating on Mg[(OH).sub.2] was it possible to generate yield stress values higher than those of the unfilled MDPE.
Figure 3 describes the ability of the MDPE compounds to deform under uniaxial tensile stress and absorb strain energy before ultimate failure. In comparison to unfilled MDPE, the addition of uncoated Mg[(OH).sub.2] filler results in a small decrease in energy absorption. Use of coatings decreases failure energy much further, but higher coating addition has no appreciable effects. Use of behenic acid coating has produced compounds of lowest ductility and with exceptionally low variability (Fig. 3). Different compound viscosities and subsequent molecular orientation effects are likely to be responsible for the apparent lack of ductility in compounds containing well-dispersed coated fillers, a point that is discussed in a later section.
Properties Under Impact Loading--Instrumented Tests
Impact testing was performed under excess energy conditions, for which zero deceleration of the impact head is assumed so that force-time can be transposed directly into force-deflection data, allowing energy absorption to be computed by integration. Izod specimens contained single, edge-notches that were molded-in, rather than being machined from unnotched specimens. Inevitably, therefore, there will be a degree of interaction between the direct effects of composition and the influence of processing on molecular structure. For example, the magnitude and direction of molecular orientation at the defect tip will have a direct influence on the failure process. Following the impact tests, the fracture surfaces of selected specimens were observed visually (macroscopic failure mode) and by SEM (microscopic effects) in order to characterize features relating to the mechanisms of failure.
Instrumented impact traces initially mirror the force required to deform the rectangular-section specimens under flexural stress; this reaches a maximum before dropping suddenly to zero if fast fracture occurs, or if failure is predominantly by ductile yielding, force decreases more gradually and over a much wider deformation span. The effect of filler coatings on peak (maximum) force is shown in Fig. 4. With the addition of 30% uncoated magnesium hydroxide, peak force decreases dramatically compared to the unfilled MDPE, which behaves as an inherently ductile polymer under these conditions. However, the introduction of filler coatings modifies the qualitative response to attain greater plastic yielding and ductility, which is reflected to some extent in the enhanced peak force data. For the three short-chain fatty acids, the level of coating has no significant effect. Compounds containing Mg[(OH).sub.2] coated with ATPE were generally more brittle, yet still generated peak force data greater than that assoc iated with uncoated filler. There is some similarity between the respective plots in Fig. 2 (tensile yield stress data from low strain rate tests) and Fig. 4 (net section yielding data in high strain rate flexure); the main qualitative difference is the response of compounds containing ATPE-coated filler, which are characterized by the highest tensile yield stress (Fig. 2) but exhibit a brittle-dominated failure process under high velocity impact loads (Fig. 4).
Since brittleness is often induced in filled thermoplastics as a result of strain rate sensitive yielding behavior, it is often the energy absorbing characteristics that govern suitability for load-bearing applications and that are more meaningful for comparative purposes. Therefore, of greater relevance to the overall impact resistance and applications-related toughness are the ultimate failure-energy absorption data presented in Fig. 5. This parameter describes the degree of plastic yielding that has occurred on impact and is considerably more discriminating between the responses of all filled MDPE compounds. Addition of 30% uncoated filler causes a large reduction in the overall energy absorption, but the application of filler coatings modifies these trends in a systematic manner, according to which coating is used and to what extent. The shortest chain coatings (decanoic acid) enhance energy absorption to the greatest extent, so that the toughness of these compounds is not significantly less than unfille d MDPE; retention of ductile-mode impact response in compounds containing 30% MgP[(OH).sub.2] filler is a highly significant result. As observed elsewhere (for example the tensile yield stress data shown in Fig. 2), property maxima occur at the lowest coating levels of those studied, corresponding approximately to a monolayer addition level of surface coverage for the ultrafine grade of Mg[(OH).sub.2] used. The yield stress data assist in the interpretation of the impact failure energy results (Fig. 5), since the compounds with lowest yield stress are clearly more likely to exhibit microscopic yielding and large-scale plastic deformation when subjected to triaxial stress under impact loading conditions.
Impact Failure Mechanism and Fracture Surface Morphology
The qualitative failure mode summary shown in Table 1 reflects the impact testing results shown in Figs. 4-5, and a clear correlation exists between impact failure energy and the overall failure mechanism. Mode I failure represents energy absorption levels in excess of 7 Joules (see also Fig. 5), Modes III and IV characterize predominantly brittle failures (energy absorption less than 2 Joules) and Mode II failures are intermediate to these extremes. Clearly, the propensity to undergo plastic yielding is key to achieving acceptable impact resistance in injection molded MDPE- Mg[(OH).sub.2] composites.
Figure 6 shows macroscopic failure surfaces, which give qualitative confirmation of the data trends from the instrumented impact tests. Unfilled MDPE is inherently ductile (Mode I failure), shows gross plastic yielding around the tip of the defect, and impacted specimens do not break completely. in contrast, the fracture surfaces for compounds containing uncoated Mg[(OH).sub.2] and ATPE-coated filler show yielding only in the plane stress dominated zones close to the edges, but the majority of the crack propagation area is relatively featureless at low magnification, showing a predominantly brittle response. The other samples (fatty-acid coated fillers) show an intermediate response, but the most striking feature is the retention of ductile-mode impact failure observed in the specimen containing filler coated by 6% decanoic acid. The extreme levels of light scattering are created by the plastic deformation in the matrix MDPE polymer, an observation that correlates with the failure energy data shown in Fig. 5.
More precise, small-scale interpretations were derived from scanning electron micrographs of failure surface topographies. Respectively, Figs. 7-8 are micrographs of unfilled MDPE and MDPE filled with 30% uncoated Mg[(OH).sub.2] obtained at a magnification of 40X. Observations at low magnification create very clear distinctions between the fibrillar morphology of the unfilled MDPE in Fig. 7 (characteristic of ductile polymers developing high localized plastic deformations as part of yielding) and the contrasting brittle-mode failure topography in Fig. 8. In the latter micrograph, yielding is evident only on a microscopic scale (see also Figs. 9-11) in locations specific to the dispersed particulate additives.
Further, more detailed discrimination between brittle fracture surface morphologies can be derived from high magnification electron micrographs (Figs. 9-11). Small scale yielding of the polymer around the filler particles is visible in Figs. 9 and 11 (Mode m and IV, respectively), resulting in a pseudo-network structure of plastically deformed material in each of these examples. It is likely that the plastic deformation zones around individual filler particles are able to interact if high levels of dispersion are present, according to the energy absorption mechanism described by Fu and Wang (10, 23). The volume of polymer that undergoes yielding by this mechanism determines the total energy absorption and the ultimate mode of failure. Fig. 10 exhibits the fractography of a Mode II impact failure [(mg(OH).sub.2] coated with 6% behenic acid) for which more highly developed localized strains are made evident by the fibrillar polymer network. Significantly, there are also more detached filler particles observed in Fig. 10 (in comparison to Fig. 9), which reinforces the theoretical and practical observations of reduced polymer-particle adhesion when coated fillers are used. Different levels of particle-matrix interaction influence the amount of immobilized PE on the filler surface; this has previously been shown to control the yield stress of particulate-modified PP composites and the critical volume fraction associated with ductile-brittle transitions (19, 20).
Fracture Surface Analysis by XPS
XPS analysis gives semi-quantitative elemental data from spectra, in this case based upon carbon is, oxygen is and magnesium 2p peaks. We have chosen to use magnesium 2p peak data to characterize the concentration of exposed [Mg(OH).sub.2] filler on the composite fracture surfaces. Some comparative data are given in Table 2, based upon averages of fifteen individual measurements. Statistical T-test analysis has given a 99.8% confidence level to suggest that these results and the specimen fracture surfaces are significantly different. Assumptions made in the calculations are that the particles are spherical and are perfectly dispersed (27). The reduced magnesium concentration on the fracture surface of compounds containing coated filler might initially be attributed directly to the presence of the organic coating agent on the magnesium hydroxide particles. This scale of difference cannot be entirely due to different levels of filler alignment. In addition, lower interfacial adhesion in the compounds containin g coated filler [immersion calorimetry data (1)] creates loosely bound particles that are more easily detached during impact. Higher concentrations of magnesium on the surface of the specimen containing uncoated filler suggests that the preferred locus of failure during impact is more predominantly at the polymer-filler interface and that crack growth occurs across any filler agglomerates that lie across the crack propagation path.
PROCESSING-INDUCED MICROSTRUCTURAL CHANGES
Mechanical properties of filled polymer composites not only are dependent upon composition but are also influenced by a number of related factors induced by processing, thermomechanical history and structure development. These include changes in molecular weight, orientation of polymer chains and particulate additives, crystallinity development and other structural and interfacial phenomena (1). In order to make a more detailed interpretation of the mechanical property changes presented earlier, further characterization of the MDPE polymer and filled compounds was therefore carried out.
The thermo-oxidative stability of polyolefins is known to be related to specific process conditions and thermomechanical history. Effects of twin-screw compounding and injection molding have therefore been characterized by GPC and gel content analysis, in order to determine changes in molecular weight and crosslinking: results are presented in Table 3. Weight average molecular weight ([M.sub.W]) decreases following processing, particularly after the shear intensive twin-screw extrusion process, in which high shear stress distributions are developed as a result of the high molar mass and shear viscosity of the MDPE polymer. The reduction in ([M.sub.w]) following injection molding appears to be greater for compounded material, suggesting that an autocatalytic degradation mechanism is active.
Measured concentrations of insoluble gel (Table 3) give an indication of crosslinking in high polymers; the gel content of MDPE increases slightly following processing. The overall amounts of crosslinked material are relatively low in all samples, yet these are likely to be sufficient to influence the GPC (and apparent ([M.sub.w]) data and flow properties of the filled compounds. Changes in molecular structure have a subsequent influence on melt flow properties, which have been investigated using melt flow index (MFI) and capillary rheometry. The low-shear MFI results (Table 3) are consistently very low, confirming that the polymer has a high shear viscosity under these conditions. MFI subsequently decreases further following each processing phase, which suggests that the developed crosslinking is more influential than the direct reductions in ([M.sub.w]). Cogswell has suggested that the dependence of shear viscosity upon ([M.sub.w]) or [M.sub.n] is determined by the chain backbone flexibility of the polymer (29).
Capillary rheometry data for each MDPE are shown in Figs. 12 and 13, in terms of both shear and extensional flow properties. Fig. 12 is consistent with the MFI data (Table 3), and suggests that the virgin MDPE has a slightly lower shear viscosity than the three processed samples, at low shear rates. However, no significant differences are evident between the shear flow behavior of each processed sample of MDPE, nor between any compound at high shear rate. In contrast, the extensional viscosity data (Fig. ] are more discriminating between the samples, with the twin-screw extruded polymers of lowest [M.sub.2] ([MDPE.sup.2] and [MDPE.sup.4]) giving the lowest elongational viscosity. Elongational viscosity is dependent upon the pressure drop for melt convergence into orifice dies, which is known to be sensitive to structural changes, particularly to [M.sub.w] (29). Clear differences therefore exist in the molecular structure and flow behavior of the MDPE polymers, following different phases of processing, that c an be interpreted in terms of modified molecular structure. These are not always systematic changes and the exact interpretations will require further research; however, it is likely that these variations will influence the microstructure and mechanical properties of filled polymer composites to some extent.
Filler Particle Orientation
Wide-angle X-ray diffraction traces were obtained from a systematic series of injection molded [Mg(OH).sub.2] filled MDPE specimens, to characterize the influence of stearic acid filler coatings. Some analogous experiments were also performed on compression molded sheets from the same compounds, for comparative purposes. From the raw XRD (intensity versus Bragg angle) traces, peak intensity (I) data were recorded for the 001 and 101 crystallographic planes in the filler (at angles of 18.4[degrees] and 37.8[degrees] respectively) and also for the 110 and 200 planes in MDPE. Table 4 gives a summary of all XRD data, expressed relative to the most intense diffraction peak (110 planes in the orthorhombic unit cell for PE). The data were normalized to eliminate the effects of peak intensity variations between samples, which occur as a result of differences in sample preparation. XRD data for unfilled MDPE show that the 200 diffraction peak becomes less intense (relative to 110 peak) following injection molding, an observation that is also true for most of the Mg[(OH).sub.2]-PE specimens. The plane of view is parallel to the molded films (compression molding) and to the flow direction in the injection molded specimens. This result describes different orientation of the PE crystallites, since the molecular chains lie along the c-axis of the unit cell.
The most striking observation in the particle-filled compounds is the increased peak intensity of the 001 crystallographic planes ([I.sub.001]) of the filler, in the injection molded specimens: peak intensities of the 101 planes ([I.sub.101]) are correspondingly less intense. Peak intensity ratios ([I.sub.001]/[I.sub.101]) vary between 1.3 and 3.9 for compression molded specimens, but are typically an order of magnitude higher (ranging from 31.6 to 47.9) for injection molded samples. The 001 plane represents the hexagonal part of the crystallographic unit cell, and increased ([I.sub.001])/([I.sub.101]) ratios indicate higher alignment of the plate-like filler particles, in the flow direction of the injection molded specimens. Further information on this experimental technique and the dependence of particle orientation upon wall slip phenomena have been published elsewhere . Since particle orientation is flow-induced, there are clearly similarities between the filler alignment data presented here and the polym er orientation data presented in Part 1 of this communication, with respect to the importance of compound rheology during injection molding. Peak intensity ratio reduces at coating levels beyond 5%, consistent with the reduced molecular orientation shown by thermal reversion data . Such alignment effects for anisometric fillers are known to have a strong influence on the mechanical properties of molded artifacts [19, 24, 31, 32] and inevitably contribute to the property differences exhibited here.
Differential Thermal Analysis (DTA) was used to investigate the influence of composition on the crystallizability of the filled MDPE compounds. In the subsequent results relating to the influence of coatings on Mg[(OH).sub.2] fillers and contrary to other reported instances of heterogeneous nucleation in particle-filled compounds [13, 25, 26], it can be seen how the melting temperature ([T.sub.m]), the onset temperature for recrystallization ([T.sub.sc]) and the recrystallization peak temperature ([T.sub.c]) of MDPE are largely unaffected by the presence of Mg[(OH).sub.2] fillers or fatty acid coatings (Table 5). In contrast, filler coatings significantly modify the recrystallization, subsequent heat of fusion and PE crystallinity data. Trends have emerged with respect to both the type and addition level of fatty acids, and the effect of these on MDPE crystallinity is summarized graphically in Fig. 14. Decanoic acid (the shortest chain coating considered) produces the highest levels of PE crystallinity, whil e behenic acid (the longest of the 'short chain' coatings) induces levels of crystallinity lower than that in the compound containing uncoated filler. Crystallinity data for the high molar mass ATPE coating is intermediate to the trends for decanoic and behenic acids. Increases in PE crystallinity with the incorporation of Mg[(OH).sub.2] particulate additives have been reported elsewhere [33-35], although it was suggested that the mineral structure of the filler is more important than addition level . Magnesium hydroxide has been found to act as a heterogeneous nucleating agent in PP, for which crystallite growth occurs on preferred crystallographic planes . In contrast to the results presented here, Mitsuishi  concluded that polymer crystallinity is independent of coating chain length.
Despite the clear trends with respect to filler coatings that have emerged (Fig. 14), it is very unlikely that the measured crystallinity changes are direct determinants of the tensile and instrumented impact properties of mg[(OH).sub.2]-filled MDPE compounds. For example, decanoic acid induces the highest levels of PE crystallinity (Fig. 14), which might be expected to enhance stiffness and strength but reduce ductility. In practice, this is opposite to the measured responses (Figs. 1, 2-5). Therefore, other mechanisms dearly have more influence on mechanical properties and override the direct effects of crystallinity variations due to the presence of coated fillers.
A MECHANISM FOR DISPERSION AND PROPERTY ENHANCEMENT, BASED UPON SURFACE ENERGIES OF COATED FILLERS
It is known that the type, concentration, and morphology of particulate additives have a strong influence on the mechanical properties of molded thermoplastic composites, to an extent determined by the interfacial characteristics, the presence of coatings, and filler dispersion (e.g. 19, 20). In this section, we propose a new mechanism to interpret filler dispersion in terms of surface energy, in order to identify a molecular basis upon which the mechanical property trends relating to dispersion effects can be more clearly understood.
It is well known that at a molecular level, coating the filler particles with an aliphatic chain will reduce the energy of interaction between filler and polymer. For a non-polar polymer, dispersion forces dominate these interactions and the thermodynamic work of adhesion is given by:
[W.sub.A] = 2 [([[gamma].sub.F] * [[gamma].sub.P]).sup.1/2] (1)
where [[gamma].sub.F] and [[gamma].sub.P] are the dispersion components of the surface excess free energy of the filler and polymer respectively. [[gamma].sub.P] is about 32 [mJm.sup.-2] ; however, it is more difficult to obtain a value for [[gamma].sub.F]. Filler surfaces that have been exposed to air will have some hydroxylation and adsorbed impurities. A value of 58 [mJm.sup.-2] has been measured by inverse gas chromatography . This is significantly lower than that calculated for ideal, clean surfaces but is probably more representative of the situation in practice. Similarly, values of about 28 [mJm.sup.-2] have been obtained for stearic acid coated fillers , which is in good agreement with theory and other measurements . Using these values, the thermodynamic work of adhesion for uncoated and coated filler are estimated to be 86 [mJm.sup.-2] and 60 [mJm.sup.-2] respectively. Thus, coating reduces the thermodynamic work of adhesion between filler and polymer by an estimated 30%. [W.sub.A] o f a coated filler is now slightly below that expected for polymer/polymer interactions of 64 [mJm.sup.-2].
The improvement in properties when using fillers with inert coatings is often attributed to improved dispersion of the filler particles arising from reduced filler/filler interactions. While these interactions will be substantially reduced on coating, the geometry of the filler will mean that there will be very few points of contact. It is believed that the following mechanism is probably more important. It is known that an immobilized layer of polymer of several nanometers in thickness exists around an uncoated filler particle owing to the strong filler-polymer interaction. A typical polymer molecule of molar mass 140,000 g * [mol.sup.-1] will have an overall chain length well in excess of the size of the individual magnesium hydroxide particles. For the uncoated filler, molecules would be expected to adsorb in trains, loops and tails, as shown schematically in Fig. 15a. Molecules that span more than one individual filler particle in the aggregate would tend to bind the aggregate together. In the case of a coated filler (Fig. 15b), the energies of interaction are now such that the formation of trains on the surface is no longer favored. The aggregate is no longer bound together, making it easier to disperse the particles by decoupling. Improved dispersion is thought to be responsible for some of the effects seen in this work, notably the yield stress data (Fig. 2) and impact behavior (Figs. 5, 6), where the lower yield stress of compounds containing decanoic acid coated Mg[(OH).sub.2] filler promotes plastic yielding and ductile-mode failure under high velocity loading.
The chain lengths of the organic acids used here are well below that required for chain entanglement with the matrix polymer, although they may still have some influence on the mechanical properties near the interface. The reason for the very different behavior of the acid functionalized PE is that it can chain entangle, giving higher yield stress (Fig. 2) and more brittle behavior under impact loads (Fig. 5), despite the high levels of dispersion in these compounds (1).
From consideration of the energies of interaction alone, all acids (excluding AT-PE) would be expected to behave in a similar way up to the point where their surfaces are completely covered, i.e., theoretical mono-layer coverage. For the most part, this would appear to be the case. After monolayer coverage has been established, there will be salts of the various acids present in the polymer. The possibility therefore exists of these salts functioning as internal lubricants and affecting the mixing process, dispersibility and subsequent flow behavior and orientation effects. This additional complexity accounts for the differences in mechanical behavior when coatings are added in quantities beyond the monolayer range.
Having outlined a new mechanism of dispersion based upon surface energy and consideration of scale effects to account for different levels of interaction, it is useful to consider the implications on mechanical properties. Theoretically, as the degree of dispersion increases, the inter-particle ligament thickness decreases and approaches the critical thickness that promotes yielding. A mechanism to relate improved dispersion of rigid particles to enhanced toughness in composites has been presented by Suetsuga (22) and is consistent with the results and interpretations presented here. For compounds containing coated Mg(OH)2 fillers (characterized by a relatively low work of adhesion and by excellent dispersion), microvoid formation around the filler particles is facilitated, which favors inter-particle matrix yielding. Once multiple yielding becomes feasible in these composites, ductilemode impact failure is promoted to a much greater extent not only because of the relative absence of agglomerates, but also b ecause the driving force for crack growth is reduced when the deformation energy is dissipated.
INFLUENCES ON MECHANICAL PROPERTIES--AN OVERVIEW
Most research studies investigating physical properties of filled polymer composites inevitably focus upon effects due to the concentration (volume fraction) of filler particles, and there have been many attempts to model relationships between composition and physical property changes (5, 6, 12, 16, 40). Empirical models based upon filler volume fraction are not directly relevant to the research reported here, however, since nominal filler concentration was kept constant throughout. The investigations described in this two-part communication extend research knowledge in this area by studying new coating materials, a series of fatty-acids that react on to the surface of magnesium hydroxide filler, but have no direct chemical interaction with the matrix polymer. In this section, it is the intention to review the most important trends that have emerged from the research into coatings of variable aliphatic chain length and also to interpret the mechanical property variations in terms of the measurements and char acterization data reported in both Parts 1 and 2 of this communication.
Filler Coating--Analysis and Dispersion
Initially, it has been demonstrated how the coating application process can be characterized by spectroscopic analysis (1); only by optimizing this initial processing stage can filler dispersion be enhanced during compounding, an important prerequisite for subsequent processing and mechanical property determination. Once coated, there is a reduction in interaction energy and work of adhesion between mineral fillers and the organic polymer matrix, as detailed by the mechanism presented in the previous section and confirmed in practice by immersion calorimetry data (1). Improvements in mechanical properties obtained following coating (especially modulus, yield strength and impact energy data; Figs. 1, 2, 5) cannot therefore be attributed to improved interfacial adhesion. On the contrary, the coating has an altogether different effect, in promoting dis-agglomeration (particle decoupling) and wetting by the non-polar PE matrix, caused by a reduction in thermodynamic work of adhesion (19). Therefore, the importan ce of improved filler dispersion should be emphasized, and the observed improvements in properties can be directly attributed to the enhancements in particle dispersion created by the use of filler coatings. Further evidence supporting the importance of dispersion and process history arises from the lack of any clear correlation in Fig. 16, which demonstrates that the property enhancements are not directly attributable to the concentration of acid groups, even though the filler surface coating mechanism is identical in each case.
The improved dispersion of coated fillers has also been demonstrated by qualitative analysis of topographical images of compound sections etched in HCI (1). Moreover, when coated fillers are compounded into MDPE, specific energy data derived from the twin-screw extruder generally decrease, demonstrating that the improved dispersion can be achieved with lower energy input. Polymeric ATPE coatings are exceptional in this respect; while excellent levels of dispersion are achieved, specific energy increases relative to the use of uncoated filler, as a result of greater physical interaction between the long aliphatic chain coating and the MDPE polymer.
At coating levels exceeding monolayer, FTIR analysis has demonstrated the existence of a continuous coating reaction, made possible by particle attrition as the organic salt detaches from the filler surfaces (1). This continued reactivity of the fatty acids in a mechanically intensive mixer results in property changes at the highest coating levels. Coating thickness measurements made by XPS suggest that the aliphatic chains of low molar mass fatty acids lie predominantly perpendicular to the filler surface, following coating, a result that carries implications regarding the definition of 'monolayer,' in these systems.
Individual Fatty-Acid Coatings
Decanoic acid, the coating studied with shortest aliphatic chains, gives relatively low modulus and yield strength relative to other compounds containing coated fillers (Figs. 1, 2) but the impact resistance of MDPE composites modified by decanoic acid coated fillers is particularly high. Mode I failures were recorded, and total energy absorption is not significantly less than for the high molar mass matrix MDPE polymer (Fig. 5). Excellent dispersion is thought to be key to this property enhancement, since the existence of low cohesive strength agglomerates would otherwise promote rapid crack growth under these conditions. The relatively low yield stress of these compounds allows net-section yielding at the root of the notch in Izod specimens, in spite of the triaxial stress state. Attainment of high impact energy by enhanced particle dispersion created by use of coating agents is consistent with the theories proposed by Fu and Wang (10). However, it does not support a suggestion that impact strength is inde pendent of dispersion, when testing notched specimens (22), possibly because the notches were moldedin, for the research reported here. Average coating thickness for decanoic acid are relatively low [10-14 A depending upon concentration (1)] in comparison with the other coatings, an observation that influences the degree of internal lubrication and shear viscosity of these compounds.
As coating chain length increases, the general trends are for the maximum modulus and yield stress to increase, when properties are studied as a function of coating concentration (Figs. 1, 2), offset by a reduction in impact energy (Fig. 5) and a qualitative failure mechanism, which becomes more dominated by brittle crack growth. With respect to the three short-chain acids, there is a clear correlation between mechanical properties data and particle dispersion, if it is assumed that the specific energy data from compounding (1) are a direct indicator of dispersive mixing.
Application of the polymeric ATPE coating has produced results that are in some respects significantly different from those relating to short chain fatty acid coatings, even though the reaction mechanism with the inorganic filler is identical. Coating thickness cannot be determined, as a result of the existence of much longer aliphatic chains and their interaction with MDPE. These physical interactions create greater entanglement densities that increase shear viscosity and specific energy during compounding; as a result, excellent levels of filler dispersion are achieved (1). However, solid-state properties such as yield stress are also significantly increased by such physical interactions, although thermal analysis data have generated no direct evidence to support the existence of co-crystallization. This is a similar effect, though from a purely physical origin, to the influence of maleated PP in the PP/Ca[CO.sub.3] composites reported by Pukanszky et al. (18), who also suggest that compatibility and physi cal interaction must be attained in order that high molar mass coatings function effectively. These criteria appear to be fulfilled when using ATPE-coated Mg[(OH).sub.2] in MDPE; it is the enhanced yield stress of these compounds (shown in Fig. 2) that determines the failure mode and reduced impact resistance (Fig. 5) that have been observed.
Effects of Processing, Orientation and Anisotropy
Despite the importance of particle dispersion, solid-state properties of artifacts manufactured from filled polymer composites are also influenced by thermomechanical (melt-state process) history and by development of microstructure during the non-isothermal flow and cooling phases of injection molding. Rheological analysis has shown that shear viscosity increases when [Mg(OH).sub.2] fillers are incorporated but in a nonsystematic manner, since coatings have been shown to induce internal lubrication effects in [Mg(OH).sub.2]-filled MDPE compounds. The occurrence of wall slip also modifies shear flow behavior of MDPE containing uncoated [Mg(OH).sub.2]. Such rheological changes modify the development and distribution of shear stress during injection molding, resulting in systematic variations in polymer orientation, as demonstrated by thermal reversion data . An earlier section of this paper also described the changes to molecular structure, particle orientation and crystallinity arising as a consequence of process history.
It is known that the degree of polymer and/or filler particle orientation is likely to have a considerable effect on the mechanical properties of injection molded components. A summary graph (Fig. 17) of yield stress plotted against molecular orientation (expressed by the reversion data initially presented in Part 1) demonstrates only a limited degree of correlation between these parameters; reversion is clearly neither the only nor the dominant factor on yielding behavior. Yield stress for this grade of isotropic MDPE would be expected to lie in the range between 16 and 18 MPa, depending upon extension rate. The degree of correlation in Fig. 17 is not high, showing that other factors (formulation, polymer-filler interaction, dispersion) are predominant. An equivalent assessment of polymer orientation on impact energy absorption reveals a stronger correlation (Fig. 18), due to the orientation (melt flow) direction being perpendicular to the defect growth direction during impact. The data points are linked ac ross a brittle-ductile transition, for all except two series of compounds. First, MDPE compounds modified by decanoic acid coated fillers exhibit ductile-mode impact failure and high energy absorption, regardless of the level of reversion. As discussed earlier, this is due to filler dispersion and a relatively low yield stress. In contrast, ATPE-modified compounds have low energy absorption due to a higher yield stress (Fig. 2) and show a high degree of reversion (regardless of ATPE addition level) because of the enhanced viscosity and melt elasticity induced by the physical interactions between MDPE and the aliphatic ATPE chains.
Anisotropic filler alignment effects have also been observed by XRD, following injection molding (Table 4); these are induced by rheological variations during in-cavity flow and are not considered to be as important as the direct enhancement of particle dispersion created by filler coatings. It should be recognized that for high aspect ratio particles, dispersion effects and anisotropic properties are interdependent.
Overall, a series of new fatty acid coatings has been studied, and it has been established that a number of important, interactive parameters influence the mechanical properties of [Mg(OH).sub.2]-filled MDPE composites. These originate from each successive phase of processing, involving filler coating, compounding dynamics and flow-induced structure development during injection molding. It is not appropriate to relate formulation variables directly to specific mechanical properties. Instead, it is essential to recognize that different types and concentrations of coatings will modify the rheological behavior of the compound during component manufacture, an effect that will then have its own additional (but indirect) influence on the final physical properties.
Balances between yield strength and toughness are summarized in Fig. 19, which illustrates that coatings may be selected to specifically tailor mechanical property variations according to end-use applications. These principles have important implications for the use of magnesium hydroxide in flame retardant applications, for which higher filler volume fractions are usually required and which might be expected to further exaggerate some of the trends observed here.
The addition level and aliphatic chain length of fatty acid coatings each influence the yield stress and impact energy absorption of Mg[(OH).sub.2]-filled MDPE composites. Property maxima are observed at coating levels close to the theoretical monolayer coverage, and ductile mode impact failure (similar to unfilled MDPE) can be retained when using short-chain decanoic acid coatings. Compounds exhibit greater strength and stiffness, but increased brittleness when the coating chain length is increased. ATPE chains exceed a critical chain length, which enables physical interaction with MDPE, as reflected by modified melt flow and mechanical properties. With Judicious selection of coating systems, it is possible to tailor an appropriate balance of yield strength, modulus and toughness to meet performance criteria demanded in different end-applications.
A mechanism based upon surface energy has been proposed to account for the improved dispersion evident for fillers modified by fatty acid coatings, which have no chemical interaction with the polymer. High dispersion and low particle-polymer interaction are fundamental to the development of matrix yielding and plastic deformation adjacent to the filler particles. This represents the origin of enhanced impact resistance, and as confirmed by fractographic analysis, modifies the macroscopic failure modes in MDPE-Mg[(OH).sub.2] composites. For uncoated fillers, XPS analysis of fracture surfaces has confirmed that the locus of brittle-mode failure is more concentrated at the polymer-filler interface.
Thermal history during processing influences rheological behavior and creates differences in MDPE molecular structure ([M.sub.w] and low levels of crosslinking) to some extent. Systematic variations in crystallinity induced by filler coatings are evident but do not control the mechanical properties of Mg[(OH).sub.2]-filled MDPE composites. Anistoropic effects are created by high shear stress distributions developed during processing; there is a correlation between molecular chain orientation and impact energy absorption, while XRD data indicate that significant alignment of plate-like particulate additives also occurs during injection molding.
The authors would like to acknowledge the source funding for this research, which was provided by the Engineering and Physical Sciences Research Council (EPSRC) of Great Britain, together with additional support from an industrial consortium: BP Chemicals, BOG International, Alcan Chemicals, Rothon Consultants, Cookson Group, APV Baker Ltd (Industrial Extruder Division), Rosand Ltd. and Stewarts & Lloyds. Contributions from colleagues at Loughtborough University (Dr. M. Gilbert and Mr. J. F. Harper) are also gratefully acknowledged.
(*.) Corresponding author.
Institute of Polymer Technology and Materials Engineering (IPTME)
(**.) Department of Chemistry
(1.) B. Haworth, C. L. Raymond, and I. Sutherland, Polym. Eng. Sci., in press (2000].
(2.) F. Molesky, in Plastics Additives and Modifiers, J. Edenbaum, ed., Van Nostrand, New York (1985).
(3.) L Jilken, G. Malhammar, and R. Selden. Polym. Testing, 10, 329 (1991).
(4.) M. Arroyo Ramos, M. Sanchez Berna, and J. P. Vigo Matheu, Polym. Eng. Sci., 31, 245 (1991).
(5.) S. N. Maiti and R. Jeyakumar, J. Polym. Mat., 7. 29 (1990).
(6.) B. Pukanszky and F. Tudos, Proc. Conf. Composite Interfaces, 491, H. Ishida, ed., Cleveland (1990).
(7.) A. C. Moloney, H. H. Kausch, T. Kaiser, and H. R. Beer, Conf. Proc. Fillers, 17/1, (1986).
(8.) J. A. Radosta, SPEA ANTEC, 22,465 (1976).
(9.) G. Kirschbaum, Kunststoffe, 79, 62 (1989).
(10.) Q. Fu and G. Wang, Polym. Eng. Sci., 32, 94 (1992).
(11.) J. Dansie, Performance Chemicals, 6, 39 (1991).
(12.) S. Miyata, T. Imahashi, and H. Anabuki, J. Appl. Polym Sci., 25, 415 (1980).
(13.) M. W. Darlington and R. S. V. Nascimento, Conf. Proc. Fillers, 15/1 (1986).
(14.) M. Cook and J. F. Harper, Plast. Rubb. Camp. Proc. Appl., 25, 99 (1996).
(15.) M. Cook and J. F. Harper, Adv. Poly. Tech., 17, 53 (1998).
(16.) R. N. Rothon, ed., Particulate-Filled Polymer Composites, Longman Scientific, U. K. (1995).
(17.) R G. Raj and B. V. Kokta, Polym. Eng. Sci., 31, 1358 (1991).
(18.) B. Pukanszky, F. Tudos, J. Jancar, and J. Kolarik, J. Mat. Sci; Lett., 8, 1040 (1989).
(19.) J. Jancar and J. Kucera, Polym. Eng. Sci., 30, 707 (1990).
(20.) J. Jancar and J. Kucera, Polym. Eng. Sci., 30, 714 (1990).
(21.) A. J. Kinloch and R. J. Young, Fracture Behaviour of Polymers, Applied Science, London (1983).
(22.) Y. Suetsuga, Conf. Proc. MOFFIS, 37, Namur, Belgium (1993).
(23.) Q. Fu and G. Wang. Polymer. Intl., 30, 714 (1990).
(24.) A. M. Riley, C. D. Paynter, P. M. McGenity, and J. M. Adams, Plast Rubb. Proc. App., 14, 85 (1990).
(25.) P. M. McGenity, J. J. Hooper, C. D. Paynter, A. M. Riley, C. Nutbeem, N. J. Elton, and J. M. Adams, Polymer, 33, 5215 (1992).
(26.) M. W. Darlington and T. J. Hutley, Polym, Commun., 26, 264 (1985).
(27.) C. L. Raymond, PhD Thesis, Loughboraugh University, U. K. (1997).
(28.) L. Mandelkern, Crystallisatian of Polymers, McGraw-Hill, New York (1964).
(29.) F. N. Cogswell, Polymer Melt Rheology, 2nd Ed., Wood-head Press, Cambridge, U. K. (1984).
(30.) B. Haworth, C. L. Raymond, and M. Gilbert, J. Mat. Sci. Lett., 19, in press (2000).
(31.) L. Jilken, G. Malhammar, and R. Selden, Polym. Testing., 10, 329 (1991).
(32.) T. M. Malik, M. I. Farooki, and C. Vachet, Polym. Compos., 13, 174 (1992).
(33.) J. P. Trotignon, L. Demdoum, and J. Verdu, Composites, 23, 313 (1992).
(34.) M. W. Darlington and J. L. Hardy, Coal. Proc. Fillers, 19/1, BPF, London (1989).
(35.) C. M. Liauw, S. J. Hurst, G. C. Lees, R. N. Rothon, and D. C. Dobsan, Progress in Rubb. & Plast. Technol., 11. 137 (1995).
(36.) K. Mitsuishi, S. Ueno, S. Kodama, and H. Kawasaki, J. Appl. Polym. Sci., 43, 2043 (1991).
(37.) E. Sheng, I. Sutherland, R. H. Bradley, and P. K. Freakley, Materials Chem. Phys., 50, 25 (1997).
(38.) E. Papirer, J. Schultz, and C. Turchi, European Polym. J., 20, 1155 (1984).
(39.) J. Israelachvili, Intermolecular and Surface Forces, Academic Press, New York (1991).
(40.) L. Nicolais and M. Narkis, Polym. Eng. Sci., 11, 194 (1971).
Table 1. Failure Modes for Instrumented Impact Tests (MDPE/Mg[(OH).sub.2]). Mode I Mode II Mode III Unfilled MDPE Stearic acid Uncoated filler coated fillers Decanoic acid Behenic acid Behenic acid coated fillers coated (6% only) coated (11% and 14%) Mode I Mode IV Unfilled MDPE ATPE coated fillers Decanoic acid coated fillers Key Mode I Ductile Gross yielding, incomplete break, high-energy absorption. Mode II Ductile-brittle Less yielding but predominantly ductile. Mode III Mixed-mode Little yielding; predominantly brittle fracture surface. Mode IV Brittle No macroscopic yielding. Table 2. XPS Analysis of Fracture Surfaces. Fracture Surface Magnesium (%) (2p peak) MDPE + 30% uncoated Mg[(OH).sub.2] 4.95 (0.97) MDPE + 30% Mg[(OH).sub.2] (14% Behenic acid) 3.32 (1.15) Table 3. Effects of Processing on Molecular Structure. Weight Average Number Average Molecular Weight Molecular Weight Molecular Weight Polymer (g/mol.) (g/mol.) Distribution [MDPE.sup.1] 169,000 10,650 15.9 [MDPE.sup.2] 158,000 9650 16.4 [MDPE.sup.3] 164,000 12,250 13.4 [MDPE.sup.5] 142,000 10,400 13.7 Gel Content Melt Flow Index Polymer (%) (dg/min.) [MDPE.sup.1] 2.03 (0.10) 0.117 (0.003) [MDPE.sup.2] 4.72 (0.62) 0.063 (0.010) [MDPE.sup.3] 5.64 (1.01) 0.036 (0.002) [MDPE.sup.5] 6.04 (0.91) 0.045 (0.002) Key: [MDPE.sup.1] Virgin MOPE, as supplied. [MDPE.sup.2] Processed by twin-screw extrusion. [MDPE.sup.3] Virgin MOPE, processed by injection molding. [MDPE.sup.4] Processed by twin-screw extrusion and by injection molding. (average values are given in all cases, with standard deviations in brackets) Table 4. Wide-Angle X-Ray Diffraction Data for MDPE/Mg [(OH).sub.2]. Peak Intensity (I) Data MDPE Compounds PE [I.sub.110] Unfilled MDPE (*) 1.0 Unfilled MDPE (C) 1.0 Unfilled MDPE (I) 1.0 Mg[(OH).sub.2] Uncoated (C) 1.0 Mg[(OH).sub.2] Uncoated (I) 1.0 Mg[(OH).sub.2] 6% Stearic Acid (C) 1.0 Mg[(OH).sub.2] 6% Stearic Acid (I) 1.0 Mg[(OH).sub.2] 10% Stearic Acid (I) 1.0 Mg[(OH).sub.2] 15% Stearic Acid (I) 1.0 MDPE Mg[(OH).sub.2] Compounds PE [I.sub.200] [I.sub.001] Unfilled MDPE (*) 0.331 -- Unfilled MDPE (C) 0.270 -- Unfilled MDPE (I) 0.175 -- Mg[(OH).sub.2] Uncoated (C) 0.301 0.848 Mg[(OH).sub.2] Uncoated (I) 0.257 4.484 Mg[(OH).sub.2] 6% Stearic Acid (C) 0.206 1.109 Mg[(OH).sub.2] 6% Stearic Acid (I) 0.287 4.008 Mg[(OH).sub.2] 10% Stearic Acid (I) 0.271 1.990 Mg[(OH).sub.2] 15% Stearic Acid (I) 0.340 3.846 MDPE Mg[(OH).sub.2] Compounds [I.sub.101] Unfilled MDPE (*) -- Unfilled MDPE (C) -- Unfilled MDPE (I) -- Mg[(OH).sub.2] Uncoated (C) 0.644 Mg[(OH).sub.2] Uncoated (I) 0.142 Mg[(OH).sub.2] 6% Stearic Acid (C) 0.283 Mg[(OH).sub.2] 6% Stearic Acid (I) 0.085 Mg[(OH).sub.2] 10% Stearic Acid (I) 0.059 Mg[(OH).sub.2] 15% Stearic Acid (I) 0.080 Key: (All data are normalized relative to the most intense diffraction peak, PE [I.sub.110]) (*)Virgin MDPE, as supplied in coarse powder form. (C)MDPE polymer/compound processed by compression molding. (I)MDPE polymer/compound processed by injection molding. Table 5. Thermal Analysis Data for MDPE-Mg[(OH).sub.2] Composites. Compound Tm ([degrees]C) Effects of Coated Mg[(OH).sub.2] Fillers [MDPE.sup.2] 128.1 MDPE +30% Mg[(OH).sub.2] 128.3 +6% decanoic acid 127.6 +11% decanoic acid 128.3 +16% decanoic acid 127.0 +5% stearic acid 128.0 +10% stearic acid 128.2 +15% stearic acid 128.5 +6% behenic acid 126.6 +11% behenic acid 128.2 +14% behenic acid 128.6 +5% ATPE 127.7 +10% ATPE 128.2 +15% ATPE 130.6 Compound Tsc (onset) ([degrees]C) Effects of Coated Mg[(OH).sub.2] Fillers [MDPE.sup.2] 118.0 MDPE +30% Mg[(OH).sub.2] 117.9 +6% decanoic acid 115.9 +11% decanoic acid 115.8 +16% decanoic acid 117.0 +5% stearic acid 117.8 +10% stearic acid 117.1 +15% stearic acid 117.9 +6% behenic acid 117.2 +11% behenic acid 117.3 +14% behenic acid 117.5 +5% ATPE 117.7 +10% ATPE 117.9 +15% ATPE 116.9 Compound Tc ([degrees]C) Crystallinity (%) Effects of Coated Mg[(OH).sub.2] Fillers [MDPE.sup.2] 110.3 52.3 MDPE +30% Mg[(OH).sub.2] 112.7 51.4 +6% decanoic acid 112.1 60.3 +11% decanoic acid 111.3 60.7 +16% decanoic acid 111.9 59.9 +5% stearic acid 112.6 55.0 +10% stearic acid 111.9 56.4 +15% stearic acid 112.4 57.8 +6% behenic acid 112.8 47.6 +11% behenic acid 112.5 48.2 +14% behenic acid 112.9 48.9 +5% ATPE 114.4 56.9 +10% ATPE 114.5 55.5 +15% ATPE 111.5 52.5 Key: [MDPE.sup.2] See key to Table 3 Tm Crystalline melting temperature (endotherm peak). Tsc Recrystallization temperature (onset of exotherm). Tc Recrystallization temperature (exothermic peak). (See also Figure 14 for graphical plots of crystallinity versus coating concentration).
|Printer friendly Cite/link Email Feedback|
|Author:||HAWORTH, B.; RAYMOND, C. L.; SUTHERLAND, I.|
|Publication:||Polymer Engineering and Science|
|Article Type:||Statistical Data Included|
|Date:||Aug 1, 2001|
|Previous Article:||Nonlinear Viscoelastic and Viscoplastic Response of Glassy Polymers.|
|Next Article:||Effect of Starch Content on Viscosity of Starch-Filled Poly(Hydroxy Ester Ether) Composites.|