Polyethylene Compounds Containing Mineral Fillers Modified by Acid Coatings. 1: Characterization and Processing.
Research has been carried out to determine the effect of filler coatings on the processing properties of medium density polyethylene (MDPE] modified by an ultrafine grade of flame-retardant magnesium hydroxide ([Mg(OH).sub.2]) filler. Selected filler coatings were acid-group terminated and were of varying aliphatic chain length. The filler dry-coating process has been optimized by characterizing the reaction between [Mg(OH).sub.2] filler surface and the acid group, using spectroscopic techniques including Fourier transform infrared (FTIR) and X-ray photoelectron spectroscopy (XPS). Using immersion calorimetry, the interactions between the fillers and the polyolefin matrix were shown to decrease on the addition of a fatty-acid coating. Compounding torque and specific energy data relate to filler dispersion: qualitative analysis has demonstrated how the coatings provide a reduction in both size and number of particle agglomerates. MDPE compound processability was assessed by capillary rheometry; wall slip was evident in compounds containing uncoated [Mg(OH).sub.2] filler. Consequently, the development of molecular orientation of the polymer during injection mold filling, quantified by a reversion analysis, is modified by the effects of filler coating chain length and addition level, an effect that has important implications to link the mechanical properties of MDPE-[Mg(OH).sub.2] composites to processing history.
An important factor that determines the physical properties of particle-reinforced polymer composites is the degree of interaction between the polymer and filler, in addition to the surface contact area between the two components (1). This interaction can be modified by surface treatment of the filler; many surface treatments exist and their selection is dependent upon the chemical nature of the matrix into which the treated filler will be incorporated, together with the intended application. According to Rothon (2), surface treatments can generally be categorized into two main types: coating and coupling agents. The main difference is that whereas a coupling agent chemically bonds the filler to the polymer, a coating is compatible but not reactive with the matrix polymer. Improved filler dispersion is often achieved as a result of using surface coatings on particulate mineral fillers. In this context, stearic acid has been widely used as a low-cost additive to achieve enhanced processability in PVC compounds and also acts as an effective coating for fillers in many particle-filled thermoplastic polymers (3), in order to achieve enhancements in physical properties, which are dependent upon improved filler dispersion.
Most literature citations featuring the application of coatings to mineral fillers do not mention the experimental details of the coating procedures that are used because fillers are frequently supplied in pre-coated form by proprietary manufacturing processes. In addition, it may often be assumed that the coating technique is totally effective, involving complete reaction onto the surface of the filler. Fekete et al. (4) reported a solution coating method for applying stearic acid onto calcium carbonate. A solution stirring time of 30 minutes was quoted, although the external source of heating was not specified. However, Fourier transform infrared (FTIR) analysis was subsequently carried out to provide a quantitative determination of the levels of coating bonded to the filler surface.
It has been reported (5) that the method and temperature of application can influence the effectiveness of a coating. Filler coating techniques include spraying, batch mixing, fluid bed applications, and fluid milling. Factors that determine the optimum method include the temperature-sensitivity of the coating and filler, the form of introduction of the coating and whether the filler can be subjected to intensive mechanical (shear-induced) energy, which could modify the surface energy and particle size (6). High-speed mixers are frequently used to treat the filler, prior to compounding into a thermoplastic matrix. The fillers can be solution-coated and subsequently dried, or dry-coated directly. Most filler manufacturers consider their coating processes to be proprietary (7), yet it is well known that stearic acid coated fillers can be produced in a high intensity mixer by heating the mineral to the melting point of the coating before adding the fatty acid. However, more specific details relating to the proc ess (for example, coating time required for complete reaction and its dependence upon temperature, particle morphology and mechanical agitation) are rarely quoted in the technical literature.
The introduction of fillers to molten polymers results in a complex rheological fluid, with variations in particle size, morphology and degree of agglomeration. Since processing conditions will strongly affect microstructure development and the ultimate properties and performance of filled polymer composites, rheological analyses of particle-filled polymer melts become very important. Plueddemann and Stark (8) have discussed the rheology of filled polymers, specifically in terms of how the application of a coating to a filler aids dispersion of the rigid particles during melt-state processing. Once the fillers are dispersed, or de-agglomerated effectively under shear deformation, polymer molecules interact with the filler particles but are also able to undergo motion between them. According to Choplin (9), fillers can be thought of as particles suspended in a non-Newtonian fluid, which have the effect of increasing the viscosity of the polymer when added at increasing concentration.
Additives at the polymer-filler interface can act as coupling agents (where increased adhesion promotes stress transfer under load), or as wetting (lubricating) agents. Modifiers in the latter category, for which the term "coating" usually applies, decrease the surface energy of the filler. This generally improves the dispersion of particulates in the polymer melt and improves the ultimate mechanical properties of the composite, whereas an internal lubricant acts primarily as a processing aid by reducing melt viscosity. In contrast, an additive acting as an external lubricant causes wall slip at the boundary between the flowing melt and the process equipment (die/mold flow channel surfaces) and modifies the rheological parameters characterizing the flow process.
The term "wetting agent" may be inappropriate for some surface treatments used with mineral fillers, since thermodynamic wetting will occur only if the specific surface excess free energy of the filler is greater than that of the polymer. More appropriate terms might be "disagglomeration agent" or "dispersion-promoter," since these more accurately describe the effect of the modifier. Thermodynamic wetting is able to take place in the majority of systems containing non-polar polymers, modified by uncoated polar fillers. The addition of fatty acid coatings may reduce the specific surface excess free energy of the filler (10), which may occur such that the surface energy of the coated-filler is similar to, or marginally higher than, that of the polymer, allowing the polymer to wet the surface of the filler. The interaction between the filler and polymer is therefore significantly reduced on the addition of a coating and it is important to emphasize the clear distinction between coating and coupling agents and mechanisms, in this respect.
A reduction in filler/polymer interaction can be shown theoretically with reference to the work of adhesion ([W.sub.A]), which is defined as the amount of reversible work required to separate unit areas of two phases. It has been suggested (11) that the work of adhesion between a filler and polymer ([W.sub.FP]) can be expressed by:
[W.sub.FP] = [2([[[gamma].sup.d].sub.F] * [[[gamma].sup.d].sub.P])].sup.1/2] + [2([[[gamma].sup.P].sub.F] * [[[gamma].sup.P].sub.P])].sup.1/2] (1)
Subscripts F and P indicate filler and polymer, while superscripts d and p indicate the respective dispersive and polar components of surface energy ([gamma]). In the case of a non-polar polymer such as PE, ([[[gamma].sup.P].sub.P] = 0, so that the second term in the above equation disappears. It is therefore clear that the dispersion component of the Van der Waals force is dominant for non-polar polymers. Since the specific surface excess free energy of the filler ([[[gamma].sup.d].sub.F]) is reduced on the addition of a coating, the work of adhesion and hence the degree of interaction between the filler and polymer is also decreased.
The effect of adding a stearic acid coating to [CaCO.sub.3] in low density polyethylene (LDPE) has been shown to decrease the compound viscosity with increasing coating addition, up to a level equivalent to a theoretical monolayer coverage (12). Similarly, a magnesium stearate coating on magnesium hydroxide in polypropylene (PP) caused a reduction in shear viscosity, relative to the incorporation of equivalent quantities of uncoated filler (13). In each example, the reduction in melt viscosity was more marked at lower shear rates, and is influenced by improved filler dispersion, whereas at higher shear rates the response is dominated by the pseudoplastic characteristics of the polymer to a much greater extent.
Overall, the main objective of this research was to incorporate magnesium hydroxide filler into polyethylene and to optimize a range of mechanical properties as a result of treating the filler surface with organic acid coatings of variable chain length. The results are presented in a two-part paper. Part 1 discusses the filler coating process, spectroscopic characterization of coated fillers and an analysis of some rheological properties of MDPE containing the coated fillers. Part 2 considers structure and mechanical properties of components molded from these compounds and illustrates interactions from all variables along the chain of knowledge from chemical characterization of the mineral additives to the ultimate physical properties of molded polymer composites.
In order to facilitate the FTIR interpretations, the selected grade of MDPE polymer (BP Chemicals, Grangemouth, UK) contained no additives other than a minimum amount of antioxidant, but was otherwise similar to commercial high molecular weight grades used in applications such as extruded pressure pipe. A scanning electron micrograph of the magnesium hydroxide fillers used is shown in Fig. 1; the particles have a high aspect ratio, planar "plate-like" character, but are not acicular. Technical specification of the fillers provided by the supplier (Premier Periclase Ltd., Ireland) suggests a primary particle size of around 0.8 [micro]m (corresponding to a specific surface area of 13[m.sup.2]/g), which was in close agreement with measurements carried out by particle size analysis (see also Fig. 1). Specifications for all materials used throughout this study are given in Table 1.
Fatty acid filler coatings of varying aliphatic chain length were applied to the inorganic fillers. The coating of lowest molar mass was decanoic acid (with ten carbon atoms along the aliphatic chain, hence "C10"), and the longest-chain coating was an acid-terminated polyethylene (ATPE) of molecular weight 5000 g/mol., supplied by AlliedSignal Corporation (Belgium). Other coatings selected for study were stearic acid (C-18), which is used in many similar commercial applications and behenic acid (C-22). This selection of coatings was sufficient to investigate the effects of fatty acid chain length. In addition, while the carboxylic acid group on the functionalized (acid-group terminated) PE is likely to react similarly with the [Mg(OH).sub.2] filler surface, the aliphatic chains are of sufficient length and are of identical chemical structure to enable physical interaction with the MDPE matrix.
Filler Coating Processes
Magnesium hydroxide fillers were coated by dry blending using a benchtop Waring mixer, which was used to produce small samples of products for subsequent spectroscopic analysis. A series of screening experiments was initially carried out, in order to select the optimum mixing process variables (power input, temperature and blending time) to give a complete coating reaction. A scaled-up dry blending unit (T.K. Fielder high-speed mixer of 8-liter capacity) was then used to produce batches of coated [Mg(OH).sub.2] fillers for subsequent compounding into MDPE. Maximum rotor speed was 3000 rpm and the coating time (determined in accordance with the FTIR characterization) was for a minimum period beyond the time taken for the mixer temperature to reach the melting point of the respective fatty-acids.
Characterization of Coated Fillers
Spectroscopic analysis by FTIR (Nicolet 20DXC FTIR spectrometer) was used to characterize the coating process as a function of mixing time and coating type/addition level. For each type of acid, samples were removed from the mixer at frequent intervals throughout the coating experiment. These were mixed with finely ground potassium bromide (KBr) at a dilution level of 2.5%, to control the degree of absorption and ensure accurate, quantitative results. From the IR spectra, the integrated area of the CH stretching peak due to the reacted acid (wavenumber range 2747-3026 [cm.sup.-1]) was measured and is expressed as a ratio relative to the OH absorption peak from the [Mg(OH).sub.2] filler (3605-3750 [cm.sup.-1] wavenumbers). This ratioing procedure was carried out to eliminate any variation due to specimen preparation.
X-ray photoelectron spectroscopy (XPS) was used to determine the thickness of the coatings on the [Mg(OH).sub.2] fillers, using a technique described elsewhere . Filler-polymer interactions were also studied using a Parr solution calorimeter; by immersing samples of filler into a liquid (representing the polymer matrix) and by measuring the resulting temperature change using micro-calorimetry, the energy of interaction between filler and polymer can be determined. The heat evolved on immersing 5 g of filler into 100 ml of heptane was accurately measured for uncoated [Mg(OH).sub.2] and also for filler coated with different levels of stearic acid. Heptane was used as the immersing liquid to represent the polyethylene, since it is also non-polar and the changes in enthalpy are likely to be similar.
Compounding and Analysis of Filled MDPE
MDPE compounds containing 30% by weight of [Mg(OH).sub.2] filler (which had been dried for 24 hours in an air-circulating oven at 60[degrees]C, to remove surface moisture) were prepared using a conventional co-rotating twin screw extruder (APV Model MP2030, 30 mm diameter, 15:1 length/diameter ratio and 7.5 kW maximum drive power) fitted with K-Tron volumetric feeders and a downstream water-bath and pelletizer unit. All compounds were produced at an identical mass output rate of 15.6 kg/hr., using a rotational screw speed of 250 rpm at a maximum melt temperature setting of 200[degrees]C; full details of the compounding process and screw design are given elsewhere . Steady-state measurements of machine torque were recorded throughout, to enable specific energy data to be estimated.
Thermogravimetric analysis (TGA) (Rheometric Scientific Instruments Model TG760) was conducted on selected compounds to investigate thermal decomposition of the mineral fillers and to determine the most appropriate conditions for subsequent ashing tests used to determine exact filler concentrations. A heating rate of 10[degrees]/min. was used, up to a maximum temperature of 900[degrees]C. Ashing tests were carried out in a furnace according to BS2782 (Part 4, Method 454A) to a maximum temperature of 850[degrees]C.
Shear flow processability of selected compounds was analyzed using a Davenport capillary rheometer operating at a temperature of 220[degrees]C. Flow analysis was carried out firstly to determine conventional shear flow behavior as a function of shear rate and also to investigate the occurrence of interfacial wall slip at the compound-die wall boundary. For the latter experiments, a set of six capillary dies of variable length (between 8 mm and 20 mm) but with a fixed length to diameter ratio (of 10:1) were designed to investigate wall slip behavior using the technique first attributed to Mooney .
Molding and Mechanical Testing
Test pieces for mechanical property determination were produced using a Negri Bossi NB55 (55 tonnes maximum lock) injection molding machine, using a melt temperature of 220[degrees]C and a controlled mold surface temperature of 15[degrees]C. Specimens machined from the linear (gauge) portion of the tensile bars were also used to determine the degree of polymer orientation induced by non-isothermal mold filling, using a reversion analysis at elevated temperature . Samples were placed on a tray coated with slip agent (an oil/talc mixture) and were heated to a temperature of 120[degrees]C for 30 minutes; this temperature was predetermined to obtain elastic recovery data, which are independent of small changes in heating time and temperature. Thermal reversion data are calculated from specimen dimensions and shrinkage measured both before and after the exposure to elevated temperature.
Additive Dispersion Analysis
Dispersive mixing analysis of the filler particles was carried out on selected compounds using a cryogenic polishing technique described elsewhere . Izod-type impact test-bars were cut normal to the injection flow-axis and mounted in a toughened epoxy resin, to reveal sections across the center of the bar. These were wet-polished in a liquid nitrogen environment firstly by using successively finer grades of silicon carbide paper, followed by alumina powders. After polishing, the compound sections were etched in 1M HCl solution for 24 hours, rinsed, then analyzed by scanning electron microscopy (SEM).
RESULTS AND DISCUSSION
Filler Coating Process
Temperature-time data were captured from the laboratory-scale mixing unit, in order to establish the optimum blending conditions required to increase the temperature above the melting point of each organic coating in a controlled manner. During the initial stages of filler coating analysis, results were obtained showing the transient-state internal mix temperatures attained in the Waring mixer vessel, as a function of heat input setting (voltage), for each of the different coating types. The data in Fig. 2 are typical of the initial temperature-time profiles; this particular set of data is from [Mg(OH).sub.2] filler coated with 7% ATPE. It is clear how the voltage applied to the heater coil increases the internal mix temperature in the coating chamber of the Waring blender. In the example where no external heating was applied, the temperature increases to an equilibrium value of around 50[degrees]C, which is due to frictional heating effects in the mixer. Internal temperature increases with input voltage sett ing at any given mixing time and reaches around 150[degrees]C after around twenty minutes' agitation, for the experiment operated at the maximum setting of 80V. This information has therefore been used to specify the optimum coating time for each given type of filler coating.
Subsequently, different levels of fatty-acid coatings were added to the [Mg(OH).sub.2] fillers, to assess the effect of coating on the temperature attained in the chamber at a given voltage setting. Figure 3 shows a typical set of results, in which the coatings appear to have an additional and clear effect on the measured heating rates. After twenty minutes mixing time, the uncoated [Mg(OH).sub.2] reaches a temperature of 120[degrees]C, while fillers mixed with ATPE develop a significantly higher temperature of around 150[degrees]C. The [Mg(OH).sub.2] modified by the higher coating level (10.2% ATPE) reaches the equilibrium temperature more rapidly. This analysis was also repeated with the other three types of coating and results in the same format were obtained.
It was observed that filler agglomeration occurred if the coating was added too rapidly, or while the heater was activated. Optimum coating was therefore achieved by introducing the filler at room temperature and by blending the filler and coating without external heating for several minutes, before subsequently switching on the external source of heat energy. This was a particular requirement for decanoic acid, since its melt temperature is relatively low (32[degrees]C) and therefore decanoic acid-coated filler is very sensitive to agglomeration if heated too rapidly. A premixing period of five minutes was therefore used for decanoic acid coatings, whereas three minutes was sufficient for fillers modified by the other coatings, which were less sensitive to this effect owing to their higher melting temperatures.
FTIR DRIFT Spectroscopy
Small samples of coated filler were taken intermittently from the Waring mixer during the coating experiments and were analyzed using FTIR DRIFTS analysis. The results for [Mg(OH).sub.2] coated with 8% behenic acid in Fig. 4 show the increase in the CH-stretching peak (relative to OH) over a period of 70 minutes, after which a plateau is reached where the reaction can be assumed to be complete. It is evident in Fig. 4 that the reaction did not begin until after around 15 minutes, when the melting temperature of the behenic acid coating (80[degrees]C) was attained. Similar analyses for the other coatings also revealed that it was necessary to exceed the melting temperature of the coating, then to hold the system above this temperature for a specified period of time in order to complete the reaction. The optimum coating times were determined to be 40 minutes for decanoic acid, 60 minutes for stearic acid and 70 minutes (approx.) for both behenic acid and ATPE.
Having established the optimum reaction times required for each coating, different levels of all coating materials were then applied to the [Mg(OH).sub.2] in the Waring blender, with the appropriate input voltage set to give sufficient heat energy for the coating to reach its melting temperature and react with the filler surface (for example, 80V setting for 70 minutes coating time). Figure 5 shows how the CH/OH ratio increases with coating addition level, for three different fatty-acid coatings. For behenic acid, this absorption ratio reaches a maximum at an addition level of around 12--13% (by weight). Higher coating levels do not give an appreciable increase in CH/OH ratio, which reaches a limiting value of 0.28. This represents the experimental monolayer coverage of this specific grade of filler with behenic acid. For stearic acid, the results were very similar to those of the behenic acid described above.
In contrast, the CH/OH ratio for ATPE coating continued to increase on the addition of higher coating levels and a limiting value of CH/OH was not achieved (Fig. 5). This observation is attributed to a continuous reduction in the OH absorption peak, giving the appreciably higher CH/OH ratios displayed, which prevents direct comparability with the other coatings. In effect, this occurs as a result of the masking of filler surface by the aliphatic chain structure of ATPE, which has a much higher molar mass than the other acid coatings. In the case of the decanoic acid, CH/OH ratio continued to increase with increasing acid coating level (up to a maximum addition of 23%), without reaching a plateau (Fig. 5). More detailed interpretation from the IR spectra revealed that no unreacted acid was present in the system. Furthermore, the COO- peaks due to the metal salt (e.g. magnesium decanoate) at 1868-1305 [cm.sup.-1] wavenumbers were measured, and demonstrated a continued increase when plotted against the decanoic acid coating addition level. Therefore a continuous, ongoing reaction in the mixer has occurred in this system, even though the maximum coating addition level far exceeds a "theoretical monolayer coverage" for this grade of [Mg(OH).sub.2] filler. In consequence, we must conclude that particle breakdown occurs in the mixer, thereby creating additional surface area of magnesium hydroxide which becomes progressively coated during the ongoing reaction in the process cycle.
Having fully characterized the coating procedures necessary for the small-scale Waring blender to produce a complete surface coating on magnesium hydroxide, larger batches of coated filler were also produced using a Fielder high speed mixer. A sample from each batch was analyzed using the DRIFT technique and the results are summarized in Fig. 6. It is acknowledged that while mixing conditions for each acid coating were not identical, because of their dissimilar melting points, qualitatively similar trends were obtained. As was the case in the laboratory-scale blending unit, the ATPE coating was characterized by significantly higher CH/OH ratios, owing to the long chain nature of the polymeric coating reducing the OH-absorption peak. The results in Fig. 6 add further confidence to the overall trends that have been determined experimentally and allow the possibility of scale-up to predict optimum coating cycles in other dry-mixing systems. A summary of the filler coating data has been compiled in
X-Ray Photoelectron Spectroscopy
The importance of the interphase in a filler-polymer system has been recognized previously (19, 20) but owing to experimental constraints, coating thickness data on mineral fillers have not been widely reported. XPS analysis was therefore used to determine the thickness of the acid coatings on the filler surface, using a technique described previously by Sutherland et al (14). Exact acid addition levels are determined by spectroscopic analysis in all cases, rather than by taking the nominal coating concentrations that were added to the mixer. The results presented in Fig. 7 show how the thickness increases with increasing coating level and also at any given addition level, coating thickness increases with aliphatic chain length. XPS analysis is specific to the over layer on the outer surface of the filler and does not detect carbon from the fragments of magnesium fatty-acid salts. The coating thickness data in Fig. 7 are of the same magnitude as the theoretical extended chain lengths of the coating materials , which vary between 12.5A (decanoic acid) and 27.5A (behenic acid), suggesting that the aliphatic chains lie predominantly perpendicular to the filler surfaces.
If uncoated fillers are immersed in either water or in heptane, the reactions are purely exothermic. When coatings are applied, the reactions are still "net exothermic," but a small initial endothermic component can be detected in a microcalorimeter of sufficient sensitivity, and this can be used to characterize an interaction energy. Figure 8 illustrates how the heat of immersion for magnesium hydroxide filler decreases on the addition of a fatty-acid coating, for both water and heptane. A rapid decrease in enthalpy is evident on the addition of 1% stearic acid coating for the water immersion experiment and immersion data for water are consistently higher than the AH values determined in heptane.
The reduction of specific surface excess free energy of a filler when a coating is added has a significant effect on the interpretation of many other results in this research. Rather than providing greater interaction between filler and matrix polymer, the thermodynamic work of adhesion will decrease on the addition of a coating. Hence the improvements in mechanical properties of filled polymer compounds, which are often seen on the application of coatings, are attributable to factors other than increased filler-polymer interaction. Notably, it is likely that this decrease in interaction between filler and polymer will result in improved filler dispersion within the polymer matrix at the compounding stage; this point is addressed further in a later section of the paper.
Proposed Mechanism for Filler Coating
From the coating analysis data in Figs. 2-6 several points become apparent. It is an immediate prerequisite to generate sufficient heat to raise the mixer temperature above the coating melting point in order to initiate a reaction. Then, the mixer is held above the coating melting temperature for a specified amount of time in order to progress the reaction between the acid group and the filler surface (see below). It is also necessary to provide an intensive mixing environment to achieve good dispersion of the coating material within the bulk of the filler. Heat generation and mechanical agitation are interrelated: it is initially important to raise the temperature gradually while mixing, so that the coating becomes well dispersed in the filler before reaching its melting temperature, in order to prevent agglomeration of the filler coating species.
Interaction occurs between the [Mg(OH).sub.2] filler surface and the acid group on the coating, according to the following reaction:
Mg - OH + R - COOH [right arrow] Mg - COOR + [H.sub.2]O (2)
Formation of the organic salt carries practical significance to the result, since this may form as part of a progressive and continuous reaction as discussed earlier (Fig. 5) and in addition, even small quantities of fatty-acid salts are known to induce a lubricating characteristic when processing specific types of polymers.
The FTIR DRIFT technique is a valuable method for determining the experimental monolayer coverage for the stearic and other fatty acids on mineral fillers such as magnesium hydroxide, which is attained when the CH/OH ratio reaches a plateau. This was determined to be around 6% (by weight) for stearic acid on this grade of magnesium hydroxide and around 12% for the behenic acid (see Fig. B and Table 2). Comparing these values to the theoretical monolayer coverage (as determined by the "footprint" method used by Papirer , which would give theoretical values of 3% and 3.5% for stearic and behenic acids), it can be seen that the experimentally determined monolayer coverages are considerably higher than those determined by theory . It is suggested that the intense mixing environment of the blenders promotes an ongoing "continuous" reaction, whereby the acid group first reacts on to the filler surface. Subsequently, owing to the high stress created by the intensive mixing process, some "reacted filler" (rich in magnesium salt) is spalled-off from the surface of the original mineral particles, revealing fresh (uncoated) filler surface, with which the coating can continuously react. On studying the infrared spectra, no absorption peak has been detected at 1710 [cm.sup.-1] wavenumbers, a result that confirms that there was no unreacted acid in the system.
Coating thickness results calculated from the XPS data are shown in Fig. 7 and further support this theory. Thickness data derived by XPS are less than, but of the same order of magnitude as theoretical extended chain lengths of the fatty-acid coatings, which are approximately 12.5A for decanoic acid, 22.5A for stearic acid and 27.5A for behenic acid. The increase in measured coating thickness with increased coating application is relatively modest, further supporting the suggested mechanism of particle breakdown at the filler surface, during the blending (coating) process.
A similar representation of FTIR data for ATPE coating (using CH/OH peak height ratios) is inappropriate. The CH/OH ratio continues to increase and does not form a plateau at the experimental monolayer level (Fig. 5). This is due to the OH peak height decreasing at high coating levels, as a result of the significantly longer chains of ATPE masking the filler particles from the incident infrared beam, hence the reflectance due to the OH groups decreases. The CH-peak height increases no further after a coating addition level of around 10% (by weight). Because of this "masking" effect, the value of the CH/OH ratio for ATPE-coated [Mg(OH).sub.2] at the plateau was not a valid consideration.
At coating levels higher than 10% ATPE, it was noted that the blended samples were "gritty" in nature, which is due to the chains of the ATPE having been melted and subsequently cooled during the coating process, producing small particles of what are, in effect, "highly-filled polymer composites." Also, the filler coated with ATPE is more difficult to disperse in the KBr diluent, since the long PE chains have a tendency to bind the filler particles together to form pseudo-agglomerates. These agglomerates can be broken down only by re-heating to the melting temperature of the ATPE and are therefore not reduced in size by physical vibration when the KBr is mixed. The theoretical monolayer coverage for ATPE on this grade of magnesium hydroxide is approximately 50% by weight, if the calculation assumes that the acid groups are all close-packed on the surface. This is clearly not a valid determination of the monolayer coverage for ATPE as a filler coating; the high molecular weight of the coating would mean that there is far more coating material than filler, if considered on a volumetric basis.
The relationship between CH/OH peak ratio and coating concentration has been the subject of detailed study for samples produced in a scaled-up mixer, for all coatings used (Fig. 6). It would appear that a similar ongoing reaction between filler surface and acid group occurs for all short-chain acids, but in the case of behenic acid, a decrease in gradient is evident. It is likely that the longer aliphatic chains of behenic acid block the availability of reaction sites on the filler surface, an effect that is not observed in the other short-chain fatty acids.
Finally, the apparent ongoing reaction between the filler surface and the acid groups will result in a significant quantity of magnesium salts, which will exist as small crystals in the filler-coating mix and will have a potentially important influence on the physical properties of the polyethylene compounds. This effect will be discussed further in Part 2 of this communication.
MDPE polymer has subsequently been compounded with a selection of the [Mg(OH).sub.2] fillers described above, by twin-screw extrusion. The torque (M) readings recorded during steady state processing were used to determine the specific energy ([U.sub.SP]) values for each compound, according to the following expression :
[U.sub.SP] = [P.sub.MAX] * (N/[N.sub.MAX]) M/[Q.sub.M] (3)
P represents power input, N is rotational screw speed and [Q.sub.M] is the steady-state mass output. The maximum drive power ([P.sub.MAX]) of the twin screw compounding unit is 7.5 kW, the maximum screw speed ([N.sub.MAX]) is 500 rpm and the compounds were all produced using a screw speed (N) of 250 rpm at a mass output rate ([Q.sub.M]) of 15.6 kg/hr. Figure 9 shows how the specific energy increases on the addition of the magnesium hydroxide filler and with the exception of the ATPE compounds, specific energy decreases progressively when coatings are applied to the filler in increasing quantities. For the three fatty-acid coatings, increasing chain length produces a decrease in specific energy, for a given coating addition level. When ATPE coating is applied, the specific energy initially increases, before decreasing in a similar manner to the compounds containing fillers modified by low molecular weight coatings.
Figure 9 illustrates specific energy data calculated from steady-state torque readings, according to equation (3). This is a direct measurement of energy required to operate the compounding unit and is often assumed to indicate the intensity of processing and specifically, the goodness of dispersive mixing in multiphase thermoplastic compounds. However, several parameters contribute to the specific energy values obtained by this route, which include energy requirements to:
* melt the polymer and to raise the compound temperature to the set processing condition (this process is usually adiabatic, under steady-state conditions)
* achieve particle breakdown to primary size by dispersive mixing
* transport the molten compound under pressure along the melt-pumping sections and mixing zones in the extruder and through the strand-die
* rotate the screws
The specific energy parameter is therefore a composite property, which depends upon each of these contributions to some extent. Specific energy data for unfilled MDPE is included in Fig. 9, so that the energy components that do not include dispersive mixing can be estimated. The practical data displayed show how the addition of a fatty-acid coating reduces the specific energy. If the mixing parameter dominates the overall response, the coating is likely to reduce the work input (hence specific energy) required to achieve a given degree of filler dispersion as a result of decreased specific excess interaction energy, and is therefore an important consideration for this study. Similar conclusions regarding the ability of coatings to enhance filler dispersion in thermoplastic melts have been made previously [5, 8, 9, 12, 13, 19, 22-25]. This characteristic is clearly desirable in order to reduce agglomeration, thereby giving the potential to increase mechanical properties of manufactured artifacts.
The incorporation of the acid coatings also appears to be contributing to an internal lubrication effect on the compounds, which causes a reduction in the torque as a result of reduced shear viscosity. Greater quantities of acid increase the lubricating effect and torque decreases progressively as coating levels increase (Fig. 9). The internal lubrication effect induced by the coatings is characterized by the reduced shear viscosity of the compounds, which will be illustrated and discussed further in the next section. Notwithstanding the point made earlier regarding dispersive mixing contributions to machine torque measurements, lower torque (and hence lower specific energy values) is also indicative of a lower compound viscosity, to some extent.
The differing melting temperatures of the fatty acids also contribute to variations in torque. In the initial stages of compounding, species of lowest melt temperature (short-chain acids and salts) will melt more rapidly than the longer-chain systems. If one component of the polymer/coated filler is in its molten state, the material will be transported to the screw flight and barrel surfaces, thus facilitating shear deformation. Moreover, if the melting temperature is reached at an earlier stage in the process, the total shear deformation experienced by the material will be increased and hence the torque will increase correspondingly. Overall, however, the decrease in torque observed with the compounds containing the fatty-acid coatings is attributed to the coated fillers requiring less energy to be dispersed within the thermoplastic MDPE polymer matrix.
In comparison to the compound containing uncoated [Mg(OH).sub.2] filler, the ATPE coated fillers show an initial increase in the torque (hence specific energy; Fig. 9), that subsequently decreases with increasing coating levels. The aliphatic chains of the ATPE polymeric coatings are substantially longer than the fatty-acid chains, so that physical interaction and entanglement with the matrix polymer is more likely. This will give rise to a higher shear viscosity in the melt-state (hence a greater energy requirement to transport the compound in the extruder, at a given output rate), an increased power contribution to disperse the fillers to a given level, and/or it may also indicate that the fillers have been dispersed to a higher degree, in the final compound. The micrographs shown in Figs. 10-12 show qualitatively that while all the MDPE-[Mg(OH).sub.2] compounds exhibit consistently high degrees of filler dispersion, the ATPE coated fillers show even greater levels of dispersion, consistent with the implica tion from the compounding data in Fig. 9.
Filler Dispersion Analysis
The scanning electron micrographs in Figs. 10-12 show the dispersion analysis of the [Mg(OH).sub.2] filler particles within the MDPE matrix. Note the lower magnification of 1k, in Fig. 11. It can be seen clearly in Fig. 10 how the uncoated [Mg(OH).sub.2] contains a significantly greater level of agglomerated filler particles of around 2 [micro]m. There are fewer visible agglomerates for the compound containing 11% decanoic acid coated filler (Fig. 11); the lateral dimension of these agglomerates (approximately 1 [micro]m) is smaller than in the uncoated sample and corresponds to the width of two primary particles in the plane of observation. Figure 12 shows how the degree of agglomeration of [Mg(OH).sub.2] is reduced further when the filler is coated with 10% ATPE; very few surface defects are present, confirming the excellent levels of dispersion achieved, and moreover, the average size of visible particles corresponds closely to an individual filler primary particle size of 0.7 [micro]m. The compounds containing AT PE coated fillers (Fig. 12) generally exhibited much better dispersion than the decanoic acid coated samples, exemplified by fewer large surface defects after etching and polishing. Physical interaction may occur between the long aliphatic chains of ATPE coating and the matrix polymer, although no evidence has been generated to support the existence of co-crystallization in the solid-state. Filler particles are therefore more likely to be encased subsurface within the polymer matrix and cannot therefore be etched. Alternatively, they may be dispersed to such a large extent that no agglomerates are present, only primary particles. These differences can be put into a clearer perspective when considering that the specific energy values for all three compounds shown in Figs. 10--12 were similar, at around 0.18 kWhr./kg (Fig. 9). The observed surface whiteness in Fig. ills thought to be due to an uneven and partially roughened surface, rather than due to a particle dispersion effect.
Dispersion analysis has produced results in the form of scanning electron micrographs of surfaces that have been sectioned and then polished to a high degree. The exposed filler particles have subsequently been etched by reaction with HC1 to reveal an array of surface holes and defects, the size of which is related to the original lateral dimension of the filler particles or agglomerates. Excellent contrast is achieved and the technique clearly distinguishes the effects of changing formulation or processing conditions. However, the main disadvantage of this technique, in addition to the time-consuming polishing method, is that the apparent size of the surface defects may not correlate directly with the corresponding filler particle or agglomerates (see Fig. 13). When an image is viewed normal to the plane of the polished surface, the implication is that the surface area associated with each of the three surface defects is approximately the same (i.e. lateral dimensions [x.sub.1] = [x.sub.2] = [x.sub.3]). However, the amount of material etched from each of these agglomerates may be quite different. From an extreme perspective,
defect 2 might be a single filler particle, but 1 and 3 may represent sections from agglomerates, or much larger particles. Statistically, over large surface areas, however, this technique is considered to be useful to compare filler dispersion levels in a qualitative manner, and can be used for comparative purposes to optimize filled thermoplastic formulations and compounding conditions adopted for manufacture.
The high levels of dispersion achieved are notable, since plate-like fillers such as this ultrafine grade of [Mg(OH).sub.2], have a greater tendency to agglomerate than "blocky," pseudo-spherical fillers, owing to their higher specific surface area. The effect of coating is to reduce the agglomeration , as is evident in the micrographs. Specific energy data from twin-screw compounding (Fig. 9) have shown consistently higher values for the compounds containing ATPE-coated [Mg(OH).sub.2] fillers, which reflects, in part, the significantly greater levels of filler dispersion achieved in these compounds. Overall, the improved dispersion of the filler particles on the addition of coatings occurs because of a lowering of the specific surface excess free energy of the filler, as exemplified by the immersion calorimetry data (Fig. 8). Thermodynamically, the coated fillers will have a reduced interaction with the polymer in comparison to uncoated filler, but the different coatings studied here will have equivalen t specific surface excess free energies, since an identical reaction with the filler particles occurs for each system. Therefore, the energy of interaction between the coating and the polymer will be constant, even if the coatings used have quite different aliphatic chain length. In turn, it can be concluded that the differences observed in filler dispersion are likely to be due to the influence of the coatings on the melt processing behavior of the compounds, in terms of shear viscosity and internal/external lubrication mechanisms such as wall slip.
The rheological analysis was carried out on a selection of MDPE compounds containing 30% (by weight) of DP393 magnesium hydroxide filler. Coated fillers studied were modified by decanoic acid (6%) and by behenic acid (6% and 14%), together with control compounds containing uncoated filler and unfilled MDPE.
The first analyses carried out on the rate-imposed rheometer determined conventional shear flow characteristics, in the form of flow curves (shear stress versus apparent shear rate); Figs. 14-15. Most unfilled high polymers behave in a pseudoplastic manner where the shear viscosity is a decreasing function of shear rate. In cases where the Power Law model  of shear flow behavior is applicable, biogarithmic plots of shear stress ([tau]) against shear rate ([gamma]) are linear:
[tau] = k . [([gamma]).sup.n] hence: log[tau] =logk + n.log [gamma] (4)
In terms of shear viscosity ([eta]):
[eta] = k . [([gamma]).sup.n-1] hence: log [eta] = logk + (n-1). log [gamma] (5)
The plots in Figs. 14-15 are based upon the power law interpretation, which is valid for most compounds studied over the shear rate range selected (40-2000 [s.sup.-1]). Gradient n is the power law index ("pseudoplasticity index") and k is the consistency index, which relates to viscosity at low shear rate for a material of prespecified pseudoplasticity index. The flow curves were generally found to be qualitatively similar for all filled-MDPE compounds; a power law interpretation has been adopted and a data summary of power law indices is given in Table 3. Various capillary dies of constant length-diameter (L/D) ratio were used, for wall-slip analysis (see below). The flow curves shown in Figs. 14-15 have been constructed from data derived from the (18 X 1.8) mm die, and are considered typical of the results obtained from all other dies. It should be stated that no attempts were made to apply end corrections to the shear flow data, to overcome sources of inaccuracy due to convergent flow at the die entrance. In this case, a Bagley correction technique  is not applicable, since the die length/radius (L/R) ratio is constant for the series of dies studied.
Rheological analysis has indicated several important results: first, the data are dependent upon die geometry to some extent, as illustrated by the differences in flow behavior that emerged when using dies of different capillary length. This point is addressed in the subsequent paragraphs, in terms of wall slip phenomena. Coating concentration is also significant, since the compound containing filler coated with 14% behenic acid consistently produces a lower shear viscosity, in comparison to lower coating content compounds (Fig. 14). In addition, the compound containing [Mg(OH).sub.2] filler modified by decanoic acid (Fig. 15) has a higher shear viscosity than those containing behenic acid coatings. These results are in agreement with the specific energy data obtained from the APV twin screw extruder (Fig. 9), which reveals how the reduction in viscosity with increasing coating chain length and concentration contributes to the compounder torque readings, power consumption and specific energy calculations. It is therefore concluded that higher fatty-acid coating levels have a lubricating effect on the shear flow behavior of filled MDPE compounds, in common with the influence of behenate salt (data for 14% behenic acid in Fig. 14), which is known to be present in significant quantities when behenic acid coatings are used at super-monolayer level. Also, the viscosity increase observed when MDPE contains uncoated [Mg(OH).sub.2] filler was relatively modest and was characterized by a lower index of pseudoplasticity (Table 3). This point is addressed in the next paragraph, in terms of finite velocity at the flow boundary.
Flow analysis was carried out using dies of variable length but constant length/diameter ratio, in order to investigate wall slip behavior according to the method first attributed to Mooney [16, 28]. Apparent shear rate ([gamma]) data, taken over a series of constant shear stress ([tau]) levels, are plotted against reciprocal die radius (1/R) from viscometric measurements taken from the range of capillary dies. A positive gradient indicates the existence of die wall slippage, indicative of finite flow velocity at the capillary flow boundary. Figures 16-18 show a selection of the wall slip data determined from the flow curves obtained from these dies. A relatively high level of scatter is not uncommon for this type of analysis, especially since in this case, the data points were taken from a range of flow curves that have been assumed to exhibit Power Law behavior. There is no evidence to suggest wall slip has occurred with the unfilled MDPE polymer (Fig. 16) at shear stress levels up to 435 kPa, However, fo r the compound containing uncoated filler (Fig. 17), there is clear evidence of wall slip, since the lines of constant shear stress have slopes greater than zero, for all shear stress levels greater than 300 kPa. The higher the shear stress, the greater the wall slip velocity. This observation accounts for the apparently anomalous flow behavior of this compound in the flow curves discussed earlier (Figs. 14-15), since wall slip induces a higher apparent shear rate at any given shear stress. In contrast, none of the compounds containing coated fillers showed a tendency to slip at the flow boundary, even when organo-acid coatings were added at concentrations of up to 14% (Fig. 18).
The mechanism by which wall-slip occurs in compounds containing uncoated filler is not clear and is the subject of ongoing research. By using surface coatings it is possible to control the energy of interaction between the filler and polymer. The implication from these results is that an increased interaction energy associated with uncoated filler (Fig. 8) appears to give finite wall slip at the flow boundary. An argument for wall slip occurring by migration of small molecules to the die wall is often made. However, the existence of magnesium behenate salt in the compounds containing high levels of behenic acid coated filler (see the conclusions made in the section on FTIR spectroscopy) does not appear to induce an external lubrication effect in the studies presented here. Other proposed mechanisms are a failure of adhesion between the polymer and the capillary wall [29, 30], or adhesive failure between the polymer chains in the melt near to the wall . Hill et al.  have developed a relationship to rel ate the critical wall shear stress for wall slip ([[tau].sub.cr]) to the work of adhesion ([W.sub.ad]) between the polymer and the die wall:
[[tau].sub.cr] = 40 . ([W.sub.ad]/R) (6)
Further research is required to determine the properties that would be required to validate the application of this approach, or a modification from it that would account for the different adhesion characteristics of heterogeneous systems such as polymers containing acid-modified filler particles, to include small-scale agglomeration effects.
Reversion Analysis and Orientation Effects
The variable degrees of filler dispersion and compound viscosity are shown to influence the molecular orientation of the MDPE polymer induced during injection mold filling, as shown by the experimental reversion data in Fig. 19. These data were estimated from the degree of axial shrinkage (elastic recovery) following a period of static heating at 120[degrees]C in an air oven. Overall, measured reversion values for all samples, particularly unfilled MDPE, are very high, as would be anticipated for a high molar mass material. However, these decrease substantially on the addition of an uncoated filler. If the fillers are subsequently coated a significant increase in the reversion is noted, which then progressively reduces when increasing amounts of short-chain fatty-acid coatings are applied. For higher levels of behenic acid coating, for example (greater than 10%), the reversion value is reduced to below that of the MDPE compound containing uncoated filler. When ATPE coating is used on the [Mg(OH).sub.2] filler , the reversion appears to increase progressively with increasing coating level, to a limiting value of around 75%, similar to that of unfilled MDPE.
The reversion analysis is a measure of the relaxation of polymer chain orientation as a result of partial recovery of the elastic component of shear strain developed in the injection mold-filling phase. The initial and most striking observation from Fig. 19 is the remarkably high levels of reversion that have been measured. This is attributed to the high molecular weight of the pipe extrusion grade MDPE polymer used, which requires fast injection speeds and high hold pressures to fill and adequately pack the mold. These conditions will therefore generate high shear stress in the flowing melt, which is responsible for the anisotropic effects that have been observed. In these experiments, all independently set injection molding conditions remained constant throughout, so that only the effects of formulation are being compared in the results. Reversion analysis was carried out on the center-waisted section of the tensile dumbbell bars, corresponding to positions where flow was relatively uniform, such that effec ts due to excessive packing in locations close to the gate were avoided. A high degree of orientation and alignment of chain-extended crystallites is induced when injection molding semicrystalline polymers, especially in the regions of high shear stress and rapid cooling adjacent to the cavity walls. Subsequent reversion occurs when the sample is subsequently raised to an elevated temperature, as previously oriented polymer chains relax towards a random-coil conformation, their preferred state of lowest free energy.
For the unfilled MDPE, the reversion reaches a value of 75%, well in excess of the reduced level of 55% observed when 30% uncoated [Mg(OH).sub.2] is present in the compound. A reduction of melt elasticity on the addition of mineral filler to a viscoelastic polymer melt would be anticipated, though this effect will be offset by the more rapid cooling of the filled polymer, owing to its enhanced thermal conductivity. The subsequent increase in reversion observed in compounds containing coated fillers is linked to the modified rheological behavior (Figs. 17-18). A clear correlation is therefore demonstrated between the occurrence of wall slip observed in MDPE modified by uncoated filler and the subsequently lower degree of reversion observed in this material.
For a given injection (volumetric flow) rate in injection molding, the occurrence of wall slip at the wall of the cavity reduces the developed shear stress and orientation induced in the polymer chains. In addition, when uncoated filler is present, the enhanced polymer-filler interaction may resist the driving force for elastic recovery, to some extent. Addition of around 6% coating to the [Mg(OH).sub.2] filler gives a reversion maximum, yet further addition of fatty-acid coatings then reduces the reversion, probably because of the same lubrication effect that also accounts for the reduction in shear viscosity in these compounds (Figs. 14-15), which was discussed in a previous section. In contrast, the addition of an ATPE coating increases the orientation relative to compounds containing fillers modified by other types of coating. Chain entanglement and physical interaction between the polymer matrix and the aliphatic ATPE chains contribute to a higher viscosity and enhanced melt elasticity of the compound, leading to higher reversion.
Relevance to Product Properties
The addition of a fatty-acid coating to the filler facilitates polymer chain relaxation in thermoplastic compounds at elevated temperature, which is particularly noticeable at addition levels of around 6% by weight. This coating concentration is related to the experimental monolayer coverage, and will be shown in terms of the mechanical testing program (see Part 2 of this paper) to represent a critical addition level for many of the compounds studied. Molecular orientation levels exemplified by the measured reversion are also highly influential on the mechanical properties, and will also be discussed in this context in the second communication. Shear stress distributions generated during injection mold filling not only are responsible for the alignment of the macromolecular chains of the MDPE polymer, but, in addition, create preferential orientation of the high aspect ratio [Mg(OH).sub.2] filler particles. X-ray diffraction (XRD) analysis, also shown in Part 2 of this communication, reveals that the plate-l ike fillers align parallel to the principal flow direction, during injection molding. This alignment is shown to be very high at a coating level of 6% stearic acid. When the platelets are preferentially aligned in the plane of injection molding, the polymer chains show a higher degree of reversion, as the shrinkage direction coincides with the direction of alignment.
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, in combination with the direct effects of filler content and dispersion. Further discussion in this context will be made in Part 2 of this communication. For filled thermoplastic compounds, therefore, 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.
Dry blending process data have allowed optimum mixing times to be determined for the coating of magnesium hydroxide filler by fatty-acids. These depend upon the type and amount of coating that is added to the filler and upon the degree of external heating. FTIR analysis has shown that filler-coating reaction is initiated once the mixer temperature exceeds the coating melting point. Optimum mixing time beyond this point is specific to the coating type. An experimental monolayer can be determined for stearic and behenic acids, but in the case of decanoic acid, continuous reaction occurs as a result of particle attrition in the mixer. Significant quantities of organic magnesium salts are formed, as a result. Consistent experimental trends allow the possibility to predict optimum coating cycles in other, scaled-up dry mixing systems. Coating thickness data have been measured by XPS and are shown to vary with coating chain length and addition level. Detected thicknesses between 10-20A suggest that the aliphatic c hains lie perpendicular to the mineral filler surfaces. Immersion calorimetry data has confirmed the reduction in interaction energy that occurs between mineral fillers and the organic polymer matrix when fatty-acid coatings are applied.
A specific energy parameter from twin screw compounding data reduces on the addition of filler coatings. as a result of the decreased interaction energy that results in enhanced filler dispersion, as confirmed by electron micrographs of cryogenically polished and acid-etched surfaces. ATPE polymeric coatings increase specific energy, which is thought to be attributable to a greater physical interaction with the MDPE chains. All filled MDPE compounds exhibit pseudo-plastic shear flow behavior; fillers increase the shear viscosity of MDPE and fatty-acid coatings have an internal lubrication effect on the materials. Shear flow data are dependent upon die geometry because of the occurrence of wall slip in compounds containing uncoated magnesium hydroxide, the mechanism of which is unclear. Differences observed in filler dispersion when using coatings of variable chain length are therefore due to the influence of the coatings on the melt processing behavior of the compounds, in terms of shear viscosity, internal lubrication and external effects such as wall slip.
Molecular orientation in injection molded specimens has been assessed by heat reversion, which decreases on addition of uncoated Mg[(OH).sub.2], and then increases when coated fillers are incorporated. A clear correlation therefore exists between wall-slip and the observed reversion in compounds containing uncoated filler.
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, ECC International, Alcan Chemicals, Rothon Consultants, Cookson Group, APV Baker Ltd (Industrial Extruder Division), Rosand Ltd. and Stewarts & Lloyds. Contributions from colleagues at Loughborough University (Dr. M. Gilbert and Mr. J.F. Harper) are also gratefully acknowledged.
Institute of Polymer Technology and Materials Engineering (IPTME)
(**.) (Department of Chemistry)
Leicestershire, LE11 3TU, United Kingdom
(1.) J. A. Brydson, Plastics Materials, 4th Ed., Butterworth Scientific, London (1982).
(2.) R N. Rothon, ed., Particulate-Filled Polymer Composites, Longman Scientific (1995).
(3.) T. H. Ferrigno and E. J. Wickson, in Handbook of PVC Forrulating, E. J. Wickson, ed., Wiley, New York (1993).
(4.) E. Fekete, B. Pukanszky, A. Toth, and I. Bertoti, J. Calloid & Interface Sci., 135, 200(1990).
(5.) P. R. Hornsby and C. L. Watson, J. Mat Sci., 30, 5347 (1995).
(6.) J. B. Griffiths, Plast. Rubb. Proc. & Appl., 13, 3(1990].
(7.) H. S. Katz and J. V. Milewski, Handbook of Fillers for Plastics, 2nd Ed., Van Nostrand Reinhold, New York (1978).
(8.) E. P. Plueddemann and G. L. Stark, 32nd Annual Technical Conference, Reinforced Plastics/Composites Institute, 4-C, 1 (1977).
(9.) L. Choplin, Proc. MOFFIS Conf., 319, Namur, Belgium (1993).
(10.) H. Balard and E. Papirer, Proc. MOFFIS Conf. 213, Namur, Belgium (1993).
(11.) M. J. Jaycock and G. D. Partfitt, Chemistry of Interfaces, Wiley, New York (1986).
(12.) Y. Bomal and P. Godard, Polym. Eng. Sci., 36, 237 (1996).
(13.) P. R. Hornsby and A. Mthupha, J. Mat. Sci., 29, 5293 (1994).
(14.) E. Sheng and I. Sutherland, Surface Sci., 314, 325 (1994).
(15.) C. L. Raymond, PhD thesis, Loughborough University, U.K. (1997).
(16.) M. Mooney, J. Rheol, 2, 210 (1931).
(17.) British Standard, BS 2782: Part 9, Method 941, BSI, London (1989).
(18.) C. Liauw, PhD thesis, Manchester Metropolitan University, U.K. (1994).
(19.) J. Jancar and J. Kucera, Polym, Eng. Sci., 30, 707 (1990).
(20.) B. Pukanszky, E. Fekete, and F. Tudos, Makromol. Chem., Makromol. Symp., 28, 165(1989).
(21.) E. Papirer, J. Schultz, and C. Turchi, European. Polym. J., 20, 1155 (1984).
(22.) M. Dillon, Plast. Rubb. Comp. Proc. Appl., 24, 267 (1995).
(23.) H. P. Schlumpf in Plastics Additives, 4th Ed., R. Gachter and H. Muller, eds., Hanser, Munich (1993).
(24.) Q. Fu and G. Wang, Polym. Eng. Sci., 32,94(1992).
(25.) Y. N. Sharma, R. D. Patel, I. H. Dhimmar, and I. S. Bhardwaj. J. Appl. Polym. Sci., 27, 97(1982).
(26.) M. Cook, PhD thesis, Loughborough University, U.K. (1996).
(27.) A. W. Birley, B. Haworth, and J. Batchelor, Physics of Plastics, Hanser, Munich (1991).
(28.) J. A. Brydson, Flow Properties of Polymer Melts, 2nd Ed., Godwin, London (1981).
(29.) S. Hatzikiriakos, C. Stewart, and J. M. Dealy, Intern. Polym. Proc., 8, 30 (1993).
(30.) S. Hatzikiriakos and J. M. Dealy, Intern. Polym. Proc., 8, 36 (1993).
(31.) J. Tordella, J. Appl. Polym. Sci. 7, 215 (1963).
(32.) D. Hill, T. Hasegawa, and M. Denn, J. Rheol, 34. 891 (1990).
|Printer friendly Cite/link Email Feedback|
|Author:||HAWORTH, B.; RAYMOND, C. L.; SUTHERLAND, I.|
|Publication:||Polymer Engineering and Science|
|Date:||Sep 1, 2000|
|Previous Article:||Recycling of Poly(Ethylene Terephthalate) Into Closed-Cell Foams.|
|Next Article:||Barrier Properties of Blends Based on Liquid Crystalline Polymers and Polyethylene.|