Barrier and mechanical properties of injection molded montmorillonite/polyesteramide nanocomposites.
Nanocomposites have attracted considerable attention [1-3] since the discovery of their potential in improving polymer properties [4-6]. Two major findings have triggered interest in these materials. First, the report from the Toyota research group  showing the thermal and mechanical properties achieved with a polyamide 6/montmorillonite composite, and second, the report from Vaia et al.  that it is possible to melt mix polymers with silicates.
The clay minerals commonly used in nanocomposites belong to the structural family known as the 2:1 phyllosilicates . The crystal structure of these clay minerals is described as two tetrahedral sheets sandwiching an edge-shared octahedral sheet . The tetrahedral sheets consist of silicon-oxygen tetrahedra and they are arranged in a hexagonal network, in with each tetrahedron shares three oxygen atoms with other tetrahedra and with the central octahedral sheet. The edge-shared octahedral layers consist of oxygen and hydroxyl ions that form octahedral sites in which aluminum and magnesium are located .
The layers in montmorillonite are arranged in stacks, which lead to a periodic van der Waals gap or interlayer. This so-called gallery is normally 3-5 [Angstrom] thick. [Al.sup.3+] is replaced with [Mg.sup.2+] in the octahedral sheet, and a negative charge is thus generated within the layer. To counterbalance the negative charge, [Ca.sup.2+] and [Na.sup.+] are located in the gallery [1, 11]. By replacing the charge-balancing interlayer-cations with, e.g., alkyl-ammonium ions, the compatibility between the clay mineral and the organic phase may be substantially enhanced. The alkyl-ammonium cations lower the surface energy of the inorganic host and improve the wetting characteristics with the polymer .
In general, the polymer-clay composites can be divided into three categories: conventional composites, intercalated nanocomposites, and delaminated nanocomposites . In conventional composites/immiscible blends, the clay acts as a conventional filler. In intercalated nanocomposites, the polymer chains are located between the clay layers resulting in a well-ordered multilayer with alternating polymer/clay layers and a periodic distance of a few nanometers. In delaminated nanocomposites, the clay layers are dispersed in a continuous polymer matrix. Delaminated nanocomposites have stiffness, strength, and barrier properties comparable to those of conventional glass- or mineral-reinforced polymers but with far less filler content [13-15]. The efficiency of the clay in modifying the properties of the polymer is primarily determined by its degree of dispersion in the polymer matrix . In the delaminated nanocomposite, the entire surface is available to interact with the polymer.
Biodegradable polymers are of interest for environmental protection and conservation. Waste management is extremely important and there is a need for environmentally compatible solutions to reduce the use of non-degradable components in packaging. One of the factors that limits the use of currently available biodegradable polymers in packaging applications is their inferior barrier properties [17-19]. However, by incorporating nano-scale particles it may be possible to enhance the barrier properties, which would make these biodegradable polymers useful in packaging applications.
The properties and synthesis of biodegradable polyesters with nano-particles have been reported elsewhere [6, 20]. However, data on their barrier properties have not been found in the literature.
As reported in a previous study , the extrusion and extrusion/compression molding of polyesteramide/octadecylamine-treated montmorillonite clay showed interesting barrier and mechanical properties. The objective of the work described in this paper was to investigate whether it is possible to achieve an even greater improvement in barrier properties by using a tougher process treatment. The morphology of the nanocomposites was assessed by X-ray scattering, differential scanning calorimetry, dynamic mechanical thermal analysis, transmission electron microscopy (TEM), and density measurements. These data were compared with the transport and mechanical properties of the extruded nanocomposites obtained in the previous study .
C18-Amine Organoclay. Nanocor Inc. kindly supplied the C18-Amine Organoclay, named NANOMER I.30E, which was delivered in powder form. The clay was an octadecylamine-treated montmorillonite with a montmorillonite clay content of ~68 wt%. The mineral purity was stated by the producer to be > 98.43%. The density of the C18-Amine Organoclay was 290 kg/[m.sup.3] (Ohaus free fall method). The density of pure montmorillonite is 2608 kg/[m.sup.3] . The C18-Amine Organoclay is herein after denoted "filler" or ODA-clay.
Polyesteramide. Bayer kindly supplied an experimental grade of the polyesteramide, BAK 404-004 (referred to as BAK). The polymer had a density of 1160 kg/[m.sup.3] at 20[degrees]C and [bar.M.sub.w] = 37,000 g/mol and a melting peak temperature in the range of 115-125[degrees]C. The polymer is considered to be biodegradable according to DIN 54900 ([T.sub.g] [approximately equal to] -10[degrees]C).
Masterbatch. A masterbatch was produced by Polykemi AB, Ystad, Sweden. The mixing was performed on a Werner & Pfleiderer twin-screw extruder at a rotation speed of 250 rpm. The temperature profile was: 170[degrees]C (zone 1), 180[degrees]C (zone 2), 190[degrees]C (zone 3), 190[degrees]C (zone 4), and 200[degrees]C (zone 5). The polymer and filler were mixed manually in a dry condition before the melt mixing in the extruder. The filler content of the master batch was 13 wt% (SD = 0.3 wt%) as revealed by thermogravimetry in a nitrogen atmosphere.
Extrusion. The 5 wt% samples were prepared by mixing the masterbatch with pure polyesteramide through a meltextrusion step prior to the injection molding. A Brabender counter-rotating twin-screw extruder DSK 35/9 D equipped with a Brabender adjustable flat sheet die head (100 * 0-1.5 mm) was used. The processing conditions were monitored by a Brabender data processing Plasti-Corder PL 2000 coupled to the extruder. Prior to extrusion, the masterbatch and the pure BAK were dried at 100[degrees]C for 2 hr in a Piovan hopper (Model T 10 IX) coupled to a Piovan air-drier (Model DSN 403). A temperature sensor placed up-stream of the die recorded the temperature of the melt. The recorded temperature represents the temperature at the polymer-metal interface. The samples were cooled in air.
Injection Molding. The injection-molding machine used was a Battenfeld BA500/125 CDK UNILOG TC140[degrees]scl equipped with a three-zone-screw (outer diameter = 25 mm; L/D ratio = 21.6; compression ratio = 1.8). Prior to injection molding, the materials were dried at 90[degrees]C for 2 hr in a Motan air-drier (Model MD 11). The mold consisted of two stainless steel blocks. The diameter of the circular cavity in the movable mould part was 135 mm and the thickness was 1 mm. The gate with a diameter of 4.5 mm was located in the center of the circular cavity. The following process parameter settings were used: temperature gradient: [T.sub.1] = 180[degrees]C, [T.sub.2] = 170[degrees]C, and [T.sub.3] = 170[degrees]C; nozzle temperature = 170[degrees]C; and hot channel attachment temperature = 170[degrees]C. Injection pressure 1300 bar during 0.4 s, holding pressure 1100 bar during 1 s and injection rate 55 [cm.sup.3]/s. The mold temperature was 35[degrees]C. The cooling period was 10 s.
Thermogravimetric Analysis. To assess the thermal degradation of the surface treatments, isothermal and dynamic thermogravimetric analyses were carried out on a Mettler-Toledo thermobalance (TGA/SDTA 85[1.sup.e]). Samples with a weight between 250 [+ or -] 1 mg were inserted in an aluminum oxide crucible and the weight was measured in air at 200[degrees]C and 850[degrees]C during 10 min. The dynamic thermogravimetric measurements were made by heating 10.53 [+ or -] 0.02 mg of the material in an aluminum oxide crucible from 35[degrees]C to 830[degrees]C (10[degrees]C/min) in air. The large sample sizes provided a more representative sampling of the "average" material. However, care must be taken when evaluating the physically true degradation temperature from the degradation kinetics of these large samples. That is why a smaller sample size was used for the dynamic measurements. All data are based on two measurements on each sample.
Dynamic Mechanical Thermal Analysis. The dynamic mechanical measurements were performed on a DMTA Mark II, from Polymer Laboratories, working in the bending mode (single cantilever) at 1 Hz. The strained samples had cross-sections of approximately 1*10 mm and a length of 35 mm. The temperature scans were performed at a heating rate of 2[degrees]C/min from -40[degrees]C to 40[degrees]C. Liquid nitrogen was used as cooling medium.
Oxygen Permeability. The oxygen transmission rate was determined at 23[degrees]C and 0% RH, using a Mocon Ox-Tran Twin apparatus, according to ASTM D 3985-95. The specimens were mounted in isolated diffusion cells and subsequently purged with nitrogen gas (2% hydrogen) in order to measure the background oxygen leakage of the instrument. Each sample was tightly sandwiched between two aluminum foils providing a 5 [cm.sup.2] active area for the measurements. One side of the sample was exposed to flowing oxygen (99.95%) at atmospheric pressure after the background measurements. The oxygen transmission rate was normalized with respect to the oxygen pressure and the film thickness to yield the oxygen permeability (OP). Measurements were made on two replicates of each sample.
Density Measurements. The densities of the materials were determined by applying the Archimedes principle using a Mettler Toledo AE 100 balance and a Mettler-density determination kit 33360. The weight of the specimen was recorded in air and in n-hexane ([rho] = 659 kg/[m.sup.3]).
Differential Scanning Calorimetry (DSC). The melting endotherms were obtained using a Mettler DSC 820 working under a nitrogen atmosphere (50 ml/min). Samples (5 mg [+ or -] 0.3 mg) were sealed into 40 [micro]l aluminum pans, and thermograms were taken from -50 to 220[degrees]C at a heating rate of 10[degrees]C/min.
X-ray Diffraction (XRD). X-ray diffractograms were obtained in a Siemens D5000 diffractometer with Cu radiation (50 kV, 40 mA). The scanning speed and the step size were 0.15[degrees]/min and 0.02[degrees]. The rotation speed of the sample was 15 rpm. The samples had an average thickness of 1 mm and were cut into 10* 10 m[m.sup.2] squares.
Mechanical Properties. The mechanical properties were measured using an Alwetron TCT 10 tensile tester. Dumbbell shaped specimens (thickness: 1 mm; length of narrow portion: 20 mm; width of narrow portion: 4 mm; width of ends: 8 mm; gauge length: 20 mm) were used. The samples were stamped out from the sheets using an Elastocon stamping knife Model EP 04 equipped with discharging stamp. The samples were conditioned and tested at 23[degrees]C and 50% RH. The strain rate was 100 mm/min, and the strain was measured as the separation of the clamps. The reported results are based on the averages of data from 10 specimens for each material.
[FIGURE 1 OMITTED]
TEM. The sample morphology was studied by TEM using a JEOL 1200. Sections with a thickness of 80 nm were obtained by cryo-sectioning at -40[degrees]C with a cutting speed of 25 mm/s. The samples were collected using a sucrose droplet in a metal loop.
Scanning Electron Microscopy (SEM). SEM was performed on specimens cracked at liquid nitrogen temperature. The samples were gold/palladium-coated and examined in a JEOL JSM-5400.
RESULTS AND DISCUSSION
The density of injection-molded composites and extruded composites  as well as the "compact" unmodified density of the filler are plotted in Fig. 1 as a function of the filler content. A closer examination of density data revealed that the injection-molded materials were denser than the extruded materials at the same filler content, although the difference was small, ~0.6% in density. The extrapolation of the density data to 100% filler content revealed that voids were probably present in the composite. On average, the void content in the injection-molded composites was 31 vol% per 100 vol% of compact filler compared to 36 vol% in the extruded composites . Thus, the higher shear rates and tougher process conditions associated with injection molding had only a small effect on the reduction in the void content.
[FIGURE 2 OMITTED]
The diffraction of the 001 planes in the filler stacks, normally observed as an XRD peak at 2[theta] = 3.7 (23.7 [Angstrom]) , appeared for all materials. In fact, the X-ray patterns of the injection-molded samples were very similar to those of the extruded/compression molded samples reported previously  (Fig. 2). The injection-molded samples contained clay stacks that had been intercalated to yield a periodic distance of 34-35 [Angstrom]. However, TEM revealed that the clay layers were to a great degree delaminated, although intercalated or stack-like structures were observed, especially in the 13 wt% system (Figs. 3 and 4). TEM showed that the clay layers were oriented "unidirectionally" over several microns. This was also observed for the extruded/compression-molded samples but mainly for the 13 wt% system. This was probably, at least to some extent, a consequence of steric effects (large aspect ratio and high concentration of clay). However, the melt flow characteristics and higher shear rates in injection molding not only favored the local orientation of the clay layers shown in Figs. 3 and 4, but also yielded a macroscopic orientation. Although two knitlines were present, the radially flowing melt yielded an "average structure" that consisted of radially oriented polymer chains and fillers. This was observed by heating sections of the molded specimens. The sections shrunk in the radial direction and grew in the tangential and thickness directions. Such an anisotropic dimensional behavior during heating was not observed for the extruded or extruded/compression-molded sheets. It should be noted that the distinct skin-core morphology typical of injection-molded materials was not observed using SEM. However, when the 13 wt% sample was cut, it cracked in a brittle manner and delaminated sheets of material oriented along the plane of the sheet were observed in the cut and crack surfaces. This did not occur with the pure polymer, the 5 wt% injection-molded sample, or the extruded and extruded/compression-molded samples. As is shown below, the 13 wt% sample was ductile in the plane of the sheet, emphasizing that the injection-molded sheets were more anisotropic than the extruded and extruded/compression-molded sheets. These data suggest that the clay layers were oriented, on average, with their broad surfaces in the plane of the sheet. This could explain the improvement in barrier properties compared to those of the extruded and extruded/compression-molded sheets reported below.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
No major difference in melting enthalpy was detected between the unfilled and filled specimens. A comparison of the melting enthalpy data from previous work (extruded composites)  and the melting enthalpy of the injection-molded composites reveals an increase from 32 J/g (extruded) to 45.3 J/g (injection-molded). This difference may be due to the fact that cooling from the melt was faster in the case of extrusion than in injection molding. The injection process also promotes a higher degree of orientation, and creates a higher crystallinity.
The glass transition temperature did not change with increasing filler content, but a broadening of the transition peak was observed with increasing filler content. The polymer thus seemed to be more constrained in the presence of nanoparticles.
Dynamic mechanical measurements showed that the tan [delta] peak was shifted slightly towards higher temperatures in both the 5 wt% and the 13 wt% systems. The resolved tan [delta] peak in the vicinity of 13[degrees]C probably originated from a constrained polymer layer next to the filler surface (Fig. 5).
According to TGA-data, the degradation of ODA during heating at 10[degrees]C/min started above 200[degrees]C (Fig. 6a). On the other hand, when the weight loss was recorded isothermally at 200[degrees]C, ODA was basically unaffected over a period of several minutes (Fig. 6b). Thus, it was concluded that the melt temperature during injection molding (170[degrees]C) had no detrimental effect on the ODA-surface. For comparison, the extensive surface degradation at 850[degrees]C is also shown in Fig. 6b.
[FIGURE 6 OMITTED]
The transmission rates through the injection-molded materials decreased more strongly with increasing filler content than those of the extruded and extruded/compression molded materials . The oxygen transmission rates of the composites were 20% (13 wt% filler) and 40% (5 wt% filler) of those of the unfilled material.
In order to model the reduction in permeability with the presence of filler, the permeability of the composite [P.sub.c] is defined as:
[P.sub.c] = [P.sub.m][v.sub.m]/[tau] (1)
where [P.sub.m] is the permeability of the polymer matrix and [v.sub.m] is the volume fraction of polymer in the composite. [tau] is the tortuosity factor, which increases with increasing impedance efficiency of the filler. Provided that the montmorillonite layers were completely delaminated and that the reduction in permeability was a function solely of the content of filler and its width (w) and thickness (l), the tortuosity factor could be estimated using the following relationship for oblate spheroids, based on the Fricke theory [23, 24]:
[FIGURE 7 OMITTED]
X = [1 - [v.sub.m]/[tau] - 1] (2)
where X is obtained from
[w/l] = [1/[0.785 - [square root of (0.616 - [X/[X + 3]])]]]. (3)
The expected permeability of composites with particles having aspect ratios of 200 and 600 is plotted in Fig. 7. These aspect ratios were estimated from information given in reference . Fitting Eqs. 1-3 to the experimental transmission rates in Fig. 7 yielded aspect ratios of the order of 270-280. These calculated aspect ratios agree well with the aspect ratios estimated from TEM data in Figs. 3 and 4.
The TEM data obtained for the injection-molded samples agreed well with the TEM data for the extruded and extruded/compression-molded samples presented in the previous paper . Assuming that the sheets were completely delaminated, the aspect ratios ranged from 40 to 400 with an average value close to 200. The discrepancy between the calculated and experimental aspect ratios could be attributed to the fact that the Fricke model is applicable to isotropic systems whereas the injection-molded system was anisotropic. In contrast to the modest decrease in oxygen permeability with increasing filler content in the extruded and extruded/compression-molded composites, the decrease was greater for the injection-molded composites and in accordance with the "expected"/predicted permeability.
The greater improvement in the barrier properties achieved by injection molding, compared to extrusion and extrusion/compression molding, was probably a combined effect of the higher crystallinity (DSC data), the lower void content (density data), and the greater degree of filler-polymer orientation.
Table 1 presents data for the Young's modulus, yield stress, and fracture strain of the injection-molded sheets. For comparison, the table also includes data from extruded and extruded/compression-molded sheets reported previously . The modulus and yield stress increased with increasing filler content. For the most highly filled composite (13 wt%) the Young's modulus was 4.1 times and the yield stress was 1.8 times greater than for the unfilled polymer. The injection-molded specimens were stiffer and yielded at a higher stress level than the corresponding extruded and extruded/compression-molded materials. This is expected, because the injection-molded material had a higher crystallinity. The stiffness increased by a factor of 2 to 4.1 and the yield stress increased by a factor of 1.2 to 1.8 for the filled systems compared to the pure polymer. Interestingly, the filled injection-molded samples, in contrast to the extruded and extruded/compression-molded samples, were tougher than the pure polymer. Apparently, the lower void content and the actual composite morphology of the injection-molded materials contributed to the enhanced toughness. Provided that the clay layers were oriented mainly with their widths in the plane of the sheet, this means that the toughness was enhanced along the width of the clay layers and, as shown earlier, brittle perpendicular to it. It may be speculated that the reason is that clay layer gliding can occur during plastic deformation in the former direction whereas a similar gliding is prohibited in the latter direction.
As with the barrier properties, the greater improvement in the mechanical properties obtained with injection molding was a combined effect of the higher crystallinity, the lower void content, and the greater degree filler-polymer orientation.
In contrast to extruded and extruded/compression-molded octadecylammonium-treated montmorillonite clay and polyesteramide composites, the injection-molded samples showed significantly improved oxygen barrier properties. This is attributed to a combined effect of a lower void content and a higher polymer crystallinity. In addition, the polymer chain and filler were more oriented in the plane of the sheet in the injection-molded samples. This anisotropy evidently yielded a very ductile specimen in the plane of the sheet. However, the sheet was brittle in cutting operations. Both X-ray and TEM studies indicated that clay stacks were still present in the sheets but that the periodic layer distance was 10 [Angstrom] greater than that of the ODA-fillers.
TABLE 1. Mechanical properties of ODA/BAK.* Sample E(0%) (a) E(5%) (a) E(13%) (a) [sigma](0%) (b) Extruded 165 354 (2.2) 436 (2.6) 11 Extruded/ compression 203 364 (1.8) 542 (2.7) 13 Injection 328 651 (2) 1358 (4.1) 17 Sample [sigma](5%) (b) [sigma](13%) (b) [epsilon](0%) (c) Extruded 15 (1.4) 18 (1.6) 518 Extruded/ compression 15 (1.1) 18 (1.4) 497 Injection 21 (1.2) 30 (1.8) 188 Sample [epsilon](5%) (c) [epsilon](13%) (c) Extruded 480 (.93) 277 (.54) Extruded/ compression 390 (.79) 147 (.30) Injection 289 (1.5) 221 (1.18) *The data for extruded and extruded/compression molded samples were taken from Ref. 21. The values within parentheses in the heading are filler contents. The values within parentheses below the heading are the ratios of the measured value to those of the unfilled polymer. (a) Young's modulus (MPa). The maximum relative errors in the data were 15% for the extruded and extruded/compression-molded and 11% for the injection molded-samples. (b) Yield stress (MPa). The maximum relative errors in the data were 12%. (c) Fracture strain (%). The maximum relative errors in the data were 47% for the extruded and extruded/compression-molded and 14% for the injection-molded.
M. Gallstedt, STFI-Packforsk, Stockholm, is thanked for experimental work. B. Lindskog, STFI-Packforsk, Stockholm, is thanked for valuable help. R. Olsson, Royal Institute of Technology, is gratefully acknowledged for his experimental work and valuable comments.
1. S.S. Ray and M. Okamoto, Progress Polymer Science, 28, 1539 (2003).
2. E. Manias, Nanocomposites 2001, Delivering New Value to Plastics, A. Golovoy, editor, Executive Conference Management, Inc., Plymouth, MI (2001).
3. M. Alexandre and P. Dubois, Materials Science Engineering, 28, 1 (2000).
4. T. Lan, P.D. Kaviratna, and T.J. Pinnavaia, Chemistry Materials, 6, 573 (1994).
5. K.E. Strawhecker and E. Manias, Chemistry Materials, 12, 2943 (2000).
6. S.T. Lim, Y.H. Hyun, and H.J. Choi, Chemistry Materials, 14. 1839 (2002).
7. A. Okada, M. Kawasumi, T. Kurauchi, and O. Kamigaito, Polymer Preprints, 28, 447 (1987).
8. R.A. Vaia, H. Ishii, and E.P. Giannelis, Chemistry Materials, 5, 1694 (1993).
9. U. Hoffman, K. Endell, and D. Wilm, Z. Krist, 86, 340 (1933).
10. Montmorillonite, Encyclopedia Britannica Online (2000).
11. E.P. Giannelis, Advanced Materials, 8, 29 (1996).
12. E.P. Giannelis, R. Krishnamoorti, and E. Manias, Advances Polymer Science, 138, 107 (1999).
13. Y. Kojima, A. Usuki, M. Kawasumi, A. Okada, Y. Fukushima, T. Kurauchi, and O. Kamigaito, J. Materials Research, 8, 1185 (1993).
14. Y. Kojima, K. Fukumori, A. Usuki, A. Okada, and T. Kurauchi, J, Material Science, Letters, 12, 889 (1993).
15. S.D. Burnside and E.P. Giannelis, Chemistry Materials, 7, 1597 (1995).
16. A. Akelah, N. Salah El-Deen, A. Hiltner, E. Baer, and A. Moet, Materials Letters, 22, 97 (1995).
17. PLLA, Technical Data Sheet from Neste (1997).
18. CPLA, Technical Data Sheet from DIC (1997).
19. BAK, Technical Data Sheet from Bayer (1997).
20. S.-R. Lee, H.-M. Park, H. Lim, T. Kang, X. Li, W.-J. Cho, and C.-S. Ha, Polymer, 43, 2495 (2002).
21. M. Krook, A.-C. Albertsson, U.W. Gedde, and M.S. Hedenqvist, Polymer Engineering Science, 42, 1238 (2002).
22. R.E. Grim, Clay Mineralogy, 2nd ed., McGraw-Hill, New York (1968).
23. A.S. Michaels and H.J. Bixler, J. Polymer Science, 5, 413 (1961).
24. H. Fricke, Physics Review, 24, 575 (1924).
25. R. Krishnamoorti, R.A. Vaia, and E.P. Giannelis, Chemistry Materials, 8, 1728 (1996).
STFI-Packforsk, Box 9, SE-164 93 Kista, Sweden
ISS Group Services Limited, Pellow House, Francis Road, Withington, Manchester M20 4XP, UK
Royal Institute of Technology, Department of Fibre and Polymer Technology, SE-100 44 Stockholm, Sweden
Correspondence to: M.S. Hedenqvist, e-mail: email@example.com
|Printer friendly Cite/link Email Feedback|
|Author:||Krook, M.; Morgan, G.; Hedenqvist, M.S.|
|Publication:||Polymer Engineering and Science|
|Date:||Jan 1, 2005|
|Previous Article:||Dynamic finite element analysis of nonlinear isotropic hyperelastic and viscoelastic materials for thermoforming applications.|
|Next Article:||In situ generated diphenylsiloxane-polyimide adduct-based nanocomposites.|