Progress in Polymer Science 33 (2008) 1119–1198

Contents lists available at ScienceDirect

Progress in Polymer Science journal homepage: www.elsevier.com/locate/ppolysci

A review on polymer–layered silicate nanocomposites S. Pavlidou a , C.D. Papaspyrides b,∗ a b

CLOTEFI, Clothing, Textile and Fibre Technological Development, 4 El. Venizelou Str., Kallithea, Athens 176 76, Greece Laboratory of Polymer Technology, Department of Chemical Engineering, National Technical University of Athens, Zographou, Athens 157 80, Greece

a r t i c l e

i n f o

Article history: Received 25 June 2007 Received in revised form 11 July 2008 Accepted 25 July 2008 Available online 25 September 2008 Keywords: Nanocomposites Polymers Layered silicates Clays

a b s t r a c t This review reports recent advances in the field of polymer–layered silicate nanocomposites. These materials have attracted both academic and industrial attention because they exhibit dramatic improvement in properties at very low filler contents. Herein, the structure, preparation and properties of polymer–layered silicate nanocomposites are discussed in general, and detailed examples are also drawn from the scientific literature. © 2008 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3.

4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Milestones in the research on polymer–layered silicate nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Layered silicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Structure and characteristics of layered silicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Organic modification of layered silicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanocomposite structures and characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Nanocomposite structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Nanocomposite structural characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preparation of nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1. Template synthesis (sol–gel technology) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2. Intercalation of polymer or prepolymer from solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.3. In situ intercalative polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4. Melt intercalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Intercalation of polymer from solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. In situ intercalative polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1. In situ intercalative polymerization of thermoplastic polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2. In situ intercalative polymerization of thermosetting polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. Polymer melt intercalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1. Introduction and advantages of the technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2. Factors affecting polymer melt intercalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author. E-mail address: [email protected] (C.D. Papaspyrides). 0079-6700/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.progpolymsci.2008.07.008

1120 1121 1122 1122 1123 1126 1126 1127 1129 1129 1129 1130 1130 1130 1130 1131 1131 1137 1142 1142 1144

1120

6.

7. 8.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

5.4.3. Compatibility issues in non-polar polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4. Degradation problems encountered during melt intercalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanocomposite properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1. The reinforcing mechanism of layered silicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2. Modulus and strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3. Toughness and strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.4. Comparison and synergistic effects of clays and conventional reinforcements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Dynamic mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3. Barrier properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4. Thermal stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5. Flame retardance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1. Flame retardance of polymer–layered silicate nanocomposites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2. Synergism between nanocomposites and flame retardants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6. Heat distortion temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7. Rheological properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.8. Crystallinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.9. Biodegradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.10. Photo-degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.11. Optical clarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanocomposites: advantages and applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Traditionally, polymeric materials have been filled with synthetic or natural inorganic compounds in order to improve their properties, or simply to reduce cost. Conventional fillers are materials in the form of particles (e.g. calcium carbonate), fibers (e.g. glass fibers) or plate-shaped particles (e.g. mica). However, although conventionally filled or reinforced polymeric materials are widely used in various fields, it is often reported that the addition of these fillers imparts drawbacks to the resulting materials, such as weight increase, brittleness and opacity [1–5]. Nanocomposites, on the other hand, are a new class of composites, for which at least one dimension of the dispersed particles is in the nanometer range. Depending on how many dimensions are in the nanometer range, one can distinguish isodimensional nanoparticles when the three dimensions are on the order of nanometers, nanotubes or whiskers when two dimensions are on the nanometer scale and the third is larger, thus forming an elongated structure, and, finally, layered crystals or clays, present in the form of sheets of one to a few nanometers thick and hundreds to thousands nanometers in extent [1,4]. Among all the potential nanocomposite precursors, those based on clay and layered silicates have been most widely investigated, probably because the starting clay materials are easily available and because their intercalation chemistry has been studied for a long time [6]. Polymer–layered silicate nanocomposites, which are the subject of the present contribution, are prepared by incorporating finely dispersed layered silicate materials in a polymer matrix [2]. However, the nanolayers are not easily dispersed in most polymers due to their preferred faceto-face stacking in agglomerated tactoids. Dispersion of the tactoids into discrete monolayers is further hindered by the intrinsic incompatibility of hydrophilic layered sil-

1151 1154 1156 1156 1156 1157 1164 1168 1169 1170 1173 1178 1178 1183 1185 1185 1186 1188 1189 1190 1190 1193 1193

icates and hydrophobic engineering plastics. Therefore, layered silicates first need to be organically modified to produce polymer–compatible clay (organoclay). In fact, it has been well-demonstrated that the replacement of the inorganic exchange cations in the cavities or “galleries” of the native clay silicate structure by alkylammonium surfactants can compatibilize the surface chemistry of the clay and a hydrophobic polymer matrix [7]. Thereafter, different approaches can be applied to incorporate the ion-exchanged layered silicates in polymer hosts by in situ polymerization, solution intercalation or simple melt mixing. In any case, nanoparticles are added to the matrix or matrix precursors as 1–100 ␮m powders, containing associated nanoparticles. Engineering the correct interfacial chemistry between nanoparticles and the polymer host, as described previously, is critical but not sufficient to transform the micron-scale compositional heterogeneity of the initial powder into nanoscale homogenization of nanoparticles within a polymeric nanocomposite [8]. Therefore, appropriate conditions have to be established during the nanocomposite preparation stage. The resulting polymer–layered silicates hybrids possess unique properties – typically not shared by their more conventional microscopic counterparts – which are attributed to their nanometer size features and the extraordinarily high surface area of the dispersed clay [1,4]. In fact, it is well established that dramatic improvements in physical properties, such as tensile strength and modulus, heat distortion temperature (HDT) and gas permeability, can be achieved by adding just a small fraction of clay to a polymer matrix, without impairing the optical homogeneity of the material. Most notable are the unexpected properties obtained from the addition of stiff filler to a polymer matrix, e.g. the often reported retention (or even improvement) of the impact strength. Since the weight fraction of the inor-

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1121

Nomenclature AIBN N,N -azobis(isobutyronitrile) CEC cation exchange capacity CX-clay X is the number of carbon atoms in clay organic modifier d spacing between diffractional lattice planes D diffusivity DETDA diethyl toluene diamine DGEBA diglycidyl ether of bisphenol A DM dioctadecyldimethyl ammonium chloride DMA dynamic mechanical analysis DSC differential scanning calorimetry DTGA differential thermogravimetric analysis E storage modulus under bending mode E loss modulus under bending mode EVA ethyl-vinyl-acetate copolymer FTIR Fourier transform infrared spectroscopy fwhm full-width-at-half-maximum G storage modulus under tensile mode G loss modulus under tensile mode GPC gas-permeation chromatography HDPE high density polyethylene HDT heat distortion temperature HRR heat release rate MMT montmorillonite NMR nuclear magnetic resonance o-Clay organo-modified clay OMLS organo-modified layered silicate, organosilicate, or organoclay P permeability PA polyamide PAA poly(acrylic acid) PBO polybenzoxale PCL polycaprolactone PCN polymer–clay nanocomposite PDMS polydimethylsiloxane PE polyethylene PEI poly(ether imide) PEO poly(ethylene oxide) PET poly(ethylene terephthalate) PHB poly(3-hydroxybutyrate) PHRR peak heat release rate PI polyimide PA polylactide PLS polymer–layered silicate nanocomposite PMMA polymethyl methacrylate PP polypropylene PP-MA or PP-g-MA maleic anhydride-grafted polypropylene PS polystyrene PSF polysulfone PU polyurethane PVA poly(vinyl acetate) PVC poly(vinyl chloride) PVE poly(vinyl ethylene) PVOH poly(vinyl alcohol) PVP poly(vinyl pyrrilidone) S solubility

SBS SEA Tan ı Tc TEM Tg TGA TGAP TGDDM

poly(styrene-butadiene-styrene) specific extinction area G /G crystallization temperature transmission electron microscopy glass transition temperature thermogravimetric analysis triglycidyl p-aminophenol tetrafunctional tetraglycidyldiaminodiphenylmethane TPO thermoplastic olefin UP unsaturated polyester WAXD or WAXS wide angle X-ray diffraction or scattering XRD X-ray diffraction

ganic additive is typically below 10%, the materials are also lighter than most conventional composites [2,9–12]. These unique properties make the nanocomposites ideal materials for products ranging from high-barrier packaging for food and electronics to strong, heat-resistant automotive components [11]. Additionally, polymer–layered silicate nanocomposites have been proposed as model systems to examine polymer structure and dynamics in confined environments [12,13]. However, despite the recent progress in polymer nanocomposite technology, there are many fundamental questions that have not been answered. For example, how do changes in polymer crystalline structure induced by the clay affect overall composite properties? How does one tailor organoclay chemistry to achieve high degrees of exfoliation reproducibility for a given polymer system? How do process parameters and fabrication affect composite properties? Further research is needed that addresses such issues [14]. The objective of this work is to review recent scientific and technological advances in the field of polymer–layered silicate nanocomposite materials and to develop a better understanding of how superior nanocomposites are formed. 2. Milestones in the research on polymer–layered silicate nanocomposites The incorporation of layered silicates into polymer matrices has been known for over 50 years [15]. In fact, one of the earliest systematic studies of the interaction between a clay mineral and a macromolecule dates back to 1949, when Bower [16] described the absorption of DNA by montmorillonite. Even in the absence of X-ray diffraction (XRD) evidence, this finding implied insertion of the macromolecule in the lamellar structure of the silicate. In 1950, Carter et al. [17] developed organoclays with several organic onium bases to reinforce latex-based elastomers and in 1958, Hauser and Kollman [18] were granted a patent for “clay complexes with conjugated unsaturated aliphatic compounds of four to five carbon atoms”. Uskov [19], in 1960, found that the softening point of poly(methyl methacrylate) derived by polymerization of

1122

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

methyl methacrylate was raised by montmorillonite modified with octadecylammonium, while in the following year Blumstein [20] obtained a polymer inserted in the structure of a montmorillonite by polymerizing a previously inserted vinyl monomer. Two years later, Greenland [21] used a poly(vinyl alcohol)/montmorillonite system to show that a polymer could be directly inserted in a clay in an aqueous solution. The same year, the incorporation of organoclay into a thermoplastic polyolefin matrix was disclosed by Nahin and Backlund [22] of Union Oil Co. They obtained organoclay composites with strong solvent resistance and high tensile strength by irradiation-induced cross linking. However, they did not focus on the intercalation characteristics of the organoclay or the potential properties of the composites. In 1975, Tanihara and Nakagawa [23] reached a similar result by intercalating polyacrylamide and poly(ethylene oxide) from an aqueous solution. In 1976 Fujiwara and Sakamoto [24] of the Unichika Co. described the first organoclay hybrid polyamide nanocomposite. However, it was not until Toyota researchers began a detailed examination of polymer–layered silicate composites that nanocomposites became more widely studied in academic, government and industrial laboratories [25–28]. The Toyota research group disclosed improved methods for producing nylon 6 clay nanocomposites using in situ polymerization similar to the Unichika process. They reported that these polymer–clay nanocomposites exhibit superior strength, modulus, heat distortion temperature, water and gas barrier properties, with comparable impact strength as neat nylon 6. They also reported on various other types of polymer–clay hybrid nanocomposites based on epoxy resin and polystyrene, acrylic polymer, rubber, and polyimides formed using a similar approach. On the other hand, work by Giannelis and co-workers [29,30] revealed that intercalation of polymer chains into the galleries of an organoclay can occur spontaneously on heating a mixture of polymer and silicate clay powder above the polymer glass transition or melt temperature. Once sufficient polymer mobility is achieved, chains diffuse into the host silicate clay galleries, thereby producing an expanded polymer–silicate structure. To summarize: although the intercalation chemistry of polymers when mixed with appropriately modified layered silicates and synthetic layered silicates has long been known, two major findings have stimulated the revival of interest in polymer–layered silicate nanocomposite materials. First, the report from the Toyota research group on a nylon 6/montmorillonite nanocomposite, in which very small amounts of layered silicate loadings resulted in pronounced improvements of thermal and mechanical properties; and second, the observation by Giannelis and his co-workers that it is possible to melt-mix polymers with layered silicates, without the use of organic solvents. Since then, the high promise for industrial applications has motivated vigorous research, and today efforts are being conducted globally, using almost all types of polymer matrices. In fact, nanocomposites have been demonstrated with many thermoplastic and thermosetting polymers of different polarities including, among others, polystyrene, polycaprolactone, polypropylene, poly(ethylene oxide), epoxy resin, polysiloxane and polyurethane [7,14,15,31–33].

It must be noted, however, that so far most of these materials have been produced only on the laboratory scale; and until a short time ago, research tended to center on proof of exfoliation of the clay [32]. 3. Layered silicates 3.1. Structure and characteristics of layered silicates Layered silicates used in the synthesis of nanocomposites are natural or synthetic minerals, consisting of very thin layers that are usually bound together with counter-ions. Their basic building blocks are tetrahedral sheets in which silicon is surrounded by four oxygen atoms, and octahedral sheets in which a metal like aluminum is surrounded by eight oxygen atoms. Therefore, in 1:1 layered structures (e.g. in kaolinite) a tetrahedral sheet is fused with an octahedral sheet, whereby the oxygen atoms are shared [34]. On the other hand, the crystal lattice of 2:1 layered silicates (or 2:1 phyllosilicates), consists of two-dimensional layers where a central octahedral sheet of alumina is fused to two external silica tetrahedra by the tip, so that the oxygen ions of the octahedral sheet also belong to the tetrahedral sheets, as shown in Fig. 1. The layer thickness is around 1 nm and the lateral dimensions may vary from 300 Å to several microns, and even larger, depending on the particulate silicate, the source of the clay and the method of preparation (e.g. clays prepared by milling typically have lateral platelet dimensions of approximately 0.1–1.0 ␮m). Therefore, the aspect ratio of these layers (ratio length/thickness) is particularly high, with values greater than 1000 [1,35–37]. The basic 2:1 structure with silicon in the tetrahedral sheets and aluminum in the octahedral sheet, without any substitution of atoms, is called pyrophyllite. Since the layers do not expand in water, pyrophyllite has only an external surface area and essentially no internal one. When silicon in the tetrahedral sheet is substituted by aluminum, the resulting structure is called mica. Due to this substitution the mineral is characterized by a negative surface charge, which is balanced by interlayer potassium cations. However, because the size of the potassium ions matches the hexagonal hole created by the Si/Al tetrahedral layer, it is able to fit very tightly between the layers. Consequently, the interlayers collapse and the layers are held together by the electrostatic attraction between the negatively charged tetrahedral layer and the potassium cations. Therefore, micas do not swell in water and, like pyrophyllite, have no internal surface area [38]. On the other hand, if in the original pyrophyllite structure the trivalent Al-cation in the octahedral layer is partially substituted by the divalent Mg-cation, the structure of montmorillonite is formed, which is the best-known member of a group of clay minerals, called “smectites” or “smectite clays”. In this case the overall negative charge is balanced by sodium and calcium ions, which exist hydrated in the interlayer [39]. A particular feature of the resulting structure is that, since these ions do not fit in the tetrahedral layer, as in mica, and the layers are held together by relatively weak forces, water and other polar molecules can enter between the unit layers, causing the lattice to expand [33].

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1123

Fig. 1. The structure of a 2:1 layered silicate [35]. Reproduced from Beyer by permission of Elsevier Science Ltd., UK.

Along with montmorillonite, hectorite and saponite are the layered silicates that are most commonly used in nanocomposite materials. Their chemical formula is given in Table 1 [1]. The reason why these materials have received a great deal of attention recently, as reinforcing materials for polymers, is their potentially high aspect ratio and the unique intercalation/exfoliation characteristics that will be discussed later [15]. In general, it is well established that structural perfection is more and more nearly reached as the reinforcing elements become smaller and that the ultimate properties of reinforcing composite elements may be expected if their dimensions reach atomic or molecular levels. For example, carbon nanotubes display the so far highest known values of elastic modulus (ca. 1.7 TPa!). Similarly, individual clay sheets, being only 1 nm thick, display a perfect crystalline structure. However, the smaller the reinforcing elements are, the larger is their internal surface and hence their tendency to agglomerate rather than to disTable 1 Chemical structure of commonly used 2:1 phyllosilicatesa [1]. 2:1 Phyllosilicates

General formula

Montmorillonite Hectorite Saponite

Mx (Al4−x Mgx )Si8 O20 (OH)4 Mx (Mg6−x Lix )Si8 O20 (OH)4 Mx Mg6 (Si8−x Alx )O20 (OH)4

a M: monovalent cation; x: degree of isomorphous substitution (between 0.5 and 1.3).

perse homogeneously in a matrix [2]. In fact, the silicate layers have the tendency to organize themselves to form stacks with a regular van der Waals gap between them, called an “interlayer” or “gallery” [1,35,36]. The interlayer dimension is determined by the crystal structure of the silicate (for dehydrated Na–montmorillonite this dimension is approximately 1 nm) [37]. Analysis of layered silicates has shown that there are several levels of organization within the clay minerals. The smallest particles, primary particles, are on the order of 10 nm and are composed of stacks of parallel lamellae. Micro-aggregates are formed by lateral joining of several primary particles, and aggregates are composed of several primary particles and micro-aggregates [40]. 3.2. Organic modification of layered silicates Since, in their pristine state layered silicates are only miscible with hydrophilic polymers, such as poly(ethylene oxide) and poly(vinyl alcohol), in order to render them miscible with other polymers, one must exchange the alkali counter-ions with a cationic-organic surfactant, as shown in Fig. 2. Alkylammonium ions are mostly used, although other “onium” salts can be used, such as sulfonium and phosphonium [1,39,41]. This can be readily achieved through ion-exchange reactions that render the clay organophilic [42]. In order to obtain the exchange of the onium ions with the cations in the galleries, water

1124

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 2. Schematic picture of an ion-exchange reaction. The inorganic, relatively small (sodium) ions are exchanged against more voluminous organic onium cations. This ion-exchange reaction has two consequences: firstly, the gap between the single sheets is widened, enabling polymer chains to move in between them and secondly, the surface properties of each single sheet are changed from being hydrophilic to hydrophobic [2]. Reproduced from Fischer by permission of Elsevier Science Ltd., UK.

swelling of the silicate is needed. For this reason alkali cations are preferred in the galleries because 2-valent and higher valent cations prevent swelling by water. Indeed, the hydrate formation of monovalent intergallery cations is the driving force for water swelling. Natural clays may contain divalent cations such as calcium and require exchange procedures with sodium prior to further treatment with onium salts [41]. The alkali cations, as they are not structural, can be easily replaced by other positively charged atoms or molecules, and thus are called exchangeable cations [43]. The organic cations lower the surface energy of the silicate surface and improve wetting with the polymer matrix [4,42]. Moreover, the long organic chains of such surfactants, with positively charged ends, are tethered to the surface of the negatively charged silicate layers, resulting in an increase of the gallery height [44]. It then becomes possible for organic species (i.e. polymers or prepolymers) to diffuse between the layers and eventually separate them [42,45]. Sometimes, the alkylammonium cations may even provide functional groups that can react with the polymer or initiate polymerization of monomers [33]. The microchemical environment in the galleries is, therefore, appropriate to the intercalation of polymer molecules [46]. Conclusively, the surface modification both increases the basal spacing of clays and serves as a compatibilizer between the hydrophilic clay and the hydrophobic polymer [45]. The excess negative charge of layered silicates and their capability to exchange ions can be quantified by a specific property known as the cation-exchange capacity (CEC) and expressed in mequiv./g [1,39]. This property is highly dependent on the nature of the isomorphous substitutions in the tetrahedral and octahedral layers and therefore on the nature of the soil where the clay was formed. This explains, for example, why montmorillonites from different origins show differences in CEC, ranging from approximately 0.9–1.2 mequiv./g [39,42]. The charge of the layer is not locally constant, as it varies from layer to layer, and must rather be considered as an average value over the whole crystal. Proportionally, even if a small part of the charge

balancing cations is located on the external crystallite surface, the majority of these exchangeable cations are located inside the galleries [1]. Depending on the functionality, packing density, and length of the organic modifiers, the organo-modified layered silicates (OMLSs, organosilicates or organoclays) may be engineered to optimize their compatibility with a given polymer [43,47]. It is worth noticing that, on the basis of the CEC of the clay, the content of the surfactant is usually about 35–45 wt.% [48]. Actually, one way to measure the clay CEC is by determining the amount of alkylammonium salt retained by the organoclays. That is, dried clays that have been subjected to organo-modification along with a sample of the corresponding untreated clay are ignited at 1000 ◦ C. From the differences in the loss on ignition of the sample and blank and the molecular weight of the alkylammonium salt, the milliequivalents of the organic substance retained by the clays are calculated and those values are taken as their CEC. Alternatively, chemical analysis of the clay can also be applied for CEC determination [42]. In general, the longer the surfactant chain length, and the higher the charge density of the clay, the further apart the clay layers will be forced. This is expected since both of these parameters contribute to increasing the volume occupied by the intragallery surfactant [7]. For example, Wang et al. prepared organoclays with different alkylammonium chain lengths and also used an organophilic clay, Cloisite 20A, which has two long alkyl chains. They found that the interlayer spacing increases with the increase in size of alkylamine chain length. The interlayer spacings of C12 M, C16 M, C18 M (with 12, 16 and 18 carbon atoms in the alkylammonium chain) and 20 A were 1.36, 1.79, 1.85 and 2.47 nm, respectively [49]. However, the interlayer distance also depends on the way the onium ion chains organize themselves in the organoclay. In order to describe the structure of the interlayer in organoclays, one has to know that, as the negative charge originates in the silicate layer, the cationic head group of the alkylammonium molecule preferentially resides at the layer surface, leaving the organic tail radiating away from the surface [1,41].

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1125

Fig. 3. Alkyl chain aggregation in layered silicates: (a) lateral monolayer; (b) lateral bilayer; (c) paraffin-type monolayer and (d) paraffin-type bilayer [1]. Reproduced from Alexandre and Dubois by permission of Elsevier Science Ltd., UK.

Initially, the orientation of surfactant chains was deduced from infrared and XRD measurements, according to which the organic chains have been long thought to lie either parallel to the silicate layer, forming mono or bilayers or, depending on the packing density and the chain length, to radiate away from the surface, forming mono or even a bimolecular tilted “paraffinic” arrangement, as shown in Fig. 3 [1,7]. A more realistic description has been proposed by Vaia et al., based on FTIR experiments. By monitoring frequency shifts of the asymmetric CH2 stretching and bending vibrations, they showed that alkyl chains can vary from liquid-like to solid-like, with the liquid-like structure dominating as the interlayer density or chain length decreases, or as the temperature increases. When the available surface area per molecule is within a certain range, the chains are not completely disordered but retain some orientational order similar to that in the liquid crystalline state (Fig. 4). As the chain length increases, the interlayer structure appears to evolve in a stepwise fashion, from a disordered to more ordered monolayer, then “jumping” to a more disordered pseudo-bilayer. In addition, an NMR study reported by Wang et al. indicated the coexistence of ordered trans and disordered gauche conformations [50]. Fornes et al. conducted WAXS scans for different organoclays and for pristine sodium montmorillonite, and plotted basal spacing values obtained vs. the mass of

Fig. 5. WAXS results for organoclays: mass of organic per unit mass of montmorillonite [51]. Reproduced from Fornes, Yoon, Hunter, Keskkula and Paul by permission of Elsevier Science Ltd., UK.

organic component per unit mass of inorganic MMT for each organoclay, as shown in Fig. 5. They further analyzed the data by expressing the mass of organic material per unit volume of gallery, or gallery density, as gallery =

mass organic 1 mass organic/mass MMT = gallery volume d − d0 (area/side)/mass MMT

Fig. 4. Alkyl chain aggregation models: (a) short alkyl chains: isolated molecules, lateral monolayer; (b) intermediate chain lengths: in-plane disorder and interdigitation to form quasi bilayers and (c) longer chain length: increased interlayer order, liquid crystalline-type environment [1]. Reproduced from Alexandre and Dubois by permission of Elsevier Science Ltd., UK.

1126

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

where d − d0 is the gallery height, as indicated by Fig. 5. The slope of the linear relation between gallery height and the second quantity in the above equation gives the density of the organic material in the gallery, gallery . Forcing the fit through an intercept d0 = 9.6 Å, which is the basal spacing of pristine sodium montmorillonite, produces a calculated density of 1.07 g/cm3 . This range of densities encompasses what might be expected for organic liquids or solids. Therefore, the densities calculated in that work agree with conclusions made by Vaia et al. about the conformation and structure of the organic interlayer of similar organoclays, suggesting that organoclay galleries exhibit molecular environments ranging from solid- to liquid-like [51]. Summarizing this section, there are two particular characteristics of layered silicates that are exploited in polymer–layered silicate nanocomposites. The first is the ability of the silicate particles to disperse into individual layers. Since dispersing a layered silicate can be pictured like opening a book, an aspect ratio as high as 1000 for fully dispersed individual layers can be obtained (contrast that to an aspect ratio of about 10 for undispersed or poorly dispersed particles). The second characteristic is the ability to fine-tune their surface chemistry through ion exchange reactions with organic and inorganic cations. These two characteristics are, of course, interrelated since the degree of dispersion in a given matrix that, in turn, determines aspect ratio, depends on the interlayer cation [4,40]. 4. Nanocomposite structures and characterization 4.1. Nanocomposite structures Any physical mixture of a polymer and silicate (or inorganic material in general) does not necessarily form a nanocomposite. The situation is analogous to polymer blends. In most cases, separation into discrete phases normally takes place. In immiscible systems, the poor physical attraction between the organic and the inorganic components leads to relatively poor mechanical properties. Furthermore, particle agglomeration tends to reduce strength and produce weaker materials [4]. Thus, when the polymer is unable to intercalate between the silicate sheets, a phase-separated composite is obtained, whose properties are in the same range as for traditional microcomposites [1,35]. Beyond this traditional class of polymer-filler composites, two types of nanocomposites can be obtained, depending on the preparation method and the nature of the components used, including polymer matrix, layered silicate and organic cation [1,35]. These two types of polymer–layered silicate nanocomposites are depicted in Fig. 6 [50]. Intercalated structures are formed when a single (or sometimes more) extended polymer chain is intercalated between the silicate layers. The result is a well ordered multilayer structure of alternating polymeric and inorganic layers, with a repeat distance between them. Intercalation causes less than 20–30 A´˚ separation between the platelets [1,4,33,35,44,52].

On the other hand, exfoliated or delaminated structures are obtained when the clay layers are well separated from one another and individually dispersed in the continuous polymer matrix [1,4,33,44]. In this case, the polymer separates the clay platelets by 80–100 A´˚ or more [52]. That is, the interlayer expansion is comparable to the radius of gyration of the polymer rather than that of an extended chain, as in the case of intercalated hybrids [4]. The exfoliation or delamination configuration is of particular interest because it maximizes the polymer–clay interactions making the entire surface of layers available for the polymer. This should lead to the most significant changes in mechanical and physical properties [35]. In fact, it is generally accepted that exfoliated systems give better mechanical properties than intercalated ones [5,33]. The complete dispersion of clay nanolayers in a polymer optimizes the number of available reinforcing elements for carrying an applied load and deflecting cracks. The coupling between the tremendous surface area of the clay and the polymer matrix facilitates stress transfer to the reinforcement phase, allowing for mechanical property improvements [35,53]. However, it is not easy to achieve complete exfoliation of clays and, indeed with few exceptions, the majority of the polymer nanocomposites reported in the literature were found to have intercalated or mixed intercalatedexfoliated nanostructures [33]. This is because the silicate layers are highly anisotropic, with lateral dimensions ranging from 100 to 1000 nm, and even when separated by large distances (i.e. when delaminated) cannot be placed completely randomly in the sea of polymer. Furthermore, the majority of the polymer chains in the hybrids are tethered to the surface of the silicate layers. Thus, it can be expected that there are domains in these materials, even above the melting temperature of the constituent polymers, wherein some long-range order is preserved and the silicate layers are oriented in some preferred direction. This long-range order and domain structure is likely to become better defined at the higher silicate contents, where the geometrically imposed mean distance between the layers becomes less than the lateral dimensions of the silicate layers, thus forcing some preferential orientation between the layers. However, there might be considerable polydispersity effects in terms of the orientation and the distance between the silicate layers. Many such randomly oriented grains make up the entire sample leading to the presence of disordered material. Thus, in general the material possesses a layered structure, with grains wherein the silicate layers are oriented in a preferred direction leading to the presence of grain boundaries and concomitant defects [54]. At this point, it is worth mentioning six interrelated structural characteristics, distinguishing polymer–silicate nanocomposites from conventional filled systems. These characteristics, which are attributed to the nanoscopic dimensions and the extreme aspect ratios of layered silicates, are [8]: • Low percolation threshold (ca. 0,1–2 vol.%). • Particle–particle correlation (orientation and position) arising at low volume fractions.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1127

Fig. 6. Schematic illustration of two different types of thermodynamically achievable polymer/layered silicate nanocomposites [50]. Reproduced from Ray and Bousima by permission of Elsevier Science Ltd., UK.

• Large number of particles per particle volume (106 –108 particles/␮m3 ). • Extensive interfacial area per volume of particles (103 –104 m2 /ml). • Short distances between particles (10–50 nm at  ∼ 1–8 vol.%). • Comparable size scales among the rigid nanoparticle inclusions, distance between particles, and the relaxation volume of polymer chains. 4.2. Nanocomposite structural characterization Two complementary techniques are generally used to characterize the structures of nanocomposites: XRD and transmission electron microscopy (TEM) [1,35,55,56]. Due to its ease of use and availability, XRD is most commonly used to probe the nanocomposite structure and occasionally to study the kinetics of polymer melt intercalation [55]. This technique allows the determination of the spaces between structural layers of the silicate utilizing Bragg’s law: sin  = n/2d, where  corresponds to the wave length of the X-ray radiation used in the diffraction experiment, d the spacing between diffractional lattice planes and  is the measured diffraction angle or glancing angle [1,56]. By monitoring the position, shape and intensity of the basal reflections from the distributed silicate layers, the nanocomposite structure may be identified [55]. For immiscible polymer/OMLS mixtures, the structure of the silicate is not affected, and thus, the characteristics of the OMLS basal reflections do not change. On the other hand, in comparison with the spacing of the organoclay used, the intercalation of the polymer chains increases the interlayer spacing, leading to a shift of the diffraction peak towards lower angle, according to Bragg’s law. In such intercalated nanocomposites, the repetitive multilayer structure is well preserved, allowing the interlayer spacing to be determined (Fig. 7). In contrast, the extensive layer separation associated with exfoliated structures disrupts the coherent layer stacking and results in a featureless diffraction pattern. Thus, for exfoliated structures no more diffraction peaks are visible in the XRD diffractograms either because of a much too large spacing between the layers (i.e. exceeding 8 nm in the case of ordered exfoliated structure) or because the nanocomposite does not present ordering [1,35,47].

The influence of polymer intercalation on the order of the OMLS layers may be monitored by changes in the fullwidth-at-half-maximum (fwhm) and intensity of the basal reflections. An increase in the degree of coherent layer stacking (i.e. a more ordered system) results in a relative decrease in the fwhm of the basal reflections upon hybrid formation. On the other hand, a decrease in the degree of coherent layer stacking (i.e. a more disordered system) results in peak broadening and intensity loss [47]. However, although XRD offers a conventional method to determine the interlayer spacing of the silicate layers in the original layered silicates and the intercalated nanocomposites (within 1–4 nm), little can be said about the spatial distribution of the silicate layers or any structural inhomogeneities in nanocomposites. Additionally, some layered silicates initially do not exhibit well-defined basal reflections. Thus, peak broadening and intensity decreases are very difficult to study systematically. Therefore, conclusions concerning the mechanism of nanocomposite

Fig. 7. Typical XRD patterns from polymer/layered silicates: (a) PE + organoclay → no formation of a nanocomposite, (b) PS + organoclay → intercalated nanocomposite, (c) siloxane + organoclay → delaminated nanocomposite [35]. Reproduced from Beyer by permission of Elsevier Science Ltd., UK.

1128

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

formation and structure based solely on XRD patterns are only tentative. On the other hand, TEM allows a qualitative understanding of the internal structure and can directly provide information in real space, in a localized area, on morphology and defect structures [57,58]. Since the silicate layers are composed of heavier elements (Al, Si and O) than the interlayer and surrounding matrix (C, H and N), they appear darker in bright-field images. Therefore, when nanocomposites are formed, the intersections of the silicate sheets are seen as dark lines which are the cross sections of the silicate layers, measuring 1 nm thick. However, special care must be exercised to guarantee a representative cross-section of the sample [56,57]. Fig. 8 shows the TEM micrographs obtained for an intercalated and an exfoliated nanocomposite. As already mentioned, besides these two well defined structures other intermediate organizations can exist presenting both intercalation and exfoliation. In this case, a broadening of the diffraction peak is often observed and one must rely on TEM observation to define the overall structure [1]. Characterization of polymer/layered silicate nanocomposites by 13 C solid-state nuclear magnetic resonance (13 C NMR) has also been proposed. VanderHart et al. first used this technique as a tool for gaining greater insight about the morphology, surface chemistry, and to a very limited extent, the dynamics of exfoliated polymer clay nanocomposites. They were especially interested in developing NMR methods to quantify the level of clay exfoliation, a very important facet of nanocomposite characterization [59]. The main objective in solid-state NMR measurement is to connect the measured longitudinal relaxations, T1 s, of proton (and 13 C nuclei) with the quality of clay dispersion [55]. The surfaces of naturally occurring layered silicates such as MMT are mainly made of tetrahedral silica, while the

central plane of the layers contains octahedrally coordinated Al3+ with frequent non-stoichiometric substitutions, where an Al3+ is replaced by Mg2+ and, somewhat less frequently, by Fe3+ . The concentration of the later ion is very important because Fe3+ is strongly paramagnetic in this distorted octahedral environment. Typical concentrations of Fe3+ in naturally occurring clays produce nearest-neighbor Fe–Fe distances of about 1.0–1.4 nm, and at such distances, the spin-exchange interaction between the unpaired electrons on different Fe atoms is expected to produce magnetic fluctuations in the vicinity of the Larmor frequencies for protons or 13 C nuclei. The spectral density of these fluctuations is important because the T1 H of protons (and 13 C nuclei) within about 1 nm of the clay surface can be directly shortened. For protons, if that mechanism is efficient, relaxation will also propagate into the bulk of the polymer by spin diffusion. Thus the paramagnetically induced relaxation will influence the overall measured T1 H to an extent that will depend both on the Fe concentration in the clay layer and, more importantly, on the average distances between clay layers. The latter dependence suggests a potential relationship between measured T1 H values and the quality of the clay dispersion. If the clay particles are stacked and poorly dispersed in the polymer matrix, the average distances between polymer/clay interfaces are greater, and the average paramagnetic contribution to T1 H is weaker. VanderHart et al. also employed the same arguments in order to understand the stability of a particular OMLS under different processing conditions [55,60,61]. Some authors also used Fourier transform infrared spectroscopy (FTIR) to elucidate the structure of the nanocomposites [62,63]. FTIR may be able to identify differences between the bonding in a mixture and the bonding in a related nanocomposite, but as these variations are

Fig. 8. TEM micrographs of poly(styrene)-based nanocomposites: (a) intercalated nanocomposite and (b) exfoliated nanocomposite [1]. Reproduced from Alexandre and Dubois by permission of Elsevier Science Ltd., UK.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 9. DSC traces of pure PS, a physical mixture of PS/OLS, and PSintercalated OLS [41]. Reproduced from Zanetti, Lomakin and Camino by permission of Wiley-VCH, Germany.

minute, even when intercalation has taken place, at present FTIR is an unreliable method of characterization in most cases [56]. Finally, differential scanning calorimetry (DSC) provides further information concerning intercalation. The many interactions the intercalated chains of the polymer form with the host species greatly reduce their rotational and translational mobility. The situation is similar to that in a reticulated polymer, where restrictions on its mobility increase its glass transition temperature (Tg ). A similar increase is anticipated to occur in a nanocomposite due to elevation of the energy threshold needed for the transition. This effect is readily detected by DSC. Fig. 9 presents DSC traces of polystyrene (PS), a PS/OMLS mixture and an intercalated PS/OMLS nanocomposite. The PS and PS/OMLS mixture curves clearly display the characteristic peak due to glass transition of the polymer. The presence of this peak in the mixture is evidence of the absence of interactions between the organic and the inorganic phases. The transition is absent in the nanocomposite curve and in fact occurs at temperatures higher than those shown in Fig. 9. In addition to being an interesting analytical datum, the considerable increase in Tg is an important property of these materials that enables them to be employed at higher temperatures compared with the original polymer and thus extends their fields of application [41]. To date, the aforementioned subsidiary methods have only been used to confirm the evidence from the primary methods. However, building a clearer picture of the changes that occur when a nanocomposite forms is important, as it not only helps to characterize the material, but in principle could indicate novel methods of synthesis [56]. Concerning the evaluation of other nanocomposites structural characteristics, it should be noted that the amount of clay present in a sample may be estimated, as for conventional composites, i.e. by placing pre-dried nanocomposite pellets in a furnace at ca. 900 ◦ C for approx-

1129

imately 45 min. The resulting ash is then weighed and corrected for loss of structural water [64]. However, unlike conventional fiber composites, the determination of filler aspect ratio for layered silicate nanocomposites is not straightforward. Good estimates require a thorough analysis of TEM photomicrographs at different magnifications. Fig. 10 depicts various complications of calculating an aspect ratio from TEM photomicrographs that arise from variations in both length/diameter, and thickness. Clay platelets intrinsically have a distribution of lateral dimensions. The recovery, refinement, chemical treatment, and post-treatment of these clays may contribute to the variation in filler geometry. Furthermore, extrusion of these clays with polymer and any additional melt processing steps that follow, e.g. injection molding, will amplify the range of particle shapes and sizes, particularly when the layered silicate is not completely exfoliated, as illustrated in Fig. 10. Finally, microtoming of the nanocomposite sample into thin sections for TEM analysis will also result in an apparent distribution of observed particle sizes even if all disk-like platelets were the same size [64]. It becomes obvious from this section that a major issue when synthesizing polymer–layered silicate nanocomposites is the characterization of the product. In fact, many of the studies conducted so far in this field are solely dedicated to structural characterization of the nanocomposites, without reporting properties of the products. 5. Preparation of nanocomposites 5.1. Introduction At present there are four principal methods for producing polymer–layered silicate nanocomposites: (1) in situ template synthesis, (2) intercalation of polymer or prepolymer from solution, (3) in situ intercalative polymerization and (4) melt intercalation [1,14,35,44,52,65]. 5.1.1. Template synthesis (sol–gel technology) In this technique, the clay minerals are synthesized within the polymer matrix, using an aqueous solution (or gel) containing the polymer and the silicate building blocks. As precursors for the clay silica sol, magnesium hydroxide sol and lithium fluoride are used. During the process, the polymer aids the nucleation and growth of the inorganic host crystals and gets trapped within the layers as they grow. Although theoretically this method has the potential of promoting the dispersion of the silicate layers in a one-step process, without needing the presence of the onium ion, it presents serious disadvantages. First of all, the synthesis of clay minerals generally requires high temperatures, which decompose the polymers. An exception is the synthesis of hectorite-type clay minerals which can be performed under relatively mild conditions. Another problem is the aggregation tendency of the growing silicate layers. Therefore, this technique, although widely used for the synthesis of double-layer hydroxide-based nanocomposites, is far less developed for layered silicates and will not be considered in the following discussion. However, it should be mentioned that several workers have suc-

1130

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 10. Examples of complications in the determination of the aspect ratio of layered silicate fillers within polymer nanocomposites [64]. Reproduced from Fornes and Paul by permission of Elsevier Science Ltd., UK.

cessfully applied it for the preparation of nanocomposite materials. For example, Carrado et al. and Carrado and Xu synthesized hectorites from gels consisting of silica, magnesium hydroxide, lithium fluoride and polymers like poly(vinyl alcohol), polyaniline and polyacrylonitrile. Evidently, some silicate layers aggregated, but most of them remained uniformly distributed in the polymer matrix [1,3,41] 5.1.2. Intercalation of polymer or prepolymer from solution Following this technique, the layered silicate is exfoliated into single layers using a solvent in which the polymer (or prepolymer in case of insoluble polymers, such as polyimide) is soluble. It is well known that such layered silicates, owing to the weak forces that stack the layers together can be easily dispersed in an adequate solvent. After the organoclay has swollen in the solvent, the polymer is added to the solution and intercalates between the clay layers. The final step consists of removing the solvent, either by vaporization, usually under vacuum, or by precipitation. Upon solvent removal the sheets reassemble, sandwiching the polymer to form a nanocomposite structure. Under this process are also gathered the nanocomposites obtained through emulsion polymerization where the layered silicate is dispersed in the aqueous phase. The major advantage of this method is that intercalated nanocomposites can be synthesized that are based on polymers with low or even no polarity. However, the solvent approach is difficult to apply in industry owing to problems associated with the use of large quantities of solvents [1,35]. 5.1.3. In situ intercalative polymerization In situ-polymerization was the first method used to synthesize polymer–clay nanocomposites based on polyamide 6. In this technique, the modified layered silicate is swollen by a liquid monomer or a monomer solution. The monomer migrates into the galleries of the layered silicate, so that

the polymerization reaction can occur between the intercalated sheets. The reaction can be initiated either by heat or radiation, by the diffusion of a suitable initiator or by an organic initiator or catalyst fixed through cationic exchange inside the interlayer before the swelling step by the monomer. Polymerization produces long-chain polymers within the clay galleries. Under conditions in which intra- and extra-gallery polymerization rates are properly balanced, the clay layers are delaminated and the resulting material possesses a disordered structure [1,35,37]. 5.1.4. Melt intercalation This technique consists of blending the layered silicate with the polymer matrix in the molten state. Under such conditions – if the layer surfaces are sufficiently compatible with the chosen polymer – the polymer can crawl into the interlayer space and form either an intercalated or an exfoliated nanocomposite [1,35,37]. Among the aforementioned methods, in situ polymerization and melt intercalation are considered as commercially attractive approaches for preparing polymer/clay nanocomposites. Melt intercalation, in particular, is especially of practical interest, since it presents significant advantages that will be discussed in the corresponding paragraph. 5.2. Intercalation of polymer from solution Intercalation of a polymer from a solution is a two-stage process in which the polymer replaces an appropriate, previously intercalated solvent, as shown in Fig. 11. Such a replacement requires a negative variation in the Gibbs free energy. It is thought that the diminished entropy due to the confinement of the polymer is compensated by an increase due to desorption of intercalated solvent molecules. In other words, the entropy gained by desorption of solvent molecules is the driving force for polymer intercalation from solution [66–71].

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1131

Fig. 11. Schematic representation of PLS obtained by intercalation from solution.

Even though this technique has been mostly used with water soluble polymers, such as PEO, PVE, PVP and PAA [3,21,72–77], intercalation from non-aqueous solutions has also been reported [78–81]. For example, HDPE-based nanocomposites have been prepared by dissolving HDPE in a mixture of xylene and benzonitrile with dispersed OMLS. The nanocomposite was then recovered by precipitation from THF [79]. PS/OMLS exfoliated nanocomposites have also been prepared by the solution intercalation technique, by mixing pure PS and organophilic clay with adsorbed cetyl pyridium chloride [82]. Similarly, several studies have focused on the preparation of PLA-layered silicate nanocomposites using intercalation from solution. The first attempts by Ogata [78], involved dissolving the polymer in hot chloroform in the presence of organomodified MMT. However, TEM and WAXD analyses revealed that only microcomposites were formed and that an intercalated morphology was not achieved. In a later study, Krikorian and Pochan [83] prepared PLA nanocomposites using dichloromethane as the polymer solvent and as the OMLS dispersion medium. The authors obtained intercalated or exfoliated nanocomposites, depending on the type of OMLS used. That is, exfoliated nanocomposites were formed when diols were present in the organic modifier of the clay, due to the favorable enthalpic interaction between these diols and the C O bonds in the PLA backbone. Chang et al. [84] reported the preparation of PLA-based nanocomposites with different kinds of OMLS via solution intercalation using N,N -dimethylacetamide (DMA). In the case of polymeric materials that are infusible and insoluble even in organic solvents, the only possible route to produce nanocomposites with this method is to use polymeric precursors that can be intercalated in the layered silicate and then thermally or chemically converted to the desired polymer [1,85]. Summarizing the above: although a number of nanocomposites have been produced by intercalation from solution (representative examples are presented in Table 2), it is important to note that, in using this method, intercalation only occurs for certain polymer/clay/solvent systems, meaning that for a given polymer one has to find the right clay, organic modifier and solvents [1,50]. Moreover, from

the industrial point of view, this method may involve the copious use of organic solvents, which is usually environmentally unfriendly and economically prohibitive [50]. 5.3. In situ intercalative polymerization 5.3.1. In situ intercalative polymerization of thermoplastic polymers The Toyota research group first reported the ability of ␣,␻-amino acid (COOH–(CH2 )n−1 –NH2 + , with n = 2, 3, 4, 5, 6, 8, 11, 12, 18) modified Na+ -MMT to be swollen by ␧-caprolactam monomer at 100 ◦ C and subsequently initiate ring opening polymerization to obtain PA6/MMT nanocomposites [25]. The number of carbon atoms in the ␣,␻-amino acid was found to have a strong effect on the swelling behavior as reported in Fig. 12, indicating that the extent of intercalation of ␧-caprolactam monomer is high when the number of carbon atoms in the ␻-amino acid is large [93]. Moreover, it was found from a comparison of different types of inorganic silicates that clays having higher CEC lead to more efficient exfoliation of the silicate platelets [51]. Fig. 13 represents the conceptual view of the swelling behavior of ␣,␻-amino acid modified Na+ -MMT by ␧-caprolactam [93]. In a typical synthesis, 12-aminolauric acid-modified MMT (12-MMT) was Table 2 PLS nanocomposites prepared by intercalation from solution. Nanocomposite +

PVOH/Na -MMT PVA/Na+ -MMT TPU/OMLS PEO/Na+ -MMT or Na+ -hectorite PEO/MMT PLA/OMLS PLA/OMLS HDPE/protonated dodecylamine modified MMT PSF/OMLS PI/dodecylammonium modified MMT

Solvent(s)

Ref.

Water Water Toluene/DMAc Acetonitrile

[86] [87] [88] [89]

Chloroform Dichloromethane DMAc Xylene/benzonitrile (80/20 wt.%)

[90] [83] [84] [91]

DMAC DMAC

[92] [85]

1132

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198 Table 3 Peak intensity (Im ) and interlayer spacing (d) of nylon-6-based nanocomposites prepared in presence of different acid derivatives by the one-pot technique [95].

Fig. 12. XRD patterns of ␻-amino acid [NH2 (CH2 )n−1 COOH] modified Na+ MMT [93]. Reproduced from Usuki, Kawasumi, Kojima, Okada, Kurauchi and Kamigaito by permission of Materials Research Society, USA.

mixed with ␧-caprolactam and the mixture was heated at 250–270 ◦ C for 48 h to polymerize ␧-caprolactam, using 12MMT as a catalyst (when the relative amount of 12-MMT in the mixture was less than 8 wt.%, 6-aminocaproic acid was added as a polymerization accelerator and the heating profile was slightly modified). Depending on the amount of 12-MMT introduced, either exfoliated (for less then 15 wt.%) or intercalated structures (from 15 to 70 wt.%) were obtained, as evidenced by XRD and TEM. Comparison of the titrated amounts of COOH and NH2 end groups present in the synthesized nanocomposites with given values, such as the CEC of the montmorillonite used (119 mequiv./100 g), have led to the conclusion that the COOH end groups present along the 12-MMT surface are responsible for the polymerization initiation [25]. Further work demonstrated that intercalative polymerization of ␧-caprolactam could be realized without

Fig. 13. Swelling behavior of ␻-amino acid modified MMT by ␧caprolactam [93]. Reproduced from Usuki, Kawasumi, Kojima, Okada, Kurauchi and Kamigaito by permission of Materials Research Society, USA.

Acid

Im (cps)

d (Å)

Phosphoric acid Hydrochloric acid Isophtalic acid Benzenesulfonic acid Acetic acid Trichloroacetic acid No acid

0 200 255 280 555 585 1840

0 21.7 20.2 19.3 20.3 21.3 18.6

modifying the MMT surface. Indeed, this monomer was able to directly intercalate the Na+ -MMT in water in the presence of hydrochloric acid, as proved by the increase in interlayer spacing from 10 to 15.1 Å. At high temperature (200 ◦ C), in the presence of excess ␧-caprolactam, the clay so modified can be swollen again, allowing the ring opening polymerization to proceed at 260 ◦ C when 6-aminocaproic acid is added as an accelerator. The resulting composite does not present a diffraction peak in XRD, and TEM observation agrees with a molecular dispersion of the silicate sheets [94]. In attempts to carry out the whole synthesis in one pot, the system proved to be sensitive to the nature of the acid used to promote the intercalation of ␧-caprolactam. Table 3 gives results obtained for different acids in relation to the intensity (Im ) of the XRD intercalation peak that might be present in the nanocomposites obtained (Fig. 14). These results show that, for unclear reasons, only phosphoric acid allows for the preparation of a truly exfoliated nanocomposite, the other acids tending to promote the formation of partially exfoliated-partially intercalated structures. One can also point out that an intercalated structure is obtained even if no acid is added [95]. Another polyamide, nylon 12, has also been reported to form nanocomposites via in situ intercalative polymerization. Reichert et al. [96] used 12-aminolauric acid (ALA) as both the layered silicate modifier and the monomer. They first studied by XRD the dependence of the clay swelling process on ALA concentration in HCl, and found that it can be separated in two regimes: a cation-exchange of inorganic cations by protonated ALA at low ALA concentra-

Fig. 14. XRD intensity curve of injection molded nylon-6 nanocomposite as obtained by the one-pot intercalation polymerization process in the presence of acetic acid [95]. Reproduced from Kojima, Usuki, Kawasumi, Okada, Kurauchi and Kamigaito by permission of John Wiley & Sons, Inc.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 15. Interlayer distance of fluoro-modified talc (ME 100) in function of an increasing amount of aminolauric acid used as the organic modifier [96]. Reproduced from Reichert, Kressler, Thomann, Mulhaupt and Stoppelmann by permission of Wiley-VCH, Germany.

tion and a further diffusion of zwitterionic 12-aminolauric acid into the interlayer space, when the ALA concentration exceeds the amount of HCl in the medium (Figs. 15 and 16). The swelling was found to be independent of the swelling temperature, the layered silicate concentration and the type of acid used to protonate ALA (HCl, H2 SO4 , H3 PO4 ). ALA was then polymerized at high temperature (280 ◦ C) and under elevated pressure (ca. 20 bar) with both types of swollen clay. XRD and TEM, coupled with energy dispersive X-ray (EDX), as well as atomic force microscopy (AFM), confirmed that the resulting structures were partially exfoliated and otherwise intercalated nanocomposites. However, although in situ polymerization was successfully applied for the preparation of PA6 and PA12/clay nanocomposites, few publications focused on the preparation of polyamide from diamine and diacid. In one of these studies, Wu et al. [97] investigated the preparation of PA1012 nanocomposite by polycondensation polymerization. A dispersion of organoclay in absolute alcohol was added to 1,10 diaminodecane in absolute alcohol. Then, this mixture was added to an absolute alcohol solution of 1,10-decanedicarboxylic acid under vigorous stirring, resulting in the immediate precipitation of a diaminodecane–decanedicarboxylic acid salt. The salt was recrystallized from a mixture of alcohol and water and

1133

was obtained as a white powder. It was then added with a slight excess of diaminodecane to a U-shaped glass tube which was purged with nitrogen before the reaction. The tube was immersed in an oil bath and the temperature was quickly raised to 200 ◦ C to start the reaction. After maintaining the autoclave for 2 h at 200 ◦ C, the temperature was increased to 215 ◦ C and held for 1.5 h under these conditions. The glass tube was flushed with nitrogen each time to remove the water produced in polycondensation. In the last step a vacuum ( MMW > LMW composites. As a result, tensile tests revealed superior performance for the higher molecular weight nylon 6 composites, particularly those based on HMW, as will be discussed in Section 6.1.2. However, probably the most critical condition for the formation of intercalated and especially exfoliated hybrids via polymer melt intercalation, is the presence of polar type interactions (i.e. other than Van der Waals forces). Therefore, polar polymers containing groups capable of associative type interactions, such as Lewis acid-base interactions or hydrogen bonding, favor the intercalation of macromolecular chains into the silicate galleries [47,49], while in the case of apolar polymer matrices, clay delamination typically requires the use of compatibilizers, as discussed in the following paragraph. A good example demonstrating the importance of polar interactions is the formation of EVA-based nanocomposites via melt intercalation. It is well established that the presence of polar groups (ester groups of the vinyl acetate moieties) all along the chains improves the ability of these polymers to intercalate in organo-modified montmorillonites [1]. Therefore, several studies have focused on the effect of the vinyl acetate content on the dispersion of clay nanoplatelets. In general, it has been observed that the higher the vinyl acetate content the larger is the basal spacing increase of the clay, inducing the formation of intercalated to exfoliated nanostructures. In a representative study, Chaudhary et al. [171] prepared EVA-based nanocomposites via melt intercalation using an intermeshing counter-rotating twin screw extruder. They used EVA copolymers with 9, 18 and 28% VA (vinyl acetate) and two organo-modified clays: Cloisite 15A (C15A), which is more suitable for the less polar EVA9 due to long aliphatic chains in C15A, and Cloisite 30B (C30B), suitable for the more polar polymers, like EVA18 or EVA28. They prepared composites with filler level varying from 2.5 to 7.5 wt.%, and subsequently characterized the structures obtained by WAXD and TEM. The results indicated that only intercalation occurred in the case of the less polar EVA9, while the clay was exfoliated in the more polar EVA18 and EVA28. Therefore, the authors concluded that an increase in the content of polar VA groups in EVA 18 and EVA28 as compared to EVA9, which lowered the thermodynamic energy barrier for clay–polymer interaction, possibly allowed a relatively higher number of polymer chains to migrate and stabilize within the clay platelet and form partially exfoliated and/or disordered intercalated states. It is worth mentioning that despite the presence of polar groups in EVAs, polymer compatibilizers have been used in the preparation of EVA-based nanocomposites. For example, Li and Ha [172] selected a maleic anhydride grafted EVA containing 18 mol% VA to process its nanocomposites with organo-modified Cloisite through melt blending at 175 ◦ C and found that the dispersion of Cloisite in the maleic anhy-

1150

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

dride grafted EVA was much better than in the simple EVA matrix. 5.4.2.4. The effect of melt intercalation processing conditions. Melt processing conditions play a key role in achieving high levels of exfoliation. Indeed, nanocomposites have been formed using a variety of shear devices (e.g. extruders, mixers, ultrasonicators, etc.), among which twin screw extruders have proven to be most effective for the exfoliation and dispersion of silicate layers [14]. The screws in twin screw extruders intermesh so that the relative motion of the flight of one screw inside the channel of the other acts as a paddle that pushes the material from screw to screw and from flight to flight. Two different patterns for intermeshing twin-screw extruders are possible. In the co-rotating pattern the screws rotate in the same direction and the material is passed from one screw to another and follows a path over and under the screws. This gives high contact with the extruder barrel, which improves the efficiency of heating. The path also ensures that most of the resin will be subjected to the same amount of shear as it passes between the screws and the barrel. The self-wiping nature of the co-rotating screws is much more complete than in the counter-rotating system, thus in the co-rotating case there is less likelihood that material will become stagnant. In the counter-rotating pattern, on the other hand, the screws rotate counter to each other and the material is brought to the junction of the two screws, building up in what is called the material bank on the top of the junction. This buildup of material is conveyed along the length of the screw by the screw flights. As the material passes between the screws, high shear is created, but shear elsewhere is very low. Since only a small amount of material passes between the screws, total shear is lower than in single-screw extruders and in co-rotating twin-screw extruders. Therefore, co-rotating systems are more effective than either counter-rotating or single-screw extruders [173]. On the other hand, although it is often stated that twin screw extruders favor intercalation when compared to single screw systems, this may not be the case for counter-rotating extrusion systems, for the aforementioned reason. Focusing on the effect of the extrusion system on the degree of intercalation, Cho and Paul [15] prepared nylon 6/o-MMT nanocomposites using either an intermeshing corotating twin screw extruder or a single screw extruder. They found that for the composite prepared by single screw extrusion, full exfoliation is not achieved, which was attributed to insufficient amount of shear and short residence time. On the other hand, by mixing in the twin screw extruder, the organoclay is uniformly dispersed into nylon 6 and the individual layers are aligned along the flow axis. However, other researchers have reported on nanocomposite preparation using single-screw extruders. For example, McNally et al. [36] successfully prepared PA12/clay nanocomposites using conventional single-screw melt blending. Moreover, it is important to notice that, even for a given extruder, processing conditions may determine the outcome. More specifically, increasing the mean residence time in the extruder generally improves the delamination

and dispersion. However, there appears to be an optimum extent of back mixing and an optimum shear intensity; excessive shear intensity or backmixing apparently causes poorer delamination and dispersion [52]. Often, special screw designs, including provisions for additional mixing, or static mixers at the end of the screw are used to enhance mixing and thus the silicate dispersion [173]. Li et al. [174] developed an ultrasonic extrusion technology, which organically combines extruder and ultrasound power. The authors claimed that the introduction of ultrasonic irradiation in extrusion processing of polymer can improve the processibility of polymer materials, and also reduce the size and size distribution of dispersed particles in polymer blends. They used the ultrasonic oscillation extrusion system developed to prepare polymer/MMT nanocomposites. The system, consisting of an extruder and a cylinder die connected to a generator of ultrasonic oscillations in the direction parallel to the flow of the polymer melt, was found to improve the exfoliation of the clay, though only for specific matrices. It should be noted at this point that, apart of the various extrusion systems, internal mixers (i.e. batch mixing devices where mixing occurs in a closed chamber) may also be successfully used for the preparation of exfoliated nanocomposites, as demonstrated, for example, in the case of PEI matrix [46,52]. However, these devices appear to be much less popular in nanocomposite preparation. Another factor affecting the resulting structures in the case of crystalline polymer matrices, is the crystallization temperature. For example, Okamoto and co-workers [175,176] observed through X-ray analyses that the intergallery spacing of PP-MA based nanocomposites increases with the crystallization temperature Tc for any amount of clay content in the nanocomposites. The microstructure of the nanocomposites, observed directly by TEM, showed that the clay particles are well dispersed at low Tc and that segregation of silicate layers occurs at high Tc . This implies that, by controlling intercalation through crystallization at a suitable temperature, one can control the fine structure, the morphology—and thus the properties of crystalline polymer/clay nanocomposites. Conclusively, a number of factors affect the outcome of melt intercalation. In this context, Dennis et al. [52] presented a simplified scheme to underline the conditions under which clay exfoliation into a polymer occurs during melt blending. The proposed scheme (Fig. 29a) is based on the relationship between the compatibility of the chemistry of the clay treatment and the matrix and the process conditions used to make a nanocomposite. It distinguishes three cases. The first case is chemistry-dependent. When the clay chemical treatment and the resin are compatible, almost any set of processing conditions can be used to form an exfoliated nanocomposite. In case 2, clay chemical treatment and polymer are marginally compatible. In this situation, the process conditions can be optimized to give an exfoliated nanocomposite. That is, the organoclay chemical treatment and the matrix are compatible enough that processing conditions can be tailored to optimize delamination and dispersion. Finally, in case 3, there is no apparent compatibility of the clay chemical treatment and the polymer. Processing conditions can be optimized to

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 29. Proposed mechanism of how the organoclay particles disperse into polymers during melt processing. Part (a) shows three cases involving the interplay between chemistry and process conditions in the mixing device. Part (b) illustrates schematically how platelets peel apart under the action of shear [52]. Reproduced from Dennis, Hunter, Chang, Kim, White, Cho and Paul DR by permission of Elsevier Science Ltd., UK.

give intercalants or tactoids that are minimized in size, but even partial exfoliation does not occur. These authors also presented possible clay delamination paths (Fig. 29b) to demonstrate that increasing shear intensity is not enough to achieve exfoliation. In pathway 1, stacks of platelets are decreased in height by sliding platelets apart from each other, a pathway that requires shear intensity. Pathway 2 shows polymer chains entering the clay galleries pushing the ends of platelets apart. This pathway does not require high shear intensity, but involves diffusion of polymer into the clay galleries (driven by either physical or chemical affinity of the polymer for the organoclay surface) and is, thus, facilitated by residence time in the mixing device. As more polymer enters and goes further in between clay platelets, especially near the edge of the clay galleries, the platelets appear to peel from the edge, since they are able to bend [14,52]. 5.4.3. Compatibility issues in non-polar polymers In contrast to polar polymers, like polyamides, that can effectively exfoliate organically modified clays using conventional melt processing techniques, for non-polar polymers, such as the most widely used polyolefins, PE or PP, synthesis of well exfoliated nanocomposites appears to be more difficult, because these polymers are so hydropho-

1151

bic and lack suitable interactions with the clay surface, even after it has been organically modified [49,162,177–179]. However, the development of polyolefin/clay nanocomposites is a field of rapidly growing industrial relevance due to their promise of improved performance in packaging and engineering applications [111]. Therefore, ways to resolve the difference in polarity between polyolefins and clays, in order to prepare nanocomposites by conventional melt compounding, have been proposed. More specifically, initial attempts to create non-polar polymer/clay nanocomposites by simple melt mixing were based on the introduction of a modified oligomer to mediate the polarity between the clay surface and the polymer [2,49,177]. The most promising strategy at the present time is to add a small amount of a maleic anhydride grafted polyolefin that is miscible with the base polyolefin. It is believed that the polar character of the anhydride has an affinity for the clay materials, such that the maleated polyolefin can serve as a “compatibilizer” between the matrix and the filler [162,180,181]. In fact, Zhai et al. [179] showed that two kinds of hybrids are formed by melt mixing: an o-MMT with PE and with PE-g-MA, respectively. For the PE/o-MMT system the intercalate effect is limited and the dispersion of clay is unsatisfactory. However, for the PE-g-MA/o-MMT nanocomposites, MMT was exfoliated in the matrix, as testified by both XRD and TEM. Wang et al. [49] prepared several types of nanocomposites with different compositions of the organically modified clay and maleated polyethylene by melt compounding at 140 ◦ C using a Brabender mixer operating at 60 rpm for 20 min. They found that the alkylammonium chain length may change the degree of interaction between clay and polyethylene and that the original basal reflection peak of the clay disappeared completely above a certain grafting level of MA, about 0.1 wt.%. Quite interestingly, Zhang and Wilkie [182] obtained PEorganoclay nanocomposites by adding maleic anhydride directly as a compatibilizer during the melt blending. As the authors suggested, the maleic anhydride probably reacted with PE during the high temperature blending in air, leading up to the formation of a graft copolymer in which maleic anhydride units are attached to the PE chains. In another study, Tang et al. [183] described the preparation of PP-based nanocomposites through a successful combination of clay modification and intercalation in one step. The authors mixed pristine MMT with the surfactant (C16) and PP with or without PP-MA using a high speed mixer. The mixed powder was then processed in a twin screw extruder and nanocomposites were obtained. The results of this study showed that the structure of PPclay nanocomposites is sensitive to the compatibilizer and the surfactant, since their increasing concentrations will reduce the free energy of the system, which is favorable for thermodynamic stability. The dispersion mechanism proposed is the following. At first, some surfactant chains diffuse into the interlayer under physical absorption and shear, rendering the clay organophilic. However, some surfactant remains in the polymer matrix, which may enhance the compatibility when the matrix is intercalated into the interlayer. In fact, the authors suggested that there is some

1152

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 13 Structures of the polar polymers used as composite matrices. Polymer

Structure

LDPE

EMA (y = 0.215)

EVA −18, −28 (x = 0.18, 0.28)

EMAAA (y = 0.18, z = 0.06)

PE-g-MAH (y ≈ 0.02)

interaction between the polymer matrix and the surfactant, just as in the interaction between the surfactant and the silicates. On the other hand, some PP-MA may be intercalated into the interlayer of MMT, after the surfactant makes the clay sufficiently organophilic. At the same time, PP-MA may act as a high molecular weight surfactant. The interlayer spacing of the clay increases and, if the miscibility of PP-MA with PP is good enough to allow dispersion at the molecular level, the exfoliation of intercalated MMT should take place. In addition to maleic anhydride and maleic anhydride grafted PE, EVA has also been used as a compatibilizer to prepare PE-based nanocomposites. For example, Zanetti and Costa [177] prepared several types of composites with different PE/EVA ratios and 5 wt.% organoclay by melt compounding at 150 ◦ C using a Brabender internal mixer with a screw speed of 60 rpm for 10 min. The polymer EVA contained 19 wt.% VA. No interaction was obtained by compounding the PE with the clay in absence of a compatibilizer. However, 1 wt.% EVA was enough to intercalate all the organoclay. Further increasing the amount of EVA above 10 wt.% caused a decrease in the degree of coherent layer stacking (i.e. a more disordered system). Zhao et al. [184] investigated chlorosilane-modified montmorillonites and their results showed that intercalated PE nanocomposites were obtained by melt intercalation using common alkylammonium intercalated clay, which was pretreated with chlorosilane. In a later work, the authors used directly a reactive intercalating agent (N-␥-trimethoxyl-silanepropyl) octadecyldimethylammonium chloride (abbreviated JSAc) to modify the montmorillonite clay, so that the chemical reaction with hydroxyl groups at the edge of the clay layers and the interlayer ion exchange were carried out in one step. PE/clay nanocomposites were then directly prepared by melt inter-

calating PE and the above mentioned clay in a twin screw extruder at 180 ◦ C and 200 rpm, whereas only microcomposites were formed when using common intercalating agent. It is also worth mentioning the work of Preston et al. [178], who prepared nanocomposites using the following matrices: two EVAs with different vinyl acetate contents, poly(ethylene-co-methyl acrylate) (EMA), poly(ethyleneco-methyl acrylate-co-acrylic acid) (EMAAA), and a blend of LDPE with maleated ethylene copolymer (PE-g-MA). Structures for each of these materials are given in Table 13. The organoclay they used was organically modified bentonite clay. Composites were prepared by melt mixing in a twin screw extruder operating at 380 rpm with a screw configured for intensive mixing. Through XRD measurements, the authors concluded that no interaction was likely between the LDPE and the silicate, whereas intercalation of the organoclay occurred in the presence of the four polar polymers. As in the case of PE, it is difficult to get exfoliated and homogenous dispersion of the silicate layer at the nanometer level in polypropylene, due to its low polarity. Consequently, PP also is usually modified with polar oligomers prior to introduction of modified clay, in order to achieve nanometric dispersion of the clay [161]. One typical example is the PP/clay nanocomposite system described by Toyota. The Toyota research group prepared PP/clay nanocomposites by direct melt compounding of PP with organo-modified MMT, in the presence of a maleic anhydride grafted PP (PP-g-MA). They added three times as much PP-g-MA as the clay by weight to prepare well mixed PP/clay nanocomposites, and pointed out that the miscibility between maleated oligomer and matrix polymer played a key role in composite properties [2,49,177,185]. In fact, it has been suggested that the relative

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1153

Fig. 30. Schematic illustration of OMLS dispersion process in PP-g-MA matrix [187]. Reproduced from Hasegawa, Okamoto, Kawasumi, Kato, Tsukigase and Usuki by permission of Wiley-VCH, Germany.

content in maleic anhydride cannot exceed a given value, in order to retain some miscibility between PP-MA and PP chains. When too many carboxyl groups were spread along the polyolefin chains, no further increase in the interlayer spacing was obtained in PP/PP-g-MA/clay blends, leading rather to the dispersion of PP-g-MA intercalated clay in the PP matrix [1]. Similarly, Hasegawa et al. [186] reported the preparation of exfoliated PP-based nanocomposite by melt blending PP-g-MA and organically modified MMT at 200 ◦ C, using a twin screw extruder. Fig. 30 shows a schematic representation of the clay dispersion process in PP-MA-based nanocomposites. According to the authors, the driving force of exfoliation originates from the strong hydrogen bonding between the MA groups and the polar clay surface. Kato et al. [188] prepared PP-based nanocomposites by the melt intercalation of PP chains modified by either maleic anhydride (PP-g-MA) or hydroxyl groups (PP-OH) in o-MMT. For both matrices, intercalated nanocomposites were recovered after melt blending at 200 ◦ C for 15 min. However, a PP-g-MA matrix with a lower maleic anhydride content did not intercalate under the same conditions, indicating that a minimal functionalization of PP chains has to be reached for intercalation to proceed. The authors also noticed that intercalation increased with the polymer-toclay ratio, i.e. when the PP-g-MA fraction was increased. Using the same method, Okamoto et al. [189] prepared PP/MMT nanocomposites. The authors mixed PP-g-MA (0.2 wt.% MA) and different amounts (2, 4 and 7.5 wt.%) of C18-MMT in a twin screw extruder at 200 ◦ C and obtained nearly exfoliated structures when 2 wt.% clay was added. However, addition of 4 and 7.5 wt.% clay led to disordered intercalated nanocomposites and ordered intercalated structures, respectively. Lopez et al. [161] used two different polar coupling agents, diethyl maleate grafted PP (PP-g-DEM) and commercial maleic anhydride grafted PP (PP-g-MA) to prepare

PP-based nanocomposites. DEM was chosen as the compatibilizing agent, because of its high thermal stability, high boiling point, and good compatibilization with polyolefins, compared to other compatibilizing agents. Furthermore, the low homopolymerization behavior of DEM allows better control of the functionalization reaction. Maleic anhydride was used as reference, since it has been widely used as compatibilizer for this kind of system. The PP/clay hybrids were prepared by melt compounding with two different clays, commercial modified montmorillonite and sodium bentonite modified with octadecylammonium ions. The results showed that although the commercial clay outperforms the octadecylammonium treated bentonite, differences in mechanical properties when using different clays are smaller if DEM is used instead of MAH. This is a consequence of the very low degree of compatibilization between the polymer matrix and the clay. In fact, this study proves that clay dispersion and interfacial adhesion are greatly affected by the kind of matrix modification. DEM has a lower polarity compared to MAH, providing a less effective interaction with the polar components of the clay. The authors therefore, concluded that clay and matrix modification are synergistic factors which need to be properly modulated in order to obtain the desired final properties in this kind of non-polar polymer-based nanocomposite. Finally, as in the case of PE and PP matrices, compatibilization is a critical issue also for other polymers, such as PS. Therefore, Wang and Wilkie [190] prepared PS/clay nanocomposites by in situ reactive blending with both the organically modified clays and the pristine inorganic clay in the presence of maleic anhydride, and found that maleic anhydride increases the possibility of nanocomposite formation. Also, Hasegawa et al. [191] produced partially exfoliated PS/clay nanocomposites by compounding in a twin-screw extruder organically modified MMT with an blend of PS and ≥50% of another compatibilizer, namely poly(styrene-co-methyl vinyl oxazoline).

1154

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 14 Molecular weight of host polymers. SPU

As-received PU Solvent cast Melt compounded

HPU

Mn

PDI

Mn

PDI

216.000 121.000 66.000

1.8 1.7 2.0

85.000 82.000 59.000

2.0 2.1 1.9

5.4.4. Degradation problems encountered during melt intercalation Despite the aforementioned advantages of polymer melt intercalation, this technique may involve polymer degradation problems since, when preparing clay/polymer nanocomposites using melt blending, a certain temperature is needed in the processing. Also, apart from the polymer matrix degradation, if the processing temperature to make the PLS is beyond the thermal stability of the organic treatment on the OMLS, some decomposition will take place. The onset temperature of decomposition of the organic modifier is, therefore, important in the process to make a polymer/clay nanocomposite, since polymer processing is normally done above 150 ◦ C. Moreover, in addition to common detrimental aspects of degradation, the resulting products may play a major but yet to be determined role in the formation of exfoliated nanostructures [43]. Therefore, the degradation issues encountered during melt intercalation have been addressed in several studies. For example, Finnigan et al. [88] prepared TPU nanocomposites by both twin screw extrusion and solvent casting, in order to compare the outcomes of these methods. The authors employed two TPUs: a soft elastomer (SPU) and a hard elastomer (HPU) consisting of the same soft and hard segments, but in different relative amounts. As demonstrated by WAXD analysis, both processing routes led to delaminated structures, which illustrates that if there is a good driving force for intercalation between the polymer and organosilicate, the need for an optimized processing route is diminished. Although melt compounding offered slightly better silicate dispersion than solvent casting, the authors suggested that solvent casting must be the preferred processing route for these materials, owing to elimination of PU and surfactant degradation. In fact, as shown in Table 14, it was found that the number average molecular weight (Mn ) of PU significantly decreases during melt compounding and, to a smaller extent, during solvent casting (due to an ultrasonic probe that was applied). In this particular case, the authors identified as additional causes of thermal degradation the small size of the extruder (and thus the large surface to volume ratio) as well as the absence of additives to reduce degradation. Xie et al. [43] focused on the effect of organic modifiers on the thermal decomposition of OMLSs and found that, while different long alkyl substituents have no effect or very little effect, the thermal degradation of organically modified montmorillonite is quite different compared to that of pure montmorillonite. The DTGA thermal curve shown in Fig. 31 for the OMLS was considered in four parts: (a) the free water region below 200 ◦ C; (b) the region where organic substances evolve in the tempera-

Fig. 31. Comparison of DTGA curves of various OMLS [43]. Reproduced from Xie, Gao, Liu, Pan, Vaia, Hunter and Singh by permission of Elsevier Science Ltd., UK.

ture range 200–500 ◦ C; (c) the structural water region in the temperature range 500–800 ◦ C; (d) a region between 800 and 1000 ◦ C where organic carbon reacts in some yet unknown way. In OMLS sample the free water disappears by 40 ◦ C. There is no interlayer water in OMLS as the quaternary ammonium salt has been exchanged for the hydrated sodium cation. The most distinguishing difference between the sodium montmorillonite and the organically modified montmorillonite is in the temperature range 200–500 ◦ C, as the organic constituent in the organo-clay starts to decompose somewhat above 200 ◦ C. Another distinguishing difference between sodium montmorillonite and the organically modified montmorillonite is in the temperature range from 800 to 1000 ◦ C. Sodium montmorillonite is very stable when the temperature is higher than 800 ◦ C, however, the OMLS continues to lose weight and a larger amount of CO2 is released at temperatures over 800 ◦ C. Davis et al. [192] found that during melt blending MMT/PA6 nanocomposites in a twin-screw extruder at 240 ◦ C, a particular quaternary alkyl ammonium treatment of the montmorillonite clay degraded it to an extent correlated with extruder residence time. To address this issue, they conducted an investigation on the processing degradation of PA6/MMT nanocomposites and clay organic modifier. The results led them to the following conclusions: 1. PA6 nanocomposites significantly degraded during processing at 300 ◦ C. Within experimental uncertainty, drying at 120 ◦ C rather than 80 ◦ C prior to processing had little effect on the degree of degradation. Virgin PA6 did not degrade under identical processing conditions. 2. Thermal decomposition of PA6 nanocomposite may have resulted from hydrolytic peptide scission. The catalytic activity of MMT was not investigated in this particular study; however, on the basis of previous research, it appeared that MMT could be involved in PA6 thermal degradation. 3. Heating at 120 ◦ C for 4 h thoroughly dried virgin PA6; but drying at 80 ◦ C resulted in no water loss. The amount of volatile water in PA6 nanocomposites was greater than was observed in virgin PA-6. Longer drying times and

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1155

Fig. 32. Synthesis of the thermally stable organoclay [193]. Reproduced from Chang, Kim, Joo and Im by permission of Elsevier Science Ltd., UK.

higher temperatures resulted in drier PA6 nanocomposites. 4. MMT and water are responsible for the degradation of PA6 nanocomposites. 5. When PA6 was processed at 300 ◦ C some water was present, however, little degradation was observed. This means that: (a) water itself may not be sufficient to cause degradation, (b) water escapes from PA-6 faster than from the nanocomposite, (c) clay and water are a special catalyst combination and/or (d) clay is a source of high-temperature reactive water or hydroxyls. In addition to the aforementioned studies, which focus mainly on the degradation mechanism, others are exploring ways to overcome or limit degradation during polymer melt intercalation. In general, when a material is subjected to extrusion, degradation is detected as discoloration and lowered physical or mechanical properties. A strong odor may also indicate degradation. If the degradation is general, that is, the entire extrudate is affected, as shown by discoloration throughout, although darker streaks may also be present, the most likely cause is that the heat is too high for the speed of extrusion. The obvious solutions are to reduce the heat or to increase the extrusion speed. Some combina-

tion of these two variables are likely feasible since the speed of the extruder affects mechanical heating of the material [173]. However, this needs to be done carefully since, as mentioned above, processing conditions may affect the morphology of the resulting material. On the other hand, a number of researchers have developed and applied clays exhibiting high thermal stability. In this context, Chang et al. [193] developed a thermally stable montmorillonite through an ion exchange reaction between Na+ -MMT and dodecyl triphenyl phosphonium chloride (C12 PPh-Cl− ) (Fig. 32). Gilman et al. [194] described the preparation of PA6-based nanocomposites of MMT modified with trialkylimidazolium cations to obtain high stability OMLS at high processing temperatures. A surprising result reported in another study was that poly(3-hydroxybutyrate) (PHB) nanocomposites prepared via melt intercalation showed severe degradation as testified by GPC, when an organically modified MMT was used, whereas no degradation was found with nanocomposites based on organically modified fluoromica. Even though there is no explanation on how organically modified fluoromica acted to protect the system, the authors suggest that the presence of Al Lewis acid sites, which catalyze the

Table 15 PLS nanocomposites prepared by melt intercalation. Nanocomposite PA6/[(HE)2 M1 R1 ] modified MMT PA6/Na+ -MMT water slurry PS/alkylammonium modified MMT PEI/hexadecylamine modified MMT PEO/Li+ or Na+ -MMT PLA/C18-MMT PP/stearylammonium modified clay PP-MA/C18-MMT PP/o-MMT modified using an organic swelling agent (Tb = 100–200 ◦ C) EVA/dimethyl-dioctadecyl ammonium modified MMT PET/1,2-dimethyl-3-N-alkyl imidazolium salt modified MMT

Mixing device and conditions

Ref. ◦

Co-rotating twin screw extruder, 240 C, 280 rpm Extrusion and drying Statically heating at 165 ◦ C, vacuum, 25 h Internal mixer, 370 ◦ C, 30 min Statically annealing, 80 ◦ C, 6 h Twin screw extruder, 190 ◦ C Twin screw extruder, PP-MA compatibilizer Twin screw extruder, 200 ◦ C Twin screw extruder, 250 ◦ C 130 ◦ C Co-rotating mini twin screw extruder, 285 ◦ C

[14] [160] [29] [46] [196] [153] [197] [186] [166] [166] [148]

1156

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 16 Predicted reinforcing factors per number of platelets per stack. No. of platelets per stack

1 2 3 4 5 10

Reinforcing factor (RF) Halpin–Tsai equation

Mori–Tanaka theory

d0 0 1 = 0.96 nm

d0 0 1 = 1.8 nm

d0 0 1 = 0.96 nm

d0 0 1 = 1.8 nm

49.2 39.7 33.5 29.0 25.6 16.3

49.2 27.5 21.1 17.5 15.1 9.3

34.8 23.8 18.3 14.9 12.7 7.5

34.8 16.6 11.7 9.3 7.8 4.6

hydrolysis of ester linkages at high temperature, may be one reason [195]. Finally, it is worth noting at this point that, despite the aforementioned degradation problems encountered during melt intercalation, very few authors have used stabilization systems in the preparation of polymeric nanocomposites. In Table 15 several PLS nanocomposites prepared via melt intercalation are presented as typical examples. 6. Nanocomposite properties 6.1. Mechanical properties 6.1.1. The reinforcing mechanism of layered silicates The first mechanism that has been put forward to explain the reinforcing action of layered silicates is one also valid for conventional reinforcements, such as fibers, which is schematically depicted in Fig. 33. That is, rigid fillers are naturally resistant to straining due to their high moduli. Therefore, when a relatively softer matrix is reinforced with such fillers, the polymer, particularly that adjacent to the filler particles, becomes highly restrained mechanically. This enables a significant portion of an applied load to be carried by the filler, assuming, of course, that the bonding between the two phases is adequate [64]. From this mechanism it becomes obvious that the larger the surface of the filler in contact with the polymer, the greater the reinforcing effect will be. This could partly explain why layered silicates, having an extremely high specific surface area (on the order of 800 m2 /g) impart dramatic improvements of modulus even when present in very small amounts in a polymer. In fact, the low silicate loading required in nanocomposites to effect significant property improve-

ments, is probably their most distinguishing characteristic. In most conventionally filled polymer systems, the modulus increases linearly with the filler volume fraction, whereas for nanocomposites much lower filler concentrations increase the modulus sharply and to a much larger extent [55]. However, some authors have argued that the dramatic improvement of modulus for such extremely low clay concentrations (i.e. 2–5 wt.%) cannot be attributed simply to the introduction of the higher modulus inorganic filler layers. A proposed theoretical approach assumes a layer of affected polymer on the filler surface, with a much higher modulus than the bulk equivalent polymer. This affected polymer can be thought of as a region of the polymer matrix that is physisorbed on the silicate surface, and is thus stiffened through its affinity for and adhesion to the filler surface. Obviously, for such high aspect ratio fillers as the layered silicate layers, the surface area exposed to the polymer is huge and, therefore, the significant increases in the modulus with very low filler content are not surprising. Furthermore, beyond the percolation limit, the additional silicate layers are incorporated in polymer regions that are already affected by other silicate layers, and thus it is expected that the enhancement of modulus will become much less dramatic [198]. In order to prove the effect of degree of exfoliation on nanocomposite mechanical properties, Fornes and Paul [64] used an analytical approach to elucidate how incomplete exfoliation influences nanocomposite stiffness. They expressed the modulus of a simple clay stack in the direction parallel to its platelets, by using the rule of mixtures: Estack = MMT EMMT + gallery Egallery where MMT is the volume fraction of silicate layers in the stack, EMMT is the modulus of MMT, gallery is the volume fraction of gallery space and Egallery is the modulus of the material in the gallery, which is expected to be much less than EMMT . The volume fraction occupied by gallery space, gallery can be expressed in terms of X-ray d-spacings, as gallery =

Fig. 33. Reinforcement mechanism in composite materials.

(n − 1)(d0 0 1 − tplatelet ) d0 0 1 (n − 1) + tplatelet

where n is the number of platelets per stack, d0 0 1 is the repeat spacing between silicate particles, and tplatelet is the thickness of a silicate platelet. Obviously, when the number of platelets in a stack is equal to one, the system represents an individual exfoliated platelet. Table 16 shows how the number of platelets in a stack affects the reinforce-

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1157

Fig. 34. The effect of the number of platelets per stack on the (a) modulus and (b) aspect ratio for the simplified arrangement of platelets [64]. Reproduced from Fornes and Paul by permission of Elsevier Science Ltd., UK.

ment factor (RF in an unexchanged, non-expandable clay (d0 0 1 = 0.96 nm) as well as in an intercalated or organically modified clay (d0 0 1 = 1.8 nm). Increasing n in both stacking scenarios leads to lower reinforcement efficiencies, especially for the intercalated clay. Interestingly, the largest drop in reinforcement is experienced when going from one to two platelets per stack. Overall, the trends in Table 16 show the high sensitivity of nanocomposite stiffness to the level of exfoliation. Fornes et al. concluded that stacks of platelets reduce stiffness of nanocomposites through lower effective filler moduli and reduced aspect ratio, the effects shown separately in Fig. 34. 6.1.2. Modulus and strength In general, the addition of an organically modified layered silicate in a polymer matrix results in significant improvements of Young’s modulus, as can be seen in Table 17 for a number of different materials. For example, Gorrasi et al. [156] reported an increase from 216 to 390 MPa for a PCL nanocomposite containing 10 wt.% ammonium-treated montmorillonite, while in another study [201], Young’s modulus was increased from 120 to 445 MPa with addition of 8 wt.% ammonium treated clay in PCL. Similarly, in the case of nylon 6 nanocomposites obtained through the intercalative ring opening polymerization of ␧-caprolactam, a large increase in the Young’s modulus at rather low filler content has been reported, whatever the method of preparation: polymerization within organo-modified montmorillonite, polymerization within protonated ␧-caprolactam swollen montmorillonite or polymerization within natural montmorillonite in the presence of ␧-caprolactam and an acid catalyst [45]. However, exceptions to this general trend have been reported. As shown in Fig. 35, in crosslinked polyester/OMLS nanocomposites, the modulus decreases with increasing clay content; in fact, the drop for the 2.5 wt.% nanocomposite was greater than expected. To explain this phenomenon, it was proposed that the intercalation and exfoliation of the clay in the polyester resin serve to effectively decrease the number of crosslinks from a topological perspective. The origin of the greater drop in properties of the 2.5 wt.% nanocomposites may be traced to the morphology; i.e. it was observed that the sam-

ple showed exfoliation on a global scale compared to the nanocomposite containing 10 wt.% clay, indicating that the crosslinking density is inversely proportional to the degree of exfoliation [140]. Apart from the modulus, the addition of OMLS in a polymer matrix usually also increases the tensile strength compared to that of the neat polymer material. For example, Shelley et al. [32] reported a 175% improvement in yield stress accompanied by a 200% increase in tensile modulus for a nylon 6 nanocomposite containing 5 wt.% clay. However, it should be emphasized that the effect of nanocomposite formation on tensile strength is not as clear as in the case of the modulus since reductions of tensile strength upon nanocomposite formation have also been reported. Such examples are included in Table 18, which lists the tensile strengths of a number of nanocomposite materials and compares them with the corresponding values for the neat polymers. Most polymer–clay nanocomposite studies report tensile properties, such as modulus, as a function of clay content [31], as in Fig. 36. This plot of Young’s modulus of nylon 6 nanocomposite vs. filler weight content, shows a

Fig. 35. Tensile modulus vs. clay concentration for crosslinked polyester nanocomposites [140]. Reproduced from Bharadawaj, Mehrabi, Hamilton, Trujillo, Murga, Fan, Chavira and Thompson by permission of Elsevier Science Ltd., UK.

1158

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 17 Young modulus of various PLS nanocomposites. Nanocomposite

Clay content (wt.%)

Young modulus (GPa)

Ref.

PA6/MMT (in situ polymerization)

0 4.7 5.3

1.11 1.87 2.04

[199]

PA6(LMW)/MMT (melt intercalation)

0 3.2 6.4

2.82 3.65 4.92

[14]

PA6(MMW)/MMT (melt intercalation)

0 3.1 7.1

2.71 3.66 5.61

[14]

PA6(HMW)/MMT (melt intercalation)

0 3.2 7.2

2.75 3.92 5.70

[14]

PP(7.2% PP-g-MA)/OMLS

0 7.2

0.714 0.838

[186]

PP(21.6% PP-g-MA)/OMLS

0 7.2

0.760 1.010

[186]

EVA/Cloisite Na

0 3

0.0122 0.0135

[167]

EVA/Cloisite 20A

0 3

0.0122 0.0249

[167]

EVA/Cloisite 25A

0 3

0.0122 0.0220

[167]

EVA/Cloisite 30B

0 3

0.0122 0.0228

[167]

EVA/Nanofil 757

0 3

0.0122 0.0116

[167]

EVA/Nanofil 15

0 3

0.0122 0.0240

[167]

EVA/Somasif ME100

0 3

0.0122 0.0124

[167]

EVA/Somasif MAE

0 3

0.0122 0.021

[167]

Soft PU/30B (solution intercalation)

0 3 7

0.0075 0.0138 0.024

[88]

Soft PU/30B (melt intercalation)

0 3 7

0.0072 0.0114 0.0193

[88]

Hard PU/30B (solution intercalation)

0 3 7

0.050 0.086 0.134

[88]

Hard PU/30B (melt intercalation)

0 3 7

0.061 0.081 0.119

[88]

HDPE/o-MMT

0 0.9 1.8 2.8 4.0

1.020 1.060 1.250 1.380 1.360

[200]

constant large rate of increase of modulus up to ca. 10 wt.% of nanoclay, whereas above this threshold the aforementioned levelling-off of Young’s modulus is observed. This change corresponds to the passage from totally exfoliated structure (below 10 wt.%) to partially exfoliated—partially intercalated structure (for 10 wt.% clay and above), as determined by XRD and TEM [1,55].

In another study, Liu and Wu [146] studied the mechanical performance of PA66 nanocomposites prepared via melt intercalation, using epoxy co-intercalated clay. The tensile strength increases rapidly from 78 MPa for PA66 up to 98 MPa for PA66CN5, but the increasing amplitude decreases when the clay content is above 5 wt.%. A similar phenomenon is observed in the dependence of tensile

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1159

Table 18 Tensile strength of various PLS nanocomposites. Nanocomposite

Tensile strength (MPa)

Ref.

PA6/MMT (in situ polymerization)

Clay content (wt.%) 0 4.7 5.3

68.6 97.2 97.3

[199]

PA6(LMW)/MMT (melt intercalation)

0 3.2 6.4

69.2 78.9 83.6

[14]

PA6(MMW)/MMT (melt intercalation)

0 3.1 7.1

70.2 86.6 95.2

[14]

PA6(HMW)/MMT (melt intercalation)

0 3.2 7.2

69.7 84.9 97.6

[14]

PMMA/OMLS

0 12.6

53.9 62.0

[1]

PS/OMLS

0 17.2 24.6

28.7 23.4 16.6

[1]

EVA EVA/Cloisite Na EVA/Cloisite 20A EVA/Cloisite 25A EVA/Cloisite 30B EVA/Nanofil 757 EVA/Nanofil 15 EVA/Somasif ME100 EVA/Somasif MAE

0 3 3 3 3 3 3 3 3

28.4 25.9 25.8 26.2 30.7 27.6 26.7 24.5 25.1

[167]

Soft PU/30B (solution intercalation)

0 3 7

45 31 21

[88]

Hard PU/30B (solution intercalation)

0 3 7

58 44 34

[88]

PU/MMT

0 5 10 21.5

PE/JS

0 5 10 15

22 25 27 28

[184]

PE/DM

0 5 10 15

22 21 23 24

[184]

27 26 26 26 25

[200]

HDPE/o-MMT

0 0.9 1.8 2.8 4.0

modulus of PA66CN on clay content. The smaller increase in amplitude observed with a clay loading above 5 wt.% was again attributed to the inevitable aggregation of the layers at high clay content. Another example (Fig. 37) shows both the tensile modulus and the yield strength of neat PA12 and the nanocomposites, which increased steadily with increasing organoclay loading. Compared to the virgin polymer, the tensile modulus of PA12/clay systems was improved by about 40% upon adding only 5 wt.% of clay, while lim-

5.9 6.2 6.5 8.3

[142]

ited improvement of the tensile strength was observed by incorporating clay in the matrix. Again, it was suggested that there is an optimum clay concentration for nanocomposite tensile strength improvement. With further increase in clay loading a moderate decrease of tensile strength was observed, suggesting that the relative amount of intercalation/exfoliation of the clay morphology gradually increases with increasing clay content, since the tensile strength is usually sensitive to the degree of dispersion [147].

1160

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198 Table 19 XRD peak intensity (Im ) and Young’s modulus of various nylon-6-based nanocomposites obtained by a one-step in situ intercalative polymerization of ␧-caprolactam with Na–montmorillonite in the presence of different acids.

Fig. 36. Effect of clay content on tensile modulus, measured at room temperature, of organo-modified montmorillonite/nylon-6-based nanocomposite obtained by melt intercalation [170]. Reproduced from Liu, Qi and Zhu by permission of John Wiley & Sons, Inc.

Similarly, other factors that influence the degree of exfoliation, apart from the clay content, also have an impact on nanocomposite modulus and strength. This explains the variations observed in moduli of PA6 nanocomposites prepared by intercalative ring opening polymerization of ␧-caprolactam, with different kinds of acids to catalyze the polymerization (Table 19). The WAXD peak intensity Im , which is inversely related to the amount of exfoliated layers in the nanocomposite, also depends on the nature of the acid used to catalyze the polymerization process. For an increase in Im , a parallel decrease in Young’s modulus is observed, indicating that exfoliated layers are the main factor responsible for the improvement in stiffness, while intercalated particles, having a smaller aspect ratio, play a rather minor role [1,55]. Cho and Paul [15] studied the effect of mixing device and processing parameters on the mechanical properties

Fig. 37. Tensile modulus and yield strength of PA12/clay nanocomposites as a function of clay concentration [147]. Reproduced from Phang, Liu, Mohamed, Pramoda, Chen, Shen, Chow, He, Lu and Hu by permission of John Wiley & Sons Ltd. on behalf of the SCI.

Acid

Im (cps)

Young’s modulus (GPa)

Phosphoric acid Hydrochloric acid Isophtalic acid Benzenesulfonic acid Acetic acid Trichloroacetic acid

0 200 255 280 555 585

2.25 2.05 1.74 1.74 1.63 1.67

of polyamide nanocomposites. In the case of composites formed by single-screw extrusion, the exfoliation of the clay platelets is not extensive. Even after a second pass through this extruder, undispersed tactoids are still easily observed with naked eye. However, the tensile strength and modulus were slightly improved by the second pass. On the other hand, nylon 6 nanocomposites with good properties can be obtained over a broad range of processing conditions in the twin screw extruder. The final nanocomposite properties are almost independent of the barrel temperature over the range of typical nylon 6 processing, but they are slightly improved by increasing the screw speed or by a second pass through the extruder. Therefore, processing conditions need to be optimized to allow greater exfoliation of the clay platelets and, thus, greater improvement in mechanical properties. The effect of PA6 molecular weight and MMT content on nanocomposite tensile modulus is shown in Fig. 38. The addition of organoclay leads to a substantial improvement in stiffness for the composites based on each of the three PA6 samples examined, i.e. LMW, MMW and HMW (low, medium and high molecular weight, respectively). Interestingly, the stiffness increases with increasing matrix molecular weight at any given concentration, even though the moduli of the neat PA6s are all quite similar. Similar trends with respect to the level of organoclay content and molecular weight are evident in the yield strength results (Fig. 39). Yield strength increases with MMT content; how-

Fig. 38. Effect of MMT content on tensile modulus for LMW, MMW, and HMW based nanocomposites [14]. Reproduced from Fornes, Yoon, Keskkula and Paul by permission of Elsevier Science Ltd., UK.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1161

Table 20 Influence of maleic anhydride-modified polypropylene content on the stiffness of PP matrices and PP/clay nanocompositesa . Sample

Filler content (wt.%)

PP-MA content (wt.%)

Young’s modulus (MPa)

PP PP/PP-MA 7 PP/PP-MA 22 PPCC PPCH 1/1 PPCH 1/2 PPCH 1/3

0 0 0 6.9 7.2 7.2 7.2

0 7.2 21.6 0 7.2 14.4 21.6

780 714 760 830 838 964 1010

a PP = polypropylene; PP-MA x: polypropylene modified by maleic anhydride (x = wt.% of PP-MA in the blend); PPCC = polypropylene-based microcomposite, PPCH y/z = polypropylene based nanocomposite (y/z = weight ratio between y parts of filler and z parts of PP-MA).

Fig. 39. Effect of MMT content on yield strength for LMW, MMW, and HMW based nanocomposites [14]. Reproduced from Fornes, Yoon, Keskkula and Paul by permission of Elsevier Science Ltd., UK.

ever, while the HMW and MMW-based nanocomposites show a steady increase in strength with clay content, the LMW-based nanocomposites show a less pronounced effect. These differences with respect to molecular weight are attributed to the better exfoliation achieved for the higher molecular weight matrices [14]. Other factors that may play a crucial role in improvement of nanocomposite mechanical properties include the organic modification of the clay and the addition of compatibilizers to the polymer matrix. As a representative example, Young’s modulus values of PP/PP-MA nanocomposites are listed in Table 20 and compared with the corresponding microcomposite as well as simple PP-MA/PP polymer blends. It is readily observed that increasing the amount of PP-MA increases the modulus, while comparison of PP with the simple PP-MA/PP blends rules out any

possible effect of matrix modification due to the presence of increasing amounts of PP-MA [1]. In another study, Hotta and Paul [162] performed tensile tests on various PE and PE-MA nanocomposites based on organoclays with one or two alkyl tails. The increase in modulus with addition of MMT is much stronger for the organoclay with two alkyl tails than for the one with a single tail, as would be expected on the basis of the much better dispersion of clay platelets for the surfactant with two alkyl tails. Similar trends were observed also for nanocomposite yield strength. Interestingly, the authors noted that there is no advantage in adding PE-MA for building modulus or strength at low MMT content (≤2.5 wt.%), in spite of the morphological differences seen. On the contrary, there is a clear advantage in adding PE-MA at higher MMT contents. Even though the benefit for modulus is not as great as might be expected, in the absence of PE-MA, the yield strength actually decreases on addition of MMT beyond 2.5 wt.%. Table 21 lists the strength and modulus values for PE-based nanocomposites, in which the initial montmorillonite was modified by two intercalating agents: the commonly used dioctadecyldimethyl ammonium chloride (DM) and the reactive N-␥-trimethoxysilanepropyl octadecyldimethyl ammonium chloride (JS). Both strength and modulus are higher in the case of the reactive intercalating agent, owing to the better dispersion of the organoclay [184]. The effect of clay organic modification on nanocomposite mechanical properties is also demonstrated in Fig. 40, which presents the ultimate strength of PUnanocomposites with different contents of two organically treated montmorillonites: MO-MMT, treated with a thermally stable, aromatic amine modifier containing active groups, and C16 -MMT, treated with a quaternary alkyl ammonium salt. As can be seen the ultimate strength increased dramatically with clay content and reached a maximum at 5 wt.% MMT, where the ultimate strength of

Table 21 Mechanical properties of PE and PE/clay composites. Sample

Tensile strength (MPa)

Flexural strength (MPa)

Flexural modulus (MPa)

Izod impact strength (J/m)

PE PE/JS5 PE/JS10 PE/JS15 PE/DM5 PE/DM10 PE/DM15

22 25 27 28 21 23 24

26 28 33 38 26 31 30

710 780 1050 1330 750 980 1030

20 16 16 12 22 16 14

1162

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 22 Tensile test results of Polypox H205 nanocomposites processed with low shear. Specimen

Average ultimate tensile stress (N/mm2 )

Average tensile modulus (N/mm2 )

Relative ultimate stress

Relative tensile modulus

0% I30E 5% I30E 10% I30E

59.0 ± 0.6 54.3 ± 9.2 53.6 ± 6.5

2565 ± 11 2796 ± 31 3075 ± 56

1.000 0.920 0.909

1.000 1.090 1.199

Fig. 40. Effect of organic-MMT loading on the tensile strength of (a) PU/MO-MMT and (b) PU/C16-MMT [202]. Reproduced from Xiong, Liu, Yang and Wang by permission of Elsevier Science Ltd., UK.

the nanocomposites increased by about 450% for C16 -MMT and 600% for MO-MMT, compared with that of pure PU, indicating that the improved mechanical strength depends on the characteristics of the modifier [202]. The extent of improvement of nanocomposite mechanical properties will also depend directly upon the average length of the dispersed clay particles, since this determines their aspect ratio and, hence, their surface area [55,203]. At this point we note that several authors have also pointed out factors that have an adverse effect on nanocomposite modulus and/or strength and need to be taken into consideration when preparing nanocomposite materials. Quite interestingly, Gopakumar et al. [151] found that the exfoliation of 5 and 10 wt.% clay in PE-MA increased Young’s modulus by 30 and 53%, respectively, whereas the tensile stress at yield showed only a marginal increase, up to a maximum of 15% for the 10 wt.% clay composition. The authors noted that the greatly enhanced interfacial area derived from exfoliation of the clay improves the mechanical reinforcement potential of the filler. However, given that the mechanical properties of a filled system depend on two principal factors, i.e. crystallinity of the polymer matrix and

the extent of filler reinforcement, the degree of crystallinity must also be considered. In another study dealing with the effect of matrix variations on mechanical properties of nanocomposites, Chaudhary et al. [171] studied the tensile properties of nanocomposites based on EVAs with various VA contents and two alternative organoclays. Since in EVA with increasing VA content the crystallinity of the polymer decreases (and will lower the stiffness), while the polarity increases (and will increase the intercalation), the authors suggested that in their system, the stiffness and toughness responses would reflect an interplay of two factors: (a) an increase in the “rigid” amorphous phase due to polymer–clay intercalation and (b) an increase in the “mobile” amorphous phase due to the increasing VA content. Experimental results showed that the influence of increasing clay concentration on the tensile behavior of EVA matrices was significant only with a low or moderately polar EVA matrix (9 and 18% VA). Thus, a linear proportionality was found between clay concentration and tensile modulus for EVA-9 and EVA-18, a relation not observed with EVA-28. In fact, it is very difficult to compare the extent of the improvement of the mechanical properties of different EVA/clay nanocomposites reported so far, because EVAs of different vinyl acetate contents have been processed into the nanocomposites with different clays and different modifying agents by different methods [81]. In the case of high Tg thermosets, it has been suggested that neither intercalated nor exfoliated nanosilicates lead to an improvement of the tensile stress at break, but rather make the materials more brittle. This effect appears to be generally more pronounced for intercalated structures than for exfoliated ones [1]. The results of tensile tests conducted by Hackman and Hollaway [134] on epoxy nanocomposites conventionally prepared under low-shear (stirring for 5 h at 90 ◦ C) are highlighted in Table 22. The tensile modulus increased by 9.0 and 19.9% with 5 and 10 wt.% clay loading respectively. However, the ultimate tensile stress decreased with increasing clay content, although the variation was large. The authors attributed this phenomenon to the fact that large clay particles act as impurities and increase stress concentrations. Flexural tests were also conducted and the results are outlined in Table 23. As can be seen, the flexural modulus and ultimate flexural stress increased by 19.6 and

Table 23 Flexural test results of Polypox H205 nanocomposites processed with low shear. Specimen

Average flexural modulus (N/mm2 )

0% I30E 5% I30E 7.5% I30E 10% I30E

2755 2966 3101 3294

± ± ± ±

83 90 85 76

Average ultimate flexural stress (N/mm2 ) 95.0 97.3 102.2 102.4

± ± ± ±

1.8 2.0 3.2 3.1

Relative flexural modulus 1.000 1.076 1.126 1.196

Relative flexural stress 1.000 1.034 1.055 1.077

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1163

Table 24 Summary of tensile properties of thermoplastic PU-based nanocomposites. 30B content (wt.%)

Solvent cast

Melt compounded

Young’s modulus (MPa) SPU 0 3 7 HPU 0 3 7

7.5 13.8 24 50 86 134

Tensile strength (MPa)

Fail strain (%)

45 31 21

1136 1109 1030

58 44 34

898 808 704

7.7%, respectively, for specimens containing 10 wt.% clay. For nanocomposites processed under high shear, the tensile modulus and ultimate tensile stress increased by 18.7 and 9.3%, respectively, when 5 wt.% clay loading was applied. In this case, the improvement in ultimate tensile strength was attributed to the smaller particles not generating stress concentrations leading to premature failure. A summary of the tensile properties of soft (SPU) and hard (HPU) polyurethane elastomers and of the corresponding nanocomposites, prepared by either solvent casting or melt compounding, is provided in Table 24. As can be seen, upon silicate addition large improvements in stiffness were observed, which however were accompanied by a decrease in tensile strength and elongation [88]. Similar trends have been reported by Tortora et al. [204]. Both exfoliated and intercalated PU/o-MMT nanocomposites showed an improvement in the elastic modulus upon increasing the clay content, but a decrease in the stress and strain at break. In general, it has been argued that in the presence of polar or ionic interactions between the polymer and the silicate layers, the stress at break is usually increased, whereas when there is lack of interfacial adhesion, no or very slight tensile strength enhancement is recorded [1]. Pegoretti et al. [149] found that the yield strength was not reduced by the addition of clay to recycled PET and considered

Young’s modulus (MPa) 7.2 11.4 19.3 61 81 119

Tensile strength (MPa)

Fail strain (%)

21 22 7

1445 1163 568

44 20 15

776 283 100

this to be a sign of good interfacial adhesion; however, in the same study, a slight decrease of stress at break and a dramatic reduction of strain were reported. On the other hand, in PS intercalated nanocomposites the ultimate tensile stress was found to decrease compared to that of the PS matrix and dropped further at higher filler content. This lack of strength was attributed to the fact that only weak interactions exist at the PS/clay interface, contrary to other compositions in which polar interactions may prevail, strengthening the matrix interface [205]. It should be noted that several authors have reported inability to measure nanocomposite yield stress, because the materials often become brittle and fail before reaching the yield point. Such remarks were made by Gorrasi et al. [6], who conducted tensile tests on PCL nanocomposite, containing 30 wt.% clay, and also on blends of this nanocomposite with HMW PCL. For the blend containing 15 wt.% clay only the elastic modulus could be evaluated since the sample did not reach the yield point, while lower clay concentrations in the blend led to better mechanical properties in terms of flexibility and drawability. For the initial nanocomposite, however, it was not even possible to draw the sample because of its brittleness. An interesting study was performed by Chang et al. [193] who prepared PET-based nanocomposites through in situ intercalative polymerization, and subsequently pro-

Table 25 Tensile properties of PET hybrid fibers. Organoclay (wt.%)

DRa

Ultimate strength (MPa)

0 (Pure PET)

1 3 10 16

46 47 51 51

2.21 2.24 2.28 2.39

3 3 3 2

1

1 3 10 16

58 56 50 48

2.88 2.80 2.63 2.47

3 3 3 3

2

1 3 10 16

68 55 54 51

3.31 2.63 2.51 2.29

3 3 3 3

3

1 3 10 16

71 68 62 55

4.10 3.40 3.12 3.08

3 3 2 3

a b

Draw ratio. Elongation percent at break.

Modulus (GPa)

E.B.b

1164

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 26 Bending modulus of various PLS nanocomposites. Nanocomposite PLA/OMLS

Clay content (wt.%) 0 4 5 7

Bending modulus (GPa)

Ref.

4.8 5.5 5.6 5.8

[208]

PE/JS

0 5 10 15

0.71 0.78 1.050 1.330

[184]

PE/DM

0 5 10 15

0.71 0.75 0.98 1.030

[184]

duced nano-hybrid fibers by extrusion through the die of a capillary rheometer. The hot extrudates were stretched through the die of a capillary rheometer at 270 ◦ C and immediately drawn to various draw ratios (DR). As is evident from Table 25, the tensile properties of the fibers formed increased with increasing amount of organoclay at DR = 1. When the organoclay was increased from 0 to 3 wt.% in hybrids at DR = 1, the strength linearly improved from 46 to 71 MPa, and the modulus from 2.21 to 4.10 GPa. On the other hand, it is quite interesting to note the effect of DR on the tensile strength and modulus of PET and PET nanocomposite fibers. As shown in Table 25, for pure PET, the strength and modulus increased from 46 to 51 MPa and 2.21 to 2.39 GPa, respectively, as the DR was increased from 1 to 16. However, the ultimate strength and modulus of the hybrid fibers decreased markedly with increasing DR. An increase in the mechanical tensile strength with increasing DR is very common for engineering plastics and is usually observed in flexible coil-like polymers. However, nanocomposite fibers did not follow this trend. Chang et al. suggested that higher stretching of the fiber leads to debonding and creation of voids in the hybrid, which reduce the tensile mechanical properties. This study clearly illustrates that nanocomposite materials may have a different response to mechanical loads than the corresponding neat polymer matrices. Finally, even though nanocomposite researchers are generally interested in the tensile properties of the final materials, there are a few reports concerning the flexural properties of PLS nanocomposites [206,207]. Some results

obtained by bending tests on nanocomposite materials are presented in Tables 26 and 27. 6.1.3. Toughness and strain The brittle behavior often exhibited by nanocomposites probably originates from the formation of microvoids due to debonding of clay platelets from the polymer matrix upon failure. This has been testified through careful inspection of fracture surfaces and is also correlated to observations by in situ deformation experiments using TEM [147,181]. In fact, the observation of nanocomposite fracture surfaces is quite interesting. Fig. 41(a) shows a typical fracture morphology in virgin nylon 12 and a ductile fracture as evidenced by plastic deformation. Fig. 41(b) and (c) show fracture surfaces of the nanocomposites containing 1 and 5 wt.% clay, respectively. No distinct clay agglomerates are observed by scanning electron microscopy (SEM) even at high magnification, as shown in Fig. 41(d). For 1 wt.% clay addition (Fig. 41(b)), the fracture surface became smoother compared with that of neat PA12; an even more brittle feature for clay concentration of 5 wt.% was observed in Fig. 41(c). Careful inspection of the fracture surface at higher magnification of nanocomposite with 5 wt.% clay (Fig. 41(d)) verifies the formation of microvoids due to the debonding of clay platelets from the matrix. Usually, microvoids are formed around the large inhomogeneities, which become evident especially at high clay loadings. These microvoids will coalesce with formation of larger cracks causing embrittlement, ultimately resulting in reduced toughness [147].

Table 27 Bending strength of various PLS nanocomposites. Nanocomposite

Bending strength (MPa)

Ref.

0 4 5 7

86 134 122 105

[208]

PE/JS

0 5 10 15

26 28 33 38

[184]

PE/DM

0 5 10 15

26 26 31 30

[184]

PLA/OMLS

Clay content (wt.%)

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1165

Fig. 41. SEM images showing fracture surfaces after impact tests: (a) neat PA12; (b) and (c) PA12 nanocomposites containing 1 and 5 wt.% clay, respectively; (d) high magnification of (c) [147]. Reproduced from Phang, Liu, Mohamed, Pramoda, Chen, Shen, Chow, He, Lu and Hu by permission of John Wiley & Sons Ltd. on behalf of the SCI.

In the case of nylon 12 nanocomposites, Fig. 42 shows that the Izod impact strength monotonically decreases as the clay concentration increases. The toughness (representing the energy absorption during the fracture process) decreases by about 25% with 5 wt.% of clay. Similar obser-

Fig. 42. Izod impact strength of PA12/clay nanocomposites as a function of clay concentration [147]. Reproduced from Phang, Liu, Mohamed, Pramoda, Chen, Shen, Chow, He, Lu and Hu by permission of John Wiley & Sons Ltd. on behalf of the SCI.

vations of reduction in impact strength are also reported in nylon 6/clay nanocomposites and PE-based nanocomposites, indicating that the incorporation of clay into semicrystalline thermoplastics usually results in toughness reduction, i.e. the aforementioned embrittlement effect from clay addition [147]. On the other hand, some studies report little or no change of toughness upon clay intercalation/exfoliation. For example, while the tensile strength and modulus of PP nanocomposites increased rapidly with increasing clay content from 0 to 5 wt.%, the notched Izod impact strength was constant, within experimental error, in the clay content range between 0 and 7 wt.% [158]. Another study reports the impact properties for exfoliated nylon 6-based nanocomposites prepared either by in situ intercalative polymerization or by melt intercalation. In that study marginal reductions in impact properties are reported, whatever the exfoliation process used. In the case of in situ intercalative polymerization, the Izod impact strength is reduced from 20.6 to 18.1 J/m when 4.7 wt.% clay is incorporated. Charpy impact tests show similar reduction in the impact strength, with a drop from 6.21 kJ/m2 for the filler free matrix, down to 6.06 kJ/m2 for the 4.7 wt.% nanocomposite. Fig. 43 shows that the decrease in the Izod impact strength of melt-intercalated nylon 6 nanocomposites is not very pronounced over a relatively large range of filler content [170]. Furthermore, toughness improvements upon clay dispersion have also been reported—a remarkable result,

1166

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 43. Effect of clay content on notched Izod impact strength of nylon-6/clay nanocomposites obtained trough melt intercalation [170]. Reproduced from Liu, Qi and Zhu by permission of John Wiley & Sons, Inc.

considering that conventional polymer–clay composites, containing aggregated nanolayer tactoids ordinarily improve rigidity but sacrifice toughness and elongation [7]. As an example, Liu and Wu [146] observed a toughening effect in PA66CN. The notched Izod impact strength increased from 96 to 146 J/m upon 5 wt.% clay addition, and remained higher than that of PA66, even with higher clay content. Such results are particularly surprising, considering that from length-scale arguments it is known that toughening occurs over a specific size range; effective toughening necessitates a filler size greater than 0.1 ␮m and may not be energetically favorable at nano-length scales. Also, the sizes of the nanoparticles are generally too small to provide toughening through a crack-bridging mechanism and cannot effectively enhance crack-trajectory tortuosity. Therefore, the extremely reduced scale of a fully exfoliated nanocomposite does not lend itself to a toughening application. However, in an intercalated system there is considerable interaction between silicate layers that might alleviate this concern [45]. For example, Zerda and Lesser [45] showed that the gross yielding behavior of a glassy thermoset was substantially modified upon the formation of intercalated nanocomposites, with void formation within clay aggregates leading to the evolution of a visible shear-banded

zone in compression samples. The fracture behavior appears to be most dramatically improved in the intercalated system. The fracture energy of the composites was increased by 100% at clay concentration of 5 wt.%. By investigating the surface roughness and crack propagation under subcritical loading, it has been hypothesized that the creation of additional surface area on crack propagation is the primary means for toughening intercalated systems. The morphology of the system plays an important role in the toughening mechanism because the spacing of regions of intercalated clay is important to toughening. It is believed, therefore, that the intercalated morphology can afford some property improvements that are unavailable to the fully exfoliated systems. Concerning the fracture behavior of EVA-based nanocomposites, Peeterbroeck et al. [167] concluded that it is independent of the origin of the clay, while it appears to be related to the nature of the clay organo-modifier and the state of nanocomposite dispersion. On the other hand, Kim et al. [44] attributed the enhanced toughness they observed for intercalated PA12 nanocomposites to the fact that some amount of applied energy is dissipated by splitting, sliding or opening of the separated bundles in the stacked layers. Also, LePluart et al. [136] reported that the incorporation of a benzyl dimethyl tallow alkyl ammonium montmorillonite in rubbery and glassy epoxy matrices leads to promising improvement of mechanical properties. They obtained an interesting stiffness/toughness balance for very low filler contents and without reducing the Tg of the matrix, which is particularly interesting considering how the brittleness of epoxies limits their use in technological areas where their high Tg is often highly appreciated. Quite interestingly, Fornes et al. [14] investigated how the matrix molecular weight as well as the extension rate during tensile tests affect the ductility of PA6-based nanocomposites. Fig. 44 presents the relationship between MMT content and elongation at break for PA6 matrices of different molecular weights, for two different rates of extension. As shown in Fig. 44a, the virgin polyamides are very ductile at a test rate of 0.51 cm/min. With increasing clay content, the ductility gradually decreases, however, the HMW and MMW-based composites attain reasonable lev-

Fig. 44. Effect of MMT content on elongation at break for LMW, MMW, and HMW based nanocomposites at a crosshead speed of (a) 0.51 cm/min and (b) 5.1 cm/min [14]. Reproduced from Fornes, Yoon, Keskkula and Paul by permission of Elsevier Science Ltd., UK.

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1167

Table 28 Elongation at break of various PLS nanocomposites. Nanocomposite

Clay content (wt.%)

Elongation at break (%)

Ref.

PA6(LMW)/MMT

0 3.2 6.4

28 11 4.8

[14]

PA6(MMW)/MMT

0 3.1 7.1

101 18 5.0

[14]

PA6(HMW)/MMT (melt intercalation)

0 3.2 7.2

129 27 6.1

[14]

EVA EVA/Cloisite Na EVA/Cloisite 20A EVA/Cloisite 25A EVA/Cloisite 30B EVA/Nanofil 757 EVA/Nanofil 15 EVA/Somasif ME100 EVA/Somasif MAE

0 3 3 3 3 3 3 3 3

1406 1403 1231 1259 1266 1358 1291 1312 1270

[167]

Soft PU/30B (solution intercalation)

0 3 7

1136 1109 1030

[88]

Soft PU/30B (melt intercalation)

0 3 7

1445 1163 568

[88]

Hard PU/30B (solution intercalation)

0 3 7

898 808 704

[88]

Hard PU/30B (melt intercalation)

0 3 7

776 283 100

[88]

0 5 10 21.5

950 1020 1065 1160

[142]

0 0.9 1.8 2.8 4.0

36 25 20 14 15

[200]

PU/MMT

HDPE/o-MMT

els of ductility at MMT concentrations as high as 3.5 wt.%, while the elongation at break for the LMW based nanocomposites decreases rapidly at low MMT content (around 1 wt.%). Even though the opposite result could have been anticipated, considering that high molecular weight matrices favor clay exfoliation, the authors attribute the larger reduction of elongation at break in the LMW-based systems to the presence of stacked silicate layers, as seen in TEM photographs. On the other hand, the higher testing rate of 5.1 cm/min yields similar trends, as shown in Fig. 44b, but the absolute level of elongation at break is significantly lower. Interestingly, the strain at break for LMW composites is relatively independent of the rate of extension, similar to what has been observed in glass fiber reinforced composites. Even at the highest clay content, the HMW composite exhibits ductile fracture, whereas the LMW and MMW based nanocomposites fracture in a brittle manner at the highest clay content. As in the case of toughness, contradictory results have also been presented concerning the effect of nanocom-

posite formation on the elongation at break, as can be seen in Table 28. Even though in most cases this property deteriorates when a layered silicate is dispersed into a polymer matrix, nanocomposites have been reported exhibiting similar or even higher elongations at break than the neat matrices. For example, in the case of EVA-12/MMT nanocomposites, a significant increase of both the strength and the elongation has been reported with the introduction of the organoclay into the EVA-12 matrix. However, this enhancement is a maximum when the clay concentration is only 2 wt.%. Further increase of clay content causes reduction in mechanical properties, probably due to aggregation of clay layers, as already discussed [209]. Thellen et al. [152] conducted tensile tests on PLA-based nanocomposite blown films and recorded improvements up to 40% for both the modulus and the elongation. Yao et al. [142] also reported improvements in strain at break. Data of tensile strength and strain-at-break vs. clay content are shown in Table 29. The authors suggested that the improved elasticity is due

1168

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Table 29 Tensile strength and strain-at-break vs. the loading of layered clay. Weight fraction of the clay (%)

Tensile strength (MPa)

Strain-at-break (%)

0 5 10 21.5

5.9 6.2 6.5 8.3

950 1020 1065 1160

in part to the plasticizing effect of gallery onium ions, which contribute to dangling chain formation in the matrix. Accordingly, Chen and Evans [210] observed a dramatic improvement in tensile elongation at break in the presence of clay. At low clay loadings, test pieces underwent yielding during tension, similar to pristine PCL, but with dramatic increases in ductility, quite the opposite of the usual effect of adding a particulate filler to a polymer. When the clay loading was high, typically higher than 20 wt.%, the composites became brittle and did not reach the yield point. Finally, it is worth summarizing the work of Hong et al. [185] on PP-based RTPO/clay nanocomposites, prepared by using PP-MA as a compatibilizer. PP-based RTPO (or in reactor made TPO) is a blend of PP and poly(ethyleneco-propylene) (EPR), produced by the bulk polymerization of propylene, followed by gas-phase copolymerization of ethylene and propylene driven by the TiCl4 /MgCl2 -based catalyst system. Such materials, like the conventional blends of PP/EPR prepared by mechanical blending, exhibit improved flexibility and toughness compared to neat PP. Moreover, because the rubber phase can be dispersed uniformly and reach a high degree of dispersion in these in situ blends, it is possible to achieve more intimate interaction between the matrix and the rubber phase. The compositions and tensile properties of polypropylene-based RTPO/clay nanocomposites are reported in Table 30. As can be seen, the tensile moduli of the nanocomposites became higher as the clay content increases. On the other hand, the elongation at break decreases as the clay content increases, but the value of nanocomposites containing 10 wt.% clay is 437%, which is much higher than that of PP/clay nanocomposites reported elsewhere. As the authors claim, these elongational properties of PP based RTPO/clay nanocomposites are unique and promising for many applications. In fact, for reasons of comparison, Hong et al. also prepared and tested nanocomposites using PP/EPR mechanical blend matrix, modified with PP-MA. For these materials, the elongation at break values were about 50%, which are much lower than those of RTPO clay nanocomposites and is not suitable for industrial application. The authors attributed this discrepancy to the difference of dispersion homogeneity and domain size of ethylene copolymer between RTPO and PP/EPR mechanical blends.

Fig. 45. Stress–strain curves for nylon 6 and 95/05 composites at a crosshead speed of 5.08 cm/min [15]. Reproduced from Cho and Paul by permission of Elsevier Science Ltd., UK.

6.1.4. Comparison and synergistic effects of clays and conventional reinforcements Typically, layered silicates are incorporated in polymeric materials as the sole reinforcing element. However, several studies have investigated the potential synergistic effects of clays and conventional reinforcements, such as glass fibers. In this context, Wu et al. [52] studied the effect of adding glass fibers to PA6 and PA6-based nanocomposites containing 3 wt.% montmorillonite. They found that the tensile strength of PA6/clay containing 30 wt.% glass fibers is 11% higher than that of PA6 containing 30 wt.% glass fibers, while the tensile modulus of the nanocomposites increases by 42%. Bending strength and bending modulus of neat PA6/clay are similar to PA6 reinforced with 20% glass fibers. However, the notched Izod impact strength of the nanocomposite is lower than that of neat polyamide 6, and is further decreased with the addition of fibers. In another study, typical stress-strain diagrams for PA6 and composites containing 5 wt.% of fillers are compared, as shown in Fig. 45 (at 5.08 cm/min) and Fig. 46 (at 0.5 cm/min). A summary of the mechanical properties of these materials is shown in Table 31. As can be seen from the table, regardless of the type of filler, the strength and modulus are substantially increased relative to the neat PA6, without significant variation in toughness or impact strength, as measured by the standard Izod test. Furthermore, nanocomposites show superior mechanical properties, especially modulus, as compared with conven-

Table 30 Compounding formulations and tensile properties of PP-based RTPO/PpgMA/clay nanocomposites. Sample

PP-based RTPO (wt.%)

PpgMA (wt.%)

Clay (wt.%)

Tensile strength at yield (MPa)

Tensile strength at break (MPa)

Elongation at break (%)

RTPO RTPO NC3 RTPO NC5 RTPO NC10

100 88 80 60

0 9 15 30

0 3 5 10

5.1 6.4 8.1 14.2

20.6 16.6 16.8 16.6

1390 980 859 437

Tensile modulus (MPa) 46.0 71.2 78.3 251

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

1169

elongation at break remain at the levels of neat PA6 up to about 5 wt.% of the organoclay, and decrease thereafter [15]. 6.2. Dynamic mechanical properties

Fig. 46. Stress–strain curves for nylon 6 and 95/05 composites at a crosshead speed of 0.51 cm/min [15]. Reproduced from Cho and Paul by permission of Elsevier Science Ltd., UK.

tional PA6 composites formed by compounding with glass fibers or untreated clay. The elongation at break for the nanocomposites is more or less the same as that of the neat PA6, whereas, values for the conventional composites are dramatically decreased. Also, the elongation at break for the nanocomposites is greatly affected by the crosshead speed, as is the case for neat PA6. On the other hand, rate of extension has little effect on the elongation of glass fiber composites. It is noteworthy that the composite of PA6 with untreated Na+ -MMT shows higher strength and modulus than neat PA6, which is quite contrary to the results from other investigators, who claim that untreated clay composites with PA6 are inferior to neat polymer in terms of some mechanical properties. Interestingly, a synergistic effect on the tensile strength and modulus is again observed when the exfoliated nanocomposite is used as the matrix for a glass fiber reinforced composite. As shown in Table 31, the modulus of the nanocomposite with 5 wt.% loading of the organoclay is improved about 38% relative to neat PA6 and the glass fiber composite shows a 22% improvement. When glass fibers are added to the nanocomposite, the modulus is 81% higher than that of PA6. This exceeds what is expected on the basis of simple additivity. Stiffness and strength are dramatically improved as the amount of organoclay increases. On the other hand, the impact strength and

Dynamic mechanical analysis (DMA) measures the response of a material to a cyclic deformation (usually tension or three-point bending type deformation) as a function of the temperature. DMA results are expressed by three main parameters: (i) the storage modulus (E or G ), corresponding to the elastic response to the deformation; (ii) the loss modulus (E or G ), corresponding to the plastic response to the deformation and (iii) tan ı, that is, the E /E (or G /G ) ratio, useful for determining the occurrence of molecular mobility transitions such as the glass transition temperature [1]. Indicatively, the temperature dependence of G , G and tan ı of a nylon 6 matrix and various nanocomposites is presented in Fig. 47. In the case of nanocomposites, the main conclusion derived from dynamic mechanical studies is that the storage modulus increases upon dispersion of a layered silicate in a polymer. This increase is generally larger above the glass transition temperature, and for exfoliated PLS nanocomposite structures is probably due to the creation of a three-dimensional network of interconnected long silicate layers, strengthening the material through mechanical percolation [1]. Above the glass transition temperature, when materials become soft, the reinforcement effect of the clay particles becomes more prominent, due to the restricted movement of the polymer chains. This results in the observed enhancement of G [55]. For example, an epoxy-based nanocomposite, containing 4 vol.% silicates, showed a 60% increase in G in the glassy region, compared to the unfilled epoxy, while the equivalent increase in the rubbery region was 450% [135]. Similar results have also been reported in the case of PP- [189], PCL- [80], SBS[211], PA- [64,212], PLA- [83,153,208,213], and epoxy-based nanocomposites [135]. Enhancement of the loss modulus, G , has also been reported for nanocomposite materials, however this aspect of dynamic mechanical performance is far less discussed in the literature. Finally, the tan ı values are affected in different ways by nanocomposite formation, depending on the polymer matrix. For example, in PS based nanocomposites, a shift of tan ı to higher temperatures has been observed, accompanied by a broadening of this transition [205], while the opposite effect was reported in the case of PP-based nanocomposites [189]. Some authors observed a decrease

Table 31 Mechanical properties of polyamide 6 composites. Polyamide composites

Clay content (%)

Izod impact strength (J/m)

Modulus (GPa)

Yield strength (MPa)

Elongation at break (%) Crosshead speed 0.51 cm/min

Polyamide 6 PA6/glass fiber PA6/MMT PA6/organoclay PA6/organoclay/glass fiber

0 5 5 3.16 8

38 53 40 38 44

± ± ± ± ±

4 8 2 3 3

2.66 ± 0.2 3.26 ± 0.1 3.01 ± 0.1 3.66 ± 0.1 4.82 ± 0.1

64.2 ± 0.8 72.6 ± 0.8 75.4 ± 0.3 83.4 ± 0.7 95.0 ± 0.9

200 18 22 126 8

± ± ± ± ±

30 1.3 6.0 25 0.5

Crosshead speed 5.08 cm/min 40 14 14 38 7

± ± ± ± ±

8 4 3 19 4

1170

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 48. Proposed model for the torturous zigzag diffusion path in an exfoliated polymer–clay nanocomposite when used as a gas barrier [85]. Reproduced from Yano, Usuki, Okada, Kurauchi and Kamigaito by permission of John Wiley & Sons Inc.

this indicates a weakening of the thermomechanical stability of these materials at high temperature. 6.3. Barrier properties

Fig. 47. . Temperature dependence of G ; G and tan ı for N6 matrix and various N6CNs [55]. Reproduced from Ray and Okamoto by permission of Elsevier Science Ltd., UK.

of tan ı peaks, and considered this indicative of a glass transition suppression by the presence of the clay. However, Fornes and Paul [64] pointed out that this conclusion is a misinterpretation, since the low values for the nanocomposites are simply the result of dividing the relatively constant loss modulus values in the Tg region, by larger and larger storage modulus values. Quite surprisingly, DMA showed that above Tg , the moduli for the pure PU and the PU/o-MMT nanocomposites show no obvious difference, while below Tg , addition of o-MMT strongly influences the modulus values. Interestingly, the authors found that E and E of the PU/o-MMT decrease in comparison with values for the PU, for unclear reasons. On the other hand, significant enhancements of E and E were seen for the nanocomposite prepared using a particular modified clay [202]. In the case of PLA-based nanocomposites, it was observed that PLACNs with a very small amount of o-PCL as a compatibilizer exhibited a very large enhancement of mechanical properties compared to that of PLACN with comparable clay loading [153]. Krikorian and Pochan [83] also studied the dynamic mechanical properties of neat PLA and nanocomposites prepared with OMLS. These authors found that at high temperatures the reinforcement effect of OMLS weakens, and suggested that

Generally, polymer/layered silicate nanocomposites are characterized by very strong enhancements of their barrier properties. Polymers ranging from epoxies and good sealants (like siloxanes) to semi-permeable (e.g. polyureas) and highly hydrophilic (e.g. PVA) are all improved up to an order of magnitude by low clay loadings [31]. The dramatic improvement of barrier properties can be explained by the concept of tortuous paths. That is, when impermeable nanoparticles are incorporated into a polymer, the permeating molecules are forced to wiggle around them in a random walk, and hence diffuse by a tortuous pathway, as shown in Fig. 48 [4,7,55,206,214–216]. The tortuosity factor is defined as the ratio of the actual distance, d , that the penetrant must travel to the shortest distance d that it would travel in the absence of barriers. It is expressed in terms of the length L, the width W and the volume fraction of the sheets s as =

L d =1+ s 2W d

It becomes obvious from this expression that a sheet-like morphology is particularly efficient at maximizing the path length, due to the large length-to-width ratio, as compared to other filler shapes [1,50,55]. According to the model proposed by Nielsen, the effect of tortuosity on the permeability may, in turn, be expressed as PPCN 1 − s = Pp  where PPCN and PP represent the permeability of the nanocomposite and the pure polymer, respectively and s is the clay content [50,55,217]. Although the above equations were developed to model the diffusion of small molecules in conventional composites, they have also been used in reproducing experimental results for the relative permeability in PLS nanocomposites. Discrepancies between the experimental data and the theoretical line may be attributed either to inadequacies of the model or to incomplete orientation of the particles

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

within the nanocomposite film plane [50,162]. In fact, the key assumption of the Nielsen model is that the sheets are placed in an arrangement such that the direction of diffusion is normal to the direction of the sheets. Clearly, this arrangement results in the highest tortuosity, and any deviation from it would, in fact, lead to deterioration of the barrier properties [50,55]. Moreover, the tortuous path theory, including the Nielsen equation as well as other phenomenological relations (e.g. the Cussler [218] formula, the Barrel [219] formula and the power law equation [220]), is grounded on the assumption that the presence of nanoparticles does not affect the diffusivity of the polymer matrix. However, experimental observations demonstrate that molecular mobility in a polymer matrix, which is intimately connected to the mass transport properties, diminished by clay incorporation. This reduction should be accompanied by a decrease in diffusivity of small molecules, which is not considered in the concept of tortuous paths. Messersmith and Giannelis [118] studied the permeability of liquids and gases in nanocomposites and they observed that water permeability in PCL nanocomposites is dramatically reduced compared to the unfilled polymer. They also noted how the decrease in permeability is much more pronounced in the nanocomposites compared to conventionally filled polymers with much higher filler content. Liu and Wu [146], recorded the water absorption curves of PA66 and corresponding nanocomposites. They found that with increasing clay content, the water absorption at saturation decreases rapidly from 7.6% for PA66 to 5.2% for the nanocomposite containing 5 wt.% clay. They attributed this reduction to the presence of immobilized polymer in the amorphous phase. However, above 5 wt.% clay content, the decrease in the saturation amount of water is not so obvious, probably because of aggregation of silicate layers. Also, the diffusion coefficient values decrease greatly with increasing clay loading, but after 5 wt.% clay content, the amplitude of the decrease is obviously slower. In addition to the immobilized phase explanation, and the increased average diffusion path length, since an epoxy-co-intercalated clay was used in their study, the authors assumed that the epoxy groups between silicate layers have a strong interaction with amino and amide groups of the PA66 matrix, to some extent preventing them from forming hydrogen bonds with water. Significant reductions in diffusivity and maximum water uptake were also reported by Liu et al. [130] in epoxybased nanocomposites. Here, however, the decreased maximum water uptake was attributed to the low maximum water uptake of the nanoclays (ca. 2.8 wt.%) compared to the epoxy resin system (ca. 7.5 wt.%). Drozdov et al. [221] conducted moisture diffusion tests on vinylester resin-MMT clay nanocomposites and demonstrated that the clay content affects in a similar way the diffusion coefficient and the constants expressing the elastoplastic behavior, indicating that moisture diffusion and elastoplasticity may reflect the same phenomena at the microlevel, associated with molecular mobility of the polymeric matrix. Moreover, their experimental data demonstrated that moisture diffusion in the neat polymer resin is nearly Fickian, but is transformed to an anoma-

1171

Fig. 49. Equilibrium concentration of water vapor, Ceq (g/100 g), as a function of activity a = p/p0 for samples OMont ( ), NPU0 (䊉), NPU4 (), NPU20 ( ), NPU40 (♦) [204]. Reproduced from Tortora, Gorrasi, Vittoria, Galli, Ritrovati and Chiellini by permission of Elsevier Science Ltd., UK.

lous mode of transport of the penetrant molecules with an increase in clay concentration. The authors explained the anomalous moisture uptake by immobilization of water molecules on the surfaces of the hydrophilic MMT clay layers. In fact, they pointed out that, after a nanocomposite plate is immersed in water, three processes occur in the nanocomposite: (1) sorption of water molecules on the sample surfaces, (2) diffusion of water into the plate, and (3) adsorption of water molecules on the hydrophilic surfaces of clay layers, where these molecules become immobilized. Many studies reported in the literature have focused on nanocomposite barrier properties against gases and vapors. As an example, Tortora et al. [204] measured the transport properties of PU/o-MMT nanocomposites (prepared using a PCL nanocomposite “master-batch”) using water vapor as hydrophilic permeant and dichloromethane as hydrophobic one. For both vapors, the sorption behavior changed in the presence of the clay, as can be seen for example in Fig. 49, where the equilibrium concentration, Ceq (g/100 g), of water vapor is represented as a function of the vapor activity for all nanocomposites and for the oMMT. The sorption curve of water vapor for o-MMT follows the Langmuir sorption isotherm, in which the sorption of solvent molecules occurs at specific sites; therefore, when all the sites are saturated, a plateau is reached. On the other hand, the sorption of neat PU shows a linear dependence of equilibrium concentration on activity, while nanocomposites show a dual sorption shape, that is a downward concavity, an inflection point and an upward curvature. The prevailing mechanism in the first zone is the sorption of solvent molecules on specific sites, due to interacting groups. Tortora et al. inferred that this type of sorption is due to the presence of clay in the polymers. At higher activities, the plasticization of the polymeric matrix determines a more than linear increase of vapor concentration and a transition in the curve is observed, from a dual type to a Flory-Huggins behavior. From the calculated values of

1172

S. Pavlidou, C.D. Papaspyrides / Progress in Polymer Science 33 (2008) 1119–1198

Fig. 50. The O2 permeability of PET/o-MMT [222]. Reproduced from Ke and Yongping by permission of Elsevier Science Ltd., UK.

the sorption parameters, defined as: S = d(Ceq )/dp, and the zero-concentration diffusion coefficients for water sorption and dichloromethane vapor, the authors concluded that the sorption did not drastically change on increasing the clay content, whereas the zero-concentration diffusion coefficient D0 strongly decreased with increasing inorganic content. The permeability calculated as the product SD0 , was largely dominated by the diffusion parameter; it showed a remarkable decrease up to 20 wt.% of clay and a levelling off at higher contents. Ke and Yongping [222] tested the O2 permeability of intercalated PET nanocomposites. As demonstrated in Fig. 50, a small amount of clay effectively reduced the permeability of the PET film. When the content of o-MMT reached 3 wt.% the permeation of O2 was reduced to half that of the pure PET film. Further examples of barrier property improvement for PET nanocomposites designated for packaging applications are given in Table 32 [216]. Ogasawara et al. [223] reported on improved helium gas barrier properties of epoxy/MMT nanocomposites, compared to the pure resin. The estimated diffusivity, D, solubility, S, and permeability P are shown in Figs. 51–53, as functions of montmorillonite weight fraction. Dispersing MMT particles in the epoxy decreased the diffusion coefficient D. For example, the diffusion coefficient of the nanocomposite with 6 wt.% clay was approximately onetenth that of the base polymer. On the other hand, the solubility increased with montmorillonite dispersion and permeability remained almost constant due to the balance of diffusivity and solubility. On the other hand, Ray et al. [208] found that the O2 gas permeability of PLA nanocomposites with 4, 5 and 7 wt.%

Fig. 51. Effect of montmorillonite weight fraction on diffusivity D of montmorillonite/epoxy nanocomposites [223]. Reproduced from Ogasawara, Ishida, Ishikawa, Aoki and Ogura by permission of Elsevier Science Ltd., UK.

Fig. 52. Effect of montmorillonite weight fraction on solubility, S, of montmorillonite/epoxy nanocomposites. The numerical curve based on rule-of-mixture is superimposed [223]. Reproduced from Ogasawara, Ishida, Ishikawa, Aoki and Ogura by permission of Elsevier Science Ltd., UK.

Table 32 Relative barrier performance of newly developed rigid packaging based on PET. Container composition (supplier)

Relative oxygen transmission rate at 23 ◦ C 50% RH

PET PET nanocomposite (Tetrapak) PET/PA nanocomposite (Eastman)

1