Study of nanocomposites prepared from polyamides and biodegradable polyesters and poly(ester amide)s

ADVERTIMENT . La consulta d’aquesta tesi queda condicionada a l’acceptació de les següents condicions d'ús: La difusió d’aquesta tesi per mitjà del servei TDX (www.tesisenxarxa.net) ha estat autoritzada pels titulars dels drets de propietat intel·lectual únicament per a usos privats emmarcats en activitats d’investigació i docència. No s’autoritza la seva reproducció amb finalitats de lucre ni la seva difusió i posada a disposició des d’un lloc aliè al servei TDX. No s’autoritza la presentació del seu contingut en una finestra o marc aliè a TDX (framing). Aquesta reserva de drets afecta tant al resum de presentació de la tesi com als seus continguts. En la utilització o cita de parts de la tesi és obligat indicar el nom de la persona autora. ADVERTENCIA. La consulta de esta tesis queda condicionada a la aceptación de las siguientes condiciones de uso: La difusión de esta tesis por medio del servicio TDR (www.tesisenred.net) ha sido autorizada por los titulares de los derechos de propiedad intelectual únicamente para usos privados enmarcados en actividades de investigación y docencia. No se autoriza su reproducción con finalidades de lucro ni su difusión y puesta a disposición desde un sitio ajeno al servicio TDR. No se autoriza la presentación de su contenido en una ventana o marco ajeno a TDR (framing). Esta reserva de derechos afecta tanto al resumen de presentación de la tesis como a sus contenidos. En la utilización o cita de partes de la tesis es obligado indicar el nombre de la persona autora. WARNING. On having consulted this thesis you’re accepting the following use conditions: Spreading this thesis by the TDX (www.tesisenxarxa.net) service has been authorized by the titular of the intellectual property rights only for private uses placed in investigation and teaching activities. Reproduction with lucrative aims is not authorized neither its spreading and availability from a site foreign to the TDX service. Introducing its content in a window or frame foreign to the TDX service is not authorized (framing). This rights affect to the presentation summary of the thesis as well as to its contents. In the using or citation of parts of the thesis it’s obliged to indicate the name of the author “Study of Nanocomposites Prepared from Polyamides and Biodegradable Polyesters and Poly(ester amide)s”  Laura Teresa Morales Gámez Advisors: Dr. Jordi Puiggalí Bellalta Dra. Mª Lourdes Franco García DEPARTAMENT D’ENGINYERIA QUÍMICA UNIVERSITAT POLITÈCNICA DE CATALUNYA 2011 iii Once you make a decision, the universe conspires to make it happen Ralph Waldo Emerson iv v Abstract Polymer clay nanocomposites of polyamides and biodegradable polymers with three kinds of organomodified clays were prepared by different techniques (in situ polymerization, solution casting, and melt mixing). The polymers used in this research were nylons 56, 65 and 47 and the biodegradable polymers: poly (glycolic acid-alt-6-hydrohexanoic acid) and poly(glycolic acid-alt-6-aminohexanoic acid). The development of biodegradable nanocomposites with improved or modified material properties is an interesting topic since these new materials are expected to replace already existing biodegradable and non-biodegradable commodity plastics in some specific applications. This project aims to study the influence of clay particles incorporated in a polymer matrix on the crystallization processes, the study of the in situ polymerization kinetics of mixtures of clays and monomers of biodegradable polymers, as well as the influence of nanoparticles on the thermal behavior and morphologic parameters. Even-odd, and odd-even polyamides were chosen to study the Brill transition and to prepare nanocomposites with organomodified clays. These polyamides have a peculiar structure where hydrogen bonds are established along two different directions. X-ray diffraction as well as SAXS-WAXD synchrotron experiments were employed to study the structural changes induced by temperature, during heating and cooling. Different organomodified clays were used to prepare nanocomposites, which final structure was found to be dependent on the preparation method. Nanocomposites derived from biodegradable polymers were characterized by means of X-ray diffraction and transmission electron microscopy. Morphological studies showed that the extent of clay dispersion depended on the clay type and on the preparation technique. Hence, exfoliated and intercalated nanocomposites could be obtained. The final nanocomposite structure was found to have a great influence on both cold and hot crystallization processes. Hence, the crystallization rate increased and decreased with respect to the neat polymer when intercalated and exfoliated structures were respectively obtained. The kinetics of the polymerization process was also studied by means of FTIR and SAXS-WAXD. The results indicate that the presence of the organomodified clay had a remarkable effect on the kinetic parameters. vi vii Acknowledgements I would like to start by thanking the research group that received me at the Universitat Politècnica de Catalunya, especially Dr. Jordi Puiggalí for his support and guidance throughout these years, whose dedications have made this work possible and who has not just only helped me to complete my research work but also to overcome all kinds of problems that arise during this time. I would also like to express my gratitude to Dr. Lourdes Franco for her help, especially in the experimental part at the beginning of my research. To all the members of the research group professors and students, for all the support, help and friendly words including Drs. Teresa Casas, Alfonso Rodriguez, Luis Javier del Valle, Meritxel Martinez and to my PhD colleague Elena Diaz. I also want to thank our “neighbors”: MACROM research group for their help and friendship, particularly to Francisco Acosta and Drs. Juan Antonio Subirana, Lourdes Campos and Daniela de Luchi. I also want to acknowledge to the National Council for Science and Technology CONACYT, from whom I received the first funding for this PhD, for the excellent job they perform supporting young Mexicans that want to start a scientific career. I would also like to thank to the Leibniz Institute of Polymer Research in Dresden, for the opportunity to do a research stay, specially Dr. Manfred Stamm who accepted my application and Dr. Leonid Ionov for receiving me in his group, for all the help and all the kind people I met, I am thankful to all the members of the research group specially Georgi Stoychev for his time and explanations, and to my office colleagues Anja, Ksenia, Alexander and Falk for all the good moments and friendship. I am grateful to Dr. François Fauth, for all the help during the performance of synchrotron experiments that was always beyond the expected, for his quick and efficient answers to my e-mail question. Finally thanks most of all to my beloved Family, for their support and company no mattering the distance. To my mother and my brother Luis Fernando for the almost everyday conversations that kept me close to them all these years, and to my friends that either in Mexico, Dallas or even here in Spain always sent me their best wishes. You all made my time away from my homeland easier and better, I am especially thankful to my dearest friends Veronica and Alejandra and to Arturo for all the nice moments we spent together. viii xv 5.4 CRYSTALLIZATION BEHAVIOR OF CLAY NANOCOMPOSITES PREPARED FORM A DEGRADABLE ALTERNATING COPOLYESTER CONSTITUTED BY GLYCOLIC ACID AND 6-HYDROXYHEXANOIC ACID. .................................................. 221 5.4.1 Introduction ................................................................................................................................................... 222 5.4.2 Experimental section .................................................................................................................................. 223  Materials ............................................................................................................................................................... 223  Preparation of nanocomposite .................................................................................................................... 224  Measurements .................................................................................................................................................... 224 5.4.3 Results and discussion ................................................................................................................................ 226  Dispersion structure of the C25A clay in the composite with Poly(glc-alt-6HH) ..................... 226  Thermal stability ............................................................................................................................................... 228  Calorimetric data of the poly(glc-alt-6HH)/C25A nanocomposite ............................................... 229  Optical microscopy studies ............................................................................................................................ 231  Crystalline morphology and isothermal crystallization data of poly(glc-alt-6HH)/C25A from SAXS/WAXD data ......................................................................................................................................................... 235  Isothermal crystallization kinetics of poly(glc-alt-6HH) and its C25A nanocomposite from FTIR analyses ................................................................................................................................................................. 242 5.4.4 Conclusions ..................................................................................................................................................... 248 5.4.5 References ....................................................................................................................................................... 249 5.5 THERMAL STABILITY ON CLAY NANOCOMPOSITES PREPARED FROM A DEGRADABLE POLY(ESTER AMIDE) CONSTITUTED BY GLYCOLIC ACID AND 6-AMINOHEXANOIC ACID. ................................................................................. 251 5.5.1 Introduction ................................................................................................................................................... 252 5.5.2 Experimental section .................................................................................................................................. 253  Materials ............................................................................................................................................................... 253  Preparation of nanocomposite .................................................................................................................... 254  Measurements .................................................................................................................................................... 254 5.5.3 Results and discussion ................................................................................................................................ 254  Dispersion structure of the C25A clay in the composite with poly(glc-alt-amh) ..................... 254  Thermal stability of the poly(glc-alt-amh)/C25A nanocomposite ................................................ 255  Evaluation of the activation energy for the thermal degradation of the poly(glc-alt- amh)/C25A nanocomposite ..................................................................................................................................... 258  Thermal degradation mechanisms of the poly(glc-alt-amh)/C25A nanocomposite ............. 264  Invariant activation parameters for the thermal decomposition of the poly(glc-alt- amh)/C25A nanocomposite ..................................................................................................................................... 268  Modeling of degradation kinetics ............................................................................................................... 270 5.5.4 Conclusion ....................................................................................................................................................... 272 5.5.5 References ....................................................................................................................................................... 273 6 CONCLUSIONS ........................................................................................................................................... 275 Polyamides ...................................................................................................................................................................... 277  Structural transitions ...................................................................................................................................... 277 xvi  Spherulitic morphology .................................................................................................................................. 277  FTIR ........................................................................................................................................................................ 278  Nanocomposites ................................................................................................................................................. 278 Poly (glycolic acid-6-hydrohexanoic acid) ......................................................................................................... 279  Thermal analysis ............................................................................................................................................... 279  Crystallization .................................................................................................................................................... 279  Morphology .......................................................................................................................................................... 279 Poly(glycolic acid-alt-6-aminohexanoic acid) .................................................................................................. 280  Nanocomposite preparation and structure. ........................................................................................... 280  Polymerization kinetics .................................................................................................................................. 280  Crystallization .................................................................................................................................................... 280  Thermal Analysis ............................................................................................................................................... 281 APPENDIX A ........................................................................................................................................................ 283 ALIPHATIC POLYESTER AND POLY(ESTER AMIDE) CLAY NANOCOMPOSITES BY IN-SITU POLYMERIZATION .......................................................................................................................................................... 283 A.1. Introduction. Biodegradable polymers and their nanocomposites ................................................. 283 A.2. Aliphatic polyester clay nanocomposites by in-situ polymerization ............................................... 284  2.1. Poly(ε-caprolactone) based nanocomposites ................................................................................ 284  2.2. Polylactide based nanocomposites..................................................................................................... 291  2.3. Poly(butylen succinate) based nanocomposites ........................................................................... 296  2.4. Poly(p-dioxanone) based nanocomposites ..................................................................................... 298 A.3. Poly(ester amide)s clay nanocomposites by in-situ polymerization ............................................... 299 A.4. Conclusions ............................................................................................................................................................. 301 A.5. References ............................................................................................................................................................... 301 1 1 INTRODUCTION 2 3 1.1 Biodegradable Polymers Nowadays, biodegradable polymers are becoming an important research focus due to the increasing demand of biodegradable materials for novel biomedical technologies including tissue engineering, regenerative medicine, gene therapy, controlled drug delivery, biothecnology [1] and also for applications in which plastics are used for short time periods and then disposed [2]. These so called biodegradable polymers can be degraded by bioactive environments (such as those containing bacteria, fungi, or algae), or by hydrolysis in water or even in buffered solutions [3]. For this reason, biodegradable polymers can be used as attractive substitutes for many synthetic materials, thereby alleviating problems associated with solid waste disposal. Biodegradable polymers can be classified into tree major categories: (1) polyesters produced by microorganisms, (2) natural polysaccharides or other biopolymers, (3) synthetic polymers, particularly aliphatic polyesters. In this last category, industrially produced polymers can be found, such as poly(ε-caprolactone), poly(L-lactide), poly(butylene succinate), besides these aliphatic polyesters, various types of synthetic biodegradable polymers have been designated and tested for practical applications. For example, polyesters containing aromatic rings or cyclic ether moieties, poly(ester amide)s, poly(ester carbonate)s, poly(ester urethanes)s etc. Some attempts have been made to introduce ester groups into vinyl polymer chains, to make vinyl polymers biodegradable, but not effective and practical method have been developed yet [4]. Scheme 1.1.1 shows the classification of more important biodegradable polymers. 4 Biodegradable polymers can be defined as the polymers that can undergo microbial induced chain scission leading to mineralization. Some of them have comparable properties as the petroleum based polymers, and the clear advantage of being biodegradable. However, in general, it seems necessary to improve properties in order to increase both the number of disposable polymers and the range of applications. In this sense, efforts are nowadays focused to prepare nanocomposites by addition of different types of organomodified clays and other nanoparticles; moreover development of new biodegradable and bioabsorbable polymers with temporary function is also of great interest in the field of biomedical materials [5]. For these applications some specific properties are required. The polymer has not just only be biocompatible-nontoxic and not rejected by the intended organism but also it must have the ability to maintain mechanical properties for certain period of time, flexibility and an adequate Biodegradable Polymers Biomass products from agro resources Polysacarides Starches: Wheat, Potatoes, maize Lignocellulosic products: Wood, Straw Others: Pectine, Chitosan Proteins, Lipids Animals: Casein, Whey, Gelatine Plant: Soya, Zeln From microorganisms (obtained by extraction) PHA PHB, PHBV From biotechnology Polylactides PLA Conventional Synthesis (from synthetic monomers) PCL PEA PGA Aliphatic copolyesters Aromatic copolyesters PHA poly(hydroxyalkanoate)s PHB poly(hydroxybutyrate) PHBV poly(hydroxybutyrate co poly(hydroxyvalerate) PLA polylactic acid PCL polycaprolactone PEA poly(estar amide)s PGA poly(glycolic acid) Scheme 1.1.1 Biodegradable Polymers Classification [5]. 5 absorption rate. In this sense, the study of how the performance of these materials can be modified by incorporation of nanoparticles becomes again an interesting topic. Polyesters constitute the main family of biodegradable polymers due to their high degradation rate. In general, the derivatives of Poly(glycolic-acid) are the most commonly used for biomedical applications including drug delivery systems, wound treatment applications and implants [6]. Poly(ester amide)s constitute a new promising family of materials which has some advantages associated to the hydrophilic character of their amide groups and the capability to establish strong hydrogen bond interactions that influence on both thermal and mechanical properties. Furthermore, the presence of ester groups should ensure degradability, although in this case the hydrolysis proceeds at a lower rate than in parent polyesters which have a higher rate of hydrolyzable ester bonds. Our group has recently developed a synthesis procedure that allows to get polyesters and poly(ester amide)s constituted by glycolic acid units and -amino acid or -hydroxy acid units with a regular sequence distribution. This kind of polymers can be obtained by a classical methodology based on a selective protection of reactive groups. However in this case, the process is highly more complicated and have lower yields respect to the proposed one. The new synthesis is based on a thermal polycondensation reaction where the formation of a metal halide salt becomes the driving force of the process. The high simplicity of this method opens again the interest towards these families of polymers characterized by a semicrystalline character that contrasts with the irregular sequence distribution of commercial copolymers prepared by ring opening polymerization. In the present work, we have selected a representative polyester derived from glycolic acid and 6-hydroxyhexanoic acid, and a representative poly(ester amide) derived from glycolic acid and 6-aminohexanoic acid, which will be thereafter named as poly(glc-alt-6HH) and poly(glc-alt- amh), respectively. Their chemical repeat units are consequently similar since only differ in the substitution of an ester group by an amide group (Scheme 1.1.2). -[OCH2CO-O(CH2)5CO]- Poly(glc-alt-6HH) -[OCH2CO-NH(CH2)5CO]- Poly(glc-alt-amh) Scheme 1.1.2 polyester derived from glycolic acid and 6-hydroxyhexanoic acid, and a representative poly(ester amide) derived from glycolic acid and 6- aminohexanoic acid. 6 7 1.2 Nanocomposites Nanocomposites have emerged in the last two decades as an efficient strategy to upgrade the structural and functional properties of synthetic polymers. Aliphatic polyesters such as polylactide (PLA), polyglycolide (PGL) and poly (ε-caprolactone) (PCL) have attracted wide attention for their biodegradability and biocompatibility in the human body. The incorporation of nanofillers (organic and inorganic) into biodegradable polymers has been a consequence of the willing to prepare new biomaterials with enhanced properties [7]. Nanocomposites are hybrid materials consisting of a polymer matrix in which nano-sized particles are homogeneously dispersed. In fact, to be called nanocomposite at least one dimension of the added particle must be in the nanometer scale [8]. Due to their small size, dispersed structures in the polymer matrix, have relatively huge surface areas per unit weight, and often these surface areas dominate the behavior of these materials. Some important nanostructures include, carbon nanotubes, biomolecules such as proteins, silica nanoparticles and montmorillonite type clays [9]. Considering the nano-sized particles that can be added to the polymer matrix, phyllosilicates are of particular interest, especially montmorillonite due to their abundance, low cost and geometrical features [10]. The indicated clay is present typically at concentrations less than 5% [11-22]. The resulting interactions with the polymer matrix can improve substantially many physical properties such as mechanical performance [23], barrier resistance [24] and flammability [25]. In the case of biodegradable polymers, biodegradability could indeed be improved [26] . The natural montmorillonite clays consist of several hundred individual plate-like particles of dimensions 1m 1m 1nm, held together by electrostatic forces with a gap of approximately 8 0.3 nm between two adjacent particles. Figure 1.2.1 shows the structure of this clay at atomic level. In a such layered material, the bonds between the atoms in the layer are very strong, but the bonds between the layers become weaker. This feature allows an easy separation of the constitutive layers. This clay consists of three subunits: an octahedral center layer consisting of aluminum cations (Al+3); two tetrahedral layers consisting mainly of silica (Si) and oxygen (O) atoms. In the octahedral layer some of the Al3+ cations are substituted by Mg2+ cations which gives rise to a net negative charge on the layer. Similarly, some of the Si4+ cations may be substituted by Al3+ cations resulting again in a net negative charge in the tetrahedral layer. In natural clays the charge balancing cations in the gap between the silicate layers are mainly Na+, K+ and Ca+2 [27]. The gap between the silicate leyers is widely known as a gallery or an interlayer. Figure 1.2.1. Typical clay layer structure: octahedral center layer consisting of aluminium (Al+3) cations; two tetrahedral layers consisting mainly of silica (Si) and oxygen (O) atoms [28]. Natural clays mixed with polymers lead to the formation of nanocomposites very rarely. Homogeneous dispersion in the organic polymer phase is hindered by the hydrophilic nature of the clay. Figure 1.2.2 shows a general procedure to increase compatibility: the cations present 15  Adjacent re-entry chain-folded models (regular folding) Two possibilities have been considered: The smooth surface model which is a very idealized visualization of the chain folding process and tries to be consistent with the highly ordered molecular arrangement expected for a crystal. The rough surface model the reentry of the chain is still in the nearest growth plane, though large variations in the fold length may exist on a local scale. Multiple nucleation and chain-end defects will further contribute to a rough surface [42] [43]. Figure 1.3.4 (a) Adjacent reentry model with smooth, regular chain folds b) Adjacent reentry model with rough fold surface [40] .  Solidification model This model explain the constancy of the radios of gyration in crystalline state, as detected by small angle neutron scattering [37]. The model is visualized in terms of an alignment of chains without a long-range diffusion process to give rise to a lamellar morphology. The chain sequences in proper conformations are incorporated into the crystal without significant reorganization of the chain conformation. Figure 1.3.5 Solidification model of crystallization process, showing how a chain can be incorporated into a lamellar structure without significant change of overall shape [43] (b) (a) 16  Isothermal Crystallization Crystallization in polymers can be described by the process of nucleation and crystal growth. Primary nucleation can be heterogeneous, when the nucleation sites are foreign substances (dust, impurities, nucleating agents, residual catalyst and any existing surfaces). Homogeneous nucleation involves the aggregation of polymer chains, chain segments of parent material to certain size and order. Nucleation is a time-dependent process and takes place as a rate even under isothermal conditions [38]. Thermodynamically, crystallization will be favored if the entropy penalty is outweighed by the enthalpy change. Crystallization is an exothermic process whereas melting is endothermic as energy is required to overcome the intermolecular interactions established in the crystal. Tm is usually higher than Tc due to the high viscosity of molten polymers. Crystallization is therefore determined by kinetics as well as thermodynamics. The crystal growth process is explained by two theories: surface nucleation theory by Hoffman and Lauritzen [44-46] and the surface roughing theory by Sandler and Gimler [47-49]  Surface Nucleation This theory and its modifications is the most widely used methodology to interpret and model the crystallization behavior of a large number of polymers. It describes the process of crystal growth by surface nucleation events: primary, secondary and tertiary nucleation. Primary nucleation can be seen as a process of crystal formation of six new surfaces, secondary and tertiary nucleation events, involve the formation of four and two new surfaces, respectively, see Scheme 1.3.1. The model describes the crystal growth process as a combination of secondary nucleation rate “i” and the substrate completion rate “g” . In the classical secondary nucleation theory, the nucleation rate is taken as a rate determining step for crystal growth where “g” is very fast compared to “i”[50]. Scheme 1.3.1 Primary , secondary and tertiary nuecleation events [51]. 17 In the Hoffman and Lauritzen theory for polymers the crystal growth process is described via three regimes so far depending on the relative values of the nucleation (i) and the substrate completion (g) rates [42]. Figure 1.3.6 shows schematically three regimes of crystal growth. Regime I is from classical secondary nucleation theory where the rate of spreading (g) is much faster than rate of secondary nucleation (i). In regime II, both the rates are comparable and in regime III, the rate of secondary nucleation (i) is very high compared to rate of spreading (g). This model also assumes the single stem nucleation, i.e. deposition of a single stem on a surface (a primary nucleus) starts the crystal growth process. Figure 1.3.6 a) Schematic for crystal growth regimes for polymers according to Lauritzen and Hofmman, i is the rate of secondary nucleation and g is the substrate completion rate. b) Schematic showing the dependence of crystal growth rate with temperature for three regimes [42]  Surface Roughening Theory Sandler and Gilmer developed a theory that is based in roughening at atomic length scales [47- 49], due to the observation of curved surfaces for the solution grown single crystals, this is different from the Hoffman and Lauritzen assumption of the presence of a flat grown facet. The theory is able to predict regime transition as well as the curved grown surfaces of crystals. A secondary nucleation step is not required according to this theory, nevertheless surface roughening can lead to secondary nucleation. 18 19 1.4 Polyamides Aliphatic polyamides also called nylons are important industrial materials, valued for their good physical properties and processability. They belong to the wide family of synthetic polymer materials containing amide linkages in their backbones [52]. Nylons are used both, as plastics and as fibres. These polymers generally exhibit high impact strength, toughness, good flexibility, and abrasion resistance [53]. Phenols, cresols and formic acid dissolve the polyamides at room temperature.  Nomenclature Nylons can be synthesized by ring opening polymerization of lactams, by condensation of - amino acids, and also by condensation of diamines and dicarboxylic acids. The polymers formed by the two former methods, are called nylon n, where n is the number of carbon atoms in the repeating unit, (e.g polycaprolactam is nylon 6). Nylons from diamines and dibasic acids are designated as nylons m n where m represents the number of carbon atoms in the diamine and n is the number of carbon atoms in the dicarboxylic acid [53], (e.g. poly(hexamethylene adipamide) is named nylon 6 6)  Structure Crystallinity and orientation are the most important features of polyamides affecting the final properties. One determining factor for the their structure is the ability of the NH group to form strong hydrogen bonds (H-bonds) with the CO group [54]. Thus, molecular chains must be oriented in such a way that hydrogen bonding becomes maximized. Intermolecular H-bonds connect neighboring chains or chain segments and form extended planar sheets that contain these H-bonds. Formation of extended sheets usually characterizes the structure of aliphatic 20 polyamides, and depends on the directionality of the molecular chain and the parity (odd/even) of the involved monomers [54]. In nylons n, all the amide groups lie in the same direction whereas in nylons m n two consecutive amide groups are in opposite directions. Thus, in the former case, packing is established between directional molecular chains whereas in the second case implies nondirectional chains. Carbonyl as well as amine groups can point out to the same side of the molecular chain or in opposite sides depending on the number of carbon atoms (even or odd) of the repeat unit, which clearly influences the hydrogen bonding geometry. The more energetically favorable structure should be obtained when NH and C=O groups of neighboring chains face each other, allowing to attain an ideal H-bond geometry. Nylons are divided in two main types of stable crystal structures, the α and γ structures [54]. The most important features related to the α structure are the formation of planar sheets of hydrogen bonded molecules with a fully extended (planar zig-zag) conformation. These sheets are stacked upon one another giving rise to the three-dimensional unit cell arrangement. Thus, nylon 66 was characterized by triclinic unit cell and a P space group defined by center of symmetry in both the diamine and diacid moieties (Figure 1.4.1). Figure 1.4.1 a) Packing of nylon 66 molecules in the triclinic unit cell [55], b) Various stacking schemes of H-bonded sheets [56]. The γ – form corresponds to a pseudohexagonal arrangement, which is favoured when the amide groups are tilted ca. 60º off the sheet plane. As a consequence, a characteristic shortening in the chain axis repeat is noticed when compared with the values of the extended conformation. a) b) 21 However, hydrogen bonds remain in a single direction. The γ -structure is considered to be a less ordered phase than the α -form. It is characteristic of nylons with a high methylene content in their chemical repeat units (nylons 11 or 12) or nylons for which linear hydrogen bonds between adjacent chains cannot be established when an extended conformation is considered. Figure 1.4.2 Structures of the α and γ forms of nylon 6 and nylon 6 6. The left side shows the view of the hydrogen-bonding planes, and the right side shows the view down the chain axis. For the α form of nylon 6, the adjacent chains are antiparallel and the hydrogen bonding is between adjacent chains within the same sheet (bisecting the CH2 angles). For the γ form of nylon 6, the chains are parallel and the hydrogen-bonding is between chains in adjacent sheets. In nylon 6 6, the chains have no directionality [57] 22  Brill Transition The Brill Transition was reported in 1942 by R. Brill, who first observed from X-ray diffraction patterns, that as nylon 6 6 crystals were heated, the two characteristic reflections move together and meet [58][59] (see figure Figure 1.4.3 b and c). This behavior is due to the structural changes in the lattice parameters during heating, in which basically the two basic equatorial spacings characteristic of the α-form (i.e. those corresponding to intrasheet and intersheet spacings at 0.440 and 0.380 nm, respectively), merge into a single one indicative of a pseudohexagonal modification, which is presumably related to a γ-form [60]. This feature is usually observed in even-even nylons with the logical variations caused by the differences on melting points and density of amide groups along the polymer chain. The structural change is often called “transition”, but it is not certain that it has is a themodynamic explanation. Brill transition occurs in a wide temperature range and is not detected in the calorimetric scans. Usually, is assigned as the lowest temperature for which the spacings of the two characteristic equatorial reflections are identical, above the Brill temperature the single spacing increases slightly as a consequence of thermal expansion. Therefore, the crystal structure at equilibrium below this temperature is triclinic and pseudohexagonal above it. Although different theories have been postulated to explain the Brill transition, nowadays it is assumed that it is only the consequence of the conformational motion of methylene groups (i.e. amide interactions remain unaltered) due to the temperature increase that gives rise to a packing change within the crystal [61], The pseudohexagonal structure is actually a special class of triclinic structures in which the projection on a plane normal to the chain axis is metrically hexagonal (see Figure 1.4.3 a) The Brill transition is most clearly displayed in X-ray diffration studies, as the two strongest reflections of for example nylon 6 6, the 100 and 010/110 reflections, merge into a single reflection at the transition, as it si shown in Figure 1.4.3 b and c. Other techniques such as DSC are less sensitive, and in general do not show the Brill transition by way of a distinct endothermic peak [61]. Some explanations for this phenomenon have been given [60]: a) anisotropy of the thermal expansion, b) the development of a tree dimensional network of hydrogen bonds between the chains induced by rotational molecular jumps of 60º at elevated temperatures, c) a transition involving a greater mobility of the methylene groups, while hydrogen bonds remain arranged in a single direction. A complete understanding of the Brill transition does not yet exist, some features need to be further explained like, the transition temperature to pseudohexagonal phase is not constant for a specific polyamide since it clearly depends on the thermal history of the sample. A pseudohexagonal phase can also be observed for some even nylons in quenched samples, and it has been interpreted as a frozen state arising from the high temperature modification, however , such state is not stable since it reverts quickly to the α-modification when crystals are heated 23 above their Tg [60], in the present research work some studies were performed in order to understand better this phenomenon. Figure 1.4.3 a) Example of pseudohexagonal structure [62] b) Variation of d100 and d010/110 spacings with temperature on heating nylon 66 from room temperature to melting [61] c) Thee-dimensional view of the X-ray diffraction patterns of nylon 66 on heating from room temperature to melting [61]. 1.4.1 References [1] L. S. Nair and C. T. Laurencin, “Biodegradable polymers as biomaterials,” Progress in Polymer Science, vol. 32, no. 8-9, pp. 762-798, Aug. 2007. [2] R. a Gross and B. Kalra, “Biodegradable polymers for the environment.,” Science (New York, N.Y.), vol. 297, no. 5582, pp. 803-7, Aug. 2002. [3] R. J. Müller, “Biodegradability of polymers: regulations and methods for testing,” in Biopolymers Online, Wiley Online Library, 2003, pp. 365-374. [4] M. Okada, “Chemical syntheses of biodegradable polymers,” Synthesis, vol. 27, pp. 87-133, 2002. [5] M. Vera, L. Franco, and J. Puiggalí, “Synthesis and Characterization of Poly(glycolic acid-alt-6- aminohexanoic acid) and Poly(glycolic acid-alt-11-aminoundecanoic acid),” Macromolecular Chemistry and Physics, vol. 205, no. 13, pp. 1782-1792, Aug. 2004. [6] E. S. Stevens, Green plastics: an introduction to the new science of biodegradable plastics. Princeton University Press, 2002. 24 [7] I. Armentano, M. Dottori, E. Fortunati, S. Mattioli, and J. M. Kenny, “Biodegradable polymer matrix nanocomposites for tissue engineering: A review,” Polymer Degradation and Stability, vol. 95, no. 11, pp. 2126-2146, Jun. 2010. [8] M. Alexandre, “Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials,” Materials Science and Engineering: R: Reports, vol. 28, no. 1-2, pp. 1-63, Jun. 2000. [9] L. H. Sperling, Introduction to Physical Polymer Science. Hoboken, NJ, USA: John Wiley & Sons, Inc., 2005. [10] S. Marras, I. Zuburtikudis, and C. Panayiotou, “Nanostructure vs. microstructure: Morphological and thermomechanical characterization of poly(l-lactic acid)/layered silicate hybrids,” European Polymer Journal, vol. 43, no. 6, pp. 2191-2206, Jun. 2007. [11] H. Yang et al., “Largely improved toughness of PP/EPDM blends by adding nano-SiO2 particles,” Polymer, vol. 48, no. 3, pp. 860-869, Jan. 2007. [12] A. Vermogen, K. Masenelli-Varlot, R. Séguéla, J. Duchet-Rumeau, S. Boucard, and P. Prele, “Evaluation of the Structure and Dispersion in Polymer-Layered Silicate Nanocomposites,” Macromolecules, vol. 38, no. 23, pp. 9661-9669, Nov. 2005. [13] M. Zanetti, S. Lomakin, and G. Camino, “Polymer layered silicate nanocomposites,” Most, vol. 9, pp. 1-9, 2000. [14] B. Lepoittevin, N. Pantoustier, M. Alexandre, C. Calberg, R. J??r??me, and P. Dubois, “Polyester layered silicate nanohybrids by controlled grafting polymerization,” Journal of Materials Chemistry, vol. 12, no. 12, pp. 3528-3532, Nov. 2002. [15] P. Bordes, E. Pollet, and L. Averous, “Nano-biocomposites: Biodegradable polyester/nanoclay systems,” Progress in Polymer Science, vol. 34, no. 2, pp. 125-155, Feb. 2009. [16] S. Y. Hwang, E. S. Yoo, and S. S. Im, “Effect of the urethane group on treated clay surfaces for high-performance poly(butylene succinate)/montmorillonite nanocomposites,” Polymer Degradation and Stability, vol. 94, no. 12, pp. 2163-2169, Dec. 2009. [17] D. Lincoln, R. Vaia, Z. G. Wang, and B. Hsiao, “Secondary structure and elevated temperature crystallite morphology of nylon-6/layered silicate nanocomposites,” Polymer, vol. 42, no. 4, pp. 1621–1631, 2001. [18] J. H. Chang, B. S. Seo, and D. H. Hwang, “An exfoliation of organoclay in thermotropic liquid crystalline polyester nanocomposites,” Polymer, vol. 43, no. 10, pp. 2969–2974, 2002. [19] R. K. Shah and D. R. Paul, “Organoclay degradation in melt processed polyethylene nanocomposites,” Polymer, vol. 47, no. 11, pp. 4075-4084, May. 2006. [20] C. J. G. Plummer, L. Garamszegi, Y. Leterrier, M. Rodlert, and J.-A. E. Månson, “Hyperbranched Polymer Layered Silicate Nanocomposites,” Chemistry of Materials, vol. 14, no. 2, pp. 486-488, Feb. 2002. [21] M. Shibata, Y. Someya, M. Orihara, and M. Miyoshi, “Thermal and mechanical properties of plasticized poly(L-lactide) nanocomposites with organo-modified montmorillonites,” Journal of Applied Polymer Science, vol. 99, no. 5, pp. 2594-2602, Mar. 2006. [22] E. Pollet, C. Delcourt, M. Alexandre, and P. Dubois, “Transesterification catalysts to improve clay exfoliation in synthetic biodegradable polyester nanocomposites,” European Polymer Journal, vol. 42, no. 6, pp. 1330-1341, 2006. [23] J. W. Cho and D. R. Paul, “Nylon 6 nanocomposites by melt compounding,” Polymer, vol. 42, no. 3, pp. 1083-1094, Feb. 2001. [24] C. Lu and Y.-W. Mai, “Influence of Aspect Ratio on Barrier Properties of Polymer-Clay Nanocomposites,” Physical Review Letters, vol. 95, no. 8, pp. 1-4, Aug. 2005. [25] J. W. Gilman et al., “Flammability Properties of Polymer−Layered-Silicate Nanocomposites. Polypropylene and Polystyrene Nanocomposites †,” Chemistry of Materials, vol. 12, no. 7, pp. 1866-1873, Jul. 2000. 3 3 EXPERIMENTAL 32 33 3.1 Characterization Techniques 3.1.1 Transmission Electron Microscopy (TEM)  Method In this technique an electron beam is passed through a very thin section of the sample and an image is obtained due to the differences in electron density of the materials. Since there is sufficient difference in electron density between the polymer and the clay to provide a contrast between the two materials it is possible to see the clay dispersion. TEM was carried out with a Philips TECNAI 10 at an accelerating voltage of 100 kV. The specimens were prepared by embedding in a low viscosity modified Spurr epoxy resin and curing them at 40 ºC for a few days and then at 60 ºC for some hours. Ultrathin sections (less than 100nm) were cut at room temperature using Sorvall Porter-Blum microtome equipped with a diamond knife. Finally, the sections were collected in a though filled with water and lifted onto carbon coated copper grids. To prevent diffusion of the epoxy resin into the polymer film, a thin layer of carbon was evaporated over the film surface. 3.1.2 X-ray Scattering  Experimental Technique X-rays are electromagnetic waves of very short wavelength; The X-rays used in polymer characterization have wavelengths of about 0.1-0.2 nm. Two types of X-ray scattering are used in the study of polymers, wide-angle X-ray diffraction (WAXD) and small-angle X-ray scattering (SAXS) depending on scale of the features studied. SAXS studies are performed on polymers for the investigation on structures on a much larger scale than the separations of crystal planes, which implies scattering angles much smaller than those used in WAXD. 34 Scattering from structures of any size takes place at well-defined angles. The scattering is usually called diffraction only when the structures are periodic. The most important periodic structures suitable for WAXD investigations are crystals, which are periodic in three dimensions [43] . X-rays are produced by bombarding a metal target with a beam or high voltage electrons. This is done inside a vacuum tube. The target metal as well as the applied voltage determines the wavelength of X-rays produced. The diffracted X-rays may be detected by their action on photographic films or plates, or by means of radiation counter and electronic equipment feeding data to a computer [63]. It is possible to obtain X-ray reflections from a series of planes inside the crystal. The orientation and interplanar spacings of these planes are defined by the three integers h, k, l called Miller indices of a plane or a face [64]. A given set of planes with indices h, k, l cut the a-axis of the unit cell in h sections, the b axis in k sections and the c axis in l sections. A zero indicates that the planes are parallel to the corresponding axis e. g. Figure 3.1.1 shows the 2 0 0 planes which cut the a axis in half but are parallel to b and c axes. Figure 3.1.1 Example of three dimensional diffraction, where three indices hkl become the order of diffraction along the unit cell axes a, b and c respectively. Diffraction can easily be understood in terms of the reflection of the incident beam by the different crystallographic planes. Thus, the intensity of rays reflected by a pair of planes with an interplanar spacing d (Figure 3.1.2) is maximum when the waves are in phase. Equation 3.1 corresponds to the Bragg’s law, which relates the angle of the incident beam, the interplanar spacing and the wavelength of the radiation. The geometric derivation is shown in Figure 3.1.2. (3.1)  sin2dn  35 Figure 3.1.2 Geometric derivation of Bragg's law: Constructive interference occurs when the delay between waves scattered from adjacent lattice planes given by 1 + 1' and 2 + 2’ is an integer multiple of the wavelenght  [64]. The process of reflection is described here in terms of incident and diffracted rays, each making an angle  with a fixed crystal plane. Reflections occur from planes set at angle  with respect to the incident beam and generates a reflected beam at an angle 2  from incident beam. The possible d spacing defined by the indices hkl (dhkl) are determined by the shape of the unit cell. Rewriting Bragg’s law: (3.2)  sin2 hkl d The spacing dhkl is easily calculated for a given measured value of  and with a set of experimental spacings is possible to determine the dimensions of the unit cell (a, b, c, α, β and γ) through the indexing process. For example, the following equation relates indices, cell parameters and spacings for an orthorhombic unit cell (α, β and γ are 90º): √ Figure 3.1.3 shows some important planes for the special case of a lattice with a rectangular projection on a plane perpendicular to the c-axis. Integers hkl label the points of intersection of three sets of equally spaced parallel planes. These planes can be chosen so that this new lattice (called reciprocal lattice) has the next property: for all values of h, k and l the line joining the origin of the reciprocal lattice to the point hkl is of length 1/ dhkl and is normal to the hkl planes of the real lattice. The reciprocal-lattice plane for a given value of l and all values of h and k is perpendicular to the c-axis and is distant l/c from the origin of the reciprocal lattice. 36 Figure 3.1.3 Planes for a lattice with a rectangular projection on a plane perpendicular to the c-axis [43]. However, the intensities of the reflections are determined by the distribution of the electrons in the cell. The highest electron density are found around atoms. Therefore, the intensities depend on what kind of atoms are present and where in the unit cell they are located. Planes going through areas with high electron density will reflect strongly, planes with low electron density will give weak intensities.  Method In X-ray polymer diffraction it is normally distinguish among single crystal, polycrystalline or powder applications and fiber pattern diffraction. The single crystal sample is a perfect (all unit cells aligned in a perfect extended pattern) crystal with a cross section of about 0.3 mm. The single crystal diffractometer and associated computer package is used mainly to elucidate the molecular structure of novel compounds, either natural products or synthetic molecules [65]. However this kind of samples are very difficult to obtain for polymers. Powder diffraction is mainly used for “finger print identification” of various solid materials, e.g. asbestos, quartz. In powder or polycrystalline diffraction it is important to have a sample with a smooth plane surface. The ideal sample has a random distribution of all possible hkl planes. Only crystallites having reflecting planes (hkl) parallel to the specimen surface will contribute to the reflected intensities. If a sample is truly random, each possible reflection from a given set of h, k, l planes will have an equal number of crystallites contributing to it. The specimen must be rocked through the glancing angle  in order to produce all possible reflections. Fiber diffraction patterns are good option to analyze structure of polymer crystallites since a fiber pattern contains information about the crystal structure of the polymer. It also contains information about the size of the crystallites and about their degree of alignment [66]. Crystalline polymers fibers must be stretched to be oriented in order to obtain a fiber diffraction 37 pattern. Thus highly oriented fiber consists of a very large number of crystallites, which all have one particular crystallographic direction oriented almost parallel to the fiber axis (usually the chain axis, the c-axis) and the remaining directions are oriented randomly around this direction. Assuming that the axis of the fiber is normal to the incident X-ray beam, the scattering expected is therefore almost exactly the same as that which would be observed from a single crystal with its c-axis parallel to the fiber axis if this crystal were rotated continuously around the c-axis during the exposure of the X-rays. Highly oriented polymer fiber diffraction pattern shows the same features that a rotating crystal, when planes are parallel to fiber axis: only four diffraction spots lying at particular points on the imaginary circle of the corresponding powder-pattern would be observed. These points are symmetrically placed with respect to the plane that contains the incident X-ray beam and normal to the rotation axis at it is shown in Figure 3.1.4. The radius of the (imaginary) powder circle and the positions of the four spots on it for a particular type of crystal plane depend on the indices of the planes. Figure 3.1.4 (a) and (b) show the production of four diffraction spots corresponding to a given set of planes for a rotation pattern: (a) the starting position, where 0 is greater than the Bragg angle; and (b) the location of the four diffraction spots corresponding to a given set of planes. c) is a schematic diagram showing the relationship among layer lines, powder rings and diffraction spots in a fiber diagram. For simplicity the layer lines are shown straight and the powder rings as circles [43]. Powder ring corresponds to a particular set of values h, k and l and it follows that diffraction spots can be seen only at those places where the lth layer line crosses the position where a powder circle corresponding to the same value of l would have been seen, as shown 38 schematically in Figure 3.1.4 c. Each circle and pair of layer lines (i.e. for ±l) gives rise to the four spots previously shown to arise from any particular set of planes. Because there can be various sets of planes with different values of h and k but the same value of l there will be several pairs of spots on each layer line. The layer line for l = 0 is called the equator and the normal to this through the point where the incident X-ray beam would strike the film is called the meridian. Figure 3.1.5 shows, as an example, the X-ray scattering pattern obtained for an oriented fiber of syndiotactic propylene. Figure 3.1.5 Fibre pattern from oriented syndiotactic polypropylene, drawn to a draw ratio of about 5 at 109ºC [43]. X-ray diffraction (XRD) is useful to characterize the morphology of the polymer nanocomposites as it enables the average basal spacing (distance between two clay platelets) to be calculated. This spacing is often referred as the d001 spacing where d refers to the spacing between the planes in a lattice and 001 refers to the indics of the involved reflection. An increase in the spacing indicates an increase in the extent of intercalation and the point where the XRD peak can no longer be observed. The clay is through to become fully exfoliated within the polymer matrix, since in an exfoliated nanocomposite the clay platelets will be at larger distances from each other with random orientation. This means that there will be no average distance between the platelets and therefore no XRD peak will be observed. Figure 3.1.6 shows an example of intercalated nanocomposite, and an exfoliated one. The XRD patterns that have to be corroborated with TEM analysis (Figure 3.1.6 b) to draw a conclusion about nanocomposite structure. 39 Figure 3.1.6 (a) WAXD patterns and (b) TEM images of three different types of nanocomposites . 3.1.3 Synchrotron Radiation  Experimental Technique Synchrotron radiation is the electromagnetic radiation emitted by high-speed electrons spiraling along the lines of force of a magnetic field. Depending on the electron’s energy and the strength of the magnetic field, the maximum intensity will occur as radio waves, visible light or X-rays, the radiation is highly polarized and the intensity greatly exceed other sources (Figure 3.1.7 shows the large spectral region covered by the synchrotron radiation). These properties make synchrotron radiation a recognized and powerful research tool for all scientific areas. In this research work simultaneous small and wide angle X-ray diffraction (SAXS-WAXD) techniques were used. These techniques allow to investigate at different length scales the structure and dynamics of the material of interest. Simultaneous SAXS/WAXD permits studding structural and morphological changes in real time. During the experiment, two position-sensitive detectors are placed in different locations covering a wide angular range, of about four orders of magnitude of scattering angle. The scheme of the experimental setup for 40 simultaneous SAXS-WAXD it is shown in Figure 3.1.8. The WAXD provides information about the molecular and atomic ordering of materials, while SAXS is sensitive to heterogeneities in the electron density on a larger scale (1-102 nm). Figure 3.1.7 Energy and wavelength scales for a very large range of the electromagnetic wave field. Showing the large spectral region covered by the synchrotron radiation [67] . Figure 3.1.8 Scheme of the experimental setup for simultaneous SAXS-WAXS experiments at the BM16 beam line. The main parts of the goniometer are rotation unit (1), arm (2), counterweight (3), and hot stage (4). The vacuum chamber is made of cylindrical parts of different lengths and a square based truncated pyramid at the front [68].  Method The collected data was then treated with different scientific software. For WAXD experiments, deconvolution was performed and the overall crystallinity of the samples as well as morphology were evaluated. On the other hand for SAXS experiments the correlation function analysis was 47 3.1.6 Fourier Transform Infrared Spectroscopy (FTIR)  Experimental Technique Infrared spectroscopy is a technique based on the vibration of the atoms of a molecule. An infrared spectrum is obtained by passing infrared radiation though a sample and determining what fraction of the incident radiation is absorbed at a particular energy. This energy of absorption corresponds to the frequency of a vibration of a part of a sample molecule [81]. For a molecule to show infrared absorption it must possess a specific feature, an electric dipole moment of the molecule must change during the vibration. A molecule containing N atoms has 3N normal vibration modes, including rotational and transitional motions of the entire molecule. In polymers the infrared absorption spectrum are of the very simple due to the occurrence of normal vibrations at almost same frequency. The main type of molecular vibrations are stretching (symmetrical and asymmetrical) and bending (scissoring, wagging, twisting and rocking). As different kinds of bonds, and thus different functional groups absorbed infrared radiation of different wave length, analysis of absorption reveals details about the molecular structure of the sample.  Method This technique was used to perform crystallization and polymerization dynamic studies, identifying functional groups that appear while temperature was varied, the methodology developed to treat the data obtained during this experiments allowed to evaluate polymerization and crystallization kinetics. 3.1.7 Nuclear Magnetic Resonance Spectroscopy (NMR)  Experimental Technique Nuclear Magnetic Resonance (NMR) spectroscopy can be used to study chain configuration, sequence distribution, and microstructure in polymers. It utilizes the property of spin (angular momentum and its associated magnetic moment) possessed by nuclei whose atomic number and mass number are not both even, such as isotopes of hydrogen and 13C, 15N, 17O and 19F. When a strong magnetic field is applied to the material containing such nuclei, the energy level splits into two, representing states with spin parallel and antiparallel to the field. Transitions between the states lead to absorption or emission of an energy [63] The most common nuclei examined by NMR 1H and 13C, since these are the most abundant NMR sensitive nuclei [71]. The resonant frequencies can be used to determine molecular structures. 1H resonances are fairly specific for the types of carbon they are attached to, these 48 resonances may be split into multiples. The magnitude of splittings, and the multiplicity, can be used to better determine the chemical structure in the vicinity of hydrogen. Since only hydrogen is observed, any feature in the molecule without attached hydrogen can only be inferred, and this turns out impossible to resolve complex structures or molecules. 13C resonance can be used to determine skeleton of an organic molecule. NMR is a very powerful tool, to characterize compound structure, and may provide a general characterization by functional groups. In this research work NMR was often used to verify purity and non – degradation of sample material. 49 3.1.8 References [1] L. S. Nair and C. T. Laurencin, “Biodegradable polymers as biomaterials,” Progress in Polymer Science, vol. 32, no. 8-9, pp. 762-798, Aug. 2007. [2] R. a Gross and B. Kalra, “Biodegradable polymers for the environment.,” Science (New York, N.Y.), vol. 297, no. 5582, pp. 803-7, Aug. 2002. [3] R. J. Müller, “Biodegradability of polymers: regulations and methods for testing,” in Biopolymers Online, Wiley Online Library, 2003, pp. 365-374. [4] M. Okada, “Chemical syntheses of biodegradable polymers,” Synthesis, vol. 27, pp. 87-133, 2002. [5] M. Vera, L. Franco, and J. Puiggalí, “Synthesis and Characterization of Poly(glycolic acid-alt-6- aminohexanoic acid) and Poly(glycolic acid-alt-11-aminoundecanoic acid),” Macromolecular Chemistry and Physics, vol. 205, no. 13, pp. 1782-1792, Aug. 2004. [6] E. S. Stevens, Green plastics: an introduction to the new science of biodegradable plastics. Princeton University Press, 2002. [7] I. Armentano, M. Dottori, E. Fortunati, S. Mattioli, and J. M. Kenny, “Biodegradable polymer matrix nanocomposites for tissue engineering: A review,” Polymer Degradation and Stability, vol. 95, no. 11, pp. 2126-2146, Jun. 2010. [8] M. Alexandre, “Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials,” Materials Science and Engineering: R: Reports, vol. 28, no. 1-2, pp. 1-63, Jun. 2000. [9] L. H. Sperling, Introduction to Physical Polymer Science. Hoboken, NJ, USA: John Wiley & Sons, Inc., 2005. [10] S. Marras, I. Zuburtikudis, and C. Panayiotou, “Nanostructure vs. microstructure: Morphological and thermomechanical characterization of poly(l-lactic acid)/layered silicate hybrids,” European Polymer Journal, vol. 43, no. 6, pp. 2191-2206, Jun. 2007. [11] H. Yang et al., “Largely improved toughness of PP/EPDM blends by adding nano-SiO2 particles,” Polymer, vol. 48, no. 3, pp. 860-869, Jan. 2007. [12] A. Vermogen, K. Masenelli-Varlot, R. Séguéla, J. Duchet-Rumeau, S. Boucard, and P. Prele, “Evaluation of the Structure and Dispersion in Polymer-Layered Silicate Nanocomposites,” Macromolecules, vol. 38, no. 23, pp. 9661-9669, Nov. 2005. [13] M. Zanetti, S. Lomakin, and G. Camino, “Polymer layered silicate nanocomposites,” Most, vol. 9, pp. 1-9, 2000. [14] B. Lepoittevin, N. Pantoustier, M. Alexandre, C. Calberg, R. J??r??me, and P. Dubois, “Polyester layered silicate nanohybrids by controlled grafting polymerization,” Journal of Materials Chemistry, vol. 12, no. 12, pp. 3528-3532, Nov. 2002. [15] P. Bordes, E. Pollet, and L. Averous, “Nano-biocomposites: Biodegradable polyester/nanoclay systems,” Progress in Polymer Science, vol. 34, no. 2, pp. 125-155, Feb. 2009. [16] S. Y. Hwang, E. S. Yoo, and S. S. Im, “Effect of the urethane group on treated clay surfaces for high-performance poly(butylene succinate)/montmorillonite nanocomposites,” Polymer Degradation and Stability, vol. 94, no. 12, pp. 2163-2169, Dec. 2009. [17] D. Lincoln, R. Vaia, Z. G. Wang, and B. Hsiao, “Secondary structure and elevated temperature crystallite morphology of nylon-6/layered silicate nanocomposites,” Polymer, vol. 42, no. 4, pp. 1621–1631, 2001. [18] J. H. Chang, B. S. Seo, and D. H. Hwang, “An exfoliation of organoclay in thermotropic liquid crystalline polyester nanocomposites,” Polymer, vol. 43, no. 10, pp. 2969–2974, 2002. [19] R. K. Shah and D. R. Paul, “Organoclay degradation in melt processed polyethylene nanocomposites,” Polymer, vol. 47, no. 11, pp. 4075-4084, May. 2006. [20] C. J. G. Plummer, L. Garamszegi, Y. Leterrier, M. Rodlert, and J.-A. E. Månson, “Hyperbranched Polymer Layered Silicate Nanocomposites,” Chemistry of Materials, vol. 14, no. 2, pp. 486-488, Feb. 2002. 50 [21] M. Shibata, Y. Someya, M. Orihara, and M. Miyoshi, “Thermal and mechanical properties of plasticized poly(L-lactide) nanocomposites with organo-modified montmorillonites,” Journal of Applied Polymer Science, vol. 99, no. 5, pp. 2594-2602, Mar. 2006. [22] E. Pollet, C. Delcourt, M. Alexandre, and P. Dubois, “Transesterification catalysts to improve clay exfoliation in synthetic biodegradable polyester nanocomposites,” European Polymer Journal, vol. 42, no. 6, pp. 1330-1341, 2006. [23] J. W. Cho and D. R. Paul, “Nylon 6 nanocomposites by melt compounding,” Polymer, vol. 42, no. 3, pp. 1083-1094, Feb. 2001. [24] C. Lu and Y.-W. Mai, “Influence of Aspect Ratio on Barrier Properties of Polymer-Clay Nanocomposites,” Physical Review Letters, vol. 95, no. 8, pp. 1-4, Aug. 2005. [25] J. W. Gilman et al., “Flammability Properties of Polymer−Layered-Silicate Nanocomposites. Polypropylene and Polystyrene Nanocomposites †,” Chemistry of Materials, vol. 12, no. 7, pp. 1866-1873, Jul. 2000. [26] B. Jang and C. Wilkie, “The effect of clay on the thermal degradation of polyamide 6 in polyamide 6/clay nanocomposites,” Polymer, vol. 46, no. 10, pp. 3264-3274, Apr. 2005. [27] S. Sinha Ray, “Polymer/layered silicate nanocomposites: a review from preparation to processing,” Progress in Polymer Science, vol. 28, no. 11, pp. 1539-1641, Nov. 2003. [28] M. Okamoto, “Biodegradable polymer/layered silicate nanocomposites: a review,” J. Ind. Eng. Chem, vol. 10, pp. 1156–1181, 2004. [29] T. H. Kim, S. T. Lim, C. H. Lee, H. J. Choi, and M. S. Jhon, “Preparation and rheological characterization of intercalated polystyrene/organophilic montmorillonite nanocomposite,” Journal of Applied Polymer Science, vol. 87, no. 13, pp. 2106-2112, Mar. 2003. [30] S. Sinharay and M. Bousmina, “Biodegradable polymers and their layered silicate nanocomposites: In greening the 21st century materials world,” Progress in Materials Science, vol. 50, no. 8, pp. 962-1079, Nov. 2005. [31] S. Sinharay and M. Bousmina, “Biodegradable polymers and their layered silicate nanocomposites: In greening the 21st century materials world,” Progress in Materials Science, vol. 50, no. 8, pp. 962-1079, Nov. 2005. [32] G. G. Odian, Principles of polymerization. Wiley-Interscience, 2004. [33] C. E. J. Carraher, Introduction to Polymer Chemistry, Second Edition. Boca Raton , Fl 33487- 2742: CRC Press, 2010, p. 534. [34] P. C. Painter and M. M. Coleman, Fundamentals of Polymer Science: An Introductory Text, Second Edition. Lancaster, Pennsylvania 17064 U.S.A: Technomic , 1994, p. 433. [35] E. MURAYAMA, “Optical Properties of Ringed Spherulites,” Polym. Prepr. Jpn, vol. 51, p. 460, 2002. [36] L. Mandelkern, “Crystallization of polymers,” in Journal of Polymer Science, Second., Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo: Cambridge University Press, 2002. [37] M. Stamm, E. W. Fischer, M. Dettenmaier, and P. Convert, “Chain conformation in the crystalline state by means of neutron scattering methods,” Faraday Discussions of the Chemical Society, vol. 68, p. 263, 1979. [38] P. H. Geil, Polymer single crystals. Krieger, 1973. [39] P. J. Flory and D. Y. Yoon, “Molecular morphology in semicrystalline polymers,” Nature, vol. 272, no. 5650, pp. 226-229, Mar. 1978. [40] P. Flory, “On the morphology of the crystalline state in polymers,” Journal of the American Chemical Society, vol. 721, no. 1959, 1962. [41] A. N. Wilkinson and A. J. Ryan, Polymer processing and structure development. Kluwer Academic Publishers, 1998. [42] J. D. Hoffman and R. L. Miller, “Kinetic of crystallization from the melt and chain folding in polyethylene fractions revisited: theory and experiment,” Polymer, vol. 38, no. 13, pp. 3151-3212, Jan. 1997. 51 [43] D. I. Bower, An introduction to polymer physics. Cambridge University Press, 2002, p. 124. [44] J. Lauritzen and J. Hoffman, “Theory of formation of polymer crystals with folded chains in dilute solution,” Journal of Research of the National Bureau of Standards Section a- Physics and Chemistry, vol. 64, pp. 73-102, 1960. [45] J. D. Hoffman, “Polymer Single Crystals. Philip H. Geil. Interscience (Wiley), New York, 1963. xii + 560 pp. Illus. $16,” Science, vol. 143, no. 3602, pp. 121-121, Jan. 1964. [46] J. Hoffman and J. Lauritzen, “Crystallization of bulk polymers with chain folding theory of growth of lamellar spherulites,” Journal of Research of the National Bureau of Standards A, vol. 65, pp. 297-336, 1961. [47] D. M. Sadler and G. H. Gilmer, “A model for chain folding in polymer crystals: rough growth faces are consistent with the observed growth rates,” Polymer, vol. 25, no. 10, pp. 1446-1452, 1984. [48] D. M. Sadler, “Roughness of growth faces of polymer crystals: Evidence from morphology and implications for growth mechanisms and types of folding,” Polymer, vol. 24, no. 11, pp. 1401- 1409, Nov. 1983. [49] D. Sadler and G. Gilmer, “Rate-Theory Model of Polymer Crystallization,” Physical Review Letters, vol. 56, no. 25, pp. 2708-2711, Jun. 1986. [50] B. Wunderlich, Macromolecular physics, no. 3. Academic Press, 1980. [51] J. D. H. HOFFMAN and R. L. MILLER, “Organic Polymers,” in Advancing Materials Research, s P. A. Psara and D. H. Langford, Eds. National Academy of Engineering, 1987, p. 251. [52] A. Ravve, Principles of polymer chemistry, no. 1. Kluwer Academic/Plenum Publishers, 2000. [53] A. Rudin, The elements of polymer science and engineering: an introductory text and reference for engineers and chemists. Academic Press, 1999. [54] X. Alex and C. Edward S, “Nylon Plastics Handbook,” M. I. Kohan, Ed. Hanser Publishers: Munich, Vienna and New York, 1995, pp. 108-137. [55] C. W. Bunn and E. V. Garner, “The Crystal Structures of Two Polyamides (’Nylons'),” Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 189, no. 1016, pp. 39-68, Mar. 1947. [56] Y. Li and W. a Goddard, “Nylon 6 Crystal Structures, Folds, and Lamellae from Theory,” Macromolecules, vol. 35, no. 22, pp. 8440-8455, Oct. 2002. [57] S. Dasgupta, W. B. Hammond, and W. a Goddard, “Crystal Structures and Properties of Nylon Polymers from Theory,” Journal of the American Chemical Society, vol. 118, no. 49, pp. 12291- 12301, Jan. 1996. [58] N. a Jones, S. J. Cooper, E. D. T. Atkins, M. J. Hill, and L. Franco, “Temperature-induced changes in chain-folded lamellar crystals of aliphatic polyamides. Investigation of nylons 2 6, 2 8, 2 10, and 2 12,” Journal of Polymer Science Part B: Polymer Physics, vol. 35, no. 4, pp. 675-688, Mar. 1997. [59] N. a Jones, E. D. T. Atkins, M. J. Hill, S. J. Cooper, and L. Franco, “Chain-Folded Lamellar Crystals of Aliphatic Polyamides. Comparisons between Nylons 4 4, 6 4, 8 4, 10 4, and 12 4,” Macromolecules, vol. 29, no. 18, pp. 6011-6018, Jan. 1996. [60] E. Navarro, L. Franco, J. a Subirana, and J. Puiggali, “Nylon 65 has a Unique Structure with Two Directions of Hydrogen Bonds,” Macromolecules, vol. 28, no. 26, pp. 8742-8750, Dec. 1995. [61] C. Ramesh, A. Keller, and S. J. E. A. Eltink, “Studies on the crystallization and melting of nylon- 6,6: 1. The dependence of the Brill transition on the crystallization temperature,” Polymer, vol. 35, no. 12, pp. 2483-2487, Jun. 1994. [62] C. Sclar and L. Carrison, “Optical crystallography of coesite,” Am. Mineral, vol. 47, pp. 1292- 1302, 1962. [63] F. W. Billmeyer, Textbook of polymer science. Wiley, 1984, p. 578. [64] H. Stanjek and W. Häusler, “Basics of X-ray Diffraction,” Hyperfine Interactions, vol. 154, no. 1- 4, pp. 107-119, 2004. [65] U. W. Gedde, Polymer physics. Chapman & Hall, 1995. 52 [66] G. R. Strobl, The physics of polymers: concepts for understanding their structures and behavior. Springer, 1997. [67] M. García-Gutierrez and D. Rueda, “Bases of Synchrotron Radiation, Light Sources, and Features of X-Ray Scattering Beamlines,” in Applications of Synchrotron Light to Scattering and Diffraction in Materials, Springer, 2009, p. 2. [68] D. R. Rueda et al., “Versatile wide angle diffraction setup for simultaneous wide and small angle x-ray scattering measurements with synchrotron radiation,” Review of Scientific Instruments, vol. 77, no. 3, p. 033904, 2006. [69] “Software for small angle scattering,” Light, Source Source Diamond Neutron, AND the STFC ISI, 2004. [Online]. Available: http://www.small-angle.ac.uk/small-angle/Software/CORFUNC.html. [70] V. B. F. Mathot and L. Benoist, Calorimetry and thermal analysis of polymers. Hanser Publishers, 1994. [71] N. P. P. Cheremisinoff, Polymer characterization: laboratory techniques and analysis. Noyes Publications, 1996. [72] J. D. Menczel and R. B. Prime, Thermal analysis of polymers: fundamentals and applications. John Wiley, 2009. [73] A. W. Coats and J. P. Redfern, “Kinetic Parameters from Thermogravimetric Data,” Nature, vol. 201, no. 4914, pp. 68-69, 1964. [74] H. E. Kissinger, “Reaction Kinetics in Differential Thermal Analysis,” Analytical Chemistry, vol. 29, no. 11, pp. 1702-1706, 1957. [75] H. L. Friedman, “Kinetics of thermal degradation of char-forming plastics from thermogravimetry. Application to a phenolic plastic,” Journal of Polymer Science Part C, vol. 6, no. 1, pp. 183-195, 1964. [76] T. Ozawa, “A New Method of Analyzing Thermogravimetric Data,” Bulletin of the Chemical Society of Japan, vol. 38, no. 11, pp. 1881–1886, 1965. [77] J. H. Flynn and L. A. Wall, “A quick, direct method for the determination of activation energy from thermogravimetric data,” Journal Of Polymer Science Part B Polymer Letters, vol. 4, no. 5, pp. 323-328, 1966. [78] A. Mianowski, “The kissinger law and isokinetic effect,” Journal of Thermal Analysis and Calorimetry, vol. 74, no. 3, pp. 953-973, Dec. 2003. [79] A. I. Lesnikovich and S. V. Levchik, “A method of finding invariant values of kinetic parameters,” Journal of Thermal Analysis, vol. 27, no. 1, pp. 89-93, May. 1983. [80] A. I. Lesnikovich and S. V. Levchik, “Isoparametric kinetic relations for chemical transformations in condensed substances (analytical survey). I,” Journal of Thermal Analysis, vol. 30, no. 1, pp. 237-262, Jan. 1985. [81] S. Barbara, “Infrared spectroscopy: fundamentals and applications,” Analytical Techniques in the Science, 2004. [82] D. K. Platt and R. T. Limited, Biodegradable polymers: market report. Rapra Technology, 2006. [83] S. Pavlidou and C. Papaspyrides, “A review on polymer–layered silicate nanocomposites,” Progress in Polymer Science, vol. 33, no. 12, pp. 1119-1198, Dec. 2008. 4 4 ODD-EVEN AND EVEN-ODD POLYAMIDES:STRUCTURE AND NANOCOMPOSITES 54 55 The work described in this chapter previously appeared in: [1] Morales-Gámez, L. Ricart, A. Franco, L.; Puiggalí, J. European Polymer Journal 2010 , 46, 2063-2077. [2] Morales-Gámez, L. Soto, D. Franco, L.; Puiggalí, J. Polymer 2010, 51, 5788-5798. [3] Morales-Gámez, L., Casas, M.T. Artigas, A. Franco, L. Puiggalí, J. Submitted Papper [4] Ricart, A, Soto, D. Franco, L. Morales, L. T.; Puiggalí, J. IOP Conference Series: Materials Science and Engineering 2010, 14, 012006. 56 63 Nylon 56 crystallized easily during cooling runs from the melt state (e.g. an exothermic peak at 224 ºC was detected at a cooling rate of 10 ºC/min) giving rise to samples with a different melting behaviour. Thus, a posterior heating trace showed clearly as the broad high temperature peak was split in two peaks (peak 2 at 238 ºC and peak 3 at 251 ºC) and that an exothermic peak indicative of a recrystallization process appeared (241 ºC). The low temperature melting peak (peak 1 at ca. 232 ºC) could still be observed although with a very low intensity. Heating traces of melt quenched samples clearly indicate that a completely amorphous sample could not be obtained at the maximum cooling rate allowed by the equipment. However, the glass transition temperature was detected at a temperature close to 55 ºC. It is relevant that the melting behaviour was slightly different than observed for melt and solution crystallized samples since peak 2 was not detected. This feature suggests that peak 3 could be associated to thickest lamellae mainly produced during the heating process as a consequence of a melt/recrystallization of the thinner lamellae. This peak should be enhanced when more imperfect lamellae susceptible of reorganization were obtained as presumable in the melt quenched samples. The nature of peak 1 is more intriguing since at this stage two alternatives may be considered: a) the existence of very defective crystals and b) a polymorphic transition around 225-230 ºC. 235 240 245 250 255 260 265 230 240 250 260 270 280 230 240 250 260 270 280 Figure 4.1.3 Melting peaks for isothermally melt crystallized samples. Deconvoluted profile is only shown for the sample crystallized at 236 ºC (dashed box). The inset shows the Hoffman-Weeks plot drawn for the crystallization temperature dependent melting peak (peak 2). Tc (ºC) Temperature (ºC) Heat Flow (a.u.) Tc = 233 ºC Tc = 235 ºC Tc = 236 ºC Tc = 237 ºC Tc = 240 ºC Tc = 238 ºC Tc ( ºC ) Tm ( ºC ) Tm 0 : 268 ºC Peak 2 Peak 3 Peak 2 Peak 1 64 Figure 4.1.3 shows the heating traces of samples previously isothermally crystallized from the melt state at different temperatures. Peaks 2 and 3 appear generally overlapped and consequently is difficult to differentiate the two melting processes. However, it can be stated that peak 2 increased on intensity and shifted to higher temperatures when crystallization temperature did, whereas peak 3 remained at a practically constant temperature. This feature is consistent with the indicated melt/reorganization process where thinner lamellae convert into thicker ones. Furthermore, it is possible to infer the equilibrium melting temperature of nylon 56 by considering the temperature evolution of peak 2 with crystallization temperature. In this way, the Hoffman-Weeks plot [25] displayed in the inset of Figure 4.1.3 indicates an extrapolated equilibrium temperature of 268 ºC, which is close to the value of 266 ºC previously postulated [26] from theoretical considerations based on the spherulite grown model forwarded by Hoffmann-Weeks [25]. Heating runs showed also the presence of the low temperature peak 1, but only when samples were isothermally crystallized at temperatures equal or lower than 235 ºC. Thus, this value is a limit for a possible crystalline transition or for the development of the indicated defective crystals. Isothermal experiments allowed the determination of the overall crystallization kinetics, which depends on primary nucleation and crystal growth, for a very restrictive temperature range due to the experimental limitations caused by the high speed of the crystallization process. The time evolution of the relative degree of crystallinity,  (t), was determined from hot crystallization exotherms (Figure 4.1.4 a) through the ratio area of the exotherm up to time t divided by the total exotherm area, i.e.:  (t) = t tdtdtdH 0)/( /  0)/( tdtdtdH (4.1.1) where dH/dt is the heat flow rate and t0 the induction time. The development of crystallinity always showed a characteristic sigmoidal dependence on time, as plotted in the inset of Figure 4.1.4a for six hot crystallization experiments. Kinetic crystallization data were analyzed assuming the well known Avrami equation [27,28] for primary crystallization: 1 -  (t) = exp[-Z (t-t0)n] (4.1.2) where Z is the temperature-dependent rate constant and n the Avrami exponent whose value varies according to the crystallization mechanism. A normalized rate constant, k = Z1/n, is usually evaluated for comparison purposes since its dimension (time-1) is independent of the value of the Avrami exponent. Table 4.1.1 Isothermal crystallization kinetic parameters deduced from DSC experiments for nylon 56.summarizes the main kinetic parameters of the primary crystallization process, which were deduced from the plots of log{-ln[1-  (t)]} against log (t - t0). The values of the Avrami 65 exponent for the hot isothermal crystallizations lie in a narrow range, from 2.15 to 2.80, 2.50 being the average value. This suggests a predetermined (heterogeneous) nucleation with spherical growth that occurred under slight geometric constraints since the theoretical value should be equal to 3. Both sporadic (heterogeneous) and homogeneous nucleation can be clearly discarded as a higher exponent, close to 4, should be derived and furthermore these nucleation mechanisms should mainly be favoured at high undercoolings. The values of the reciprocal of the crystallization half-time, 1/  1/2, are also summarized in Table 1. This parameter is a direct measure of the crystallization process, and could therefore be used to check the accuracy of the Avrami analyses. In this way, a similar dependence with the crystallization temperature was found for this parameter and the kinetic rate constant, demonstrating the suitability of the deduced Avrami values. Figure 4.1.4 b shows the crystallization exotherms obtained during cooling runs performed at the different rates used in the synchrotron radiation experiments. A well defined peak is always observed within a narrow temperature range, which obviously shifts to lower temperatures by increasing the cooling rate. However, it is interesting to note that peaks had a long tail which could be associated to a secondary crystallization process and which was more clearly observed at high cooling rates. Table 4.1.1 Isothermal crystallization kinetic parameters deduced from DSC experiments for nylon 56. Tc (ºC) n Z·108 (s-n) k·103 (s-1) 1/ τ1/2 . 103 (s-1) 233 2.34 217.1 3.80 4.17 235 2.57 17.8 2.36 2.33 236 2.80 0.396 0.998 1.17 237 2.64 0.511 0.723 0.83 238 2.52 0.601 0.547 0.55 240 2.15 3.79 0.353 0.42 66 Figure 4.1.4 a) Exothermic DSC peaks corresponding to the hot isothermal crystallizations performed between 233 and 240 ºC. Inset shows the development of relative crystallinity over time for isothermal crystallizations performed between 233 and 240 ºC. b) Dynamic DSC curves obtained at the indicated rates for the hot crystallization of nylon 56. Temperature (ºC) Heat Flow (a.u.) 15 ºC/min 20 ºC/min 12 ºC/min 8 ºC/min Secondary crystallization t – t0 (min) 233 ºC 236 ºC 237ºC 235ºC 238 ºC 240 ºC Heat flow (a.u.) Crystallinity t-t0 ( min) 0 1 0.5 240 ºC 238 ºC 237 ºC 236 ºC 235 ºC 233 ºC a) b) 0 10 20 30 40 50 60 70 80 90 100 110 0 15 30 45 60 75 90 100 125 150 175 200 225 250 67  Brill transition of nylon 56 on heating/cooling process Fiber diffraction patterns of nylon 56 were mainly characterized by strong equatorial reflections at 0.432 and 0.375 nm and an off meridional reflection at 1.272 nm (inset of Figure 4.1.5), which were indexed as the (020) and (110) reflexions on the basis of a monoclinic unit cell with a = 0.512 nm, b = 0.864 nm, c (chain axis) = 3.133 nm and  = 125.7º (form I) [4]. Structural modelling based on the diffraction data and energy calculations pointed towards the indicated model based on the establishment of two hydrogen bonding directions [4]. Figure 4.1.5 shows the X-ray fiber diffraction patterns of a nylon 56 sample taken under stress at 200 ºC and 220 ºC. At 200 ºC the pattern shows only one strong and diffuse equatorial reflection at 0.423 nm, which is an indication that the Brill transition took place. It is interesting to note that 00l reflections still appeared with an off meridional orientation which is an indication that the structure obtained at 200 ºC differed from a pseudohexagonal structure usually postulated for conventional polyamides. These 00l reflections seemed to have a close to meridional orientation in the patterns obtained at 220 ºC, although it is difficult to determine the cc* angle due to their arched appearance and the overlapping between 00l and 00 l reflections. 200 ºC 220 ºC 25 ºC 25 ºC c* (220 ºC) c* (200 ºC) Figure 4.1.5 X-ray fiber diffraction patterns of nylon 56 at 200 ºC (left) and 220 ºC (right). Insets show the equatorial reflections observed at 25, 200 and 220 ºC and the 002 reflections (second layer line) observed at 25 ºC. 68 In this case, the new additional equatorial reflections observed at 0.454 nm and 0.436 nm are highly significant since allow to discard again a pseudohexagonal structure. Previous works suggested that at high temperature a monoclinic structure (form II) with a = 0.551 nm, b = 0.846 nm, c (chain axis) = 3.133 nm and  = 112.6º was achieved [4]. Note that the cc* angle was 35.7º at room temperature whereas it decreased to 22.6º at 220 ºC justifying the close meridional orientation detected for the 00l reflections. 200 270 230 ºC T ºC 25 250 200 ºC T ºC b) a) Figure 4.1.6 Three-dimensional representations of WAXD profiles of nylon 56 during cooling (12 ºC/min) from the melt to room temperature. All the temperature range is showed in a), whereas a different view covering only the last frames (up to 200 ºC) is shown in b). 69 Figure 4.1.6 shows three-dimensional representations of WAXD profiles obtained by synchrotron radiation during a heating process performed at 12 ºC/min from room temperature to fusion (q is the scattering vector given by [4/  ] sin (  ) or 2  / dB where  and dB are the scattering angle and the Bragg spacing, respectively). Similar temperature dependent profiles were observed at heating rates of 8, 12, 15 and 20 ºC/min. Profiles showed that the spacings of the two equatorial reflections at 0.433 and 0.374 nm gradually merged into a single peak at 0.423 nm that was reached at a temperature close to 200 ºC. This process seems a typical Brill transition where a pseudohexagonal packing (  *-form) is favoured at a temperature slightly lower than the melting point. It is worth mentioning that the Brill transition temperature of nylon 56 was practically independent of the heating rate as shown in Figure 4.1.7. After the Brill transition, new peaks (e.g. those above indicated at 0.454 nm and 0.436 nm) started to appear as shown in Figure 4.1.6 b and a transition towards the indicated form II took place. All equatorial reflections became narrower (Figure 4.1.5) and increased on intensity (Figure 4.1.6mb) during heating above the Brill transition temperature and before to start the melting process. Transition to form II occurred in a temperature range that was slightly lower (5 ºC) than the endothermic peak 1 observed in the calorimetric analyses. In this way, this small melting peak seems to be related to highly defective crystals formed between bundles of lamellae, a conclusion that has been reported for different polyamides [29-31]. Figure 4.1.8 compares the deconvoluted WAXD profiles representative of the structures attained at room temperature, at the Brill transition temperature and at a temperature close to 0.37 0.38 0.39 0.40 0.41 0.42 0.43 0.44 050 100 150 200 250 300 Distance (nm) T(ºC) 0.437 nm 0.422nm 0.374nm Tm = 263 ºC 0 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 Q (a.u) t-t0 (min) 8 ºC/min 12 ºC/min 15 ºC/min 20 ºC/min 230 200 170 140 110 80 50 20 0.433 nm 0.374 nm T (ºC) Spacing (nm) 0.423 nm 200-210 ºC Figure 4.1.7 Plot showing the temperature evolution of the spacings corresponding to the two strongest equatorial reflections at different heating rates. 70 fusion. In all cases, two amorphous halos (average values of 0.420 nm and 0.375 nm) were detected. However, the position of the maxima changed with temperature (i.e. the maximum of the first halo appeared at 0.409 and 0.430 nm in the patterns taken at 25 and 230 ºC, respectively). This feature suggests that the amorphous phase has a more compact molecular arrangement when temperature decreases, as it will be discussed in the next section. In fact, a similar increase in the average interchain distance in the amorphous phase above the Brill transition temperature was reported and analyzed in detail for nylon 66 [17]. It is worth to pointing out that Bragg reflections were very broad while temperature was lower or equal than the Brill transition temperature. Assuming that the low temperature structure (form I) is defined by a molecular arrangement where hydrogen bonds are established along two directions, it seems reasonable to expect clear differences on heating between conventional polyamides and nylon 56. Thus, the pseudohexagonal structure [32-34] (  ’ form) attained with nylons characterized by a single hydrogen bond direction could not be observed in the diffraction patterns of nylon 56. Transitions induced by temperature on this polyamide may involve only slight changes in the torsional angles vicinal to amide groups or even an increase in the mobility of polymethylene segments without disrupting the initial hydrogen-bonding scheme. Note that the chain axis projection may correspond to a pseudohexagonal packing, as deduced from the single equatorial reflection at the Brill transition temperature, but a chain axis shift still remained between neighbouring chains. In this sense, fiber patterns with nonmeridional 00l reflections are essential to support the finding that the Brill structure is different from the conventional  ’ form. Figure 4.1.9 a shows the WAXD profiles acquired during a cooling run (10 ºC/min) from the melt state. It is clear that nylon 56 crystallized into the form II characterized as above indicated by multiple narrow reflections with an equatorial or close equatorial orientation. Note that the profile showed in Figure 4.1.9 is practically identical to that attained during the heating process (Figure 4.1.8 c) just at some degrees before fusion. Figure 4.1.9 a shows also that the reflection at ca. 0.423 nm does not split when temperature is lowered up to room temperature and consequently it could be deduced that the Brill transition is not reversible on cooling. WAXD profiles showed also that characteristic reflections of form II moves to lower spacings by decreasing the temperature and overlapped the main equatorial reflection at ca. 120 ºC. Thus, the intensity of the reflection at 0.423 nm increased during cooling as well as the peak became broader. 71 0 100 200 300 400 500 10 12 14 16 18 20 I(a.u.) q(nm-1) 0 100 200 300 400 500 600 10 12 14 16 18 20 I (a.u.) q (nm-1) 0 100 200 300 400 500 600 10 12 14 16 18 20 I (a.u) q (nm-1) b) c) a) q (nm-1) 0.433 0.374 0.419 0.423 0.436 0.454 Figure 4.1.8 One-dimensional WAXD profiles for nylon 56 taken at room temperature (a), 200 ºC (b) and 230 ºC (c) during a heating scan (12 ºC/min). Spacings of main reflections are indicated together with the deconvoluted peaks. 72 0 100 200 300 400 500 600 700 800 11 13 15 17 19 21 I(a.u) q (nm-1) 0 100 200 300 400 500 600 10 12 14 16 18 20 I(a.u.) q (nm-1) b) c) a) q (nm-1) I (a.u) 25 270 T ºC 120 ºC 0.423 0.436 0.454 0.419 0.383 0.440 0.405 Figure 4.1.9 . a) Three-dimensional representation of WAXD profiles of nylon 56 during cooling (12 ºC/min) from the melt to room temperature. b) and c) One-dimensional WAXD profiles for nylon 56 taken at 220 ºC (b) and at room temperature (c) during a cooling run (12 ºC/min) from the melt state. Spacings of main reflections are indicated together with the deconvoluted peaks. 79 the development of the three different birefringent zones with changes that took place at well defined temperatures (237 and 233 ºC). It is interesting to note that the indicated birefringence changes are different to those observed in conventional polyamides like nylon 66 where birefringence changed from negative to positive by decreasing the crystallization temperature. In this case, the change in the optical properties was explained considering the structure based on the stacking of hydrogen-bonded sheets and different growth geometries [41,42]. Thus, positive and negative spherulites were interpreted as a consequence of the establishment of hydrogen bonds along a radial or a tangential spherulitic direction, respectively. The birefringence sign was directly associated with how lamellae with a single structure grow in the spherulite. However, the reason for such a drastic change in the growth mechanism at a well defined temperature remains unclear. The peculiar structure found for the high temperature form of the studied odd-even nylon where two hydrogen-bonding directions seem to exist may be one of the reasons for the unusual formation of positive spherulites at higher crystallization temperature. In any way, the synchrotron data acquired during cooling runs allowed discarding a direct relation between the change on the birefringence sign and possible polymorphic transitions. Furthermore, no changes on both texture and birefringence could be detected when the different spherulites were heated until fusion. Thus, the morphologies developed during crystallization of nylon 56 were not reversible. Spherulitic growth rates were determined from isothermal experiments by following the change of the spherulite radius with time up to impingement (Figure 4.1.15 a) within the studied temperature intervals. The measured radial growth rates, G, varied from a minimum value of 0.08 m/s at 239 ºC to a maximum value close to 1.3 m/s at 225 ºC. Non-isothermal procedures were also applied to study the temperature dependence of the spherulitic growth rate during hot crystallization. Thus, the spherulitic growth rate (G) can be estimated [43-45] by measuring the change of the spherulite radius (R) with temperature (T) when experiments are performed at a constant cooling rate (dT / dt): G = dR / dt = (dR / dT) (dT / dt) (4.1.4) 80 237 235 230 a) b) 237 235 225 c) 237 233 25 25 70 40 40 Figure 4.1.14 a) Optical micrographs of nylon 56 spherulites isothermally crystallized at 237 ºC (left), 235 ºC (middle) and 230 ºC (right). b) Optical micrograph of a nylon 56 spherulite that was isothermally crystallized at three different temperatures: Firstly at 237 ºC, secondly at 235 ºC and finally at 225 ºC. Inset shows a black and white micrograph where the low birefringence zone corresponding to the polymer crystallized at the intermediate temperature appeared as a black ring. c) Optical micrograph of a nylon 56 spherulite nonisother-mally crystallized at a cooling rate of 1 ºC/min. 81 Experimental problems lie in the choice of the cooling rate required to maximize the crystallization temperature range where radii can be well measured. For this reason, the use of various rates is highly effective in expanding this range. The plot of the radius versus temperature (Figure 4.1.15 b) can be fitted to polynomial equations with a good regression coefficient (r) that allows the calculation of the value of its first derivative (dR / dT) for each cooling rate as a function of the crystallization temperature. Third-order equations were always chosen since the regression coefficients (≥ 0.998) were slightly better than those calculated for lower-order equations and remained practically constant when higher orders were assayed. R ( m ) R ( m ) a) b) Figure 4.1.15 a) Plots of the radius of nylon 56 spherulites versus crystallization time for isothermal hot crystallizations performed at temperatures ranging between 225 and 237 ºC. b) Variation in spherulite radius with temperature during cooling at the indicated rates. 82 Figure 4.1.16 b plots the deduced G values from non-isothermal data and those measured from isothermal experiments. It should be pointed out that a good agreement was found and that nonisothermal experiments had several advantages: a continuous evolution could be determined and measures were less time consuming. Experimental data defined the right side of the typical bell shaped curve that describes the temperature dependence of the growth rate, i.e. the zone controlled by secondary nucleation. Both isothermal and non-isothermal measures suggests the existence of a shoulder at high temperatures (> 233 ºC) which may be a consequence of a different secondary nucleation constant. Thus, at least experiments pointed out to the existence of two crystallization regimes which could be associated to different spherulites, e.g. positive at temperatures lower than 233 ºC and negative at higher temperatures. 4.1.4 Conclusions Nylon 56 crystallized from solution according to a peculiar monoclinic structure (form I) where hydrogen bonds were established along two directions and where neighbouring chains were shifted along their chain axis direction. On heating, this structure showed a Brill transition resulting in a pseudohexagonal chain axis projected unit cell and a structure where the chain Positive spherulites Negative spherulites Figure 4.1.16 Spherulitic growth rates determined by the equations deduced for cooling runs of 8 ( ▲), 1 (▲) and 0.5 ºC/min (∆ ). For the sake of completeness, experimental data deduced from isothermal experiments are also plotted ( ◊). 83 axis shift was kept in order to optimize the hydrogen bonding interactions. At some degrees before fusion, the diffraction patterns showed new narrow reflections which could be indexed according to a new monoclinic unit cell (form II). Brill transition was not reversible since nylon 56 mainly crystallized from the melt into form II, which on cooling gave rise to a pseudohexagonal packing. A minor crystallization into form I could also be detected and accounted into a significant ratio of this form when room temperature was achieved. Nylon 56 crystallized on cooling into fibrillar spherulites with optical properties that were depended on the crystallization temperature and differed from those found in nylons having conventional sheet structures. During crystallization thinner lamellae inserted into the loosely stacked bundles of primary lamellae and the interlamellar amorphous regions became more compact. 4.1.5 References [1] Bunn CW, Garner EV, Proc. R. Soc. London Ser. A 1947;189:39-68. [2] Xenopoulos A, Clark ES. In Nylon Plastics Handbook; Kohan MI Ed.; Hanser Publishers: Munich, Vienna and New York, 1995; Chapter 5:108-137. [3] Kinoshita Y, Makromol. Chem. 1959;33:1-20. [4] Puiggalí J, Franco L, Alemán C, and Subirana JA. Macromolecules 1998;31: 8540-48. [5] Franco L, Subirana JA, Puiggalí J. Macromolecules 1998;31:3912-24. [6] Villaseñor P, Franco L, Subirana JA, Puiggali J. J Polym Sci Part B, Polym Phys Ed 1999;37:2383- 95. [7] Navarro E, Franco L, Subirana JA, Puiggalí J. Macromolecules 1995;28:8742- 50. [8] Franco L, Cooper SJ, Atkins A DT, Hill M, Jones NA. Macromolecules 1998;36: 1153-65. [9] Holmes DE, Bunn CW and Smith D. J Polym Sci Part A, General Papers 1955;17:159-177. [10] Brill R. Makromol Chem 1956;18:294-309. [11] Schmidt GF, and Stuart HA, Naturforsch Z. 1958;13A:222-26. [12] Hirschinger J, Miura H, Gardner KH, EnglishAD, Macromolecules 1990;23:2153-59. [13] Wendoloski JJ, Gardner KH, Hirschinger J, Miura H, English AD. Science 1990;247:431-436. [14] Ramesh C, Keller A and Eltink S J E A. Polymer 1994;35:2483-87. [15] Hill MJ, Atkins EDT, Macromolecules 1995;28(2):604-9. [16] Vasanthan N, Murthy NS, Bray RG. Macromolecules 1998;31:8433-35. [17] Murthy N S, Wang Z, Hsiao BS, Macromolecules 1999;32:5594-99. [18] Ramesh C, and Gowd EB, Macromolecules 1999;32:3721-26. [19] Jones NA, Atkins EDT, Hill MJJ, Polym. Sci. Part B, Polym. Phys. 2000;38:1209-21. [20] Feldman A Y, Wachtel E, Vaughan G B M, Weinberg A and Marom G. Macromolecules 2006;39:4455-59. [21] Tashiro K, Yoshioka Y, Polymer 2004;45:6349-55. [22] Yoshioka Y, Tashiro K, Ramesh C, Polymer 2003;44:6407-17. 84 [23] Cui X, Yan D, Eur Polym J, 2005;41:863-870. [24] Rueda DR, García-Gutiérrez MC, Nogales A, Capitán MJ, Ezquerra TA, Labrador A, et al. Rev Sci Instrum 2006;(77)Art. No. 033904 Part 1. [25] Hoffman JD, Weeks JJ. J. Chem. Phys. 1962;37:1723-46. [26] Magill JH. J. Polym. Sci. part A 1965;3:1195-1219. [27] Avrami, M. J Chem Phys 1939;7:1103-12. [28] Avrami, M. J Chem Phys 1940;8:212-24. [29] Wunderlich B. Macromolecular Physics. Crystal Melting, vol. 3. New York : Academic Press ; 1980. [30] Liu M, Zhao Q, Wang Y, Zhang C, Mo Z, Cao S. Polymer 2003;44:2537-45. [31] Cui X, Qing S, Yan D. Eur Polym J 2005;41:3060-68. [32] Biangardi JJ, Macromol Sci 1990;29:139-153. [33] Jones NA, Atkins EDT, Hill M, Cooper SJ, Franco L. Polymer 1997; 38:2689-99. [34] Feldman AY, Wachtel E, Vaughan GBM, Weinberg A, Marom G. Macromolecules 2006;39:4455-59. [35] Vonk, C. G.; Kortleve, G. Kolloid Z Z Polym 1967;220:19-24. [36] Vonk, C. G. J Appl Cryst 1975;8:340-341. [37] Hsiao, B. S.; Wang, Z.; Yeh, F.; Yan, G.; Sheth, K. C. Polymer 1999;40:3515-23. [38] Hsiao, B. S.; Gardner, K. H.; Wu, D. Q.; Chu, B. Polymer 1993;34:3986-95. [39] Ikada, Y.; Jamshida, K.; Tsuji, H.; Hyoan, S. H. Macromolecules 1987; 20: 904-6. [40] Dreyfuss, P. J. Polym Sci. Part B: Phys. Ed. 1973; 11:201-16. [41] Lovinger AJ. J. Appl. Phys 1978;49:5003-13. [42] Lovinger AJ. J. Appl. Phys 1978;49:5014-28. [43] Chen M, Chung CT, J Polym Sci Part B: Polym Phys 1998;36:2393-99. [44] di Lorenzo, M. L.; Cimmino, S.; Silvestre, C. Macromolecules 2000;33:3828-32. [45] di Lorenzo, M. L. Polymer 2001;42:9441-46. 85 4.2 Study on the Brill transition and melt crystallization of Nylon 65: A polymer able to adopt a structure with two hydrogen-bonding directions. Real time temperature dependence of X ray diffraction patterns and infrared spectra for nylon 65, a representative polymer of the even-odd nylon series, was studied. A particular structure based on the establishment of two hydrogen-bonding directions had previously been postulated for this polymer. Therefore, the determination of its temperature-induced transitions is a relevant topic. Results indicate that nylon 65 undergoes a reversible Brill transition at high temperature, leading to a pseudohexagonal chain axis projected unit cell. Furthermore, this polyamide shows a polymorphic transition around 100 ºC which is not completely reversible on cooling. Crystallization of nylon 65 was also analyzed by simultaneous WAXD and SAXS synchrotron radiation experiments to determine the evolution of the degree of crystallinity and morphological parameters on cooling. Optical microscopy studies were also performed under isothermal and non-isothermal conditions to distinguish the different spherulitic morphologies. Results reveal that the optical properties of nylon 65 spherulites are different from those of conventional even-even nylon spherulites. Multiple melting peaks associated with lamellae of different thicknesses were observed in the calorimetric heating scan of melt-crystallized samples. 86 4.2.1 Introduction It is well known that the structures of aliphatic polyamides are usually based on the stacking of sheets composed of hydrogen-bonded molecular chains with a planar zig-zag conformation ( and forms).1 Nylons derived from -aminoacids with an even number of carbon atoms (e.g. nylon 62) or even diamines and even dicarboxylic acids (e.g. nylon 663) are the most representative examples. X-ray fiber diffraction patterns of polymers with the above structures are characterized by two strong equatorial reflections that appear at spacings close to 0.44 and 0.37 nm. These reflections are associated with interchain distances within and between layers, respectively. Energy considerations indicate that sheet structures are favored when NH and CO groups of neighboring chains face each other and form all possible hydrogen bonds with an appropriate geometry (e.g. angles close to 180º for N-H…O and H…OC interactions). However, depending on the number of methylene groups of the constitutive units, this cannot be achieved with an all trans molecular conformation. Thus, new structures should be favored as the  form firstly postulated for nylon 774. Furthermore, the  form seems to be stabilized when the number of methylene groups is high, even if the hydrogen-bond geometry can be well established with an all trans conformation.1,5 The  form is characterized by a pseudohexagonal molecular packing that gives rise to a characteristic diffraction pattern with a strong equatorial reflection at 0.415 nm. In this case, the torsional angles of the bonds adjacent to the amide groups tend to ± 120º, causing the amide plane to tilt by approximately 60º. The chain is shortened and the establishment of good hydrogen-bonding interactions along a single direction becomes possible. Some polyamides, like nylon 6, show polymorphism between the  and forms. The first is commonly caused by slow cooling from the melt while the second occurs in melt-spun fibers or by rapid crystallization from the melt.6,7 Stretching or annealing may favor the  to  conversion7,8 whereas the opposite is observed by treatment with iodine/potassium iodide aqueous solutions.9 Nylon 65 is also a polyamide that cannot form all favorable hydrogen-bonding interactions when molecular chains have an all trans conformation (Figure 4.2.1 a). Recent diffraction data surprisingly revealed that solution-crystallized single crystals had characteristic reflections at 0.432 and 0.375 nm instead of those expected at ~0.415 nm (  -form). In fact, a new structural model characterized by the establishment of hydrogen bonds along two different directions was inferred10 (Figure 4.2.1 b). The molecular conformation is close to the all trans one since only a slight deviation towards 150º (or -150º) for the two CH2CH2-CONH torsional angles of the dicarboxylic moiety was postulated. This conformation causes amide groups of the odd glutaric 87 unit to rotate in opposite senses from the plane defined by the methylene carbon atoms. Hydrogen bonds along two directions can be well established when neighboring chains are conveniently shifted, giving rise to a monoclinic unit cell. Aliphatic polyamides usually show a not completely well understood phase transition that occurs on heating/cooling. It is named Brill transition and was discovered in nylon 66, with the detection of a reversible change from a triclinic to a pseudohexagonal structure.11 Basically, on heating the two characteristic packing reflections of the triclinic sheet structure (0.44 and 0.37 nm) gradually merge into a single reflection (0.42 nm) indicative of a pseudohexagonal packing. b) a) // // Figure 4.2.1 a) Scheme of the unfavorable hydrogen-bond geometry between nylon 65 molecular chains with an all trans conformation. b) Scheme of the establishment of hydrogen bonds along two directions when consecutive amide planes of a molecular chain slightly rotate in opposite directions from the plane defined by the methylene carbon atoms. External chains (ball and stick representation) should be shifted along the chain axis direction (see arrows) with respect to the central chain (stick representation), thus giving rise to a monoclinic unit cell. Color code: nitrogen, blue; oxygen, red; carbon, gray; hydrogen, brown. 88 Although a large number of studies have been reported for several nylons,12-26 the phenomenon is not yet fully understood and some points still deserve special attention: a) The temperature (TB) at which the Brill transition is considered complete is variable since it depends on many factors (e.g. crystallization conditions, thermal history and heating rate). b) The Brill transition is reversible (i.e. on cooling the single peak splits again into the two indicated packing spacings). A hysteresis effect is usually detected since TB is higher on heating than on cooling. c) Gradual crystallographic changes occur over a wide temperature range before and after the transition for heating and cooling processes, respectively. These changes reflect a variation in the dimensions of the unit cell associated with the layered structure. The occurrence of a phase transition at TB cannot be corroborated by DSC experiments or optical microscopy observations since no additional endothermic/exothermic peaks or changes in birefringence were respectively detected. The aim of the present work is to gain insight into the structural transitions induced by temperature observed in even-odd polyamides having the new structure with two hydrogenbonding directions. Nylon 65 was specifically chosen as a representative polymer. It should also be pointed out that the Brill transition has not yet been studied for any even-odd polyamide. 4.2.2 Experimental section  Materials Nylon 65 was synthesized, as previously described,10 by interfacial polycondensation of 1,5- diaminopentane and adipoyl dichloride using toluene as organic solvent and sodium hydroxide as proton acceptor. An intrinsic viscosity of 0.85 dL/g was determined in dichloroacetic acid at 25 ºC.  Measurements Calorimetric data were obtained by differential scanning calorimetry using a TA Instruments Q100 series with Tzero technology and equipped with a refrigerated cooling system (RCS) operating at temperatures from -90 ºC to 550 ºC. Experiments were conducted under a flow of dry nitrogen with a sample weight of approximately 5 mg, while calibration was performed with indium. The Tzero calibration involved two experiments: the first was done without samples and the second was performed with sapphire disks. The spherulite growth rate was determined by optical microscopy using a Zeiss Axioskop 40 Pol light polarizing microscope equipped with a Linkam temperature control system configured by a THMS 600 heating and freezing stage connected to a LNP 94 liquid nitrogen cooling 95  Brill transition studies on nylon 65 Figure 4.2.5 a shows the X-ray fiber diffraction pattern of a nylon 65 sample taken under stress at 190 ºC. It is clear that only one strong equatorial reflection (ca. 0.425 nm) was detected and that 0k0 reflections had an off-meridional orientation. Thus, the high temperature structure does not correspond to a typical hexagonal unit cell, although the existence of pseudohexagonal packing was suspected by considering only the h0l reflections. In fact, a monoclinic unit cell with a = 0.523 nm, b (chain axis) = 3.055 nm, c = 0.85 nm,  =  = 90º and  = 110º was previously reported.10 The inset of Figure 4.2.5 a contains the initial fiber pattern taken at room temperature, that is, before the start of the heating process. In this case, two equatorial reflections can be clearly distinguished as well as the 0k0 reflections defining a bb* angle slightly higher than that measured in the pattern taken at 190 ºC (24º versus 20º). At intermediate temperatures (e.g. 100 ºC) equatorial spots appeared at intermediate positions, as shown in the insets of Figure 4.2.5 a as well. It is also remarkable that the high temperature pattern suggests a highly crystalline sample since some new non-equatorial reflections appeared (see arrow in Figure 4.2.1 a). Figure 4.2.6 a shows a three-dimensional representation of WAXD profiles obtained by synchrotron radiation during a heating process performed at 10 ºC/min from room temperature to fusion. The same temperature dependent profiles were observed at heating rates of 13, 15 and 20 ºC/min. Representative profiles taken at selected temperatures are included in Figure 4.2.6 b for comparison of the spacings of the main equatorial reflections. Diffraction profiles and fiber patterns obtained during the heating runs allow the inference of three highly significant features: a) The spacings of the two equatorial reflections at 0.432 and 0.375 nm remained practically constant up to a temperature of approximately 70 ºC. Then, the intensity of these reflections diminished whereas new ones near 0.426 and 0.400 nm appeared with increasing intensity. A polymorphic transition seemed to occur which practically ended when a temperature of 120 ºC was reached. No significant changes were observed for the 0k0 reflections (as deduced from the fiber patterns), and consequently the transition seemed to involve only a modification of the dimensions of the chain axis projected unit cell. Thus, this rectangular cell is defined at room temperature by parameters of 0.412 and 0.862 nm which changed to 0.453 and 0.844 nm when the temperature reached 120º. This change implies a less compact structure since the packing surface increases from 0.355 nm2 to 0.383 nm2. The two structures found at low and high temperature will hereafter be called forms I and II, respectively. b) After 120 ºC the two reflections at 0.426 and 0.400 nm gradually merged into a single peak at 0.425 nm that was obtained at a temperature of 190 ºC. This process seems a typical Brill transition where a pseudohexagonal packing (*-form) is favored at a 96 temperature slightly lower than the melting point (Figure 4.2.4). It is worthing to point out that the Brill transition of nylon 65 was always observed at the same temperature (TB) despite variations in the heating rate. c) New peaks (e.g. 0.379 nm), which suggest an increase of the crystalline order, were detected when the temperature was higher than 190 ºC. b) 12 13 14 15 16 17 18 19 I (a.u.) q (nm-1) 174.5 ºC 99.0 ºC 93.7 ºC 90.2 ºC 86.7 ºC 69.2 ºC 206.1 ºC 139.4 ºC 20 ºC T I (a.u.) q (nm-1) a) 250 ºC 25 ºC Figure 4.2.6 a) Three-dimensional representation of WAXD profiles of nylon 65 during heating (10 ºC/min) from room temperature to fusion. b) One-dimensional WAXD profiles of nylon 65 samples taken at selected temperatures during a heating scan (10 ºC/min). Spacings of main reflections are indicated. 97 Figure 4.2.7 a) Three-dimensional representation of WAXD profiles of nylon 65 during cooling (10 ºC/min) from the melt to room temperature. Inset shows a different view of the temperature evolution of characteristic reflections. b) Onedimensional WAXD profile for nylon 65 taken at room temperature after a cooling run (10 ºC/min) from the melt state. Spacings of main reflections are indicated together with the deconvoluted peaks. b) q (nm-1) I (a.u.) T 25 ºC 250 ºC a) 0.379 nm 0.403 nm 0.430 nm 0.422 nm 98 The form I to form II transition takes place in the temperature range where a broad endothermic peak (110 ºC) was observed in the DSC heating run of the sample coming directly from synthesis (Figure 4.2.2), whereas no calorimetric peak was detected around the Brill transition temperature, which is common in polyamides. Thus, for example DSC heating traces of nylon 66 do not show any endothermic peak at TB with the exception of samples crystallized from solution.14,20,30 Assuming that form I corresponds to a structure with two hydrogen-bonding directions, it seems reasonable that thermal treatments result in structures different from the conventional  form, which is characterized by a single hydrogen-bonding direction. Thus, transitions may involve only slight changes in the torsional angles vicinal to amide groups or even an increase in the mobility of polymethylene segments, which could lead to a pseudohexagonal packing without disrupting the initial hydrogenbonding scheme. In this sense, fiber patterns with nonmeridional 0k0 reflections are essential to support the finding that the Brill structure is different from the conventional  form. Figure 4.2.7 a shows the WAXD profiles acquired during a cooling run (10 ºC/min) from the melt state. It is clear that a narrow intense peak appeared around 0.420 nm during crystallization and that it progressively broadened and split into different peaks when the temperature decreased. Figure 4.2.7 b shows the deconvolution of the diffraction profile obtained at room temperature where two amorphous halos and four crystalline peaks appearing at 0.430 and 0.379 nm (form I) and 0.422 and 0.403 nm (form II) could be observed. Thus, a Brill transition seemed to occur at a high temperature during the cooling process and was followed by a phase transition from form II to form I. Conversion between these two structures is only partial since the intensities of reflections associated with form I are similar to those related to form II (e.g. a form I to form II ratio of 0.45:0.55 was determined from the deconvoluted profile). The ratio between both forms was kept practically constant even if the cooling rate was decreased. Therefore, a value of 0.47:0.53 was determined for a rate of 4 ºC/min. It is also remarkable that the profile obtained at the end of crystallization (~ 165 ºC) is identical to that recorded at the end of the heating process, which was associated with a highly ordered structure. The variation in intensity of the strongest peak (0.422-0.420 nm) could be useful in monitoring the different processes that occur on cooling, as is shown in Figure 4.2.8. Thus, this peak appeared and increased in intensity within the temperature range of 210-170 ºC, where crystallization took place. Two zones could be distinguished, i.e. 210-195 ºC and 195-170 ºC, with a quick and a slow increase in intensity, respectively. The maximum change was observed at approximately 203 ºC, in full agreement with DSC calorimetric data. Note also that the exothermic crystallization peak (Figure 4.2.2) is highly asymmetric and that the line base was only recovered when the temperature decreased to approximately 170 ºC. Thus, DSC and 99 WAXD data show a slow secondary crystallization process extending over the 195-170 ºC interval. The peak intensity remained practically constant in the temperature range between 170 and 130 ºC. Then, it slightly decreased again with decreasing temperature due to the split of the peak caused by the Brill transition. Thus, this transition occurred at a lower temperature on cooling than on heating (190 ºC), indicating a clear hysteresis effect, as usually found in polyamides.20 The transition was undetectable in the corresponding DSC cooling trace, in agreement with the lack of signal in the previous heating scan. At 100 ºC the decrease in the peak intensity was more pronounced as a consequence, in this case, of the new splitting caused by the form II to form I transition. This could not be detected by DSC because it was a partial conversion occurring in a broad temperature interval. However, it must be pointed out that, in a subsequent heating run (Figure 4.2.2) some endothermic signal was envisaged just after the glass transition. In any case, interpretation of endothermic peaks in this temperature region seems conflictive due to the overlapping with the broad endotherm corresponding to the evaporation of adsorbed water, which is clearly detected in the initial scanning run despite the sample was previously dried under vacuum. 2070120170220 T(ºC) I(ua) I (a.u.) T (ºC) 220 170 120 70 20 TB II → I Crystallization Figure 4.2.8 Temperature evolution of the peak intensity at 0.422-0.420 nm during a cooling run (10 ºC/min) from the melt state. 100  Crystallization of Nylon 65 The crystallization process was simultaneously monitored by time-resolved WAXD and SAXS non-isothermal experiments. In this way, the evolution of the mass fraction of the crystalline phase in the sample, Xc WAXD, was determined from the different WAXD deconvoluted profiles as the ratio between the total intensities of the crystalline reflections Ic and the overall intensity IT. Values at the end of crystallization ranged between 0.21 and 0.27 and increased with decreasing the cooling rate. SAXS patterns showed a long period peak at a value of the scattering vector, q = [4/] sin(), close to 0.4 nm-1 after subtraction of the empty sample background observed near the beam stop (Figure 4.2.9). This peak, which can be attributed to the lamellar structure of the spherulites, started to appear at the same temperature than crystalline reflections in the WAXD patterns (Figure 4.2.10), as presumable for a crystallization process controlled by nucleation and crystal growth. This temperature obviously decreased with increasing the cooling rate. The intensity of the SAXS peak increased during primary crystallization, then remained practically constant over a short temperature range and finally decreased. Thus, the SAXS peak practically disappeared before secondary crystallization was complete (Figure 4.2.10). This observation is important because it suggests a change in the amorphous phase since the intensity of SAXS peaks depends on the degree of crystallinity but also on the difference between the electronic densities of amorphous and crystalline phases. It is clear that on cooling the amorphous interlamellar component should adopt a more compact molecular arrangement, probably as a result of the improved hydrogen-bonding interactions. I. q2 (a.u.) q (nm-1) 250 20 T (ºC) Frame corresponding to the temperature of 188 Figure 4.2.9 Three-dimensional representation of SAXS profiles of nylon 65 during cooling (10 ºC/min) from 250 ºC (melt state) to room temperature. 101 SAXS data were analyzed by the normalized one-dimensional correlation function,31  (r), which corresponds to the Fourier transform of the Lorentz-corrected SAXS profile:  (r) =  0 2)cos()( dqqrqIq /  0 2)( dqqIq (1) The scattering intensity was extrapolated to both low and high q values using Vonk’s model32 and Porod’s law, respectively. evaluate the peak intensity evolution during crystallization (Figure 4.2.10), and morphological parameters like the long period, L  , crystalline lamellar thickness, lc, and amorphous layer thickness, la. The evolution of these parameters during crystallization (Figure 4.2.10) shows a slight change in the long period (e.g. from 11.0 to 9.0 nm in the cooling performed at 12 ºC/min), which is mainly due to the decrease in crystalline lamellar thickness (e.g. from 8.7 to 7.4 nm). The latter was more significant during the secondary crystallization step (Figure 4.2.10) and indicates that new secondary lamellae inserted into the loosely stacked bundles of primary lamellae. New lamellae suffer spatial restrictions, leading to thinner defective crystals. 0 20 40 60 80 100 120 2070120170220 Temperature (ºC) 0 5 10 15 20 25 Distance (nm) 25 20 15 10 5 XcWAXD (%), Q (a.u.) LB lc la Q Secondary crystallization Primary crystallization 12 10 8 6 4 2 0 Distance (nm)  c WAX D L  Figure 4.2.10 Temperature evolution of the long period, Lγ, crystal thickness, lc, amorphous thickness, la, scattering invariant, Q, and degree of crystallinity, ΧWAXD, during a non-isothermal melt crystallization performed at a cooling rate of 12 ºC/min. 102 Correlation functions were used to determine the scattering invariant, Q, which allows to evaluate the peak intensity evolution during crystallization (Figure 4.2.10), and morphological parameters like the long period, L  , crystalline lamellar thickness, lc, and amorphous layer thickness, la. The evolution of these parameters during crystallization (Figure 4.2.10) shows a slight change in the long period (e.g. from 11.0 to 9.0 nm in the cooling performed at 12 ºC/min), which is mainly due to the decrease in crystalline lamellar thickness (e.g. from 8.7 to 7.4 nm). The latter was more significant during the secondary crystallization step (Figure 4.2.10) and indicates that new secondary lamellae inserted into the loosely stacked bundles of primary lamellae. New lamellae suffer spatial restrictions, leading to thinner defective crystals. Figure 4.2.11 compares the correlation functions calculated for the SAXS profiles obtained at different cooling rates and at the temperature (time) corresponding to the maximum peak intensity. Differences in lamellar spacings are minimal due to the balance between two counter factors: enhanced insertion mechanism producing thinner secondary lamellae, and increased crystallization temperature resulting in thicker primary lamellae with decreasing the cooling rate. However, the slight increase observed for the lamellar spacing with the cooling run 050 100 150 200 250 300 r (Ả) γ (r) (a.u.) → 20 ºC/min 15 ºC/min 12 ºC/min 5 10 15 20 25 30 r (nm) 9.7 nm 10.0 nm 10.3 nm  ( r ) 0 1 2 Figure 4.2.11 Correlation functions corresponding to the maximum intensity SAXS profile obtained during cooling runs of nylon 65 at the indicated rates. 103 indicates the prevalence of the lamellar insertion effect. Figure 4.2.11 also shows that the L  value associated with the most probable distance between the centers of gravity of two adjacent crystals (abscise of the first maximum of the correlation function) is greater than the long period determined from twice the abscise value of the first minimum of the correlation function, which is interpreted as the most probable distance between the centers of gravity of a crystal and its adjacent amorphous layer. This indicates a broader distribution of the layer widths of the major component,33 which corresponds to the crystal phase. SAXS crystallinities,  SAXS, in the 0.80-0.83 range were calculated at the end of primary crystallization from the values of the morphological parameters (lc/(lc + la)). These crystallinities were considerably higher than those estimated from WAXD experiments, suggesting that amorphous phase domains exist between the lamellar stacks.  Spherulitic morphology of nylon 65 Spherulites of even-even nylons (e.g. nylon 66) have been widely studied and their optical properties have been interpreted.1,34,35 These polymers render negative spherulites at crystallization temperatures slightly lower than their melting point and positive spherulites at lower temperatures. The change in the optical properties is explained by a well established structure based on the stacking of hydrogen-bonded sheets. In fact, X-ray microbeam diffraction patterns suggested that positive and negative spherulites differ in the radial or tangential spherulitic direction where hydrogen bonds respectively form. Thus, the birefringence sign is directly associated with how lamellae with a single structure grow in the spherulite. However, the reason for such a drastic change in the growth mechanism at a well defined temperature remains unclear. Surprisingly, the spherulitic morphology of even-odd nylons has been little studied. A detailed phenomenological description has only been reported for nylons 49, 67 and 69 by Magill.36 In this case, rather puzzling observations suggesting a complex crystallization behavior were made. The main points of this work can be summarized as follows: The three polymers exhibited a wide variety of spherulitic structures. Thus, a sample could render spherulites with a different birefringence sign and even a different texture (e.g. fibrilar or ringed) under certain crystallization conditions. The birefringence sign often changed in the negative-positive-negative sequence with decreasing the crystallization temperature. The difference in behavior compared with the above even-even nylons, and in particular the different sign obtained at the lowest crystallization temperature, is worth noting. Microbeam diffraction patterns indicated that spherulites always had a pseudohexagonal structure. Thus, no further investigation was undertaken to relate the variability observed in the spherulitic morphologies to different crystalline structures. 104 The birefringence sign of some spherulites (e.g. nylon 49 crystallized at low undercooling) could change in a reversible way by heating and cooling processes. Isothermal crystallization of nylon 65 from the melt rendered spherulites of appreciable size over the narrow temperature range of 241-227 ºC. This crystallization proceeded according to a heterogeneous and thermal nucleation since spherulites of non-homogeneous size formed at a given crystallization temperature. The nucleation density increased exponentially with decreasing temperature in such a way that morphologies were difficult to examine at temperatures lower than 227 ºC. The induction time required for the first nuclei to be active at each temperature behaved oppositely to the nucleation density, as shown in Figure 4.2.12. Spherulites exhibited a negative birefringence sign over the studied temperature range and a ringed texture (Figure 4.2.13 a). The spacing between rings increased from ~ 0.15 nm at 238 ºC to ~ 0.30 nm at 232 ºC, where the ringed texture was better displayed. At 227 ºC the ringed texture was very difficult to be detected, although very close rings could be envisaged in the blue sectors, suggesting a trend towards a fibrillar texture. 0 100 200 300 400 500 600 700 800 900 229 234 239 244 Temperature (ºC) Nuclei /mm2 0 20 40 60 80 100 120 140 160 180 Induction Time (min) Nuclei / mm2 Nuclei / mm2 Figure 4.2.12 Variation in nucleation density and induction time with isothermal crystallization temperature. 111 Figure 4.2.16 Change of the infrared absorption bands in the 1380-1350 cm-1 (a) and 1220-1120 cm-1 ranges during a heating run of a nylon 65 sample coming directly from synthesis. 1350135513601365137013751380 Absorbance (a.u.) Wavenumber (cm-1) 112011401160118012001220 Absorbance (a.u.) Wavenumber (cm-1) 180 ºC 180 ºC 25 ºC a) b) 25 ºC 112 4.2.4 Conclusions Several conclusions can be drawn: 1. Nylon 65 samples coming directly from synthesis crystallize according to a peculiar structure (form I) that can be interpreted in terms of a packing where hydrogen bonds are established along two directions. On heating, this structure converts into a less compact structure (form II) whose X-ray diffraction pattern differs from the initial one in the closeness of the two strong equatorial reflections. A Brill transition occurs at some degrees before fusion, resulting in a pseudohexagonal chain axis projected unit cell. This high temperature structure is peculiar and is clearly different from the hexagonal arrangement found in conventional nylons since reflections related to the chain repeat have an unusual non-meridional orientation. 2. Nylon 65 crystallizes from the melt according to the high temperature structure obtained after the Brill transition. On cooling this structure transforms into form II, showing a hysteresis effect. The transition from form II to form I occurs at a lower temperature although it cannot be completed during the cooling run. 3. Crystallization from the melt gives rise to spherulites constituted by lamellae of different thicknesses, which accounts for the multiple melting peaks observed in the calorimetric heating runs. During crystallization thinner lamellae insert into the loosely stacked bundles of primary lamellae and the interlamellar amorphous regions become more compact. 4. Spherulites with different textures (ringed or fibrilar) and birefringences can be obtained by varying the crystallization conditions. Negative spherulites form in the low temperature region, indicating a molecular arrangement different from that found in conventional even-even nylons, whose low temperature spherulites show a positive birefringence. 5. Absorption bands observed in the room temperature infrared spectra suggest that nylon 65 has a structure related to conventional  /  forms despite its different hydrogenbonding scheme. Spectra are sensitive to the structural changes; specifically the temperature evolution of the amide A band allows the form I to form II transition to be detected. 113 4.2.5 References [1] Xenopoulos A, Clark ES. In Kohan MI, editor. Nylon plastics handbook. Hanser Publishers: Munich, Vienna and New York, 1995; Chapter 5. [2] Holmes DE, Bunn CW, Smith DJ. J Polym Sci Part A, General Papers 1955;17:159-77. [3] Bunn CW, Garner, EV. Proc R Soc London Ser A 1947;189:39-68. [4] Kinoshita, Y. Makromol. Chem. 1959;33:1-20. [5] Aharoni, SM. n-Nylons: their synthesis, structure and properties. John Wiley and Sons: New York, 1997. [6] Murthy NS, Aharoni SM, Szollosi AB. J Polym Sci Part B, Polym Phys 1985;23:2549-65. [7] Lincoln DM, Vaia RA. Macromolecules 2004;37:4554-61. [8] Miyasaka K, Ishikawa K. J Polym Sci Part B, Polym Phys 1968;6:1317-29. [9] Arimoto H, Ishibashi M, Hirai M. J Polym Sci Part A2 Polym Phys 1965;3:317-26. [10] Navarro E, Franco L, Subirana JA, Puiggalí J. Macromolecules 1995;28:8742-50. [11] Brill R. Makromol Chem. 1956;18:294-309. [12] Itoh T. Jpn J Appl Phys. 1976;15:2295-2311. [13] Newman BA, Sham TP, Pae KD. J Appl Phys 1976;48:4092-8. [14] Starkweather HW, Jones GA. J Polym Sci Part B, Polym Phys 1981;19:467-77. [15] Kim KG, Newman BA, Scheinbeim JI. J Polym Sci Part B, Polym Phys 1985;23:2477-82. [16] Biangardi HJ. J Macromol Sci Phys B 1990;29:139-53. [17] Hirschinger J, Miura H, Gardner KH, English AD. Macromolecules 1990;23:2153-2169. [18] Wendoloski JJ, Gardner KH, Hirschinger J, Miura H, English AD. Science 1990;247:431-36. [19] Radusch HJ, Stolp M, Androsch R. Polymer 1994;35:3568-71. [20] Ramesh C, Keller A, Eltink SJEA. Polymer 1994;35:2483-87. [21] Hill MJ, Atkins EDT. Macromolecules 1995;28:604-9. [22] Vasanthan N, Murthy NS, Bray RG. Macromolecules 1998;31:8433-5. [23] Murthy NS, Wang Z, Hsiao BS. Macromolecules 1999;32:5594-9. [24] Ramesh C, Gowd EB. Macromolecules 1999;32:3721-6. [25] Jones NA, Atkins EDT, Hill MJ. J Polym Sci Part B, Polym Phys 2000;38:1209-21. [26] Feldman AY, Wachtel E, Vaughan GBM, Weinberg A, Marom G. Macromolecules 2006;39:4455-9. [27] Rueda, DR, García-Gutiérrez MC, Nogales A, Capitán MJ, Ezquerra TA, Labrador A, Fraga E, Beltrán D, Juanhuix J, Herranz JF, Bordas, J. Rev Sci Instrum 2006, 77, Art. No. 033904 Part 1. [28] Murthy NS, Curran SA, Aharoni SM, Minor H. Macromolecules 1991;24:3215-20. [29] Hoffman JD, Weeks JJ. J Chem Phys 1962;37:1723-41. [30] Xenopoulos A, Wunderlich B. Colloid Polym Sci 1991;269:375-91. [31] Vonk CG, Kortleve G. Kolloid Z Z Polym 1967;220:19-24. [32] Vonk CG. J Appl Cryst 1975;8:340-1. [33] Hsiao BS, Wang Z, Yeh F, Yan G, Sheth KC. Polymer 1999;40:3515-23. [34] Lovinger AJ. J. Appl. Phys 1978:49:5003-5013. [35] Lovinger AJ. J. Appl. Phys 1978:49:5014-28. 114 [36] Magill JH. J. Polym. Sci. part A 1969;7:123-142. [37] Sibila JP, Sanjeeva NS, Gabriel MK, McDonnell ME, Bray RG, Curran SA. In Nylon Plastics Handbook; Kohan MI Ed.; Hanser Publishers: Munich, Vienna and New York, 1995; Chapter 4. [38] Skrovanek DJ, Painter PC, Coleman MM. Macromolecules 1986,19:699. 115 4.3 Crystallization studies on clay nanocomposites of nylon 47 having exfoliated or intercalated structures Basic structural data on nylon 47 were obtained from X-ray diffraction of powder, and fiber samples and electron diffraction of thin spherulitic samples. The studied even-odd polyamide was characterized by a peculiar structure where hydrogen bonds were established along two directions. Nylon 47 showed reversible polymorphic transitions on heating/cooling processes that were analyzed by real time synchrotron WAXD experiments. Results indicate that nylon 47 had a first structural transition at low temperature and then underwent a gradual Brill transition towards a pseudohexagonal packing. Optical and electron microscopy studies were also performed under isothermal conditions to distinguish the different spherulitic morphologies and changes on optical properties. Results revealed a different behaviour from that of spherulites of conventional even-even nylons. Interestingly, spherulites crystallized under a low supercooling had a reversible change of birefringence with temperature. This was due to the peculiar morphology attained at high temperature and the reversible structural changes that take place with temperature. Intercalated and exfoliated nanocomposites based of nylon 47 were prepared by using Cloisites 25A and 30B, and different preparation methods (i.e. solution intercalation and melt mixing). The influence of the final silicate layer morphology on the hot crystallization behaviour was investigated by optical microscopy and differential scanning calorimetry. Crystallization rates of the neat polymer and its two nanocomposites were significantly different, mainly as a consequence of variations on primary nucleation. 116 4.3.1 Introduction Aliphatic polyamides (nylons) constitute a family of polymers with exceptional properties because of their capability to establish strong intermolecular hydrogen bonding interactions.1 Molecular conformation and packing preferences are basically conditioned to favour a close arrangement between amide groups. In this way, the crystalline structure of conventional eveneven nylons (e.g. nylons 66 and 6-10) are based on a stacking of sheets composed of hydrogenbonded molecular chains with a planar zig-zag conformation ( and  forms).1,2 Similar arrangements with hydrogen bonds established along a single direction are also commonly found in some even nylons (i.e. nylon 6).1,3 The corresponding X-ray fiber diffraction patterns of such structures are characterized by the presence of two strong equatorial reflections at spacings close to 0.440 and 0.380 nm which are associated to intrasheet and intersheet spacings, respectively. Nylons can however crystallize according to other arrangements and molecular conformations depending on the parity of the constituent monomers (e.g. the pseudohexagonal  phase postulated for some odd-odd nylons1,4) and indeed the presence of special units like glycine5,6 (e.g. nylons 2/3, 2/6) and malonic acid7 (e.g. nylons n,3). Furthermore, nylons can experiment phase transitions during heating and cooling processes, as for example the not completely well understood reversible structural change detected with nylons having conventional / forms at room temperature.8-18 In this case, the evolution of the diffraction patterns on heating shows that the two characteristic equatorial reflections gradually merge on a single one (ca. 0.421 nm) indicative of a pseudohexagonal arrangement at the so called Brill transition temperature. On cooling from the melt state, the polymer firstly crystallized in the indicated pseudohexagonal packing and then the characteristic single reflection splits into the two above indicated reflections at a temperature lower than observed in the heating process. A peculiar structure based on the establishment of hydrogen bonds along two different directions has lately been postulated for some even-odd and odd-even nylons (e.g. nylons 69,19 65,20,21 12-5,22 56,23,24 5-1025 and 9226) which fibers rendered two strong equatorial reflections at similar spacings than reported for the / conventional structures. The new structure was postulated since good intermolecular hydrogen bonding interactions could not be established when nylons had a planar zig-zag molecular conformation and were derived from diamine and dicarboxylic acid units with different parity (i.e. even-odd and odd-even nylons). This feature is illustrated in Figure 1a for nylon 47 which is the polymer object of the present work. 117 X b) a) Figure 4.3.1 a) Scheme showing the unfavourable hydrogen-bonding geometry between pimelamide units having an all trans conformation. b) Scheme showing as hydrogen bonds could be well established along two directions when the two amide planes of the pimelamide unit rotate in opposite directions from the plane defined by its methylene carbons. External chains (stick representation) should be shifted along the chain axis direction (see arrows) with respect to the central chain (ball and stick representation), thus giving rise to a monoclinic unit cell. Color code: nitrogen, blue; oxygen, red; carbon, gray; hydrogen, brown. 118 Basically, the distinctive feature of the new molecular arrangement is the capability of establishing good hydrogen bonding interactions along two different directions with molecules having a practically all trans conformation. Thus, a slight deviation towards 150º (or -150º) for the two torsional angles vicinal to the odd diamide units seems necessary to face all NH and CO groups of neighbouring chains. The two amide groups of the odd unit rotated in opposite senses from the plane defined by the methylene carbon atoms allowing a good hydrogen bonding geometry when neighbouring chains became conveniently shifted along the chain axis direction (Figure 4.3.1). In this way, a monoclinic unit cell containing two molecular segments was derived and the chain axis projection corresponded to a rectangular unit cell.19-26 Polymers filled with a layered clay (or phyllosilicate) give rise to microstructural dispersions different from conventional ones obtained from inorganic fillers. Incorporation of nanoelements may provide commodity materials with a suite of characteristics that organic chemistry and traditional polymer-blending approaches cannot supply from an economical point of view.27-29 Pioneering works concerning clay nanocomposites were precisely carried out with polyamides having the conventional sheet structure (e.g. nylon 6).30,31 Crystallization processes are also influenced by the incorporation of phyllosilicate particles since they may have an impact on the overall crystallization rate, crystal growth, nucleation type and morphological features. In fact, crystallization is determined by different factors which may be favoured or disfavoured when the clay particles are incorporated. In this way, published results suggested that the crystallization process highly depends of the type of clay dispersion (e.g. exfoliated or intercalated) and even on the interactions between clay and polymer matrix.32-35 It seems therefore interesting to bring new data on the crystallization behaviour of composites based on polyamides having different intermolecular interactions than conventional nylons. In this way, the present work is focused in the structural characterization of a new even-odd polyamide (i.e. nylon 47), the preparation of nanocomposites with different structures and finally the evaluation of the crystallization behaviour of the neat polymer and its nanocomposites. 4.3.2 Experimental section  Materials Nylon 47 was synthesized by interfacial polycondensation of 1,4-diaminobutane and pimeloyl chloride using toluene as organic solvent and sodium hydroxide as proton acceptor following the procedure previously described for similar nylons.20 The polymer was purified by precipitation with water of a formic acid solution. Nylon 47 was obtained with a yield of 55% and an intrinsic viscosity of 0.85 dL/g (determined in dichloroacetic acid at 25 ºC). 119 Dimethyl hydrogenated-tallow 2-ethylhexyl ammonium montmorillonite (Cloisite 25A, Southern Clay Products, 2MHTEX) and methyl tallow bis(2-hydroxyethyl) ammonium montmorillonite (Cloisite 30B, Southern Clay Products, MT2EH) (tallow (65% C18, 30% C16, 5% C14)) were used as received. The chemical structure of the specific surfactant of the organo-modified layered phyllosilicates are shown in Table 4.3.1. Table 4.3.1 Characteristics of Organoclaysa. Clay Type Chemical Structure of Organic Modifier Cloisite 30B Cloisite 25A a HT is the hydrogenated-tallow. T 65% C18, 30% C16, 5% C14.  Preparation of nanocomposites Nanocomposites containing 3% of C25A or C30B clay particles were prepared by melt mixing in two steps using a co-rotating tightly intermeshed twin-screw extruder (DSM Xplore 5ml microcompounder). All materials were dried under vacuum prior to mixing. The processing temperature, screw rotation and cycle time were 260 ºC, 100 rpm and 3 minutes, respectively. Alternatively, nanocomposites were also prepared by the solution-intercalation film-casting technique. For each final nanocomposite composition 100 mg of nylon 47 was dissolved in 10 mL of 1,1,1,3,3,3-hexafluoroisopropanol. Clay dispersions (<0.1 wt %) were obtained by suspension of clay in a separate beaker of 1,1,1,3,3,3-hexafluroisopropanol. Both the Nylon 47 solution and clay suspension were agitated separately for 30 min. The final mixture was further sonicated for 120 min with a Sonorex Super 10P sonicator. The amount of OMMT loading was fixed at 3 wt%. The mixture was then cast on a glass surface and the solvent was removed in a vacuum oven at 40 ºC. Eventually, optically clear nanocomposite films with thicknesses ranging from 20 to 35 m were obtained. N+ T CH2CH2OH CH2CH2OH H3C N+ CH3 CH2CHCH2CH2CH2CH3 HT H3C CH2CH3 120  Measurements X-ray fiber and power diffraction patterns of nylon 47 were obtained with Ni-filtered CuK radiation of wavelength 0.1542 nm from an Enraf Nonius rotating anode X-ray generator and using a modified Statton camera (W. H. Warhus, Wilmington, DE). Oriented fiber samples were obtained by melt drawing. Time resolved WAXD experiments were carried out at the CRG beamline (BM16) of the European Synchrotron Radiation Facility of Grenoble. The beam was monochromatized to a wavelength of 0.098 nm. Polymer samples were confined between Kapton films and then held on a Linkam hot stage with temperature control within  0.1 ºC. WAXD profiles were acquired during heating and cooling runs in time frames of 12 s and rates of 3 ºC/min. The WAXD detector was calibrated with diffractions of a standard of an alumina (Al2O3) sample. The diffraction profiles were normalized to the beam intensity and corrected considering the empty sample background. Deconvolution of WAXD peaks was performed with the PeakFit v4 program by Jandel Scientific Software using a mathematical function known as “Gaussian area”. Spherulites of nylon 47 were grown from homogeneous melt-crystallized thin films placed between two cover glasses. These films were produced by evaporation of a dilute solution of the polymer in 1,1,1,6,6,6-hexafluoroisopropanol. Samples were crystallized isothermally at different temperatures below the melting point using a Linkam temperature control system configured by a THMS 600 heating and freezing stage connected to an LNP 94 liquid nitrogen cooling system. Additionally, non-isothermal experiments were carried out at cooling/heating rates of 1 ºC/min. The experimental procedure allowed films with thickness lower than 10 m to be obtained. Optical photographs were taken using a Zeiss AxioCam MRC5 digital camera mounted on a Zeiss Axioskop 40 Pol light polarizing microscope. A first-order red tint plate was employed to determine the sign of spherulite birefringence under crossed polarizers. Thin spherulites were also observed with a Philips TECNAI 10 electron microscope operating at 80 and 100 kV for bright field and electron diffraction modes, respectively. After manual separation of the two glasses, the spherulites attached to the cover-slip were covered with a thin carbon film, floated off on water, picked up on copper grips and shadowed with Pt-Carbon at an angle of 15º. Bright field micrographs were taken with a SIS MegaView II digital camera. Selected area electron diffraction patterns were recorded on Maco EM films from not shadowed samples. The patterns were internally calibrated with gold (d111 = 0.235 nm). Interlayer spacing of clay nanocomposites was studied by wide angle X-ray scattering (WAXD) using a PANalytical X´Pert diffractometer with Cu Kα radiation (λ = 0.1542 nm) using a silicium monocrystal sample holder. The structure and distribution of Cloisite in the 127 11 12 13 14 15 16 17 18 19 I(a.u.) Q(nm-1) Figure 4.3.5 a) Onedimensional WAXD profiles of nylon 47 taken at room temperature (down) and just before starting the melting process (middle) and at the end of the cooling process (up). Deconvoluted Bragg and amorphous peaks are showed for both profiles. b) Onedimensional WAXD profiles of nylon 47 taken at selected temperatures during heating and cooling scans (3 ºC/min). Spacings of main reflections associated to forms II and I are indicated. 0.443 nm 0.432 nm 0.422 nm 0.414 nm Bragg 0.430 nm 0.386 nm 0.408 nm 0.349 nm Amorphous 0.345 nm Amorphous Phase a) Bragg Bragg Amorphous Phase 0.433 nm 0.345 nm 0.430 nm 0.415 nm 0.387 nm 11 13 15 17 19 I (a.u.) Q 31 ºC 38 ºC 98 ºC 197 ºC 231 ºC 110 ºC Heating Cooling 0.430 nm 0.389 nm 0.407 nm 0.422 nm Mel Q (nm -1) b) q q 128 Figure 4.3.6 a shows the WAXD profiles acquired during a cooling run (3 ºC/min) from the melt state. It is clear that a narrow intense peak appeared around 0.420 nm during crystallization and that it progressively broadened and split into different peaks when the temperature decreased. The variation in intensity of the strongest peak (0.422-0.420 nm) was useful in monitoring the different processes that occur on cooling, as is shown in Figure 4.3.6 b. Thus, this peak appeared and increased in intensity within the temperature range of 240-197 ºC, where crystallization took place. The peak intensity remained practically constant in the temperature range between 197 and 168 ºC and then, it slightly decreased again with decreasing temperature due to the split of the peak caused by the Brill transition. Thus, this transition occurred at a lower temperature on cooling than on heating (190 ºC), indicating a clear hysteresis effect, as usually found in polyamides.16 At 130 ºC the decrease in the peak intensity was more pronounced as a consequence, in this case, of the new splitting caused by the form II to form I transition. This could not be detected by DSC because it was a partial conversion occurring in a broad temperature interval. Structural changes could not take place when temperature became close to the glass transition temperature and consequently the peak intensity was constant at temperatures lower than 65 ºC. It is important to note that form I could not be completely recovered during the cooling process and that the X-ray profile attained at room temperature corresponded to a mixture of structures. Thus, the final diffraction profile became clearly different to that obtained from the as-synthesized sample. Figure 4.3.5 a compares also the deconvoluted profiles taken at room temperature with the initial and the hot crystallized samples. Specifically, the major difference correspond to the peak close to 0.415-0.408 nm that appeared with a different intensity. It may correspond to the weak 111 reflection of form I and also to the characteristic strong reflection of form II. Logically the intensity of the peak at 0.415-0.407 nm will depend on the ratio between form I and form II that is achieved after the cooling process. Figure 4.3.5 a shows also as the amorphous halos slightly changed with temperature as a consequence of thermal dilatation and specifically peaks shifted to higher a spacing with increasing temperature. Figure 4.3.5 b shows for the sake of completeness some representative X-ray profiles taken during heating and cooling runs. Specifically, those corresponding to the achievement of form II (110 ºC on heating and 98 ºC on cooling), those corresponding to the similar structures attained on heating just before melting (231 º) and on cooling at the end of crystallization (206 ºC), and finally the dissimilar profiles taken at room temperature with the as-synthesized sample before starting and at the end of the process, respectively. 129  Spherulitic morphologies of nylon 47 Scarce works concern the study of optical properties and morphologic characteristics of spherulites obtained from even-odd nylons. A detailed phenomenological description has been reported for nylons 49, 67 and 69 by Magill38 and some additional data has also recently been given for nylon 65.37 Rather puzzling observations were reported but in general a wide variety of spherulitic structures with different optical birefringence were found. The birefringence sign often changed in the negative-positive-negative sequence with decreasing the crystallization temperature. This behavior was clearly different to that reported for even-even nylons which are characterized by a stacking of hydrogen-bonded sheets. In this case, polymers rendered negative (nm -1) a) b) 050100150200250300 I (a.u.) T (ºC) Crystallization II → I TB Q (nm-1) q Figure 4.3.6 a) Three-dimensional representation of WAXD profiles of nylon 47 during cooling (3 ºC/min) from the melt to room temperature. b) Temperature evolution of the peak intensity at ca. 0.422 nm during a cooling run (3 ºC/min) from the melt state. 130 spherulites at crystallization temperatures slightly lower than their melting point and positive spherulites at lower temperatures. The change in the optical properties was explained assuming that positive and negative spherulites differ in the radial or tangential spherulitic direction where hydrogen bonds respectively form. Thus, the birefringence sign is directly associated with how lamellae with a single structure grow in the spherulite.38,39 However, the reason for such a drastic change in the growth mechanism at a well defined temperature remains unclear. Isothermal crystallizations of nylon 47 from the melt state rendered different kinds of spherulites depending specifically on the selected crystallization temperature. Thus, peculiar spherulites with well defined crystalline domains and a negative birefringence were observed (Figure 4.3.7 a) when crystallization temperature was higher than 228 ºC. Electron micrographs (Figure 4.3.8) revealed a complex morphology where an arborescent growth with curved arms develop from the primary nucleus (Figure 4.3.8 a and b). Micrographs taken at higher magnifications showed the flat appearance of constitutive crystals (Figure 4.3.8 c) and indeed the presence of single crystals which in some cases had a lozenge shape (e.g. see the dashed area of Figure 4.3.8 d). This crystals gave rise to hk0 electron diffraction patterns with intense reflections (Figure 4.3.2 b) that corresponded to form I as expected since patterns were recorded at at room temperature . Heating Isothermal 232 ºC Non-isothermal 232 ºC 25 ºC Isothermal 232 ºC Isothermal 220 ºC Isothermal 228 ºC a) b) c) d) 25 m Figure 4.3.7 Optical micrographs taken at the respective crystallization temperature of nylon 47 spherulites obtained at 232 ºC (a), 228 (b) and 220 ºC (c). d) Optical micrographs taken at room temperature and after heating up to 232 ºC of a nylon 47 spherulite that was firstly isothermally crystallized at 232 ºC and then non-isothermally crystallized until room temperature at a cooling rate of 1 ºC/min. 131 At lower temperatures than 228 ºC spherulites tended to have a fibrilar texture and kept a negative birefringence. In fact, ringed spherulites with a small inter-ring spacing could be envisaged in the optical micrographs when temperature was close to 228 ºC (Figure 4.3.7 c). Electron micrographs clearly revealed (Figure 4.3.9 a) that samples crystallized at 218 ºC had a regular banding produced by a lamellar twisting with a spacing close in this case to 0.5 m. Bright zones consisted of lamellae lying practically flat whereas dark zones were associated to lamellae standing on edge. Spherulites were sufficiently thin to get hk0 electron diffraction patterns (e.g. Figure 4.3.2 a) from zones corresponding to the flat lamellae (Figure 4.3.9 b). Again these patterns corresponded to form I (i.e. the most stable structure at room temperature). It is also highly interesting that spherulites grown at the specific temperature of 228 ºC were clearly not birefringent (Figure 4.3.7 b). 2.5 m 1 m 4 2 m a) b) c) d) Figure 4.3.8 Transmission electron micrographs of nylon 47 spherulites obtained by isothermal crystallization at 232 ºC. Different magnifications are given to show specific morphologic details. Dashed area in d) shows the presence of lozenge crystals (see arrows). 132 Figure 4.3.7 d shows polarized optical micrographs taken at room temperature and at 232 ºC (i.e. after a subsequent heating) of a spherulite that was firstly isothermally crystallized at 232 ºC and then non-isothermall crystallized by cooling up to room temperature at a rate of 0.5 ºC/min. It is clear that the central zone isothermally crystallized and constituted by flat crystal domains have a reversible change of the birefringence sign with temperature. Note that it was negative at the crystallization temperature (Figure 4.3.7 a), became positive at room temperature and finally negative after the heating process. By contrast the outer nonisothermally crystallized zone was characterized by a fibrilar texture and had always a negative birefringence sign. This feature was clearly distinctive to typical even-even nylons which show a positive birefringence when they were crystallized at low temperatures. The reversible birefringence change observed for nylon 47 seems a consequence of a structural transition (from the Brill or form II structure to the form I expected at high and low temperatures, respectively). Probably, the structural transition involved small changes on the molecular conformation and a variation on the angle defined by the two hydrogen bonding directions. This change on the packing mode should greatly influence on optical properties when crystals had a flat disposition that emphasized the molecular and hydrogen bonding arrangements. In this way, reversible changes could only be envisaged in the spherulitic textures developed at higher temperatures (i.e. > 228 ºC).  Dispersion structure of C25A and C30B organomodified clays in their nanocomposites with nylon 47 The nanocomposite structures were first analyzed by reflection X-ray diffraction of film samples. The diffraction patterns show evidence of intercalation of polymer chains into the silicate galleries in the range of 2  = 1-10º (  is the scattering angle) when the characteristic 001 silicate diffraction peak appears at a lower diffraction angle (larger spacing) than in the pattern of the neat clay. Similarly, the absence of this peak may suggest an exfoliated structure. Cloisite C30B has hydroxyl polar groups (Table 1) that can interact with the amide groups of nylon 47 and consequently can favour the achievement of an exfoliated structure. This structure was found to be enhanced when nanocomposites were prepared by the melt mixing technique instead of solution-intercalation, probably as a consequence of the high melting temperature of nylon 47 that could favour the exfoliation of silicate layers. Hence, nylon 47/C25A and nylon 47/C30B nanocomposites obtained by solution-intercalation and melt mixing methods, respectively, were selected as the best preparations corresponding to intercalated and exfoliated structures. Figure 4.3.10 a shows the shift of the characteristic 001 reflection of the C25A clay from 1.94 nm (2  = 4.55º) to 2.25 nm (2  = 3.93º) when the nanocomposite sample was prepared. This 133 observation indicates the achievement of a regular intercalated structure with an increase of the interlayer spacing caused by the insertion of polymer molecular chains. Direct observation of the morphology and phase distribution of ultrathin sections of nylon 47 / C25A specimens by transmission electron microscopy (Figure 4.3.10 a) revealed the presence of well-ordered layered structures as presumable when an intercalated structure was predominant. Figure 4.3.10 b clearly demonstrates that an exfoliated structure was characteristic of the nylon 47/C30B nanocomposite. Thus, the 001 reflection of the C30B clay (1.80 nm, 2  = 4.91º) disappeared in the X-ray profile of the nanocomposite, whereas TEM micrographs revealed the presence of practically dispersed silicate layers. The structure of nylon 47 changed also according to the nanocomposite preparation method as shown in the X-ray diffractograms of Figure 4.3.11. Hence, samples obtained from solution showed the characteristic reflections of form I (e.g. those appearing at 0.430 and 0.388 nm), whereas samples coming from the melt had reflections of both structures: form I (e.g. 0.388 nm) and form II (e.g. 0.419 nm). Note also that in this case, the 002 reflection shifted to a lower spacing (i.e. 0.696 nm) which may indicate a shortening of the chain axis repeat or alternatively an increase of the shift between neighbouring chains along the c crystallographic axis. 1 a) b) 0.5 m Figure 4.3.9 a) Transmission electron micrograph of nylon 47 spherulites obtained by isothermal crystallization at 218 ºC. b) Specific zone of a non shadowed sample where a hk0 electron diffraction pattern of form I was obtained. 134 150 nm 150 nm a) b) -300 1200 2700 4200 5700 2 4 6 8 10 12 q (nm-1) Intensity (a.u.) 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 2 4 6 8 10 12 q (nm-1) Intensity (a.u.) 1.80 nm 1.94 nm 2.25 nm C30B C25A 2  (degrees) 2  (degrees) Figure 4.3.10 Transmission electron micrographs and X-ray diffraction patterns showing the morphology and the zone corresponding to the 001 clay reflection of the Nylon 47/C25A (a) and Nylon 47/C30B (b) nanocomposites with a Cloisite concentration of 3%. X-ray diffraction patterns of the neat clays are also shown. 0 1000 2000 3000 4000 5000 913 17 21 25 29 2  (degrees) Intensity (a.u.) 0.430 nm 0.388 nm 0.419 nm 0.406 nm 0.735 nm 0.696 nm C25A C30B Figure 4.3.11 X-ray diffraction profiles showing the main reflections of nylon 47 in nanocomposites with Cloisite 25A and 30B prepared by solvent casting and melt mixing methods, respectively. 135  Influence of C25A and C30B clay particles in the thermal behaviour of nylon 47 Calorimetric analyses showed differences between the non-isothermal crystallization of the neat polymer and its nanocomposites with the C25A and C30B clays. Thus, the peak crystallization temperature increased and the exothermic peak became broader respect to the neat polymer when the nanocomposite had an intercalated structure as shown in Figure 4.3.10 for a representative cooling rate (i.e. 10 ºC/min). Crystallization was consequently favoured by the presence on the layered structure (nucleation effect) although the corresponding crystallization rate decreased. The exfoliated structure led to a disfavoured crystallization process due to both the decrease of the peak crystallization temperature and the slight peak broadening. Differences on the crystallization process were clearer when isothermal experiments (Figure 4.3.13) were carried out. Note that in this case, the crystallization peak of nanocomposites with an exfoliated structure appeared at later times than required for the neat polymer, and also that this peak became clearly broader. Differences were enhanced when experiments were conducted at the higher temperatures (i.e. when the overall crystallization rate was slower). Figure 4.3.12 DSC cooling runs (10 ºC/min) from the melt state of a nylon 47/C25A (●), nylon 47 (▲) and nylon 47/C30B (■) samples. Temperature (ºC) Heat Flow (a.u.) -3 -2.5 -2 -1.5 -1 -0.5 0 190 200 210 220 230 240 250 136 Crystallization exotherms started earlier when nanocomposites had the intercalated structure as expected for a favoured nucleation. However, the overall crystallization rate decreased again respect to the neat polyamide since the corresponding exothermic peak became slightly broader. In summary, both isothermal and non-isothermal crystallizations suggest that primary nucleation was enhanced or disfavoured when intercalated or exfoliated silicate layers were respectively present. The crystallization rate decreased in both cases suggesting that the crystal growth was hindered as will be described in the next section. Note that the neat polymer and its nanocomposite with the C25A clay had similar DSC crystallization curves (Figure 4.3.10 and Figure 4.3.11) due to the two opposing effects observed for the intercalated nanocomposites.  Optical microscopy studies on nylon 47/C25A and nylon47/C30B nanocomposites Accurate measurements of the evolution of the spherulitic radius with crystallization time were only feasible in the restricted temperature range between 218 and 243 ºC. Thus, the kinetic analysis was limited to a region close to the polymer melting point and consequently crystallization was mainly governed by the secondary nucleation process. Figure 4.3.13 DSC curves of isothermal crystallization at selected temperatures of the nylon 47/C30B (dashed lines) and nylon 47/C25A (solid lines) nanocomposites. The inset compares the isothermal curves at 230 ºC of the neat polymer (▲) and its nanocomposites with C30B (■) and C25A (●). -0.35 -0.25 -0.15 -0.05 0.05 0.15 0.25 0.35 010 20 30 40 50 60 0.15 0.25 0.35 020 40 60 Heat Flow (W/g) Time (min) 230 ºC 224 ºC 230 ºC 226 ºC 228 ºC Temperature (ºC) Heat Flow (a.u.) 239 crystallinity since, owing to possible distortions in the crystal lattice and thermal disorder, the measured value of Ic might underestimate the true value of crystallinity. The time evolution of WAXD crystallinity at the isothermal crystallization temperature of 50 ºC is displayed in Figure 5.4.10 b for the neat polyester and its nanocomposite. It is clear that the overall crystallization proceeded faster in the neat polyester and also that a higher degree of crystallinity was attained in this sample (45% versus 33%). These absolute crystallinities did not change significantly (from 45 to 47% and from 33 to 35%) with crystallization temperature in the studied range of 45-55 ºC for the neat polyester and 48-54 ºC for the nanocomposite sample. There was also good agreement between the evolution of the SAXS invariant, Q, and the degree of crystallinity evaluated by WAXD for both samples. Figure 5.4.9 a) Time evolution of main morphological parameters and the invariant during isothermal crystallization at 54 ºC of poly(glc-alt-6HH)/C25A. b) Final lamellar spacings at various crystallization temperatures for the neat polyester (empty symbols) and its nanocomposite (full symbols). 0 40 80 120 160 0 500 1000 1500 2000 2500 t-t 0 (s) Distance (Å) 0 0.001 0.002 0.003 0.004 Invariant Q (a.u.) Primary crystallization t1 t2 LB L  lc la a) b) 0 20 40 60 80 100 120 140 40 45 50 55 60 Temperature (ºC) Distance ( Å ) Lγ la lc Q 240 0 400 800 1200 1600 33.544.55 d (Å) I(a.u.) 33.544.554 a) (11 ) (110) (020) (021) (022 + 111) clay 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 500 1000 1500 tt 0 (s) X c WAXD b) Combinated SAXS and WAXD data can be used to verify the assignment of la and lc thicknesses, which cannot be distinguished from the analysis of the correlation function.41-43 Thus, the ratio between XcWAXD and XcSAXS is an estimate of the volume-filling fraction of the lamellar stacks, XS. Ratios of 0.58 and 0.39 were determined for the neat polyester and the nanocomposite at the end of crystallization, respectively. Note that the opposite assignment of the amorphous and crystalline layer thicknesses should render an unrealistic ratio greater than one. The determined ratios point to the existence of amorphous domains between lamellar stacks of the studied samples, which appear less significant for the neat polyester. Figure 5.4.10 a) Deconvolution of the WAXD profile corresponding to the end of isothermal crystallization at 50ºC of the nanocomposite. Inset shows the profiles obtained after eliminatio n of the amorphous halo in the early (850 s) and final (2900 s) stages of this crystallization. b) Evolution of crystallinity determined from WAXD data during isothermal crystallization a t 50ºC of the neat polyester (▲) and its nanocomposite (□). 241 0 0.1 0.2 0.3 0.4 0.5 10001100120013001400 Wavenumber (cm -1 ) Absorbance 0 (min) 40 (min) 142 0 1234 1222 1196 128 2  Isothermal crystallization kinetics of poly(glc-alt-6HH) and its C25A nanocomposite from FTIR analyses FTIR is highly sensitive to molecular conformation and packing density, hence its usefulness in polymer crystallization studies. Characteristic bands can be correlated to the crystalline and amorphous phases of the bulk and typically remain distinguishable over the course of crystallization. Isothermal studies are preferred to avoid shape susceptibility and intensity of FTIR bands with temperature. Figure 5.4.11 compares the absorption infrared spectra (1050-1000 cm-1) of the neat polyester in the molten state and at the end of an isothermal crystallization performed at 50 ºC. It is clear that different bands in this region can be assigned to the crystalline (1420, 1234 and 1222 cm-1) or the amorphous phase (1282 and 1196 cm-1) since their absorptions increase or decrease, respectively, during the crystallization process. The continuous evolution of the absorption bands is shown in Figure 5.4.12 for a representative example corresponding to the isothermal crystallization at 50 ºC of the neat polyester. The opposite behavior (not shown) is logically observed during a heating run of a semicrystalline sample. Figure 5.4.13a shows that the five selected bands exhibit a similar behavior for a given sample and crystallization temperature. Note that in all cases absorption values started and stopped changing at the same crystallization time and that the maximum absorption change was also detected at a similar time. Thus, it seems feasible to perform a crystallization kinetic analysis by considering the evolution of these absorption bands. Figure 5.4.11 Absorption FTIR spectra (1500-1000 cm-1) of the nea t polyester at the beginning and the end of isothermal crystallization a t 50ºC. 242 0.05 0.09 0.13 0.17 12501270129013101330 Absorbance Wavenumber (cm -1 ) b) Figure 5.4.12 Changes in the infrared absorption bands at 1420 cm-1 (a), 1282 cm-1 (b) and 1234, 1222 and 1196 cm-1 (c) of the neat polyester during isothermal crystallization at 50ºC. 0.1 0.16 0.22 0.28 1182120212221242 Absorbance Wavenumber (cm -1 ) c) 0.05 0.09 0.13 0.17 134013901440 Absorbance Wavenumber (cm -1 ) a) 243 Figure 5.4.13 a) Time evolution of the absorption of selected infrared bands for isothermal crystallization at 50ºC of the neat polyester. Characteristic bands of the amorphous (1282 and 1196 cm-1) and crystalline phases (1420, 1234 and 1222 cm-1) were chosen. b) Time evolution of relative crystallinity deduced from FTIR data (1420 cm-1 b and) for isothermal crystallization at 50ºC of the neat polyester (▲) and its nanocomposite (♦). c) Avrami plot considering FTIR data (1420 cm-1 band) for isothermal crystallization at 50ºC of the neat polyester (▲) and its nanocomposite (♦). -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.8 1.0 1.2 1.4 1.6 log (t-t 0 ) log[-ln(1-  (t)] c) a) b) 0 0.2 0.4 0.6 0.8 1 0 10203040 Relative Crystallinity t-t 0 (min) 0.1 0.14 0.18 0.22 0.26 0.3 0 10203040 Absorbance t (min) 1420 1282 1234 1222 1196 244 A relative degree of crystallinity,  (t), can be defined and measured for any characteristic absorption band associated with the crystalline (equation 5.4.2) or the amorphous phase (equation 5.4.3):  (t) = (At - A0) / (A  - A0) (5.4.2)  (t) = (At - A  ) / (A0 - A  ) (5.4.3) where At is the absorption measured at a crystallization time t, and A0 and A  are the initial and final absorptions of the considered band. Figure 5.4.13 b compares the evolution of crystallinity, evaluated through absorption measurements of the band at 1420 cm-1 for the neat polyester and the nanocomposite at a crystallization temperature of 50 ºC. It is again clear that crystallization of the neat polyester proceeds faster. Kinetic crystallization data were analyzed by the Avrami equation44-46 for primary crystallization, i.e.: 1 -  (t) = exp[-Z(t-t0)n] (5.4.4) where Z is the temperature-dependent rate constant and n the Avrami exponent, whose value varies according to the crystallization mechanism. This mechanism is composed of two steps: nucleation, which can be either homogeneous or heterogeneous depending on how nuclei are formed, and crystal growth geometry. A normalized rate constant, k = Z1/n, is usually evaluated for comparison purposes since its dimension (time-1) is independent of the Avrami exponent value. Plots of log {-ln[1-  (t)]} against log (t - t0) for a given sample and crystallization temperature give straight lines (Figure 5.4.13 c) with slopes corresponding to the Avrami exponents and intercepts at log (t - t0) = 0 equal to log Z. The results were practically independent of the absorption band considered, as can be seen in Table 5.4.3. Avrami exponents of both samples were practically constant in the studied range of crystallization temperatures. Furthermore, the exponents did not change significantly with the addition of nanoparticles, suggesting that crystal growth nucleation and dimensionality remained unaffected. The average value of 2.16-2.12 suggests a predetermined (heterogeneous) nucleation with spherical growth geometry, the ideal of n for such a situation being 3. The alternative interpretation of n = 3, i.e. sporadic (homogeneous) nucleation and disk-like growth geometry, was discarded since this is expected to occur at large supercoolings and for thin films. Optical microscopy observations indicated a spherulite growth and an athermal nucleation, in full agreement with the given interpretation of the Avrami exponent. It should also be noted that the Avrami exponent reflects the geometry of spherulites during growth and can be affected by numerous factors. Several explanations have been provided to justify such a fractional Avrami exponent (e.g. truncation effects between spherulites). 245 Figure 5.4.14 shows that a good correlation between the overall crystallization rate and the reciprocal of the crystallization half time,  1/2, is found for the neat polyester and its nanocomposite. These time values, summarized in Table 5.4.2 for all experiments, can be easily estimated from the conversion curves. Note that they were measured without using kinetic equations; hence, the indicated fit demonstrates the goodness of the Avrami analysis. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 49 50 51 52 53 54 55 56 57 Temperature (ºC) k (min -1 ) 0 0.04 0.08 0.12 0.16 0.2  1/2 -1 (min) -1 Figure 5.4.14 Comparison between the overall crystallization rate (□) and the reciprocal of the crystallization half time (○) determined at different isothermal crystallizatio n temperatures for the neat polyester (full symbols) and its nanocomposite (empty symbols). 246 Table 5.4.3 . Kinetic parameters derived from the FTIR analysis of the isothermal hot crystallization of the neat poly(glc-alt-6HH) sample and its nanocomposite with the C25A clay. Temperature (ºC) Parameter Neat Polyester Nanocomposite Wavenumber (cm-1) Wavenumber (cm-1) 1420 1234 1222 1196 Average 1420 1234 1222 1196 Average 50 n 2.38 2.33 2.79 1.71 2.30 2.53 2.54 2.52 1.75 2.34 1/2 (min) 5.54 5.60 6.14 5.54 5.70 12.02 11.86 11.32 12.02 11.81 k (min-1) 0.15 0.14 0.15 0.13 0.14 0.074 0.074 0.076 0.078 0.076 52 n 2.38 2.36 2.39 2.27 2.35 2.37 2.38 2.66 1.83 2.31 1/2 (min) 10.74 11.02 10.821 11.56 11.04 15.57 16.53 16.51 16.13 16.18 k (min-1) 0.075 0.074 0.074 0.071 0.073 0.053 0.058 0.056 0.049 0.054 54 n 2.09 2.07 2.13 2.06 2.09 1.66 1.65 1.66 1.54 1.62 1/2 (min) 16.53 16.67 15.94 16.98 16.53 21.77 21.39 21.18 23.83 22.04 k (min-1) 0.049 0.048 0.050 0.047 0.0485 0.033 0.034 0.034 0.030 0.033 56 n 1.74 2.05 1.80 2.41 2.00 2.10 2.56 2.17 1.97 2.20 1/2 (min) 30.58 31.58 30.20 37.51 32.47 36.85 36.52 36.25 37.80 36.86 k (min-1) 0.033 0.034 0.040 0.028 0.034 0.027 0.026 0.029 0.024 0.027 247 5.4.4 Conclusions Nanocomposites of C25A organo-modified clay and a new biodegradable polyester characterized by an alternating distribution of glycolic acid and 6-hydroxyhexanoic acid units were prepared by the solvent-casting technique. X-ray and TEM observations revealed full dispersion of silicate layers and suggested high miscibility between the polymer matrix and the clay. Thermogravimetric analyses showed that the addition of clay particles had a small stabilization effect at the beginning of the degradation process, which is currently interpreted as a consequence of polymer chain nanoconfinement. The semicrystalline character of the new polyester allowed the study of the influence of clay particles on crystallization kinetics and crystal morphology. Thus, incorporation of C25A decelerated the mechanism of primary nucleation and crystal growth of poly(glc-alt-6HH), a trend commonly observed when high homogeneous dispersion of silicate layers occurs. A slight increase in the secondary nucleation constant was also inferred for the nanocomposite by considering the Lauritzen and Hoffman treatment. FTIR and WAXD data obtained during isothermal crystallization were consistent and indicated a decrease in the overall crystallization rate when silicate layers were added. Optical microscopy revealed differences in spherulite morphology between the neat polyester and its nanocomposite. Furthermore, SAXS data showed significant changes in the morphology of constitutive lamellae since a dramatic decrease in amorphous layer thickness was observed for the nanocomposite. Despite this feature, the degree of crystallinity was higher for the neat polyester, suggesting an increase of the amorphous domains between lamellar stacks when the miscible silicate layers were added to the polymer matrix. 248 5.4.5 References [1] Gross, R. A.; Kalra, B. Science 2002, 297, 803-807. [2] Moore, G. F.; Saunders, S. M. In Advances in Biodegradable Polymers; Rapra Review Reports, 1997, vol 9, no. 2. [3] Bayer, A. G. Anwendungstechnische Information ATI 968 d, e. [4] Grigat, E.; Koch, R.; Timmermann, R. Polym Degrad Stab 1998, 59, 223-226. [5] Yamamoto, M.; Witt, U.; Skupin, G.; Beimborn, D.; Muller, R. J. In Biopolymers; Steinbüchel, A.; Doi, Y., Eds.; Weinheim: Wiley-VCH; Weinheim, 2002; Vol. 4, Chapter 3, p. 299. [6] Witt, U.; Yamamoto, M.; Seeliger, U.; Muller, R. J.; Warzelhan, V. Angew Chem Int Ed 1999, 38, 1438-1442. [7] Usuki, A.; Kojima, Y.; Kawasumi, M.; Okada, A.; Fukushima, Y.; Kurauchi, T.; Kamigato, O. J Mater Res 1993, 8, 1179-1184. [8] Yano, K.; Usuki, A.; Okada, A.; Kurauchi, T.; Kamigaito, O. J Polym Sci, Part A: Polym Chem 1993, 31, 2493-2498. [9] Kojima, Y.; Usuki, A.; Kawasumi, M.; Okada, A.; Fukushima, Y.; Kurauchi, T.; Kamigaito, O. J Mater Res 1993, 8, 1185-1189. [10] Kojima, Y.; Usuki, A.; Kawasumi, M.; Okada, A.; Kurauchi, T.; Kamigaito, O. Mater Life 1993, 5, 13-18. [11] Kawasumi, M. J Polym Sci Part A: Polym Chem 2004, 42, 819-824. [12] Giannelis, E. P. Adv Mater 1996, 8, 29-35. [13] Jog, J. P. Mater Sci Tech 2006, 22, 797-804. [14] Nam, J. Y.; Sinha Ray, S.; Okamoto, M. Macromolecules 2003, 36, 7126-7131. [15] Krikorian, V.; Pochan, D. J. Macromolecules 2004, 37, 6480-6491. [16] Sinha Ray, S.; Bousmina, M. Macromol Chem Phys 2006, 207, 1207-1219. [17] Kennedy, M.; Brown, G.; Stpierre, L. Polym Eng Sci 1990, 30, 769-775. [18] Nitta, K.; Asuka, K.; Boping, L.; Terano, M. Polymer 2006, 47, 6457-6463. [19] Somwangthanaroj, A.; Lee, E. C.; Solomon, M. J. Macromolecules 2003, 36, 2333-2342. [20] Nowackia, R.; Monasseb, B.; Piorkowskaa, E.; Galeskia, A.; Haudinb, J. M. Polymer 2004, 45, 4877-4892. [21] Burke, M.; Young, R.; Standford, J. Polym Bull 1993, 30, 361-368. [22] Wang, K.; Wu, J.; Zeng, H. Eur Polym J 2003, 39, 1647-1652. [23] Middleton, J. C.; Tipton, A. J. Med Plast Biomater 1998, 5, 30-39. [24] Martínez-Palau, M.; Franco, L.; Puiggalí, J. Macromol Chem Phys 2008, 209, 393-403. [25] Martínez-Palau, M.; Franco, L.; Puiggalí, J. Polymer 2007, 48, 6018-6028. [26] del Valle, L.; Martínez-Palau, M.; Gámez, A.; Sepulcro, F.; Puiggalí, J. Curr Trends Polymer Sci 2008, 12, 33-41. [27] Martínez-Palau, M.; Franco, L.; Puiggalí, J. J Appl Polym Sci 2008, 110, 2127-2138. [28] Rueda, D. R.; García-Gutiérrez, M. C.; Nogales, A.; Capitán, M. J.; Ezquerra, T. A.; Labrador, A.; et al. Rev Sci Instrum 2006, 77, Art. No. 033904 Part 1. [29] http://www.ccp13.ac.uk/software/program/corfunc/corfunc.htm. [30] Chen, K.; Wilkie, C. A.; Vyazovkin, S. J Phys Chem 2007, 111, 12685-12692. [31] Hoffmann, J. D.; Weeks, J. J. J Chem Phys 1962, 37, 1723-1741. 255  Thermal stability of the poly(glc-alt-amh)/C25A nanocomposite Thermogravimetric scans showed clear differences between the neat polymer and its nanocomposite with the C25A organo-modified clay (Figure 5.5.2 a). In this way, the nanocomposite showed a lower onset degradation temperature for all the assayed heating rates (Table 5.5.1), probably due to the lower stability of the organo-modifier compound, and in general a shift of its degradation curve to lower temperatures. However, curves approached to each other at the last stages of degradation suggesting that the neat polymer decomposition at high temperature proceeded faster than in the nanocomposite. Both samples reached a constant weight percentage, the remaining residue was logically greater for the nanocomposite (13% versus to 9%) due its clay content. The degree of degradation or conversion,  at a given temperature (Figure 5.5.2 b) was calculated as:    WW WW 0 0  (5.5.1) where W0, W and  Wwere the initial weight, the weight at the considered temperature and the final weight at the end of the degradation process, respectively. 100 nm 2.68 nm 1.97 nm C25A Nanocom p osite Figure 5.5.1 Transmission electron micrograph showing the morphology of the poly(glc-alt-amh)/C25A nanocomposite with a Cloisite concentration of 3%. Inse t shows the diffraction peak associated to the interlayer spacing observed in the C25A organo-modified clay (dashed line) and the nanocomposite sample (solid line). 256 a) b) 257 Figure 5.5.2 b plots the degree of conversion versus temperature (TG curve) of the nanocomposite sample at all the assayed heating rates together with the corresponding derivative curves (DTG). For the sake of completeness, curves of the neat polymer are also shown for a representative heating rate (40 ºC/min). The characteristic TG and DTG temperatures for the nanocomposite and the neat polymer sample are summarized in Table 5.5.1. Table 5.5.1 Thermogravimetric data of the nanocomposite and pristine samples. Sample  (ºC/min) Tonset (ºC) T20% (ºC) T50% (ºC) T70% (ºC) Tmax(ºC) 2 207 319 359 411 326/412 5 220 339 370 423 345/434 Poly(glc-alt-amh)a 10 232 56 388 433 357/442 20 240 375 408 451 382/460 40 260 397 429 471 404/485 2 194 314 378 414 315/387/415 5 215 332 386 426 333/400/363 Poly(glc-alt-amh)/C25A 10 223 348 404 440 349/420/447 20 227 362 414 456  40 249 378 424 470 379/455/480 a From reference 11. Both samples showed a clear first degradation step which approximately corresponded to a conversion of 0.45-0.50. This step can be mainly associated to the decomposition of the glycolic acid residues as it was previously determined from the study of a series of copolymers derived from different -amino acids and consequently with different weight percentages of glycolic acid units [11]. Thermogravimetric traces clearly indicated that this process always ended at lower temperatures for the nanocomposite than for the pristine sample. DTG curves showed a second degradation step for the neat polymer which was associated to the decomposition of the rich -amino acid fraction [11]. This process appeared rather more complicated in the nanocomposite sample since at least two DTG additional peaks were detected at a temperature that increased with the heating rate. Thus, incorporation of clay particles had a remarkable influence on the degradation process which took consequently place according to three differentiated steps. These clear differences between the decomposition of the pristine and the nanocomposite samples demonstrate that the degradation mechanism changed when the organo-modified clay was added and justify undertaking a more detailed kinetic analysis. Each degradation step of the nanocomposite was then analyzed by mathematical deconvolution of the DTG curves as previously performed with the neat polymer [11]. Figure 5.5.3 shows the 258 separation in three and two peaks of the DTG curves obtained at a representative heating rate of 5 ºC/min for the nanocomposite and the pristine polymer, respectively. In all cases, the sum of the separated curves reproduced quite well the experimental signal.  Evaluation of the activation energy for the thermal degradation of the poly(glc-alt- amh)/C25A nanocomposite According to non-isothermal kinetic theory, thermal degradation of a sample can be expressed by the following function: 1exp ( ) dE Af dT RT       (5.5.2) Figure 5.5.3 Deconvolution of DTG curves corresponding to the thermal decomposition of the pristine (a) and the nanocomposite (b) samples at 5 ºC/min. d   d   a) b) 259 where  is the heating rate, T is the absolute temperature, R the gas constant, f (  ) the differential conversion function, and A and E the preexponential and the activation energy for the studied decomposition reaction step. Activation energies for the three degradation steps were determined by using the Kissinger method20, and advanced isoconversional methods such as Kissinger-Akahira-Sunose (KAS)20,21 , Friedman22,23 and Flynn-Wall-Ozawa (FWO)24,25 which have the advantage that don’t need the knowledge of the exact thermodegradation mechanism. Integral (KAS and FWO) and differential (Friedman) isoconversional methods make use of the isoconversional principle which states that at a constant extent of conversion the reaction rate is a function only of the temperature. The Kissinger method [20] gives the associated activation energy, E, only at the maximum of the DTG curve for each degradation step and is based on the equation: 1 max 2 max max ln ln ln (1 )n A RE n TE RT        (5.5.3) where  is the heating rate, Tmax is the temperature at the maximum reaction rates, max is the conversion at this Tmax temperature, n is the reaction order and A the frequency factor. From a plot of ln(/Tmax2) versus 1/Tmax and fitting the data to a straight line (Figure 5.5.4 a and Figure 5.5.5 a), the activation energy was calculated from the slopes for each degradation step as summarized in ¡Error! No se encuentra el origen de la referencia. A good linearity was always observed with correlation coefficients of 0.9999, 0.9968 and 0.9832 for the first, second and third degradation step, respectively. It should be pointed out that Kissinger is not an isoconversional method since the peak temperature is obtained at different heating rates, and the extent of conversion related to the peak is known to change with the heating rate26,27. Moreover, the determined activation energy may loss sense if it varies throughout the degradation process. In order to calculate the activation energy during the whole process, the KAS method20,21 was applied. This is based on the integration of equation 5.5.2, which after a subsequent reordering leads to the expression. RT E Eg AR T       )( lnln 2   (5.5.4) where g(  ) is the integral conversion function (i.e. 0 () () d gf     ). 260 Table 5.5.2 Activation energies of the nanocomposite and pristine samples determined by isoconversional methods. Sample Stepa E (kJ/mol) Kissingerb E( kJ/mol) KASc E( kJ/mol) Friedmanc E (kJ/mol) FWOc Poly(glc-alt-amh)d 1 118 116 120 120 2 176 179 185 181 Poly(glc-alt-amh)/C25A 1 127 111 133 124 2 146 139 159 157 3 168 150 168 161 a The different steps of degradation are referred to 1, 2 and 3 in increasing order of temperature. b Calculated at the temperature corresponding to the maximum of each step in the DTG curve. c Summarized energies correspond to mean values obtained from different degrees of conversion (from 0.1 to 0.95). d From reference 11. For each degree of conversion and degradation process the activation energy was obtained from the slope of the linear representation of ln (  / T2) versus 1000/T. A good linearity was obtained with correlation coefficients not less than 0.97, 0.98 and 0.95 for the first, second and third degradation step, respectively. In each case the worst coefficient were found for the lowest conversion degree (0.1), whereas at a conversion of 0.95 the coefficient was in the three cases higher than 0.99. As shown in Figure 5.5.4 b (step 1) and Figure 5.5.5 b (steps 2 and 3). Activation energies slightly increased with the conversion degree (Figure 5.5.6) and also when degradation progressed from step 1 to step 3. Average values are summarized in Table 5.5.2. 261 -11 -10.5 -10 -9.5 -9 -8.5 -8 1.5 1.55 1.6 1.65 1.7 ln β/TX2) 1000/Tmax (K-1) 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 1.41.51.61.71.81.9 lnβ 1000/T (K-1) -6 -5 -4 -3 -2 -1 0 1.4 1.5 1.6 1.7 1.8 1.9 ln β* (dα/dT) 1000/T (K-1) -12.5 -12 -11.5 -11 -10.5 -10 -9.5 -9 1.41.51.61.71.81.9 ln ( β/T2) 1000/T(K-1) Figure 5.5.4 Kissinger (a), Kas (b), Friedman (c) and FWO (d) plots for the first thermal decomposition step of the poly(glc-alt-amh)/C25A nanocomposite sample. From left to right lines correspond to conversion ranging from 0.1 to 0.9 in steps of 0.1 and the conversion of 0.95. 262 -12 -11.5 -11 -10.5 -10 -9.5 -9 -8.5 -8 1.2 1.3 1.4 1.5 ln (β/Tmax 2 ) 1000/T max (K -1 ) 1.21.41.61.8 2 -13 -12.5 -12 -11.5 -11 -10.5 -10 -9.5 -9 1.2 1.3 1.4 1.5 1.6 1.7 1000/T(K -1 ) ln ( β/T 2 ) 1000/T (K -1 ) 1.2 1.4 1.6 1.8 2 -7 -6 -5 -4 -3 -2 -1 0 1.2 1.3 1.4 1.5 1.6 1.7 1000/T (K -1 ) ln β* (dα/dT) 1000/T (K -1 ) 1.2 1.4 1.6 1.8 2 0 0.5 1 1.5 2 2.5 3 3.5 4 1.2 1.3 1.4 1.5 1.6 1.7 1000/T(K -1 ) ln β 1000/T(K -1 ) Figure 5.5.5 Kissinger (a), Kas (b), Friedman (c) and FWO (d) plots for the second (♦)and the third (◊) thermal decomposition steps of the poly(glc-alt-amh)/C25A nanocomposite sample. From left to right lines correspond to conversion ranging from 0.1 to 0.9 in steps of 0.1 and the conversion of 0.95. 263 The Friedman method [22,23] (equation 5.5.5) derives from the logarithmic form of the rate equation 5.5.2 and enables also to get the values of activation energies over a wide range of conversions by plotting ln(  d  /dT) versus 1000/T from thermograms recorded at several heating rates. ln ln ln(1 ) dE An dT RT        (5.5.5) Figure 5.5.4 c (first degradation step) and Figure 5.5.5 c (second and third degradation steps) show straight lines whose slopes allowed the evaluation of activation energies (Table 5.5.2 for average values). A good linearity was obtained for each conversion with correlation coefficients not less than 0.98, 0.95 and 0.98 for the first, second and third degradation step, respectively. Finally, the integral Flynn-Wall-Ozawa method24,25, based on equation 5.5.6, allows also determining the activation energies for each degradation step and conversion degree. 0.0048 ln ln 1.0516 () AE E gR RT   (5.5.6) Activation energies (Table 5.5.2 for average values) were in this case calculated from the linear plots of ln  versus 1000/T (Figure 5.5.4 d and Figure 5.5.5 d). Table 5.5.2summarizes the values of the activation energies calculated for the poly(glc-alt- amh)/C25A nanocomposite by using the four indicated methods. For the sake of completeness previous results determined for the neat polymer are also included11. Figure 5.5.6 plots also the variation of the activation energy deduced from the Friedman method with the conversion for the three degradation steps. Activation energy was practically constant for the first degradation step and shows a slightly greater fluctuation for the second one, which had an overlapping with the other two steps and particularly with the third one (at higher conversions). Activation energy progressively increased with the conversion degree for the last degradation step as also found in the KAS analysis. Average values of the activation energy for the three considered steps are summarized in Table 5.5.2. It can be observed that these average values were close to the single activation energy determined by the Kissinger plot, and higher than the average activation energies determined by the KAS method. Differential isoconversional methods, like Friedman, are recommended over integral when discrepancies exists21, 22, 23. Thus, we obtained energies of 130, 153 and 168 kJ/mol by averaging Kissinger and Friedman data for the first, second and third degradation steps, respectively. 264 The first degradation step had the lowest activation energy for both the neat polymer and the nanocomposite sample. The incorporation of clay particles slightly increased this activation energy. It can also be emphasized that the second stage of degradation of the neat polymer had an activation energy considerably higher than those determined in either the second or the third degradation steps of the nanocomposite sample.  Thermal degradation mechanisms of the poly(glc-alt-amh)/C25A nanocomposite The Coats-Redfern method28 was chosen in order to determine the thermal degradation mechanism involved in the three different degradation steps of the nanocomposite. Conventional g (  ) functions29,30 were considered and for each one, the activation energy was calculated according to equation 5.5.7, which was derived considering an asymptotic approximation (2RT/E <<1). 2 () ln ln gARE TERT       (5.5.7) Figure 5.5.6 Dependence of the kinetic rate constant, k, on temperature for the differen t degradation steps (1, 2 and 3) of the pristine (dashed lines) and the poly(glc-alt-amh)/C25A nanocomposite (solid lines) sample. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 550 600 650 700 750 800 850 k (s-1) k (s-1) Temperature (ºC) 1122 271 experimental and the simulated TG curves for a representative heating rate. Results indicated that the best agreement was obtained when the autocatalytic (n = 1.5; m = 0.5) model was considered for the first step and consequently the other two proposed mechanisms could be clearly discarded. As shown in Figure 5.5.8, the fitting to experimental data is good for almost all whole area, with slight divergences in the regions corresponding to the lowest and the highest weight loss. A similar procedure was performed to choose the autocatalytic model over the R2 model for the third degradation step. Finally, the invariant values of the activation energies and the pre-exponential factors allowed us to calculate the rate constant, k, of the different non-isothermal degradation steps for the nanocomposite and the neat polymer samples. Figure 5.5.6 compares the temperature dependence of the rate constant for the two studied samples and the different degradation steps. Although for the first degradation step the nanocomposite showed higher activation energy than the neat polymer, it is clear that its pre-exponential factor was also higher (Table 5.5.4). In this way, degradation can proceed faster in the nanocomposite during the first stage of thermal decomposition. It has been claimed that an increase on the activation energy cannot be rationalized in terms of the formation of a surface silicate barrier On the other hand, the high pre-exponential factor calculated for the second degradation step of the neat polymer justified its higher degradation rate during the last stages of decomposition despite the activation energies of the two samples were quite similar (i.e. 167 kJ/mol for the neat polymer and 151 kJ/mol and 165 kJ/mol for the nanocomposite). Figure 5.5.9 Comparison between experimental (- - -) and simulated TG curves for the poly(glc-alt- amh)/C25A nanocomposite sample at representative heating rate of 5 ºC/min. Curves were calculate d considering the autocatalytic, A3/2 and R2 models for the first degradation step. 272 5.5.4 Conclusion Thermal decomposition in a nitrogen atmosphere of the alternating biodegradable poly(ester amide) constituted by glycolic acid and 6-aminohexanoic acid units was significantly changed by the incorporation of a small percentage (3 wt %) of the C25A organo-modified clay, which rendered an intercalated structure when the sample was prepared by the melt-mixing technique. The polymer showed a first degradation step, which mainly involved the decomposition of glycolic acid units, and which proceeded faster when the clay was added. The low stability of the organo-modifier compound had a determinant role and contributed to modify the degradation mechanism associated to this step. Results derived from isoconversional analyses, and the Coats-Redfern and IKP methods were highly consistent and allowed postulating an autocatalytic mechanism when simulated degradation data were compared with the experimental curves. Last stages of degradation proceeded slowly for the nanocomposite, which furthermore showed a complex degradation process due to existence of two additional decomposition mechanisms. Thus, the second degradation step observed in the pristine sample was split due to the influence of clay particles which probably enhanced the performance of the char formed. The preexponential factor of the last degradation step of the pristine sample was higher than the calculated value for the nanocomposite suggesting that motion of reactive groups was hindered by the presence of silicate layers. 273 5.5.5 References [1] R. A. Gross, B. Kalra, Science 297 (2002) 803. [2] G. F. Moore, S. M. Saunders, In Advances in Biodegradable Polymers; Rapra Review Reports 9 (1997) 2. [3] I. Arvanitoyannis, N. Kawasaki, N. Yamamoto, Polymer 36 (1995) 857. [4] N. Paredes, A. Rodríguez-Galán, J. Puiggalí, J. Polym. Sci. Part A: Polym. Chem. 36 (1998) 1271. [5] R. Katsavara, V. Beridze, N. Arbuli, D. Kharadze, C. C. Chu, C. Y. Won, J. Polym. Sci. Part A: Polym. Chem. 37 (1999) 391. [6] H. R. Stapert, A. W. Bouwens, P. J. Dijkstra, J. Feijen, Macromol. Chem. Phys. 200 (1999) 1921. [7] M. Vera, A. Rodríguez-Galán, J. Puiggalí, Macromol. Rapid Commun. 25 (2004) 812. [8] M. Vera, A. L. Franco, J. Puiggalí, Macromol. Chem. Phys. 205 (2004) 1782. [9] E. Botines, M. T. Casas, J. Puiggalí, J. Polym. Sci. Part B: Polym. Phys. 45 (2007) 815. [10] E. Botines, J. Puiggalí, Eur. Polym. J. 42 (2006) 1595. [11] E. Botines, L. Franco, J. Puiggalí, J. Appl. Polym. Sci. 102 (2006) 5545. [12] L. del Valle, F. Sepulcre, A. Gámez, A. Rodríguez-Galán, J. Puiggalí, Current Trends in Polymer Science 12 (2008) 27. [13] L. T. Morales, L. Franco, M. T. Casas, J. Puiggalí, J. Polym. Sci. Part A: Polym. Chem. 47 (2009) 3616. [14] L. T. Morales, L. Franco, M. T. Casas, J. Puiggalí. Polym. Eng. Sci., submitted. [15] B. N. Jang, C. A.Wilkie, Polymer 46 (2005) 2933. [16] X. Yuan, C. Li, G. Guan, Y. Xiao, D. Zhang, Polym. Degrad. Stab. 93 (2008) 466. [17] S. Sinha Ray, M. Bousmina, Prog. Mater. Sci. 50 (2005) 962. [18] K. Chen, C. A.Wilkie, S.Vyazovkin, J. Phys. Chem. B 111 (2007) 12685. [19] J. H. Chang, Y. Uk.An, G.S. Sur, J. Polym. Sci. Part B: Polym. Phys. 41 (2003) 94. [20] H. E. Kissinger, Anal. Chem. 29 (1957) 1702. [21] T. Akahira, T. Sunose, Res. Report Chiba Inst. Technol. 16 (1971) 22. [22] H. J. Friedman, Polym. Sci. Part C 6 (1964) 183. [23] H. L. Friedman, J. Polym. Lett. 4 (1966) 323. [24] T. Ozawa, Bull. Chem. Soc. Jpn. 38 (1965) 1881. [25] J. H. Flynn, L. A. Wall, J. Polym. Sci. Polym. Lett. 4 (1966) 323. [26] N. Sbirrazzuoli, Y. Girault, L. Elégant, Thermochim. Acta 25 (1997) 293. [27] S. Vyazovkin, N. Sbirrazzuoli, Macromol. Rapid Commun. 27 (2006) 1515. [28] A. W. Coats, J. P. Redfern, Nature 201 (1964) 68. [29] A. B. Phadnis, Thermochimica Acta 62 (1983) 361. [30] S. Vyazovkin, D. J Dollimore, Chem. Inform. Comput. Sci. 36 (1996) 42. [31] A. I. Lesnikovich, S. V. Levchik, J. Therm. Anal. 27 (1983) 89. [32] A. I. Lesnikovich, S. V.Levchik, J. Therm. Anal. 30 (1985) 677. 274 6 CONCLUSIONS 276 277 Polyamides  Structural transitions  Nylons 56, 65 and 47 have similar structures when crystallized directly from synthesis. The molecular arrangement is characterized by a monoclinic unit cell (form I) that can be interpreted in terms of a peculiar packing where hydrogen bonds are established along two directions and where neighboring chains are shifted along their chain axis direction.  Although a similar structure was postulated at room temperature for the studied evenodd and odd-even nylons.  On heating, nylons 65 and 47 first showed a transition towards a less compact structure (form II) characterized by the presence of two close equatorial reflections (0.422 and 0.401 nm) and subsequently underwent a second transition where the two indicated reflections progressively merged into a single one.  On heating, nylon 56 showed only a transition towards a pseudohexagonal packing, although an additional rearrangement of amide groups was detected at some degrees before fusion.  Nylons 65, 47 and 56 crystallized from the melt according to the Brill high temperature structure. › In the case of nylons 65 and 47, the structure first reverted to form II on cooling, showing a hysteresis effect and then this form II underwent a partial transition to the low temperature structure › Nylon 56 did not show a Brill transition during the cooling process alghough a progressive, minor crystallization into the low temperature structure also took place.  Spherulitic morphology  Nylon 56 crystallized on cooling into fibrillar spherulites with optical properties that were depended on the crystallization temperature and differed from those found in nylons having conventional sheet structures. During crystallization thinner lamellae inserted into the loosely stacked bundles of primary lamellae and the interlamellar amorphous regions became more compact  Nylon 65 crystallization from the melt gives rise to spherulites constituted by lamellae of different thicknesses, which accounts for the multiple melting peaks observed in the calorimetric heating runs. During crystallization thinner lamellae insert into the loosely stacked bundles of primary lamellae and the interlamellar amorphous regions become more compact. Spherulites with different textures (ringed or fibrilar) and birefringences 278 can be obtained by varying the crystallization conditions. Negative spherulites form in the low temperature region, indicating a molecular arrangement different from that found in conventional even-even nylons, whose low temperature spherulites show a positive birefringence.  Nylon 47 showed spherulites with different textures and birefringences depending on the crystallization temperature. Negative spherulites formed in the low temperature region, indicating a molecular arrangement different from that found in conventional even-even nylons that gave rise to spherulites with a positive birefringence. Spherulites obtained at low supercoolings were mainly constituted by flat micro-crystals that showed reversible optical properties with temperature. This peculiar behaviour may be a consequence of the postulated structure with two hydrogen bonding directions since small changes on the torsional angles of the odd diamide units could induce a variation on the angle between the two hydrogen boding directions and on the birefringence sign of the crystalline micro-domains.  FTIR  Nylon 65 absorption bands observed in the room temperature infrared spectra suggests that this polyamide has a structure related to conventional / forms despite its different hydrogen-bonding scheme. Spectra are sensitive to the structural changes; specifically the temperature evolution of the amide A band allows the form I to form II transition to be detected.  Nanocomposites  Nanocomposites with intercalated and exfoliated structures could be obtained from nylon 47 and the C25A and C30B organomodified clays. The final structure mainly depended on the preparation method. Specifically, melt mixing favoured the exfoliated distribution whereas intercalated structures were obtained by solution intercalation. Incorporation of clay particles influenced the overall crystallization rate under both isothermal and non-isothermal conditions. In all cases, clay particles decelerated the crystal growth process and had a strong influence on the primary nucleation which could be enhanced or disfavoured when intercalated or exfoliated structures were respectively achieved. 279 Poly (glycolic acid-6-hydrohexanoic acid) The solvent casting technique was appropriate to prepare nanocomposites with an exfoliated structure from the C25A organo-modified clay and a new biodegradable polyester characterized by an alternating distribution of glycolic acid and 6-hydroxyhexanoic acid units. The nanocomposite had interesting characteristic to remark:  Thermal analysis Thermogravimetric analyses showed that the addition of clay particles had a small stabilization effect at the beginning of the degradation process, which is currently interpreted as a consequence of polymer chain nanoconfinement.  Crystallization The semicrystalline character of the new polyester allowed the study of the influence of clay particles on crystallization kinetics and crystal morphology. Thus, incorporation of C25A decelerated the mechanism of primary nucleation and crystal growth of the neat polyester, a trend commonly observed when high homogeneous dispersion of silicate layers occurs. A slight increase in the secondary nucleation constant was also inferred for the nanocomposite by considering the Lauritzen and Hoffman treatment. FTIR and WAXD data obtained during isothermal crystallization were consistent and indicated a decrease in the overall crystallization rate when silicate layers were added.  Morphology Optical microscopy revealed differences in spherulite morphology between the neat polyester and its nanocomposite. Furthermore, SAXS data showed significant changes in the morphology of constitutive lamellae since a dramatic decrease in amorphous layer thickness was observed for the nanocomposite. Despite this feature, the degree of crystallinity was higher for the neat polyester, suggesting an increase of the amorphous domains between lamellar stacks when the miscible silicate layers were added to the polymer matrix. 280 Poly(glycolic acid-alt-6-aminohexanoic acid)  Nanocomposite preparation and structure.  The C25A organo-modified montmorillonite has proved to be effective for the preparation of nanocomposites of the degradable alternating poly(ester amide) constituted by glycolic acid and 6-aminohexanoic acid units by the in situ polymerization technique.  The in situ polymerization of sodium chloroacetylaminohexanoate in presence of C20A or C30B organo-modified clays rendered intercalated silicate structures, as determined by X-ray diffraction and transmission electron microscopy.  Nanocomposites prepared by the melt mixing technique using a 3% content of C25A organo-modified clay and the selected poly(ester amide) rendered an intercalated structure.  Polymerization kinetics  Polymerization kinetics of sodium chloroacetylaminohexanoate was strongly influenced by the presence of organo-modified montmorillonites under both nonisothermal and isothermal conditions. The reaction process was rather complicated when polymerization temperatures lower than 145 ºC were selected since polymer crystallization occurred before the polymerization reaction was finished. WAXD profiles revealed changes in the monomer structure and the range where the reaction proceeded in a liquefied state. › FTIR and WAXD experiments were appropriate techniques to evaluate the in situ polymerization kinetics and in particular the estimation of the activation energy and the pre-exponential factor. This factor was found to decrease when nanocomposites had an exfoliated structure probably as a consequence of a decrease of the chain mobility and the frequency at which reactive groups were close enough to facilitate the condensation reaction. Nanocomposites with intercalated structures showed also a decrease of the pre-exponential factor but also in the activation energy, which in this case suggest a favoured condensation process caused by the nanoconfinement of monomers in the silicate galleries.  Crystallization  The nanocomposite with an exfoliated structure prepared by in-situ polymerization thechnique, crystallized at lower rate that the neat polymer since both nucleation density and crystal growth rate were lower. On the contrary the nanocomposite having an 287 polymer layers thoroughly coating the platelets (Figure A.2). Direct visualization of the polymer grafted onto the clay platelets gave relevant information on the structure of the obtained nanohybrid materials. First, the grafting density increased drastically as the proportion of OH-substituted alkylammonium cations used to organo-modify the clay was raised. Second, the polymer deposit was not simply a continuous film growing in thickness with increased OH content. Instead, separate polymer islands formed in the low-OH-content systems, probably as a result of a phase separation process between the ammonium ions induced by the polymerization reaction. Homogeneous coverage and subsequent thickening only took place from 50% OH content. When this situation was achieved, adjacent platelets became fully independent of each other since they were fully covered by the polymer and exfoliation was greatly favored. Bulk polymerizations of -caprolactone were also conducted at 170 ºC in the presence of catalytic traces of water and 10, 30 and 50 wt% of hydrated synthetic montmorillonite SOMASIF ME100 without additional catalysts [28]. 1H NMR and GPC analyses suggested that the montmorillonite present in the system induced both significantly higher lactone hydrolysis and polymer chain growth rates. All systems gave rise to low molecular weights (Mw = 5360– 22,432), which seemed to indicate a hindrance effect in diffusion caused by increasing amounts of silicate in the system. Wide-angle X-ray diffraction data revealed that within the interlamellar regions of silicate, -caprolactone was arranged in weakly ordered bimolecular pseudo-layers. It is worth noting that the interlayer spacings of montmorillonite in the nanocomposites were slightly smaller than for silicate dispersed in -caprolactone. Such results were attibuted to an unfavorable polymerization of -caprolactone in the interlayer region of the silicate (i.e. chains remained small and unable to open the clay layers to give rise a delaminated structure). The comparison between measured values of gallery height and calculated dimensions of the poly(- caprolactone) chain indicated that polymer chains were flatly arranged on each side of the silicate platelet, creating pseudo-bilayers inside the montmorillonite gallery. 288 OH OH OH OH OH OO O OH OOO Polymerization low grafting density high grafting density > 50% OH grups < 50% OH grups Polymer patch, ~ 2-3 nm high Grafted polymer layer, ~ 5-8 nm thick Polymerization MMT surface MMT surface Long alkyl chain amoniun anions MMT surface MMT surface MMT surface ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| ||| || ||| ||| Figure A.2 Scheme of the polymer surface grafting onto individual clay platelets and concomitant phase separation, as explained in detail in reference [26]. Comprehension of the nanocomposite structure and interactions at the polymer/nanofiller interface is crucial in controlling nanocomposite properties. In all cases, nanocomposites are characterized by the presence of a polymer layer at the interface with the inorganic surface whose properties clearly differ from those of the bulk polymer. Pucciariello et al. [29] conducted a detailed surface analysis of poly(-caprolactone)-montmorillonite clay nanocomposites obtained by in-situ polymerization. The organophilic clay was prepared by treating the natural montmorillonite with a solution of protonated 12-aminolauric acid. In-situ polymerization of -caprolactone was performed under a nitrogen atmosphere at room 289 temperature for 2 h and then at 170 °C for 48 h. The organo-modified montmorillonite concentration was close to 18% as this was previously determined as the maximum content at which the silicate was exfoliated [30,31]. The X-ray photoelectron spectroscopy (XPS) technique gave information on atom concentrations in the surface layer, the valence state of these atoms and the bonding of their nearest neighbors. Spectroscopic data indicated a great polarization of Si-O and Al-O bonds in the organo-modified montmorillonite, which was attributed to the electron-attracting and electron-donor effects of the NH3+ and COOH groups of the aminolauric ion, which efficiently coordinated the partially negatively charged oxygens and the partially positively charged silicons (aluminums) of the clay (Figure A.3 a). XPS data of the nanocomposite revealed lower binding energies of Si-O and Al-O bonds, and consequently a reduced ionicity of such bonds with respect to the organophilic clay. Thus, the presence of the polymer limited the polarizing effect of the aminolauric cation because of the electrondonor/electron-attracting effects of the ester group oxygens (Figure A.3 b). N O=C HO Si  H H H OSi O   ..... ..... ..... N O=C HO Si  H HH OSi O   ..... ..... ..... Si  O  N H H H COOH ..... ..... NH HH COOH Si O  N O=C HO Si  H H H OSi O   ..... ..... ..... O OC O O C ...... ..... Figure A.3 Schemes explained in detail in Ref. [30] showing: a) Coordination of the aminolauric cation HOCO-(CH2)11-NH3+ to the clay. b) Model of the interactions occurring in the polymer-alkylammonium cation-clay system. Poly(-caprolactone) was studied as a polymer biodegradable matrix which adds potential to the use of derived nanocomposites as biodegradable packaging materials. The synthesis of a series of montmorillonite-PCL nanocomposites in which the content of the inorganic material was varied regularly from 0 to 44 wt.-% was performed to find some basic structure/property correlations of the multiphase nanocomposites and investigate the permeability to organic (e.g. dichloromethane) and inorganic (e.g. water) solvents of the multiphase polymers [32]. In-situ polymerization was carried out at 85 ºC using a montmorillonite modified with protonated 12- aminolauric acid. Permeability to water and dichloromethane was found to decrease significantly with increasing clay content. In particular, the water permeability behavior was largely dominated by the diffusion parameter. The diffusion path of the polar molecules of 290 water was assumed to be slowed down with respect to dichloromethane vapor because of not only the physical barrier of the clay layers but also the hydrophilic character of the platelets. Barrier properties associated with the biodegradability of poly(-caprolactone) play an important role in enhancing interest in these nanocomposites as biodegradable packaging materials. PCL is thermodynamically miscible with many other polymers (e.g. PVC or SAN copolymer) and can be used as an environmentally decomposable additive that facilitates plastic degradation [33]. A study was recently undertaken to dissolve PCL-clay systems prepared by in-situ polymerization in a SAN matrix and evaluate the structure and mechanical properties of these ternary nanocomposites [34]. The in-situ polymerization was performed at 170 ºC using a commercial montmorillonite modified with hexadecyltrimethylammonium bromide. Clay content was of 10, 30 or 50 wt.% since nanocomposites were then used as masterbatches to obtain SAN/(PCL/clay) ternary blends containing 0.66–5.65 wt.% of the organo-modified clay. Blends of PCL nanocomposites with SAN copolymer gave rise to intercalated and semiexfoliated structures, as deduced from XRD and TEM techniques. The increase of clay in the system led to higher values of Young’s modulus. Thus, compared to unmodified SAN the addition of only 0.66 wt.% resulted in approximately a 10% increase in stiffness modulus while at 5.65 wt.% a 40% increase was obtained. Supercritical CO2 has been studied as an alternative polymerization medium (e.g. for ringopening polymerization of lactones [35,36]) for more conventional organic solvents due to its well known advantages [37]. Thus, it is environmentally friendly, non toxic, non flammable and economical since its critical parameters are relatively easily obtained. Furthermore, its fluid/solvent properties can be tuned by small changes in temperature. Preparation of nanocomposites by redispersion of masterbatches of exfoliated PCL- nanocomposites into polymers miscible with PCL, such as the styrene/acrylonitrile copolymer [34], showed encouraging results but also drawbacks, e.g. difficulty in recovering the aggregated bulk masterbatch and need to purify it before use. Detrembleur et al. [38] proposed an alternative method based on the preparation of PCL/clay masterbatches by in-situ intercalative polymerization in supercritical carbon dioxide. The specific properties of this solvent allowed in-situ polymerization at high clay content without viscosity problems. This is never the case for bulk polymerization processes, where clay loading is usually limited to 30 wt %. Furthermore, the product obtained after depressurization was an easily recoverable fine powder and supercritical CO2 was able to extract the residual monomer during depressurization, directly providing a ready-to-use dry powder. The ring-opening polymerization of - caprolactone in supercritical carbon dioxide occurred as a dispersion polymerization since only the monomer was soluble in this medium under typical supercritical conditions [39]. Catalysts 291 like Sn(Oct)2 were preferably used as they should have low sensitivity to the carbonation reaction and protic impurities. Polymerizations of -caprolactone were carried out in supercritical CO2 (85 ºC, 28 MPa) using the natural montmorillonite and the clay organomodified with a non-functional (Cloisite 20A) or a functional (Cloisite 30B) quaternary ammonium salt. In all cases, nanocomposites were qualified as ‘‘pre-exfoliated’’ masterbatches since true exfoliation could not be reached at high clay contents. Redispersion of organomodified clay/polymer masterbatches into chlorinated polyethylene was proved to be more efficient in terms of quality of clay delamination than direct blending of commercial clay. Starch, a natural polymer constituted by linear -glucan amylose and highly branched amylopectin, is considered one of the most promising candidates to be used as environmentally friendly materials. It has an attractive combination of availability, price and performance, but its mechanical properties are poorer than those of synthetic polymers mainly due to its hydrophilic nature, which makes it sensitive to moisture content. For this reason, starch has been modified by blending with synthetic polymers such as poly(-caprolactone) and even with layered silicates [40,41]. Nazami et al. [42] have recently prepared starch-g-polycaprolactone by in-situ ring-opening polymerization of -caprolactone in the presence of starch, Sn(Oct)2 and an organo-modified montmorillonite (i.e. Cloisite 15A) at reaction temperatures between 100-150 ºC. Results suggested a slight improvement in thermal stability but intercalation of the copolymer into clay galleries was less effective than the solution intercalation method.  2.2. Polylactide based nanocomposites Polylactide (PLA) is a well-known green polymer which receives great attention from the polymer industry since it can be produced from renewable resources (e.g. corn starch and other carbohydrate-rich substances like maize, sugar or wheat) and is biodegradable and compostable. These advantages make PLA an attractive alternative to classical commodity polymers (e.g. in the production of loose-fill packaging, compost bags, food packaging and disposable tableware). However, several properties need to be improved to widen its range of applications. For example, PLA is too brittle and permeable to gases to render optimal application performances, especially for packaging purposes. Despite this, PLA is currently one of the most widely used speciality polymers in the biomedical field (e.g. for sutures, stents, dialysis media, drug delivery devices and even for tissue engineering). The introduction of a few percent of nanofillers, such as layered aluminosilicate clays, has been extensively considered to enhance PLA properties. However, the greatest improvement is usually achieved when nanoparticles are fully and uniformly delaminated (exfoliated) in the polymer matrix, a challenge that cannot be fully met by direct melt blending of the clay as this usually leads to intercalated nanocomposites [43,44]. The use of in-situ polymerization 292 techniques appears fully justified since the monomer can penetrate and then polymerize inside the clay sheets, enhancing the delamination efficiency. CH 3 N OH HO + CH 3 N OH HO + R R CH 3 N O O + R CH 3 N O O + R Al Al Et Et Et Et CH 3 N O O + R Al Al Et Et Et Et O OCH 3 Al Et Et n O OCH 3 Al Et Et n CH 3 N O O + R O OCH 3 Al Et Et n O OCH 3 Al Et Et n AlEt 3 L,L-LA bulk, 120ºC, 48h nOH / nAl = 1 Silicate layer Silicate layer Silicate layer N O CH 3 +CH 3 N O O +CH 3 N O +CH 3 N OH + R Al Et HO O Al Et OO Al N O CH 3 + CH 3 N OH HO +CH 3 N OH HO + R R AlEt 3 L,L-LA bulk, 120ºC, 48h n OH / n Al = 3 Al Et Et Silicate layer Silicate layer Figure A.4 a) Scheme of the L,L-lactide in-situ polymerization performed from Cloisite 30B using triethylaluminium (AlEt3) as the initiator (R stands for tallow alkyl chain). b) Scheme of a mixture of aluminium mono-, di-, and tri-alkoxides produced by addition of AlEt3 onto Cloisite 30B in a nOH/nAl = 3 molar ratio. In-situ ring-opening polymerization of the lactide monomer in the presence of clay has been extensively studied [45,46]. For instance, PLA/MMT nanocomposites were prepared in bulk with catalysts such as tin(II) octoate (Sn(Oct)2) or triethylaluminum in the presence of montmorillonite clays (e.g. Cloisites 25A and 30B) organo-modified with an ammonium salt functionalized or not with hydroxyl groups. When Cloisite 30B was used, polymerization was co-initiated by a molar equivalent of AlEt3 or Sn(Oct)2 with respect to the hydroxyl groups borne by the ammonium cations of the filler, which was added before the L,L-lactide monomer. The presence of hydroxyl groups was found to be critical because these led to aluminium alkoxide or tin alkoxide active species (Figure A.4and Equation A.1), which acted as initiators and led to polylactide chains grafted onto the clay surface. Sn (Oct)2 + n OH Sn(Oct)2-n(OR)n + n OctH (A.1) 293 The molar ratio between hydroxyl groups and aluminium cations was crucial in promoting a controlled polymerization. Thus, a defect of aluminium cations should lead to more efficient aluminium trialkoxides but this was not the case when polymerizations were performed from lactide rings [46]. It was claimed that the probability to form a large amount of trialkoxide species from the anchored hydroxyl groups was rather low, and most probably a mixture of aluminium mono-, di-, and tri-alkoxides was formed (Figure 6.4 b). This distribution was responsible for the observed low monomer conversion, loss of polymerization control and bimodality of the molecular weight distribution. The grafted chains pushed the lamellar sheets apart from each other and led to the achievement of an excellent degree of clay exfoliation, which for example leads to improved thermal stability compared to unfilled PLA and even intercalated counterparts. Moreover, while the Tg and Tm of the PLA matrix were not influenced by the nanofiller, the degree of crystallinity of the polyester in the exfoliated structure was significantly higher than in the intercalated nanocomposite [45]. The grafting reaction was confirmed by the impossibility of dissolving the so-produced PLA chains in good solvents like toluene, THF or CHCl3. In fact, a specific cationic exchange reaction with LiCl was required to recover the PLA chains, suggesting that polyester chains were attached to the ammonium cations localized in close vicinity and in electrostatic interaction with the montmorillonite surface. Interestingly, polymerization conditions (e.g. bulk with 3 wt% of Cloisite) allowed the synthesis of PLA grafts with a number average molecular weight close to 14,000 g/mol and a relatively narrow distribution for such a heterogeneous initiation process (i.e. polydispersity index was close to 1.5). Dubois et al. [46] also used in-situ polymerization to prepare a PLA-MMT masterbatch and confirmed that its dilution with the neat PLA during melt processing was an effective technique for improving clay dispersion. It is well known that brittleness is a strong limitation for the application of PLA. The use of plasticizing agents is a usual procedure that may allow PLA to fulfil mechanical requirements. Considerable efforts have been made to improve PLA brittleness to make it competitive with low-cost flexible commodity polymers (e.g. polyethylene and polypropylene). Several types of compounds such as citrate ester, poly(ethylene glycol) (PEG), glucose monoesters, partial fatty acid esters, oligomeric lactic acid and glycerol have been studied as plasticizers for PLA [47- 49]. However, the addition of a plasticizer generally reduces strength and modulus. Moreover, despite the increase of deformation, PLA-based materials having a good balance of stiffness and high deformation are still required for wide applications. For this reason, the possibility of preparing nanocomposites based on the PLA matrix and adding a plasticizer compound has been considered, specifically, the preparation of nanocomposites by melt blending a PLLA matrix, a plasticizer like poly(ethylene glycol) and a nanofiller. 294 Nevertheless, PEG tends to diffuse out of the material and accumulates at the nanocomposite surface, leading to structural matrix changes upon ageing [50]. Thus, in-situ polymerization of L,L-lactide in the presence of both dihydroxylated PEG (Mw =1 000) and Cloisite 30B has been studied as an interesting alternative since it leads to nanocomposites based on a triblock copolymer matrix where the central polyethylene block affords flexibility and is not susceptible to diffusing out of the material [45]. Results showed that intensive clay platelet destructuration was achieved independent of the PEG weight ratio. The plasticizing effect of the PEG sequence (entrapped in the triblock copolymer) was highlighted by the significant Tg decrease of the nanocomposite (e.g. from 60 to 12 ºC at 16.2 wt.-% content in PEG). Moreover, the thermal degradation of the resulting nanocomposites was dependent on the relative content in PEG blocks and decreased as the polyether level increased within the triblock copolymer. Polylactide/vermiculite nanocomposites were also prepared by in-situ intercalative polymerization of L,L-lactide in the presence of vermiculite clay particles (VMT) that were treated with an alkylammonium surfactant in order to decrease their highly hydrophilic character and favor diffusion of the cyclic monomer into the clay interlayer spaces [51]. XRD suggested that exfoliated structures were attained and TEM observations revealed that VMT layers were exfoliated and dispersed uniformly in the polylactide matrix. TGA indicated a slight improvement in thermal stability (i.e. the onset temperature increased from 279 to 314 ºC when 5% of organo-modified clay was added). Dynamo-mechanical analyses showed an increase in storage and loss moduli caused by the reinforcing effect of the nanoscale VMT layers, as well as a slight increase of Tg with increasing the clay content. Nanocomposites based on anionic clays or layered double hydroxides (LDHs) have also been widely studied. These clays are comprised of positively charged layers with anions and water molecules in the interlayer region. Compounds were defined with the general formula (Mg1- xAlx(OH)2)x+(A-)x nH2O, in which A- represents an interlamellar anion (e.g. carbonate) [52]. Although LDHs occur naturally, they are usually synthesized under controlled conditions in order to obtain materials with a known, homogeneous composition [53]. LDH platelets have a high aspect ratio, tuneable layer charge density and can be prepared by low-cost processes. The potential suitability of organo-LDHs for intercalation of hydrophobic polymers, the potentially greater susceptibility to complete exfoliation than that of cationic clays and the catalytic activity for polymerization of lactides in the interlayer are also worth mentioning [54,55]. Taviot-Gueho and Leroux [56] suggested in-situ polymerization as an appropriate method for preparing polymer-LDH compounds because of the more confined interlayer spacing compared to a typical unmodified montmorillonite clay (0.78 nm versus 1.26 nm). The high charge density in LDHs and their dense packing favor the insertion of small molecules such as the lactide 295 dimer over the insertion of long PLA chains. This should lead to an effective dispersion of LDH platelets in the growing polymer matrix. Efforts have also focused on the insertion of new organic anionic species in LDH interlayers to enhance/modify hydrophobicity of LDHs [56,57]. Plackett et al. [58] studied in-situ polymerization of the lactide dimmer in the presence of LDHs modified with carbonate (LDH-CO3) or laurate units LDH (LDH-C12). LDH-CO3 was synthesized by a conventional co-precipitation method [59] whereas LDH-C12 was prepared using the reconstruction method. In this procedure the previously synthesized LDH-CO3 was calcined to form a mixed metal oxide (MMO) and then dispersed into an ethanol/water solution containing sodium laureate. This process involved fast rehydration of the layered structure, followed by a slower anion exchange reaction, giving rise to the so-called “memory effect” mechanism [60]. X-ray diffraction, scanning electron microscopy and transmission electron microscopy revealed that exfoliated nanocomposites were obtained when using LDH-C12 but that LDH-CO3 gave a partly phase-separated morphology (Figure A.5). OO O O OO O OO O O O OO O O OO O OO O O O OO O O OO O OO O O O OO O O OO O O O O O O OO O OO O O O OO O O OOO OOOO O O O O O O O OOO OO O O O O O OOOO O O O O OOO OO O O O O O OOO O O O O O O O O O O O OO OO O O O O OO O O OO O O O O O O O O OO OO O O O OO O O O Melt-ROP Sn based catalyst OO O O OO O O Carbonate anion: Water: Lauric acid anions: Lactide (L-Lactic cyclic dimer): OO O O LDH with Lauric anions LDH with carbonate anions and water Intercalated LDH/PLA nanocomposite Exfoliated LDH/PLA nanocomposite Figure A.5 Ring opening polymerization of lactide rings in the presence of layered double hydroxides modified with carbonate or laurate units. Thermogravimetric analysis showed that PLA-LDH combinations exhibited higher degradation onset temperatures than unfilled PLA. Differential scanning calorimetry indicated that both crystallinity and temperature of crystallization increased on adding LDH-C12 or LDH-CO3, suggesting that these additives have a nucleating effect. Although in-situ polymerization of lactide in the presence of 1-5% LDH-C12 could be a promising method for producing nanocomposites with an exfoliated structure, the molecular weight was significantly reduced when compared with the polymer synthesized in the absence of LDHs. Since this phenomenon 296 also occurred when Mg(OH)2 was used instead of LDH, a chain-termination mechanism via LDH surface hydroxyl groups and/or metal-catalyzed degradation was proposed. From an industrial point of view, the melt-intercalation technique is usually preferred to in-situ polymerization because it is simpler and uses already existing technologies. In this sense, methods that combine the efficiency of the in-situ polymerization approach and the practicability of the melt-intercalation technique are currently being applied. Thus, a highly filled polymer/clay masterbatch was first synthesized using an appropriate solvent and then dispersed into the commercial polymer by melt blending. Detrembleur et al. [61] used supercritical carbon dioxide as a polymerization medium for in-situ polymerization of D,L-lactide in the presence of different organo-modified clays (e.g. C20A and C30B). Polymerizations were performed at 85 ºC and 240 bar where the lactide monomer was partially soluble whereas the formed polymer precipitated. The polymerization kinetic rate was observed to decrease with increasing the clay amount due to significant hindrance of the clay. A slightly higher conversion was observed in the C30B-based systems since its hydroxyl groups acted efficiently as polymerization initiators. Studies carried out at low clay levels (3 wt%) demonstrated that final structures were also strongly influenced by the nature of the organomodifier compound since intercalated or exfoliated nanocomposites were obtained using C20A and C30B, respectively. Nanocomposites were also successfully obtained with clay levels as high as 35-50% and were then useful as masterbatches to be mixed with the commercial PLA matrix. In this way, well-delaminated nanocomposites with 3 wt% of Cloisite 30B were attained, as deduced from TEM and XRD analysis. The nanocomposites showed significant improvement in both stiffness (up to 20%) and toughness (e.g. from 5.1 to 6.0 kJ/m2) compared with the unfilled matrix.  2.3. Poly(butylen succinate) based nanocomposites Poly(butylene succinate) (PBS) is currently one of the most commonly applied biodegradable polyesters of the poly(alkylene dicarboxylate) family. It is commercialized by Showa Highpolymer as BIONOLLE and is usually copolymerized and blended to improve mechanical properties and biodegradability [62-65]. However, these methods (i.e. copolymerization and blending) often cause problems that affect crystallinity and melting point since the incorporation of a second component results in imperfect packing and/or isomorphism, and adversely affects the temperature range over which the resulting materials can be used [66,67]. Preparation of nanocomposites appears as a promising alternative method to the production of commercial PBS-based polymers. In this case, the aggregation trend of inorganic materials caused by strong hydrophilic interactions should be avoided. 303 [29] Pucciariello, R., Villani, V., Langerame, F., Gorrasi, G., Vittoria, V. (2004) Interfacial Effects in Organophilic Montmorillonite–Poly(α-caprolactone) Nanocomposites. J. Polym. Sci.: Part B: Polym. Phys., 42 (21), 3907–3919. [30] Gorrasi, G., Tortora, M., Vittoria, V., Pollet, E., Lepoittevin, B., Alexandre, M., Dubois, P. (2003) Vapor barrier properties of polycaprolactone montmorillonite nanocomposites: effect of clay dispersion. Polymer, 44 (8), 2271-2279. [31] Pucciariello, R., Villani, V., Belviso, S., Gorrasi, G., Tortora, M., Vittoria, V. J. (2004) Phase behavior of modifired montmorillonite-poly(-caprolactone) nanocomposites. J. Polym. Sci. Part B: Polym. Phys., 42(7), 1321-1332. [32] Tortora, M., Vittoria, V., Galli, G., Ritrovati, S., Chiellini, E. (2002) Transport Properties of Modified Montmorillonite-Poly(-caprolactone) Nanocomposites. Macromol. Mater. Eng., 287 (4), 243-249. [33] Eastmond, G.C. (1999) Poly(caprolactone) Blends . Adv Polym Sci., 149, 59-223. [34] Kiersnowski, A., Piglowski, J. (2004) Polymer-layered silicate nanocomposites based on poly(- caprolactone). Eur. Polym. J., 40 (6), 1199–1207. [35] Stassin, F., Halleux, O., Jérôme, R. (2001) Ring-opening polymerization of caprolactone in supercritical carbon dioxide. Macromolecules, 34 (22), 775–781. [36] Bratton, D., Brown, M., Howdle, S.M. (2002) Suspension polymerization of L-lactide in supercritical carbon dioxide in the presence of a triblock copolymer stabilizer. Macromolecules, 36 (16), 5908–5911. [37] Wells, S.L., DeSimone, J.M. (2001) CO2 technology platform: an important tool for environmental problem solving. Angew. Chem. Int. Ed., 40 (3), 518–527. [38] Urbanczyk, L., Calberg, C., Stassin, F., Alexandre, M., Jérôme, R., Jérôme, C., Detrembleur, C. (2008) Synthesis of PCL/clay masterbatches in supercritical carbon dioxide. Polymer, 49 (18), 3979–3986. [39] Stassin, F., Jérôme, R. (2001) Ring-opening polymerization of epsilon-caprolactone in supercritical carbon dioxide. Macromolecules, 34 (4), 775–781. [40] Avella, M., Vlieger, J.J.D., Errico, M.E., Fischer, S., Vacca, P., Volpe, M.G. (2005). Biodegradable starch/clay nanocomposite films for food packaging applications. Food Chem., 93 (3), 467–474. [41] Pandey, J.K., Kumar, A.P., Misra, M., Mohanty, A.K., Drzal, L.T., Singh, R.P. (2005) Recent advances in biodegradable nanocomposites. J. Nanosci. Nanotechnol., 5 (4), 497–525. [42] Namazi, H., Mosadegh, M., Dadkhah, A. (2009) New intercalated layer silicate nanocomposites based on synthesized starch-g-PCL prepared via solution intercalation and in situ polymerization methods: As a comparative study. Carbohyd. Polym., 75 (4), 665–669. [43] Ray, S.S, Maiti, P., Okamoto, M., Yamada, K., Ueda, K. (2002) New polylactide/layered silicate nanocomposites. 1. Preparation, characterization, and properties. Macromolecules, 35 (8), 3104– 3110. [44] Ray, S.S, Okamoto, K., Yamada, K., Okamoto, M. (2002) Novel porous ceramic material via burning of polylactide/layered silicate nanocomposite. Nano. Lett. 2 (4), 423–425. [45] Paul, M.A., Alexandre, M., Degée, P., Calberg, C., Jérôme, R., Dubois, P. (2003) Exfoliated polylactide/clay nanocomposites by in situ coordination-insertion polymerization. Macromol. Rapid. Commun. 24 (6), 66–561. [46] Paul, M.A., Delcourt, C., Alexandre, M., Degée, P., Monteverde, F., Rulmont, A., Dubois, P. (2005) (Plasticized) polylactide/(organo-)clay nanocomposites by in situ intercalative polymerization. Macromol. Chem. Phys., 206 (4), 484–498. [47] Jacobsen, S.; Fritz, H.G. (1999) Plasticizing polylactide - The effect of different plasticizers on the mechanical properties. Polym. Eng. Sci., 39 (7), 1303-1310. [48] Martin, O.; Ave´rous, L. (2001) Poly(lactic acid): plasticization and properties of biodegradable multiphase systems. Polymer, 42 (14), 6209-6219. 304 [49]. Baiardo, M., Frisoni, G., Scandola, M., Rimelen, M., Lips, D., Ruffieux, K., Wintermantel, E. (2003) Thermal and mechanical properties of plasticized poly(L-lactic acid). J. Appl. Polym. Sci., 90 (7), 1731-1738. [50] Hu, Y., Rogunova, M., Topolkaraev, V., Hiltner, A., Baer, E. (2003) Aging of poly(lactide)/poly(ethylene glycol) blends. Part 1. Poly(lactide) with low stereoregularity. Polymer, 44 (19), 5701-5710. [51] Zhang, J.H., Zhuang, W., Zhang, Q., Liu, B., Wang, W., Hu, B.X., Shen, J. (2007) Novel polylactide/vermiculite nanocomposites by in situ intercalative polymerization. I. preparation, characterization, and properties. Polym. Composite., 28 (4), 545-550. [52] Saber, O., Tagaya, H. (2008) Preparation and intercalation reactions of nano-structural materials, Zn-Al-Ti LDH. Mater. Chem. Phys., 108 (2-3), 449-455. [53] Pluta, M. (2004) Morphology and properties of polylactide modified by thermal treatment, filling with layered silicates and plasticization. Polymer, 45 (24), 8239-8251. [54] Li, L., Ma, R.Z., Ebina, Y., Iyi, N., Sasaki, T. (2005) Positively charged nanosheets derived via total delamination of layered double hydroxides. Chem. Mat., 17 (17), 4386-4391. [55] Zammarano, M., Bellayer, S., Gilman, J.W., Franceschi, M., Beyer, F.L., Harris, R.H., Meriani, S. (2006) Delamination of organo-modified layered double hydroxides in polyamide 6 by melt processing. Polymer, 47 (2), 652-662. [56] Sabbar, E.M., Roy, M.E., Leroux, F. (2006) Probing the interaction between di- and trifunctionalized carboxy-phosphonic acid and LDH layer structure. J. Phys. Chem. Solids, 67 (11), 2419-2429. [57] Jaubertie, C., Holgado, M.J., San Roman, M.S., Rives, V. (2006) Structural characterization and delamination of lactate-intercalated Zn, Allayered double hydroxides. Chem. Mater, 18 (13), 3114-3121. [58] Katiyar, V., Gerds, N., Koch, C.B., Risbo, J., Hansen, H.C.B., Plackett, D. (2010) Poly L-lactide- layered double hydroxide nanocomposites via in situ polymerization of L-lactide. Polym. Degrad. Stabil., 95 (12), 2563-2573. [59] Miyata, S. (1975) The syntheses of hydrotalcite-like compounds and their structures and physicochemical properties 1. systems Mg2+-Al3+-NO-3, Mg2+ -Al3+-Cl- ,Mg2+-Al3+-ClO-4, Ni2+- Al3+-Cl- and Zn2+-Al3+-Cl-. Clays Clay Miner., 23(5), 369-375. [60] Cavani, F., Clause, O., Trifiro, F., Vaccari, A. (1991) Anionic clays with hydrotalcite-like structure as precursors of hydrogenation catalysts. Adv. Catal. Des. 186-190. [61] Urbanczyk, L., Ngoundjo, F., Alexandre, M., Jérôme, C., Detrembleur, C., Calberg, C. (2009) Synthesis of polylactide/clay nanocomposites by in situ intercalative polymerization in supercritical carbon dioxide. Eur. Polym. J., 45 (3), 643–648. [62] Fujimaki, T. (1998) Processability and properties of aliphatic polyesters, ‘BIONOLLE’, synthesized by polycondensation reaction. Polym. Degrad. Stabil., 59 (1-3), 209-214. [63] Yoo, Y.T., Lee, B.J., Han, S.I., Im, S.S., Kim, D.K. (2003) Physical properties and biodegradation of poly(butylene adipate) ionomers. Polym. Degrad. Stabil., 79 (2), 257-264. [64] Han, S.I., Kang, S.W., Kim, B.S., Im, S.S. (2005) A novel polymeric ionomer as a potential biomaterial: crystallization behavior, degradation, and in-vitro cellular interactions. Adv. Funct. Mater., 15 (3), 367-374. [65] Han, S.I., Yoo, Y., Kim, D.K., Im, S,S. (2004) Biodegradable aliphatic polyester ionomers. Macromol. Biosci., 4 (3), 199-207. [66] Takiyama, E., Fujimaki, T. (1992) Characteristics of biodegradable aliphatic polymer e bionolle. Plastics, 43, 87. [67] Abe, H., Doi, Y., Hori, Y., Hagiwara, T. (1998) Physical properties and enzymatic degradability of copolymers of (R)-3-hydroxybutyric acid and (S,S )- lactide. Polymer, 39 (1), 59-67. [68] Kim, H.S., Chen, G.X., Jin, H.J., Yoon, J.S. (2008) In situ copolymerization of butylene succinate with twice functionalized organoclay: Thermal stability. Colloid Surface A, 313-314, 56-59. 305 [69] Han, S., Lim, J.S., Kim, D.K., Kim, M.N., Im, S.S. (2008) In situ polymerized poly(butylene succinate)/silica nanocomposites: Physical properties and biodegradation. Polym. Degrad. Stabil., 93 (5), 889-895. [70] Hwang, S., Yoo, E.S., Im, S.S. (2009) Effect of the urethane group on treated clay surfaces for high-performance poly(butylene succinate)/montmorillonite nanocomposites. Polym. Degrad. Stabil., 94 (12), 2163–2169. [71] Pollet, E., Delcourt, C., Alexandre, M., Dubois, Ph. (2006) Transesterification catalysts to improve clay exfoliation in synthetic biodegradable polyester nanocomposites. Eur. Polym. J., 42 (6), 1330– 1341. [72] Yang, K.K., Wang, X.L., Wang, Y.Z. (2002) Poly(p-Dioxanone) and its copolymers. J. Macromol. Sci. Part C Polym. Rev., 42 (3), 373-398. [73] Yang, K.K., Wang, X.L., Wang, Y.Z. (2007) Progress in Nanocomposite of Biodegradable Polymer. J. Ind. Eng. Chem., 13 (4), 485-500. [74] Huang, F.Y., Wang, Y.Z., Wang, X.L., Yang, K.K., Zhou, Q., Huang, F.Y. (2005) Preparation and characterization of a novel biodegradable poly(p-dioxanone)/montmorillonite nanocomposite. J. Polym. Sci. Polym. Chem., 43 (11), 2298-2303. [75] Morales, L; Franco, L; Casas, MT; Puiggalí, J. (2009) Poly(ester amide)/Clay Nanocomposites Prepared by In Situ Polymerization of the Sodium Salt of N-Chloroacetyl-6-Aminohexanoic Acid. J. Polym. Sci. Part A: Polym. Chem., 47 (14), 3616-3629. [76] Epple, M.; Kirschnick, H. (1996) The Thermally Induced Solid-State Polymerization Reaction in Halogenoacetates. Chem. Ber., 129 (9), 1123-1129. [77] Herzberg, O.; Epple, M. (2001) Formation of Polyesters by Thermally Induced Polymerization Reactions of Molecular Solids. Eur. J. Inorg. Chem., 2001 (6), 1395-1406. [78] Morales, L; Franco, L; Casas, MT; Puiggalí, J. (in press) Crystallization studies on a clay nanocomposite prepared from a degradable poly(ester amide) constituted by glycolic acid and 6- aminohexanoic acid. Polym. Eng. Sci., 2010, in press.