Final Report Summary - SMARTPIEZOCOMPOSITE (Crystallisation of polymers in the presence of inorganic nanoparticles - a way towards piezoelectric 'smart' nanocomposites)
The research was aimed at finding how inorganic nanoparticles influence structural as well as kinetic aspects of crystallisation of polymers in selected polymer-inorganic nanocomposites. The project was essentially divided into three sub-streams:
(i) investigation of poly(vinylidene fluoride) (PVDF) nanocomposites;
(ii) investigation of poly(L-lactic acid) (PLLA) nanocomposites;
(iii) determination how geometry and surface properties of nanoparticles affect the phase behaviour of selected model systems.
Results from these studies were supposed to clarify whether it is possible to use inorganic nanoparticles to control polymorphism and orientational structure of polymers having potential in the field of flexible piezoelectrics. To optimise the overall morphology of the materials, several formation strategies were suggested in the project proposal.
Melt blending of unmodified, hydrophilic smectite silicates, namely either montmorillonite or fluoromica, with PVDF resulted, as expected, in the formation of phase-separated microcomposites. Both the silicates caused however strong heterogeneous nucleation effects manifested, amongst others, by threefold faster crystallisation rate of PVDF in the composites and noticeably increased crystallinity in comparison to the values recorded for the pure polymer. The crystallisation of PVDF in the presence of the unmodified silicates leads to formation the most common, simple monoclinic system with chains adopting TGTG' conformation, known as the alpha-polymorph. Uni- or biaxial orientation of smectite tactoids (i.e. planar stacks consisted of 10 - 20 platelets having approximately 15 - 30 nm in thickness) does not change the crystalline structure of the polymer at the unit cell level but was found to exert an influence on its orientational arrangement. Interestingly, in the solution-processed films having analogous compositions to those melt-extruded, the in-plane orientation of silicate particles practically did not affect the anisotropy of the polymer. Instead, the oriented tactoids caused formation the gamma-polymorph being a monoclinic arrangement of PVDF chains in TTTG conformation. The nanocomposites made of PVDF and organoclays were found most promising for applications as plastic piezoelectric materials since, under certain conditions, they were able to promote the formation of the desired, polar beta-polymorph (orthorhombic, TTTT). The main factor governing the formation of the beta-polymorph of PVDF was a careful control of the particles surface properties.
Application of commercially available silicate nanofillers as Nanomer (types 1.44P or 1.28E) or Cloisite (C20A) for the formation of the intercalated nanocomposites caused crystallisation of PVDF into mixed crystal systems containing mainly beta and gamma polymorphs. It is important to mention here, that commercial organoclays are typically obtained by the non-stoichiometric ion exchange of the natural sodium montmorillonite with mixtures of various amphiphilic compounds. Replacing these organoclays with the ones obtained either by stoichiometric ion-exchange with selected single quaternary alkylammonium compounds or synthesised by grafting the silicate with poly(methyl methacrylate) (PMMA) via RAFT polymerisation technique allowed PVDF crystallising almost entirely in the beta-polymorph. Moreover, in the case of alkylammonium-exchanged silicates it was found that both the crystallinity degree and crystallisation rate were clearly controlled by the length of alkyl chains of ammonium molecules: the longer the chains the higher the crystallinity and the lower crystallisation rate. Unlike the polymorphism, the orientational structure of the nanocomposites was found independent of surface chemistry of the organoclays. X-ray scattering studies on the extruded fibres and filaments revealed that the extrusion of PVDF-based nanocomposites causes a strong alignment of either montmorillonite or fluoromica particles along the extrusion axis. While, however, in the PVDF / organomontmorillonite systems the orientation of polymer crystals was only insignificantly higher than in the extruded, the orientation of the polymer in PVDF / fluoromica systems was found as high as in commercially available, biaxially stretched PVDF films designed for piezoelectric applications. Time-resolved X-ray diffraction (XRD) studies performed at the European Synchrotron Radiation Facility revealed that the oriented arrays of PVDF crystals are formed when the polymer solidifies from a stretched melt. Comparison of results obtained for the nanocomposites with different fillers indicated that that the main factor governing the orientational structure of the polymer is the aspect ratio of nanoparticles. Those with high aspect ratios, as the fluoromica whose particle diameter (ca. 1.2 microns) is fourfold higher than of montmorillonite, more effectively restrict the relaxation (coiling) of polymer chains in the direction transverse to the extrusion flow. Hence, the solidification occurs by incorporating the oriented polymer chains to growing lamellar crystals before isotropisation of the polymer melt. Further studies on PVDF-based systems revealed that organomontmorillonites can also be used to control the anisotropy of the polymer in the solution-processed films. The orientation effects were however achieved at much higher clay loadings in comparison to the systems with the fluoromica.
Formation of the oriented nanocomposites with PLLA matrix was found much more challenging task. Both melt blending and an in-situ polymerisation of L,L-lactide allowed formation of intercalated nanocomposites, but were found ineffective in obtaining materials having controllable crystallisation behaviour. In the first case the polymerisation resulted in non-stereoregular, low-molecular products due to isomerisation and premature termination processes. The main problem in the melt-compounding was a limited thermal stability of the polymer under the processing conditions. In terms of formation of nanocomposites, the most effective strategies to intercalate and exfoliate silicate particles were those involving poly(methyl methacrylate) as a co-intercalation agent. This is due to high affinity of PMMA to organoclays and miscibility of this polymer with PLLA above the upper critical solubility temperature (~ 220 degrees of Celsius). Unfortunately, like in the case of PVDF, the presence of PMMA in the systems was found to severely decrease the crystallisation rate of the polymer. The slow crystallisation of the melt-processed ternary nanocomposites resulted in formation of almost amorphous, isotropic systems. Unlike in the PVDF-based systems and contrary to the case of PMMA-grafted silicates, the presence of organoclays obtained by ion-exchange in the PLLA matrix exerts only marginal influence on kinetics of crystallisation upon cooling. The presence of organoclays had, however, a clear influence on the cold crystallisation behavior of PLLA. The pure polymer slowly heated after a dynamic cooling cycle revealed only a single regime of the cold crystallisation at approx. 100 degrees of Celsius. In the case of nanocomposites two crystallisation modes were discerned at 100 and at 155 degrees of Celsius. In either of these regimes different crystalline forms were developed: in the first, the cold crystallisation leads to the orthorhombic beta-polymorph, while at the higher temperature formation of the alpha-polymorph comprising the 103 helix conformation was dominating. The contribution of the second crystallisation process and thus formation of the alpha-polymorph was found increasing with the length of alkyl chains of the organoclay particles. The results of XRD studies of PLLA-based systems suggest that the isotropisation of the polymer melt is much faster than crystallisation process, which, together with the negligible heterogeneous nucleation effects, leads to isotropic structure of the polymer. In order to utilise cold crystallisation effects to obtain oriented structures, the systems were prepared by spin-coating of mixtures of the polymer with organosilicates modified with long-tailed ammonium salts. Although such a processing has never led to materials having structures truly integrated at the nanoscale, it allowed, to some extent, controlling the orientational morphology of PLLA-based films. Typically the films directly after formation were amorphous. Grazing-incidence XRD studies revealed, however, that annealing at 150 - 155 degrees of Celsius (i.e. just below the melting point) leads to partial dewetting of the substrate and formation the lamellar structures of polymer chains packed in orthorhombic unit cells. As a result of such processing the polymer backbones showed a preferred orientation in the direction normal to the substrate, though a considerable amount of PLLA remained isotropic. In the case of the systems containing more than 20 wt. % of organically modified silicates, due to high viscosity of the material deposited on the surface no dewetting was observed. In these composites, the polymer chains were aligned mostly parallel to the surface. Structural analysis suggested that polymer crystallises in the form of fibrils rather than lamellar systems. Similarly like in the case of PVDF-based systems, the degree of orientation in PLLA composites was found depending on the aspect ratio of the organoclays. Generally, the systems containing the fluoromica tended to have a clear texture at lower loadings than those based on montmorillonite.
In conclusion - the proposed method to control the morphology of polymers can, under some conditions, be applied in relation to linear semicrystalline polymers including those having a potential in the field of piezoelectric materials. The main restriction here stems from the kinetic aspects of polymer crystallisation: a requirement to control the anisotropy of the polymer is the crystallisation rate higher than structural relaxation (isotropisation) of the melt. The role of nanoparticles is to prevent polymer macromolecules from diffusion in the direction transverse to the flow. This is why the particles with higher aspect ratios were found more effective as anisotropy-driving agents. The polymorphic behavior of polymers can be controlled by changing the surface properties of nanofillers. Despite however numerous experimental results, at this stage it is impossible to propose a plausible and consistent explanation of this effect.
In spite of the results discussed above indicate that obtaining structures molecularly oriented in the same way as commercially available polymer piezoproducts, the formation of nanocomposites revealing piezoelectric effect at the macroscopic scale was found not possible. This was because none of the methods applied for making the nanocomposites did allow controlling the domain structure of polymers. Poling the materials during the processing, in the similar way as in the case of conventional plastic piezoelectrics, would allow obtaining materials with all domains having unidirectional dipole moments. Such additional, electrical orientation could be imparted to the materials by melt extrusion in the high electric field. This, however, would demand designing and constructing additional electromechanical devices, as e.g. a slit die in which the extrudate would flow through a constant electric field zone. Designing of such devices was however beyond the scope of this project although will be considered as a topic for future research. It is worth stressing here that, despite of failure in formation of piezoelectric nanocomposites in this project, the approach as studied can still be considered as advantageous over the conventional methods of fabrication of plastic piezoelectric materials. The formation of e.g. PVDF-based piezoelectric materials requires achieving three goals: (i) Crystallisation of the polymer in the beta-form; (ii) uniaxial (structural) orientation of polymer backbones in the crystalline systems; (iii) unidirectional arrangement of molecular dipoles in the electric domains. Typically to satisfy these requirements three separate processes are necessary. Here we offer a strategy that allows controlling the polymorphism and the orientation in a single step. This, in turn, really opens an avenue to obtain piezoelectric materials having complex shapes, though the processing methodology requires additional improvements.
The key to understand the physical phenomena governing crystallisation processes was a thorough understanding of structure and properties of materials investigated in the project. Special emphasis was placed on analysis of the layered silicates. Structure of these complex systems was found not clear, even though a large number of publications tend to imply this is a solved problem. Deeper understanding of the possible variations in arrangements of structural units and ways to quantify are crucial to clarifying the broad empirical and many times contradictory reports on crystallisation of polymer nanocomposites. In order to gain a deeper insight in this issue as well as other properties of the systems were analysed using a multi-technique approach providing information about different features of the analysed structures. Combination of conventional XRD techniques with the solid-state nuclear magnetic resonance spectroscopy and quantum-chemical calculations allowed analysing the structures at different scales of assembly and creating a full structural model of the analysed materials. This approach was found extremely efficient in analysing layered nanostructures as well as packing the polymer chains in the crystalline domains. Combination of such an approach with advanced XRD techniques allowed studying different materials in the form of fibres as well as films deposited on solid surfaces or free standing. In addition, development of X-ray scattering tools at the host institution - the Max Planck Institute for Polymer Research - provided opportunities for both external and internal collaborations and resulted in a number of new collaborative projects that will be continued in future.
(i) investigation of poly(vinylidene fluoride) (PVDF) nanocomposites;
(ii) investigation of poly(L-lactic acid) (PLLA) nanocomposites;
(iii) determination how geometry and surface properties of nanoparticles affect the phase behaviour of selected model systems.
Results from these studies were supposed to clarify whether it is possible to use inorganic nanoparticles to control polymorphism and orientational structure of polymers having potential in the field of flexible piezoelectrics. To optimise the overall morphology of the materials, several formation strategies were suggested in the project proposal.
Melt blending of unmodified, hydrophilic smectite silicates, namely either montmorillonite or fluoromica, with PVDF resulted, as expected, in the formation of phase-separated microcomposites. Both the silicates caused however strong heterogeneous nucleation effects manifested, amongst others, by threefold faster crystallisation rate of PVDF in the composites and noticeably increased crystallinity in comparison to the values recorded for the pure polymer. The crystallisation of PVDF in the presence of the unmodified silicates leads to formation the most common, simple monoclinic system with chains adopting TGTG' conformation, known as the alpha-polymorph. Uni- or biaxial orientation of smectite tactoids (i.e. planar stacks consisted of 10 - 20 platelets having approximately 15 - 30 nm in thickness) does not change the crystalline structure of the polymer at the unit cell level but was found to exert an influence on its orientational arrangement. Interestingly, in the solution-processed films having analogous compositions to those melt-extruded, the in-plane orientation of silicate particles practically did not affect the anisotropy of the polymer. Instead, the oriented tactoids caused formation the gamma-polymorph being a monoclinic arrangement of PVDF chains in TTTG conformation. The nanocomposites made of PVDF and organoclays were found most promising for applications as plastic piezoelectric materials since, under certain conditions, they were able to promote the formation of the desired, polar beta-polymorph (orthorhombic, TTTT). The main factor governing the formation of the beta-polymorph of PVDF was a careful control of the particles surface properties.
Application of commercially available silicate nanofillers as Nanomer (types 1.44P or 1.28E) or Cloisite (C20A) for the formation of the intercalated nanocomposites caused crystallisation of PVDF into mixed crystal systems containing mainly beta and gamma polymorphs. It is important to mention here, that commercial organoclays are typically obtained by the non-stoichiometric ion exchange of the natural sodium montmorillonite with mixtures of various amphiphilic compounds. Replacing these organoclays with the ones obtained either by stoichiometric ion-exchange with selected single quaternary alkylammonium compounds or synthesised by grafting the silicate with poly(methyl methacrylate) (PMMA) via RAFT polymerisation technique allowed PVDF crystallising almost entirely in the beta-polymorph. Moreover, in the case of alkylammonium-exchanged silicates it was found that both the crystallinity degree and crystallisation rate were clearly controlled by the length of alkyl chains of ammonium molecules: the longer the chains the higher the crystallinity and the lower crystallisation rate. Unlike the polymorphism, the orientational structure of the nanocomposites was found independent of surface chemistry of the organoclays. X-ray scattering studies on the extruded fibres and filaments revealed that the extrusion of PVDF-based nanocomposites causes a strong alignment of either montmorillonite or fluoromica particles along the extrusion axis. While, however, in the PVDF / organomontmorillonite systems the orientation of polymer crystals was only insignificantly higher than in the extruded, the orientation of the polymer in PVDF / fluoromica systems was found as high as in commercially available, biaxially stretched PVDF films designed for piezoelectric applications. Time-resolved X-ray diffraction (XRD) studies performed at the European Synchrotron Radiation Facility revealed that the oriented arrays of PVDF crystals are formed when the polymer solidifies from a stretched melt. Comparison of results obtained for the nanocomposites with different fillers indicated that that the main factor governing the orientational structure of the polymer is the aspect ratio of nanoparticles. Those with high aspect ratios, as the fluoromica whose particle diameter (ca. 1.2 microns) is fourfold higher than of montmorillonite, more effectively restrict the relaxation (coiling) of polymer chains in the direction transverse to the extrusion flow. Hence, the solidification occurs by incorporating the oriented polymer chains to growing lamellar crystals before isotropisation of the polymer melt. Further studies on PVDF-based systems revealed that organomontmorillonites can also be used to control the anisotropy of the polymer in the solution-processed films. The orientation effects were however achieved at much higher clay loadings in comparison to the systems with the fluoromica.
Formation of the oriented nanocomposites with PLLA matrix was found much more challenging task. Both melt blending and an in-situ polymerisation of L,L-lactide allowed formation of intercalated nanocomposites, but were found ineffective in obtaining materials having controllable crystallisation behaviour. In the first case the polymerisation resulted in non-stereoregular, low-molecular products due to isomerisation and premature termination processes. The main problem in the melt-compounding was a limited thermal stability of the polymer under the processing conditions. In terms of formation of nanocomposites, the most effective strategies to intercalate and exfoliate silicate particles were those involving poly(methyl methacrylate) as a co-intercalation agent. This is due to high affinity of PMMA to organoclays and miscibility of this polymer with PLLA above the upper critical solubility temperature (~ 220 degrees of Celsius). Unfortunately, like in the case of PVDF, the presence of PMMA in the systems was found to severely decrease the crystallisation rate of the polymer. The slow crystallisation of the melt-processed ternary nanocomposites resulted in formation of almost amorphous, isotropic systems. Unlike in the PVDF-based systems and contrary to the case of PMMA-grafted silicates, the presence of organoclays obtained by ion-exchange in the PLLA matrix exerts only marginal influence on kinetics of crystallisation upon cooling. The presence of organoclays had, however, a clear influence on the cold crystallisation behavior of PLLA. The pure polymer slowly heated after a dynamic cooling cycle revealed only a single regime of the cold crystallisation at approx. 100 degrees of Celsius. In the case of nanocomposites two crystallisation modes were discerned at 100 and at 155 degrees of Celsius. In either of these regimes different crystalline forms were developed: in the first, the cold crystallisation leads to the orthorhombic beta-polymorph, while at the higher temperature formation of the alpha-polymorph comprising the 103 helix conformation was dominating. The contribution of the second crystallisation process and thus formation of the alpha-polymorph was found increasing with the length of alkyl chains of the organoclay particles. The results of XRD studies of PLLA-based systems suggest that the isotropisation of the polymer melt is much faster than crystallisation process, which, together with the negligible heterogeneous nucleation effects, leads to isotropic structure of the polymer. In order to utilise cold crystallisation effects to obtain oriented structures, the systems were prepared by spin-coating of mixtures of the polymer with organosilicates modified with long-tailed ammonium salts. Although such a processing has never led to materials having structures truly integrated at the nanoscale, it allowed, to some extent, controlling the orientational morphology of PLLA-based films. Typically the films directly after formation were amorphous. Grazing-incidence XRD studies revealed, however, that annealing at 150 - 155 degrees of Celsius (i.e. just below the melting point) leads to partial dewetting of the substrate and formation the lamellar structures of polymer chains packed in orthorhombic unit cells. As a result of such processing the polymer backbones showed a preferred orientation in the direction normal to the substrate, though a considerable amount of PLLA remained isotropic. In the case of the systems containing more than 20 wt. % of organically modified silicates, due to high viscosity of the material deposited on the surface no dewetting was observed. In these composites, the polymer chains were aligned mostly parallel to the surface. Structural analysis suggested that polymer crystallises in the form of fibrils rather than lamellar systems. Similarly like in the case of PVDF-based systems, the degree of orientation in PLLA composites was found depending on the aspect ratio of the organoclays. Generally, the systems containing the fluoromica tended to have a clear texture at lower loadings than those based on montmorillonite.
In conclusion - the proposed method to control the morphology of polymers can, under some conditions, be applied in relation to linear semicrystalline polymers including those having a potential in the field of piezoelectric materials. The main restriction here stems from the kinetic aspects of polymer crystallisation: a requirement to control the anisotropy of the polymer is the crystallisation rate higher than structural relaxation (isotropisation) of the melt. The role of nanoparticles is to prevent polymer macromolecules from diffusion in the direction transverse to the flow. This is why the particles with higher aspect ratios were found more effective as anisotropy-driving agents. The polymorphic behavior of polymers can be controlled by changing the surface properties of nanofillers. Despite however numerous experimental results, at this stage it is impossible to propose a plausible and consistent explanation of this effect.
In spite of the results discussed above indicate that obtaining structures molecularly oriented in the same way as commercially available polymer piezoproducts, the formation of nanocomposites revealing piezoelectric effect at the macroscopic scale was found not possible. This was because none of the methods applied for making the nanocomposites did allow controlling the domain structure of polymers. Poling the materials during the processing, in the similar way as in the case of conventional plastic piezoelectrics, would allow obtaining materials with all domains having unidirectional dipole moments. Such additional, electrical orientation could be imparted to the materials by melt extrusion in the high electric field. This, however, would demand designing and constructing additional electromechanical devices, as e.g. a slit die in which the extrudate would flow through a constant electric field zone. Designing of such devices was however beyond the scope of this project although will be considered as a topic for future research. It is worth stressing here that, despite of failure in formation of piezoelectric nanocomposites in this project, the approach as studied can still be considered as advantageous over the conventional methods of fabrication of plastic piezoelectric materials. The formation of e.g. PVDF-based piezoelectric materials requires achieving three goals: (i) Crystallisation of the polymer in the beta-form; (ii) uniaxial (structural) orientation of polymer backbones in the crystalline systems; (iii) unidirectional arrangement of molecular dipoles in the electric domains. Typically to satisfy these requirements three separate processes are necessary. Here we offer a strategy that allows controlling the polymorphism and the orientation in a single step. This, in turn, really opens an avenue to obtain piezoelectric materials having complex shapes, though the processing methodology requires additional improvements.
The key to understand the physical phenomena governing crystallisation processes was a thorough understanding of structure and properties of materials investigated in the project. Special emphasis was placed on analysis of the layered silicates. Structure of these complex systems was found not clear, even though a large number of publications tend to imply this is a solved problem. Deeper understanding of the possible variations in arrangements of structural units and ways to quantify are crucial to clarifying the broad empirical and many times contradictory reports on crystallisation of polymer nanocomposites. In order to gain a deeper insight in this issue as well as other properties of the systems were analysed using a multi-technique approach providing information about different features of the analysed structures. Combination of conventional XRD techniques with the solid-state nuclear magnetic resonance spectroscopy and quantum-chemical calculations allowed analysing the structures at different scales of assembly and creating a full structural model of the analysed materials. This approach was found extremely efficient in analysing layered nanostructures as well as packing the polymer chains in the crystalline domains. Combination of such an approach with advanced XRD techniques allowed studying different materials in the form of fibres as well as films deposited on solid surfaces or free standing. In addition, development of X-ray scattering tools at the host institution - the Max Planck Institute for Polymer Research - provided opportunities for both external and internal collaborations and resulted in a number of new collaborative projects that will be continued in future.