Abstract
The development of shape-memory polymer composites (SMPCs) enables high recovery stress levels as well as novel functions such as electrical conductivity, magnetism, and biofunctionality. In this review chapter the substantial enhancement in mechanical properties of shape-memory polymers (SMPs) by incorporating small amounts of stiff fillers will be highlighted exemplarily for clay and polyhedral oligomeric silsesquioxanes (POSS). Three different functions resulting from adding functional fillers to SMP-matrices will be introduced and discussed: magnetic SMPCs with different types of magnetic nanoparticles, conductive SMPCs based on carbon nanotubes (CNTs), carbon black (CB), short carbon fiber (SCF), and biofunctional SMPCs containing hydroxyapatite (HA). Indirect induction of the shape-memory effect (SME) was realized for magnetic and conductive SMPCs either by exposure to an alternating magnetic field or by application of electrical current. Major challenges in design and fundamental understanding of polymer composites are the complexity of the composite structure, and the relationship between structural parameters and properties/functions, which is essential for tailoring SMPCs for specific applications. Therefore the novel functions and enhanced properties of SMPCs will be described considering the micro-/nanostructural parameters, such as dimension, shape, distribution, volume fraction, and alignment of fillers as well as interfacial interaction between the polymer matrix and dispersed fillers. Finally, an outlook is given describing the future challenges of this exciting research field as well as potential applications including automotive, aerospace, sensors, and biomedical applications.
Access provided by Autonomous University of Puebla. Download chapter PDF
Similar content being viewed by others
Keywords
- Carbon nanotubes
- Nanoparticle
- Shape-memory effect
- Shape-memory polymer composite
- Stimuli-sensitive polymer
1 Introduction to Shape-Memory Polymer Composites
Polymer composites are a combination of a polymer matrix and micro/nano-sized fillers such as particles, fibers, platelets, or tubes. The polymer matrix can be an amorphous or crystalline thermoplastic material or a crosslinked three-dimensional polymer network. The matrix holds or binds the fillers together and protects them from damage by distributing any stress through the whole specimen. Polymer composites have received considerable attention over the last decade because of their potential to enhance dramatically properties relative to the neat polymer matrix [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]. Incorporation of small amounts of filler leads to an improvement in material properties, such as modulus, strength, heat resistance, flame retardancy, and lowered gas permeability [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]. In addition, polymer composites could also yield novel functions including electrical [15, 16, 17, 18], magnetic [19, 20, 21, 22, 23, 24, 25], and optical functions [26, 27, 28, 29], as well as biofunctionality [30, 31, 32, 33]. The enhancement of material properties and creation of novel functions has been linked to the interfacial interaction between the polymer matrix and fillers as well as the formation of a network of interconnected filler particles. This network of interconnected particles can conduct heat and electrical current [15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]. Development and tailoring of polymer composites offer the possibility to promote their use in automotive, aerospace, building, electrical, optoelectronic, and biomedical applications [34, 35, 36]. The novel functions and properties enhancement of polymer composites can also be controlled by micro-/nanostructural parameters such as dimension, shape, distribution, volume fraction, alignment, and packing arrangement of fillers. Anisotropic properties could be obtained from fiber filled composites, except for the very short, randomly distributed fibers, whereas physical properties of polymers with randomly distributed particles are isotropic. Furthermore, composites with nanometer-sized fillers had different properties compared to those filled with macro-sized fillers. Some of the properties of nanocomposites, such as increased tensile strength, may be achieved by using higher macro-sized filler concentration at the expense of increased weight and decreased gloss. Other properties of nanocomposites such as optical clarity or improved barrier properties could not be achieved by high concentration of macro-sized fillers. The filler concentration required for substantial improvement in the overall material properties is called critical filler volume fraction. The fillers could be classified either by their geometry or by their size. Three different categories of filler materials could be found, namely particles (e.g., silica, metal, POSS, and other organic and inorganic particles), layered materials (e.g., graphite and layered silicate), and fibrous materials (e.g., nanofibers and single-walled and multi-walled nanotubes).
A morphological characteristic, which is of fundamental importance to the understanding of the structure–property relationship of nanocomposites, is the surface area/volume ratio of the fillers [37]. As illustrated in Fig.1, the change in particle diameter, layer thickness, or fibrous material diameter from micrometer to nanometer changes the surface area/volume ratio by three orders of magnitude. At this scale, there often is a distinct size dependence of the material properties. In addition, the properties of the composite became dominated by the properties of the interface or interphase when the interfacial area drastically increased.
Major challenges in design and fundamental understanding of polymer composites are related to the complexity of the composite structure, dispersibility of fillers, and the relationship between dispersion and optimal properties. Uniform dispersion of nanoparticles and nanotubes against their agglomeration due to van der Waals bonding is the first step in the processing of nanocomposites [38, 39, 40]. Exfoliation of clays and graphitic layers are essential. Several strategies have been studied to achieve well-dispersed fillers in polymer matrix, including melt processing, solvent casting method (often with surface functionalization and/or sonication pretreatment), and in-situ polymerization. Melt processing by itself often led to limited filler dispersion in the polymer matrix. Blending polymer and fillers in solvent or in-situ polymerization resulted in a better dispersion. It is well established that, the modification of the filler surface by grafting of macromolecules onto its surface is preferable. In this case the filler is highly compatible to the polymer matrix. This is often desirable since it provides the best possible adhesion and allows for optimal transfer of stress from the matrix to the fiber.
Shape-memory polymers (SMPs) are mechanically active or smart materials, which can be fabricated in a specific permanent shape, deformed and fixed in a second, temporary shape. They are able to return to their original, permanent shape when exposed to a suitable external stimulus. Examples of external stimuli, which have been applied to trigger the shape-memory effect (SME), are heat and light [41, 42, 43, 44, 45, 46]. On the molecular/morphological level SMPs consist of at least two components: switching domains and permanent netpoints [41]. The switching domains act as molecular switch with a well-defined melting temperature (T m) or a glass transition temperature (T g) and enable the fixation of the temporary shape. The permanent netpoints are physical netpoints (hard domain) associated with a high thermal transition temperature (T m or T g) or covalent netpoints (covalently crosslinked polymer network). They determine the permanent shape of the SMP. SMPs can be folded, rolled, or otherwise packaged in different shapes for storage and later recover the original as-manufactured shapes, without loss of performance. The active movement of SMP largely depends on the nature of the polymer chains, molecular weight, microphase separation between hard and switching domains, and degree of physical or chemical crosslinking.
The light weight, low cost, easy processibility and very high recoverable strain (several hundred percentages) compared to shape-memory metallic alloys (SMAs) (maximum 8%) and ceramics, make SMPs good candidates for many potential applications [47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57]. Despite the demonstrated merits, the relatively low recovery stress of SMPs (3±2MPa) compared to that of SMAs (0.5±0.25GPa) [58] limits the applications of SMPs to a certain extent under constraint conditions. Incorporating especially stiff fillers or fibers is a common way to increase the stiffness and recovery stress of SMPs [59, 60, 61, 62, 63]. In the case of micro-sized fillers, such as chopped carbon, glass or Kevlar fibers, 40–50wt% of filler was necessary to achieve an improvement in polymer stiffness. In the case of nanofillers such as SiC, only 20wt% was required to increase the constrained bending recovery force of epoxy SMP by 50% [59]. However, the incorporation of stiff fillers for improving the recovery force was in most cases accompanied by a decrease in recovery strain. An example is the addition of 30wt% carbon black (CB) to shape-memory polyurethane (SMPU), which led to a reduction in the shape recovery rate from 98% to 65% [60]. Nanocomposites of SMPU with carbon nanotubes (CNT) showed promising improvement in both strain and stress recovery [64]. Similarly, 25% increase in the shape recovery stress of SMPU by adding only 1wt% nanoclay with only a slight decrease in the recovery strain ratio was reported [65].
Another motivation to generate shape-memory polymer composites (SMPCs) besides the substantial improvement in the stress recovery was the introduction of novel functions, which could be obtained by incorporating small amounts of active fillers in the polymer matrix, such as unique electrical and magnetical properties as well as biofunctionality. In most cases the pure SMP can be thermally-actuated by increasing the environmental temperature (direct heating). Mixing SMP with magnetically or electrically active fillers extended the range of suitable stimuli. Composite from SMPs and carbon fillers, such as carbon black (CB) or carbon nanotubes (CNTs), could be actuated through an electric field. The carbon particles significantly reduced the electric resistance and resulted in conductive SMPC, which could be actuated by means of Joule heat. Similarly, the SMPC containing magnetic particles, such as iron oxide or nickel zinc ferrite, could be inductively-actuated by exposure to an alternating magnetic field. The latter approach had the advantage of wireless/remote operation. Example of the use of this versatile combination of features included trusses and torus-shaped structures for lightweight satellite supports, antenna reflectors, and deployable wings for unmanned aerial vehicles. These SMPC allowed users to pack tightly large, lightweight structures into small volumes for later use in orbit or in the atmosphere [66].
In this review chapter the substantial enhancement in mechanical properties will be highlighted exemplarily for clay and polyhedral oligomeric silsesquioxanes (POSS) containing SMPCs (Sects.2 and 3). Three different functions resulting from adding functional fillers to SMP-matrices will be introduced and discussed: magnetic SMPCs (Sect.4), electrically conductive SMPCs based on CNT, CB, or short carbon fiber (SCF) (Sect.5), and biofunctional SMPCs containing hydroxyapatite (HA) (Sect.6). Finally, an outlook is given describing the future challenges of this exciting research field as well as potential applications.
2 Enhancing Mechanical Properties of SMP by Incorporation of Layered Silicate
Nanocomposites from polymers and layered silicates have received great interest, as they often exhibit a remarkable improvement of material properties such as, high moduli [67] and increased strength. Furthermore, heat resistance [68] was increased, while gas permeability [69] and flammability [70] could be decreased. The biodegradability of polymers could be influenced as well [71]. Such composites were also investigated as model systems to study the structure and dynamics of polymers in confined environments [72, 73]. Layered silicates have layer thicknesses on the order of 1nm with very high aspect ratios (e.g., 10–1,000). Three different classes of nanostructural morphologies were described based on the degree of interfacial interactions between the layered silicate and the polymer matrix [73] (see Fig.2). The first nanostructure was characterized by the polymer chains being inserted into the layered silicate structure in a crystallographically regular fashion with a repeat distance of a few nanometers. This morphological structure was called intercalated nanocomposite. The second morphological structure was named flocculated nanocomposites, where intercalated stacked silicate layers some time flocculated caused by hydroxylated edge–edge interactions. The third morphological structure was called exfoliated nanocomposites, where the individual silicate layers are separated in the polymer matrix by average distances, which completely depended on the relative clay content. The lack of affinity between hydrophilic silicate and hydrophobic polymer caused agglomeration of the mineral in the polymer matrix. Surface modification of clay particles facilitated the compatibility with organic polymers.
The interfacial interaction between layered silicate and polymer matrix largely influenced the substantial improvement in mechanical properties, which was observed for different clay/polymer nanocomposites. Clay of high aspect ratio enabled large surface areas (700m2∕g) to be in contact and bond with the polymer matrix. The high degree of polymer-clay surface interaction, which resulted from the high aspect-ratio of the clay platelets impart superior mechanical and barrier properties as compared to those of the base polymer matrix [74]. The increase in mechanical properties could be attributed to the high strength and stiffness clay layers acting as short randomly dispersed fibers, transmitting stress through the specimen and strongly influence the polymer chain mobility [75, 76]. It is well established that exfoliation of clay in polymer matrix causes a reduction in mobility and degree of short-range chain alignment, thus offering resistance to the movement of polymeric chains under stress and increasing the modulus [76].
2.1 SMPU as Polymer Matrix
Nanocomposites from SMPU and clay were prepared to investigate the reinforcement effect of reactive clay by determining the modulus or strength of the composites as well as the shape-memory properties [77]. It was also aimed to understand the negative impact of clay on the hydrogen bonding of the hard segment and its effect on the SME of the nanocomposites. SMPU was normally comprised of alternating soft polyester or polyether and hard polyurethane–urea segments. These two segments underwent microphase separation into hard and switching domains [78, 79, 80]. The microphase separation was responsible for the excellent SME and elastomeric properties of PU. Figure3 depicts schematically the structure of PU elastomers as multiblock copolymers with hard (urethane) and switching (polyester or polyether diol) segments. The TEM shows how these two segments can segregate in different domains having a nanoscale morphology. The very small dark particles in the micrograph are the hard domains, and the bright matrix is the switching domain of the PU [78].
SMPU/clay nanocomposites were synthesized from PCL-diol, methylene diisocyanate, and butanediol [77]. The desired amount of reactive clay (CloisiteⓇ 30B) was added to the reaction mixture just 5min after chain extender and the whole mixture was stirred at the same temperature for additional 20min. This allowed the reaction between –OH groups in the clay with residual –NCO groups in chain-extended PU chains as was studied earlier [81, 82, 83, 84]. The crystalline PCL soft segment (67wt%) was chosen to trigger the SME by melting of the switching segment crystals. This nanocomposite offered networks constructed by the tethered PU chains onto reactive clay, which added additional constraints to chain motion on top of hydrogen bonds in the hard domains. The morphology of the obtained nanocomposites was studied by wide angle X-ray diffraction (WAXD). Figure4 shows typical WAXD patterns obtained for clay and SMPU/clay nanocomposites. The WAXD of pure clay (CloisiteⓇ 30B) showed a peak at 2θ=4.9∘ indicating an interlayer spacing d 001 of 1.8nm. No peak could be observed for a composite containing 1wt% clay (PU-01) content, indicating a fully exfoliated state, which was also confirmed by TEM (see Fig.5a). In Fig.5a, clay particles were dispersed on the scale of single clay layers, which are marked by arrows. Increased concentrations of clay resulted in small shoulders in the WAXD (Fig.4). The TEM image in Fig.5b, corresponding to 3wt% clay content, shows both individual clay layers and sparse clay stacks, the latter with expanded d-spacing compared to d 001=1.8nm.
The WAXD pattern for the composite containing 5wt% clay content (Fig.4) showed a very small peak at 2θ=6.5∘, and a shoulder at 2θ=3.4∘, which was more obvious than in the case of the composite with 3wt% clay content. In this case, the TEM image (Fig.5c) revealed a combination of individual clay layers and some agglomerates of multiple clay sheets. Here, the clay layers were not completely separated from the clay tactoids, although the size of tactoids was significantly reduced, e.g., 2–3 clay sheets, compared to 200–500 clay sheets in original clay particles. For that reason it was apparent that the layered silicates were well exfoliated and well dispersed in the polymer matrix. The clay platelets with their large surface area per unit volume could potentially interact with both hard and switching domains and may interfere with the crystallinity of both segments.
The influence of filler on the thermally-induced shape-memory properties was quantified by the shape fixity rate R f(N) and the shape recovery rate R r(N). R f(N) is the ability of the switching segment to fix the mechanical deformation during the programming process. R f(N) was calculated for cycle N from the ratio of elongation in the tension-free state after cooling εu(N) to the value of extension εm during programming:
The value of R r(N) can be calculated from εu(N) and the extension at the tension-free states εp(N−1) and εp(N) while expanding the sample in two subsequent cycles N−1 and N based on the following equation [41, 85]:
The temperature dependence of R r for nanocomposites was shown for different clay contents (Fig.6a). The melting temperature range of the switching domains is also indicated in this figure. The shape recovery started in all cases at 30∘C, (10∘C lower than the T m) and approximately 70% of the deformation was recovered as the temperature reached 50∘C. This behavior was attributed to the fact that, once the crystals started melting, the amorphous chains relaxed and shape recovery began. The continued shape recovery above 50∘C was possibly related to the rearrangement of amorphous molecular chains since they required enough time to relax to the original shape. The finally reached R r was almost 100% for pure SMPU and around 90% for nanocomposites with 1 and 3wt% clay content. For a clay content of 5wt% an R r of 85% was obtained. The presence of nanoclay in SMPU decreased R r was observed for other fillers as well [86, 87].
Figure6b shows a typical behavior for the recovery stress as a function of temperature under constant strain condition. As the switching domain crystals melted, the sample attempted to shrink and applied a compressive stress on the clamps of the DMA setup. In this case the machine applied the same amount of stress (recovery stress) to balance the shrinkage stress and to keep the sample length constant. At the same time, the switching segments became more flexible as the switching domain crystals were melted. The recovery stress declined after reaching a maximum. The addition of clay particles influenced the recovery stress level. The magnitude of recovery stress increased by 25% in the presence of 1wt% clay. However, the peak recovery stresses of 3 and 5wt% nanocomposites were lower than those of 1wt%, although they were slightly larger than that of pure SMPU. It was surprising to find that the value of recovery stress for composite of 5wt% clay content was lower than those of 1 and 3wt% clay content. This was probably caused by stress relaxation during the cooling step [77]. Stress relaxation could also occur during heating at a constant strain, especially during shape recovery under constant strain. In addition, it was found that the clay was completely exfoliated in the case of 1wt% clay content and the relaxation rate was the lowest, resulting in the highest recovery stress. A similar improvement in mechanical properties and SME of SMPU clay nanocomposites was reported by Cho and Lee [88].
Fibers of SMPU/clay (CloisiteⓇ 30B) nanocomposite were obtained by melt spinning and their shape-memory behavior was investigated [89]. This composite was initially prepared with 20wt% clay content through a solvent cast process with DMF as solvent. The dry product was used as master batch to prepare nanocomposite fibers with 0.25, 0.5, 0.75, and 1.0wt% clay content. The WAXD patterns confirmed that the clay is completely exfoliated in the SMPU matrix. No diffraction peak of clay was detected for all nanocomposites having different clay contents. Both, melting and crystallisation temperatures, associated to PCL segments were shifted to higher temperatures with increasing clay content, indicating that clay acted as nucleating agent, increasing the degree of crystallinity of PCL.
The clay content also significantly influenced the shape fixity rate of fibers; however the shape recovery rate decreased significantly with increasing clay content. This behavior was attributed to the increase in crystallinity of PCL. It might also be related to the fact that clay was interfering with the switching segments resulting in poor shape recovery rates. The stress recovery was increased with increasing clay content up to 0.5wt% and then decreased again for high clay contents for different εm (see Fig.7). This effect was attributed to the poor dispersion of the high clay content in the polymer matrix.
2.2 Poly(Ethyl Methacrylate) (PEMA) as Polymer Matrix
A macroazoinitiator (MAI) containing a poly(ethylene glycol) (PEG) segment was intercalated in the gallery of sodium montmorillonite (Na-MMT). This intercalated MAI was used in the preparation of Na-MMT/poly(ethyl methacrylate) (PEMA) nanocomposites via in situ radical polymerization of ethyl methacrylate [90]. This processing technique allowed a partial compatibility between Na-MMT and the polymer matrix based on the compatibility of PEG and PEMA. The WAXD pattern and the morphology observed with TEM revealed that the clay intercalated with PEG segments was heterogeneously dispersed in the PEMA matrix. Thus the intercalated clay effectively enhanced the mechanical properties of PEMA. The amount of MAI intercalated at the gallery of Na-MMT, determined by thermogravimetry, was 0.22g-MAI/g-Na-MMT. The clay/PEG building block linked to PEMA should act as netpoint determining the permanent shape and enhancing the mechanical properties of the material [90]. Both T g and modulus of the polymer matrix were increased by adding nanofillers because chain mobility was reduced to a great extent [91, 92, 93, 94].
Figure8 shows DMTA measurements of PEMA/clay nanocomposites for different clay contents. The tensile storage modulus E ′ increased and the peak maximum of tanδ (glass relaxation process) shifted to higher temperatures with increasing clay content. This behavior was attributed to the fact that the intercalated clay with PEG segments could significantly reduce the chain mobility of PEMA segments and reinforced effectively the PEMA matrix. The rubbery plateau shifted to higher temperatures with increasing clay content. The intercalated clay with PEG segment created additional physical crosslinks, which provided mechanical stability of PEMA chains during deformation process.
Cyclic, thermomechanical experiments were performed to examine the SME of PEMA/clay nanocomposites. In these tests, the sample was elongated at 80∘C to εm=50% at a deformation rate of 10mm min−1. While maintaining the strain at εm, the samples were quenched to 25∘C for 10min, where the switching domains became glassy. Upon removing the stress at 25∘C, a small strain recovery to εu occurred. The sample was subsequently heated to 80∘C again and remained at that temperature for the next 10min allowing recovery of strain. The temperature of 80∘C was selected for the elongation and recovery in the thermomechanical cycle, which was higher than the T gs measured by DSC, although it was lower than the peak temperatures of tanδ in Fig.8. The system had amorphous switching domains with T gs in the range of 55–72∘C based on the different molecular weight and clay content. At least seven thermomechanical cycles could be performed without failure under this experimental condition. After completing the cycle (N=1) a residual strain εp remained where the next cycle (N=2) started [95, 96, 97, 98]. εp of the pure polymer increased with the number of cycles to a value of approximately 17% for N=6. The slipping and disentangling of the polymer chains caused a permanent deformation, which increased by repeated deformations. For nanocomposites with 1.2 and 3.3wt% clay content, εp was less than 3% even when N=6 cycles were performed. The nanoclay could significantly stabilize the permanent shape of the nanocomposite by formation of physical crosslinks. εp slightly increased for nanocomposites with high clay contents. With increasing hard segment content, the probability to deform both the hard domains and the switching domains by external force increased [99, 100, 101]. For pure PEMA, εu was 49% and 48% for N=1 and 6, respectively. For nanocomposite with 9.3wt% clay content, εu was 43% and 39% for N=1, and 6, respectively. These results suggested that the temporary shape was efficiently fixed by cooling to 25∘C with a minor shape recovery. This minor shape recovery after cooling was found to be increased with increasing clay content.
2.3 Low Density Polyethylene (LDPE) as Polymer Matrix
Crosslinked low density polyethylene (LDPE) could be deformed easily above its melting point (switching segments) and this deformation could be fixed by cooling. The PE crystallites formed the switching domains and acted as temporary netpoints. On heating, the deformed sample restored its original, permanent shape. Crosslinked LDPE is used in many heat-shrinkable products (e.g., foils and tubes), which are applied in packaging, and electrical insulators. The addition of clay into polyolefin matrix enhanced the mechanical properties, crystallinity, and dynamic mechanical properties [102, 103, 104]. Therefore, the SME could be improved as well. The effect of the incorporation of modified clay (CloisiteⓇ 15A) into crosslinked LDPE on mechanical and shape-memory properties was investigated by Rezanejad and Kokabi [105]. LDPE and 3, 5, 8 and 10wt% CloisiteⓇ 15A were melt blended at 170∘C in a first processing step. The obtained nanocomposites were further mixed with 0.5wt% dicumyl peroxide (DCP) as a thermally-triggered initiator and 0.1wt% phenolic primary antioxidant (Irganox 1010) to ensure the thermal stability of the material at 130∘C in an extruder at 50rpm for 8min.
The permanent shape of the composite was obtained by curing under a hot press at 180∘C and 80ton pressure. The WAXD pattern showed a strong peak at the position of 2θ=2.75∘ for clay, which corresponded to a d-spacing of 3.15nm. After melt blending with LDPE the position of the (001) peak shifted to lower angles (higher d 001). The composite viscosity increased with increasing clay content, and thus the exerted shear stress to the layers should also increase and clay layers separated easier up to 5wt% clay content, which has 3.7nm d-spacing. For higher clay content, the layered silicates were stacked and separation of these layers became more difficult. The switching temperature shifted to higher values and the shape recovery rate decreased with increasing clay content. The increase in the switching temperature is attributed to the increase in T m associated with the switching domains by adding clay. The layered silicates significantly hindered the motion of the polymer chains, thus preventing them fully recovering to the permanent shape.
The generation of stress under constant strain condition was explored as well. Increasing clay content increased recovery stress as clearly seen in Fig.9. This behavior was attributed to the reinforcement effect caused by the clay. Another interesting feature in this figure is that the slope of the initial part of stress vs temperature increased with increasing clay content. This finding was related to the high number of crosslinks per unit volume for high clay content. In addition, the generated stress decreased at high temperature after reaching its maximum value for nanocomposites with clay content≤5wt% caused by slippage of polymeric chains from each other and their disentanglement [106]. This decrease was not detected with 8wt% clay (see Fig.9). A schematic presentation of molecules in neat polymer and nanocomposite is given in Fig.10 (a: neat polymer, b: nanocomposite). Crosslinks could not be formed in the surface of the clays, and thus the presence of clay layers caused the crosslinks to distribute more efficiently in the matrix [105]. In a well-distributed system, the likelihood of entanglements between two crosslinks was higher than neat polymer; thus, by stretching, disentanglement was less probable and only configurationally changes would occur in the nanocomposites.
3 Enhancing Mechanical Properties of SMPs by Incorporation of Polyhedral Oligomeric Silsesquioxanes (POSS)
Polyhedral oligomeric silsesquioxanes (POSS) belong to a special class of functional nanoscale fillers consisting of an eight-corner, –(SiO1.5) n-based cage bearing one or more functional groups [107, 108, 109, 110, 111] (see Fig.11a). POSS-based chemicals bridge the gap between fillers and monomers in offering shrinkage control and reinforcement of polymeric materials. When appropriately functionalized they also bridge the gap between polymers and plasticizers without plasticizer migration. The size of the pendant POSS cage (1–3nm) is comparable with the dimensions of the linear polymer, enabling POSS to control the motions of the chains at the molecular level (see Fig.11b). Therefore an enhancement in the physical properties was being expected, while the processability and mechanical properties of the polymer matrix are retained. Other property enhancements such as gas permeability might also be realized. A higher thermal stability, a better environmental durability under special conditions (i.e., exposure to atomic oxygen and fire resistance), and improvement in mechanical properties (reinforcement) were expected for POSS modified polymer hybrids [78, 79, 80, 112, 113, 114]. POSS allowed the creation of materials exhibiting hybrid properties. Conceptually, POSS might be thought of as an organic-inorganic hybrid (Fig.11c) [107]. POSS could be incorporated into the polymer matrix by two different methods, namely mechanical incorporation as nanofiller particles by melt blending and chemical linkage through introduction of one or more functional groups to the corners of the POSS chemical structure. These functional groups could react chemically with the polymer matrix to produce polymer-graft-POSS composites. For mechanical blending the eight corners of the cage structure of POSS had all non-reactive R groups. When POSS was linked chemically to the polymer matrix, a reinforcement of the system on the molecular level was obtained [115]. POSS is inert and rigid whereas the surrounding organic groups provide compatibility with the matrix and processability.
3.1 SMPC from PCL Networks and POSS Nanoparticles
Biodegradable shape-memory polymer networks with single POSS moieties located in the center of the network chains would promote POSS crystallization even within a constraining network structure. Successful synthesis of POSS initiated poly(ε-caprolactone) (PCL) telechelic diols, utilizing a POSS diol as initiator, was reported by Lee et al. [116]. The POSS–PCL diols were terminated with acrylate groups and photocured in the presence of a tetrathiol crosslinker. Scheme1 shows the chemical reaction for the synthesis of POSS–PCL network.
Typical DSC thermograms for POSS–PCL networks having different POSS contents are shown in Fig.12. All polymer networks showed a T g at approximately−50∘C regardless of POSS content. This T g is higher than that of the original telechelics and of pure PCL homopolymer (−60∘C), possibly due to constraints from the crosslinking. For sample a modest melting endotherm was observed at a temperature above 60∘C for composites with 42 and 34wt% POSS. This melting endotherm was related to the crystalline phase of POSS. Networks with POSS content<35wt% exhibited a melting peak around 40∘C associated to PCL crystals.
The SME of the network nanocomposite with 42wt% POSS content was investigated. Here POSS crystallites were used to fix the temporary shape. The cyclic thermomechanical test started with heating the composite to 110∘C. At this temperature, the sample was deformed by applying a load of 0.23N and then cooled under this load to 30∘C (POSS crystallization occurred according to DSC measurements at about 80∘C, see Fig.12). After releasing the load, the sample was heated to 110∘C again. A good shape fixity rate was observed since the sample did not retract upon removal of the load at 30∘C. The shape recovery rate was improved from 97.7% (first cycle) to 99% in the third cycle.
The same cyclic experiment was also performed without the unloading step. Typical results obtained from thermomechanical experiments for a POSS–PCL composite with 42wt% POSS content are shown in Fig.13. Application of the external load led to an elongation of about 25% at 110∘C. The relatively small reversible change in strain during cooling to 30∘C and reheating to 110∘C was caused by crystallization/melting of POSS moieties and the thermal expansion. The initial increase was caused by the negative thermal expansion coefficient of an elongated amorphous network. The increase in strain at around 55∘C was correlating with the crystallization of POSS moieties. Further cooling occurred under a positive thermal expansion coefficient as the switching domains solidified. Upon heating, the decrease in strain beginning at 80∘C was caused by the melting of POSS domains. The dramatic decrease in strain from 80∘C to 110∘C was again caused by thermal shrinking.
3.2 SMPC from PU and POSS Nanoparticles
SMPU/POSS nanocomposites were prepared by in situ polymerization of POSS-diol with polylactide diol and lysine derived diisocyanate using a two-step synthesis route. The initial step, synthesis of the lactide-based switching segment, allowed for manipulation of T g through controlled PEG inclusion. Synthesis of the subsequent PU using POSS as the hard segment contributing mechanical stability by forming physical crosslinks in the form of crystalline POSS domains. In this system, the biodegradable poly(rac-lactide) segment formed the switching domains [117]. The chemical formula of the obtained thermoplastic SMPU/POSS nanocomposite is shown in Fig.14.
The nanocomposites were elastic at temperatures above the T g of PDLLA-domains as determined by DMTA measurements. The SMPU/POSS nanocomposites had storage tensile moduli greater than 2GPa at low temperatures (see Fig.15a). The modulus dropped to a value of about 10MPa upon heating above T g of the switching domains. This rubbery plateau bridged the temperature range from PDLLA domains associated T g and POSS related T m. Pure SMPU without POSS was in melt at T>T g (i.e., no elastic plateau was observed at temperatures higher than T g). The elastic recovery was enabled by the physical crosslinks formed by the crystallites of POSS moieties as revealed with WAXD [117]. Increasing the POSS content in the SMPU increased the crystallinity and also the rigidity of the material.
Cyclic, thermomechanical tensile tests were performed for the nanocomposites with POSS/polyol ratio=2.63 (see Fig.15b). The sample was firstly heated to 80∘C (T>T g) and deformed (1) by ramping to a load of 0.3N. The sample was cooled under this load (2) to 10∘C, to fix the temporary, elongated shape. After unloading (3) the sample was heated (4) to 80∘C to recover the permanent shape. The first cycle showed about 5% creep occurring between the elongation and fixing step over a period of 5min. A shape fixity rate of higher than 99% was observed for the first and each subsequent cycle. For the first cycle shape recovery rate reached only 71% of the original strain. Both creep and recovery improved on each subsequent cycle: creep decreased to 3% and 2.5% and shape recovery rate increased to 89% and 93% for cycles two and three.
4 Magnetic SMPCs Obtained by Incorporation of Magnetic Particles
Magnetic particles, such as iron, nickel, cobalt, and some of their alloys, are called ferromagnetic or ferrimagnetic materials. These materials exhibit a strong attraction to magnetic fields and are able to retain their magnetic properties after the external field has been removed. The domains of ferromagnetic material are nearly randomly organized in unmagnitized state and the net magnetic field for the part as a whole is zero. When a magnetizing force is applied, the domains become aligned to produce a strong magnetic field within the part. The ferromagnetic particles can generate heat in an alternating magnetic field via hysteresis loss, eddy current, and/or additional mechanisms based on the different kinds of magnetic particles and their sizes [118]. The particles interact with the external magnetic field via a Zeeman term, and with SMP matrix via elastic deformation, as well as with each other via the demagnetization field [119]. When the programmed SMP composite with a certain ferromagnetic particle content was exposed to an alternating magnetic field, the temperature of the material increased. If the temperature exceeded the T sw of SMP matrix, the original permanent shape could be recovered. This indirect, non-contact heating method could be used to trigger the SME if the SMP could not be actuated by direct heating methods through increasing environmental temperature.
The amount of heat generated in an alternating magnetic field was directly related to the size of the magnetic particles. It was possible to evaluate the concentration of magnetic nanoparticle and its core size using vibrating sample magnetometry (VSM) measurements at room temperature [120, 121]. Particles in the nanoscale could generate enough amount of heat required for induced SME only at high magnetization frequency [122]. Here the applied frequency was an appropriate parameter controlling the amount of heat that will be generated. For particle diameters in the micrometer range, hystersis loss and eddy current loss occurred when the magnetic field was applied. The substantial contribution of eddy current loss and the microsize domain surfaces generated more heat in the low frequency range. Therefore, large particle sizes or aggregation of particles are not recommended for inductively actuating the SME particularly for medical applications. Here overheating could cause severe damage to the surrounding tissues.
Certain ferromagnetic particles enabled an innate thermoregulation, which was caused by the Curie temperature (T C). This effect occurred when the magnetic particles with an appropriate diameter generate heat in an alternating magnetic field by only hysteresis loss mechanism instead of an eddy current mechanism [123, 124]. Based on this phenomenon the ferromagnetic particles were able to heat a material up to T C (i.e., the magnetic material becomes paramagnetic and loses its ability to generate heat via a hysteresis loss mechanism). Magnetic materials with suitable T C could significantly reduce the danger of overheating in biomedical applications as the ferromagnetic material worked as a thermostat. Some magnetic materials composed of small ferromagnetic clusters (e.g., crystallites of 1–10nm diameter) might exhibit a behavior similar to paramagnetic at temperatures below T C. These materials were called superparamagnetic. In this case the energy required to change the direction of the magnetic moment of a particle is comparable to the ambient thermal energy and the rate at which the particles will randomly reverse direction becomes significant. Unlike ferromagnetic materials, superparamagnetic materials did not retain any significant amount of magnetization in the absence of an externally applied magnetic field and thus did not form aggregates. When a magnetic field was applied to superparamagnetic nanospheres, the external magnetic field oriented all the crystals of magnetic particles in its path in the same direction. This alignment locally increased the amplitude (strength) of the external magnetic field. Once the field was removed, Brownian motion mixing magnetic domains demagnetized the material.
The inductive heating equipment normally consisted of a high-frequency generator, a water cooled coil, and a temperature detector such as IR pyrometer for non-contact measuring of the sample temperature as clearly seen in Fig.16. In the next two sections some selected SMPCs with two different types of ferromagnetic particles, namely nickel zinc ferrite and iron(II,III) oxide or magnetite will be discussed. Adjusting the applied magnetic field strength, frequency, and T C one could control the amount of heat generated for different potential applications.
4.1 Nickel Zinc Ferrite Particles as Magnetic Fillers
T C of nickel zinc ferrite magnetic particles (Ni1−x Zn x Fe2O4)(III) could be varied in a wide range. T C decreased from approximately 370∘C to 150∘C by increasing zinc substitution (x) from 0.4 to 0.75 [123, 125]. In addition these particles had a relatively high electrical resistivity and a very high environmental stability. These particles were reported to have an average size of approximately 50μm and a spherical shape [123]. As the contribution of eddy current heat was neglectable, the SMPCs with ferromagnetic particles will mainly be heated inductively via hysteresis loss mechanism. A prototype of therapeutic devices having a complex shape was prepared from nickel zinc ferrite particles (10wt%) embedded in an SMPU matrix [123].
The SMPU matrix was polyester-based thermoset (Mitsubishi Heavy Industries, Ltd). This SMPU is an amorphous material with two T gs at 55∘C and 115∘C for the switching and hard domains, respectively. The first prototype was a flower-shaped endovascular thrombectomy device, which was intended for stroke treatment. The second prototype device was an expandable SMPU foam device for potential application in aneurysm embolization. Both devices are presented in their collapsed and deployed forms in Fig.17. Actuation was achieved by inductively heating the samples above the T g of the switching domains (55∘C) by an alternating magnetic field of 12.2MHz and a magnetic field strength of approximately 400Am−1 (center of the inductive coil) in air at room temperature.
The influence of the magnetic particles on the molecular dynamics of the switching and hard domains of the SMPU was investigated using DMTA measurements. Figure18 shows typical DMTA data for the temperature dependence of dynamic storage and loss moduli for pure SMPU and the composite from SMPU and 10wt% nickel zinc ferrite particles. The storage shear modulus increased significantly by adding the magnetic particles. At the same time, the loss peak maximum of the switching domains is clearly shifted to higher temperatures. In this case the magnetic particles significantly improved the mechanical properties of the SMPU matrix.
The thermoregulation and hystersis loss for two nickel zinc ferrite types having different particle diameters were investigated using thermosensitive wax gauges [123]. Both particle types stopped heating when their T C was reached. The results obtained suggested that the heating mechanisms were hystersis loss and not eddy currents at the applied alternating field with a frequency of 12.2MHz and a magnetic field strength of 545Am−1. The eddy current power dissipation in conductive particles was directly proportional to the square of the frequency of the applied field and the particle diameter. The observed decrease in power dissipation as particle diameter decreased was a predictable result according to multidomain magnetic particle theory that shows coercivity to be proportional to the inverse of the diameter [123, 126]. Therefore, lower frequencies and smaller particles would reduce the presence of eddy currents, thereby maintaining and even enhancing the Curie thermoregulation mechanism.
4.2 Magnetite or Iron Oxide Particles as Magnetic Fillers
Magnetite is the most magnetic of all the naturally occurring minerals on earth. It is a type of iron oxide with natural magnetic properties. The chemical name of magnetite is ferrous-ferric oxide, and its chemical formula is Fe3O4. This refers to the different oxidation states of the iron in one structure. Arrangement of different oxidation states causes a transfer of electrons between the different irons in a structured path or vector. This electric vector generates the magnetic field. Magnetite is a ferromagnetic material with a high T C=768∘C and a high degree of spin polarization.
Magnetically-induced SME for composites from thermoplastic SMPU and aggregated micro-sized Fe3O4 particles for different particle contents up to 40vol.% was reported [127]. The SMPU (MHI Mitsubishi Heavy Industries) was synthesized from diphenylmethane-4,4-diisocyante, adipic acid, ethylene oxide, polypropylene oxide, 1,4-butanediol, and bisphenol A with T g of approximately 45∘C (T sw). The magnetic particles were mechanically mixed with the SMPU at 200∘C for 10min. Alternating magnetic field with a frequency of 50Hz and a magnetic field strength of 4.4kAm−1 were employed to induce the SME. Under this condition a programmed helically bended stripe consisting of SMPU and 20vol.% magnetic particles recovered to its original plane stripe shape within 20min [127]. The relatively slow recovery rate was attributed to the low magnetic field strength and magnetizing frequency.
Two different series of nanocomposites with magnetic particles were investigated using different thermoplastic SMPs as matrix components [85]. The first SMP was a thermoplastic aliphatic polyetherurethane, TFX. The T g of the switching domains was about 74∘C, while the T g of the hard domains was approximately 120∘C. The second SMP was a biodegradable multiblock copolymer (PDC) comprising of poly(p-dioxanone) (PPDO) hard segments, poly(ε-caprolactone) (PCL) switching segments, and 2,2(4),4-trimethylhexane-diisocyanate (TMDI) junction unit. The T trans associated to the switching domains was the T m of PCL segments. The T m was selected slightly higher than body temperature to avoid any damage to the surrounding tissues when the composite was heated to induce the SME in biomedical applications. The chemical structures of TFX and PDC are shown in Fig.19a, b, respectively. The magnetic nanoparticles consisted of an iron(III) oxide (Fe2O3) core in a silica (SiO2) matrix. The mean aggregate size (photon correlation spectroscopy of an aqueous dispersion) was 90nm, the mean domain size (X-ray diffraction) was 20–26nm, and the domain content (X-ray fluorescence analysis) was 50–60wt%. The magnetic nanoparticles were mixed in the melt with SMPs at approximately 170∘C. A homogeneous distribution of the magnetic nanoparticles in the polymeric matrix was obtained as clearly seen in Fig.20. This figure shows TEM micrographs of TFX with a 10wt% content of magnetic nanoparticles. The silica apparently improved the compatibility between the particles and the polymer matrix. The black particles in Fig.20b are the iron oxide domains embedded into silica matrix (dark gray wrapping).
As soon as the magnetic field was switched on, the temperature of the nanocomposites began to increase. Within a few minutes the temperature measured at the sample surface reached a constant level, which was characterized by the maximum achievable temperature T max. The value of T max increases strongly with increasing nanoparticle concentration and magnetic field strength. The magnetically actuated SME (for the nanocomposite TFX with 10wt% magnetic nanoparticles) is illustrated in Fig.21. The temporary shape (corkscrew like spiral) changed quickly to a permanent shape (plane stripe) within just 22s as shown in Fig.21.
The effect of magnetic nanoparticles on the cyclic, thermomechanical tensile tests of TFX nanocomposites is shown in Fig.22. Here TFX and a nanocomposite from TFX and 7.5wt% magnetic particles were compared. In these tests, the samples were elongated at a temperature T high, which was higher than T sw but lower than T trans of the hard domains. Strain was kept constant for a certain time interval to allow relaxation. The elongated samples were cooled to fix the temporary shape. This step was performed under stress-control, which resulted in an increase of strain as a consequence of entropy elasticity. The SME was initiated by reheating the composite to T high.
The influence of magnetic nanoparticles on the thermally and magnetically-induced shape-memory properties was quantified by R f and R r. For TFX composites, R f values were in the range between 100% and 118% [85].
TFX materials had R r values of≈80% in the first cycle (under stress-free condition) regardless of the concentration of the magnetic nanoparticles. The low elasticity of PDC compared to TFX nanocomposites at T high required an alternative programming procedure enabling the realization of comparable deformations εu to those applied for TFX. The composites were elongated at T d=25∘C, which was below T trans of the switching domains. Fixation of the temporary shape was caused by strain-induced crystallization and strain-oriented reorganization. R f-values between 50% and 60% were reached for PDC materials. Thus R f achieved by applying of cold-drawing for pure PDC polymer was≈40–50% lower compared to that programming by deformation at T high, which was caused by the loss of the rubber elastic part of the deformation after external stress was released. The recovery experiments were performed at T high=55∘C. R r-values were between 47% and 65% for samples programmed by cold-drawing. The comparison between thermally- and magnetically-induced recovery for both TFX and PDC nanocomposites are summarized in Table1 for different concentrations.
The sample geometry was an important parameter influencing the maximum achievable temperature (T max). The higher the value of surface/volume ratio (S∕V ), the lower is T max (see Fig.23a). Another very important factor was the surrounding environment, which influenced the heat loss and consequently T max. The SME of the magnetic composites could be triggered by an alternating magnetic field in air. As the thermal conductivity of air was relatively low, the amount of heat loss from the sample surface to the environment did not significantly decrease the temperature of the sample. For many potential applications including in vivo medical application, the medium was aqueous in nature having a higher thermal conductivity and heat capacity than air. Figure23b shows the heating curves of nanocomposites from TFX and 18wt% magnetic nanoparticles, which were investigated in air and in 10wt% gelatin-water solution as well as in distilled water. Obviously, only in air the sample temperature exceeded T sw (74∘C) of TFX while the increase in temperature in water and gelatin solution was very little [128].
5 Electrically Conductive SMPCs Obtained by Incorporation of Carbon Fillers or Ni
Polymer composites with carbon compounds as fillers such as carbon black (CB), carbon fibers, carbon nanotubes, or graphite were highly conductive materials. The carbon compounds significantly reduced the electric resistance and resulted in conductive SMPC, which could be triggered by means of Joule heat as an indirect actuation method. The just mentioned carbon compounds could conduct electricity in the plane of each covalently bonded sheet due to the delocalization of outer electrons to form π-cloud. This resulted in a lower bulk electrical conductivity for carbon than for most metals. Although metal powders were intrinsically more conductive than carbon compounds, metal had a tendency to oxidize forming an insulating layer on its surface. At atmospheric pressure carbon took the form of graphite, in which each atom was bonded trigonally to three others in a plane composed of fused hexagonal rings, just like those in aromatic hydrocarbons. The resulting network is two-dimensional, and the resulting flat sheets were stacked and loosely bonded through weak van der Waals forces (see Fig.24a). This gave graphite its softness and its cleaving properties (the sheets slip easily off one another). Carbon nanotubes (CNTs) could be considered as the result of folding graphite layers into carbon cylinders and might be composed of a single shell-single wall nanotubes (SWCNT) (Fig.24b) or of several shells-multi-wall nanotubes (MWCNTs) (Fig.24c) [132, 133]. Carbon nanofibers (CNFs) are a form of MWCNTs, which were synthesized by the catalytic decomposition of aliphatic hydrocarbons. They differed from CNTs in their diameter (dia.∼100nm), which was much larger than that of either SWCNT (dia.∼1nm) or MWCNT (dia.∼10nm). In recent years, due to advances in technology, CNFs became readily available in large quantities and at a price, which was significantly lower than that of CNTs. Carbon fiber (CF) was a material consisting of thin fibers about 5–10μm in diameter and composed mostly of carbon atoms. The carbon atoms were bonded together in microscopic crystals, which were more or less aligned parallel to the long axis of the fiber. The crystal alignment made the fiber very strong for its dimension.
5.1 Carbon Nanotubes as Conductive Fillers
Composites from SMPs and CNTs (SWCNTs or MWCNTs) received substantial attention during the last decade. Low concentrations of CNTs significantly enhanced the mechanical and thermal properties, imparted unique electrical properties as well as improved the shape-memory behavior of the polymer matrix [134, 135, 136, 137, 138, 139, 140, 141, 142, 143]. CNTs have an anisotropic nature enabling a percolative behavior at low volume fractions in the SMP matrix. The nanoscale size of CNTs and their exceptional electrical and mechanical properties provided an excellent opportunity to improve the thermal management and structural reinforcement of a polymer matrix. Conductive SMPCs containing CNTs provided actuation of SME by applying electric current rather than heating by raising the environmental temperature. It was found that SMPCs containing CNTs could be used as electroactive actuators for controlling microaerial vehicles [141]. The compatibility and interfacial adhesion as well as homogenous distribution of CNTs in a polymer matrix could be achieved by surface modification in nitric acid and sulfuric acid mixture [144, 145, 146, 147]. Generally, inclusion of CNTs in SMPs, such as thermoplastic elastomers could significantly modify the polymer behavior, impart new or enhance characteristic properties from the associative networking of the nanoelements.
5.1.1 SMPU and Carbon Nanotubes Composites
Impact of SWCNTs on the thermal, mechanical, and electrical behavior of thermoplastic SMPU has been studied [134]. The pure SMPU matrix (Morthane PS455–203 Huntsman Polyurethanes) was synthesized from∼9% aromatic diisocyanate,∼43% aliphatic and cycloaliphatic dicarboxylic acids, and∼48% aliphatic diols as confirmed by proton NMR in CDCl3. DSC measurements indicated that crystallites acting as switches melted around 48∘C. A weak endothermic peak at approximately 150∘C was attributed to melting of crystallites formed by hard segments. This SMPU displayed high elasticity (∼700%) and strain-induced crystallization. Large deformation in the rubbery state at room temperature could easily induce crystallization of the switching segments, creating an additional physical crosslink besides the ubiquitous hard segment crystallites, which restricted strain recovery on removal of the applied stress. Therefore the SME of this material could be created by cold drawing at room temperature. Melting of the crystallites, which were formed by strain-induced crystallization, released the constrained polymer chains, which entropically recoiled, driving the polymer back to a stress-free conformation. SWCNTs could be uniformly dispersed in SMPU up to 19.5vol.% (30wt%). Figure25 shows fracture surface SEM micrographs for SMPU/SWCTN nanocomposite with 5.9vol.% (10wt%) SWCNT. Clearly the nanotubes were dispersed homogenously in the SMPU matrix. It was also found that only 5vol.% SWCNT were necessary to increase the rubbery modulus by a factor of 5 at room temperature. Furthermore, the SWCNTs provided a conductivity of 1Scm−1 (percolation threshold∼0.9vol.%).
Shape- and stress-recovery characteristics of the SMPU/SWCNT nanocomposite (0.57vol.% or 1wt%) are qualitatively depicted in Fig.26. The SME of this nanocomposite was actuated thermally, optically, and electrically. Recovery of substantial deformation, in excess of 300%, was possible, as shown by the complete closing of a loose knot by heating the material to 55∘C (Fig.26a). Exposed to near-infrared irradiation (NIR), the nanocomposite deformed to 300% exerting∼19J to lift a 60 g weight more than 3cm. Pure SMPU was transparent to NIR light and remained unaffected (Fig.26c). Heating caused by the finite resistivity of the nanocomposite enabled current-induced actuation (Fig.26d). The resistivity of the nanocomposite was directly related to the concentration of CNT, its aspect ratio, and orientation distribution (which was dependent on deformation history [135]). The critical concentration of the CNT was also a function of the percolation threshold (ϕc∼0.5–1vol.%) [136]. The constrained recovery of nanocomposite by remote actuation using NIR, provided stresses, which were within 2–5% of that demonstrated for heat actuation indicating sufficient internal heat generation by non-radiative energy decay from the SWCNTs to completely melt the polymer crystallites, which had been formed by strain-induced crystallization. For neat SMPU only a minor recovery (0.2MPa) occurred under the same experimental conditions, probably arising from direct radiative heating from the lamp. The shape fixity rate and recovery constraint stress of a nanocomposite with 2.9vol.% (5wt%) SWCNT increased from approximately 0.56 to 0.70 and from 0.6 to 1.4MPa compared with pure SMPU, respectively.
The enhanced recovery characteristics were thought to be associated with the synergism between the anisotropic SWCNTs and the crystallizable switching segments of the polymer. The extent of SWCNT orientation in the direction of deformation (tube axis parallel with applied stress) increased with SWCNT concentration. In conjunction, the polymer crystallite fraction increased with SWCNT concentration. The addition of SWCNTs also slightly increased initial polymer crystallinity by serving as heterogeneous nucleation sites [137]. Simultaneous X-ray diffraction–deformation experiments revealed a complex interplay between crystallization of switching segments, their alignment, and the orientation of SWCNTs during elongation [134]. On removal of the stress, the SWCNT alignment relaxed and the fraction crystallinity remained constant. Heating the constrained composite resulted in melting of the crystallites without substantial change in SWCNT alignment.
MWCNTs with 10–20nm diameter and 20μm length were incorporated into a SMPU [40wt% hard segment content (MDI and BD) and 60wt% PCL switching segment] to actuate the SME by applying an electric field [138]. The MWCNTs were surface-modified in mixed solvents of nitric and sulfuric acids (3:1 molar ratio) at 140∘C for 10min, followed by high-energy sonication in ethanol for 2h. SMPC films were obtained by casting a 10wt% SMPU/MWCNT solution in mixed solvents of tetrahydrofuran (THF) and dimethylformamide (DMF). The surface modification produced a negative impact on the electrical conductivity of the nanotube. This was attributed to the fact, that the acid treatment of the surface led to an increase in the number of defects in the lattice structure of carbon–carbon bonds. For this reason, the degree of surface modification was carefully controlled to avoid a dramatic decrease in the conductivity. The electrical resistivity was approximately 580Ω for a composite with 5wt% modified-MWCNT. This concentration was sufficient to heat the sample above 35∘C at 60V in 8s [139, 140]. It was, however, impossible to heat the sample above the T sw of SMPU/MWCNT nanocomposite with a voltage lower than 40V. Electric-field-triggered shape recovery was recorded using a video camera. The sample was initially a rectangular strip, was deformed into a helix shape at 60∘C and cooled to room temperature to fix the temporary shape. The original permanent shape of the sample was recovered almost completely within 10s when an electric field of more than 40V was applied [138]. The rate of shape recovery was strongly dependent on the magnitude of the applied voltage and the MWCNT content in the samples. The SMPC with untreated MWCNT resulted in higher temperatures than that of the composite with modified MWCNT under identical voltage.
A homogenous distribution of MWCNT in SMPU could be reached by mixing the MWCNT with the prepolymer mixture prior to the addition of chain extender (BD). This in situ polymerization process normally provided a good opportunity of interaction between the polymer chains and nanofillers. Highly conductive SMPU actuators prepared by in situ polymerization of SMPU in the presence of surface modified MWCNT were reported [141]. Different concentrations of MWCNTs up to 10wt% were mixed with the prepolymer solutions. The SMPU/MWCNTs nanocomposites were also synthesized with different hard segment contents (25–40wt%) to tailor the overall physical and mechanical properties as well as shape-memory properties. The composite was elongated by 100% at 32∘C and the temporary shape fixed by cooling to 10∘C. The permanent shape was recovered by application of different voltages. This nanocomposite actuator was explored for controlling the surface of a microaerial vehicle. When the electric power was supplied the actuator shrunk increasingly and accordingly the control surface was deflected gradually. The nanocomposite could produce 30∘ actuation angle or the control surface could rotate from−15∘ to+15∘ [141].
Fibers from SMPU/MWCNTs nanocomposites were prepared using a melt spinning technique to improve the mechanical, thermal, morphological, and viscoelastic properties as well as the shape-memory properties of the polymer matrix [142]. The efficiency to produce fibers with smooth surfaces decreased with increasing concentration of MWCNTs in the composite as shown clearly in Fig.27. The fiber surface of this nanocomposite became rough and coarse with increasing MWCNTs content particularly when the MWCNTs concentration increased from 3 to 7wt%. Homogenous distribution of MWCNTs with very high tendency to align parallel to the drawing direction could be achieved with concentrations less than 7wt%. The MWCNTs were incorporated into the SMPU matrix via in-situ polymerization by mixing MWCNTs with MDI in DMF before adding the PCL-diol [142].
The SME of this nanocomposite fiber was investigated by heating the fiber to 70∘C, and stretching it to 100% strain at a speed of 10mm min−1. The fiber was cooled to ambient temperature (22∘C) while keeping the applied stress. After unloading, the fiber was heated to 70∘C under stress-free condition to recover its original permanent shape. This cycle was repeated four times for each fiber. The R f(N) and R r(N) at the Nth cycle as well as the total recovery ratio R r ⋅tot(N) after Nth cycle were calculated and are listed in Table 2.
Generally the recovery strain and stress at 100% strain were mainly controlled by the stability of the hard domains. As the MWCNTs were mixed initially with MDI in DMF prior to adding PCL-diol, it was expected that the MWCNTs were strongly linked to the hard segments. This interaction could contribute greatly to increase the mechanical stability of the fibers particularly at T prog. Therefore, the maximum stress at 100% deformation increased with increasing MWCNTs content indicating that the SMPU/MWCNTs fibers were able to withstand higher stress at the same elongation providing a higher shape-recovery force to the fiber. At the same time, the MWCNTs having good interactions with the SMPU chains, particularly with the hard-segment regions, helped to store the internal stress during stretching and shape fixation. As a result, R r value increased. At high concentrations, the MWCNTs were not homogenously distributed in the polymer matrix and aggregated leading to incompatibility of the two components and weakening of the interfacial adhesion, consequently R r decreased [142].
5.1.2 SMPC from Polyvinyl Alcohol and Carbon Nanotubes
Fibers from polyvinyl alcohol (PVA) and up to 20wt% SWCNTs were developed to exhibit an exceptional high energy to rupture the fiber composite [143, 148, 149], which was a necessary condition to store a large amount of mechanical energy. The spinning process [150] consisted of injecting a dispersion of surfactant-stabilized SWCNTs in the co-flowing stream of a coagulating polymer solution. The favorable interactions between PVA and SWCNTs provided high toughness of the composite fibers. These composites could generate a stress upon shape recovery up to two orders of magnitude higher than that generated by conventional polymers. As clearly seen in Fig.28, large values of stress were needed to deform the fibers at a low deformation temperature (T d). The fibers became softer and could be more easily deformed at higher T d. This softness was associated with a lower supply of mechanical energy. This can be estimated from Fig.28, where the area under each curve corresponds to the energy supplied to the fibers at different T d, from 70 to 180∘C, and upon mechanical stretching.
These fibers exhibited also excellent shape fixity ratios. The fiber then shrank or recovered to its original permanent shape substantially by heating. Cyclic, thermomechanical experiments were performed to characterize the thermally-induced shape-memory properties of these nanocomposite fibers quantitatively [143]. The fibers were stretched at different T d and then cooled to room temperature under the external load to fix the temporary shape. When the fibers were reheated at fixed strain the fibers generated a strong stress with a maximum at a T σ−max. This peak was also observed for other SMPs under fixed strain condition [151, 152]. In fact the recovery peak was related to the T d of the SMPs. When the PVA/SWCNTs fibers were initially deformed above T g, the peak disappeared and the stress generated by shape recovery substantially decreased. This occurs because polymer chains could relax when deformed at temperatures well above T g, thus decreasing the potential for stored mechanical energy. In this case, the peak is preserved well above the T g of the neat PVA and, more strikingly, T σ−max and T d are roughly equal. It means that the fibers could recover to their original shapes only at the temperature at which they were deformed (T d). The peak of stress generated could be observed up to 180∘C, which was∼100∘C above the T g of the neat PVA. In addition, it is observed that the maximal stress generated by the fiber was close to 150MPa. This value of stress recovery was closer to the stress generated by shape-memory metallic alloys (0.5±0.25GPa) [153, 154, 155, 156]. The broad T g of this composite was a drawback for the shape recovery behavior, where a large increase in the temperature (>100K) was required to obtain medium shape recovery values around 50%.
5.2 Carbon Black (CB), Short Carbon Fiber (SCF), or Ni as Conductive Fillers
SMPCs containing homogenously distributed carbon fillers such as CB could induce electrical conductivity to the polymer matrix. The conductive fillers (e.g., CB) were usually randomly dispersed, so that the formation of the conductive channels was not well under control but depended more or less on the dispersion and geometrical parameters of filler as well as on its concentration. Incorporation of CB with small amount of SCF or metal powder, such as aligned Ni could substantially improve electrical conductivity. The SCF or aligned Ni powder created conductive paths/chains, which were essential for good electrical conductivity, enabling the transport of electrons along the direction of the electrical field. In the next two sections the effect of CB and SCF or aligned Ni powder on the shape-memory behavior actuation by applying electric current will be presented. These ternary composites had excellent electrical properties, as fibrous fillers or aligned Ni powder cooperated with CB enhancing the formation of conductive networks.
5.2.1 Thermosetting Styrene Resin as Polymer Matrix
The electrical and thermomechanical properties of styrene-based SMPCs containing CB nanoparticle and SCF of 0.5–3mm length and 7μm diameter were investigated [157]. The SMP material used in this work was a thermosetting styrene-based resin with 75∘C curing temperature (Cornestone Research Group Inc. OH). These composites were prepared by mixing styrene-based with a crosslink agent followed by adding certain amounts of CB and SCF with continuous stirring. Solidification occurred when the composites were heated at 75∘C for 24h. Figure29 shows typical optical microscope and SEM photographs of the composite with 2wt% CB and 0.1wt% SCF. The two fillers could produce different contributions to improve the electrical conductivity of the composites. SCFs were added to provide the conductive pathways and promote relatively long distance charge transfers, leading to the formation of continuous conductive networks.
These SCFs played a more prominent role in comparison with the CB aggregate in the polymer matrix. It seemed that the CB particles and their aggregates acted as the nodes among the fibers by forming local conductive pathways, which improved the orientation of the SCF. It is also clear that many CB aggregates were adsorbed on the surface of SCFs (see Fig.29b). These enlarged the area of conductive fillers and homogeneously improved the electrical and thermal conductivity of the materials. As a result, the co-supporting two-filler system was expected to enhance conductivity and therefore synergistic effects as well.
The impact of the two fillers on the volume electrical resistivity of the SMPCs at room temperature as a function of fillers concentration is presented in Fig.30. This figure contains two curves, one for SMP/CB and another for SMP/CB/SCF nanocomposites. For the second curve the concentration of CB was kept constant at 5wt% and only the SCF content was varied. The two curves revealed that the volume resistivity of the SMP/CB/SCF systems was significantly lower than that of SMP/CB with the same CB content. This could be attributed to the fact that the inherent fibrillar form of SCF had a higher aspect ratio and orientation to form a three-dimensional network in the composites, ensuring better electrical response than that of particulate fillers.
A typical example of electrically-induced SME for nanocomposites with 5 and 2wt% CB and SCF, respectively, is shown in Fig.31. Only 50s were required to reach a complete recovery from the temporary to the permanent shape by heating the sample up to 65∘C by application of an electric field. The shape transition at 25V (see the value of volume resistivity in Fig.30) was documented with a digital camera (Fig.31) [157]. The rate of shape recovery was dependent on the magnitude of the applied voltage and the electrical resistivity of the composites.
5.2.2 SMPU as Polymer Matrix
Replacing SCF by Ni was proposed to significantly reduce the electrical resistivity of SMPU nanocomposites [158, 159]. Here, the resistivity of SMPU/CB was compared with that of SMPU/CB/Ni nanocomposites. Addition of small amount of randomly distributed Ni microparticles (0.5vol.%) in the SMPU/CB composite led to a slight decrease in the electrical resistivity. If these Ni particles were aligned into chains (by applying a low magnetic field on the SMP/CB/Ni solution before curing), the drop of the electrical resistivity could be significant. The aligned Ni chains served as conductive channels to bridge CB aggregates. Consequently the electrical conductivity was significantly increased while keeping the other properties constant. Small amount of Ni particles was required to form chains inside polymers with randomly distributed conductive CB.
Thin films of SMPU/CB/Ni nanocomposites were prepared via solvent cast process using DMF as a common solvent. The concentration of CB was varied from 4 to 10vol.%, while the concentration of Ni particles was kept constant (0.5vol%). The Ni particles aligned one after another in one line, got organized instantly upon applying a weak magnetic field in solution before drying the samples. After solidification, these chains were fixed as clearly seen in Fig.32a. This procedure could produce Ni chains of approximately 150±50μm length and 5±2μm diameter which was comparable to the dimension of SCF.
The effect of CB concentration on the resistivity of SMPU/CB/Ni (aligned or chained) and SMP/CB/Ni (random, without alignment by magnetic field), as well as SMPU/CB is demonstrated in Fig.32b. Obviously, the random distribution of Ni (0.5vol.%) slightly reduced the resistivity of the composites. The aligned Ni in chain shape obtained in the magnetic field could greatly reduce the electrical resistivity by more than ten times. This remarkable reduction in the electrical resistivity was the result of the conductive chains, which served as conductive channels to bridge those small isolated CB aggregations. The reduction in resistivity was dramatic with low concentration of CB and then decreased in its magnitude with increasing the concentration of CB. For the same sample, the resistivity measured 1month later was about the same as before as clearly seen in Fig.32b.
To investigate the SME induced by Joule heating, three samples of SMPU/CB, SMPU/CB/Ni (random), and SMPU/CB/Ni (chained) with 10vol.% CB and 0.5wt% Ni were programmed in a bent shape, 150∘ at 80∘C, and then cooled to room temperature (22∘C) [158]. An infrared video camera was applied to monitor the temperature distribution and shape recovery simultaneously. Four snap shots for each sample are presented in Fig.33. For SMPU/CB/Ni (chained) composite (a), the temperature reached approximately 80∘C, which was much higher than the T g (40∘C) of switching domains of the SMPU, so that almost full recovery is observed within 120s. The temperature of SMPU/CB sample (c) was the lowest; about 45∘C only, slightly higher than the T g of the switching domains of SMPU, hence, the shape recovery was small. For sample (b) SMP/CB/Ni (random), the temperature reached around 65∘C and the shape recovery was not completed after 120s. The power consumption was about 1.2W for sample (a). Based on the above, it was apparent that the conductivity of SMPU matrix increased substantially by the combined effect of CB and chained Ni fillers.
5.3 Thermoexpanded Graphite as Conductive Filler
Epoxy resin composition based on diglycidyl phthalic, hydrophthalic acid ester and a block oligomer from the aliphatic epoxy resin and acid oligoether cured by iso-methyltetrahydrophthalic anhydride was used to generate composites with different concentrations of thermoexpanded graphite (TEG) [160]. This filler was selected according to its low bulk density and good compatibility with the polymer matrix. This enabled a considerable effect of volume increase not only at the expense of composite loosening at the polymer matrix–filler interface, but also caused by the original volume of TEG particles. The TEG [161] was the result of the interaction between cast graphite and sulphuric acid, in the presence of chromic acid, followed by washing out until the pH value of the aqueous extract was equal to 6.5–7 and drying at 105∘C to constant weight. The dried oxidized graphite was subjected to a thermal shock at 1,000∘C for 3min to remove water from the interplanar space. The bulk density and carbon content of TEG are 0.02gcm−3 and 95.2%, respectively. The samples were prepared by mixing the epoxy composition and TEG powder components in the required proportion. Then the polymer billets were cured in a cylindrical or pipe-like mould at 120∘C for 4h. The cured billet was deformed in the high elastic state and cooled, under load, to a temperature lower than T g to “freeze” the attained non-equilibrium state. So, the products were in the form of rods of about 20mm diameter. The deformation was realized by the uniaxial compression in a cylindrical mould. The shape was recovered by heating samples to T>T g and holding them at that temperature for 15min. The degrees of deformation (ε%) and recovery (R r %) were calculated using the formulas
where l i and l are the length of the sample before and after deformation, respectively. It was found that, at specific TEG concentrations, the composite was not only well deformed, but also capable of complete recovery of the shape. At low and high concentrations of TEG (3 and 15wt%, respectively) small R r was obtained. Optimal shape recovery rates were determined at 6–8wt% of TEG.
The morphologies of the composites (6wt% TEG) were studied at different stages namely, prior to deformation, after deformation (εm=50%), and after complete recovery. Three different microstructures at the different stages are presented in Fig.34. In the first stage before deformation (Fig.34a), the composite showed a typical two-phase structure with a very large TEG particle size dispersion (∼0.157mm) in a bright polymer matrix. After deformation, the TEG particles became more compact and the polymer matrix became oriented as clearly seen in Fig.34b. After shape recovery (Fig.34c), the average particle size considerably decreased and the particles changed their location in the polymer matrix. By quantitative structural analysis, a decrease in the amount of dark phase (TEG) was explained by the crushing and redistribution of TEG particles under deformation. It was also apparent that prior to deformation the polymer matrix was light in color. After deformation and subsequent relaxation it became more gray. This could only be the result of the introduction of fine graphite particles into the matrix, since the deformation of the polymer did usually not result in a color change.
6 Enhancing Biofunctionality of SMPs by Incorporation Hydroxylapatite (HA) or β-Tricalcium Phosphate (β-TCP) Particles
Hydroxylapatite or hydroxyapatite (HA) is a mineral occurring in nature with the formula Ca5(PO4)3(OH) and usually written as Ca10(PO4)6(OH)2 to denote that the crystal unit cell comprises two molecules. HA crystallizes in the hexagonal crystal system. Natural bone and teeth contain approximately 70% of HA inorganic mineral. HA is one of just a few materials classified as bioactive, meaning that it supports bone ingrowth and osseointegration when used in orthopaedic, dental, and maxillofacial applications [162, 163, 164]. Therefore, it is commonly used as a filler to replace bone or as a coating to promote bone ingrowth into prosthetic implants. The introduction of nano-HA greatly increased the mechanical properties of the SMPCs and improved the protein adsorption capacity [165, 166, 167]. It has been found that proteins in mineralized tissues acted as nature’s crystal engineers, where they played a key role in promoting or inhibiting the growth of minerals such as hydroxyapatite.
Pure β-tricalcium phosphate (β-TCP), Ca3(PO4)2, is not found in nature or in biologic systems and could not be obtained directly by precipitation or hydrolysis methods. Pure β-TCP could be obtained by heating calcium-deficient apatite of appropriate Ca/P molar ratio above 800∘C or by heating amorphous calcium phosphate, ACP. HA and β-TCP were different in both solubility and in in vivo biodegradation. Both HA and β-TCP showed osteoconductive properties. Porous β-TCP material provided a carrier matrix for bioactive agents and could form a moldable putty composition upon the addition of a binder. The poor mechanical properties of HA and β-TCP, such as low strength and fracture toughness, limited wide applications in hard tissue implantations [168]. Therefore composites from biocompatible SMPs and HA or β-TCP could improve the mechanical properties and introduce outstanding materials for a wide range of medical applications.
6.1 SMPC from Poly(rac-Lactide) and HA Nanoparticles
The idea of using polymers as binders for particulate bioceramics to prepare biocomposites with improved processing and retention characteristics and to overcome the problem of poor mechanical properties was reported. Poly(rac-lactide) (PDLLA) is a thermoplastic biodegradable, resorbable, and biocompatible aliphatic polyester, which in recent years received significant attention in the biomedical research field. It was also generally viewed as a “polymer of the future” because cost of production keeps on decreasing and it could be disposed easily without any significant environmental penalty. PDLLA (M w:144KDa) was synthesized by ring-opening polymerization of rac-dilactide monomer. PDLLA was dissolved in CHCl2 and mixed with an HA solution in ethanol [169]. When the solvent evaporated slowly from the mixture, some floccules formed gradually at the bottom of the beaker, which mainly were PDLLA/HA composites. These floccules were dried at 30∘C under vacuum and press-molded at 105∘C.
The miscibility of different concentrations of HA in PDLLA was investigated by SEM as shown in Fig.35. HA particles were uniformly distributed with approximately 1μm average diameter in the PDLLA matrix, whereby PDLLA and HA phases have a close contact. The obtained SEM morphologies indicate that the current preparation method was an efficient way to obtain PDLLA/HA composites with homogenous particle distribution. PDLLA/HA composites combining biodegradability, biocompatibility, and shape-memory capability [169] are attracting much attention because of their good osteoconductivity, osteoinductivity, and high mechanical strength. Medical products containing PDLLA/HA composites such as screws, plates, pins, and rods have already been commercialized.
For a microscopic demonstration of the SME, seven letters forming the word “science” were processed as permanent shapes of PDLLA/HA composite (7:3). All samples were heated to the programming temperature 68∘C which is higher than T g of PDLLA switching domains. After 3min, these letters were all pulled into a straight shape, and moved to a refrigerator for fixing this deformation. When the samples were heated at 70∘C again, they recovered their permanent shapes within 100s (see Fig.36). The amorphous PDLLA matrix acted as switching domains, while the crystalline HA particles stabilized the permanent shape [169]. The shape recovery ratios of all composites were higher than 95%, whereas PDLLA itself had a relatively low R r of 81% as the permanent shapes only determined by physical chain entanglements.
Further evaluation of the shape-memory behavior were exploring the effect of sample thickness and HA content on the shape recovery ratio using a bending test (see Fig.37). The composites were programmed by 180∘ folding and R r was calculated using (5) [169]:
The sample thickness had an almost negligible effect on R r. However R r increased from approximately 95.7% to 99.4% with increasing the ratio of PDLLA/HA up to 2:1, whereas R r decreased from 99.5% to 97.2% with the increase of composite ratio from 2.5:1 to 4:1. R r reached a maximum value of 99.6% for PDLLA/HA ratios between 2.0 and 2.5. The SME of PDLLA was improved greatly by adding HA due to the hydrogen bonding between the C=O bond in PDLLA and the surface P-OH groups of HA nanocrystals as confirmed by infrared spectroscopy (FTIR) and X-ray photoelectron spectroscopy (XPS) [169].
A model of hydrogen bonding between PDLLA and HA was designed as schematically shown in Fig.38 [170]. The hydrogen bonds acted as physical net points determining the permanent shape.
6.2 SMPC from Poly(rac-Lactide) and β-TCP Nanoparticles
The nanocomposites of PDLL/β-TCP are very promising and desirable biomaterials applied in tissue engineering [171], and have been used clinically in various forms [172, 173, 174]. These nanocomposites with different β-TCP are prepared in the same way as that of PDLL/HA nanocomposites described in the previous section. The average particle size of β-TCP used in this work was approximately 720nm with particle distribution of 200–1,500nm as determined by laser diffraction particle size analyzer. The effect of in vitro degradation on the shape-memory capability of PDLL/β-TCP nanocomposites was studied [174]. The hydrolytic degradation process was investigated using phosphate buffer saline solution (PBS, pH=7.4) at 37∘C. At predetermined time intervals, the specimens were removed from degradation medium and rinsed with distilled water to remove residual buffer salts, and dried in vacuum. The SME of pure PDLLA and PDLLA/β-TCP composites with different β-TCP contents were investigated before and after immersing in the buffer solution. The shape-recovery behavior was evaluated using the bending test. The process of shape recovery for the original as well as partially degraded samples is presented in Fig.39. The pure PDLLA and PDLLA/β-TCP composites displayed significant differences in their shape-memory behavior at different degradation time intervals.
Figure40 shows degradation time dependence of shape recovery ratio calculated according to equation (5). R r decreased with increasing in vitro degradation time in PBS. The R r for PDLLA/β-TCP composites was considerably higher than that of pure PDLLA at the same degradation time. For all composites, R r (at the 21st day of degradation time) slightly increased. The reason might be related to the cleavage of PDLLA chains and changes of the crystal phases of β-TCP particles during the degradation process.
Ca2P2O7, CaHPO4, and HA phases can be produced during in vitro degradation of PDLLA/ β-TCP composites based on the following reaction formula [174]:
The formation of new inorganic phases could cause a decrease of T g because of a plasticizer effect [175, 176]. The observed slight increase in the shape-recovery ratios of the composites after 21days degradation time in Fig.39 could be caused by existing of Ca2P2O7, CaHPO4, and HA particles that can produce more dynamical constrains to the polymer chains [177]. However, with the increase of degradation time the shape-memory properties of the composites became undesirable caused by cleavage of PDLLA chains and the dissolution degradation of the inorganic phases in PDLLA/β-TCP composites.
7 Conclusion
In this chapter a comprehensive overview of recent developments and progress in shape-memory composites has been presented. Incorporating small amounts of micro/nano-sized fillers into SMPs significantly improved the overall material properties with simultaneously enhanced or retained shape-memory capabilities and processability. The technology merges the properties of fillers with the versatility and precision of polymer chemistry and engineering. The micro/nano-sized fillers provided molecular level control over polymer dynamics, surface-bulk properties, and biological function. Changing the filler type, concentration, and surface modification to control the interfacial interaction between the polymer matrix and filler enabled tailoring strength, stiffness, and elasticity of the composites. The incorporation of fillers enables the implementation of new functions such as active movements induced by different external stimuli, e.g., alternating magnetic fields or electric fields. Magnetic particles of an appropriate size generate heat only by a hysteresis loss mechanism instead of an eddy current mechanism during the implementation of magnetically actuation of SMPC. In this way the danger of overheating in in vivo applications could be reduced significantly as the magnetic particles will work as a thermostat. Actuation of shape-memory properties using an alternating magnetic field or electric field is very crucial particularly when a direct thermal actuation method is not suitable. It is also possible to increase both stress and strain recovery of the SMP matrix when the fillers were carefully modified to increase the interaction with the polymer matrix. This interfacial interaction between fillers and polymer matrix largely influenced the substantial improvement, which was observed for different material properties. The outstanding properties of SMPCs as well as their light weight, low cost, easy processibility and very high recoverable strain make SMPCs good candidates for many potential applications. SMPCs are recognized as extraordinarily versatile materials and enabling technology for future space and interplanetary missions. For example, SMPCs have been explored to prepare trusses and torus-shape structures for lightweight satellite supports, antenna reflectors, and deployable wings for unmanned aerial vehicles. In this case, the SMPC materials allowed users to pack large, lightweight structures tightly into small volumes for later use on orbit or in the atmosphere. SMPCs prepared from biodegradable matrix and active fillers are potential candidates for biomedical applications such as orthopedic, dental, and maxillofacial applications. Shape stability, excellent recovery stress and strain, biocompatibility, and biodegradability (for in vivo biomedical application) or inductively actuation as well as ease of processing are vital factors for the acceptance of SMPC in these potential applications.
Abbreviations
- β-TCP:
-
β-Tricalcium phosphate
- εm :
-
Maximum deformation, parameter in cyclic, thermomechanical tests
- εu(N):
-
Free state deformation after cooling
- ϕ:
-
Volume fraction
- ϕc :
-
Critical volume fraction
- ρ:
-
Volume resistivity
- σmax :
-
Maximum stress
- BD:
-
1,4-Butanediol
- CB:
-
Carbon black
- CNT:
-
Carbon nanotube
- DMF:
-
Dimethylformamide
- DMTA:
-
Dynamic mechanical analysis at varied temperatures
- DSC:
-
Differential scanning calorimetry
- f :
-
Frequency
- Fe3O4 :
-
Iron (II,III) oxide
- FTIR:
-
Fourier transform infrared spectroscopy
- G ′ :
-
Shear storage modules determined in dynamic mechanical test
- G ′′ :
-
Shear loss modules determined in dynamic mechanical test
- H :
-
Magnetic field strength
- HA:
-
Hydroxyapatite
- LDPE:
-
Low-density polyethylene
- MAI:
-
Macroazo initiator
- MDI:
-
Methylene bis(p-cyclohexyl isocyanate)
- MWCNT:
-
Multi-walled carbon nanotube
- N :
-
Number of thermomechanical cycle
- Na-MMT:
-
Sodium montmorillonite
- Ni 1−x Zn x Fe2O4 :
-
Nickel zinc ferrite magnetic particles
- PBS:
-
Phosphate buffer saline solution
- PCL:
-
Poly(ε-caprolactone)
- PDLLA:
-
Poly(rac-lactide)
- PE:
-
Polyethylene
- PEG:
-
Poly(ethylene glycol)
- PEMA:
-
Poly(ethyl methacrylate)
- POSS:
-
Polyhedral oligomeric silsesquioxanes
- PPDO:
-
Poly(p-dioxanone)
- PS:
-
Polystyrene
- PTMG:
-
Poly(tetramethylene glycol)
- PVA:
-
Polyvinyl alcohol
- R f(N):
-
Shape fixity rate in cycle number N
- R r(N):
-
Shape recovery rate in cycle number N
- SCF:
-
Short carbon fiber
- SEM:
-
Scanning electron microscopy
- SiC:
-
Silicon carbide
- SiO2 :
-
Silicon oxide
- SMAs:
-
Shape-memory metallic alloys
- SME:
-
Shape-memory effect
- SMP:
-
Shape-memory polymer composites
- SMPs:
-
Shape-memory polymers
- SMPU:
-
Shape-memory polyurethane
- S∕V :
-
Surface to volume ratio
- SWCNT:
-
Single-walled carbon nanotube
- T C :
-
Curie temperature
- T d :
-
Deformation temperature, parameter in cyclic, thermomechanical tests
- TEM:
-
Transmission electron microscopy
- TFX:
-
Polyetherurethane prepared from MDI, BD, and PTMG
- T g :
-
Glass transition temperature
- THF:
-
Tetrahydrofuran
- T high :
-
High temperature, parameter in cyclic, thermomechanical tests
- T low :
-
Low temperature, parameter in cyclic, thermomechanical tests
- T m :
-
Melting temperature
- T progm :
-
Programming temperature, parameter in cyclic, thermomechanical tests
- T σ−max :
-
Temperature at maximum stress during constraint strain recovery
- T trans :
-
Thermal transition temperature
- VSM:
-
Vibrating sample magnetometry
- WAXD:
-
Wide angle X-ray diffraction
- XPS:
-
X-ray photoelectron spectroscopy
References
Song HM, Kim YJ, Park JH (2008) J Phys Chem C 112:5397
Satake A, Miyajima Y, Kobuke Y (2005) Chem Mater 17:716
Huang CM, Wei KH, Jeng US, Sheu HS (2008) Macromolecules 41:6876
Bekyarova E, Thostenson ET, Yu A, Kim H, Gao J, Tang J, Hahn HT, Chou TW, Itkis ME, Haddon RC (2007) Langmuir 23:3970
Xu P, Han X, Wang C, Zhou D, Lv Z, Wen A, Wang X, Zhang B (2008) J Phys Chem B 112:10443
Bellomo EG, Deming TJ (2006) J Am Chem Soc 128:2276
Sun H, Zhang H, Ju J, Zhang J, Qian G, Wang C, Yang B, Wang ZY (2008) Chem Mater 20:6764
Chen WF, Wu JS, Kuo PL (2008) Chem Mater 20:5756
Li YJ, Shimizu H (2007) Polymer 48:2203
Kharchenko SB, Douglas JF, Obrzut J, Grulke EA, Migler KB (2004) Nat Mater 3:564
Pegel S, Potschke P, Petzold G, Alig I, Dudkin SM, Lellinger D (2008) Polymer 49:974
Schaefer DW, Justice RS (2007) Macromolecules 40:8501
Fornes TD, Paul DR (2004) Macromolecules 37:7698
Mark JE (2006) Acc Chem Res 39:881
Lu G, Tang H, Qu Y, Li L, Yang X (2007) Macromolecules 40:6579
Liu N, Shi MM, Pan XW, Qiu WM, Zhu JH, He HP, Chen HZ, Wang M (2008) J Phys Chem C 112:15865
Chen Q, Zhao L, Li C, Shi G (2007) J Phys Chem C 111:18392
Gangopadhyay R, De A (2000) Chem Mater 12:608
Yang Y, Dudley MCGL, Lawrence RW (2005) Nano Lett 5:2131
Liu G, Wang H, Yang X (2009) Polymer 50:2578
Fang FF, Kim JK, Choi HJ (2009) Polymer 50:2290
Guo F, Zhang Q, Zhang B, Zhang H, Zhang L (2009) Polymer 50:1887
Ding H, Liu ZM, Wan M, Fu SY (2008) J Phys Chem B 112:9289
Kommareddi NS, Tata M, John VT, McPherson GL, Lee MF, Akkara CK, Kaplan DL (1996) Chem Mater 8:801
Xu P, Han X, Wang C, Zhao H, Wang J, Wang X, Zhang B (2008) J Phys Chem B 112:2557
Beecroft LL, Ober CK (1997) Chem Mater 9:1302
Dennany L, O’Reilly EJ, Innis PC, Wallace GG, Forster RJ (2009) J Phys Chem B 113:7443
Sainz R, Small WR, Young NA, Vallés C, Benito AM, Maser WK, Panhuis MIH (2006) Macromolecules 39:7324
Ravindranath R, Ajikumar, Hanafiah NBM, Knoll W, Valiyaveettil S (2006) Chem Mater 18:1213
Costantino U, Bugatti V, Gorrasi G, Montanari F, Nocchetti M, Tammaro L, Vittoria V (2009) Appl Mater Interfaces 1:668
Chiellini E, Cinelli P, Ilieva VL, Martera M (2008) Biomacromolecules 9:1007
Zhang D, Kandadai MA, Cech J, Roth S, Curran SA (2006) J Phys Chem B 110:12910
Urbina MC, Zinoveva S, Miller T, Sabliov CM, Monroe WT, Kumar CSSR (2008) J Phys Chem C 112:11102
Kelly A (1985) Compos Sci Technol 23:171
Kelley A, Zweben C (2000) Comprehensive composite materials, vols 1–6. Elsevier, Oxford
Thostenson ET, Li C, Chou TW (2005) Compos Sci Technol 65:491
Hussain F, Hojjati M, Okamoto M, Gorga RE (2006) J Compos Mater 40:1511
Sun H, Zhang H, Zhang J, Ning Y, Yao Y, Bao X, Wang C, Li M, Yang B (2008) J Phys Chem C 112:2317
Koval’chuk AA, Shevchenko VG, Shchegolikhin AN, Nedorezova PM, Klyamkina AN, Aladyshev AM (2008) Macromolecules 41:7536
Iyer S, Schiraldi DA (2007) Macromolecules 40:4942
Lendlein A, Kelch S (2002) Angew Chem 41:2034
Bellin I, Kelch S, Langer R, Lendlein A (2006) Proc Natl Acad Sci USA 103:18043
Jiang H, Kelch S, Lendlein A (2006) Adv Mater 18:1471
Lendlein A, Jiang H, Juenger O, Langer R (2005) Nature 434:879
Wilson TS, Small W, Benett WJ, Bearinger JP, Maitland DJ (2005) SPIE-Int Soc Opt Eng 60070R/1–60070R/8
Liu C, Qin H, Mather PT (2007) J Mater Chem 17:1453
Meng QH, Hu JL, Zhu Y, Lu L, Liu Y (2007) Smart Mater Struct 16:1192
Wache HM, Tartakowska DJ, Hentrich A, Wagner MH (2004) J Mater Sci Mater M 14:109
Wischke C, Neffe AT, Steuer S, Lendlein A (2009) J Control Release 138:243
Metcalfe A, Desfaits AC, Salazkin I, Yahiab LH, Sokolowskic WM, Raymonda J (2003) Biomaterials 24:491
Kim BK, Lee SW, Xu M (1996) Polymer 37:5781
Jin ZX, Pramoda KP, Xu GQ, Goh SH (2001) Chem Phys Lett 337:43
Madbouly SA, Kratz K, Klein K, Lützow K, Lendlein A (2009) In: Shastri VP, Lendlein A, Gall K (eds) Active polymers. Mater Res Soc Symp Proc, vol 1190, 1190-NN04-04
Rickert D, Moses MA, Lendlein A, Kelch S, Franke RP (2003) Clin Hemorheol Micro 28:175
Rickert D, Lendlein A, Schmidt AM, Kelch S, Roehlke W, Fuhrmann R, Franke RP (2003) J Biomed Mater Res B Appl Biomater 67B:722
Neffe AT, Hanh BD, Steuer S, Lendlein A (2009) Adv Mater 21:3394
Mondal S, Hu JL, Zhu Y (2006) J Memb Sci 280:427
Mondal S, Hu JL (2006) J Elastomers Plast 38:261
Wei ZG, Sandstrom R, Miyazaki S (1998) J Mater Sci 33:3743
Gall K (2002) Acta Mater 50:5115
Liu Y, Lv H, Lan X, Leng J, Du S (2009) Compos Sci Technol 69:2064
Ratna D, Karger-Kocsis J (2008) J Mater Sci 43:254
Gunes IS, Jana SC (2008) J Nanosci Nanotechnol 8:1616
Mondal S, HU JL (2006) Iran Polym J 15:135
Gunes IS, Cao F, Jana SC (2008) Polymer 49:2223
Scarborough SE, Cadogan DP (2006) Applications of inflatable Rigidizable Structures Soc Adv Mater Proc Eng Conference SAMPE 2006, Long Beach, CA, April 30 – May 4, 2006
Biswas M, Sinha RS (2001) Adv Polym Sci 155:167
Giannelis EP (1998) Appl Organomet Chem 12:675
Xu R, Manias E, Snyder AJ, Runt J (2001) Macromolecules 34:337
Gilman JW, Jackson CL, Morgan AB, Harris R (2000) Chem Mater 12:1866
Sinha RS, Yamada K, Okamoto M, Ueda K (2002) Nano Lett 2:1093
Vaia RA, Jandt KD, Kramer EJ, Giannelis EP (1995) Macromolecules 28:8080
Ray SS, Okamoto K, Okamoto M (2003) Macromolecules 26:2355
Pinnavaia TJ, Beall GW (2000) Polymer-clay nanocomposites. Wiley, New York
Lietz S, Yang JL, Bosch E, Sandler JKW, Zhang Z, Altsta V (2007) Macromol Mater Eng 292:23
Yang JL, Zhang Z, Schlarb AK, Friedrich K (2006) Polymer 47:2791
Cao F, Jana SC (2007) Polymer 48:3790
Madbouly SA, Otaigbe JU, Nanda AK, Wicks DA (2007) Macromolecules 40:4982
Nanda AK, Wicks DA, Madbouly SA, Otaigbe JU (2006) Macromolecules 39:7037
Wicks DA, Nanda AK, Otaigbe JU, Madbouly SA (2008) U.S. Pat. Appl. Publ. 7pp. CODEN: USXXCO US 2008108773 A1 20080508 CAN 148:518339 AN 2008:555910
Pattanayak A, Jana SC (2005) Polymer 46:3275
Pattanayak A, Jana SC (2005) Polymer 46:3394
Pattanayak A, Jana SC (2005) Polymer 46:5183
Pattanayak A, Jana SC (2005) Polym Eng Sci 45:1532
Mohr R, Kratz K, Weigel T, Lucka-Gabor M, Moneke M, Lendlein A (2006) Proc Natl Acad Sci USA 103:3540
Liang C, Rogres CA, Malafeew EJ (1997) J Intell Mater Syst Struct 8:380
Ohki T, Ni Q, Ohako N, Iwamoto M (2004) Composites A35:1065
Cho JW, Lee SH (2004) Eur Polym J 40:1343
Hong SJ, Yu WR (2008) Macromol Res 16:644
Kim MS, Jun JK, Jeong HM (2008) Compos Sci Technol 68:1919
Joshi M, Butola BS (2004) J Macromol Sci Polym Rev C44:389
Liu B, Ding Q, He Q, Cai J, Hu B, Shen J (2006) J Appl Polym Sci 99:2578
Teh PL, Ishak ZAM, Hashim AS, Karger-Kocsis J, Ishiaku A (2006) J Appl Polym Sci 100:1083
Chang Y, Yang Y, Ryu S, Nah C (2002) Polym Int 51:319
Jeong HM, Ahn BK, Kim BK (2001) Eur Polym J 37:2245
Jeong HM, Lee SH, Cho KJ, Jeong YT, Kang KK, Oh JK (2002) J Appl Polym Sci 84:1709
Jeong HM, Song JH, Chi KW, Kim I, Kim KT (2002) Polym Int 51:275
Kim BK, Lee JS, Lee YM, Shin JH, Park SH (2001) J Macromol Sci Phys B40:1179
Choi N, Lendlein A (2007) Soft Matter 3:90
Lee HY, Jeong HM, Lee JS, Kim BK (2000) Polym J 32:23
Choi YJ, Kim BK, Jeong HM (1998) Polymer (Korea) 22:131
Gopakumar TG, Lee JA, Kontopoulou M, Parent JS (2002) Polymer 43:5483
Ranade A, Nayak K, Fairbrother D, D’Souza NA (2005) Polymer 46:7323
Zhang J, Jiang DD, Wilkie CA (2005) Thermochim Acta 430:107
Rezanejad S, Kokabi M (2007) Eur Polym J 43:2856
Trznadel M, Pakula T, Kryszewki M (1985) Polymer 26:1019
POSS is a trademark of hybrid plastics (www.hybridplastics.com)
Lichtenhan JD (1995) Commun Inorg Chem 17:115
Schwab JJ, Lichtenhan JD (1998) Appl Organomet Chem 32:707
Laine RM, Zhang C, Sellinger A, Viculis L (1998) Appl Organomet Chem 12:715
Lucke S, Stopperk-Langner K (1999) Appl Surf Sci 145:713
Gao F, Tong Y, Schricker SR, Culbertson BM (2001) Polym Adv Technol 12:355
Tegou E, Bellas V, Gogolides E, Argitis P (2004) Microelect Eng 73/74:238
Kickelbick G (2002) Prog Polym Sci 28:83
Lichtenhan JD, Vu NQ, Carter JA, Gilman JW, Feher FJ (1993) Macromolecules 36:2141
Lee KM, Knight PT, Chung T, Mather PT (2008) Macromolecules 41:4730
Knight PT, Lee KM, Qin H, Mather PT (2008) Biomacromolecules 9:2458
Goldman A (1990) Modern ferrite technology. Van Nostrand Reinhold, New York
Conti S, Lenz M, Rumpf M (2007) J Mech Phys Solids 55:1462
Goya GF, Berquo TS, Fonseca FC, Morales MP (2003) J Appl Phys 94:3520
Szabo D, Szeghy G, Mrinyi M (1998) Macromolecules 31:6541
Rosensweig RE (2002) J Magn Magn Mater 252:370
Buckley PR, Mckinley GH, Wilson TS, Small W, Benett WJ, Bearinger JP, McElfresh WM, Maitland D (2006) J IEEE Trans Biomed Eng 53:2075
Conti S, Lenz M, Rumpf M (2008) Mater Sci Eng 481:351
Zhang XK, Li YF, Xiao JQ, Wetzel ED (2003) J Appl Phys 93:7124
Ma M, Wu Y, Zhou J, Sun Y, Zhang Y, Gu N (2004) J Magn Magn Mater 268:33.
Razzaq MY, Anhalt M, Frormann L, Weidenfeller B (2007) Mater Sci Eng 444:227
Weigel T, Mohr R, Lendlein A (2009) Smart Mater Struct 18:025011
Oshima C, Nagashima A (1997) J Phys - Condens Mat 9:1
Cohen ML (2001) Mat Sci Eng C 15:1
Guo T, Nikolaev P, Rinzler AG, Tomanek D, Colbert DT, Smalley RE (1995) J Phys Chem 99:10694
Harris PJF (2009) Carbon Nanotube Science: Synthesis, Properties and Applications. Cambridge University Press, New York
Reich S, Thomson C, Maultzsch J (2004) Carbon nanotubes: basic concepts and physical properties. WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Koerner H, Price G, Pearce NA, Alexander M, Vaia RA (2004) Nat Mater 3:115
Taya M, Kim WJ, Ono K (1998) Mech Mater 28:53
Koerner H (2003) ACS New Orleans, 23–27 March 2003, MTLS-017
Grady BP, Pompeo F, Shambaugh RL, Resasco DE (2002) J Phys Chem B 106:5852
Cho JW, Kim JW, Jung YC, Goo NS (2005) Macromol Rapid Commun 26:412
Grunlan JC, Mehrabi AR, Bannon MV, Bahr JL (2004) Adv Mater 16:150
Allaoui A, Bai S, Cheng HM, Bai JB (2002) Compos Sci Technol 62:1993
Paik H, Goo NS, Jung YC, Cho JW (2006) Smart Mater Struct 15:1476
Meng Q, Hu J, Zhu Y (2007) J Appl Polym Sci 106:837
Miaudet P, Derre A, Maugey M, Zakri C, Piccione PM, Inoubli R, Poulin P (2007) Science 318:1294
Velasco-Santos C, Martinez-Hernandez AL, Fisher FT, Ruoff R, Castano VM (2003) Chem Mater 15:4470
Geng H, Rosen R, Zheng B, Shimoda H, Fleming L, Liu J, Zhou O (2002) Adv Mater 14:1387
Lin Y, Zhou B, Fernando KAS, Liu P, Allard LF, Sun YP (2003) Macromolecules 36:7199
Huang W, Fernando S, Allard LF, Sun YP (2003) Nano Lett 3:565
Dalton B et al. (2003) Nature 423:703
Miaudet P et al. (2005) Nano Lett 5:2212
Vigolo B et al. (2000) Science 290:1331
Miyamoto Y, Fukao K, Yamao H, Sekimoto K (2002) Phys Rev Lett 88:225504
Gall K et al. (2005) J Biomed Mater Res A 73:339
Humbeeck JV et al. (2001) Adv Eng Mater 3:837
Kornbluh RD et al. (2002) Proc SPIE 4698:254
Patoor E, Lagoudas DC, Entchev PB, Brinson LC, Gao X (2006) Mech Mater 38:391
Lagoudas DC et al. (2006) Mech Mater 38:430
Leng J, Lv H, Liu Y, Du S (2008) J Appl Phys 104:104917
Leng J, Huang W, Lan X, Liu Y, Du S (2008) Appl Phys Lett 92:204101
Leng J, Lan X, Liu Y, Du S, Huang W, Liu N, Phee S, Yuan Q (2008) Appl Phys Lett 92:014104
Beloshenko VA, Beygelzimer YE, Borzenko AP, Varyukhin VN (2002) Compos Part A 33:1001
Yaroshenko OP, Shapranov VV, Savoskin MV, Popov AF (1997) Ukraine Patent No. 18065
Jarcho M (1981) Clin Orthop 157:259
Legeros RZ (1981) Progr Cryst Growth Char Mater 4:1
Steendam R, van Steenbergen MJ, Hennink WE, Frijlink HW, Lerk CF (2001) J Control Release 70:71
Hulbert SF, Young FA, Mathews RS, Klawitter JJ, Talbert CD, Stelling FH (1970) J Biomed Mater Res 4:433
Bonfield W (1988) Ann NY Acad Sci 523:173
Labella R, Braden M, Deb S (1994) Biomaterials 15:1197
De Groot K, De Putter C, Smitt P, Driessen A (1981) Sci Ceram 1:433
Zheng X, Zhou S, Li X, Weng J (2006) Biomaterials 27:4288
Zhou S, Zheng X, Yu X, Wang J, Weng J, Li X, Feng B, Yin M (2007) Chem Mater 19:247
Hench LL, Polak JM (2000) Science 295:1014
Niemela T (2005) Polym Degrad Stab 89:492
Stamboulis A, Hench LL, Boccaccini AR (2002) J Mater Sci Mater Med 13:843
Zheng X, Zhou S, Yu X, Li X, Feng B, Qu S, Weng J (2008) J Biomed Mater Res B Appl Biomater 86B:170
Choi D, Kumta PN (2007) Mater Sci Eng C 27:377
Yang B, Huang WM, Li C, Chor JH (2005) Eur Polym J 41:1123
Prokop A, Jubel A, Hahn U (2005) Biomaterials 26:4129
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Madbouly, S.A., Lendlein, A. (2009). Shape-Memory Polymer Composites. In: Lendlein, A. (eds) Shape-Memory Polymers. Advances in Polymer Science, vol 226. Springer, Berlin, Heidelberg. https://doi.org/10.1007/12_2009_28
Download citation
DOI: https://doi.org/10.1007/12_2009_28
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-12358-0
Online ISBN: 978-3-642-12359-7
eBook Packages: Chemistry and Materials ScienceChemistry and Material Science (R0)