Get Complete Project Material File(s) Now! »
Hydrogen bonding
Hydrogen bonding is an electrostatic attraction between a hydrogen atom and electronegative atoms, such as nitrogen, oxygen or fluorine. Nitrogen and hydrogen elements are present in proteins, DNA, glycogen etc. natural products, where they induce strong hydrogen bonding interactions and are responsible for the formation of multilevel structures and double helixes. In order to imitate biological processes, various hydrogen-bonding moieties can be introduced ureidopyrimidinone (UPy) group, which is a quadruple hydrogen bonding unit. Two UPy groups are able to assemble to form dimers. When UPy groups were incorporated into the polymer chains as endgroups, each two UPy units coming from different polymer chains were able to aggregate into UPy dimers, which was called supramolecular polymerization. In addition, the polymer chains can self-assemble into fiber structures by forming dimers stacks as shown in Figure 1.3b. Similar to nanocomposites, the fibers remarkably improve the mechanical properties, which leads to solid-like behavior of materials with UPy while non-modified polymers behave as a liquid (Figure 1.3a).
Utilizing the same hydrogen bonding moieties, Guan et al.33 synthesized a cyclic UPy core and converted it to a terminal diolefin monomer, which was further used to prepare polymer materials (Figure 1.4). Compared to the brittle control sample without hydrogen bonding motifs in the main chains, the polymer exhibits a rare combination of high modulus and high toughness resulting in the dissipation of the dissociation of hydrogen bonding. Additionally, it also shows other properties, including self-healing and shape memory. They demonstrated the first biomimetic modular polymer materials displaying a great combination of high toughness and modulus. Moreover, the polymer exhibits self-healing and shape memory behavior.
Metal-ligand interaction
Generally , a metal-ligand interaction describes the coordination complexes between a transition metal and π-donor ligands. The ligand acts as a donor of an ion pair to the metal d orbitals and the metal center can form multiple coordination bonds. The coordination bond can be very stable but still responsive to light, solvent or heat. These stimuli will cause temporary dissociation of the metal-ligand motifs. When the ligand is covalently coupled into the end group of the polymer backbone, the polymer undergoes supramolecular polymerization through metal-ligand interaction. On the other way, if the ligand is located in the middle of the polymer chains, metal-ligand association will act as physical cross-linkers by forming intermolecular cross-links or intramolecular loops.
Rowan et al.34,35 modified 2,6-bis(1’-methylbenzimidazolyl) pyridine (Mebip) ligands at the end of poly (ethylene-co-butylene) chains. These end-functionalized materials showed good mechanical properties and also showed efficient defect healing under ultraviolet light. Figure1.5 illustrates how low molecular weight polymers with ligand end groups are cross-linked through metal-ion binding and form a polymer network. Under a transmission electron microscope (TEM), the materials show a pronounced microphase separation.
Extensive studies of metal-ligand super-molecular interactions have been carried out on poly(urethane) by Weng et al36. The tridentate ligand 2,6-bis(1,2,3-triazol-4-yl)pyridine was incorporated into the polymer backbone to prepare ligand macromolecules. Upon coordinating with transition metal ions (Zn2+ and Eu3+), the materials show a rare combination of high modulus, good toughness and high deformability and self-healing capability in the presence of solvent (Figure 1.6).
Host-guest interaction
Host-guest interaction is another class of supramolecular associations, where one component are host molecules and the other component are guest molecules. Typical host molecules are cyclodextrins, crown ethers, pillararenes and cucurbiturils etc. cyclic molecules forming a large cavity and guest molecules mainly including some ammonium salt frequently possess a cationic group37. This two categories associate to form complexes. They are often utilized to form a supramolecular binding motif and serve as a tool for the fabrication of various self-assembled structures, such as micelles, nanotubes nanorods, nanosheets and vesicles37. If the functional groups are introduced in the polymer chains, they can act as non-covalent supramolecular cross-linkers upon complexation. Moreover, the association is controlled by experiment conditions (temperature, pH, ionic strength). This dynamic and reversible properties of host-guest interactions brings excellent stimuli-responsive features to the resultant elastomers.
Due to their advantages described above, Scherman et al38. covalently grafted cucurbit[8]uril on the side chains of a loosely cross-linked polymer network. In this system, the host-guest interaction was used to form sacrificial bonds which improved the fracture resistance, fatigue resistance and energy dissipation, which was achieved by the dynamic disassociation/re-association of the cucurbit[8]uril complexes (Figure 1.7).
Other kinds of super-molecular interaction
Phase separation interaction and π-π stacking interactions are another two types of supramolecular interactions which have been used in polymer materials in the past few decades. These interactions trigger supramolecular assembly. They produce a variety of well-aligned anisotropic structures39, such as fibers, spheres, cylinders and lamellae etc. which form ordered micro-phase structures at longer range. These secondary structures enhanced toughness of elastomers. Weng and his coworker40 fabricated different PnBA-PS-PnBA triblock copolymers and adjusted the length of PS to tailor the mechanical properties of elastomers. Suitable length of polymer chains allows them to assemble into a spherical structure, which improves the toughness of these thermoplastic elastomers.
Using on the π-π stacking interaction of pyrene derivatives, elaborated polymers have been synthesized by Colquhoun, Rowan and their coworkers41. The polymer chain-segment is composed of two naphthalene-diimide unites, as shown in Figure 1.8a. When they increased the concentration of naphthalene-diimide units, it resulted in a progressive increase in binding constant of 2 orders of magnitude. Furthermore, an unexpected enhancement of toughness and self-healing ability was observed when the naphthalene-diimide units were separately incorporated into two polymer chains as illustrated in Figure 1.8b.
Many additional studies about π-π stacking interaction, hydrogen bonding, metal-ligand etc. supramolecular chemistry have been carried out more recently and the last several decades have witnessed great progress in the utilization of supramolecular interactions to control self-assembly and architecture of tough elastomers42. However, these supramolecular elastomers are generally strongly viscoelastic due to the dynamic association of supramolecular interaction making it difficult to obtain fully reversible elasticity and fatigue toughness.
Multiple networks
In the previous section, we described some approaches to develop tough elastomers through the incorporation of nanoparticles to form nanocomposites and the use of supramolecular or coordination chemistry to obtain self-healing properties. In this section we present an alternative approach to make unfilled elastomers with minimal viscoelasticity but a higher toughness through network design. In this type of approach covalent bonds are used as ‘sacrificial bonds’. Here, the term of sacrificial bond means that during the extension of the polymer network a fraction of the covalent bonds in the polymer network are overloaded (through network design) and can rupture and dissipate energy before the macroscopic failure of the material occurs. Compared to supramolecular interactions, the rupture of embedded covalent bonds can dissipate more energy per bond due to their higher bond association energy but typically cannot be easily reformed.
In this section, we present several approaches to design the networks in order to introduce such sacrificial bonds (Table 1.2). First, we will introduce the concept of bimodal networks which is known from the end of last century. Next, we present a general reinforcement strategy using the double network structure which was first invented for hydrogels. Finally, we focus on the extension of this method to multiple network elastomers developed by our group to synthesize soft but tough materials.
Monomers, oligomers, and cross-linkers are used to synthesize polymer networks by various polymerization reactions. In non-controlled types of polymerizations of networks, the molecular weight of the polymer chains between two neighboring cross-links varies significantly due to side reactions or to the random activation and termination of the polymerization reaction. When only one kind of monomer and cross-linker are used in polymerization (or when they possess similarly reactivity), the distribution of the length of polymer chains between two cross-links is unimodal. When the polymer network is submitted to an external stress, the shortest chains and more extended chains in the polymer network break first. These chains are considered to be the ‘culprits’ in causing fracture of elastomer materials by nucleating macroscopic defects.
The term bimodal network means an elastomer with two populations of polymer chain length between the cross-links. For the population of short chains, it is easy to reach their maximum elongation and their rupture dissipates efficiently the energy. In principle if both populations were bicontinuous, the bimodal polymer network should perform as a tough material. For example, a PDMS bimodal network was synthesized by Mark et al43. They polymerized two kinds of PDMS precursor polymers with different chain lengths with alkoxy and hydroxyl (or vinyl) end-groups, respectively. The bimodal polymer networks were prepared through the chemical reaction between alkoxy and hydroxyl end groups. Adjusting the ratio of short /long polymer chains, they compared a series of bimodal network elastomers (Figure 1.9). However the desired significant reinforcement of the ultimate properties and of the fracture toughness was not very large compared with the unimodal network. This was presumably due to the fact that the short and long chains phase separate and form distinct regions.
More recently, Cohen et al45,46. utilized vinyl-terminated precursor chains with average molecular weights of 800-8500-91000 g/mol and a tetrakis(dimethylsiloxy)silane cross-linker to synthesize trimodal networks. Compared with the unimodal networks with a similar modulus, the trimodal networks achieved a higher toughness and elongation at break (Figure 1.10), but these networks are still too brittle for industrial applications. Therefore, other strategies to design covalent networks are desired to achieve significant improvement of the fracture toughness of “unfilled” elastomers.
Double network hydrogels
Hydrogels, have attracted a great interest in the past few years due to their potential use for artificial soft tissues. Conventional hydrogels are composed of a single network of hydrophilic polymer swollen in water and are usually very brittle, like fragile jellies, which may seriously restrict the application. This is particularly true in applications in which the mechanical properties are essential, such as articular cartilage, semilunar cartilage, tendons and ligaments and other connective tissues.
During the last decade, a great breakthrough has been achieved by Gong’s group44 in the design and synthesis of a particular class of tough hydrogels defined as double networks (DN) hydrogels. A double network hydrogel is a heterogeneous network (IPN) consisting of interpenetrated polymer networks synthesized sequentially (Figure 1.11), which is different from a conventional hydrogel. A DN hydrogel typically contains 80-90 weight percent of water, yet its toughness is comparable to that of conventional unfilled elastomers. The tough DN hydrogels were synthesized for the first time in 2003 by Gong and her coworkers using two sequential steps of photo-polymerization44. At first, a densely cross-linked polyelectrolyte (poly(2-acrylamido-2-methylpropanesulfinic acid (PAMPS)) network was synthesized by a UV-initiated polymerization carried out at a very slow polymerization rate in the glove box, and then this single network was swollen by a neutral monomer (acrylamide) solution containing a very low concentration of cross-linker and initiator. When the single network was swollen to equilibrium, a double network was obtained by performing a second UV-polymerization on the swollen PAMPS single network. The swelling in the second step caused the densely cross-linked first network to become isotropically stretched and to almost reach the maximum extensibility of the polymer chains between two cross-linkers. Therefore, the first network is stiff and the second network is soft and extensible because of its low chemical cross-link density. According to Gong’s research, the essential features of tough DN hydrogels are due to the structure and the mechanical properties of first and second network6, which can be summarized into three points as follows:
The mechanical behavior of the first network is stiff and brittle, such as that of a highly stretched polyelectrolyte. On the contrary, the second network is soft and ductile such as an unstretched neutral polymer.
The second network forms the majority of the DN hydrogel (about 90%).
The DN possesses a highly stretched first network and a loosely cross-linked second network.
Following these principles, a series of tough DN hydrogels with various compositions were prepared by Gong’s group. For instance, the DN hydrogel of PAMPS/PAAm shown in Figure 1.12. Here, PAMPS, PAAm and PAMPS/PAAm stand for poly (2-acrylamido-2-methylpropanesulfonic acid), polyacrylamide and double network, respectively. The materials were tested in uniaxial compressive and tensile tests. PAMPS/PAAm DN showed much higher stress and strain at break compared to PAMPS and PAAm gels. Compression data illustrates that highly cross-linked PAMPS is brittle as expected. On the other hand, loosely cross-linked PAAm is soft and can be highly deformed. Whereas DN of PAMPS/PAAm remains quite extensible but shows obvious strain hardening leading to high stress at break. Figure 1.12b illustrates that the PAMPS gel is brittle and breaks during the compression process, while the DN gel is tough and is able to sustain a stress of 17.2 MPa under compression despite the presence of 90 wt% of water.
hardening regimes. The presence of a necking extends the maximum strain of the DN before rupture. In cyclic loading, when comparing the first and second cycle, the DN shows a large hysteresis and after the first cycle the sample becomes fairly soft, because of an internal fracture of the first network embedded in the DN. Based on the results obtained by small angle neutron scattering (SANS), the neutron scattering pattern from the DN hydrogels in the deformed state is anisotropic. Thus it is assumed that in the necking area the first network ruptures into numerous clusters which play the role of physical cross-linkers for the second network. The details of the fracture mechanism of the DN will be discussed in the next section.
The fracture of polymer networks usually starts by the breakage of the shortest chains. In order to avoid this effect, Sakai and Gong et al.7 prepared a well-defined “homogeneous” first network using an end-group reaction (Figure 1.14a), similar to the synthesis of bimodal elastomers by Mark43. To ensure the swelling ratio of the first network made from a neutral polymer, a “molecular stent” method was proposed for the synthesis of tough DN48. The molecular stent method is a technique to synthesize tough DN hydrogels based on a neutral first network. In this technique, linear, strong polyelectrolyte chains or ionic micelles are introduced in the neutral first network48. Comparing the nominal tough DN hydrogels made from a heterogeneous first network, with this DN hydrogel made from a more homogeneous first network, both possess almost the same toughness, but a different initial modulus and hysteresis ahead of the necking region as shown in Figure 1.14b. The difference in the hysteresis is due to the better homogeneity of first network. Before yielding, the DN with a homogeneous first network shows much less evidence of fracture of polymer chains. This also leads to a lower reduction of the Young’s modulus after yielding in cyclic loading tests.
Multiple network elastomers
Inspired by the interpenetrated network of DN hydrogels, our group developed a new approach to prepare tough elastomers by using interpenetrating multiple networks without solvent. Different from previous double network elastomers made by Mark49, Yoo50, Baysal et al.51, multiple network elastomers possess a highly pre-stretched first network, which is usually constituted of short, stretched polymer chains with a high cross-link density. Etienne Ducrot52 and Pierre Millereau11, two doctors who graduated from our group, studied the effect of the first network on the mechanical properties of elastomers. In Figure 1.15a, a significant improvement in Young’s modulus, the stress at break and toughness is observed for the multiple networks elastomers. The results indicate that this approach is promising to make tough elastomers. Furthermore, the elastomers toughness increases with the number of interpenetrated polymer networks. Figure 1.15b shows the mechanical performance of DN elastomers with different cross-link densities of the first network, and the results further indicate the critical relationship between the mechanical properties of elastomers and cross-linker density of the first network. The more densely cross-linked is the first network and the higher is the Young’s modulus.
Pierre tailored the level of pre-stretch in the first network by using some solvent during the preparation of the second or additional networks. The results of the extension tests illustrate that elastomers become more and more robust (Figure. 1.16) with a high modulus and fracture toughness. Moreover, although the same monomer composition is used between the first network and the other networks (second, third network), all the elastomers have outstanding mechanical properties. The comparison of the results suggests that the reinforcement of multiple network elastomers originates mostly from the pre-stretched polymer chains in the first network which can fail without a macroscopic breakage of the whole material.
Theory of crack propagation
The fracture of elastomers is difficult to model and the fracture energy is currently very hard or impossible to predict from the knowledge of its molecular structure alone due to the presence of entanglements and the heterogeneity of the polymer network. There are a large proportion of theories that are proposed based on an ideal polymer network structure as the Lake-Thomas model. Recently, several models were constructed to explain and predict the reinforcement of DN hydrogel, as listed in Table 1.3. Thus in this part we introduce several theories about the calculation of the dissipated energy during crack propagation.
Table of contents :
Chapter 1- The design of robust elastomers and mechano-chemistry
1. Traditional tough elastomers and multiple network elastomers
1.1 Nano-composite elastomers
1.2 Supramolecular interaction elastomers
1.2.1 Hydrogen bonding
1.2.2 Metal-ligand interaction
1.2.3 Host-guest interaction
1.2.4 Other kinds of super-molecular interaction
1.3 Multiple networks
1.3.1 Bimodal networks
1.3.2 Double network hydrogels
1.3.3 Multiple network elastomers
2. Theory of crack propagation
2.1 Lake-Thomas model
2.2 Brown’s model
2.3 Tanaka’s model
2.4 The fracture mechanism of multiple network elastomers
2.5 Finite element model
3. Mechanochemistry
3.1 Introduction
3.2 Optically inactive mechanophores
3.3 Mechanoluminescent polymers
3.3.1 Mechanofluorescent polymers
3.3.2 Mechanochemiluminescent polymers
3.4 Mechanochromic polymers
Chapter 2. Synthesis and characterization of mechanically responsive multiple network elastomers
1. Standard synthesis of multiple networks
1.1 Chemical reagents
1.2 Polymerization conditions of the networks
1.3 Synthesis of the spiropyran (SP) cross-linker
1.4 Synthesis of the first (or filler) network
1.5 Preparation of a family of multiple networks elastomers
2. Synthesis of various multiple network elastomers
2.1 Effects of cross-linking in the first network
2.2 Various SP concentrations in the first network
2.3 Different monomers in the first network
3. Characterization of multiple network elastomers
3.1 Tensile tests
3.1.1 Uniaxial extension
3.1.2 Cyclic loading tests
3.1.3 Step cycle elongation tests
3.1.4 Relaxation test
3.1.5 Fracture tests
3.2 Color analysis: basic principles
Chapter 3: Mechanical properties and optical response of multiple network elastomers
1. Mechanical properties
1.1 Standard family multiple network elastomers
1.2 Various stretching rates
1.3 Different cross-link densities
1.4 Behavior of samples under cyclic loading tests
1.5 Fracture energy
2. Optical response to mechanical stress
2.1 Color analysis for the EA0.5-0.05 standard family of materials
2.2 The effect of strain rate
2.3 The effect of various cross-link densities in the filler network
2.4 The effect of varying the SP concentration
3. Accurate calibration of the Stress: Toward Quantification
4. Optical response in fracture tests
4.1 Optical response around the crack in EA0.5-0.05 family
4.2 Optical response around the crack for EA0.2-0.05(2.61)EA sample
5. Quantitative Stress distribution around the crack tip before propagation
5.1 Stress distribution in standard multiple network elastomers
5.2 Stress distribution in various elastomers at the same energy release rate
6. Mapping the Strain Energy Density
7. Preliminary results of color change during unloading
7.1 Color change in cyclic loading
7.2 Color change during the relaxation process
Chapter 4: Mapping the stress in unloading process
1. Construction of a color map of the stress
1.1 Color change during the unloading process
1.1.1 Mechanism of color change during the unloading process
1.1.2 Color map of stress in standard multiple network elastomers
1.1.3 Multiple network elastomers with different mechanical property
1.1.4 Elastomers with various SP concentrations
2. Stress distribution around the crack tip during crack propagation
2.1 Color change during crack propagation
2.2 Stress distribution in crack propagation
3. Quantify the level of activation of SP near the crack tip
Chapter 5: The fracture mechanism of multiple network elastomers
2. Results
2.1. Mechanical properties of multiple network elastomers
2.2. Mechanical response in uniaxial extension
3. Discussion
3.1 Influence of the SP position in multiple networks
3.2. The effect of the connectivity between the first and the second network
2.3. Stress transfer to the matrix network during the crack propagation process
3.3 Higher magnification detection of the stress in the matrix network
Chapter 6: Construction of the strain field around the crack tip
1. Synthesis of multiple network elastomers containing fluorescent beads
1.1 The choice of fluorescent beads
1.2 Synthesis of elastomers
2. Characterization of elastomers
2.1 The effect of fluorescent beads on mechanical properties
2.2 Confocal microscope observations
2.3 Calculation
3. Preliminary results
3.1 The vector displacement field of fluorescent beads
3.2 The strain field around the crack tip
Chapter 7. Perspective and discussion
1. Discussion around the quantification of polymer chains involved in the damage
2. Discussion about the fracture of the first network after yielding
3. The combination of the stress field and strain field around the crack tip
3.1 Calibration curve of fluorescence