A novel polysiloxane elastomer based on reversible aluminum-carboxylate coordination

Abstract

Non-covalently crosslinked elastomers based on metal–ligand complexations have been comprehensively reported as functional materials, yet the coordination systems have been mostly built using nitrogen-containing aromatic ligands and transition metal ions. Inspired by the diverse nature of metal–ligand complexations, herein we report a novel network that, to the best of our knowledge, for the first time uses main-group aluminum ions to crosslink carboxyl-modified polysiloxanes [Al(III)-CPDMS]. The Al(III)-carboxylate complexation has been revealed to be effective in constructing an elastic network with self-healing and reprocessing properties. The mechanical and self-healing properties of Al(III)-CPDMS can be regulated by changing its chain length, carboxyl grafting density and Al(III) feed. We intend to present a design concept that facilely uses versatile coordination bonds to create novel functional elastic systems.

---

Introduction

Recently, reversible interactions have been increasingly adopted by polymer researchers to construct cross-linking systems with advanced functionalities that can be rarely realized by covalently cross-linked permanent networks. By incorporating dynamic covalent bonds (such as disulfide,^1–3 boronic ester and boroxine bonds^4–6 or Diels–Alder adducts^7–9) or non-covalent bonds (such as hydrogen bonds,^10–15 π–π stacking interactions,^16,17 host–guest interactions,^18,19 ionic interactions^20–22 and metal–ligand coordination^23–32) into the polymer matrix, cross-linked networks have been endowed with many novel capabilities, including self-healing, reprocessing, and stimuli-responsive properties. Reversible elastomers based on metal–ligand complexation were initially inspired by the byssus material secreted by mussels, which has abundant imidazole-containing histidines and catechol-containing DOPAs to bind metal ions, such as Zn(II) and Fe(III). These strong and reversible metal–ligand complexations help mussels attach to hard surfaces and repair their broken byssus.²⁸ Mimicking these natural metal–ligand bindings, many polymer coordinating systems have been designed mostly based on nitrogen-containing aromatic rings (such as pyridines,^23,25 triazoles²⁹ and catechols²⁷) as multi-dentate ligands and transition metals, such as Zn(II),³⁰ Fe(III),^25,27 Co(II)²³ and Cu(I),³¹ or rare earth metals, such as Eu(III),²⁹ as center ions. These choices have effectively led to the building of tight-coordination networks because the aromatic conjugation helps to stabilize the ligands and the transition/rare earth metal ions have an empty d orbit for accepting electrons. Besides this, the aromatic rings might also be able to form micro-crystalline zones due to π–π stacking interactions, which favors the materials’ mechanical performance. Although metal–ligand coordination has proven to be useful and promising in preparing reversible and tunable elastomers, most of the reported systems were synthesized via intricate molecular modification and processing methods and were limited in their choice of strongly coordinating transition or rare earth metals and aromatic ligands. Despite this design strategy, Zuo et al.³⁰ recently reported a cross-linked system based on weak but abundant Zn(II)-carboxylate coordination bonds, where Zn(II) acted as a borderline acid and carboxylate as a hard base. A thermo-healable but rigid thermoset was thus synthesized using polysiloxane backbones that have been widely used as addition-cured elastic materials,^33–35 because the exceedingly dense carboxyl grafting restrains the chains’ mobility. Inspired by natural interactions, materials scientists have well established that weak interactions, such as hydrogen bonds or even van der Waals interactions,³⁵ can be employed to build functional elastic networks.

Nevertheless, metal–ligand coordination bonds, which are usually stronger than hydrogen bonds, seem to be less reported in the creation of elastomers, particularly the use of main-group metals such as Al(III), compared with comprehensively reported transition and rare earth metals. Al(III) coordination has been studied in the environmental field and in biological systems. Martell³⁶et al. described the aquo aluminum ion as the “hardest” of the trivalent metal ions, due to the high affinity of Al(III) for anions, e.g. OH⁻, alkoxide ions, phenoxide ions, and carboxylates. These ligands can form multi-dentate or mono-dentate coordination compounds with Al(III) via their negative oxygen donors. The Al–O bonds usually have a high bond energy, but are also susceptible to hydrolysis because OH⁻ usually has a much higher affinity to Al(III). As a result, the effective building of a crosslinked network with Al(III) coordination is challenging to due to its weak hydrolytic stability, and thus to date has not been reported.

Inspired by the diverse range of complexations that occur in nature, we aim to construct novel coordinating elastomers via facile technologies and simple interactions. In this study, we adopt Al(III), a main-group metal, as the center ion, and simple carboxylates as the ligands, to build a reversibly crosslinked polysiloxane network, Al(III)-CPDMS. Such complexation has been examined in its capability of crosslinking the polysiloxane backbones to afford an elastic network and in its reversibility to endow the network with self-healing and reprocessing capabilities.

Experimental

Materials

Octamethylcyclotetrasiloxane (D₄, 99%) was supplied by Dow Corning Silicones. Tetramethyltetravinylcyclotetrasiloxane (D^Vi₄, 97%) was purchased from Hubei Xingmingtai Chemicals (Wuhan, China). Benzoin dimethyl ether (DMPA, 99%), 3-mercaptopropionic acid (98%), tetramethylammonium hydroxide pentahydrate (NMe₄OH·5H₂O, 97%), 1,3-divinyl-1,1,3,3-tetramethyldisiloxane (MM^Vi, 97%), anhydrous acetonitrile (99.9%) and anhydrous aluminum chloride (99%) were purchased from Macklin (Shanghai, China) and used as received. Chloroform, methanol, tetrahydrofuran and other solvents were supplied by Guangzhou Chemical Reagent Factory (Guangzhou, China) and dried before use.

Synthesis

Synthesis of carboxyl modified polydimethylsiloxanes (CPDMS). The carboxyl modified polydimethylsiloxanes (CPDMS) were synthesized via two reaction steps, as illustrated in Scheme 1.

Scheme 1 Synthetic route of carboxyl modified polydimethylsiloxane (CPDMS) and the cross-linking via Al(III)-carboxylate coordination.

Firstly, vinyl-terminated linear polyvinylmethylsiloxane (PVMS) was prepared via the base-catalyzed ring-opening polymerization of D^Vi₄ and D₄, with MM^Vi as a vinyl-containing capping agent and tetramethylammonium hydroxide pentahydrate (NMe₄OH·5H₂O) as the base-catalyst. Generally, MM^Vi, D^Vi₄, D₄ and NMe₄OH·5H₂O were added to a round-bottomed flask equipped with a stirring system and a condenser. The mixture was stirred for 9 h at 90 °C and then heated to 150 °C for 1 h to decompose NMe₄OH·5H₂O. The volatile compounds were removed by rotary evaporation at 150 °C/0.1 MPa to obtain the colorless and clear liquid product PVMS in yields of usually around 95%. PVMS was synthesized with varying vinyl content (the average chain length that has one vinyl containing unit), which could be facilely controlled by varying the cyclosiloxane feed. For example, PVMS with 10 kDa weight-average molecular weight and 2 kDa average siloxane units with one vinyl-containing unit, was synthesized using D^Vi₄ (0.34 g, 1.0 mmol), D₄ (9.48 g, 32.0 mmol), MM^Vi (0.18 g, 1.0 mmol) and NMe₄OH·5H₂O (0.04 g, 0.2 mmol, 0.5 wt% to cyclosiloxane). FT-IR (KBr, cm⁻¹): 3054, 2964, 2904, 1597, 1409, 1261, 1099, 1020, 868. ¹H-NMR (400 MHz, CDCl₃, δ, ppm): 5.72–6.19 (m, –CH=CH₂), 0.02–0.22 (m, Si-CH₃).

In the second step, PVMS, 1.3 molar equivalents (to –Vi groups) of 3-mercaptopropionic acid, and 0.3 molar equivalents (to –Vi groups) of DMPA were dissolved in THF and added into a quartz flask. The mixture was stirred for 1 h at room temperature under the exposure of 365 nm UV light. Then, the solvent was removed by rotary evaporation. The crude product was dissolved in chloroform and washed three times with water/methanol (v/v, 5/3). After removing the solvents by vacuum evaporation, carboxyl modified polydimethylsiloxane (CPDMS) was obtained as a light-yellow transparent viscous liquid. Yield: 92%. FT-IR (KBr, cm⁻¹): 3100, 2967, 2910, 1715, 1416, 1261, 1110, 1007, 868. ¹H-NMR (400 MHz, CDCl₃, δ, ppm): 0.03–0.25 (m, Si-CH₃), 0.87–0.96 (m, SiCH₂CH₂S), 2.62–2.70 (m, CH₂SCH₂), 2.80–2.84 (m, CH₂COOH).

Preparation of Al(III)-coordination elastomers. The Al(III)-coordination elastomers were prepared following a general procedure: CPDMS was dissolved in chloroform and AlCl₃ in anhydrous acetonitrile. It should be noted that the solvents were strictly dried before use due to the high hydrolytic activity of Al(III). The two solutions were then mixed, stirred at room temperature for 1 h and heated to 60 °C to evaporate the solvents. The resulting transparent solid was dried under vacuum at 80 °C and heat-pressed into samples for testing.

Characterization

Infrared spectra were recorded on a Bruker Vertex 70 FT-IR spectrometer. ¹H-NMR spectra were collected with a Bruker Avance III-400 (400 MHz) spectrometer using CDCl₃ as a solvent. Gel permeation chromatography (GPC) was performed using a Waters 515 HPLC pump equipped with a Shodex K-G guard column and a Shodex K-804L chromatographic column. Detection was achieved using a Waters 2414 refractive index detector, and the samples were analyzed at 35 °C using chloroform as the eluent at a flow rate of 1.0 mL min⁻¹. The instrument was calibrated using narrow-dispersed polystyrene standards.

Differential scanning calorimetry (DSC) measurements were carried out with a DSC 204C apparatus (Netzsch Instruments, Germany). The specimens were first heated and held at 150 °C for 5 min and then cooled and held at −150 °C. After that, they were heated to 150 °C. The heating and cooling rate was 10 °C min⁻¹.

X-ray diffraction analysis was performed with an X’Pert-Pro X-ray diffractometer (PANalytical X’ Pert Pro, The Netherlands) with filtered monochromatic Cu Kα radiation in the 2θ range of 5° to 90°.

Rheological measurements were performed by a MCR102 rheometer (Anton Paar, Austria) using 25 mm aluminum plates. The dynamic temperature (10–180 °C) sweep measurements were performed at a strain amplitude of 0.15% and a frequency of 1 Hz. The dynamic frequency sweep (0.01–100 rad s⁻¹) measurements were performed at a strain amplitude of 0.15% over a temperature range of 10–120 °C.

Tensile tests were performed on rectangular specimens (50 mm × 10 mm × 1 mm) using a KJ-2091 tensile testing machine (Kejian Instruments Co. Ltd, China) at a stretching rate of 200 mm min⁻¹. The Young's modulus was calculated at a small strain (2%) from the stress–strain curves. The tests and calculations were done on 3 independent samples, and the average of these results was taken as the Young's modulus of the sample. Stress relaxation tests were also performed on the KJ-2091 tensile testing machine at 10% strain at 25 °C.

The self-healing experiments on Al(III)-CPDMS were conducted according to a typical cutting-and-healing procedure: tensile rectangular specimens were cut from the middle of the sample with a razor blade, and then the cut surfaces were gently brought together immediately. To make sure the cut surface had good contact, a very small force (<0.2 N) was applied for a very short period (<15 s). After that, the contacted samples were horizontally stored without applied stress for 24 h at room temperature or 40 °C before the tensile testing. Digital photos of the self-healing experiments are presented in Fig. S1 (Page S2, ESI†). The healing efficiency (HE) was used to evaluate the self-healing properties of different samples and calculated as follows:

HE = (TS_healed/TS_virgin) × 100% (1)

where TS_healed and TS_virgin represent the average tensile strengths of the healed and virgin samples. The tests and calculations were done on 3 independent samples, and the average of these results was taken as the HE of the sample.

The reprocessing properties were evaluated by comparing the tensile profiles of the intact and reprocessed samples. The reprocessing involved cutting the samples into pieces and then heat-pressing them into films.

In addition, we have to stress that the Al(III)-CPDMS elastomers were carefully prepared, stored under dry conditions, as well as being dried prior to all the tests.

Results and discussion

Synthesis of the aluminum-carboxylate coordination elastomers [Al(III)-CPDMS]

The carboxyl-modified polysiloxanes were characterized by FT-IR, ¹H-NMR and GPC. As is shown in Fig. 1a. The absence of vinyl bands at 1400 cm⁻¹ and 3000 cm⁻¹ and the presence of C=O at 1700 cm⁻¹ and O–H at 3000 cm⁻¹ indicated the effective modification of the material through the thiol–ene reaction. This is also suggested by the disappearance of the vinyl-related protons signal (5.72–6.19 ppm) and the presence of –CH₂–S–CH₂– (2.62–2.70 ppm) and –CH₂–C=O (2.80–2.84 ppm) in the ¹H-NMR spectra of CPDMS (Fig. 1b). The GPC traces showed that the molecular weight of the polysiloxanes slightly increased after the modification (Fig. 1c). Besides this, considering the acidity of the AlCl₃ solution, the effect of AlCl₃ on the polysiloxanes was characterized by comparing the PVMS GPC curves before and after AlCl₃ solution treatment, consistent with the Al(III)-CPDMS preparation. As is presented in Fig. S2 (Page S3, ESI†), the GPC curves of PVMS before and after the AlCl₃ treatment were found to be highly overlapped, indicating that no obvious degradation of the polysiloxane backbones occurred.

Fig. 1 Structural characterization of CPDMS and Al(III)-CPDMS: (a) FT-IR, (b) ¹H-NMR and (c) GPC measurements.

A typical FT-IR spectrum of the Al(III)-CPDMS elastomers (blue line, Fig. 1a) shows a drastic decrease in the absorbance of the carboxyl C=O peak at 1715 cm⁻¹. The adsorption peaks shifted towards lower frequencies and split into two peaks at 1600 cm⁻¹ and 1384 cm⁻¹, corresponding to the asymmetric (ν_as) and symmetric (ν_s) stretching vibrations of the carboxylate groups, respectively. The difference (Δ) between carboxylate group asymmetric and symmetric stretching vibrations (Δ = ν_as − ν_s) was approximately 216 cm⁻¹, indicating that the carboxylate groups were coordinated in a unidentate fashion to Al(III) via only one oxygen atom, instead of in a bidentate or chelating manner.^37,38 These changes suggest that most of the carboxyl side-groups were deprotonated into carboxylates during the coordination with Al(III), which is assumed to afford a more stable coordination complex with stronger electrostatic attractions.

The appearance of Al(III)-CPDMS was remarkably altered by the introduction of Al(III), changing from a viscous liquid into an elastic solid. Dynamic rheological frequency sweeps were conducted on CPDMS and Al(III)-CPDMS, as is presented in Fig. S3 (Page S3, ESI†). CPDMS_2K-10K was observed to have a viscous flow over the frequency range with G′′ > G′, while Al(III)-CPDMS showed rubbery elasticity with G′ > G′′ and a significantly higher modulus. This suggests the formation of a crosslinked network based on Al(III)-carboxylate coordination.

A series of CPDMS products was synthesized with varying molecular weights and grafting densities, which were calculated from GPC and ¹H-NMR measurements, respectively. The detailed structural parameters of the CPDMS are summarized in Table 1. All of the polysiloxane ligands are denoted as CPDMS_xK-yK, where xK represents the average chain length between two grafted carboxyl-groups on the same backbone, and yK represents the molecular weight (M_w). The simple preparation of Al(III)-coordination elastomers using AlCl₃ and CPDMS gave rise to transparent films after drying and heat-pressing. All samples were denoted as Al_z-CPDMS_xk-yK, where z represents the molar ratio between the carboxyl side-groups and Al(III).

Table 1 Structural parameters of CPDMS

Sample	M w , kDa	Đ	Grafting density^b
a The Mw (weight-average molecular weight) and Đ (Mw/Mn) were calculated from the GPC results. b The grafting density was calculated from the ^1H-NMR spectra and presented as one carboxyl-group per average chain length between the adjacent two on the same backbone.
CPDMS2K-10K	11.1	1.92	1/1991
CPDMS2K-20K	24.1	1.72	1/1970
CPDMS2K-40K	39.8	1.71	1/2034
CPDMS3K-10K	12.0	1.81	1/3244
CPDMS3K-20K	22.2	1.66	1/2730
CPDMS4K-10K	12.1	1.85	1/3725
CPDMS4K-20K	23.4	1.78	1/4455
CPDMS4K-40K	43.3	1.57	1/4079

---

The crosslinked network based on Al(III)-carboxylate coordination

The FT-IR spectra of Al_z-CPDMS_2K-10K with Al(III) and –COOH molar ratios ranging from 1 : 1 to 1 : 6 are shown in Fig. 2a. With the addition of an increasing amount of Al(III), the carboxyl absorbance (1715 cm⁻¹) decreased and the carboxylate absorbance (1600 cm⁻¹, 1380 cm⁻¹) increased. The coordination number of Al(III) is usually 4 or 6, but the disappearance of the carboxyl adsorption peak only occurred in Al₁-CPDMS_2K-10K where Al(III) is in excess of 4- or 6-coordination. This demonstrated that the reactivity of the carboxyl groups in Al-coordination was evidently reduced when attached to the polymer chains and that a perfect 6-coordinated Al(III) structure was hardly constructed, even though there were enough carboxyl groups in the system to achieve this. However, the FT-IR spectra recorded at increasing temperature (Fig. 2b) show that the Al(III)-carboxylate coordination has better thermal stability compared to hydrogen interactions. When the sample (Al₄-CPDMS_2K-10K) was heated from 40 °C to 180 °C, the carboxyl absorption at 1715 cm⁻¹ shifted to 1725 cm⁻¹ due to the dissociation of hydrogen bonds. The Al(III)-coordinated carboxylate absorption peaks at 1600 cm⁻¹ and 1384 cm⁻¹ remained unchanged over the heating process. Another important factor is that heating also helped to eliminate the inevitable, but harmful, airborne moisture, and the unchanged adsorption bands of the carboxylates suggested that moisture is not the main factor hindering the formation of 4- or 6- coordinated Al(III) during preparation, even though it is the factor that harms the material mechanical properties during storage. These findings suggest that the Al(III)-CPDMS network is mainly based on Al(III)-carboxylate coordination and partially on hydrogen bonding between the carboxyl groups that did not participate in coordination. It should also be noted that the CPDMS in this work did not form an effective crosslinked network via only H-bonds, as discussed above (Fig. S3, Page S3, ESI†).

Fig. 2 FT-IR spectra of (a) Al_z-CPDMS_2K-10K with varying amounts of Al(III) and (b) Al₄-CPDMS_2K-10K at various temperatures. (c) DSC curves and (d) XRD profiles of Al(III)-CPDMS at 25 °C.

The thermal transition of Al(III)-CPDMS was also observed using DSC, as presented in Fig. 2c. For all Al(III)-CPDMS samples, a clear glass transition temperature (T_g1), attributed to the polysiloxane backbones, was observed at around −120 °C. Another glass transition temperature (T_g2), attributed to the carbon side-chains, was found at around −8 °C. Clear melting peaks at around −52 °C for the crystallized polysiloxane backbones were only observed for the Al₄-CPDMS_4K-40K and Al₆-CPDMS_4K-40K samples, where the CPDMS had the largest M_w, lowest carboxyl grafting density and a relatively lower Al(III) feed. These transitions show that Al(III)-CPDMS is a typical linear polysiloxane-based elastomer compared to those previously reported.¹¹ The higher carboxyl modification content resulted in a slight increase of approximately 3 °C in the T_g2. It should be noted that the addition of Al(III) altered the crystallization properties of CPDMS_4k-40K, reflecting that the Al(III)-carboxylate interactions restrained the mobility of the polysiloxane segments. Besides this, no melting peaks were observed above 0 °C, indicating that there were no crystalline domains present in the Al(III)-CPDMS network at room temperature. The XRD profiles (Fig. 2d) of Al(III)-CPDMS at room temperature also confirmed the generally amorphous structure of the crosslinked network, with only broad scattering peaks detected.

Viscoelastic properties

The viscoelastic properties of the Al(III)-coordination elastomers were characterized by rheological measurements. Firstly, dynamic temperature sweeps were conducted to show the temperature dependency of Al(III)-CPDMS, as presented in Fig. 3. For most samples, a crossing-point of G′ and G′′ was observed indicating the heat-induced transition from the rubbery plateau to viscous flow, which favors the material reprocessing properties.

Fig. 3 Temperature dependency of G′ and G′′ for the Al(III)-CPDMS samples with (a) varying CPDMS chain length, (b) varying carboxyl grafting density and (c) varying Al(III) feed.

The temperature dependency of the viscoelasticity of the Al(III)-CPDMS samples was found to be highly related to their cross-linked structures. The first and major factor was the CPDMS chain length, the effect of which is illustrated in Fig. 3a. For Al₄-CPDMS_2K-10K, Al₄-CPDMS_2K-20K and Al₄-CPDMS_2K-40K, which had similar carboxyl-grafting densities and Al(III) addition, the doubling in chain length resulted in nearly double the transition temperature (50 °C, 91 °C and 170 °C respectively). The higher molecular weight of CPDMS not only brought more physical entanglements to the system, but also enhanced the Al(III)-carboxylate coordination by reducing the mobility of the segments. The carboxyl-grafting density (Fig. 3b) and Al(III) feed (Fig. 3c) had a similar influence on the transition temperature, increasing it by approximately 20 °C, when these factors were elevated by one level, due to their positive correlation with the density of crosslinking based on Al(III)-carboxylate coordination.

The master curves were constructed using the isothermal dynamic frequency sweeps at various temperatures ranging from 10 to 120 °C by applying the time–temperature superposition (TTS) principle. In Fig. 4, at the reference temperature (10 °C), all the master curves show a rubbery plateau at higher shear frequency and a crossing-point of G′ and G′′, below which a viscous flow was observed. This can be explained by the Al(III)-carboxylate coordination, with a shorter lifetime compared to covalent bonds, which played a critical role in the formation of the crosslinked networks and the maintenance of the crosslinking density at high shear frequency to support the rubbery plateau of the Al(III)-CPDMS. The transition towards a viscous flow at a lower shear frequency demonstrates that Al-carboxylate coordination, as a reversible link, can only crosslink the linear CPDMS chains transiently, which is similar to many other reported systems with reversible cross-links.^11,12

Fig. 4 Frequency dependency of G′ and G′′ for the Al(III)-CPDMS samples with (a) varying CPDMS chain length, (b) varying carboxyl grafting density and (c) varying Al(III) feed.

The viscous–elastic transition induced by the increase in the frequency was also dominated by the crosslinking structure. However, the increase in the CPDMS chain length (Fig. 4a) still plays a major role in the Al(III)-CPDMS transition. Considering that Al₄-CPDMS_2K-10K, Al₄-CPDMS_2K-20K and Al₄-CPDMS_2K-40K actually had the same density of Al(III)-carboxylate coordination, the difference in their dynamic frequency sweeps reflects the chain entanglement and mobility restraining effects induced by the increased chain length. Again, both the carboxyl-grafting density (Fig. 4b) and Al(III) feed (Fig. 4c) had a positive correlation with the crosslinking density. An increase in these two factors helped the system extend the rubbery plateau to a lower frequency zone by forming a more densely crosslinked network.

Tensile and self-healing properties

The tensile properties of the Al(III)-CPDMS samples are demonstrated by the stress–strain curves and Young's modulus measurements shown in Fig. 5. Firstly, when Al(III) was in a 0.5 molar equivalent to the carboxyl groups, the tensile profiles changed drastically with the CPDMS structures. A remarkable enhancement in the breaking strength was realized by increasing the CPDMS chain length to 40 kDa (Al₂-CPDMS_2K-40K). Normally, a crosslinked network with a longer chain length and more chain entanglements would break at a greater strain rather than a higher stress, because the disentanglement would dissipate the stress and promote deformation. However, for Al₂-CPDMS_2K-10K, Al₂-CPDMS_2K-20K and Al₂-CPDMS_2K-40K, which had the same Al-coordination density and varying chain length, similar Young's moduli (Fig. 5b) were observed at the initial stretch, but elevated strain and stress values at breaking. This suggests the similar structural response of the systems with the same crosslinking density to the initial stretching. In the discussion about viscoelasticity, the chain length of CPDMS was found to significantly affect the viscous–elastic transition, that is, the increase in the CPDMS chain length made the material more elastic at a higher temperature and higher shear frequency zone. This might be reasoned by the fact that the increased chain entanglement is beneficial for stabilizing the reversible Al(III)-carboxylate coordination that reduces the chain flexibility. On the other hand, considering the situation that the Al(III)-carboxylate interaction is susceptible to moisture hydrolysis,³⁶ it is plausible to assume that the hydrophobic PDMS chain entanglements help to protect the Al(III)-carboxylate coordination from moisture. In the entangled network, reduced chain flexibility helped to stabilize the vulnerable complexation and consequently protected the effective crosslinks.

Fig. 5 (a) Stress–strain curves and (b) Young's moduli of the Al(III)-CPDMS samples.

Reduced carboxyl-grafting density helped Al(III)-CPDMS breaking at a larger strain (Al₂-CPDMS_4K-40K). The reduced Al(III) coordination number still played a role in maintaining the extending ability of the chains and the crosslinking network at stretching, while markedly lower Young's moduli at initial stretching were observed for Al₂-CPDMS_4K-10K and Al₂-CPDMS_4k-40K, which have a lower density of crosslinking.

The self-healing performance of Al(III)-CPDMS was evaluated by comparing the tensile strength before and after a cutting and 24-hour-healing process. Typically, the healing experiments of Al(III)-CPDMS were conducted as presented in Fig. S1 (Page S2, ESI†), that is, after the healing period, the cut-surfaces were recombined and could withstand a slight amount of stretching. The stress–strain curves of pristine and healed Al(III)-CPDMS are presented in Fig. 6a and their calculated healing efficiencies are shown in Fig. 6b. The tensile stress–strain curves of the pristine and healed samples were highly overlapped before breaking, which demonstrated that the interfacial healing helped the material reconstruct with similar crosslinked networks to those of the pristine sample. The Al(III)-CPDMS samples with a lower carboxyl-grafting density were found to be capable of healing at a higher efficiency compared to those with more grafted carboxyls, even without heating. However, the increase in the Al(III) feed barely affected the self-healing efficiency. This is because the healing mechanism involves not only the Al(III)-carboxylate recombination process, but also interfacial chain diffusion. Although both the carboxyl-grafting density and Al(III) feed control the crosslinking density, it is the carboxyl modification, not Al(III), that directly limits the mobility of the segment and consequently dominates the interfacial chain diffusion.

Fig. 6 (a) Stress–strain curves of the pristine and healed Al(III)-CPDMS samples and the (b) healing efficiency measurements. (c) Stress relaxation curves of the Al(III)-CPDMS samples and (d) the tensile profiles of the pristine and reprocessed Al₂-CPDMS_4K-10K sample by heat-pressing.

The structural influence on chain mobility was also observed in the stress relaxation, as presented in Fig. 6c. Most of the tensile strength of the Al(III)-CPDMS samples rapidly relaxed to zero, as in many other reversibly cross-linked networks. However, for Al₂-CPDMS_2K-40K, which has the most chain entanglement and Al(III)-carboxylate interactions, it took far more time (>700 min) to reach zero-stress as the stress was still decreasing at 700 min. Comparing Al₂-CPDMS_2K-40K to Al₂-CPDMS_2K-10K and Al₆-CPDMS_2K-40K, the difference in stress relaxation clearly reflected that the increased chain length and crosslinking density synergistically restrained the chain mobility.

The self-healing can be expectedly promoted when the reversible system is heated. By heat-pressing, the shattered Al(III)-CPDMS samples were also reprocessed. The tensile profiles of the intact and reprocessed Al₂-CPDMS_4K-10K samples are presented in Fig. 6d. The tensile strength and strain values at breakage for the reprocessed samples were similar to those of the pristine sample, indicating that a similar crosslinked structure was recovered after breaking and heat-pressing. A slight increase in the tensile strength was noted and can be explained by the fact that the breaking and re-heating process might have helped to erase microdefects and homogenize the system.

Conclusions

In conclusion, we report a novel elastic crosslinked network based on metal–ligand coordination that first adopts main-group aluminum ions and simple carboxyl groups. An elastic system can be effectively constructed and regulated by controlling the polysiloxane chain length and carboxyl grafting density, as well as the Al(III) feed. The reversibility of the Al(III)-carboxylate coordination endows the material with self-healing and thermally reprocessing abilities. However, further research is needed to improve the mechanical properties of the Al(III)-coordination network, to acquire a more accurate evaluation of the healing efficiency, and to gain a better understanding of the healing mechanism.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China under grant number 51473051, the Science and Technology Planning Project of Guangdong Province, China under grant number 2016A010103007, and the Fundamental Research Funds for the Central Universities, SCUT under grant number 2015ZZ062.

Notes and references

1. A. Rekondo, R. Martin, A. R. de Luzuriaga, G. Cabañero, H. J. Grande and I. Odriozola, Mater. Horiz., 2014, 1, 237–240.
2. R. Martin, A. Rekondo, A. R. de Luzuriaga, G. Cabañero, H. J. Grande and I. Odriozola, J. Mater. Chem. A, 2014, 2, 5710–5715.
3. Y. Lai, X. Kuang, P. Zhu, M. Huang, X. Dong and D. Wang, Adv. Mater., 2018, 1802556.
4. W. A. Ogden and Z. Guan, J. Am. Chem. Soc., 2018, 140, 6217–6220.
5. J.-C. Lai, J.-F. Mei, X.-Y. Jia, C.-H. Li, X.-Z. You and Z. Bao, Adv. Mater., 2016, 28, 8277–8282.
6. O. R. Cromwell, J. Chung and Z. Guan, J. Am. Chem. Soc., 2015, 137, 6492–6495.
7. K. K. Oehlenschlaeger, J. O. Mueller, J. Brandt, S. Hilf, A. Lederer, M. Wilhelm, R. Graf, M. L. Coote, F. G. Schmidt and C. Barner-Kowollik, Adv. Mater., 2014, 26, 3561–3566.
8. Y.-L. Liu and T.-W. Chuo, Polym. Chem., 2013, 4, 2194.
9. N. Yoshie, S. Saito and N. Oya, Polymer, 2011, 52, 6074–6079.
10. P. Cordier, F. Tournilhac, C. Soulié-Ziakovic and L. Leibler, Nature, 2008, 451, 977–980.
11. L. Yang, Y. Lin, L. Wang and A. Zhang, Polym. Chem., 2014, 5, 153–160.
12. Y. You, W. Huang, A. Zhang and Y. Lin, J. Polym. Sci., Part A: Polym. Chem., 2016, 54, 3760–3768.
13. Y. You, A. Zhang and Y. Lin, J. Appl. Polym. Sci., 2016, 133, 43385.
14. Y. Chen, A. M. Kushner, G. A. Williams and Z. Guan, Nat. Chem., 2012, 4, 467–472.
15. M. Guo, L. M. Pitet, H. M. Wyss, M. Vos, P. Y. W. Dankers and E. W. Meijer, J. Am. Chem. Soc., 2014, 136, 6969–6977.
16. J.-F. Mei, X.-Y. Jia, J.-C. Lai, Y. Sun, C.-H. Li, J.-H. Wu, Y. Cao, X.-Z. You and Z. Bao, Macromol. Rapid Commun., 2016, 37, 1667–1675.
17. S. Burattini, B. W. Greenland, W. Hayes, M. E. Mackay, S. J. Rowan and H. M. Colquhoun, Chem. Mater., 2011, 23, 6–8.
18. M. Nakahata, Y. Takashima, H. Yamaguchi and A. Harada, Nat. Commun., 2011, 2, 511.
19. T. Kakuta, Y. Takashima, M. Nakahata, M. Otsubo, H. Yamaguchi and A. Harada, Adv. Mater., 2013, 25, 2849–2853.
20. F. B. Madsen, L. Yu and A. L. Skov, ACS Macro Lett., 2016, 5, 1196–1200.
21. S. J. Dünki, Y. S. Ko, F. A. Nüesch and D. M. Opris, Adv. Funct. Mater., 2015, 25, 2467–2475.
22. M. A. Aboudzadeh, M. E. Muñoz, A. Santamaría, R. Marcilla and D. Mecerreyes, Macromol. Rapid Commun., 2012, 33, 314–318.
23. B. Sandmann, B. Happ, S. Kupfer, F. H. Schacher, M. D. Hager and U. S. Schubert, Macromol. Rapid Commun., 2015, 36, 604–609.
24. S. Bode, R. K. Bose, S. Matthes, M. Ehrhardt, A. Seifert, F. H. Schacher, R. M. Paulus, S. Stumpf, B. Sandmann, J. Vitz, A. Winter, S. Hoeppener, S. J. Garcia, S. Spange, S. van der Zwaag, M. D. Hager and U. S. Schubert, Polym. Chem., 2013, 4, 4966.
25. C.-H. Li, C. Wang, C. Keplinger, J.-L. Zuo, L. Jin, Y. Sun, P. Zheng, Y. Cao, F. Lissel, C. Linder, X.-Z. You and Z. Bao, Nat. Chem., 2016, 8, 618–624.
26. S. Basak, J. Nanda and A. Banerjee, Chem. Commun., 2014, 50, 2356–2359.
27. M. Krogsgaard, M. A. Behrens, J. S. Pedersen and H. Birkedal, Biomacromolecules, 2013, 14, 297–301.
28. J. H. Waite and C. C. Broomell, J. Exp. Biol., 2012, 215, 873–883.
29. B. Yang, H. Zhang, H. Peng, Y. Xu, B. Wu, W. Weng and L. Li, Polym. Chem., 2014, 5, 1945–1953.
30. J.-C. Lai, L. Li, D.-P. Wang, M.-H. Zhang, S.-R. Mo, X. Wang, K.-Y. Zeng, C.-H. Li, Q. Jiang, X.-Z. You and J.-L. Zuo, Nat. Commun., 2018, 9, 2725.
31. J. Zhang, M. Li, L. Cheng and T. Li, Polym. Chem., 2017, 8, 6527–6533.
32. G. Jiang, C. Liu, X. Liu, G. Zhang, M. Yang and F. Liu, Macromol. Mater. Eng., 2009, 294, 815–820.
33. S.-J. Wang, X.-D. Fan, J. Kong, X. Wang, Y.-Y. Liu and G.-B. Zhang, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 2708–2720.
34. J. Kong, X. Fan, G. Zhang, X. Xie, Q. Si and S. Wang, Polymer, 2006, 47, 1519–1525.
35. S.-J. Wang, X.-D. Fan, J. Kong and J.-R. Lu, Polymer, 2009, 50, 3587–3594.
36. A. E. Martell, R. D. Hancock, R. M. Smith and R. J. Montekaitis, Coord. Chem. Rev., 1996, 149, 311–328.
37. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, Wiley, New York, 1987.
38. P. Persson, M. Karlsson and L.-O. Öhman, Geochim. Cosmochim. Acta, 1998, 62, 3657–3668.

---

Footnote

† Electronic supplementary information (ESI) available: Photos of self-healing, GPC curves of PVMS and dynamic frequency sweeps of CPDMS and Al(III)-CPDMS. See DOI: 10.1039/c8nj04761h

---