Читать книгу Self-Healing Smart Materials - Группа авторов - Страница 14

The most widely explored extrinsic self-healing polymers are systems based on composites containing the healing agent into purposely designed containers that break when the material is damaged. Usually, their self-healing is autonomous, needing no other driving force than the damage itself, or in some cases the action of the surrounding environment. The key for the self-healing to take place autonomously is that the chemical precursors released can readily react at the material’s operation temperature. Quite often this is room temperature, and hence these polymeric composites must be designed in order to carry the appropriate amount of catalyst to complete the curing reaction at this low temperature. Besides this, other relevant aspects of the healing mechanism need to be accounted to achieve high healing efficiencies. The vessels containing the healing agents must break, and the healing agents have to flow to fill the crack before the curing reaction produce its gelation. Hence, a good load transfer from the matrix to the vessels combined with the specific fracture toughness of the vessels is needed, and the viscosity of the healing agents must be low enough to allow them to flow into the crack. These features have also been addressed in several previous works [22].

There are several monomers and catalysts that allow reaction rates high enough to achieve self-healing at room temperature. Grubbs catalysts are one of the most widely used for autonomous self-healing composites based on microencapsulated healing agents. It was first proposed for self-healing composites by White et al., to aid the crosslinking of dicyclopentadiene (DCPD) through a ring opening metathesis polymerization (ROMP) [12]. The DCPD was contained into urea microcapsules embedded within these composites, and the Grubbs catalyst was dispersed in the matrix. After healing during 48 h at room temperature, the researchers observed that the healed sample reached a load of 75% of that of the virgin sample in a fracture test. These results were very promising, and the healing efficiency values were quickly surpassed in the following years by similar systems that incorporated some improvements in their formulations and/or their processing [23, 24]. By protecting the catalyst with wax [25] a considerable raise in the healing efficiency was achieved, thanks to an enhanced stability of the catalyst, but this modification was also associated with a decrease in the damage resistance to impact. These systems were also used to study the effect of the particles’ size and load on the healing efficiency, which allowed to determine that microcapsules smaller than 30 mm are efficient for small cracks (around 3 mm), but larger capsules and higher loads are needed for larger cracks [26].

However, some shortcomings of the DCPD-Grubbs catalyst systems (low long term stability due to possible side reactions with air and the polymer matrix) [27], encouraged the search for other chemistries that could overcome these difficulties. The condensation of hydroxy capped polydimethylsiloxane (HOPDMS) and polydiethoxysiloxane (PDES) with dibutyltindilaurate (DBTDL) in a vinyl ester matrix showed an improved stability, but the healing efficiencies were rather poor [27]. Much better results were observed for a polydimethylsiloxane (PDMS) elastomer filled with poly(urea-formaldehyde) (UF) microcapsules containing the same precursor and Pt initiator used to synthesize the PDMS matrix separately [28]. The authors studied the breaking mechanism of the capsule, and provided an interesting insight from the observation of the process of deformation and failure of a microcapsule within the elastomeric matrix. The sequence is shown in Figure 1.3. For stretch values below 1.5 the micro-capsule is deformed along with the matrix, withstanding the stress, but for higher deformations, the failure of its wall releases the chemical precursors. This study focused on some very important aspects of the use of microcapsules for self-healing composites, such as the microcapsule rupture process, and how its presence affects some properties of the matrix. It clearly showed that the breaking of the capsules is a quite complex matter, and even brittle materials (when evaluated in bulk) such as UF may need relatively high strain values to fail [28, 29]. The authors also observed an improvement on tear strength with the increase in the microcapsules concentration (up to 20 wt% of microcapsules), which is in agreement with previous results for reinforced PDMS [30, 31]. To ease the flow of the healing agent off the microcapsules and into the crack, its viscosity was reduced by using heptane as diluent [28]. Using a similar system, consisting in a silanol-terminated polydimethylsiloxane (STP) as healing agent and dibutyltin dilaurate (DTDL) as catalyst, Kim et al. obtained a self-healing coating that could repair itself at temperatures as low as −20 °C. The viscoelastic material produced by the healing agent was capable of protecting the coated and scratched specimens from saline solution uptake upon immersion for 48 h at −20 °C [32].

Figure 1.3 Images of a single PDMS microcapsule subjected to uniaxial tension at different values of deformation (γ). Reprinted with permission from Ref. [28]. Copyright (2007) John Wiley & Sons, Inc.

Microencapsulated epoxy precursors were also applied as external healing agent in matrices with a dispersed latent initiator [33–37]. Yin et al. utilized UF microcapsules containing an epoxy resin based on diglycidyl ether of bisphenol A (DGEBA), and a latent catalyst consisting in a complex of CuBr₂ and 2-methylimidazole (MeIm), dispersed in the matrix [33]. Their results showed very good healing efficiencies, with healed specimens showing fracture toughness values 11% higher than those of the virgin samples. As the authors explained, this is a side effect of the higher fracture toughness of the resin used as healing agent when compared to the matrix, probably due to the higher temperature and/or the different curing mechanism of the healing reaction: the samples were synthesized at temperatures up to 100 °C, and the subsequent healing was performed at 130 °C. They also applied the self-healing matrix to the synthesis of self-healing epoxy laminates, using glass fibers [33, 34]. However, the healing efficiencies were lower in this case, and an important downside of the use of microcapsules was the reduction of the tensile properties for microcapsule loads above 30 wt% [34]. Guadagno et al. assessed the healing efficiency and the mechanical properties of self-healing epoxy systems with different flexibilizers [36, 37]. When the curing temperature of these systems was raised from 170 to 180 °C, the mechanical moduli increased, but the healing efficiency slightly decreased, probably due to an initial thermolytic decomposition of the Hoveyda–Grubbs’ first generation catalyst. Interestingly, the replacement of the Heloxy71 used as flexibilizer by the reactive diluent 1,4-butanedi-oldiglycidylether (BDE) leads to an increase in the compressive modulus, from around 0.7 to 1.4 GPa, with no loss in the healing efficiency (above 80%) [37]. Hia et al. demonstrated multiple self-healing ability in epoxy composites with alginate multicore microcapsules [38]. Three to four healing cycles were produced, though the efficiencies were around 55%, which are lower than for other systems. Moreover, the authors found that self-healing specimens have lower impact strength than the neat epoxy polymers, due to the large size of the capsules, and its high load.

Polyurethanes (PU) were explored for extrinsic self-healing systems [39, 40]. He et al. used isophorondiisocyanate (IPDI) encapsulated into polyurea capsules as self-healing agent for PU matrices and also for epoxy-based ones [41]. They show that complete healing (efficiencies around 100%) of both epoxy and PU matrices is achieved for capsules with diameters of 96 mm or higher. Smaller capsules produced poorer healing performances. Gil et al. used microencapsulated diisocyanate to improve the tensile strength of collagen [42]. They used two different diisocyanates: IPDI and 4,4’-diphenylmethane diisocyanate (MDI); the isocyanate groups react with the collagen, creating new crosslinks and mending the damages. Other healing agents employed for extrinsic self-healing coatings, including thiol-ene and azide-alkyne precursors, as well as vinyl ester and unsaturated polyesters have also been tested, as pointed out in a review by Hillewaere and Du Prez [43].

A smart approach proposes to develop self-healing systems especially designed so that the healing can be activated by the environment surrounding the material when a crack propagates. Different research groups used encapsulated isocyanate that reacts with water when released into aqueous environments, producing coatings with potential applications in offshore and other marine devices. Di Credico et al. also used encapsulated IPDI to provide self-healing capacity to a DGEBA-based epoxy matrix, thanks to the crosslinking reaction of isocyanate with the surrounding water [44]. Figure 1.4-I shows the damaged coating, and the healed coating after immersion for 48 h in salty water. The authors emphasized that the rough outer surface of the microcapsules played a key role improving the adhesion to the matrix, allowing the capsule to fail and release the IPDI. Wang et al. also used IPDI as healing agent to repair cracks in an alkyd varnish coating (AVC) [45]. Figure 1.4-II shows the aspect of the mended scratch on different substrates after different times of exposure to marine salty water. Though some seawater could penetrate into the coating and reach the substrate, the self-healing prevented a much larger damage.

Light is another possible external stimulus that can be harnessed to trigger the healing response. Sunlight was proposed by some authors to induce the self-healing of polymeric coatings. Song et al. designed the first photoinduced microcapsule-based self-healing coating from tetraethyl ortosilicate and a polysiloxane, with encapsulated methacryloxypropylterminated polydimethylsiloxane (MAT-PDMS) and benzoinisobutyl ether (BIE) photoinitiator [46]. The conversion of the healing agent upon exposure to real sunlight was somewhat lower than when an artificial UV light was used instead. However, several tests (optical and SEM examination, and measurements of water uptake and chloride ion penetration) showed that relatively short times (4 h) of sunlight exposition are enough to produce an acceptable healing. Figure 1.5 shows the results obtained, comparing the performance of the self-healing coating with the control coating (without the encapsulated healing agent). Khalaj Asadi et al. successfully encapsulated a sunlight curable silicon based resin, and studied several parameters affecting some final properties of the capsules [47]. They used polyvinylpyrrolidone as emulsifier, and established the optimum amount to obtain higher yields. These capsules showed good resistance to water and xilene, but ethanol produced a rapid deterioration of the microcapsules. The authors remarked that these microcapsules can be readily used for self-healing polymers formulations.

Figure 1.4 (I)—SEM micrographs of crack in a coating with PU/PUF microcapsules (a) before and (b) after immersion in salt water for 48 h. Reprinted from Ref. [44]; Copyright (2013) with permission from Elsevier. (II)—Alkyd varnish coatings on a titanium surface after 200 and 1,200 h of seawater immersion. Reprinted from Ref. [45] with permission from The Royal Society of Chemistry.

The use of UV light to trigger the mending reactions was explored as well by other researchers. Gao et al. encapsulated a photosensitive resin obtained from a mixture of Bisphenol A epoxy resin diacrylate ester (BAEA) and trimethylolpropane-triacrylate (TMPTA) with 1-hydroxy-cyclohexyl-phenyl-ketone into UF capsules, and embedded them into an epoxy-amine matrix [48]. The microcapsules were synthesized containing TiO₂ in its shell in order to absorb the UV light and protect the photosensitive resin into the undamaged capsules from curing before being released. Anticorrosion tests were performed after scratching the samples and healing during 30 s with UV irradiation. The neat epoxy-amine matrix and a composite with capsules without TiO₂ were used as control (CC1 and CC2 respectively). Corrosion was observed in CC1, but CC2 and the self-healing coating could protect the steel substrate. When the experiment was repeated with a new scratch on the same samples, CC2 could no longer protect the steel, and corrosion was observed. Figure 1.6 shows the images of the samples after each test. For the self-healing coatings, the experiment was repeated 5 times, and the self-healing ability was eventually lost after 5 irradiation events. SEM images of the scratch and the electric current measured for each case confirmed the previous observations.

Zhu et al. also used a UV-curable healing agent into microcapsules with a rapidly degradable inner polymeric shell and an outer TiO₂ shell that can absorb UV radiation [49]. The action of the TiO₂ shell helps to degrade the inner shell, releasing the healing agent. Hence, the self-healing composite displays a dual release mechanism that enhances its efficiency. The micro-encapsulated healing agent consisted in an epoxy silicone with a photosensitive initiator (triarylsulfonium hexafluorphosphate salt) and the matrix was based in silicone resins. Figure 1.7 shows a scratch on the coatings after 12 h of UV irradiation. The comparison was made using composites with microcapsules without the healing agent (labeled as “BS-xx”), and composites prepared with capsules filled with the healing agent but unable to fail and release it by UV irradiation, due to a low concentration of TiO₂ NPs in its outer shell (labeled as “CS-xx”). The self-healing coatings were labeled as “SH-xx”. The numbers xx represent the wt% of microcapsules. The effect of the healing agent released within the crack is very clear, and for a microcapsules load of 60 wt% the healing seems to be excellent.

Figure 1.5 (I)—Scheme of the sunlight induced healing mechanism: the crack breaks the microcapsules and release the healing agent, which undergoes the crosslinking reaction upon exposure to sunlight. (II)—Water uptake measurements for the plain mortar, and mortars coated with the control and the self-healing coating. (III)—Chloride penetration tests. Current vs. elapsed time, and accumulated charge during 6 h for the undamaged control coating (a), scribed control coating (b) and scribed and healed self-healing coating (c). Reprinted with permission from Ref. [46]. Copyright (2013) American Chemical Society.

Figure 1.6 Steel substrates coated with (a) CC1, (b) CC2 and (c) self-healing coating, after successive scribing and healing sequences. Reprinted with permission from Ref. [48]; Copyright (2015) American Chemical Society.

Some drawbacks of the use of microcapsules/hollow microfibers are worth to mention. Samadzadeh et al. [50] have mentioned some of them, including the negative side effects on the mechanical properties of the material, such as Young’s modulus and ultimate stress [50, 51]. Adhesive properties can also suffer a decrease due to the presence of microcapsules [50]. In most cases a compromise between an acceptable healing with a minor deterioration of the resistance has to be reached. Additionally, there are some aspects that should not be overlooked when designing a self-healing composite based on the dispersion of microcapsules with a healing agent in a polymeric matrix. The adhesion between the capsule and the matrix plays a very important role, since it is directly related with the load transfer to the microcapsule, and to its ability to release the healing agent [22, 49]. Another disadvantage is that once the healing agent has been consumed in one or multiple repairing events, the material loses its self-healing feature. This last disadvantage is one of the most important differences in comparison with intrinsic self-healing systems, as we will show in the next section.

Figure 1.7 SEM images of the scratch of (a) BS-55, (b) CS-55, (c) SH-55, (d) BS-60, (e) CS-60, and (f) SH-60 after 12 h of UV irradiation (wavelength: 310 nm; power: 582 W/m2). Reprinted with permission from Ref. [49]. Copyright (2019) American Chemical Society.