Supporting Information 1. Microcapsules characterization 1.1. Thermogravimetric analysis (TGA) Thermal degradation in air and nitrogen of the synthesized microcapsules is shown in figure S1. 100 Weight loss ( % ) 80 60 40 20 0 0 200 400 600 800 1000 Temperature ( C ) Figure S1. Thermogravimetric curves of microcapsules containing DCPD core in air (thick line) and nitrogen (thin line). Thermogravimetric analysis of microcapsules containing DCPD, performed either in air or under nitrogen atmosphere, show a first sudden weight loss around 200 C due to the explosion of microcapsules and rapid evaporation of DCPD. The thermal stability in air and nitrogen of the microcapsules is similar up to 280 C, and above this temperature, the degradation in air is faster with respect to nitrogen because of the effects of the oxidation. The degradation in air in the temperature range of 250 560 C occurs by fragmentation of the polymer structure shell (urea/formaldehyde) and elimination of ammonia, amines, CO 2
etc. The last two steps of degradation involve cyclisation reactions developing structures more thermally stable than the original ones as it results for thermal degradation of urea/formaldehyde polycondensate 1. In air the more thermally stable remaining structures degrade at about 500 C. 1.2. Thermal analysis The DSC thermogram of the microcapsules, figure 2S, shows two endothermic peaks; a small one between 50-120 C and a large peak centered at 260 C. Heat Flow exo -100-50 0 50 100 150 200 250 300 Temperature ( C ) Figure S2. DSC curve of microcapsules containing DCPD core. The first endothermic peak is due to impurities of water and free formaldehyde progressively eliminated during the dynamic thermal treatment between 50-120 C. In fact, formaldehyde is one of the components used for the synthesis of the microcapsule wall based on the reaction of urea with formaldehyde 1. The thermal degradation of ureaformaldehyde polycondensate was already studied in literature and it was found that below 100 C impurities of free CH 2 O and H 2 O are eliminated from the sample. The peak at higher temperatures, between 200 and 270 C corresponding to the first step of weight loss in the thermogravimetric curve, is due to the explosion of microcapsules and evaporation of
DCPD; and, although the boiling point of the healing agent is 170 C, the pressure of the microcapsule walls prevents the monomer evaporation. The synthesized microcapsules, submitted to a thermal treatment of 180 C to analyze thermal and chemical stability, remain undamaged and therefore very heat-proof. This is revealed through a morphological investigation by scanning electron microscopy. Degradation of the urea/formaldehyde shell starts at 200 C with increasing inside pressure of microcapsules. 1.3. Morphological analysis All the synthesized microcapsules were separated on sieves of different sizes. In table S1 the microcapsules size distribution is defined in terms of weight percentage corresponding to the different size ranges. Table S1. Microcapsules size distribution. Size ranges (μm) Percentage by weight (%) (%) 250 8.03 250-150 20.04 150-25 55.82 25 16.10 All the percentages corresponding to the different ranges analyzed by scanning electron micrograph (SEM) occurred without debris. An example of the type of work carried out is reported in the microcapsules fraction 25 μm. SEM images of a microcapsule fraction selected in the range 25 μm are reported at two different magnifications in figure S3.
Figure S3. Scanning electron microscope images of microcapsules selected in the range 25 μm. Microcapsules size distributions obtained from a series of 20 samples for the dimensional range 25 μm are reported in figure 4S. 20% Percentage [-] 16% 12% 8% 4% 0.5 1 1.25 1.5 1.75 2 3 3.5 4.5 5 5.5 6.5 7 7.5 8 9 10.5 11 12.5 14 16 18 19 21 22.5 25 0% Diameter [μm] Figure S4. Microcapsules size distributions for the dimensional range 25 μm
The photographs at two different magnifications in figure S3 show a spherical shape of microcapsules. All the microcapsules of different dimensional range turn out sufficiently robust to be incorporated into the epoxy formulation used without bursting, as well as it has been found during the manufacturing of the self-healing composite. 1.4. Efficiency of the microencapsulation process As shown by the FT/IR analysis reported in figure S5, the amount of DCPD encapsulated is a sufficient quantity to activate the ROMP reaction. For this investigation, all the microcapsule fractions were powdered in a mortar; a first small fraction was dried under vacuum and analyzed by FT/IR spectroscopy (red spectrum), while another fraction was treated with Grubbs catalyst powder and again analyzed by FT/IR (brown spectrum). In figure S5 FT/IR spectra of DCPD and microcapsules, as obtained from the synthesis, were also reported for comparison. The highlighted peak at 968 cm -1, present only in the brown spectrum, is characteristic of ring-opened poly(dcpd) providing evidence that the embedded DCPD is active in the metathesis reaction. Figure S5. FTIR spectra of microcapsules (blue spectrum), DCPD (green spectrum), of the microcapsules fraction dried under vacuum (red spectrum) and microcapsules fraction treated with Grubbs catalyst powder (brown spectrum).
1.5. Thermolytic stability of Grubbs catalyst as evaluated by extraction from the formulated epoxy matrices. The thermolytic stability of Grubbs catalyst was evaluated by 1 H NMR analysis of the extraction products with CHCl 3 on sample B,C,D and E. CHCl 3 solubilize both the active catalyst and its product of decomposition. In particular, as shown in Table 5, samples B and C (treated at 80 C for 9 hours and 120 C for two hours) give almost no decomposition, while the sample D (cured at 80 C for 4 hours and 150 C for 2 hours) give rise to about 40% of decomposition product. The sample E (treated at 80 C for 4 hours and 180 C for 2 hours) decomposes completely. Comparing the spectra showed in Figure S6 which report 1 H NMR of first generation Grubbs catalyst, of the sample B, of sample D and of sample E, it is possible to draw the data reported in table 5. Thus, samples B, C as well as D still contain catalyst so as to ensure, without a doubt, the self-healing. On the other hand, polymerization tests performed on the cured materials gave positive results, confirming definitively the goodness of treatment. Instead 1 H NMR sample E show that the Grubbs catalyst is completely decomposed.
Sample E Sample D Sample C Sample B Grubbs catalyst Figure S6 1 H-NMR of Grubbs first generation catalyst and samples B, C, D and E. REFERENCES 1. Camino G., Operti L., Costa L., Trossarelli L. Proceedings of the Seventh International Conference on Thermal Analysis. New York, 1982, pp. 1144-1149.