www.sciencemag.org/content/344/6180/186/suppl/DC1

Supplementary Material for

Toughening Elastomers with Sacrificial Bonds and Watching Them Break

Etienne Ducrot, Yulan Chen, Markus Bulters, Rint P. Sijbesma, Costantino Creton*

*Corresponding author. E-mail: costantino.creton@espci.fr

Published 11 April 2014, Science 344, 186 (2014)

DOI: 10.1126/science.1248494

This PDF file includes:

Materials and Methods

Supplementary Text

Figs. S1 to S7

Tables S1 to S3

Full Reference List Other Supplementary Material for this manuscript includes the following: (available at www.sciencemag.org/content/344/6180/186/suppl/DC1) Movies S1 and S2

2

Materials and Methods

Synthesis procedure for SN, DN and TN: Basic chemicals and equipment:

Unless otherwise specified, all starting materials were purchased from Aldrich or SDS and used with no further purification. Methyl acrylate (MA), ethyl acrylate (EA) and butanediol diacrylate (BDA) were filtrated through a column of basic alumina to remove the inhibitor. Every reactant was bubbled with nitrogen to remove any trace of oxygen and then stored in a freezer inside a glove box under nitrogen atmosphere. All reactions were performed under nitrogen atmosphere in a glove box unless otherwise specified. UV-initiated polymerizations were performed using a Vilbert-Lourmat UV-lamp (model VL-215.L) with a two side irradiation sample’s holder. The polymerizations were initiated by 2-hydroxyethyl-2-methylpropiophenone (HMP).

Composition of networks:

To determine the composition of the final DN and TN, we used the weight and the thickness of the sample. Before swelling, the thickness (hSN) and the weight (mSN) of the piece of first network was measured. The same characteristics were collected from the final double/triple network sample (h and m).The weight fraction ( ) of first network in the complete DN or TN was then calculated. The prestretch of chains of the first network ( ) was also determined using the thichnesses.

Eq. 1

Synthesis of first networks EAx:

First networks EAx were prepared by free radical polymerization of a solution of EA as monomer, BDA crosslinker and HMP. A 50 % solution in toluene of reactants was poured in a 1mm thick glass mold and exposed to the UV. The polymerizations were left to proceed for 1.5 h. The samples were then extracted from the mold and immersed in toluene/cyclohexane mixtures for a week to extract any unreacted species. Dialysis baths were changed every day. Swollen first networks were then dried under vacuum at 80 °C for a night. At this point we quantified the fraction of unreacted species to be less than 2 wt % of the dry sample. So called simple networks (SN) were finally stored at room temperature for later use.

3

Synthesis of double networks EAxMA:

Starting from first networks EAx, double networks were prepared following a swelling and polymerization sequence. A piece of first network was swollen in a bath composed of MA (40 g) as second monomer, BDA (4.37 µL, 0.01 mol % of monomer) as crosslinker and HMP (3.54 µL, 0.01 mol % of monomer) as UV initiator. Once swollen to equilibrium, the sample was carefully extracted from the monomer bath and placed between siliconized PET sheets and glass plates. The whole sample’s holder was then exposed to the UV for two hours to initiate and complete the polymerization. The double network was then recovered and dried under vacuum at 80 °C for a night. It was finally stored at room temperature until later use.

In the DN, a measure of the fraction of extractable components was performed by immersing the sample for two weeks in toluene. The loss of weight was quantified as less than 1 wt %.

Synthesis of triple networks EAxMAMA:

Triple networks were prepared following the same swelling/polymerization process but starting from a double network EAxMA. Pure poly(ethyl acrylate) networks:

The same strategy was implemented starting from EA0.5 first networks to go to double and triple networks made of pure poly(ethyl acrylate). Swelling baths were composed of EA (40 g) as monomer, BDA (7.52 µL, 0.01 mol % of monomer) and HMP (6.08 µL, 0.01 mol % of monomer) as UV initiator. The same swelling and polymerization sequence was followed to synthesize double and triple networks only composed of PEA. Second networks alone:

As a reference, solutions of monomers for the second and third networks preparations were polymerized between glass plates. Mixtures of monomer, crosslinker and UV initiator were poured in a glass mold and placed under UV for two hours.

After polymerization and a night at 80 °C under vacuum, the sample was stored at room temperature until later use. Those samples will be referred as MA2N and EA2N for respectively methyl or ethyl acrylate second networks.

4

Tensile tests:

Uniaxial tensile tests were performed on a standard tensile Instron machine, model 5565, equipped with an environmental chamber allowing a precise control of the temperature. We used a 100 N load cell and a video extensometer, model SVE, which can precisely follow the local displacement of markers. The relative uncertainty of the load cell and the video extensometer are respectively 0.1 % in the range 0 to 100 N and 0.11 % at the full scale of 120 mm.

Specimens were cut in a dumbbell shape using a normalized cutter with a central part of 20 mm in length (L0), 4 mm in width (w) and with a thickness (h) fixed by the sample itself between 0.6 mm and 2.5 mm.

Tensile tests were performed at a constant velocity of the crosshead of 500 µm.s-1. The force (F) and the local elongation in the tensile direction (λ), measured via the

video extensometer, were recorded all along the experiment. Nominal stress σN was defined as the tensile force per unit of initial area and true stress σT was calculated assuming constant volume deformation, thanks to the Eq. 2 and Eq. 3.

Cyclic extension:

Incremental loading and unloading cycles were performed with the same experimental setup as tensile tests at 500 µm.s-1 corresponding to a nominal strain rate of 0.025 s-1.

Cycles were applied between nearly λ=1 (F=0.1 N) to λi. Three cycles were performed with those limits, the maximum of λ was then incremented (λi+1 > λi) and three cycles were performed again. Here again, the force (F) and the strain ratio (λ) measured by video extensometer, were recorded continuously during the experiment.

To go further on the interpretation of the mechanical hysteresis, the energy dissipated under cyclic loading was calculated, from stress strain curves by integration. For every cycle from λ ~ 1 to , the dissipated energy was calculated from Eq. 4. Every potential contribution from viscous effects was removed by subtracting the hysteresis for the third cycle which is only due to viscoelastic dissipations. The cumulative dissipated energy which corresponds to the irreversible energy dissipation for a hypothetical cycle from λ = 1 to λn on a fresh sample was calculated according to Eq. 5.

Eq. 2

Eq. 3

5

Determination of the fracture energy Gc:

Fracture tests were performed using the classical single edge notch test on an Instron testing machine. A notch of 1 mm length was made in the middle of a strip of material, whose total width was 5 mm (Fig. S5).

The specimen was fixed between clamps previously spaced of l=20mm. The thickness depends on the sample itself between 0.6 mm and 2.5 mm. Samples were dotted with white paint to allow a measurement of the local elongation via the video extensometer. Force and elongation were measured while deforming the sample by moving the crosshead at 100 µm.s-1, corresponding to a nominal strain rate of 0.005 s-1

The fracture energy (GC) has been calculated using the methodology developed by Greensmith for elastomers which in essence introduces an empirical correction factor of

: to the linear solution to account for the nonlinear behavior of the rubber at large strain near the crack tip. The fracture toughness is then given by:

with c the length of the crack, λC the strain at break in single edge notch experiment

and W the strain energy density calculated by integration of the stress versus engineering strain of un-notched samples, until εc (εc=λc-1).

Mechanoluminescence experiments: Samples preparation

For mechanoluminescent samples the general synthesis procedure was followed, with a few adaptations. The crosslinker of the first network was changed to Bis(adamantyl)-1,2-dioxetane bisacrylate isomers (Fig S6). Exposure time to the UV was longer (one day) than our standard procedure to reach a complete conversion of the monomer. We observed that the kinetics of polymerization were slowed down by this special crosslinker. To avoid the thermal degradation of the crosslinker, every drying step was performed at room temperature under vacuum for one day and samples were stored in the freezer before testing.

Eq. 4

Eq. 5

Eq. 6

6

Data acquisition and analysis in cyclic extension

Tensile experiments were performed as presented previously at room temperature in the complete dark at 1 mm.s-1. Two dots of freshly exposed to the sun phosphorescent paint were made on the clamps just before the experiment.

The chemoluminescence signal was captured using a sensitive Andor Ixon Ultra EMCCD camera with a Nikon 60 mm lens. Pictures were taken at a rate of 47.8 Hz with an exposure time of 20 ms and a gain of 100.

Images were analysed with Matlab. The deformation of the sample was measured for every picture using the two phosphorescent dots as markers. The intensity was integrated on a rectangle that contains the whole sample as presented in Fig. S7A. Deformation λ and light intensity from the sample of DN and TN are presented in Fig S7 B-C as a function of time. Data acquisition and analysis in crack propagation experiment

Fracture experiments were performed as presented previously at room temperature in the complete dark at 500 µm.s-1. The sample’s geometry was 10 mm in width, 10 mm in length with a notch of 1 mm.

The chemoluminescence signal was captured using an Andor Ixon Ultra EMCCD camera with a Canon 100 mm Macro lens. Pictures were taken at a rate of 40 Hz with an exposure time of 2 ms and a gain of 300. Raw images were then analysed with Matlab.

Supplementary Text

Transfer reaction, additional crosslinks and links between networks Two samples of the second network of MA alone were prepared with (MA2N) or without (MA2N_noX) crosslinker everything else kept identical. Those two materials were tested in uniaxial extension, stress strain curves are presented on Fig. S1. There is not much difference between the two compositions. MA2N_noX shows nearly the same properties as those of MA2N. Using the Rubinstein and Panyukov equation(27) (Eq. 7), we estimated the contribution to the modulus from entanglements and from crosslinks for the two samples (Table S3).

Eq. 7

7

The best fits of the uniaxial elongation data to Eq. 7 give the same contribution to the modulus Gx for the two samples. For a sample prepared with no crosslinker, Gx should have been equal to zero and the resulting material would have been a polymer melt. This is evidence that there are transfer reactions during the polymerization which create additional crosslinks. Assuming rubber-like elasticity (Eq. 8) we estimated the real crosslinks concentration composed of BDA and transfer reaction of concentration respectively [BDA] and [Tr]. The concentration of BDA being set by the starting solution, [Tr] can be estimated for the two samples around 0.1 mol % of monomer. We have assumed here that Ex = 3Gx from incompressibility.

Eq. 8

8

Fig. S1. Transfer reaction during the second/third polymerizations Nominal Stress/Strain curves on poly(methyl acrylate) networks, second/third networks used to prepare double and triple networks. - In red, MA2N corresponds to a second network polymerized alone with a typical concentration of BDA as crosslinker ([BDA] = 0.01 mol% of monomer). - In blue, MA2N_noX is a second network alone, prepared with no crosslinker ([BDA] = 0 mol% of monomer). The similar behavior reveals the existence of transfer reactions that increase the number of crosslinks.

9

Fig. S2. Prestretching and dilution of the first network A. Prestretch of chains of the first network λPrestretch the first network in DN and TN. The reference of prestretching being taken for the first network alone is in the dry state. B. Weight fraction of the first network in DN and TN versus the concentration of crosslinker in the first network

A

B

10

Fig. S3. Various monomers for the first network Nominal Stress/Strain curves on simple and double networks, at T = 60°C. Changing the monomer of the first network while keeping the weight between crosslinks constant at 3900 g.mol-1. First networks were prepared from Methyl Acrylate (MA), Ethyl Acrylate (EA) and Butyl Acrylate (BA), while the second network of methyl acrylate was kept identical. A. First networks B. Corresponding double networks

A

B

11

Fig. S4. Damage on multiple networks under cyclic loading A. Residual elongation λmin after unloading at 0.1N versus previously achieved λmin, in cyclic extension experiments for DN and TN samples B. Evolution of the loading modulus measured during cycle i normalized by the initial modulus of the fresh sample as a function of maximum elongation previously achieved by the sample

A

B

12

Fig. S5. Geometry of samples for single edge notched test

13

OO

O O

O O OO

Fig. S6. Mechanoluminescent crosslinker (BADOBA) One isomer of Bis(adamantyl)-1,2-dioxetane bisacrylate

14

Fig. S7. Mechanoluminescence in cyclic tests A. Image analysis to measure the deformation of the samples with the phosphorescent dots (circles) and integration zone for the intensity of luminescence B. Luminescence intensity and elongation versus time for DN whose first network was crosslinked with the mechanoluminescent crosslinker C. Luminescence intensity and elongation versus time for TN whose first network was crosslinked with the mechanoluminescent crosslinker

A

B

C

15

Table S1. Formulation of EAX, PEA first networks prepared in solvent (a)[BDA]=100.nBDA/nmonomer, with nmonomer and nBDA the number of moles of monomer and BDA respectively, (b) Mc

th the theoretical weight between crosslinks as Mcth = Mmonomerx

nmonomer/(2xnBDA), with Mmonomer the molar mass of the monomer

Sample (a)[BDA]

(mol %) (b)Mc

th

(g/mol) EA (g)

Toluene (g)

BDA (µL)

HMP (µL)

EA05 1.45 3400 8.6 8.6 235.8 152.5

EA1 2.81 1700 8.6 8.6 471.5 152.5

EA2 5.81 860 8.6 8.6 943 152.5

16

Table S2. Compositions of DN and TN A. Compositions of double networks EAxMA, and of triple networks EAxMAMA B. Compositions of DN and TN of pure poly(ethyl acrylate)

Sample First network

EA05MA EA05 1.7 20 %

EA1MA EA1 1.6 30 %

EA2MA EA2 1.3 35 %

EA05MAMA EA05 2.5 6 %

EA1MAMA EA1 2.2 10 %

Sample First network

EA0.5EA EA0.5 1.6 20%

EA0.5EAEA EA0.5 2.5 5%

A

B

17

Table S3. Second networks; chemical crosslinker and transfer reactions Composition and Properties of second networks of PMA: MA2N and MA2N_NOX at 60 °C, with Mmono=86.09 g.mol-1, ρ=1.12 g.cm-3 E elastic modulus measured under small strain uniaxial elongation; Ee Ex contribution to the modulus from entanglements and crosslinks respectively; [Tr] equivalent concentration of transfer reactions

Sample [BDA] E Ee Ex [Bda]+[Tr] [Tr]

MA2N 0.01 mol % 0.9 MPa 0.78 MPa 0.25 MPa 0.11 mol % 0.1 mol %

MA2N_noX 0 mol % 0.96 MPa 0.85 MPa 0.25 MPa 0.11 mol % 0.11 mol %

18

Movie S1: Cyclic extension on a triple network where the first network is crosslinked with the mechanoluminescent crosslinker. Video accelerated five times compared to real speed. A strip of TN with a first network crosslinked with the mechanoluminescent crosslinker is stretched in the vertical direction in the dark. Initial dimensions of the sample are l = 10 mm, w = 5 mm, t = 2 mm. Blue dots are phosphorescent markers on the two clamps to visualize the position of the clamps at every moment. During the first three cycles, no light is emitted, corresponding to the perfect elasticity of the sample. Then the sample starts to glow and stops immediately when the strain decreases. For the two following cycles only a very brief flash of light is emitted at the time where the maximum elongation of the previous cycle is reached and then surpassed. The same behavior is reproduced for higher λmax until final fracture of the samples.

Movie S2: Crack propagation in a triple network using mechanoluminescence Video slowed down thirteen times compared to the real speed. A notched strip of TN is stretched vertically, the notch is to the right of the image and the crack propagates from right to left. Images corresponding to the beginning of the experiment are not shown here. Before the propagation of the crack, the sample starts to emit light at the crack tip showing some bond breaking before the crack moves. Then the crack propagates and a large zone in front of the crack tip emits a bright light. A lot of chains are broken far from the tip. The vertical line is an artifact due to the way the images are acquired when the camera is saturated. Spots of light very far from the crack are reflections at the surface of the sample.

References and Notes 1. G. J. Lake, P. B. Lindley, The mechanical fatigue limit for rubber. J. Appl. Polym. Sci. 9,

1233–1251 (1965). doi:10.1002/app.1965.070090405

2. G. J. Lake, A. G. Thomas, The strength of highly elastic materials. Proc. R. Soc. Lond. A Math. Phys. Sci. 300, 108–119 (1967). doi:10.1098/rspa.1967.0160

3. A. N. Gent, Adhesion and strength of viscoelastic solids. Is there a relationship between adhesion and bulk properties? Langmuir 12, 4492–4496 (1996). doi:10.1021/la950887q

4. G. Heinrich, M. Kluppel, T. A. Vilgis, Reinforcement of elastomers. Curr. Opin. Solid St. M. 6, 195–203 (2002). doi:10.1016/S1359-0286(02)00030-X

5. B. N. J. Persson, O. Albohr, G. Heinrich, H. Ueba, Crack propagation in rubber-like materials. J. Phys. Condens. Matter 17, R1071–R1142 (2005). doi:10.1088/0953-8984/17/44/R01

6. B. D. Viers, J. E. Mark, Elastomeric properties of polysiloxane networks. Bimodal elastomers that are spatially inhomogeneous and others that are very broadly multimodal. J. Macromol. Sci. Part A Pure Appl. Chem. 44, 131–138 (2007). doi:10.1080/10601320601030657

7. G. D. Genesky, C. Cohen, Toughness and fracture energy of PDMS bimodal and trimodal networks with widely separated precursor molar masses. Polymer (Guildf.) 51, 4152–4159 (2010). doi:10.1016/j.polymer.2010.06.054

8. G. D. Genesky, B. M. Aguilera-Mercado, D. M. Bhawe, F. A. Escobedo, C. Cohen, Experiments and simulations: Enhanced mechanical properties of end-linked bimodal elastomers. Macromolecules 41, 8231–8241 (2008). doi:10.1021/ma801065x

9. N. K. Singh, A. J. Lesser, A physical and mechanical study of prestressed competitive double network thermoplastic elastomers. Macromolecules 44, 1480–1490 (2011). doi:10.1021/ma1028054

10. G. E. Dieter, Mechanical Metallurgy (McGraw Hill, New York, ed. 2, 1976).

11. H. R. Brown, A molecular interpretation of the toughness of glassy polymers. Macromolecules 24, 2752–2756 (1991). doi:10.1021/ma00010a018

12. E. J. Kramer, L. L. Berger, Fundamental processes of craze growth and fracture. Adv. Polym. Sci. 91, 1–68 (1990). doi:10.1007/BFb0018018

13. J. P. Gong, Y. Katsuyama, T. Kurokawa, Y. Osada, Double-network hydrogels with extremely high mechanical strength. Adv. Mater. 15, 1155–1158 (2003). doi:10.1002/adma.200304907

14. Y. Tanaka, R. Kuwabara, Y. H. Na, T. Kurokawa, J. P. Gong, Y. Osada, Determination of fracture energy of high strength double network hydrogels. J. Phys. Chem. B 109, 11559–11562 (2005). doi:10.1021/jp0500790 Medline

15. R. E. Webber, C. Creton, H. R. Brown, J. P. Gong, Large strain hysteresis and mullins effect of tough double-network hydrogels. Macromolecules 40, 2919–2927 (2007). doi:10.1021/ma062924y

16. J.-Y. Sun, X. Zhao, W. R. K. Illeperuma, O. Chaudhuri, K. H. Oh, D. J. Mooney, J. J. Vlassak, Z. Suo, Highly stretchable and tough hydrogels. Nature 489, 133–136 (2012). doi:10.1038/nature11409 Medline

17. Y. H. Na, Y. Tanaka, Y. Kawauchi, H. Furukawa, T. Sumiyoshi, J. P. Gong, Y. Osada, Necking phenomenon of double-network gels. Macromolecules 39, 4641–4645 (2006). doi:10.1021/ma060568d

18. H. W. Greensmith, Rupture of rubber. X. The change in stored energy on making a small cut in a test piece held in simple extension. J. Appl. Polym. Sci. 7, 993–1002 (1963). doi:10.1002/app.1963.070070316

19. K. Sakulkaew, A. G. Thomas, J. J. C. Busfield, The effect of the rate of strain on tearing in rubber. Polym. Test. 30, 163–172 (2011). doi:10.1016/j.polymertesting.2010.11.014

20. Q. M. Yu, Y. Tanaka, H. Furukawa, T. Kurokawa, J. P. Gong, Direct observation of damage zone around crack tips in double-network gels. Macromolecules 42, 3852–3855 (2009). doi:10.1021/ma900622s

21. T. Nakajima, T. Kurokawa, S. Ahmed, W.- Wu, J. P. Gong, Characterization of internal fracture process of double network hydrogels under uniaxial elongation. Soft Matter 9, 1955–1966 (2013). doi:10.1039/c2sm27232f

22. Y. Chen, A. J. H. Spiering, S. Karthikeyan, G. W. M. Peters, E. W. Meijer, R. P. Sijbesma, Mechanically induced chemiluminescence from polymers incorporating a 1,2-dioxetane unit in the main chain. Nat. Chem. 4, 559–562 (2012). doi:10.1038/nchem.1358 Medline

23. P. Lechtken, Anwendung nicht-isothermer Kinetik auf chemilumineszente Systeme. Thermolyse von 1,2-Dioxetanen. Chem. Ber. 109, 2862–2870 (1976). doi:10.1002/cber.19761090818

24. Y. Tanaka, A local damage model for anomalous high toughness of double-network gels. Europhys. Lett. 78, 56005 (2007). doi:10.1209/0295-5075/78/56005

25. H. R. Brown, A model of the fracture of double network gels. Macromolecules 40, 3815–3818 (2007). doi:10.1021/ma062642y

26. X. Wang, W. Hong, Pseudo-elasticity of a double network gel. Soft Matter 7, 8576–8581 (2011). doi:10.1039/c1sm05787a

27. M. Rubinstein, S. Panyukov, Elasticity of polymer networks. Macromolecules 35, 6670–6686 (2002). doi:10.1021/ma0203849