Self-healing of abrasion damage on epoxy resin controlled by ionic liquid

Abstract

A conventional epoxy resin thermoset has been modified with ionic liquid concentrations from 7 to 12 wt% to obtain softer, more ductile materials with self-healing ability for surface abrasion damage. The ionic liquid accelerates the initial stage of the epoxy resin curing process and reduces the glass transition temperature. Hardness values decrease with increasing ionic liquid content. Dynamic mechanical analysis shows that the addition of ionic liquid induces a reduction in storage modulus onset and maximum loss modulus values, and a displacement of the tan delta maximum values to lower temperatures with respect to neat epoxy resin. Although the instantaneous abrasion resistance under multiple scratching is lower for the epoxy resin–ionic liquid composites due to their lower hardness, the viscoelastic recovery and healing ability of the damaged surface with time shows that the self-healing ability increases with increasing ionic liquid content. The proposed mechanism is based upon the interactions between epoxy networks and ionic liquid molecules.

---

1. Introduction

Among the wide range of potential applications of room-temperature ionic liquids (ILs), the use of ILs to modify and control the properties of polymers, and their use in the development of new polymer-ILs hybrid materials, has been established as a new field of research attracting increasing attention.¹

Since the first report² on the hardening effect of a tetrafluoroborate imidazolim IL on epoxy resin, different research groups have focused their attention in such aspects as the use of ILs as catalysts of the crosslinking process,^3–9 as curing agents,^10–15 as dispersants of nanophases,^16–22 and as modifiers of thermal, rheological or mechanical properties.^23–40

As the poor tribological performance of epoxy resins is a major limitation in many structural applications, our research group has examined the effect of the addition of less than 3 wt% concentration of ILs²³ or IL-modified nanophases^18,22,25 on the improvement of the tribological performance of epoxy resin cured with a conventional amine hardener. In this line, we have very recently described the first self-healing effect of abrasion surface damage on an epoxy resin material by the addition of a relatively high (9 wt%) concentration of the IL 1-octyl-3-methylimidazolium tetrafluoroborate.³⁷

Immediately afterwards, Hameed et al.⁴¹ described the controlled mechanical behaviour of epoxy resin from brittle to ductile and to elastomeric behaviour, when high amounts, increasing from 10 to 60%, of the ionic liquid (IL) 1-butyl-3-methylimidazolium chloride are added. These authors have proposed the formation of electron donor–electron acceptor complexes between the hydroxyl groups of the epoxy chains with the cation and anion of the IL.

In order to study the influence of the IL 1-octyl-3-methylimidazolium tetrafluoroborate on the properties, abrasion resistance and self-healing ability of the epoxy resin, in the present work we describe new epoxy resin–ionic liquid materials with IL concentrations between 7 and 12 wt%.

2. Experimental

2.1. Materials and curing process

The prepolymer and hardener products were obtained from JEMG Gazachim Composites (Spain). Diglycidylether of bisphenol A (DGEBA) (Fig. S1a†) (molecular weight < 700) was mixed with 28 wt% hardener (mix of amines: 2-piperazin-1-ylethylamine; 3,6-diazooctane-1,8-diaminetriethylenetetramine; 3,6-dioxaoctamethylenediamine and 3-aminomethyl-3,5,5-trimethylcyclohexylamine). 1-Octyl-3-methylimidazolium tetrafluoroborate ionic liquid (IL) (>98% purity) (Fluka, Germany) (Fig. S1b†) was used as received. The IL was added to the prepolymer and mixed with an IKA Digital Ultra Turrax for 30 seconds at 1600 rpm, before adding the hardener (Fig. S1c†). In the case of epoxy resin containing a 12 wt% IL, the prepolymer + IL + hardener blend was stirred under vaccum at 1050 rpm for 1 minute, before the curing step. Epoxy resin (ER) and ER-IL nanocomposites were cured in a vacuum oven at 60 °C for 2 hour followed by a post-curing at room temperature for 24 hours.

2.2. Properties

Thermal analyses were carried out using a DSC 822e (Mettler Toledo), under a N₂ flow (50 ml min⁻¹) at a heating rate of 10 °C min⁻¹. Thermogravimetric measurements were performed with a TGA 1HT (Mettler Toledo) equipment (N₂ flow 50 ml min⁻¹; heating rate of 10 °C min⁻¹), in the temperature range from room temperature to 800 °C. A dynamic-mechanical analyzer (DMA) Q800 from TA Instruments was used to determine the viscoelastic properties of the materials, using 17 × 11 × 3.5 mm test coupons under the single cantilever configuration, at a heating rate of 3 °C min⁻¹ with a frequency of 1 Hz. Hardness (Shore D) values were determined with a TH210 hardness tester. Vickers hardness (HV) were measured with a Leco M-400-T1 microhardness tester.

2.3. Abrasion tests

Abrasion tests under multiple scratching were carried out at 25 °C and 50% HR with a Microtest Scratch tester (MTR 3/50-50/NI) with a diamond tip indenter (200 μm diameter; 120° cone angle) following the previously described²⁵ experimental procedure. 15 successive scratches along a 5 mm length were performed at 5 m min⁻¹ under a constant load of 5 N. For each scratch, friction coefficient and residual depth were determined. Surface topography and cross section areas of the wear tracks were measured with a Talysurf CLI 500 (Taylor Hobson) optical profiler. Optical micrographs were obtained with a Leica DMR optical microscope. SEM micrographs were obtained with a Hitachi S-3500N.

3. Results and discussion

3.1. Curing process

Table 1 and Fig. 1 show the influence of the IL on the curing of the epoxy resin. Neat epoxy resin shows a high residual enthalpy value after 20 minutes. The IL clearly reduces the residual enthalpy with respect to neat epoxy in this initial step. However, this effect of hardening acceleration is not observed at the intermediate step (40 minutes), where the IL increases the residual enthalpy values. Finally, after 60 minutes, the residual enthalpies for the ER–IL nanocomposites are again lower than those of the neat epoxy. It can be concluded that the IL acts as a hardener. The kinetic of the process for tetrafluoroborate alkylimidazolium ILs acting as the only hardener has been described by Maka et al.¹⁹ However, in the present case, the curing process is the result of the action of the conventional amine hardener and the IL. The presence of the amine hardener seems to control the process after the initial acceleration by the IL.

Table 1 DSC residual curing enthalpies and thermal properties

Material	DSC residual curing enthalpies ΔH (J g^−1)			Thermal properties
				Td (°C) (−50 wt%)	Tg (°C)
	20 min	40 min	60 min
ER	56.7	9.3	8.5	343.86	91.85
ER + 7% IL	33.3	19.3	1.6	345.22	70.43
ER + 9% IL	31.4	16.8	5.5	352.46	68.50
ER + 12% IL	38.8	14.4	2.0	341.29	58.25

---

Fig. 1 DSC curing enthalpies with time: (a) ER + 7% IL (inset showing the integration of the area below the curve to obtain ΔH); (b) ER + 9% IL; (c) ER + 12% IL.

3.2. Materials properties

Table 1 shows the decrease in glass transition temperature values as the IL proportion increases, in agreement with the plasticizing effect of the dispersed fluid IL phase. The composites show similar degradation temperatures to that of the neat epoxy, due to the high thermal stability of the neat IL, with a degradation temperature of 416 °C.

As expected, hardness values (Table S1†) decrease significantly with the addition of the IL fluid phase.

Dynamic mechanical analysis (Table S2;† Fig. 2a and b) confirms the plasticizing effect of the IL, as can be seen by the displacement of the onset of the storage modulus (E′), and the maximum values of the loss modulus (E′′) and loss factor (tan δ) to lower temperature values. The reductions in storage modulus with the addition of IL are in agreement with a more ductile behaviour of the ER–IL nanocomposites, while the slight reduction in tan δ could be related to an increase in crosslinking density.

Fig. 2 Dynamic mechanical properties: (a) storage modulus; (b) loss factor.

3.3. Abrasion under multiple scratching and self-healing ability

Fig. 3a and b show the evolution of friction coefficients (Fig. 3a) and residual depth (Fig. 3b) values with the number of scratches for all materials. The lubricating effect of the IL is only observed for the material with a 7 wt% proportion of IL, which reduces the friction coefficient in a 10% with respect to neat epoxy. Increasing IL content strongly increases friction coefficients, with the highest values obtained for the 9 wt% IL proportion. This friction behaviour could be explained by the surface damage measured from residual depth values (Fig. 3b). Again, the materials containing the highest IL proportion (9 and 12 wt%, respectively) show the highest residual depths, in agreement with their lower hardness (Table S1†). The lower residual depth increase with respect to the neat epoxy is observed for the lower IL proportion (7 wt%), in agreement with its higher hardness compared with the softer materials containing a higher concentration of IL. These lower surface damage favours the self-lubricating behaviour and corresponding friction reduction observed for ER + 7% IL. More severe surface damage increases friction coefficients.

Fig. 3 Effect of the number of scratches: (a) friction coefficient (COF); (b) residual depth.

3.4. Self-healing of abrasion damage with time

The evolution of the wear tracks and cross section profiles with time after the scratch tests are shown in Fig. S2–S4,† 4 and 5, respectively. The corresponding results for neat ER have been previously described.³⁷ The self-healing percentages (Fig. 6) have been quantified from the reduction in the cross section areas below and above the surface as shown in Fig. S5.† After 22 hours, the materials containing the highest IL proportion also show the highest self-repairing of the surface damage. In particular, a 96.2% reduction is reached for a 9 wt% IL (previously reported³⁷), and a 90.9% for a 12 wt%. A 7 wt% proportion of IL affords a 71.7% reduction after 22 hours, while the neat epoxy presents 21.2% for the same period of time, which is even lower than the value after 30 minutes (26.7%) (Fig. S6†). These observations confirm the active participation of the added fluid phase in the healing of the scratch abrasive grooves. The interactions between the cation–anion pairs of the IL and the polar groups of the polymer chains would reduce the brittle behaviour of the resin and the permanent damage produced by crack propagation under load.

Fig. 4 (a) Surface topography and (b) cross section profiles of the evolution of the wear track on ER + 7% IL.

Fig. 5 (a) Surface topography and (b) cross section profiles of the evolution of the wear track on ER + 12% IL.

Fig. 6 Healing percentages (±1%) (calculated as reduction of cross section areas) as a function of time and IL concentration.

3.5. Effect of porosity level

Fig. 7a and b show the fracture surfaces of ER + 9% IL and ER + 12% IL, respectively. In both cases, a homogeneous distribution of the IL, as seen by the fluorine element maps, is confirmed. A relevant observation is the higher number of pores in the case of ER + 9% IL. The porosity levels have been quantified by surface topography of cross sections of both materials (Fig. S7a and S7b†) and confirm the lower porosity [7.8 (±1) pores per mm²] of ER + 12% IL, with respect to ER + 9% IL [1.8 (±0.4) pores per mm²]. For the self-healing effect to take place, the ionic liquid phase should be present at the abrasion damaged surface region. The higher porosity level of ER + 9% IL could explain its higher self-healing percentage with respect to ER + 12% IL.

Fig. 7 SEM micrographs and EDX element maps of fracture surfaces: (a) ER + 9% IL; (b) ER + 12% IL.

The materials described here still present a brittle fracture behaviour³⁷ due to their relatively low IL content. This is in agreement with the results described by Hameed et al.⁴¹ for materials with IL concentrations below 20 wt%. For materials with higher IL contents, it could be expected a combination of ductility and self-healing of abrasion damage.

4. Conclusions

The effect of the addition of 1-octyl-3-methylimidazolium to a conventional epoxy resin has been studied. For increasing ionic liquid concentrations, the ionic liquid accelerates the curing process, particularly in its early stage, and reduces hardness and glass transition temperature, thus showing a softening and plasticizing effect. The storage modulus reduction and the shift of the loss factor to lower temperatures confirm the increasingly ductile behaviour with increasing ionic liquid content. The instantaneous surface damage under abrasion is higher for the softer epoxy resin–ionic liquid materials and increases with increasing ionic liquid content. Only the material with the lowest 7 wt% ionic liquid content presents a lower friction coefficient than the neat epoxy, due to its lower instantaneous surface damage. In contrast, the self-healing ability over time increases with increasing ionic liquid concentration, to reach values higher than a 90% after 22 hours, for the materials with 9 and 12 wt% ionic liquid. The molecular interactions between the polymer chain and the charged species present in the ionic liquid phase, the increased chain mobility and the curing ability of the ionic liquid could be responsible for the observed self-healing behaviour. A new family of thermoset materials with controlled ductility and abrasion self-healing could be tailored from different ionic liquid composition and concentrations.

Acknowledgements

The authors acknowledge the Ministerio de Economía y Competitividad (MINECO, Spain), the EU FEDER Program (Grant# MAT2014-55384 P), and the Fundación Séneca Agencia de Ciencia y Tecnología de la Región de Murcia for the “Ayuda a las Unidades y Grupos de Excelencia Científica de la Región de Murcia (Programa Séneca 2014)” (Grant# 19877/GERM/14), for financial support. N. Saurín is grateful to MINECO (Spain) for a FPI research (Grant # BES-2012-056621).

References

1. S. Livi, J. Duchet-Rumeau, J. F. Gerard and T. N. Pham, Macromol. Chem. Phys., 2015, 216, 359–368.
2. K. Kowalczyk and T. Spychaj, Polimery, 2003, 48, 833–835.
3. H. Maka, T. Spychaj and R. Pilawka, Ind. Eng. Chem. Res., 2012, 51, 5197–5206.
4. H. Maka and T. Spychaj, Polimery, 2012, 57, 456–462.
5. M. S. Fedoseev, M. S. Gruzdev and L. F. Derzhavinskaya, Int. J. Polym. Sci., 2014, 607341.
6. H. Maka, T. Spychaj and K. Kowalczyk, J. Appl. Polym. Sci., 2014, 131, 40401.
7. W. M. McDanel, M. G. Cowan, T. K. Carlisle, A. K. Swanson, R. D. Noble and D. L. Gin, Polymer, 2014, 55, 3305–3313.
8. H. Maka, T. Spychaj and J. Adamus, RSC Adv., 2015, 5, 82813–82821.
9. U. Arnold, C. Altesleben, S. Behrens, S. Essig, L. Lautenschutz, D. Schild and J. Sauer, Catal. Today, 2015, 246, 116–124.
10. M. A. M. Rahmathullah, A. Jeyarajasingam, B. Merritt, M. VanLandingham, S. H. McKnight and G. R. Palmese, Macromolecules, 2009, 42, 3219–3221.
11. B. C. Guo, J. J. Wan, Y. D. Lei and D. M. Jia, J. Phys. D: Appl. Phys., 2009, 42, 145307.
12. W. Li, L. X. Hou, Q. Zhou, L. Yan and L. S. Loo, Polym. Eng. Sci., 2013, 53, 2470–2477.
13. H. Maka, T. Spychaj and W. Sikorski, Int. J. Polym. Anal. Charact., 2014, 19, 682–692.
14. W. M. McDanel, M. G. Cowan, J. A. Barton, D. L. Gin and R. D. Noble, Ind. Eng. Chem. Res., 2015, 54, 4396–4406.
15. F. Lionetto, A. Timo and M. Frigione, Thermochim. Acta, 2015, 612, 70–78.
16. N. Hameed, N. V. Salim, T. L. Hanley, M. Sona, B. L. Fox and Q. P. Guo, Phys. Chem. Chem. Phys., 2013, 15, 11696–11703.
17. X. Zhang, Y. P. Zheng, R. L. Yang and H. C. Yang, Int. J. Polym. Sci., 2014, 712637.
18. N. Saurin, J. Sanes and M. D. Bermudez, Tribol. Lett., 2014, 56, 133–142.
19. H. Maka, T. Spychaj and R. Pilawka, eXPRESS Polym. Lett., 2014, 8, 723–732.
20. X. Y. Zheng, D. K. Li, C. Y. Feng and X. T. Chen, Thermochim. Acta, 2015, 618, 18–25.
21. H. Gholami, H. Arab, M. Mokhtarifar, M. Maghrebi and M. M. Baniadam, Mater. Des., 2016, 91, 180–185.
22. N. Saurín, J. Sanes and M. D. Bermúdez, New graphene/ionic liquid nanolubricants, Materials Today Proceedings, 2016, 3, S227–S232.
23. J. Sanes, F. J. Carrion-Vilches and M. D. Bermudez, e-Polym., 2007, 005.
24. K. Matsumoto and T. Endo, Macromolecules, 2008, 41, 6981–6986.
25. J. Sanes, F. J. Carrion and M. D. Bermudez, Wear, 2010, 268, 1295–1302.
26. B. G. Soares, S. Livi, J. Duchet-Rumeau and J. F. Gerard, Macromol. Mater. Eng., 2011, 296, 826–834.
27. B. G. Soares, S. Livi, J. Duchet-Rumeau and J. F. Gerard, Polymer, 2012, 53, 60–66.
28. L. X. Hou and Y. Liu, J. Appl. Polym. Sci., 2012, 126, 1572–1579.
29. X. F. Zhang, H. Y. Sun, C. Yang, K. Zhang, M. M. F. Yuen and S. H. Yang, RSC Adv., 2013, 3, 1916–1921.
30. K. Matsumoto and T. Endo, React. Funct. Polym., 2013, 73, 278–282.
31. A. A. Silva, S. Livi, D. B. Netto, B. G. Soares, J. Duchet and J. F. Gerard, Polymer, 2013, 54, 2123–2129.
32. S. Livi, A. A. Silva, Y. Thimont, T. K. L. Nguyen, B. G. Soares, J. F. Gerard and J. Duchet-Rumeau, RSC Adv., 2014, 4, 28099–28106.
33. B. G. Soares, A. A. Silva, S. Livi, J. Duchet-Rumeau and J. F. Gerard, J. Appl. Polym. Sci., 2014, 131, 39834.
34. T. K. L. Nguyen, S. Livi, S. Pruvost, B. G. Soares and J. Duchet-Rumeau, J. Polym. Sci., Part A: Polym. Chem., 2014, 52, 3463–3471.
35. N. Hameed, J. Bavishi, J. Parameswaranpillai, N. V. Salim, J. Joseph, G. Madras and B. L. Fox, RSC Adv., 2015, 5, 52832–52836.
36. B. G. Soares, A. A. Silva, J. Pereira and S. Livi, Macromol. Mater. Eng., 2015, 300, 312–319.
37. N. Saurín, J. Sanes and M. D. Bermúdez, Tribol. Lett., 2015, 58, 4.
38. E. B. Gienger, P. A. T. Nguyen, W. Chin, K. D. Behler, J. F. Snyder and E. D. Wetzel, J. Appl. Polym. Sci., 2015, 132, 42681.
39. H. Maka, T. Spychaj and M. Zenker, J. Ind. Eng. Chem., 2015, 31, 192–198.
40. P. L. Kuo, C. H. Tsao, C. H. Hsu, S. T. Chen and H. M. Hsu, J. Membr. Sci., 2016, 499, 462–469.
41. N. Hameed, N. V. Salim, T. R. Walsh, J. S. Wiggins, P. M. Ajayan and B. L. Fox, Chem. Commun., 2015, 51, 9903–9906.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra05503f

---