Cure Behavior and Mechanical Properties of Structural Self-Healing Epoxy Resins

Cure Behavior and Mechanical Properties of Structural Self-Healing

Epoxy Resins

LIBERATA GUADAGNO,1 PASQUALE LONGO,2 MARIALUIGIA RAIMONDO,1 CARLO NADDEO,1 ANNALUISA MARICONDA,2

ANDREA SORRENTINO,1 VITTORIA VITTORIA,1 GENEROSO IANNUZZO,3 SALVATORE RUSSO3

1Dipartimento di Ingegneria Chimica e Alimentare, Universita di Salerno, Via Ponte Don Melillo 84084 Fisciano, Salerno, Italy

2Dipartimento di Chimica, Universita di Salerno, Via Ponte Don Melillo 84084 Fisciano, Salerno, Italy

3ALENIA Aeronautica SpA, Viale dell’Aeronautica 80038 Pomigliano D’Arco, Napoli, Italy

Received 9 April 2010; revised 5 August 2010; accepted 5 August 2010

DOI: 10.1002/polb.22139

Published online in Wiley Online Library (wileyonlinelibrary.com).

ABSTRACT: Advances in the growing use of polymer compo-

sites in aerospace applications explore the possibility in the

development of smart materials capable of self-repair. In this

article, we have formulated and characterized a multifunctional

autonomically healing composite inspired by the design of

White et al. Microcapsules containing dicyclopentadiene

(DCPD) and powders of Grubbs first generation catalyst were

embedded in an epoxy formulation and both the epoxy precur-

sor and the composite were cured up to the temperature of 120�C that preserves the activity of the catalyst. The results on the

cure behavior of an epoxy matrix for self-healing material and

the influence of the components related to self-healing func-

tion on the dynamic-mechanical properties are investigated.

The presence of catalyst powder causes a slight decrease in

the elastic modulus value with respect to the epoxy matrix. At

variance, a large recovery in this parameter is gained for the

self-healing specimen, proving that well-distributed microcap-

sules contribute to improve the mechanical properties. VC 2010

Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 000:

000–000, 2010

KEYWORDS: aircraft materials; composites; curing of polymers;

epoxy matrix; FTIR; mechanical properties; self-healing

materials

INTRODUCTION In recent years, polymer composites areincreasingly used in structural applications, ranging fromcivil infrastructure to high performance vehicles such as rac-ing cars and military aircraft. This popularity is due to theirlower weight and to a continuous improvement of their per-formance aided in recent years by nanotechnology. However,limited storage stability and reliability are critical for poly-mer composites designed for structural applications.1–4 Oneof the main structural problems in these materials can arisefrom the effects of impact damage on the structural integrityof the material. A most common phenomenon is the impactof hail on the fuselage crown during a storm. The strongimpact of the stones on the keel during landing, plus the ac-cidental fall of heavy objects on the structure during theassembling phases and, in the worst case, the impact of birdsduring landing and take off, can lead to substantial matrixmicrocracking, delamination, and fiber–matrix debonding ofthe composite. This can reduce the structural capability andlead to premature failure if the damage is not detected andrepaired. Damage can also provide a route for the ingress ofwater and/or several contaminants into the structure. Waterseepage into the cracks can amplify the damage because anaircraft in flight is subjected to continued changing tempera-

ture. Internal damage is difficult to detect and, once devel-oped, even more difficult to repair, and therefore, the integ-rity of the structure is greatly compromised.

In addition to conventional methods for damage detection(such as nondestructive testing techniques; ultrasonic, infra-red thermography, X-ray tomography, and computerizedvibro thermography) and common repair methods (forexample, reinforced patch bonded or bolted to the compositestructure), there has recently been the development of effi-cient remendable materials5–8 and self-healing composites,which are expected to significantly extend the life of poly-meric components by autonomically healing microcrackswhenever and wherever they develop.9–23 For structuralapplications, such as aeronautic materials, the extrinsic self-healing systems are of interest because they would allow thedifficulties connected to damage diagnosis and the followingappropriate interventions to be reduced. In the case of ex-trinsic self-healing, the matrix resin itself is not healable.24,25

The healing agent, which partially restores the mechanicalproperties after damage, has to be stored inside compositematerials in advance. In accordance with the type of the ves-sel, there are two modes of repair activity: (a) self-healing interms of healant-loaded pipelines12–15,22,23 and (b) self-

Additional Supporting Information may be found in the online version of this article. Correspondence to: L. Guadagno (E-mail: [email protected])

Journal of Polymer Science: Part B: Polymer Physics, Vol. 000, 000–000 (2010) VC 2010 Wiley Periodicals, Inc.

SELF-HEALING EPOXY RESINS, GUADAGNO ET AL. 1

healing in terms of healant-loaded microcapsules.26,27 A veryinteresting system was developed by White et al.9,27 It con-sists of incorporating a microencapsulated healing agent andcatalytic chemical trigger within an epoxy matrix. Anapproaching crack ruptures embedded microcapsules releas-ing a healing agent into the crack plane through capillaryaction. Polymerization of the healing agent is triggered bycontact with the embedded catalyst, bonding the crack faces.

This system is a challenge for epoxy structural composites:however, some drawbacks have to be re-evaluated to be fullyapplied to advanced applications. These mainly regard thethermal stability of the Grubbs’ catalyst inside epoxy resinduring the curing cycle and the impossibility to utilize pri-mary amines as hardeners, as they can poison the catalyst.This last drawback forces the use of compounds that havenot been fully explored in literature as hardener agents,especially with regard to the cure behavior and mechanicalproperties of the manufactured composites.

Here, we have formulated and characterized a multifunc-tional autonomically healing composite containing solid par-ticles of Grubbs’ first generation catalyst and poly(urea-form-aldehyde) microcapsules filled with liquid dicyclopentadiene(DCPD). The validity of this system, as self-repair material,has extensively been described in literature.9,17,18,26,27

Our experimentation has shown that the catalytic activity ofGrubbs’ catalyst can be preserved up to temperaturesbetween 120 and 130 �C. Curing temperatures equal to orhigher than 150 �C deactivate the catalyst. In this article, wehave investigated the possibility of curing an epoxy resin upto 120 �C trying to obtain the best possible mechanical prop-erties for this system, using for the curing process a tertiaryamine (Ancamine K54) that shows little chemical interactionwith the catalyst used in the complete self-healing specimen.

The matrix of the self-healing system is a blend of epoxy res-ins (Epon 828 and Heloxy 71). Differential scanning calorim-eter (DSC) and DMA data, on the cure behavior of the epoxymatrix, have shown that the highest Tg and maximum curedegree (DC) were obtained at the ‘‘curing temperature’’ of 80�C. A subsequent increase in temperature up to 120 �C isnecessary to obtain the best mechanical properties. Infraredspectroscopy was used to follow the curing progress bydetermining the decrease of the band due to the epoxygroup. In light of FTIR data, the further improvement of themechanical properties has been attributed to the completereaction of the residual epoxy groups at 120 �C.

We have obtained promising results in view of applying theWhite system to epoxy resins for structural applications.

EXPERIMENTAL

MaterialsWhite’s design for self-healing material includes the presenceof microencapsulated monomers (healing agent) and catalystparticles in an epoxy matrix. The healing agent used in thiswork was DCPD. This monomer was encapsulated by in situpolymerization using an already reported procedure.9–28 The

catalyst used in the matrix was Grubbs’ catalyst [bis(tricyclo-hexylphosphine)benzylidine ruthenium (IV) dichloride]. Theepoxy matrix used was a blend of diglycidyl ether of bisphe-nol A (trade name EPON 828—Acronym DGEBA) and a high-molecular-weight epoxy flexibilizer, Dimer Acid DiglycidylEster (trade name HELOXY 71—Acronym DADGE), whichwas used in small percentages to improve the toughness ofthe material and consequently crack growth stability.

Resin 828 is an undiluted clear difunctional bisphenol A/epi-chloroydrin derived liquid epoxy resin with the followingtypical properties:

• Epoxide equivalent weight (grams of resins containing onegram-equivalent of Epoxide—ASTM D1652) ¼ 185–192

• Viscosity, at 25 �C, poise (kinematic viscosity—determina-tion of the viscosity of liquids by Ubbelohde Viscometer—ASTM Method D445) ¼ 110–150

• Density, lbs/gal, at 25 �C ¼ 9.7• Physical form (25 �C) ¼ clear liquid

Heloxy 71 is an undiluted low-viscosity liquid, aliphatic esterresin that imparts increased flexibility and resistance to ther-mal shock when blended with conventional liquid bisphenolA based epoxy resins. It also serves as a viscosity depressingagent. This epoxy functional material reacts into the resinsystem and, once reacted, becomes permanently bound intothe polymer network and does not migrate to the interfaceunder normal conditions. The typical properties measuredfor this component are as follows:

• Epoxide equivalent weight (grams of resins containing onegram-equivalent of Epoxide—ASTM D1652) ¼ 390–470

• Viscosity, at 25 �C, poise (kinematic viscosity—determina-tion of the viscosity of liquids by Ubbehlode Viscometer—ASTM Method D445) ¼ 4–9

• Density, lbs/gal, at 25 �C ¼ 8.2• Physical form (25 �C) ¼ clear amber liquid

Epon Resin 828 and Heloxy 71 modifier were purchasedfrom Aldrich.

The curing agent investigated for this study is an anionic ini-tiator Phenol, 2,4,6-tris[(dimethylamino) methyl] (Tradename Ancamine K54).

Figure 1 shows the chemical structures of compounds usedin this work.

Self-Healing System ManufacturingSpecimens tested in this work: Sample 1 was obtained bymixing Ancamine K54 with Epon 828 at the concentration of10:100 (by wt) hardener to resin Epon 828.

Sample 2 was obtained by mixing Ancamine K54 with Hel-oxy 71 at the concentration of 10:100 (by wt) hardener toresin Heloxy 71.

Sample EHA was obtained by mixing Epon 828 with Heloxy71 flexibilizer at the concentration of 63%:37% (by wt)epoxide to flexibilizer. Ancamine K54 was added at the con-centration of 10:100 (by wt) hardener to mixture (Epon 828and Heloxy 71). The samples were polymerized by a first

JOURNAL OF POLYMER SCIENCE: PART B: POLYMER PHYSICS DOI 10.1002/POLB

2 WILEYONLINELIBRARY.COM/JOURNAL/JPOLB

isothermal stage at different temperatures between 50 and100 �C. These samples are named EHA-[X �C,Y h], where X isthe temperature and Y is the time of the curing treatment.

Sample B was obtained by dispersing finely pulverizedGrubbs’ first catalyst (5 wt %) in sample EHA and polymer-izing this last mixture by a two-stage curing cycle: a first iso-thermal stage at 80 �C for 9 h and a second isothermal stagewith the higher temperature of 120 �C for 2 h.

Sample C was obtained by dispersing finely pulverizedGrubbs’ catalyst (5 wt %) and DCPD-filled microcapsules atthe concentration of 20 wt % into the mixture EHA beforethe ‘‘curing’’ treatment which was the same as Sample B.

Sample D was obtained as Sample B, utilizing for the poly-merization of the mixture two-different stage of the curingcycle: a first isothermal stage at 80 �C for 4 h and a secondisothermal stage with the higher temperature of 150 �C for2 h.

Sample E was obtained as Sample D, but with a differentcuring cycle, consisting in a first isothermal stage at 80 �Cfor 4 h and a second isothermal stage at 180 �C for 2 h.

MethodsThermal analysis was performed with a Mettler DSC 822 dif-ferential thermal analyzer in a flowing nitrogen atmosphere.The curing reaction of the epoxy matrix was performed indynamic regime, by heating in the DSC at a 10 �C/min from�60 �C to 300 �C, and in an isothermal regime, by heatingthe mixture at the chosen temperature, T* and keeping it atthis temperature for different times of treatment. After this,the partially cured resin was cooled to �60 �C at 50 �C/minand again heated up to 300 �C at 10 �C/min, to detect the

percentage of curing obtained in the isothermal regime andthe glass transition temperature.

Thermogravimetric analysis (TGA) was carried out with aMettler TGA/SDTA 851 thermobalance. Resin samples wereheated from 25 to 1000 �C at a 10�C/min heating rate underair and nitrogen flow. The weight loss was recorded as afunction of temperature.

Scanning electron microscopy (SEM) pictures were obtainedwith a LEO 1525 microscope. The samples were coveredwith a 250-Å-thick gold film using a sputter coater (Agarmod. 108 A).

The infrared spectra were obtained at different temperaturesby using a Bruker Vertex 70 FTIR-spectrophotometer with a2 cm�1 resolution (64 scans collected).

Dynamic-mechanical properties were performed using a Q800 dynamic mechanical thermal analyzer (TA Instruments).The spectra were recorded in the flexural mode, obtainingthe loss factor, tan d, at a frequency of 1 Hz, as a function oftemperature. The heating rate was 3 �C/min in the range of�100 to 200 �C.

RESULTS AND DISCUSSIONS

Cure Behavior of the Epoxy MatrixDifferential Scanning CalorimetryBecause of its sensitivity to any reaction involving heat flowchanges, DSC is especially useful for studying the cure of re-active epoxy systems whose curing is accompanied by theliberation of heat. A combination of dynamic and isothermalexperiments can provide information on reaction rates, curerates, and degree of cure.29–31

FIGURE 1 Chemical structures of compounds used for the epoxy matrix.

ARTICLE


In this work, DSC has been used for the estimation of the DCof Sample EHA under the assumption that the exothermicheat evolved during cure is proportional to the extent ofreaction. This assumption has already been adopted by otherauthors.32,33

The DC can be determined from the total heat of reaction,DHT, of the curing reaction and the residual heat of reaction,DHresid, of the partially polymerized material as follows:

DC ¼ DHT � DHresid

DHT� 100

To obtain fraction reacted at various temperatures, whichare very useful for cure cycle and processing studies, wehave performed a series of isothermal experiments.

To secure accurate total DH values from isothermal studies,dynamic runs were made after each isothermal cure toobtain the residual heat of reaction. The total heat of reac-tion was calculated as follow:

DHT ¼ DHiso þ DHresid

where DHiso and DHresid are the areas under the isothermaland dynamic thermograms, respectively.

As was previously discussed, the thermal history of polymercuring is a critical step in achieving an efficient self-healingsystem. Indeed, the thermal stability of the Grubbs’ catalystwas analyzed up to 150 �C to verify the activation of thering-opening metathesis polymerization of the healing agentafter the curing cycle. Thermal stability of the catalyst wasevaluated under conditions of actual use for applications inself-healing structural systems. The catalyst is dispersed inthe epoxy matrix in its solid state and both the epoxy pre-cursor and the composite were cured (with the curing agent)

choosing processing conditions that contemplate permanenceat high temperatures; after which, the catalysts wereextracted from the epoxy matrix and analyzed by 1H-NMRspectroscopy. Infrared spectroscopy was also used to evalu-ate the catalytic activity directly inside epoxy resin after thecure processing. The catalytic activity of Grubbs’ first genera-tion catalyst can be preserved up to temperatures between120 and 130 �C. Curing temperatures equal to or higherthan 150 �C deactivate the catalyst (see section ‘‘Control ofthe catalyst stability’’). Owing to the exposed considerations,temperatures up to 120 �C were selected for the isothermalcuring process of the epoxy matrix.

The preliminary calorimetric analysis indicated that, in theabsence of the curing agent (Ancamine K54), Epon 828 andHeloxy 71 were not reactive up to 300 �C. The results of cal-orimetric analysis, performed on the two formulations con-taining the tertiary amine named as Sample 1 and Sample 2in the experimental part, are reported in Figure 2.

The appearance of exothermic peaks in both the DSC scans(in dynamic regime) provides an indication of the consider-able reactivity of the ancamine K54 in the curing reactionsof the two resins (Epon and Heloxy).

DSC scans (in dynamic regime) for sample EHA are shown inFigure 3 where the cycle composed of three parts isreported: first scan (heating at 10 �C/min from �60 �C to300 �C), second scan (cooling at 50 �C/min from 300 �C to�60 �C), and third scan (second heating at 10 �C/min from�60 �C to 300 �C).

The DSC curve shows two overlapped exothermic peaksbetween 90 and 230 �C. A comparison of this DSC trace withcalorimetric curves relative to Sample 1 and Sample 2 of Figure2 clarifies the nature of the complex two exothermic peaks: thefirst (maximum at 120 �C) is ascribable to the reactions of Ank-amine with Epoxide rings of Epon 828 and, therefore, to the

FIGURE 2 Comparison of DSC curves for formulations: Sample

1, Epon 828/Ancamine K54 (10/1) and Sample 2, Heloxy 71/Anc-

amine K54 (10/1).

FIGURE 3 DSC scans (in dynamic regime) for sample EHA: first

scan (heating at 10 �C/min from �60 �C to 300 �C), second scan

(cooling at 50 �C/min from 300 �C to �60 �C), and third scan

(second heating at 10 �C/min from �60 �C to 300 �C).



homopolymerization of the epoxy resin; the second, to thehomopolymerization of the Heloxy flexiblizer resin.

The flat DSC curve of the third scan indicates that, indynamic regime, the curing reactions in the formulation areclosed at the end of the first scan, DHT (calculated for thefirst scan) is 329.44 J/g.

In Figure 4, dynamic runs of EHA mixture after each isother-mal cure of 9 h at the chosen temperatures (50, 60, 70, 80,100 �C) are shown. From Figure 4, a decrease in cure exo-therm, as isothermal curing temperature increases, is clear.

The DC for each curing temperature is reported in Table 1,where it is interesting to note that curing temperaturesabove 70 �C give cure degrees higher than 96%.

Dynamic Mechanical ResultsTo compare treatments and evaluate the improvement of thephysical properties after this first cycle of the curing process,dynamic mechanical measurements between �100 �C and160 �C at constant frequency were performed, and theresults are shown in Figure 5.

Dynamical mechanical data provide useful information onthe relaxation processes that become operative in the poly-mer in a temperature range depending on the examinedsystem.

In the dynamic mechanical spectra of epoxy resins, the maxi-mum of tan d is considered as the glass transition point andits temperature designated as (Tg)DMA.

Glass transition points measured by DMA of the samplescured at the chosen temperatures are reported in Table 2.

Figure 5 and Table 2 show that a curing temperaturebetween 60 �C and 80 �C gives the highest glass-transitiontemperature, whereas the sample cured at 100 �C shows thepeak of the main transition at 58 �C indicating the presenceof a significant fraction of the material with a lower transi-tion temperature. The presence of a weak secondary peak,active at higher temperature (�110 �C), suggests the pres-ence of a small fraction with lower mobility due to a differ-ent degree of crosslinking. It can be suggest that this behav-ior is due to the high reactivity of the catalytic curing agentat 100 �C. That causes a fast hardening in some small zonesmaking the thinness of the epoxy mixture unsuitable toobtain a homogeneous cross-linked structure throughout thesample.

DSC and DMA data have allowed the choice of the best cureconditions. The chosen temperature for the ‘‘curing’’ treat-ment of the matrix designed to manufacture the self-healing

FIGURE 4 Dynamic differential scanning calorimetry for resid-

ual cure after isothermal treatments of sample EHA. [Color fig-

ure can be viewed in the online issue, which is available at

wileyonlinelibrary.com.]

TABLE 1 Degree of Cure (DC) of EHA Sample after Each

Isothermal Cure of 9 h at the Chosen Temperatures (50, 60, 70,

80, and 100 8C)

Curing Temperature (�C) Curing Degree DC (%)

50 94

60 95

70 96

80 98

100 97

FIGURE 5 DMA plot of tan d versus temperature of EHA sam-

ple cured at temperatures between 50 and 100 �C. [Color figurecan be viewed in the online issue, which is available at


ARTICLE


system is 80 �C for which the highest Tg and maximum DCwere obtained.

To further improve the mechanical parameters of the samplecured at 80 �C, a post cure at a temperature of 120 �C wasmade obtaining sample EHA-[80 �C, 9 h þ120 �C, 2 h]. Forthe second step of the curing cycle, as previously reported, atemperature lower than 150 �C has been considered becausethe Grubbs’ catalyst had shown decomposition in air after atreatment of 150 �C for 2�3 h.

In Table 3, the elastic modulus of the sample EHA [80 �C, 9h þ120 �C, 2 h] is compared with that of sample EHA-[80�C, 9 h] in the temperature range between �50 and 50 �C.

Sample EHA [80 �C, 9 h þ 120 �C, 2 h] shows a higher mod-ulus for all the investigated temperatures. The increase ofthe elastic modulus is more intense at lower temperatures.As a matter of fact, at �50 �C, the modulus is 76% higherthan sample EHA-[80 �C, 9 h]; whereas, at 50 �C an increaseof 42% is observed. It is worth noting that this strongincrease of modulus does not correspond to a similar strongincrease in the glass transition point, as evident in Figure 6,where the temperature profiles of tan d for EHA-[80 �C, 9 hþ120 �C, 2 h] and EHA-[80 �C, 9 h] samples are compared.

Spectroscopic AnalysisInfrared spectroscopy was used to follow the curing progressof the epoxy matrix (EHA sample) by determining thedecrease of the band due to the epoxy group.

In Figure 7, we show the FTIR spectra in the interval 400–1800 cm�1 of the EHA sample as a function of the reactiontime during a two stage curing process composed of a treat-ment at 80 �C for 9 h and a following treatment at 120 �Cfor 2 h.

The most remarkable changes are recorded in the ether andepoxy ring regions. The characteristic response of ethers inthe IR is associated with the stretching vibration of theCAOAC system. In the initial spectrum EHA-[80 �C, 0 h], wecan note the vibrational bands of the aryl alkyl ethers(arylAOACH2A) which give two peaks; asymmetric CAOACstretch at 1247 cm�1 and a symmetric stretch at 1039 cm�1.These two bands are due to the structure of the epoxy resinprecursor34 (see Fig. 1 in the section ‘‘Experimental’’).

In the interval 1157–1065 cm�1, thermal treatment causes anoticeable increase of the absorbance due to the appearanceand growth of the band at 1120 cm�1, characteristic of ali-phatic ethers; confirming that the curing process of the resinlead to polyetherification as looked-for in epoxy resin curedwith a tertiary amine.

The most intense signal of the epoxy ring is the band cen-tered at 916 cm�1 attributed to asymmetrical ring stretchingin which the CAC bond is stretching during contraction ofthe CAO bond. In the initial spectrum (red curve), the bandat 916 cm�1 is very evident, and it decreases with increasingreaction time and temperature. It completely disappears fora treatment time of 2 h at 120 �C, indicating the time for thecomplete consumption of the epoxy rings, due to the homo-polymerization reaction.

In Figure 8, we show the absorbance (%) peak at 916 cm�1

as a function of the reaction time during the curing processcomposed of a treatment at 80 �C for 9 h (on the left of thedotted line) and a following treatment at 120 �C for 2 h (onthe right of the dotted line). The percentage in the absorb-ance value is related to intensity of the peak at 25 �C.

It is interesting to note that the intensity of the epoxy ringshows a rapid decrease at 80 �C in the first 60 min, followedby a slower decrease up to 180 min, after which a plateau is

TABLE 2 Glass Transition Points Measured by DMA of EHA

Sample Cured at Temperatures Between 50 and 100 8C

Curing Temperature (�C) Tg DMA (�C)

50 78

60 93

70 94

80 93

100 58/110

TABLE 3 The Elastic Modulus E0 (MPa) of EHA-[80 8C, 9 h], and

EHA-[80 8C, 9 h 1 120 8C, 2 h] Samples in the Temperature

Range between 250 and 50 8C

Temperature

(�C)EHA-[80 �C, 9h]Sample

EHA-[80 �C, 9h þ120 �C, 2h] Sample

�50 705 1240

�25 666 1160

0 615 1042

25 525 820

50 345 490

FIGURE 6 DMA plot of tan d for EHA-[80 �C, 9 h þ 120 �C, 2 h]

and EHA-[80 �C, 9 h] samples.



reached and, even increasing the time up to 540 min, no fur-ther consumption of the oxirane rings is observed. Anincrease of temperature up to 120 �C, however, is necessaryto cause the reaction of the residual epoxy groups. After 2 hat 120 �C, the band at 916 cm�1 completely disappears, indi-cating the time for the complete curing process, whichexplains the higher modulus for the sample cured up to 120�C, as reported above. The described curing cycle composedof two stages up to 120 �C was chosen to formulate our self-healing specimens. Before the manufacturing of the epoxyspecimen, a control of thermolytic stability and morphologyof the catalyst was carried out.

Catalyst CharacterizationControl of the Catalyst MorphologyAs previously reported from other authors,35 the crystalmorphology has a significant effect on the dissolutionkinetics of the DCPD/Grubbs’ catalyst system. Smaller par-ticles of Grubbs’ catalyst dissolve more quickly, leading to ahigh concentration of catalyst available for the healing agentand consequently for the metathesis reaction.

For the formulation of our self-healing systems, the Grubbs’catalyst, as-received from Sigma–Aldrich, was placed in a vialand pulverized by a small magnet in rotation for 30 min.

Observing the Figure 9 that shows the crystal morphologybefore and after the described treatment, it is possible toevaluate as the above described procedure gave catalystpowder with smaller dimensions.

Control of the Catalyst StabilityThe results on the stability of Grubbs’ catalyst are reportedin Table 4.

The thermolytic stability was evaluated in air or nitrogenatmosphere with the catalysts in powder form kept at differ-ent temperatures. The catalyst powder, after the thermal

FIGURE 7 FTIR spectra of EHA resin as a function of reaction

time during a two stage curing process at 80 and 120 �C.

FIGURE 8 The absorbance (%) peak at 916 cm�1 as a function

of the reaction time during a two stage curing process com-

posed of a treatment at 80 �C for 9 h (on the left of the dotted

line) and a following treatment at 120 �C for 2 h (on the right

of the dotted line) of the sample EHA.

FIGURE 9 SEM images of the Grubbs’ catalyst crystal before

(at the top) and after treatment (at the bottom).

ARTICLE


treatment, was dissolved in the deuterated solvent immedi-ately before the 1H-NMR spectra recording.

It is interesting to note that, from data reported in Table 4,Grubbs’ catalyst shows exceptional stability after a thermaltreatment also at 120 �C in air atmosphere for 3 h when thethermal treatment is directly done in the solid state.

The stability of the catalyst was subsequently also analyzedfor the powder embedded inside epoxy system (after thecuring process at high temperatures). In this case, the cata-lyst was extracted by chloroform from the epoxy matrix andanalyzed by 1H NMR spectroscopy. In particular, the epoxymatrices containing the catalyst powder were cured at hightemperatures between 80 and 180 �C. These last results areshown in Table 5.

Thus, Grubbs’ catalyst has shown to be extraordinarily toler-ant toward the different kinds of functional groups of the ep-oxy formulations up to 120 �C; therefore, it is possible tomanufacture self-healing systems for ring-opening metathesispolymerization.

For information on the 1H-NMR spectra see the SupportingInformation (Section 1.5).

Self-Healing SystemThe self-healing mixture was obtained as reported in ‘‘Exper-imental’’ by first dispersing pulverized Grubbs’ 1st genera-tion catalyst (5%) and then dispersing microcapsules con-taining DCPD (without selecting a dimensional range of size)into the epoxy mixture. This mixture was cured by a two-stage curing cycle: a first isothermal stage at 80 �C (9 h) anda second isothermal stage at 120 �C (2 h) obtaining Sample

C. For information on the microcapsule characterization seeSupporting Information (Section 1).

Control of the Catalyst Stability Inside Self-Healing SystemTo analyze the catalytic activity directly inside self-healingspecimen; Sample C was cut by a serrated blade; the powderwhich was produced from the sample was collected in amortar and a drop of healing agent (DCPD) was addedbefore dispersing the sample powder into the KBr disk forFTIR investigation.

Figure 10 shows FTIR spectrum of Sample C treated withDCPD. From the presence of the peak at 968 cm�1 character-istic of ring-opened poly(DCPD) we deduce that the catalystshows ability to polymerize the healing agent, providing evi-dence that the embedded catalyst is active also for a selfhealing sample cured up to 120 �C.

Thermogravimetric AnalysisThermal degradation in air of the EHA [80 �C, 9 h þ 120 �C,2h] sample (black curve) and B sample is shown in Figure11, where we can observe that the catalyst does not haveany effect on the recorded curve.

Thermal degradation in air of Sample C (black curve) isshown in Figure 12, where the thermogravimetric curve ofthe microcapsules (gray curve) is also reported forcomparison.

The oxidation of both samples substantially occurs in thetemperature range of 200–560 �C. The sudden decrease inweight around 220 �C of the microcapsules is due to theexplosion of microcapsules and rapid evaporation of DCPD.The self-healing specimen (Sample C) is stable up to 220 �Cwhere it begins to degrade.

Spectroscopic AnalysisInfrared spectroscopy was also used to follow the curingprogress of the self-healing system (Sample C) to analyze theeffects of the self-healing components (microcapsules andcatalyst) on the cure behavior.

TABLE 4 Summary of Results on the Thermolytic

Decomposition of Grubbs’ Catalyst

TestaTemperature

(�C)Time

(h)

Catalyst State

(After Treatment)

1b 120 15 Not decomposed

2c 120 3 Not decomposed

3c 150 1 Decomposed

4b 150 4 Not decomposed

a All the NMR analysis were performed dissolving the catalyst in tetra-

chlorodideuterioethane (TCDE).b This test was carried out keeping the sample under nitrogen

atmosphere.c These tests were carried out heating the sample in air.

TABLE 5 Summary of Results on the Thermolytic

Decomposition of Grubbs’ Catalyst Extracted from the

Formulated Epoxy Matrices

Sample

Curing Cycle

Temperature (Time)

Catalyst State

(After Treatment)

B 80 �C (9 h) þ 120 �C (2h) Not decomposed

C 80 �C (9 h) þ 120 �C (2h) Not decomposed

D 80 �C (4 h) þ 150 �C (2h) Partially decomposed

E 80 �C (4 h) þ 180 �C (2h) Decomposed

FIGURE 10 FTIR spectrum of Sample C powder treated with

DCPD.



In Figure 13, we show the FTIR spectra in the interval 400–1800 cm�1 of Sample C as a function of the reaction timeduring a two stage curing process composed of a treatmentat 80 �C for 9 h and a following treatment at 120 �C for 2 hfollowing the same procedure already used for the epoxymatrix (sample EHA).

In Figure 14, we show the absorbance (%) peak at 916cm�1 as a function of the reaction time during the curingprocess.

These last results highlight that the cure behavior is not sub-stantially different from the epoxy matrix (sample EHA).

Morphological Analysis (SEM)A cross-sectional view of our self-healing specimen (C sam-ple) is reported in the SEM images of Figure 15 whichshows that the microcapsules in the epoxy matrix are homo-geneously dispersed as evident from the absence ofagglomerations.

FIGURE 11 TG curves of [80 �C, 9 h þ 120 �C, 2 h] and B sam-

ples. [Color figure can be viewed in the online issue, which is

available at wileyonlinelibrary.com.]

FIGURE 12 TG curves of microcapsules and Sample C.

FIGURE 13 FTIR spectra of Sample C as a function of reaction

time during a two stage curing process at 80 and 120 �C. [Colorfigure can be viewed in the online issue, which is available at


FIGURE 14 The absorbance (%) peak at 916 cm�1 as a function

of the reaction time during a two stage curing process com-

posed of a treatment at 80 �C for 9 h (on the left of the dotted

line) and a following treatment at 120 �C for 2 h (on the right of

the dotted line) of the Sample C. [Color figure can be viewed in

the online issue, which is available at wileyonlinelibrary.com.]

ARTICLE


Dynamic Mechanical ResultsWe have tested a self-healing specimen to determine the me-chanical parameters (elastic modulus and stiffness) and thencompared this system with an epoxy matrix in two differentforms: one with and another without catalyst powder. Figure16(a) shows the loss factor, tan d, of the self-healing system(C sample) compared with EHA-[80 �C, 9 h þ 120 �C, 2 h]and B samples as a function of temperature. The dynamicmechanical spectra of the three samples show (Tg)DMA in thesame range of temperature. The ‘‘storage’’ modulus valuesare high for all the samples, but in particular for the self-healing sample (C sample), being greater than 1000 MPa upto 60–70 �C as could be seen in Figure 16(b). In Table 6, the

elastic modulus of sample EHA-[80 �C, 9 h þ120 �C, 2 h] iscompared with B and C samples in the temperature rangebetween �50 and 50 �C. It is interesting to note that, for thesame curing cycle, the presence of catalyst powder (SampleB) causes a slight decrease in the elastic modulus value withrespect to the epoxy matrix. It seems likely that in a way thecatalyst powder interrupts the cross-linking reactions locallydetermining a slight decrease in the elastic modulus. Thisdecrease could be also due to embrittlement effects causedby the catalyst powder. A large recovery in this performanceis gained for the whole self-healing formulation (C sample)at all the explored temperatures, proving that well-distrib-uted microcapsules, synthesized with a described procedure,contribute to improve the mechanical characteristics of thesample.

CONCLUSIONS

In this study, cure behavior of an epoxy matrix for a struc-tural self-healing system was examined. The epoxy matrix isa blend of epoxy resins and flexibilizer (Hepon 828 and Hel-oxy 71) and an anionic initiator (curing agent) which showsvery little chemical interaction with the Grubbs’ catalyst.

The mechanical properties of the epoxy matrix have beenmonitored and used to guide the selection of the most prom-ising ‘‘curing cycle’’ for the self-healing specimen. DSC andDMA results revealed that the highest Tg and maximum DCwas obtained at the ‘‘curing temperature’’ of 80 �C. It wasshown that the best mechanical properties in terms of glasstransition temperature and elastic modulus were obtainedby a two-stage curing cycle: a first isothermal stage at 80 �Cand a second isothermal stage at 120 �C; since, a treatment

FIGURE 15 SEM image of self-healing specimen section

(C sample).

FIGURE 16 (a) DMA plot of tan d for EHA [80 �C, 9 h þ 120 �C, 2 h], B and C samples and (b) Storage modulus of EHA [80 �C, 9 h

þ 120 �C, 2 h], B and C samples. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]



up to 120 �C is necessary to obtain the best mechanicalproperties.

Infrared spectroscopy was used to follow the curing progressby determining the decrease of the band due to the epoxygroup. In light of FTIR data, the further improvement of themechanical properties has been attributed to the completereaction of the residual epoxy groups at 120 �C.

We have prepared and tested a self-healing specimen con-taining a finely pulverized Grubbs’ first generation catalystand microcapsules filled with liquid DCPD; and then, ana-lyzed their influence on the cure behavior and dynamic me-chanical properties of the epoxy matrix.

It has been found that, for the same curing cycle, the pres-ence of catalyst powder causes a slight decrease in the elas-tic modulus value. A large recovery in this performance isgained for the self-healing specimen, proving that well-dis-tributed microcapsules contribute to improve the mechanicalcharacteristics of the self-healing sample.

Research is ongoing to refine the self repairing efficiency by theuse of Grubbs’ 2nd generation catalyst and Hoveyda-Grubbscatalysts which can give improved thermal stability in the stageof the ‘‘curing’’ process of the self-healing epoxy mixtures.

REFERENCES AND NOTES

1 Riefsnider, K. L.; Schulte, K.; Duke, J. C. ASTM Spec Tech

Publ 1983, 813, 136–159.

2 Osswald, T.; Menges, G. Material Science of Polymers for

Engineers; Hanser Publishers: Munich, 2003; pp 447–519.

3 Chamis, C. C.; Sullivan, T. L. In Situ Ply Strength: An Initial

Assessment; NASA Lewis Research Center: Cleveland, OH,

1987, p 19.

4 Wilson, D. J. K.; Wells, J. N.; Hay, D.; Owens, G. A.; Johnson,

F. 18th Int SAMPE Tech Conf 1986, 18, 242–253.

5 Chen, X.; Dam, M. A.; Ono, K.; Mal, A.; Shen, H.; Nutt, S. R.;

Sheran, K.; Wudl, F. Science 2002, 295, 1698–1702.

6 Stevens, M.; Jenkins, A. J Polym Sci Polym Chem Ed 1979,

17, 3675–3685.

7 Canary, S. A.; Stevens, M. P. J Polym Sci Part A: Polym

Chem 1992, 30, 1755–1760.

8 Kuramoto, N.; Hayashi, K.; Nagai, K. J Polym Sci Part A:

Polym Chem 1994, 32, 2501–2504.

9 White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore, J. S.; Kess-

ler, M. R.; Sriram, S. R.; Brown, E. N.; Viswanathan, S. Nature

2001, 409, 794–797.

10 Wu, D. Y.; Meure, S.; Solomon, D. Prog Polym Sci 2008, 33,

479–522.

11 Jud, K.; Kausch, H. H.; Williams, J. G. J Mater Sci 1981, 16,

204–210

12 Dry, C. Int J Mod Phys 1992, B6, 2763–2771.

13 Dry, C. Smart Mater Struct 1994, 3, 118–123.

14 Dry, C.; McMillan, W. Smart Mater Struct 1996, 5,

297–300.

15 Dry, C.; Sottos, N. Proc SPIE: Smart Mater Struct 1996,

1916, 438–444.

16 Dry, C.; McMillan, W. Smart Mater Struct 1997, 6, 35–39.

17 Kessler, M. R.; White, S. R. Compos A 2001, 32, 683–699.

18 Kessler, M. R.; Sottos, N. R.; White, S. R. Compos A 2003,

34, 743–753.

19 Cho, S. H.; Andersson, H. M.; White, S. R.; Sottos, N. R.;

Braun, P. V. Adv Mater 2006, 18, 997–1000.

20 Pang, J. W. C.; Bond, I. P. Compos A 2004, 36, 183–188.

21 Pang, J. W. C.; Jody, W. C.; Bond, I. P. Comp Sci Tech 2005,

65, 1791–1799.

22 Keller, M. W.; White, S. R.; Sottos, N. R. Adv Funct Mater

2007, 17, 2399–2404.

23 Toohey, K. S.; Hansen, C. J.; Lewis, J. A.; White, S. R.; Sot-

tos, N. R. Adv Funct Mater 2009, 19, 1399–1405.

24 Yuan, Y. C.; Yin, T.; Rong, M. Z.; Zhang, M. Q. eXPRESS

Polym Lett 2008, 2, 238–250.

25 van der Zwaag, S. Self Healing Materials: An Alternative

Approach to 20 Centuries of Materials Science, Vol. 100.

Springer: New York, 2007.

26 Jung, D.; Hegeman, A.; Sottos, N. R.; Geubelle, P. H.; White,

S. R. In Proceedings of the ASME International Mechanical En-

gineering Congress and Exposition, Dallas, USA, MD-80, 1997;

pp 265–275.

27 White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore, J. S.;

Sriram, S. R.; Kessler, M. R.; Brown, E. N. U.S. Pat. 6,858,659,

USA, 2005.

28 Brown, E. N.; Kessler, M. R.; Sottos, N. R.; White, S. R. J

Microencapsul 2003, 20, 719–730.

29 Maas, T. A. M. M. Polym Eng Sci 1978, 18, 29–32.

30 Pappalardo, L. T. Soc Plast Eng 1974, 20, 13–16.

31 Sanjana, Z. N.; Sampson, R. N. Insul/Circ 1981, 27, 87–92.

32 Horie, K.; Hiura, H.; Sawada, M.; Mita, I.; Kambe, H. J Polym

Sci Part A-1: Polym Chem 1970, 8, 1357–1372.

33 Edwards, G. D.; Ng, Q. Y. J Polym Sci [C] 1968, 21, 105–117.

34 Guadagno, L.; Vertuccio, L.; Sorrentino, A.; Raimondo, M.;

Naddeo, C.; Vittoria, V.; Iannuzzo, G.; Calvi, E.; Russo, S. Car-

bon 2009, 47, 2419–2430.

35 Jones, A. S.; Rule, J. D.; Moore, J. S.; White, S. R.; Sottos,

N. R. Chem Mater 2006, 18, 1312–1317.

TABLE 6 The Elastic Modulus E0 (MPa) of EHA 80 8C 9 h 1

120 8C 2 h, B and C Samples in the Temperature Range

between 250 and 50 8C

Temperature (�C)EHA 80 �C 9 h þ120 �C 2 h Sample B Sample C Sample

�50 1240 1120 1604

�25 1160 1062 1606

0 1042 980 1647

25 820 780 1574

50 490 410 1290

ARTICLE


Cure Behavior and Mechanical Properties of Structural Self-Healing Epoxy Resins

Documents

Transcript of Cure Behavior and Mechanical Properties of Structural Self-Healing Epoxy Resins