thermal and mechanical properties of a dendritic hydroxyl

Thermal and Mechanical Properties of a Dendritic Hydroxyl-Functional

Hyperbranched Polymer and Tetrafunctional Epoxy Resin Blends

JIN ZHANG, QIPENG GUO, BRONWYN FOX

Centre for Material and Fibre Innovation, Deakin University, Waurn Ponds Campus, Geelong, Victoria 3217, Australia

Received 12 August 2009; revised 20 September 2009; accepted 4 November 2009

DOI: 10.1002/polb.21902

Published online in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: Blends of a tetrafunctional epoxy resin, tetragly-

cidyl-4,40-diaminodiphenylmethane (TGDDM), and a hydroxyl-

functionalized hyperbranched polymer (HBP), aliphatic hyper-

branched polyester Boltorn H40, were prepared using 3,30-dia-

minodiphenyl sulfone (DDS) as curing agent. The phase

behavior and morphology of the DDS-cured epoxy/HBP blends

with HBP content up to 30 phr were investigated by differential

scanning calorimetry (DSC), dynamic mechanical analysis

(DMA), and scanning electron microscopy (SEM). The phase

behavior and morphology of the DDS-cured epoxy/HBP blends

were observed to be dependent on the blend composition.

Blends with HBP content from 10 to 30 phr, show a particulate

morphology where discrete HBP-rich particles are dispersed in

the continuous cured epoxy-rich matrix. The cured blends with

15 and 20 phr exhibit a bimodal particle size distribution

whereas the cured blend with 30 phr HBP demonstrates a

monomodal particle size distribution. Mechanical measure-

ments show that at a concentration range of 0–30 phr addition,

the HBP is able to almost double the fracture toughness of the

unmodified TGDDM epoxy resin. FTIR displays the formation

of hydrogen bonding between the epoxy network and the HBP

modifier. VC 2010 Wiley Periodicals, Inc. J Polym Sci Part B:

Polym Phys 48: 417–424, 2010

KEYWORDS: blends; hyperbranched; modification; phase sepa-

ration; structure-property relations

INTRODUCTION Hyperbranched polymers (HBPs) have at-tracted much attention over the past 20 years due to theirunique chemical and physical properties. Their unique three-dimensional dentritic architecture leads to significant differ-ences in properties from linear polymers. The potentialapplications of HBPs cover versatile areas including conju-gated functional materials, polymer electrolytes, supramolec-ular chemistry, coatings, nanomaterials, modifiers and addi-tives.1,2 One of the most important properties of HBPs istheir low melt viscosities even for high molecular weights,resulting from a much lower degree of chain entanglementrelative to linear polymers.3 This characteristic has particularsignificance when HBPs are used as additives to modifyinherently brittle thermosets. Conventional modifiers such asrubber4,5 and thermoplastic6–9 toughening agents are able toenhance the fracture toughness of unmodified thermosets,however, they tend to cause significant increase in the vis-cosity of the blends with only a relatively small percentageaddition. It is well-known that liquid molding technologiesare economical alternatives to expensive autoclave process-ing of polymer matrix composites; viscosity and resin floware critical issues for obtaining good quality composite com-ponents. As a result, HBPs are promising candidates formodifying thermosets with minimal decrease in processabil-ity of the modified resins. A number of studies have reportedthe use of hydroxyl and epoxy functionalized HBPs to modifydifferent epoxy resin formulations.

Both hydroxyl and epoxy functionalized HBPs have beenshown to significantly increase the toughness of less-reactiveand lower crosslink-density difunctional epoxy resin systemscontaining the diglycidyl ether of bisphenol A (DGEBA). Twohydroxyl functionalized HBPs, Boltorn H30 and H40, werestudied for their modification result on a DGEBA epoxy(Epon 828) cured with 3,30 diamino diphenyl sulfone(DDS).10 The stress intensity factor KC for fracture toughnesstests increased from 0.93 MPa

ffiffiffiffim

pto 1.56 MPa

ffiffiffiffim

pwith 30

wt % H30 and to 1.67 MPaffiffiffiffim

pwith 30 wt % H40. Two

other works also showed similar trends when a fifth genera-tion hydroxyl functionalized HBP, Boltorn G5,11 and an epoxi-dized HBP Boltorn E112 were used as additives in DGEBAepoxy resins. For the more reactive trifunctional epoxies,HBPs have also demonstrated an excellent potential for useas toughening agents. It was reported that the HBP BoltornG513 and E114,15 modified triglycidyl p-amino phenol (TGAP)(MY0510, Ciba Speciality Chemicals) epoxy resins had 50%higher impact strength than unmodified epoxy resin with 20wt % content of Boltorn G5 and 100% higher impactstrength than unmodified epoxy resin with 20 wt % contentof Boltorn E1. Nevertheless, the modification effect of HBPson the highly reactive and highly crosslinkable tetrafunc-tional epoxy resins have not been shown to be desirable.HBP Boltorn E1 was added into an epoxy resin based on tet-raglycidyl methylene dianiline (TGDDM) (Araldite MY721,Ciba) and a cycloaliphatic diamine hardener.16 Decreased KIC

Correspondence to: J. Zhang (E-mail: [email protected])

Journal of Polymer Science: Part B: Polymer Physics, Vol. 48, 417–424 (2010) VC 2010 Wiley Periodicals, Inc.

POLYMER AND TETRAFUNCTIONAL EPOXY RESIN BLENDS, ZHANG, GUO, AND FOX 417

value was believed to be caused by the lack of plastic defor-mation of highly crosslinked epoxy network.

High service temperature and good chemical and water resist-ance provide tetrafunctional epoxies with great advantages inthe application of aerospace polymer matrix composites; in fact,they have been used as matrix materials for > 90% compo-sites.17 Modification of the brittleness of these highly crosslinkedepoxies is crucial in material development aiming at reducingthe delamination tendency of composite structures. In thiswork, a tetrafunctional epoxy resin was modified by a fourthgeneration hydroxyl functionalized HBP. Here we report theresults of an investigation of the thermomechanical property,phase behavior, morphology, and fracture toughness of theseblends. The tetrafunctional epoxy resin is tetraglycidyl methyl-ene dianiline (TGDDM) (Araldite MY720, Huntsman), and thehydroxyl functionalized HBP is an aliphatic hyperbranched poly-ester, Boltorn H40. Boltorn H40 is a commercial product of Per-storp Speciality Chemicals (Sweden).15 The cure agent is 3,30-diaminodiphenyl sulfone (DDS). The chemical structures ofTGDDM, DDS and Boltorn H40 are shown as follows:

The structure of Boltorn H40 is a perfect dendrimer analogwhich has 64 hydroxyl end groups per molecule. Thesehydroxyl groups provide a potential for hydrogen bondinginteraction with the DDS-cured epoxies and thus enhance itsmiscibility with the epoxy matrix. This work has shown thateffective toughening can be achieved by the addition of Bol-torn H40 into TGDDM epoxy resin; a twofold increase infracture toughness was found for the epoxy/HBP blend withHBP content of 30 part per 100 (phr) relative to unmodifiedTGDDM epoxy resin.

EXPERIMENTAL

Material and Sample PreparationThe uncured epoxy resin with epoxide equivalent weight117–134 was kindly donated by Huntsman Australia; thecure agent 3,30-diaminodiphenyl sulfone (DDS) was pur-chased from Sigma Aldrich (purity > 97%). Boltorn H40 waskindly donated by Perstorp Speciality Chemicals, Sweden. Itis a fourth-generation of hyperbranched polyester with theo-retically 64 primary hydroxyl groups and a molecular weightof 7316 g/mol. It is an amorphous solid at room tempera-ture with a reported glass transition temperature of 40 �C,as measured by DSC.

HBP was dissolved in acetone (0.79 g/mL) firstly; the solu-tion was then added into the epoxy which was heated andstirred at 60 �C, resulting in a homogenous solution. Thecure agent was added to this solution with continuous stir-ring. By removing the solvent under reduced pressure andthen under high vacuum at elevated temperature (50 �C to70 �C) for 100 min, a homogenous blend was obtained. Theepoxy/HBP blends with 0, 5, 10, 15, 20, and 30 phr HBPwere cast in a mold preheated at 120 �C and cured at 180�C for 3 h. The blends were further postcured at 220 �C for2 h in a vacuum oven.

Differential Scanning CalorimetryDSC was performed on the cured neat epoxy resin and ep-oxy/HBP blends by using a TA Q200 differential scanningcalorimeter under constant flow of nitrogen. Glass transitiontemperatures were determined from the midpoint of theslope change of the heat capacity plot of the second scan. Aheating rate of 10 �C/min was used.

Dynamic Mechanical AnalysisDynamic mechanical measurements were carried out for thecured neat epoxy resin and epoxy/HBP blends by using a TAQ800 analyzer at a fixed frequency of 1 Hz with 3 �C/minheating rate. A single cantilever mode was adopted for test-ing samples with dimensions of 32 mm � 12 mm � 4 mm.

Fracture Toughness MeasurementsFracture toughness tests were performed by using the sin-gle-edge-notch bending (SENB) method according to ASTM D5045-99 using a MTS 385 KN universal tester. The sampledimension was 60 mm � 14 mm � 6.5 mm. A cross-headspeed of 10 mm/min was used for testing. The critical-stress-intensity factor, KIC, was calculated by the followingequation:

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

418 INTERSCIENCE.WILEY.COM/JOURNAL/JPOLB

KIC ¼ p

BW1=2

� �f ðxÞ

Where P is the maximum load, B and W are the specimenthickness and width, respectively, and f(x) is expressed asfollows:

f ðxÞ ¼ 6x1=2 ½1:99� xð1� xÞð2:15� 3:93x þ 2:7x2Þ�

ð1þ 2xÞð1� xÞ3=2

Where x ¼ a/W and a is the crack length. To determine theKIC value, a minimum of three samples were tested.

Morphological ObservationsThe fractography of the failed SENB samples was observedto examine the fracture characteristics of the cured epoxy/HBP blends. To study the phase structure of the cured neatepoxy resin and epoxy/HBP blends, the samples were frac-tured under cryogenic conditions using liquid nitrogen andthen etched in tetrahydrofuran (THF) for 4 h. The HBP-richphase was preferentially removed by the solvent, while thecured epoxy phase remained relatively unaffected. A ZeissSupra 55VP scanning electron microscope was used for ob-servation, before which the surfaces were coated with goldof the thickness of about 13 nm.

Fourier Transform Infrared (FTIR) SpectroscopyFTIR measurements were performed with a Bruker Vertex70 FTIR spectrometer for the HBP, the cured neat epoxyresin and the cured epoxy/HBP blends. Scraps from the sam-ple materials were ground with KBr into fine powder, whichwas further pressed into thin disks for measurements. Allspectra were recorded at room temperature and signal-aver-aged over 64 scans at a resolution of 2 cm�1.

RESULTS AND DISCUSSION

Differential Scanning CalorimetryThe DDS-cured epoxy/HBP blends display a decreased trans-parency and gradually become opaque with the increase ofthe HBP content, indicating the occurrence of phase separa-tion. The DSC curves obtained for the DDS-cured neat epoxyresin and epoxy/HBP blends during the second heating scanare shown in Figure 1. The glass transition of the epoxy-richphase, TgER can be seen in all DSC curves. There is a reduc-tion in TgER with increasing HBP content, which may becaused by the residual miscibility between HBP and thecrosslinked epoxy resin. The dilution effect of HBP may alsoproduce an incomplete curing reaction, which results in alower Tg for the crosslinked epoxy resin.18,19 It is noted thatthe glass transition of the HBP-rich phase, TgHBP was notdetected by DSC for the blends with HBP content up to 20phr. However, the cured blend with 30 phr HBP exhibits twoseparate glass transitions. The glass transition at higher tem-perature in the blend containing 30 phr HBP is associatedwith the cured epoxy phase, which is TgER; the other atlower temperature is attributed to the HBP-rich phase, whichis TgHBP. The values for TgER and TgHBP measured by DSC areincluded in Table 1 along with the measurements carriedout by DMA.

Dynamic Mechanical AnalysisFigure 2 shows the dynamic mechanical spectra for the DDS-cured neat epoxy resin and epoxy/HBP blends with the HBPcontent of up to 30 phr. The DMA measurements of thecured epoxy/HBP blends with HBP content up to 20 phrshow similar storage moduli G to the cured neat epoxy resin.A reduction in G at �40 �C was observed in the blend with30 phr HBP content, which corresponds to the low-tempera-ture thermal transition of HBP. The tan d versus temperaturecurve for the cured neat epoxy resin displays a well-definedrelaxation peak at 245 �C, attributed to its glass transition.This relaxation peak shifts slightly to a lower temperature inthe cured blends with increasing HBP content, that is, areduction from 237 �C to 230 �C for TgER, corresponding tothe increase of HBP content from 5 to 30 phr. It is notedfrom Figure 2(b) that a well-defined lower-temperaturerelaxation peak starts to appear in the cured blend with HBPcontent of 15 phr. This indicates the existence of anotherphase, that is, a HBP-rich phase in the cured blends. The val-ues for TgER and TgHBP measured by DMA also can be found

FIGURE 1 DSC thermograms of the second run of the DDS-

cured neat epoxy resin and epoxy/HBP blends.

TABLE 1 TgER and TgHBP for DDS-Cured Neat Epoxy Resin and

Epoxy/HBP Blends Measured by DMA and DSC

HBP

(phr)

TgER

(DMA) (�C)

TgER

(DSC) (�C)

TgHBP

(DMA) (�C)

TgHBP

(DSC) (�C)

0 245 216

5 237 204

10 235 205

15 230 205 63

20 230 195 59

30 230 196 64 36

ARTICLE


in Table 1. As expected, the Tg values measured by DMA aregenerally 20–30 �C higher than those measured by DSC atthe same compositions. DMA is more sensitive to the detec-tion of the HBP-rich phase, which shows the cured blendswith 15–30 phr HBP are phase separated. It is interesting tosee that the glass transition temperature of the HBP-richphase (TgHBP) is decreased in the cured blends with additionof HBP. This result implies that the epoxy resin dissolved inthe HBP-rich phase was not sufficiently cured; the HBP-richphase composed of HBP and some uncured and/or insuffi-ciently cured epoxy resin thus had a reduced Tg.

3

Analogous results have been shown in other researcher’swork about the toughening of TGDDM (Araldite MY721,Ciba) by using a hyperbranched epoxy-functional HBP Bol-torn E1.16 The increasing additive content produced a grad-ual decrease in TgER, resulting in a 24 �C reduction from thecured neat epoxy resin when 15 wt % HBP is used. In con-trast, the TgER was almost unaffected when this HBP wasadded into a difunctional epoxy DGEBA (DER 331, Dow).16

Previous work also supports unchanged TgER in the DDM-

cured diglycidyl 1,2-cyclohexanedicarboxylate (DGCHD)/Boltorn H40 blends with HBP content up to 30 wt %.3

Fracture Toughness and Phase StructureThe KIC values and Young’s modulus values as a function ofthe HBP content for the cured neat epoxy resin and epoxy/HBP blends are shown in Figure 3. The addition of HBPleads to a 90% increase in the KIC values, from 0.503MPa

ffiffiffiffim

pfor the cured neat epoxy resin to 0.956 MPa

ffiffiffiffim

p

for the cured epoxy/HBP blend with 30 phr HBP. In themean time, the Young’s modulus of the cured epoxy/HBPblend shows limited decrease over that of the unmodifiedepoxy. The HBP modified tetrafunctional epoxy systems inthis work exhibit significantly higher fracture toughness thanthermoplastic modified tetrafunctional epoxies.20–22 In Fer-nandez et al.’s work,20 polyethersulfone (PES) was addedinto a DDM-cured TGDDM epoxy resin; no significantimprovement was observed in fracture toughness after modi-fication. The deficiency of modification was believed to becaused by the nanoscale phase separation; the high crosslinkdensity of the tetrafunctional epoxy resin hinders the occur-rence of higher enhancement in fracture toughness. Phenol-phthalein poly(ether ether ketone) (PEK-C) was also appliedto modify a DDM-cured TGDDM resin in Song et al.’s work.23

The fracture toughness of epoxy/thermoplastic blendsshowed a slight decreasing tendency, presumably due to thereduced crosslink density of the epoxy net work and the ho-mogenous phase structure of the cured blend.

The morphology of the cured epoxy/HBP blends were inves-tigated by SEM. Figure 4 presents the phase structure of thecured epoxy/HBP blends by cryogenically breaking the sam-ples and then immersing in THF. Heterogeneous morphologyis observed for the cured blends with 10, 15, 20, and 30 phrHBP content. In comparison with the DMA and DSC results,the cured blend with HBP content of 10 phr was alsorevealed to have a two-phase morphology. For the curedblend containing 5 phr HBP, a homogenous morphology wasobserved [Fig. 4(a)]. One of the most important criteria forobtaining increased fracture toughness in rubber or

FIGURE 2 Temperature dependence of (a) storage modulus

(G0) and (b) tan d of the DDS-cured neat epoxy resin and ep-

oxy/HBP blends.

FIGURE 3 Effect of HBP content on the fracture toughness and

Young’s modulus of DDS-cured neat epoxy resin and epoxy/

HBP blends.



thermoplastic modified epoxy resins is a two-phase morphol-ogy after phase separation.24 Similarly, with 5 phr HBP, notoughening result is shown here. With the increase of HBPcontent, a particulate morphology in which the discrete HBP-rich phases are dispersed in a continuous matrix of epoxyresin is obtained.25 The discrete HBP-rich particles with av-erage size less than 1 lm in diameter appear in the curedblend containing 10 phr HBP [Fig. 4(b,c)]. Figure 4(c) is anenlargement to reveal the small HBP-rich particles which

cannot be seen clearly in Figure 4(b) using the same scale asthe other SEM micrographs. The toughening effect is notobvious for the cured blend with 10 phr HBP content despitethe occurrence of phase separation. Inclusion of 15 phr HBPproduces much more and larger HBP-rich phases with a bi-modal particle size distribution, resulting in a 35% enhance-ment of KIC. When the HBP content increases to 20 phr, theparticle size ratio in the bimodal phase structure becomeshigher [Fig. 4(e)], leading to higher fracture toughness in the

FIGURE 4 SEM micrographs of cryogenically fractured surfaces of DDS-cured epoxies with (a) 5, (b), (c) 10, (d) 15, (e) 20, (f) 30 phr

HBP modifiers. The scale bars are 10 lm except for the micrograph (c) which is 2 lm.

ARTICLE


cured blend. Okamoto et al.’s have studied the relationshipbetween the Izod impact strength and the blend ratios oftwo different monomodal high impact polystyrene (HIPS) fortwo bimodal systems. They observed that the particle sizeratio in the bimodal system affects the extent of the toughen-ing effect. The greater the difference between the particlesize of the constituent bimodal HIPS, the higher the impactstrength is.26 The maximum toughening effect is demon-strated for the cured blend containing 30 phr HBP, when auniform distribution of monomodal HBP-rich particles isachieved.

It is noted that all the epoxy/HBP blends with HBP contentfrom 10 to 30 phr experience phase separation through thenucleation and growth (NG) mechanism.27,28 An importantfactor which may influence the phase behavior of the epoxy/HBP blends needs to be considered: as a low viscosity modi-fier, HBP does not have the semi permeable characteristicsrequired for mass transfer (usually, when the modifier is aviscous thermoplastic, the dispersed primary phase will con-tinuously receive material from the continuous primaryphase but not deliver species back29), which can be associ-ated with significantly different thermodynamic simulationsfrom thermoplastic modified epoxies. For the cured blendscontaining 15 and 20 phr HBP, both phase structures displaybimodal particle-size distributions [Fig. 4(d,e)]. The explana-tion could be based on the following: when the average dis-tance between growing particles is high and the conversionrate is significantly high, concentration profiles of the modi-fiers become very sharp. The high supersaturation at thesepoints (concentration excess with respect to equilibrium atthe particular conversion level) causes a new nucleation pro-cess to occur. The different concentration profiles arerepeated as the cure conversion proceeds, resulting in the bi-modal particle size distributions. However, the inclusion ofhigher HBP content (30 phr) gives rise to a reduced cureconversion rate when phase separation occurs, which maynot be sufficient enough to initiate the second NG process.Under these conditions, the cured blend containing 30 phrHBP exhibits a monomodal particle size distribution [Fig.4(f)].

FractographyFigure 5 presents the SEM micrographs on the fracturesurfaces of cured neat epoxy resin and epoxy/HBP blends af-ter SENB testing. For the neat epoxy resin [Fig. 5(a)], theregularly oriented cracks are developed freely, indicating typ-ical brittle characteristics of the fracture surface. The inclu-sion of 5 phr HBP does not change the brittle nature of thefracture surface [Fig. 5(b)], accounting for its unimprovedfracture toughness. As we have known earlier, phase separa-tion occurs when the HBP content increases to 10 phr.Although the dispersed HBP-rich particles cannot be dis-tinctly observed in Figure 5(c), a great amount of tortuousand fine cracks appear on the fracture surface, resultingfrom the formation of shear bands near the plastic HBP-richparticles. The plastic characteristic is further developed onthe surface of the cured blend containing 15 phr HBP [Fig.5(d)]. The dispersed HBP-rich particles create stress concen-

trations at their equators and also act as sites for initiatingshear bands. When the shear bands created by one particleinteract with another particle, they may stop propagatingand keep the yielding localized.30,31 More matrix yielding isinduced locally surrounding the particles due to a largernumber of dispersed particles. Some cracks deviate fromtheir original planes, leading to significantly increased sur-face area. All these factors contribute to more absorbed sur-face energy thereby increasing fracture toughness. For thecured blend containing 20 phr HBP [Fig. 5(e)], the crack tipsbecome obscured and large plastically deformed planes arealigned regularly on the surface. As the particle size ratio inthe bimodal phase structure increases, the surface reliefbecomes more pronounced, which absorbs even more energyand gives a higher KIC. The SEM micrograph for the 30 phrHBP blends [Fig. 5(f)] shows distinctly different characteris-tics from the others. No more lengthy cracks can bedetected; extensive plastic deformation takes place in thecontinuous matrix embedded with uniformly distributedHBP-rich particles with considerably high concentration.Apparently, the most surface energy is absorbed through thisway, contributing to the highest KIC achieved among all theHBP modified epoxies in this study. In addition to the previ-ous observations, particle cavitations are present in all theinvestigated epoxy/HBP blends which experience phase sep-aration. The cavitated HBP-rich particles induce large stressconcentrations, leading to extensive shear deformation. Thishigh energy absorbing mechanism assists with the effectivetoughening of the tetrafunctional epoxy resin.32

Hydrogen-Bonding InteractionsFigure 6 shows the FTIR spectra of the HBP and the DDS-cured epoxy/HBP blends in the stretching region from 3100to 3800 cm�1. Two components can be seen from both thespectrum for the HBP and for the DDS-cured neat epoxyresin. For the cured epoxy resin, a broad band centered at3389 cm�1 is attributed to the self-associated hydroxylgroups (i.e., hydrogen bonded hydroxyl groups) and ashoulder centered at 3557 cm�1 is assigned to nonassoci-ated, free hydroxyl groups33; for the HBP, a broad band cen-tered at 3432 cm�1 is attributed to the hydrogen bondedhydroxyl group and the shoulder centered at 3557 cm�1 isalso assigned to free hydroxyl groups. The relative intensityof the free hydroxyl band to the hydrogen bonded hydroxylband is the highest for the HBP, indicating the existence of ahigher fraction of free hydroxyl groups in the HBP than boththe cured neat epoxy resin and the epoxy/HBP blends. Thespectrum for the cured epoxy resin shows a reduced relativeintensity of the free hydroxyl band to the hydrogen bondedhydroxyl band. The inclusion of HBP to the epoxy resincauses gradually decreased relative intensity of the freehydroxyl band to the hydrogen bonded hydroxyl band, whichindicates the formation of hydrogen bonded hydroxyls in thecured epoxy/HBP blends. It also can be seen from Figure 6that with the increasing HBP content, the associatedhydroxyl band shifts to higher frequencies; however, the non-associated hydroxyl band does not move. The frequency dif-ference between the nonassociated hydroxyl absorption and



the associated hydroxyl absorption (Dv) is a measure of theaverage strength of the intermolecular interactions.34 Thisindicates that the average strength of the hydrogen bond inthe cured blends (Dv ¼ 148 cm�1 for the cured epoxy/HBPblend containing 30 phr HBP) is lower than the hydrogenbond in the cured neat epoxy resin (Dv ¼ 168 cm�1).

CONCLUSIONS

The DDS-cured epoxy/HBP blends are immiscible, with twophases clearly existing in the cured blends, that is, an epoxy-

rich phase and a HBP-rich phase, which accounts for the twoseparate glass transition temperatures demonstrated byDMA. The phase morphology is dependent on the blend com-position. In the blend with low HBP content (5 phr), there isno phase separation occurring; however, all the other studiedcured epoxy/HBP blends exhibit a particulate morphology,where discrete HBP-rich particles are dispersed in a continu-ous epoxy-rich matrix. For the blends with 15 and 20 phrHBP, a bimodal particle size distribution was observed,which is in contrast to the monomodal particle size distribu-tion found in the blend containing 30 phr HBP. The fracture

FIGURE 5 SEM micrographs on KIC fracture surfaces of DDS-cured epoxies with (a) 0, (b) 5, (c) 10, (d) 15, (e) 20, (f) 30 phr HBP

modifiers. The scale bars are 10 lm.

ARTICLE


toughness increases from 0.50 MPaffiffiffiffim

pfor the cured neat

epoxy resin to 0.96 MPaffiffiffiffim

pfor the epoxy/HBP blend with

30 phr HBP, where a twofold increase was achieved. BoltornH40 has successfully toughened the tetrafunctional TGDDMepoxy resin. Fractography on the KIC fracture surfaces indi-cates that matrix yielding and particle cavitations are themain mechanisms responsible for the improved fracturetoughness. The toughening effect of the HBP modified epoxyresin significantly depends on the phase structures. Both thebimodal and monomodal particle size distribution give riseto improved fracture toughness, whereas the monomodalparticle size distribution shows stronger toughening effect inthis work. FTIR reveals the formation of hydrogen bondedhydroxyl groups between the epoxy network and the HBPmodifier.

This work was financially supported by Deakin Universityunder a central research grants scheme. The authors thank toPerstorp Specialty Chemicals (Sweden) and Hunstman (Aus-tralia) for the kind donation of sample materials.

REFERENCES AND NOTES

1 Gao, C.; Yan, D. Prog Polym Sci 2004, 29, 183–275.

2 Mezzenga, R.; Boogh, L.; Manson, J. A. E. Compos Sci Tech-

nol 2001, 61, 787–795.

3 Guo, Q.; Habrard, A.; Park, Y.; Halley, P. J.; Simon, G. P.

J Polym Sci Part B: Polym Phys 2006, 44, 889–899.

4 Chikhi, N.; Fellahi S.; Bakar M. Eur Polym J 2002, 38,

251–264.

5 Ratna, D.; Simon, G. P. Polymer 2001, 42, 7739–7747.

6 Francis, B.; Thomas, S.; Jose, J.; Ramaswamy, R.; Lakshmana

Rao, V. Polymer 2005, 46, 12372–12385.

7 Mimura, K.; Ito, H.; Fujioka, H. Polymer 2000, 41, 4451–4459.

8 Guo, Q. Polymer 1993, 34, 70–76.

9 Zhong, Z.; Zheng, S.; Huang, J; Cheng, X.; Guo, Q; Wei, J.

Polymer 1998, 39, 1075–1080.

10 Cicala, G.; Recca, A.; Restuccia C. Polym Eng Sci 2005, 45,

225–237.

11 Ratna, D.; Simon, G. P. Polymer 2001, 42, 8833–8839.

12 Varley, R. J.; Tian, W. Polym Int 2004, 53, 69–77.

13 Ratna, D.; Simon, G. P. Polym Eng Sci 2001, 41, 1815–1822.

14 Ratna, D.; Varley, R.; Simon, G. P. J Appl Polym Sci 2003,

89, 2339–2345.

15 Ratna, D.; Varley, R.; Simon, G. P. J Appl Polym Sci 2004,

92, 1604–1610.

16 Varley, R. J. Polym Int 2004, 53, 78–84.

17 Lin, K. F.; Chen, J. C. Polym Eng Sci 1996, 36, 211–217.

18 Zheng, S.; Wang, J.; Guo, Q.; Wei, J.; Li, J. Polymer 1996,

37, 4667–4673.

19 Guo, Q.; Huang, J.; Li, B.; Chen, T.; Zhang, H.; Feng, Z. Poly-

mer 1991, 32, 58–65.

20 Fernandez, B.; Arbelaiz, A.; Diaz, E.; Mondragon, I. Polym

Compos 2004, 25, 480–488.

21 Blanco, I.; Cicala, G.; Costa, M.; Recca, A. J Appl Polym Sci

2006, 100, 4880–4887.

22 Hourston, D. J.; Lane, J. M.; MacBeath, N. A. Polym Int

1991, 26, 17–21.

23 Song, X.; Zheng, S.; Huang, J.; Zhu, P.; Guo, Q. J Appl

Polym Sci 2001, 79, 598–607.

24 Bucknall, C. B.; Patridge, I. K. Polymer 1983, 24, 639–644.

25 Guo, Q.; Huang, J.; Ge, L.; Feng, Z. Eur Polym J 1992, 28,

405–409.

26 Okamoto, Y.; Miyagi, H.; Kihara, H.; Mitsui, S. In Impact

Toughening Mechanisms in Glassy Polymers, Shonaike, G. O.;

Simon, G. P., Eds.; Polymer Blends and Alloys. CRC Press:

New York, 1999.

27 Chen, J. L.; Chang, F. C. Polymer 2001, 42, 2193–2199.

28 Inoue, T. Prog Polym Sci 1995, 20, 119–153.

29 Williams, R. J. J.; Rozenberg, B. A.; Pascault, J. P. In Reac-

tion-Induced Phase Separation in Modified Thermosetting Poly-

mers; Polymer Analysis Polymer Physics. Springer: Berlin,

1997; pp 96–156.

30 Garg, A. C.; Mai, Y. W. Compos Sci Technol 1988, 31,

179–223.

31 Atkins, A. G.; Mai, Y. W. Elastic and Plastic Fracture—Met-

als, Polymers, Ceramics, Composites, Biological materials, 2nd

ed. Ellis Horwood Limited: UK, 1988.

32 Boogh, L.; Pettersson, B.; Manson, J. E. Polymer 1999, 40,

2249–2261.

33 Guo, Q.; Harrats, C.; Groeninckx, G.; Reynaers, H.; Koch,

M. H. J. Polymer 2001, 42, 6031–6041.

34 Purcell, K.; Drago, R. J Am Chem Soc 1967, 89, 2874–2879.

FIGURE 6 FTIR spectra in the 3100–3800 cm�1 region of the

HBP (Boltorn H40) and the DDS-cured epoxy/HBP blends.



本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具

http://www.xuebalib.com/cloud/

http://www.xuebalib.com/

http://www.xuebalib.com/cloud/


http://www.xuebalib.com/vip.html

http://www.xuebalib.com/db.php

http://www.xuebalib.com/zixun/2014-08-15/44.html


thermal and mechanical properties of a dendritic hydroxyl

Documents