at SciVerse ScienceDirect

Polymer Degradation and Stability 98 (2013) 1205e1215

Contents lists available

Polymer Degradation and Stability

journal homepage: www.elsevier .com/locate /polydegstab

Stability investigation of self-healing microcapsules containing rejuvenator forbitumen

Jun-Feng Su a,b,*, Jian Qiu c, Erik Schlangen b

a Institute of Materials Science & Chemical Engineering, Tianjin University of Commerce, Tianjin 300134, PR ChinabDepartment of Materials and Environment, Faculty of Civil Engineering & Geosciences, Delft University of Technology, Stevinweg 1, 2628CN Delft, The NetherlandscDepartment of Road and Railway Engineering, Faculty of Civil Engineering & Geosciences, Delft University of Technology, Stevinweg 1, 2628CN Delft, The Netherlands

a r t i c l e i n f o

Article history:Received 26 November 2012Received in revised form8 March 2013Accepted 12 March 2013Available online 23 March 2013

Keywords:MicrocapsulesStabilitySelf-healingBitumenRejuvenator

* Corresponding author. Institute of Materials ScieTianjin University of Commerce, Tianjin 300134,26210595.

E-mail addresses: j.su@tudelft.nl, sujunfeng2000@

0141-3910/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.polymdegradstab.2013.03.0

a b s t r a c t

Preservation and renovation bitumen of pavement is a big problem for the whole world. Traditionally,application rejuvenator is the only one method that can restore the original properties of the pavements.However, some puzzles still restrict its successful usage. Microencapsulation is a promising method toapply rejuvenator in bitumen. These microcapsules can break and leak the oily-liquid rejuvenator intomicrocracks and self-healing the aged bitumen. Based on our previous work, the objective of this studywas to investigate the thermal stability, mechanical stability and interface stability of microcapsules inbitumen. The results showed that these microcapsules containing rejuvenator survived in meltingbitumen and in a violent repeated temperature changes. Microcapsules had the elasticeplastic defor-mation ability resisting the temperature changes and mixing stress. Moreover, the chemical bondsimproved the interface stability between shells and bitumen. Microcapsules containing rejuvenator willbe a promising product to realize the smart pavements.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The decline in the use of natural bitumen for road constructioncan be traced to the 1910s when the advent of vacuum distillationmade it possible to obtain artificial bitumen from crude oil.Currently, 95% of the almost 100 million tons of bitumen producedworldwide each year is applied in the paving industry, where thebitumen essentially acts as a binder for mineral aggregates to formasphalt mixes [1]. Other uses of bitumen are as emulsions, water-proof materials, or formedmaterials, but these account for less than5% of the total bitumen produced. As a widely applied material inpavements, bitumen must be sufficiently fluid at high temperature(around 160 �C) to be workable and allow for homogenous coatingof the aggregates upon mixing. Another important issue that has tobe considered is the extent towhich it ages from climate and traffic.After years of use the stiffness of asphalt concrete increases whileits relaxation capacity decreases. This causes the binder to becomemore brittle, causing the development of microcracks and ulti-mately to cracking of the interface between the aggregates and

nce & Chemical Engineering,PR China. Tel./fax: þ86 22

yahoo.com.cn (J.-F. Su).

All rights reserved.08

binder [2]. This occurs mainly as a result of oxidation of the hy-drocarbon compounds contained within the bitumen [3]. Bitumenbinders are usually categorized into two subdivisions: solids calledasphaltenes and liquids called maltenes. Maltenes can be furtherdivided into polar aromatics, naphthalene aromatics, and saturates(paraffins) [4]. The main aging mechanism of bitumen is the loss ofvolatiles and oxidation, which leads to bitumen with higher vis-cosity (stiffer) [5]. In other words, the amount of solid componentincreases and that of the liquid component decreases, thus result-ing in an increase in the rigidity of the pavement.

The aging problem of bitumen leads to pavement failure,including surface raveling and reflective cracking. It therefore in-creases the cost of renovating and preserving bituminous pave-ments [6]. Several physical and chemical methods are currentlyemployed for bitumen preservation, including the use of rejuve-nator emulsions or fog seals, and through additive modification orthin overlay technologies [1,7]. However, of these methods only thefirst, i.e., the application of rejuvenators, can restore the originalproperties of the pavement [6]. The most important goal of usingrejuvenator product is to restore the asphaltene/maltene ratio to itsoriginal balance [8]. Rejuvenating agents have the ability toreconstitute the binder’s chemical composition and they consist oflubricating and extender oils that contain a high proportion ofmaltene constituents [9]. Rejuvenator can soften the aged binderand provide comprehensive rejuvenation that replenishes the

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e12151206

volatiles and dispersing oils while simultaneously promotingadhesion.

However, for a rejuvenator to be successfully applied the diffi-culty in penetrating the pavement surface still remains a significantproblem. Shen et al. [10] reported on the use of three rejuvenatorsand found that none could penetrate more than 2 cm into theasphalt concrete. Further issues encountered when applying thesematerials include the fact that road closures are necessary for someperiod of time after their application. The rejuvenator may alsocause a high reduction in the surface friction of the pavement forvehicles. Moreover, these rejuvenators may also be harmful to theenvironment.

To overcome these issues inspiration is provided by the conceptof self-healing based on microcapsules. This particular approachinvolves incorporation of a microencapsulated healing agent and adispersed catalyst within a polymer matrix [11e13]. Upon damage-induced cracking, the microcapsules are ruptured by the propa-gating crack fronts resulting in release of the healing agent into thecracks by capillary action [14]. The method of encapsulating re-juvenators inside the bitumen may be an alternative approachworthy of consideration. The application of microcapsules con-taining rejuvenator to bitumen derives from the success observedfor some polymer self-healing materials [15e17]. García et al. [18]reported a method to prepare rejuvenator capsules by using anepoxy resin as a coating and porous sand as a skeleton. The ad-vantages of these capsules include the fact that they are strongenough to resist the mixing process, the high temperature, and allthe years in the road until they are required. However, these cap-sules have some limitations that restrict their application. Theprimary limitation is that it is hard for the rejuvenator to flow outfrom the porous sand when the shell is broken because the

Fig. 1. Illustration of the possible states of micro

rejuvenator has a high viscidity resulting from it consisting oflubricating and extender oils. The capillary action of the porousstructure also limits the rejuvenator release. Another limitation isthat the capsule size does not fit with the thickness of bitumenbetween aggregates. To realize the application of these promisingchemical products, we have developed a novel method to fabricatemicrocapsules containing rejuvenator by in situ polymerizationusing methanol-melamine-formaldehyde (MMF) prepolymer asshells [19]. A two-step coacervation process, with the aid of styrenemaleic anhydride (SMA) as a surfactant, was successfully applied toenhance the thermal stability and compactibility of the shells. It hasbeen shown that this product is an environmentally friendlypowder that encapsulates a suitably sized rejuvenator for chemicaland construction engineering.

To produce microcapsules containing rejuvenator by chemicalmeans, the factors of cost, complexity, and capacity must beconsidered for the construction industry. These coreeshell micro-capsule structures need to meet specific requirements in terms ofsize distribution, encapsulation ratio, and non-biodegradabilitybecause these factors will influence their service performance [20].As bitumen acts as thin layers between aggregates that areusually less than 50 mm, the size of the microcapsules containingrejuvenators should be smaller than 50 mm to avoid being squeezedor pulverized during asphalt forming. Besides the complex fabrica-tion process of microcapsules containing rejuvenator, theirsurvival ability is another important issue that must be addressed.Fig. 1 illustrates the possible states of the microcapsules in bitumenmaterial. First, for practical application themicrocapsulesmust havehigh thermal stability and high mechanical strength to resist themelting temperature and mixing stress of the asphalt (Fig. 1aec).The microcapsules must maintain their shape and compatibility at

capsules containing rejuvenator in bitumen.

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e1215 1207

temperatures of 160e200 �C during asphalt application. Further,breakage of the shell may occur at high temperatures owing to amismatch in thermal expansion of the core and shell materials [21].It is expected that the rejuvenator is protected, with fewer cracksand with lower permeability, during asphalt paving. However,microcapsule shells, such as inorganic or flexible organic shells, willnot break when the mechanical strength applied is very high(Fig. 1d). The result is that microcracks may be triggered, leading tofracturewithout leakage of the rejuvenator. Consequently, the shellsemployed are more commonly polymeric materials that possessgood thermal stability under high temperature andwhich exhibit anappropriate strength and toughness [22]. Second, interface stabilityand interfacial interactions are keys for all multicomponent mate-rials, irrespective of the number or type of their components, or oftheir actual structure. Thus, understanding the relationship betweenthe interfacial behaviors and mechanical properties of bitumen/microcapsule composites is important for them to be useful underrealistic engineering environments. In our previous report [23], itwas found that microcracks and interface separation may occur formicrocapsule/matrix compositeswhen a repeated, vigorous thermalabsorbing-releasing process is undertaken. During a process withrepeated temperature changes via heat transmission, expansion andshrinkage of themicrocapsules andpolymermatrixwill occur owingto the different expansion coefficients exhibited. The microcapsulevolume can also be affected by the encapsulated rejuvenator uponenvironmental temperature changes. These phenomena will causemicrocracks or fractures in the matrix during heat absorption orresealing, spoiling the thin microcapsule shells such that theencapsulated rejuvenator will lose the protection provided by theshells. Moreover, the mechanical integrity of these composites maydecrease because of internal cracking or microcapsule rupture [24].This is not expected to appear because it is well known thatmicrocapsule rupture and release due to wear and tear can beavoided by proper design of their mechanical properties [25].Theoretically, the mechanical strength of a microcapsule is deter-mined by its size, and by the shell thickness and structure [26]. Themechanical properties of the shells play an important role in manyprocesses, and therefore, understanding of these properties isdesirable for optimizing their application. The theoretical tensileyield strength and ultimate tensile strength of the composites differfor the cases of adhesion andno adhesionbetween thefiller particlesand matrix. In the case of no adhesion between the microcapsulesand bitumen, the interfacial layers are unable to transfer stress.

We are in a position to commence exploring chemical methodscapable of providing simple, cheap, robust, and environmentallyfriendly microcapsules containing rejuvenator for bitumen. In viewof the above, the objective of this work was to fabricate microcap-sules containing rejuvenator by in situ polymerization using MMF-resin shells and to then investigate their stability in bitumen. Ther-mal stability and mechanical properties were measured to ensurethat the microcapsules survived the bitumen during melting andmixing. In addition, interface stabilitywasmonitored to evaluate theinterface bonding behaviors during the service life of the bitumen.

2. Experimental

2.1. Materials

The shell material was commercial prepolymer of melamine-formaldehyde modified by methanol (solid content was 78.0%)purchased from Aonisite Chemical Trade Co., Ltd. (Tianjin, China).The rejuvenator was a commercial product. Styrene maleic anhy-dride (SMA) copolymer (Scripset� 520, Hercules, USA) was appliedas dispersant. A small percentage of the anhydride groups havebeen established with a lowmolecular weight alcohol and it is fine,

off-white, free flowing power with a faint, aromatic odor. Thebitumen used in this study was 70/100 pen obtained from KuwaitPetroleum in a 4.5% by weight [20]. The material used as rejuve-nator is dense, aromatic oil obtained from Petroplus Refining Ant-werp (800DLA, Belgium).

2.2. Microcapsules synthesis

The method of fabrication microcapsules containing rejuvenatorby coacervation proceed can be divided into three steps [19]: (1)SMA (10.0 g) and NP-10 (0.2 g) were added to 100 ml water at 50 �Cand allowed mix for 2 h. Then a solution of NaOH (10%) was addeddropwise adjusting its pH value to 10. The above surfactant solutionand rejuvenator were emulsified mechanically under a vigorousstirring rate for 10min using a high-speed dispersemachine. (2) Theencapsulation was carried out in a 500 ml three-neck round-bottomed flask equipped with a condensator and a tetrafluoro-ethylene mechanical stirrer. The above emulsion was transferred inthe bottle, which was dipped in steady temperature flume (roomtemperature). Half of MMF prepolymer (16 g) was added dropwisewith a stirring speed of 500 r min�1. After 1 h, the temperature wasincreased to 60 �C with a rate of 2 �C min�1. Then another half ofprepolymer (16 g)was dropped in a bottle at the samedropping rate.(3) The temperaturewas increased to 75 �C. After polymerization for1 h, the temperature was decreased slowly at a rate of 2 �C min�1 toambient temperature. At last, the resultant microcapsules werefiltered and washed with pure water and dried in a vacuum oven.

2.3. Morphologies observations

An optical microscope (BX41-12P02, OLYMPUS) was used tocheck the fabrication process of microcapsules in emulsion. About1 ml of the colloidal solution was extracted and spread on a cleanglass slide (1 � 3 cm). The dried microcapsules were adhered on adouble-side adhesive tape without cracking the shells. The surfacemorphologies were observed by using an Environmental ScanElectron Microscopy (ESEM, Philips XL30) at an accelerated voltageof 20 kV.

2.4. Mean size and shell thickness of microcapsules

For each microcapsule sample, the mean size is the average sizevalue of fifty microcapsules measured from the SEM morphologyimage. About 2 g MMF-shell microcapsules was mixed in 5 g epoxyresin. After the composite was dried in room temperature, it wascarefully cut to obtain the cross-section. The thickness of shells canbe measured from the SEM images of cross-section of microcap-sules [27]. At least 20 shells of each microcapsule sample weremeasured and the average data were calculated.

2.5. Thermogravimetric analysis (TGA)

The thermal stability characterization of microcapsules wasperformed on a Dupont SDT-2960 Thermogravimetric analysis(TGA) at a scanning rate of 5 �C min�1 in a flow of 40 ml min�1

nitrogen (N2).

2.6. States of microcapsules in bitumen

Various bitumen/microcapsule samples were prepared toinvestigate the thermal stability. Afluorescencemicroscope (CKX41-F32FL, OLYMPUS) was applied to investigate the interface of mate-rials using the light characters such as reflection, diffractionand refraction. A thermal absorbing-releasing process was per-formed to investigate the thermal stability of bitumen/microcapsule

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e12151208

composites using a temperature-controlled chest. A similar methodhas been successfully applied to investigate the thermal stability ofmicrocapsules containing phase changematerials in epoxy [28]. Thesamples (2 � 1 � 0.5 cm) were heated to 50 �C with the rate of2 �C min�1 and keeping for 10 min, and then decreased the tem-perature to �10 �C with the rate of 2 �C min�1. This thermal treat-ment process for each samplewas repeated for 20, 40, and 60 times.The interface morphologies of the composites were observed by afluorescence microscope. The bitumen/microcapsule samples wereadhered on a glass sheet. Itwas carefully pouredwith a drop of liquidnitrogen at one end. Microcracks were quickly generated in thissample because of the low temperature brittleness. The state ofmicrocapsules in bitumen could be observed by the fluorescencemicroscope.

2.7. Mechanical properties measurement

The mechanical properties of the MMF-shell microcapsuleswere tested according to our reported two-plate micromanipula-tion method [25]. Single microcapsule was adhered on a glass sheet(500 � 500 mm) and pressed by another glass sheet. A pressuresensor under the bottom glass measured the intensity and datawere directly recorded and yield stress was calculated automati-cally by a computer from the forceedisplacement curves.

2.8. Fourier transform infrared spectroscopy (FT-IR)

The chemical structures were analyzed by a Perkin Elmer�

Spectrum 100. FT-IR spectra in absorption modes were recordedamong the range of 400e4000 cm�1.

3. Results and discussion

3.1. Morphologies and physical structure of microcapsules

The fabrication process of microcapsules using melamine-formaldehyde (MF) and urea-formaldehyde (UF) resins as shell

Fig. 2. Morphologies of microcapsules containing rejuvenator, (a) optical morphology of micof single microcapsule, (d) SEM morphology of dried microcapsules with larger size, (e, f) opmorphology of microcapsules.

materials is usually defined as an in situ polymerization method.Emulsifying agent is used to absorb the shell materials on coredroplets. The direct polymerization of a single monomer or pre-polymer is carried out on the core-particle surface. It has beenproved that the MMF can be successfully applied to fabricate mi-crocapsules using the in situ polymerization method [25]. The shellthickness, surface morphology and average size of microcapsulescan be controlled by regulating the core/shell ratio, prepolymeradding speed and emulsion stirring rate. The core/shell ratio meansthe weight ratio of core material and shell material in originalfabrication [19]. Fig. 2(a) shows the optical morphologies of mi-crocapsules in emulsion fabricated by emulsifying rejuvenator witha stirring rate of 4000 rmin�1. Being encapsulated by shell material,the rejuvenator droplets are ultimately separated through theregulation of hydrolyzed SMAmolecules. Thesemicrocapsules haveregular globe shape with smooth surface. Fig. 2(b,c) and (d) showsthe SEM surface morphologies of dried microcapsules with core/shell ratio of 1/1 fabricated by 4000 and 2000 r min�1 emulsionstirring rates. Their mean sizes are about 10 and 20 mm. The driedmicrocapsules still keep the regular global shape. There is noadhesion and impurity substance between microcapsules. Theshells are compact without holes and cracks. To determine the shellthickness, microcapsules were embedded in epoxy resin as shownin Fig. 2(e,f). Microcapsules were uniformly dispersed in epoxy. Thecross-section SEM morphologies of a typical single microcapsuleare presented in Fig. 2(g). The shell thickness was measureddirectly. It must be noted that the shell may not be cut across itsequator, so the thickness is an average data of at least 10 shells foreach microcapsule sample.

Fig. 3(a) shows the mean sizes of microcapsules (core/shell ra-tios of 1/1, 1/2 and 1/3) under various emulsion stirring rates inrange of 1000e8000 r min�1. With the increasing of stirring rates,the size of these microcapsule samples decreased sharply from100.5 to 2.0 mm. The reason is that higher stirring rates will dispersethe core material into smaller droplets. A similar conclusion hasbeen reported by other researches [29,30]. The spread of themicrosphere size distribution was found to decrease with stirring

rocapsules in emulsion, (b) SEMmorphology of dried microcapsules, (c) smooth surfacetical morphology of microcapsules embedded in epoxy resin, and (g) cross-section SEM

Fig. 3. Physical properties of microcapsules controlled by core material stirring rates and core/shell ratios, (a) mean size and (b) shell thickness.

Fig. 4. TGA curves of microcapsules containing rejuvenator with core/shell ratio of 1/3under various temperatures for 1 h, (a) 260 �C, (b) 240 �C, (c) 220 �C, and (d) 200 �C.

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e1215 1209

speed. With the validation of the mathematical correlation, it ispossible to have a good estimate of the average microsphere sizeprior to microsphere preparation. Moreover, less core/shell ratioleads to a higher shell thickness value. This result accords withreported results and indicates that the size is mainly determined byemulsion stirring rates [29]. In addition, it can be concluded thatthe MMF prepolymer had cross-linked with a compact structureforming thin shells. Fig. 3(b) shows the data of shell thickness ofmicrocapsules (core/shell ratios of 1/1, 1/2 and 1/3) under corematerial stirring rates of 2000, 4000, 6000 and 8000 r min�1,respectively. Microcapsule samples have the shell thickness from4.51 � 0.50 to 0.55 � 0.12 mm. More shell material leads the shell toa thicker structure for each sample fabricated under the same corematerial stirring rate.

3.2. Thermal stability of microcapsules

Microcapsules containing rejuvenator are expected to keep theirshells intact, resisting thehigh temperatures of themelting bitumen.In the general encapsulation process, the core material is emulsifiedto form droplets, which are then covered by shell material to obtainthe microcapsules. The surface morphology is controlled by theoperating conditions. Fan and Zhou [31] reported that the initial pHvalue, the concentrations of the shell material and surfactant, andthe stirring rate used during the stage of microencapsulation are allfactors that can influence the surface morphology of the microcap-sules obtained by in situ polymerization.

In our previous study we also found that the surfacemorphology greatly affects the thermal stability of microcapsulescontaining rejuvenator [19]. A lower prepolymer concentration canhelp to improve the degree of surface smoothness and compacti-bility of the shells. It is worth noting that the smooth and non-porous shells are believed to result from the deposition of lowmolecular weight prepolymer at the oilewater interface, while theprepolymer remains soluble. The core/shell ratio is considered oneof the main factors to affect the surface morphology. TGA has beenwidely applied to investigate the encapsulation effect and shellcompactness of microcapsules [32]. We have previously reportedthat the decomposition temperature of microcapsules is higherthan the bitumenmelting temperature of 180 �C. This indicates thatthe cured MMF resin will not thermally decompose upon mixingwith the melting bitumen.

On the other hand, there remains a lack of understanding on thethermal stability of microcapsules under extreme temperaturechanges. The reason for this is that the polymeric shells exhibita tendency to craze and fracture under forceful thermal trans-mission [24]. Fig. 4 shows the TGA curves for microcapsules con-taining rejuvenator with a core/shell ratio of 1/3 under various

temperatures (260 �C, 240 �C, 220 �C, and 200 �C) for 1 h. It wasfound that after heating for 1 h to a temperature of 260 �C themicrocapsules showed a decomposition temperature from about150 �C. The pure rejuvenator applied in this study showed amarkedweight loss between the temperature range 337e469 �C owing tothe evaporation of the oil substance at high temperature [19].Therefore, we can confirm that theweight loss of themicrocapsulesover the temperature range of 150e337 �C is a result of the weightloss of the shells. Comparatively, after heating for 1 h to tempera-tures of 240 �C, 220 �C, or 200 �C, the microcapsules showeddecomposition temperatures of 200 �C, 240 �C, and 290 �C,respectively. Heating temperatures have been shown to greatlyinfluence the original decomposition temperatures of microcap-sules. Higher temperatures lower the temperature at which thecommencement of degradation occurs for microcapsules. Thisphenomenon can be attributed to two factors. The first is that theshells had broken under the heating temperatures over an hour.Second, the reason may be that the shells did not retain theircompact shapes. With an increase in temperature more defects onthe shells may occur. These defects will lead the microcapsules tomore rapidly decompose. In addition to the above, the initialdecomposition temperatures of the microcapsules remain higherthan the bitumen melting temperature. This indicates that themicrocapsules can survive when in melting bitumen.

Fig. 5(aef) shows the SEM morphologies of microcapsules afterthey have been heated for 1 h to 180 �C, 200 �C, 220 �C, 240 �C,

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e12151210

260 �C, and 280 �C, respectively. It can be seen that the microcap-sules retain their compact shell structures for temperatures of180 �C, 200 �C, and 220 �C for 1 h, as shown in Fig. 5(aec). Under180 �C and 200 �C, the microcapsules closely maintain the shapeand smooth surfaces of the original states. Although the thermaleffect did not break the shells under 220 �C, the shells stuck closetogether because of the softening of the shells. From Fig. 5(d) and(e) it can be seen that the shells contain holes, cracks, or evenbreaks when the temperature is increased to 240 �C and 260 �C. Theshells are almost entirely broken under 280 �C after the hour ofheating, as shown in Fig. 5(f). Under these circumstances, theencapsulated rejuvenator leaked out without the protection of theshells.

3.3. Thermal stability of microcapsules in bitumen

Microcapsules containing rejuvenator are homogeneouslydistributed in bitumen after hot-mixing. In addition to the thermalstability of the microcapsules, we also investigated the thermalstability of the microcapsules in bitumen. The mechanical proper-ties of the shells can be enhanced by modifying their chemicalstructure and by controlling the synthesis conditions [33]. How-ever, until now little has been known about shell stability inbitumen during a repeated thermal transmission process. Toinvestigate the shell stability of microcapsules in bitumen wedesigned a thermal transmission process to simulate a practicalapplication environment. To verify the thermal stability of micro-capsules in bitumen a microcapsule sample was mixed (2.0 wt.%)with melting bitumen under temperatures of 180 �C and 200 �Cfor 10 min. The microcapsule sample had the same core/shell ratio(1/1) with a mean size about 15 mm. The frozen bitumen sampleswere broken apart and the interface morphologies were observed.It can be seen in Fig. 6(a) and (b) that the microcapsules survived inthe bitumen under a temperature of 200 �C. The microcapsulesretained their global shape with no cracks or thermal decomposi-tion. These results indicate that these microcapsules can resist thethermal effects of asphalt for common applications.

An alternative thermal process was designed to test the thermalstability of the microcapsules containing rejuvenator in bitumenunder an extreme condition. The bitumen/microcapsule sampleswere repeatedly treated 20, 40, and 60 times with a thermalabsorbing-releasing process. For each condition the sample washeated to 50 �C at a rate of 2 �C min�1, with this temperaturemaintained for 10 min before being reduced to �10 �C at a rate of2 �C min�1. Fig. 7(a) shows the original state morphologies andfluorescence properties of the microcapsules (2.0 wt.%) dispersed

Fig. 5. SEM morphologies of microcapsules after heated to (a) 180 �

in bitumen. The amount of rejuvenator retained within themicrocapsule compared to the initial amount is indicative of theencapsulation efficiency (E) of the microcapsule fabrication pro-cess. We found that the highest E (70%) was obtained under theformation conditions of a core/shell ratio of 1/3, a stirring speed of3000 r min�1, and 2.0e2.5% of SMA [19]. All residues on the shellscannot be removed by water because the rejuvenator cannot betotally encapsulated by MMF shells. Bitumen is stained blue, whilethe rejuvenator is green. It is clear from our results that the mi-crocapsules have been homogeneously dispersed in the bitumen.Fig. 7(b,c) shows the microcapsules’ states in bitumen afterrepeated thermal treatment of 20 and 40 times. It is found that thearea of blue points had grown larger in comparison to the originalstate, which indicates that the rejuvenator had permeated out ofthe shells. It is well known that the release behavior of microcap-sules largely depends on the polymer structure of the shell and onthe temperature, which in turn is influenced by the preparationconditions employed. As the microcapsules possess the same meansize and core/shell ratio in all bitumen samples studied herein, thetemperature change is the only factor influencing the permeabilityof the microcapsules. The lower the temperature the more timerequired for the core material to penetrate the shell. Rejuvenator isan oily mixture with high viscosity; therefore, heating effects canaccelerate its molecular velocity through the shells. However, mostof the rejuvenator is retained in the shells under such a thermaltreatment process. This means that the rejuvenator can be pro-tected by the shells for a longer life in practical applications. InFig. 7(d) it can be seen that some microcapsules had broken afterthe treatment was repeated 60 times, as evidenced by the rejuve-nator having flowed out of the shells. In this instance the rejuve-nator permeated the bitumen as small molecules.

3.4. Deformation behaviors of microcapsules in bitumen

Another important issue to be considered for microcapsulescontaining rejuvenator when applied in bitumen is their mechan-ical properties, including their deformation behavior. In some casesbreakage of the microcapsules must be triggered by interiormicrocracks or by exterior pressure. In other cases, the release orleakage of rejuvenator in bitumen is undesirable. Microcapsulesmust be sufficiently strong to remain intact during manufactureand further processing, such as during drying, pumping, and mix-ing [23]. Mechanical properties are, therefore, fundamentallyimportant because they determine the microcapsule’s stability andintegrity. It is well known that microcapsule rupture and releasedue to wear and tear can be avoided by proper design of their

C, (b) 200 �C, (c) 220 �C, (d) 240 �C, (e) 260 �C and (f) 280 �C.

Fig. 6. SEM morphologies of microcapsules in bitumen (2.0 wt.%).

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e1215 1211

mechanical properties [34]. It has been reported that some mi-crocapsules with polymeric shells may act as viscoelastic particles(depending on their size) [35]. Several methods have been reportedto measure the mechanical properties of a single microcapsule orcell [29,30]. Because of their small structure, there has been littlefocus on the relationship between their deformation and shellstructure (size and thickness). As a soft and microscale particle, anindividual microcapsule’s mechanical properties are difficult to bequantitatively characterized. There is therefore a requirement for asystem capable of accurately measuring low-magnitude forces andmicroscopic material deformation [36]. In this study, a similarmethod was also applied to investigate the mechanical propertiesof microcapsules using two-plate micromanipulation [20].

Fig. 8(a) shows a typical loadedisplacement curve for a singlemicrocapsule, where it increases linearly upon loading untilremaining constant, then decreases sharply upon unloading. Themicrocapsule samplewas fabricated under a corematerial dispersalrate of 3000 r min�1 with a core/shell ratio of 1/1. The shape of thecurve illustrates both elastic (recovery) and plastic (inelastic)deformation behaviors. During unloading, the curve representselastic behavior with elastic displacement being recovered. Thismeans that a single microcapsule has an ability to recover itsoriginal shape. The loadingeunloading curves exhibit significanthysteresis under a large deformation. During the unloading pro-cess, the force decreases to zero before the displacement recoversto zero. This means that the microcapsules had a plastic deforma-tion. The ‘Y’ point represents the ‘yield point’ on the concaveeconvex curve. In theory, creep in polymer materials under a con-stant load is controlled by the free volume available for the polymerchains andmolecular units to move. Thinner shells may provide thecross-linked MMF polymer with more volume to perform molec-ular chain adjustments. Moreover, a higher proportion of shellmaterial leads the shells to a greater hardness, which means thatthe thicker shell has an increased ability to resist deformation [27].However, in this study we pay more interest to the plastic defor-mation behavior because it determines the rupture of microcap-sules. Interestingly, it has also previously been reported that there

Fig. 7. Fluorescence microscope morphology of microcapsules (2.0 wt.%) dispersed in bitumrepeated for 20, 40, and 60 times, each time the sample was heated to 50 �C with the rate ofthe rate of 2 �C min�1.

is a dependence of the failure strength on microcapsule diameter[37]. The ‘yield point’ was selected as a parameter by which toevaluate the plastic deformation. Fig. 8(b,c) shows the opticalmorphologies of a microcapsule with a permanent deformationwith respect to its yield point. Prior to the yield point, a singlemicrocapsule will deform elastically before returning to its originalshape when the applied stress is removed. Once the yield point ispassed, some fraction of the deformation may become permanentand nonreversible. It should be mentioned that the core materialsand environmental temperature will, at the same time, affect theyield points, but these have not been considered in the presentstudy. Because the core and shell materials have different co-efficients of expansion, changes in temperature will result in anextrusion force. Fig. 8(d) shows the ‘yield stress’ of MMF-shellmicrocapsules (under 23 �C) with various mean sizes and shellthicknesses, illustrating their plastic deformation behavior. Theyield stress of a microcapsule with a core/shell ratio of 1/3 is in therange of 1.20e0.70 MPa, whereas the yield stress of a microcapsulewith a core/shell ratio of 1/1 is in the range of 0.75e0.52 MPa. Moreshell material can provide a shell with greater thickness, thusleading to a higher stiffness of elastic microcapsule. This trend issimilar to the results provided by nanoindentation [27].

Fig. 9(a,b) shows the SEM morphologies of microcapsules inbitumenwith plastic deformation. Somemicrocapsules have a largeplastic deformation without evidence of any breaks. This defor-mation may result from a combined effect of the applied force andheat. It can therefore be concluded that the micromechanicalproperties of microcapsules depend on the thickness and micro-structure of the shells. During compression, the volume and Pois-son’s ratio of polymer microcapsules may change continuouslywith the deformation because of their spherical geometry.

3.5. Interface stability of bitumen/microcapsule samples

The above results have proved that the MMF-shell microcap-sules containing rejuvenator have satisfactory thermal stability andelasticeplastic deformation properties. It can be imaged that these

en, (a) original state, (bed) states after a thermal treatment process for each sample2 �C min�1and keeping for 10 min, and then decreased the temperature to �10 �C with

Fig. 8. Plastic deformation behaviors of microcapsules containing rejuvenator, (a) a typical loadedisplacement curve of single microcapsule, (b,c) optical morphologies of amicrocapsule with a permanent deformation over the its yield point, and (c) yield stress of microcapsules with various mean sizes and shell thicknesses.

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e12151212

microcapsules will survive during mixing with bitumen and ag-gregates. However, it is reported that microcracks and interfaceseparation may appear in a repeated vigorous thermal absorbing-releasing process for microcapsule/matrix composites [24,38].During a repeated temperature changes process with heat trans-mission, expansion and shrinkage of the microcapsules andbitumen will occur due to the different expansion coefficients.Microcapsules volume can be affected by the encapsulated reju-venator when the environmental temperature changes. Thesephenomena may cause microcracks or fractures in the bitumenduring heat absorbing or resealing, spoiling the thin microcapsuleshell, thus the encapsulated rejuvenator will lose the shells pro-tection. Theoretical tensile yield strength and ultimate tensilestrength of the composites are different for the cases of adhesionand no adhesion between the filler particles and matrix. In the caseof no adhesion, the interfacial layer cannot transfer stress. Themaincauses of internal thermal stress failure of composites are consid-ered to be the residual stresses due to expansion and shrinkage inthe thermal transmission process and mismatch of moleculemovement among the components. Therefore, it is vital to inves-tigate the bonding stability between bitumen and microcapsules. Athermal absorbing-releasing process was applied in this study toenlarge the interface change degree and interphase morphologieswere observed to give a better explanation of the interfacebehaviors.

A thermal absorbing-releasing process was applied in this studyto enlarge the interface change degrees and interphase morphol-ogies were analyzed to give a better explanation of the interfacebehaviors. The thermal treatment process bitumen samplerepeated for 30 times, each time the sample was heated to 50 �C

Fig. 9. Deformation morphologies of microcap

with the rate of 2 �C min�1 and keeping for 10 min, and thendecreased the temperature to �10 �C with the rate of 2 �C min�1.Fig. 10(a,b) shows the SEM interface morphologies of a bitumen/microcapsule sample before and after the thermal treatment pro-cess. To observe the interface easily, the bitumen were frozen andbroken. It can be seen that the microcapsules keep compactstructure and global shape in both states. There is no interfacedebonding emerged between microcapsules and bitumen after arepeated temperature change. Both composites have a compactinterphase. It means that the temperature changes may not affectthe original adhesive state between shells and bitumen. It is wellknown that strength of composites strongly depends on the stresstransfer between the particles and the matrix. For these well-bonded microcapsules, the applied stress can be effectively trans-ferred to the particles from the bitumen, which will clearly improvethe strength. An interphase region is comprised of polymer mole-cules that are bound at the filled particles surface, and they exhibitunique physical and chemical properties. Polymer moleculeswithin the interphase region are restricted from chain manage-ment, molecular cross-linking and dipole polarization, thus theylose one or more degrees of molecular rotational and/or vibrationalfreedom. The molecule at the interface had modified the aggluti-nation structure, and had moved to fit the violent thermal trans-mission. From the SEM morphologies in Fig. 10, it can be deducedthat the original close interface could resist this molecule modifi-cation and the mismatch in expansion coefficients between theshell and bitumen.

More heat-transmission cycles may generate destruction in theinterface structure. If the interface is only molecular entanglementswithout chemical bonds, the molecule connecting the shell and

sules containing rejuvenator in bitumen.

Fig. 10. SEM interface morphology of bitumen/microcapsules sample, (a) original state and (b) states after a thermal treatment process for each sample repeated for 30 times, eachtime the sample was heated to 50 �C with the rate of 2 �C min�1and keeping for 10 min, and then decreased the temperature to �10 �C with the rate of 2 �C min�1.

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e1215 1213

matrix will undergo structure modification losing its adhesivefunction under long-time repeated thermal-stress effects. There-fore, it is necessary to investigate the chemical bonding withinthe interphase region through molecular structure analysis. Fig. 11illustrates the potential chemical reactions betweenMMF resin andbitumen. In Fig. 11(a,b), further chemical cross-linking reaction willoccur between the methanol-modified MF prepolymer molecularchains. Because this step is complex, the synthetic products aredifficult to be controlled. At the end of the chains, hydroxy (eOH)will not be completely consumed. The complexity of bitumenchemistry lies in the fact that many different chemicals are present.It has been reported [2] that bitumen has three typical functionalgroups (phenolic, anhydride and carboxylic acid) as shown inFig. 11(cee). They all may have the dehydration reactions or

Fig. 11. Chemical sketches of interface bonding structures between MMF-resin shell and(phenolic, anhydride and carboxylic acid) in bitumen, and (f) the possible dehydration reac

esterification reactions with MMF resin. In Fig. 11(f), these reactionscan form chemical bonds between shells and bitumen, which willenhance the interface stability under heat and stress effects. Inaddition, these bonds help microcracks to propagate avoiding to bestopped by interface separation or interphase gaps. Microcrackswill fairly penetrate the interphase and shells and rejuvenator willoutflow the shells.

FT-IR was applied to prove the above interface structures. We donot investigate its in-situ molecular structure as the interphase istoo very thin. Fig. 12 shows the FT-IR curves of cross-linked MMF,bitumen and their mixture, respectively. Curve (a) is the FT-IRspectrum of the MMF prepolymer. Curve (b) is the FT-IR spec-trum of the cross-linked MMF resin. It can determined that thebands at 1598,1012 and 819 cm�1 are C]C stretch, CeO stretch and

bitumen, (a, b) MMF crosslinking structure, (cee) there typical functional groupstions or esterification reactions between the MMF-shell and bitumen.

Fig. 12. FT-IR curves of (a) crosslinked MMF-resin, (b) bitumen, and (c) crosslinkedMMF-resin and bitumen mixtures.

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e12151214

CeH bend, respectively. The strong peaks at 2948 and 2850 cm�1

are alkyl CeH stretch and aliphatic CeH stretch. All these bands canbe used to characterize the existing states of inside or outside ofmicrocapsules, since they do not overlap with other spectral bands

Fig. 13. Morphology of bitumen/microcapsules sample treated by liquid nitrogen, (a) cracknitrogen, (b) original fluorescence microscope morphology of microcapsules in bitumen witcracks after 2 h.

and have reasonable absorbance. The chemical structure of curedMMF prepolymer has a wide absorption peak at approximately3329 cm�1attributing to the superposition of NeH stretchingvibration. The peaks at 1551 and 815 cm�1 are assigned to thevibrations of triazine ring; and the corresponding peaks of curedMMF lie at 1559 and 815 cm�1. Curve (c) shows the FT-IR spectrumof the sample of cross-linkedMMF/bitumen (98/2, w/w) composite.The peaks at 1071e990 cm�1 are assigned to the CeN stretch (aryl).The peaks at 1680e1630 are assigned to the C]O stretch (amides).All these evidences indicate that the MMF resin and bitumen mayhave chemical reaction forming bonds in interphase.

3.6. Microcapsule states in bitumen with cracks

Besides the stability of microcapsules in bitumen, we also needto prove that these microcapsules can break by microcracks. Fig. 13shows the microcapsule states in bitumenwith cracks. As shown inFig. 13(a), microcapsules are keeping compact structure in bitumenwithout defects and microcapsule has the mean size about 10 mm.In Fig. 13(b,c), microcracks were generated by liquid nitrogenquickly with awidth of 10e20 mm.With the cracks propagation, theshell had been split by the tip stress of cracks (Fig. 13(d)). Inter-estingly, it was found that the rejuvenators had rapidly filled thecracks under the capillary action (Fig. 13(e)). Then the rejuvenatorhad permeated through both sides of cracks and the cracks werehealed (Fig. 13(f)). Future work will be carried out about the self-healing behaviors with more details.

s appeared in bitumen sample by the low temperature brittleness treatment of liquidh cracks, (c, d) fluorescence microscope morphology of microcapsules in bitumen with

J.-F. Su et al. / Polymer Degradation and Stability 98 (2013) 1205e1215 1215

4. Conclusions

In this paper, MMF resin was successfully applied to encapsu-lated rejuvenator. These microcapsules were the first productapplied for bitumen self-healing. These microcapsules can bedesigned with various mean sizes and shell thicknesses. It isnecessary to instigate the thermal stability, mechanical stabilityand interface stability of microcapsules in bitumen. Under 180 �Cand 200 �C, the microcapsules still maintained the globe shape andsmooth surface as original state. Microcapsules had survived inbitumen under temperature of 200 �C, which indicates that thesemicrocapsules can resist the thermal effect of bitumen in applica-tion. Microcapsules had the elasticeplastic deformation abilityresisting the temperature changes and mixing stress. Moreover,microcapsules had a large plastic deformation without breaks un-der a combined effect of the force and heat. The chemical bondsimproved the interface stability between shells and bitumen. Withthe microcracks propagation, the shell had been split by the tipstress of cracks and the rejuvenators had rapidly filled the cracksunder the capillary action. Later, the rejuvenator had permeatedthrough both sides of cracks and the cracks were healed. Micro-capsules containing rejuvenator will be a promising product torealize the smart pavements. In the second phase of this work,further investigations will be carried out concerning the rejuve-nator permeation rate, self-healing effective and mechanicalproperties recovery.

Acknowledgments

The authors acknowledge the financial support from the DelftCentre for Materials (DCMat) in the form of Project IOP Self-HealingMaterials SHM1036, “Encapsulated rejuvenator for asphalt”.Dr. Jun-Feng Su also thanks the previous financial support of theNational Natural Science Foundation of China (No. 50803045).

References

[1] Lesueur D. The colloidal structure of bitumen: consequences on the rheologyand on the mechanisms of bitumen modification. Adv Colloid Interf Sci2009;145:42e82.

[2] Branthaver JF, Petersen JC, Robertson RE, Duvall JJ, Kim SS, Harnsberger PM,et al. Binder characterization and evaluation. In Chemistry, vol. 2. Washington,DC: National Research Council; 1993. SHRP-A-368.

[3] Hunter RN. Bituminous mixtures in road construction. London: Thomas Tel-ford; 1997.

[4] Corbett LW. Composition of asphalt based on generic fractionation usingsolvent deasphaltening, elutioneadsorption chromatography and densimetriccharacterization. Anal Chem 1969;41(4):576e9.

[5] Lu X, Isacsson U. Chemical and rheological evaluation of ageing properties ofSBS polymer modified bitumens. Fuel 1998;77(9e10):961e72.

[6] Zargar M, Ahmadinia E, Asli H, Karim MR. Investigation of the possibility ofusing waste cooking oil as a rejuvenating agent for aged bitumen. J HazardMater 2012;233e234:254e8.

[7] Boyer RE. Asphalt rejuvenators: “fact, or fable”. In: Transportation systems2000 (TS2K) workshop, San Antonio, Texas.

[8] Karlsson R, Isacsson U. Material-related aspects of asphalt recycling-state ofthe art. J Mater Civil Eng 2006;18(1):81e92.

[9] Chiu CT, Lee MG. Effectiveness of seal rejuvenators for bituminous pavementsurfaces. J Test Eval 2006;34(5):390e4.

[10] Shen J, Amirkhanian S, Miller JA. Effects of rejuvenating agents on superpavemixtures containing reclaimed asphalt pavement. J Mater Civil Eng2007;19(5):376e84.

[11] Brown EN, Kessler MR, Sottos NR, White SR. In situ poly(urea-formaldehyde)microencapsulation of dicyclopentadiene. J Microencapsul 2003;20:719e30.

[12] Cho SH, White SR, Braun PV. Room-temperature polydimethylsiloxane-basedself-healing polymers. Chem Mater 2012;24:4209e14.

[13] Li H, Wang R, Liu W. Preparation and self-healing performance of epoxycomposites with microcapsules and tungsten (VI) chloride catalyst. J ReinfPlast Comp 2012;31:924e32.

[14] Toohey KS, Sottos NR, White SR. Characterization of microvascular-based self-healing coatings. Exp Mech 2009;49:707e17.

[15] Tong XM, Zhang M, Wang MS, Fu Y. Effects of surface modification of self-healing poly(melamine-urea-formaldehyde) microcapsules on the propertiesof unsaturated polyester composites. J Appl Polym Sci 2013;127:3954e61.

[16] Hansen CJ, White SR, Sottos NR, Lewis JA. Accelerated self-healing via ternaryinterpenetrating microvascular networks. Adv Funct Mater 2011;21(22):4320e6.

[17] Lee J, Zhang M, Bhattacharyya D, Yuan YC, Jayaraman K, Mai YW. Micro-mechanical behavior of self-healing epoxy and hardener-loaded microcap-sules by nanoindentation. Mater Lett 2012;76:62e5.

[18] García A, Schlangen E, van de Ven M, Sierra-Beltrán G. Preparation of capsulescontaining rejuvenators for their use in asphalt concrete. J Hazard Mater2010;184(1e3):603e11.

[19] Su JF, Schlangen E. Synthesis and physicochemical properties of novel highcompact microcapsules containing rejuvenator applied in asphalt. Chem Eng J2012;198e199:289e300.

[20] Su JF, Wang LX, Ren L. Fabrication and thermal properties of microPCMs: usedmelamine formaldehyde resin as shell material. J Appl Polym Sci 2006;101:1522e8.

[21] Wang XY, Su JF, Wang SB, Zhao YH. The effect of interface behaviors on themechanical properties of microPCMs/epoxy composites. Polym Compos2011;32(9):1439e50.

[22] Pan X, York D, Preece JA, Zhang Z. Size and strength distributions ofmelamine-formaldehyde microcapsules prepared by membrane emulsifica-tion. Powder Technol 2012;227:43e50.

[23] Su JF, Wang XY, Dong H. Interface stability of microencapsulated-paraffinfilled epoxy composites: effect of methylation reaction on melamine-formaldehyde shell material. Compos Interf 2011;18(8):645e59.

[24] Su JF, Zhao YH, Wang XY, Dong H, Wang SB. Effect of interface debonding onthe thermal conductivity of microencapsulated-paraffin filled epoxyematrixcomposites. Compos Pt A-Appl Sci Manuf 2012;43(3):325e32.

[25] Su JF, Wang SB, Zhang YY, Huang Z. Physicochemical properties and me-chanical characters of methanol-modified melamine-formaldehyde (MMF)shell microPCMs containing paraffin. Colloid Polym Sci 2011;289(2):111e9.

[26] Hu JF, Chen HQ, Zhang ZB. Mechanical properties of melamine formal-dehyde microcapsules for self-healing materials. Mater Chem Phys2009;118(1):63e70.

[27] Su JF, Wang XY, Dong H. Micromechanical properties of melamine-formaldehyde microcapsules by nanoindentation: effect of size and shellthickness. Mater Lett 2012;89:1e4.

[28] Su JF, Wang SB, Zhou JW, Huang Z, Zhao YH, Yuan XY, et al. Fabrication andinterfacial morphologies of microPCMs/epoxy composites. Colloid Polym Sci2011;289(2):169e77.

[29] Su JF, Schlangen E, Qiu J. Design and construction of microcapsules containingrejuvenator for asphalt. Powder Technol 2013;235:563e71.

[30] Zhu K, Qi H, Wang S, Zhou J, Zhao Y, Su J, et al. Preparation and character-ization of melamine-formaldehyde resin micro- and nanocapsules filled withn-dodecane. J Macromol Sci Phys 2012;51:1976e90.

[31] Fan CJ, Zhou XD. Influence of operating conditions on the surface morphologyof microcapsules prepared by in situ polymerization. Colloid Surf A e Phys-icochem Eng Asp 2010;363(1e3):49e55.

[32] Zhang XX, Tao XM, Yick KL, Fan YF. Expansion space and thermal stability ofmicroencapsulated n-octadecane. J Appl Polym Sci 2005;97(1):390e6.

[33] Mercadé-Prieto R, Nguyen B, Allen R, York D, Preece JA, Goodwin TE, et al.Determination of the elastic properties of single microcapsules usingmicromanipulation and finite element modeling. Chem Eng Sci 2011;66(10):2042e9.

[34] Koleva I, Rehage H. Deformation and orientation dynamics of polysiloxanemicrocapsules in linear shear flow. Soft Matter 2012;8(13):3681e93.

[35] Yan Y, Zhang Z, Stokes Jason R, Zhou QZ, Ma GH, Adams MJ. Mechanicalcharacterization of agarose micro-particles with a narrow size distribution.Powder Technol 2009;192(1):122e30.

[36] Meier M, John E, Wieckhusen D, Wirth W, Peukert W. Influence of me-chanical properties on impact fracture: prediction of the milling behaviour ofpharmaceutical powders by nanoindentation. Powder Technol 2009;188(3):301e13.

[37] Liu KK, Williams DR, Briscoe BJ. Compressive deformation of a single micro-capsule. Phys Rev E 1996;54(6):6673e80.

[38] Wu Y, Huang L, Liu Z, Yang Z. Facile preparation of curved polymer compositenanosheets. Polymer 2012;51:3075e82.