EUROPEAN

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

European Polymer Journal 43 (2007) 2866–2881

www.elsevier.com/locate/europolj

POLYMERJOURNAL

Macromolecular Nanotechnology

Thermal stability, crystallization, structure and morphologyof syndiotactic 1,2-polybutadiene/organoclay nanocomposite

Jiali Cai a,b,*, Qing Yu b, Ying Han a, Xuequan Zhang c, Liansheng Jiang c

a Key Laboratory for Ultrafine Materials of Ministry of Education, School of Materials Science and Engineering,

East China University of Science and Technology, Shanghai 200237, PR Chinab Key Laboratory of Mesoscopic Materials Science, Department of Chemistry and Chemical Engineering, Nanjing University,

Nanjing 210093, PR Chinac State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry,

Chinese Academy of Sciences, Changchun 130022, PR China

Received 1 September 2006; received in revised form 9 April 2007; accepted 23 April 2007Available online 10 May 2007

Abstract

Syndiotactic 1,2-polybutadiene/organoclay nanocomposites were prepared and characterized by thermogravimetryanalysis (TGA), X-ray diffraction (XRD), polarized optical microscopy (POM), and differential scanning calorimetry(DSC), respectively. The XRD shows that exfoliated nanocomposites are formed dominantly at lower clay concentrations(less than 2%), at higher clay contents intercalated nanocomposites dominate. At the same time, the XRD indicates thatthe crystal structures of sPB formed in the sPB/organoclay nanocomposites do not vary, only the relative intensity of thepeaks corresponding to (010) and (200)/(110) crystal planes, respectively, varies. The DSC and POM indicate thatorganoclay layers can improve cooling crystallization temperature, crystallization rate and reducing the spherulite sizesof sPB. TGA shows that under argon flow the nanocomposites exhibit slight decrease of thermal stability, while underoxygen flow the resistance of oxidation and thermal stability of sPB/organoclay nanocomposites were significantlyimproved relative to pristine sPB. The primary and secondary crystallization for pristine sPB and sPB/organoclay (2%)nanocomposites were analyzed and compared based on different approaches. The nanocomposites exhibit smaller Avramiexponent and larger crystallization rate constant, with respect to pristine sPB. Primary crystallization under isothermalconditions displays both athermal nucleation and three-dimensional spherulite growth and under nonisothermal processesthe mechanism of primary crystallization becomes very complex. Secondary crystallization shows a lower-dimensionalcrystal growth geometry for both isothermal and nonisothermal conditions. The activation energy of crystallization ofsPB and sPB/organoclay nanocomposites under isothermal and nonisothermal conditions were also calculated based ondifferent approaches.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Syndiotactic 1,2-polybutadiene; Crystallization; Clay; Differential scanning calorimetry; Polarized optical microscopy; Ther-mal stability

0014-3057/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.eurpolymj.2007.04.036

* Corresponding author. Address: Key Laboratory for Ultrafine Materials of Ministry of Education, School of Materials Science andEngineering, East China University of Science and Technology, Shanghai 200237, PR China. Fax: +86 021 64253539.

E-mail address: jialicai1023@yahoo.com.cn (J. Cai).

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2867

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

1. Introduction

Syndiotactic 1,2-polybutadiene (sPB), first syn-thesized in 1955 by Natta and Corradini [1], is athermoplastic elastomer of industry interest due toits adjustable properties from plastics to rubber. Ithas exhibited various practical applications, andcan be used as films in packaging breathing itemssuch as vegetables, fruits, crabs with puncturebecause of its higher carbon oxide gases and oxygenpermeability and better resistance against wettingand slippage, rubber goods, molding applicationssuch as molded bottles, adhesive, oil paint, photo-sensitive resin, plastics materials, tire treadings,tubes, hoses, and so on [2,3]. The various studieson sPB, thus, are useful from theoretical and indus-trial points of view. Recently, due to the discoveryof new catalyst systems that can produce syndiotac-tic polymers with controllable constitution and con-figuration, many studies have been stimulated andsome work on synthesis, thermal behavior, crystalli-zation, structure and morphology of sPB have beenpublished [4–8].

Nanocomposites consisting of polymer matrixand organoclay have gained great interest becauseof their unique mechanical, thermal, electrical andpermeability properties compared to the pristinepolymers [9,10]. Furthermore, the incorporation ofnanoscale-layer clay into semi-crystalline polymerswould dramatically influence the crystallineproperties including crystallization kinetics, crystalstructure and crystallinity. Nanoclay layer is aneffective heterogeneous nucleating agent for decreas-ing the interfacial free-energy per unit area perpen-dicular to macromolecular chains, which enhancesthe nucleating speed and the crystallization,improves the cooling crystallization temperature,decreases the spherulite size; at the same time, aphysical hindrance of nanoclay layers to the motionof polymer chains limits the crystallization of basalpolymer, resulting in a decrease of crystallinity anddegree of perfection of the crystal [10].

More often, isothermal crystallization is a com-mon practice due to the convenience of theoreticaltreatment on the obtained data, however, isother-mal crystallization are usually performed in a smal-ler temperature window. While, nonisothermalcrystallization studies of polymers are more mean-ingful than isothermal crystallization study becauseit is similar to conventional industrial processing.There have several crystallization studies on pristinesPBs with different syndiotactic regularities [5–7],

One of the authors have studied effects of atacticsPB on the crystallization of sPB [6], while pristinesPB has not been investigated in all respects of itscrystallization behavior, for example, secondarycrystallization. To the best of our knowledge, upto now, there are no studies of crystallization andthermal stability of sPB/organoclay nanocomposite.

In this study, we first prepared films of sPB andsPB/organoclay nanocomposites by in situ vacuumdistillation of the solutions. And then, thermal sta-bility, the clay dispersion states, isothermal andnonisothermal crystallization kinetics of pristinesPB and sPB/organoclay were investigated, respec-tively. The mechanisms of primary and secondarycrystallization under isothermal and nonisothermalconditions were analyzed in terms of the parametersobtained. The thermogravimetry analysis (TGA),X-ray diffraction (XRD) and polarized opticalmicroscopy (POM) were used to investigate thermalstability, crystalline structure and morphology ofsPB and sPB/organoclay nanocomposites.

2. Experimental

2.1. Preparation of pristine sPB and sPB/organoclay

samples

The synthesis and characterization of the sPBinvestigated have been described elsewhere [6].

The samples of syndiotactic 1,2-polybutadiene(sPB) were synthesized by using a new catalyst sys-tem, composed of iron(III) acetylacetonate (Fe-(acac)3), diethyl phosphite (DEP) and Al(i-Bu)3.After the sPB was dissolved in trichlorobenzene(TCB), its molecular weight was determined byhigh-temperature gel permeation chromatographyat 150 �C, to give a weight-average molecularweight (Mw) of 1.28 · 106 g mol�1 and Mw/Mn of2.42.

The organoclay in this work was cloisite 20A(Southern Clay Product Inc). It was a montmoril-lonite organically modified with dimethyl dehydro-genated tallow quaternary ammonium in order toincrease the gallery spacing of the original Na+–montmorillonite. The Cation exchange capacity(CEC) of the clay was 95 mequiv/100 g. For thepreparation of sPB/organoclay nanocomposite, weused a technique different from conventional solu-tion casting and melt mixing techniques. The sPBsamples were first weighted to prepare toluene solu-tions of 10% sPB weight concentration, after thesPB samples were dissolved in toluene at 100 �C

2868 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

(a temperature near the boiling point of toluene),organoclays with different weight ratios were addedto the solutions; a magnetic stirrer was then put intothe solution to let organoclay mechanically blendedfor 24 h at 100 �C with nitrogen purge protection,followed by vacuum distillation of the solutionsfor another 12 h without moving the magnetic stir-rer. As a matter of fact, due to high volatility of tol-uene under the conditions of vacuum distillation at100 �C, toluene in the solutions was taken away in afew minutes. The obtained sample films were thensubjected to pressing under pressure of 15 MPa at210 �C between polytetrafluoroethylene films andwere naturally cooled down to room temperature.The sheets were then used for XRD and POMmeasurements.

2.2. Thermogravimetry analysis

A Perkin–Elmer (PE) Pyris 1 TGA was used toinvestigate the thermal stability of the films of sPBand sPB/organoclay nanocomposites. The sampleswith weight 3.0 ± 0.3 mg were heated at 20 �C/minfrom 100 �C to 700 �C under air purge and argonpurge, respectively. The data analysis was per-formed with the provided software of this computer.or confirmation, repeated experiments were per-formed, good repetitions were observed.

2.3. DSC measurements

A Perkin–Elmer (PE) diamond DSC calibratedwith indium and zinc standards was used to monitorthe crystallization and melting behaviors of the sPBand sPB/organoclay. All operations were performedunder nitrogen purge. Sample weight for eachexperiment is 4.0 ± 0.3 mg.

(I), the sample films obtained from the solutionsby vacuum distillation were subjected to two scan-ning circles: first the samples was heated at 20 �C/min from 50 �C to 200 �C (first heating), kept at200 �C for 3 min, cooled at 20 �C/min from 200 �Cto 50 �C (first cooling), stayed at 50 �C for 3 min,heated at 20 �C/min from 50 �C to 200 �C (secondheating), kept at 200 �C for 3 min, and then cooledat 20 �C/min from 200 �C to 50 �C (second cooling).

(II), for studying the isothermal crystallization ofsPB and sPB/organoclay, the following proceduresare followed: first the samples were heated to200 �C, kept at 200 �C for 3 min to eliminate thethermal history of this sample, then cooled at100 �C/min to the preset temperature to let sample

crystallize, the heat flow was recorded with crystal-lization time.

(III) The experiments of nonisothermal crystalli-zation of sPB/organoclay (2%) were performed asfollows: first the samples was heated at 100 �C/minfrom 50 �C to 200 �C, kept at 200 �C for 3 min toeliminate the thermal history of this samples, fol-lowed by cooling at different rates – 5 �C/min,7.5 �C/min, 10 �C/min, 15 �C/min, 20 �C/min to50 �C.

The melting peak temperature and the heat offusion were calculated by the supplied software ofthe DSC instrument.

2.4. POM measurements

The morphology of various thin film samples ofsPB and sPB/organoclay (2%) was observed andcompared by using a polarized optical microscope(Leika) equipped with a Leika automatic thermalcontrol hot-stage (with the controlling temperatureprecision of ±0.1 �C), a JVC model digital-color-video camera, a liquid-nitrogen cold-trap, and acomputer. The sPB and sPB/organoclay sheets withthickness of 30 lm were sandwiched between micro-scope cover glass and heated to 200 �C, then rapidlycooled at 80 �C/min to the designated temperature(as an example, 148 �C was selected for morphologycomparison of pristine sPB and sPB/organoclay(2%), for other temperatures the difference of pris-tine sPB and sPB/organoclay (2%) were similar),and held there for observing morphology develop-ment of isothermal crystallization. The JVC modeldigital-color-video camera on the Leika POMinstrument and an installed software in the com-puter were used to in situ record and save the spher-ulitic morphology at due interval time in thecomputer.

2.5. X-ray measurements

X-ray diffraction experiments were performed byusing a Rigaku D/max 3C diffractometer withcurved graphite crystal-filtered CuKa1 radiation(k = 0.15406 nm). The sheet specimens were fixedon the rotating sample stage of the goniometer inorder to eliminate any anisotropic effects, and sothat suitable averages of the diffracted intensitiesin reciprocal space could be obtained. Data werecollected over the range 2h = 1–40� in the fixed timemode with a step interval of 0.02�. For betterobtaining information of data in the wide angle

2 4 6 8 10

65

432

Inte

nsity

2θ

2θ

1

10 20 30 40

Inte

nsity

1

234

5

0

6

Fig. 1. Wide angle X-ray diffraction diagrams range of 2h from1� to 10� (a) and 10–40� (b) for sPB/organoclay with differentcontent organoclay. Numbers: 0-pristine sPB; 1–0.5%; 2–1%;3–2%; 4–4%; 5–8%; 6-organoclay.

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2869

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

range of 1–40�, the data of 1–10� and 10–40� werecollected, respectively.

The X-ray diffraction experiments were per-formed under ambient conditions on the organoclayand the sheets of sPB/organoclay with differentratios subjected to melt-pressing at 200 �C to evalu-ate the dispersion states of clay sheets and the struc-ture of sPB in sPB/organoclay nanocomposites.

3. Results and discussion

3.1. X-ray diffraction (XRD)

For investigating dispersion states of organoclaynano-sheets in sPB matrix, the XRD experiments ofsPB/organoclay nanocomposites with differentorganoclay concentrations range from 0.5% to 8%were performed and shown in Fig. 1a. For compar-ison, the XRD data of Cloisite 20A organoclay werealso drawn in Fig. 1a. The 2h angle range is 1–10� inFig. 1a. For Cloisite 20A organoclay, two peakslocated at 3.7� and 7.3�, respectively, are found.The relative weaker peak at 7.3� possibly corre-sponds to the basal spacing of the clay aggregates,ca. 1.2 nm and the peak located at 3.7� correspondsto the gallery spacing of organoclay with 2.4 nm. Asweight percent of the organoclay are 0.5% and 1%,no distinct X-ray diffraction peak was observed(Fig. 1a), indicating the delamination and disper-sion of clay nanoclay into sPB matrix. As the con-centration of organoclay was increased to 2%, thepeak at 7.3� for original organoclay also disappearsand the peak at 3.7� shifts to lower angle, splittinginto two peaks, which are located at 2.5� and 3.1�,respectively. The two peaks correspond to spacingof 3.5 nm and 2.8 nm, which indicates the existenceof intercalated clay due to the entrance of sPBchains into the gallery of organoclay. As theorganoclay content was increased to 4%, a widepeak located at 2.6�, corresponding to 3.4 nm,exists. As the content of organoclay was 8%, itcan be observed that one strong peak at 2.8� witha shoulder peak at 3.9� exists. Fig. 1a also indicatesthat the positions for peaks at lower angles increasewith increasing the content of sPB samples, indicat-ing the increment of the gallery spacing of organo-clay with increasing sPB concentration. It isnoteworthy that for concentrations of organoclaywith 0.5% and 1%, although no distinct diffractionpeak was observed, there exists a wide shoulderpeak in the small-angle region, which could not

exclude the existence of intercalated structures,therefore, possibly indicating the existence of par-tially exfoliated/partially intercalated nanostruc-tures in these sPB/organoclay nanocomposites. Itseems reasonable that Kojima et al. [11] character-ized the relative proportion of exfoliated and inter-calated species with the content of nanoclay usingthe peak intensity together with the interlayer spac-ing (i.e., the peak position). Here from the XRDpatterns of sPB/organoclay, it is possible that theexfoliated structure is the dominant population asthe content of organoclay is below 2%, and aboveit, the intercalated hybrids gradually dominate.

200 400 600

0

50

100

Wei

ght,%

Temperature (oC)

pristine sPB 0.5% 1% 2% 4% 8%

Fig. 2. Thermogravimetric curves of sPB/organoclay with differ-ent contents as indicated under air purge at a heating rate of20 K/min.

2870 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

And between the two extremes, exfoliated structurewith partially structure order is likely formed [12].

Fig. 1b indicates the XRD patterns for sPB/organoclay with different content of organoclay inthe 2h angle range of 10–40�. For comparison pur-pose, the XRD data of the organoclay and pristinesPB sample were added into Fig. 1b to observe thepossible variation. For pristine sPB sample, thereexisted five distinct diffraction peaks, the peak inten-sity decreases with increment of peak position. Thediffraction angles at the maximum intensities corre-spond to interplanar distances of 0.655, 0.549,0.420, 0.380, 0.318 nm according to the sequencefrom low to high angles. The values are close tothose reported by Natta and Corradini [1]. Accord-ing to the orthorhombic cell unit packing of the sPBsample proposed by these authors, the indices forthe peaks from low to high angles correspond to(010), (200)/(1 10), (210), (111)/(201), and(120). Interestingly, as organoclay contents are0.5% and 1%, the peak of (111)/(2 01) crystal planeis very weak, with the other four peaks being clearlyobserved. While as organoclay content is increasedto 2%, the (111)/(20 1) peak again appears, above2%, the peak intensity gradually increases. In addi-tion, with increasing the concentration of organo-clay, the relative intensity of peaks correspondingto (010) and (200)/(1 10) crystal planes varies, asthe content of organoclay is below 2%, the intensityof (01 0) peak is stronger than that of (200)/(1 10)peak. While as the content of organoclay reaches4%, the relative intensity of the two peaks reverses,the intensity of (010) peak is weaker than that of(200)/(1 10) peak. In our previous work [6], it wasfound that atactic 1,2-polymer chains could affectthe relative intensity change of XRD peaks of syn-diotactic 1,2-polybutadiene. Initiated by our experi-mental work [6], Napolitano et al. [8] havecontributed the intensity change of peaks to thepresence of configurational defects in the crystalphase by comparing the experimental and simulatedXRD profiles obtained by using program Cerius2

software. While that the intensity of (111)/(2 01)peak for sPB/organoclay nanocomposites with0.5% and 1% organoclay becomes so weak is stillkept inexplicit, the mechanism study of peak inten-sity change is in progress. Although the peak inten-sity change with organoclay content was observed,but the peak positions were kept almost unchange-able, this is to say, the crystal structure of sPBsample does not change due to the addition of

organoclay, only affecting the crystal perfectionformed.

4. Thermogravimetry analysis (TGA)

The thermal stabilities of pristine sPB and nano-composites have been investigated by TGA

under Argon and air flows, respectively. Since thepractical application is under air atmosphere, here,we only show the traces under air purge (Fig. 2).The decomposition temperatures for weight loss of10% in both argon (T0.1-Argon) and air (T0.1-air) arecollected and listed in Table 1. Table 1 indicates thatunder argon flow the decomposition temperatureindicates a little decrease, especially for 8% organo-clay content, the T0.1-Argon decreases from 441.4 �Cto 424.7 �C. While under air flow the nanocompos-ites indicate improved thermal stability. T0.1-air dis-tinctly increases from 336.5 �C for pristine sPB tothe range between 430.3 �C and 444.9 �C, dependingon the content of organoclay for nanocomposites.For confirmation, the data of these samples wereperformed again, good repetitions were observed.Since practical processing is under air atmosphere,organoclay is a possible candidate to improve thethermal stability of sPB materials.

At the same time, it can be observed from Fig. 2that the TGA curve for pristine sPB is a little differ-ent from those of nanocomposites under air purge.At 183.6 �C, the weight percent of pristine sPB sam-

Table 1Thermogravimetry analysis parameters of sPB/organoclay with different contents

Content of clay (%) T0.1-Ar (�C) T0.1-Oxygen (�C) TOnset oxidation (�C) WEnd oxidation (%)

0 441.4 336.5 183.6 109.80.5 436.1 442.6 210.1 100.61 441.6 444.9 216.3 100.42 436.1 442.7 207.6 101.94 435.7 434.2 196.9 102.08 424.7 430.3 202.2 102.4

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2871

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

ple quickly increases from 100% to 109.8%, this ispossibly due to the occurrence of oxidation reactionof the vinyl group in air atmosphere (The conclu-sion was drawn from the following facts: one is that,under nitrogen flow, we could not observe the phe-nomena of weight addition with increasing temper-ature, the other is that under air atmosphere theweight-adding phenomena could be observed, indi-cating that oxidation reaction occurs on the sPBchains under air atmosphere at higher temperature.In addition, thermogravimetry data of low densitypolyethylene (LDPE) under air atmosphere (thegraph is not shown here) indicate that there is noweight change before 261 �C (>183.6 �C – the onsetof weight addition of sPB), indicating that the meth-ylene, methine on the sPB chain does not contributeto the oxidation reactions of sPB samples. And, fur-thermore, a FTIR experiment (not shown as well)was designed to indicate that after the sPB samplewas heated at 20 �C/min under air atmosphere to220 �C, stay at the temperature for 12 s, and thencooled to room temperature at 10 �C/min, at1721 cm�1, an additional peak attributed to oxida-tion reactions of vinyl groups appears in the FTIRspectroscopy relative to FTIR data of the sPBsample without subjecting to heating at high tem-perature. These facts indicate that the weight-increasing is due to oxidation of vinyl groups ofsPB chains. But further details need more experi-mental supports, the related works are in progress).And then at 264.7 �C the weight of the pristine sPBbegins to decrease, until 535 �C is reached, at whichthe sample is decomposed completely. For nano-composites, the onset temperature of the weightincrease due to oxidation was greatly delayed, fur-thermore, the increase amplitude in weightdecreases greatly from 109.8% for pristine sPB tothe range between 100.4% and 102.4% dependingon the content of organoclay. This means that dur-ing heating the chance of vinyl groups along the sPBchain being oxidized was greatly reduced due to theexistence of nanoclays. The data of the onset oxida-

tion temperature (Tonset oxidation) and the maximumweight increase (Tend oxidation) due to oxidation werealso listed in Table 1, together with T0.1-argon andT0.1-air.

5. Differential scanning calorimetry

5.1. Effects of organoclay content on crystallization

and melting

The films obtained by vacuum distillation at100 �C were subjected to 2 heating/cooling DSC cir-cles, good repeated curves for first and second cool-ing scans were observed, indicating that staying at200 �C for 3 min could eliminate the previous ther-mal history of pristine sPB and sPB/organoclaynanocomposites. The first heating curves for sPBblended with different contents organoclays (0.5%,1%, 2% 4%, 8%) were indicated in Fig. 3a. For com-parison, the heating curves for pristine sPB filmsobtained by vacuum distillation at 100 �C (indicatedas number ‘1’) and melt crystallized at 100 �C for2 h (indicated as number ‘0’), respectively, wereadded into Fig. 3a as well. It indicates that meltingtemperatures of the samples obtained from the solu-tions by vacuum distillation at 100 �C are far higherthan that obtained by melt-crystallization at 100 �C,indicating more perfect crystals formed from tolu-ene solutions at 100 �C than those obtained fromthe melt-crystallization at the same temperature,the conclusion can also be supported by the secondheating curves, which indicate apparent decrease ofmelting peak temperature relative to the first heat-ing curves (Fig. 3a and b). The melting temperatureparameters for the first and second heating curvesare collected and summarized in Table 2. Exceptthe difference of crystal perfections between solu-tion- and melt-crystallization, it was surprising tofind that the melting enthalpies obtained fromtoluene solutions are far higher (55.4 ± 3.5 J/g)than those obtained from melt-crystallization(36.8 ± 2.4 J/g), increasing by ca. 50%. The melting

120 140 160 180

5

6

4

3

2Hea

tflow

End

o U

P

1

100 120 140 160 180

3

21

5

4

6

Hea

tflow

End

o U

P

Temperature ( C)o

Temperature ( C)o

0

Fig. 3. First (a) and second (b) heating scan at 20 �C/min of sPBwith different content of organoclay. Numbers: 0, melt-crystal-lization of pure sPB at 100 �C for 2 h; 1, the samples wereprepared from pure sPB toluene solution with 10% concentrationat 100 �C through vacuum distillation with a magnetic stirrerkept working, the time was lasted for 12 h; the same operationprocedure as ‘1’ was applied for toluene solution with 10%concentration of sPB/organoclay (0.5%)–2, sPB/organoclay(1%)–3, sPB/organoclay (2%)–4, sPB/organoclay (4%)–5, sPB/organoclay (8%)–6, respectively.

Table 2Melting parameters for the sPBs with different contents oforganoclay

Content ofclay (%)

Tpeaka

(�C)DH

(J/g)aTpeak

b

(�C)DH

(J/g)b

0 168.67c 31.8c 162.62 39.4174.44 52.6 162.69 39.2

0.5 175.69 53.9 165.34 35.31 180.30 54.3 171.02 34.82 177.98 58.9 170.85 35.74 178.14 57.2 169.93 37.58 177.36 55.3 170.42 38.1

a First heating.b Second heating.c Pure sPB melt crystallized at 100 �C for 2 h.

100 120 140 160

4

3

5

2

6

Hea

tflow

End

o U

P

Temperature (oC)

1

Fig. 4. Cooling curves at 20 �C/min of sPB with differentcontents of organoclay. The numbers (1–6) have the samemeanings as those in Fig. 3.

2872 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

enthalpies were also listed in Table 2. The meltingenthalpy for pristine sPB melt crystallized at100 �C for 2 h was calculated from the higher melt-ing peak without considering the contribution of thelower temperature melting peak, which is due to theformation of thinner and less perfect crystals [13]. Inaddition, for the second heating curve of sPB with

1% organoclay, a shoulder peak, as shown with anarrow, appears; Table 2 also indicates that for boththe first and second heating curves, below 1%organoclay, melting peak temperature (Tmp)increases with content of organoclay, above 1%,Tmp decreases by 1–2 �C with increasing concentra-tion of organoclay.

Fig. 4 compares heat flow development duringcooling at 20 �C/min for sPB/organoclay nanocom-posites. The corresponding crystallization parame-ters extracted and listed in Table 3 are the onsetcrystallization temperature (Tonset), end crystalliza-tion temperature (Tend), peak temperature (Tmp),the full width at half height of the exothermic peak(TWHH), and crystallization exothermic enthalpy(DH). Table 3 indicates that as 0.5% organoclay

Table 3The corresponding crystallization parameters for sPB withdifferent contents of organoclay

Content ofclay (%)

Tonset

(�C)Tend

(�C)Tmp

(�C)TWHH

(�C)DH

(J/g)

0 145.4 108.1 133.1 9 37.50.5 149.8 118.3 141.0 7.5 34.51 151.1 118.1 142.8 7.2 34.12 151.7 118.5 144.2 5.4 33.94 151.9 118.8 144.4 4.4 34.98 152.1 118.1 144.7 4.5 33.4

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2873

LE

CU

LA

RN

AN

OT

EC

HN

OL

OG

Y

was added into sPB matrix, Tonset, Tend, Tp dis-tinctly improved, TWHH decreases from 9 �C to7.5 �C, and DH decreases slightly; increasing thecontent of organoclay in the sPB matrix, Tonset, Tp

increase slightly, Tend and DH vary little, and TWHH

decreases until the content of organoclay reaches4% and becomes level off above 4%. These data indi-cates that the organoclay affects greatly the crystal-lization of sPB during cooling from the melt.

5.2. Isothermal crystallization

For further studying effects of organoclay oncrystallization of sPB, the nanocomposites of sPB/organoclay (2%) were selected for comparison withpristine sPB. Five crystallization temperatures(146 �C, 148 �C, 150 �C, 152 �C, 154 �C) were cho-sen. Fig. 5 gives the dependence of relative crystallin-ity X(t) for sPB/organoclay (2%) on crystallizationtime at five different temperatures. About 60 data

0 5 10 15 20 25 300.0

0.5

1.0 152 C148 C 150 C 154 C

Rel

ativ

e C

ryst

allin

ity

Time (min)

146 C

Fig. 5. The relative crystallinity development with crystallizationtime for sPB/clay (2%) at different temperatures. The dotted andsolid lines are the calculated ones from the standard and series–parallel Avrami models, respectively.

MA

CR

OM

O

points for each crystallization temperature areshown as geometrical points. It is clear that the timeto reach the ultimate degree of crystallinity increaseswith increasing crystallization temperature, Tc. Thecrystallization-half time, t1/2,the important crystalli-zation parameter, is defined as the elapsed time fromthe onset of crystallization until the crystallizationreaches half of the whole crystallization contribu-tion, it can be obtained directly from Fig. 5. Table4 listed the t1/2 data from Fig. 5. At the same time,the isothermal crystallization-half time, t1/2, for pris-tine sPB at the five temperatures are summarized inTable 4. It can be observed that t1/2 values for bothpristine sPB and sPB/organoclay (2%) increase withincreasing Tc at the studied temperature range; fur-thermore, the t1/2 for pristine sPB is larger than thatfor sPB/organoclay (2%) at the same crystallizationtemperature. Generally, the rate of crystallizationcan be described as the reciprocal of t1/2. Thus, thisindicates that the crystallization rate of sPB/organo-clay (2%) is faster than that of pristine sPB at thesame crystallization temperature, and the crystalliza-tion rate for both pristine sPB and sPB/organoclay(2%) decreases with increasing crystallization tem-perature, characteristic of nucleation-controlledprocess.

The whole crystallization process of a semi-crys-talline polymer is generally separated into two mainstages: primary crystallization and secondary crys-tallization. And primary crystallization is the mac-roscopic development of crystallinity degree due totwo consecutive microscopic mechanisms, i.e., pri-mary nucleation and subsequent crystal growth.The primary process of isothermal crystallizationof a polymer can be described by the well-knownAvrami equation [14,15]:

X t ¼ 1� expð�ktnÞ ð1Þ

where Xt is the relative crystallinity at time t, k iscrystallization rate constant, and n is the Avramiexponent which indicates the nucleation mechanismand growth dimensions. Depending on the mecha-nism of nucleation and type of crystal growth, dif-ferent integer values of n can be obtained.

As the Avrami equation only hold well in theearly stages of isothermal crystallization, the dataanalysis based on the Avrami model was performedby directly fitting the early Xt for each crystalliza-tion temperature to Eq. (1). The kinetic parametersof pristine sPB and sPB/organoclay (2%) could beobtained by using the non-linear regression method.The standard deviation (SD) is defined as [16]:

Table 4Primary crystallization kinetic parameters for sPB/clay (2%) and pure sPB based on the Avrami model

Tc (�C) t1/2 (min) t��1=2 (min) n k (min�n) SD

146 0.51 (0.82) 0.52 (0.79) 2.07 (2.21) 2.71 (1.17) 7.31 · 10�3 (5.25 · 10�3)a

148 0.68 (1.13) 0.65 (1.26) 2.33 (2.75) 1.90 (0.37) 3.64 · 10�3 (4.31 · 10�3)150 1.10 (2.05) 0.89 (2.24) 2.18 (2.65) 0.89 (0.082) 8.65 · 10�3 (6.24 · 10�3)152 2.65 (4.02) 1.76 (4.23) 2.31 (2.54) 0.19 (0.018) 1.80 · 10�3 (3.26 · 10�3)154 6.23 (7.66) 5.39 (7.95) 2.13 (2.92) 0.019 (0.0016) 6.50 · 10�3 (5.96 · 10�3)

a The data in the bracket are those for pristine sPB.

2874 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

SD ¼Xn

i¼1

ðX i;cal � X i;expÞ=ðn� pÞ( )1

2

ð2Þ

where n is the number of data points and p is the to-tal number of independent and dependent variables.The subscripts, ‘cal’ and ‘exp’, indicate the calcu-lated and experimental values of the i-point, Xi,respectively.

The obtained kinetic parameters (n and k), alongwith the corresponding SD values for pristine sPBand sPB/organoclay (2%), are summarized in Table4. The fractional value of n obtained from experi-mental data is due to the various simplificationsassumed in the original derivation of Avrami equa-tion which do not necessarily applied to all polymersystems under all crystallization conditions [13]. Inthis study for pristine sPB and sPB/organoclay(2%), the n values for sPB range from 2.21 to 2.92and those for sPB/organoclay (2%) from 2.07 to2.33, indicating three-dimensional growth by ather-mal nucleation mechanism during primary crystalli-zation. For the same crystallization temperature,sPB/organoclay (2%) characterizes smaller n value,indicating that the organoclay acted as nucleationagents during primary crystallization process.

The values of the crystallization rate constant k forboth pristine sPB and sPB/organoclay (2%) decreasewith increasing Tc, and k for pristine sPB is smallerthan that for sPB/organoclay (2%) at the same tem-perature. From Arami exponent n and crystallizationrate constant k, the crystallization time t��1=2 can be cal-culated using the following formula:

t��1=2 ¼ln 2

k

� �1n

ð3Þ

the calculated values of the crystallization half time,t��1=2, for pristine sPB and sPB/organoclay (2%) arealso summarized in Table 4 for comparison. It indi-cates that the calculated t��1=2 value for pristine sPB islarger than that for sPB/organoclay (2%) at thesame temperature.

Secondary crystallization is mainly concernedwith the crystallization of interfibrillar, crystalliz-able melt, which has been rejected and trappedbetween the fibrillar structures. The model of thesecondary crystallization occurring with spheruliteswas first suggested by Hiller [17] and Price [18] toanalyze the isothermal crystallization data of highperformance polymers such as poly(ether etherketone) (PEEK), poly(ether ketone ketone) (PEKK)and poly(phenylene sulfide) (PPS). Velisaris andSeferis [19] proposed a parallel Avrami model toconsider secondary crystallization. By consideringinter- and intra- spherulite in the secondary process,a ‘series–parallel’ Avrami model was recently pro-posed by Woo and Yau [20]. The primary crystalli-zation is assumed to hold the following relationship:

X t ¼ xp½1� expð�kptnpÞ� ð4Þ

The secondary crystallization which occurs at t > td

is assumed to hold the following relationship:

X t ¼xp½1� expð�kptnpÞ�þxsf1� exp½�ksðt� tdÞns �gð5Þ

where the subscript ‘p’ indicates the primary crystal-lization starting at t = 0 and the subscript ‘s’ indi-cates the secondary crystallization starting at t =td (the latter being the onset time of secondary crys-tallization); np, ns are the Avrami exponents, kp andks are the rate constants, and xp, xs are themass fractions, in all cases for the primary and sec-ondary crystallization, respectively. Apparently,xs = 1 � xp.

For obtaining the kinetic parameters in Eq. (5),td, the onset time of secondary crystallization,should be first obtained. Lu and Hay [21] obtainedtd by adjusting the relative crystallinity at the endof the primary process until the Avrami exponentremained essentially constant. Rybnikar [22] pro-posed that td was in agreement with the value ofthe two half-times of primary crystallization, i.e.,the end of primary crystallization. Woo and Chen

-1.5 -1.0 -0.5 0.0 0.5

-2

-1

0

1

log

[-ln

(1-X

t)]

logt (min)

logtd

Fig. 6. A typical representation of the onset time of secondarycrystallization td during isothermal crystallization process of sPB/clay (2%) at 146 �C.

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2875

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

[23] suggested that td was defined as the time atwhich the discontinuity in the Avrami plot waslocated. In this study, we have adopted the lattermethod to extract the td of sPB and sPB/organoclay(2%) at different crystallization temperatures. Asshown in Fig. 6, the primary and secondary crystal-lization processes occurred consecutively and wereseparated at the critical time of td, when a suddenchange of the slope, i.e., the Avrami exponent, fromca. 2.1 to 1.3 occurred.

Data analysis based on the ‘series–parallel’Avrami model was carried out by fitting the experi-mental Xt data obtained for each temperature toEqs. (4) and (5). The kinetic parameters, along withthe SD values calculated for Xt ranging from 0.01 to0.99 are listed in Table 5. The values of np vary fromca 2.1 to 2.5 for sPB/organoclay (2%) and from 2.3to 2.7 for pristine sPB. and the values of ns range

Table 5Crystallization kinetic parameters for sPB/clay (2%) and pure sPB base

Tc (�C) xp np kp (min�np) ns

146 0.93 2.14 3.21 1.34(0.91)a (2.29) (1.19) (1.35)

148 0.80 2.34 2.45 1.46(0.76) (2.67) (0.38) (1.48)

150 0.62 2.17 1.29 1.27(0.56) (2.52) (0.085) (1.37)

152 0.56 2.09 0.36 1.16(0.52) (2.56) (0.019) (1.46)

154 0.51 2.48 0.11 1.34(0.48) (2.72) (0.0018) (1.58)

a The data in the bracket are those for pristine sPB.

from ca 1.1 to 1.5 for sPB/organoclay (2%) andfrom 1.3 to 1.6 for pristine sPB. Comparing the val-ues of np and ns between pristine sPB and sPB/organoclay (2%), pristine sPB has larger values.For both sPB and sPB/organoclay (2%), ns is alwayssmaller than the np at the same temperature, this isdue to the fact that the growth geometry of crystal isrestricted because of impingement and confinementof crystals formed during primary crystallizationprocess. Thus the crystals grow in lower dimensionbetween the fibrillar structures during the laterstage.

The values of ks for both pristine sPB and sPB/organoclay (2%) depend on the Tc and change insame trend as the values of kp for the primary crys-tallization. Whether kp or ks, the smaller values ofkp and ks for pristine sPB are characterized relativeto those for sPB/organoclay (2%). The xp decreaseswith the increasing Tc for both pristine sPB andsPB/organoclay (2%), this indicates that a faster pri-mary crystallization process was accompanied by adecrease in the mass fraction for secondary crystal-lization. It is noteworthy that the time for the onsetof secondary crystallization td increases with Tc,except for 148 �C and 150 �C of sPB/organoclay(2%), as shown in Table 5. Comparing the td andthe tpeak(the time at which the maximum crystalliza-tion rate is reached), td is larger than tpeak. This is ingood agreement with the assumption of Karmal andChu [24] that secondary crystallization starts aftertpeak for polyethylene samples.

The calculated curves using the ‘conventional’Avrami model and the ‘series–parallel’ Avramimodel were drawn, as sPB/organoclay (2%) sampleswere selected as an example, the calculated curvesand experimental data were compared (as shownin Fig. 5). It indicates that the ‘series–parallel’

d on the ‘series–parallel’ Avrami model

ks min�ns td min tpeak min SDa

1.78 0.92 0.43 2.56 · 10�3

(1.16) (1.56) (0.67) (4.25 · 10�3)1.65 0.86 0.52 4.52 · 10�3

(0.11) (2.19) (1.16) (5.23 · 10�3)0.59 0.83 0.57 2.57 · 10�3

(0.023) (2.36) (1.98) (3.65 · 10�3)0.29 1.93 1.11 5.41 · 10�3

(0.0072) (4.01) (2.96) (4.12 · 10�3)0.09 3.83 3.50 4.89 · 10�3

(0.0016) (6.32) (4.31) (3.21 · 10�3)

2876 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

Avrami model fits better with the experimental datathan the ‘conventional’ Avrami model does.

As the crystallization process is assumed to bethermally activated, the crystallization rate constantk can be approximately described in the Arreheniusform, as follows [25]:

k1=n ¼ k0 expð�DE=RT cÞ ð6aÞ

or1

nln k ¼ ln k0 � DE=RT c ð6bÞ

where k0 is the temperature-dependent pre-exponen-tial factor, DE is the total activation energy whichincludes the transport activation energy and nucle-ation activation energy at the crystallization temper-ature Tc, R is the universal gas constant. Theactivation energy DE can be determined by multi-plying �R of slope obtained by fitting linear the plotof 1/n lnk versus 1/Tc. With this technique, the ki-netic parameters for primary and secondary crystal-lization process are analyzed and listed in Table 5(Fig. 7). The values of the activation energy for pri-mary and secondary crystallization of pristine sPBand sPB/organoclay (2%) were determined to be�440 (�665) kJ/mol, �275.9 (�479) kJ/mol, respec-tively. The values in the bracket designate those ofsecondary crystallization. It is clear that whetherpristine sPB or sPB/organoclay (2%), the DE forthe secondary crystallization is larger than that forthe primary crystallization process, so indicatingthat secondary crystallization from the crystalliz-able melt in the interfibrilles is more difficult thanprimary crystallization from the melt. This observa-tion was interpreted as follows. With the crystalliza-

2.34 2.36 2.38 2.40

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1/nl

nk

1/Tc(10-3K-1)

2.34 2.36 2.38-4

-3

-2

-1

0

(1/n

)lnk

(1/Tc)(10-3K-1)

Fig. 7. Arrhenius plots for sPB/clay (2%) and pure sPB. j,primary crystallization; d, secondary crystallization.

tion process proceeding, rejection of the non-crystallizable material is also concentrated betweenthe interfibrilles, which impeded diffusion of thecrystallizable material Thus, the secondary crystalli-zation became difficult due to the impeded diffusionof the crystallizable species and by a hinderingmechanism between interfibrilles. Bertini et al. [5]have reported the value of activation energy�509 kJ/mol for primary crystallization process ofa sPB with higher syndio-regularity, based onArrhenius analysis.

In the case of isothermal crystallization, the truelink of the obtained Avrami exponent from DSCmeasurement and crystallization mechanism canbe further examined by morphological observations,for example, by polarized optical microscopy. So inthis study, we investigated the morphology of sPBand sPB/organoclay (2%) using polarized opticalmicroscopy (POM). The representative end mor-phologies of sPB and sPB/organoclay (2%) crystal-lized at 148 �C for 30 min were shown in Fig. 8.For sPB, the diameters of the spherulites range from40 to 80 lm, while, for sPB/organoclay (2%) thediameters of the spherulites become more uniformand smaller, as expected. Comparing the lineargrowth rate of the spherulites of sPB and sPB/organoclay (2%) at the same crystallization temper-ature, the growth rate of sPB/organoclay (2%) wasgreatly enhanced. For other crystallization tempera-tures, the same difference between sPB and sPB/organoclay (2%) was observed. In this study,because the thickness of DSC films is approximately300 um, far larger than the final spherulite diameter.Therefore, it is confirmed that the isothermal crys-tallization processes of sPB and sPB/organoclay(2%) investigated by DSC measurement exhibitthree-dimensional spherulite growth rather thantwo-dimensional spherulite growth.

5.3. Nonisothermal crystallization

Several nonisothermal studies of sPB with differ-ent syndio-regularity have been reported [5–7]. Inthis section, we will investigate nonisothermal crys-tallization process of sPB/organoclay (2%) to inves-tigate effects of added organoclay on crystallizationof sPB. Fig. 9 presents the nonisothermal DSCcurves of sPB/organoclay (2%). The correspondingparameters obtained from these curves are summa-rized in Table 6. To and Te represent the onsettemperature of crystallization and the end tempera-ture of crystallization, respectively, Tp is the peak

Fig. 8. Representative polarized optical micrographs of pure sPB (a) and sPB/clay (2%) isothermally crystallized at 148 �C, respectively.

120 140 160 180

20oC/min

15oC/min

10oC/min7.5oC/min

Hea

tflow

End

o U

P

Temperature (oC)

5oC/min

Fig. 9. Heat flow curves of sPB/clay (2%) under nonisothermalcrystallization processes at indicated cooling rates.

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2877

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

temperature at which the crystallization rate was atthe maximum. The time tmax corresponds to Tp, andDHc is the crystallization enthalpy. It can be seenthat the To, Te, Tp shift to a lower temperatureregion. This indicates that the mechanism of crystal-lization of sPB/organoclay (2%) is dependent on thecooling rate: at higher cooling rates, the activation

Table 6Fundamental parameters obtained from the nonisothermal crystallizati

Rate (�C/min) T0 (�C) Te (�C) Tp (�C

5 153.5 139.2 149.57.5 152.9 136.2 148.2

10 152.6 134.1 147.515 152.2 132.0 146.420 152.0 127.0 144.2

of nuclei occurs at lower temperatures, whereaswhen the samples are cooled at lower cooling rates,crystallization occurs at higher temperatures.

Fig. 10 indicates the temperature-dependence ofrelative crystallinity function Xt at different coolingrates. The relationship between the temperature T

and time t can be established by:

t ¼ T 0 � TU

ð7Þ

where T is the temperature at time t, T0 is the onsetcrystallization temperature (the temperature atwhich heat flow change of DSC exothermic curvecan be detected, t = 0). The temperature abscissaaxis shown in Fig. 10 can be transferred into a timescale (Fig. 11). It shows that increasing cooling rateresults in shorter time for crystallization to becompleted.

Nonisothermal crystallization theory and meth-ods can have practical significance both in the simu-lation of technological processes and in thedetermination of parameters of the crystallizationkinetics of polymers. For obtaining the kineticparameters of crystallization kinetics of poly-mers, several formulae have been suggested [26–29], most based on the ‘conventional’ Avrami model.Although the Avrami equation is often used to studythe isothermal crystallization behavior of polymers,

on of sPB/clay (2%)

) tmax (min) t1/2 (min) DHc (J/g)

0.78 0.90 33.40.61 0.68 33.20.50 0.57 33.50.36 0.40 34.90.27 0.37 33.9

130 140 1500

20

40

60

80

100R

elat

ive

Cry

stal

linity

, %

Temperature ( C)o

20oC/min 15oC/min 10oC/min 7.5oC/min5oC/min

Fig. 10. The relative crystallinity as a function of temperature forthe nonisothermal melt crystallization of spB/clay (2%) atdifferent cooling rates.

0.0 0.5 1.0 1.5 2.0 2.50

20

40

60

80

100

Rel

ativ

e C

ryst

allin

ity,%

Time (min)

5oC/min 7.5oC/min10oC/min15oC/min20oC/min

Fig. 11. The relative crystallinity as a function of time for thenonisothermal melt crystallization of spB/clay (2%) at differentcooling rates.

-2 -1 0-6

-4

-2

0Regime 3

Regime 2

log[

-ln(1

-Xt)]

logt (min)

5oC/min 7.5oC/min 10oC/min 15oC/min 20oC/min

Regime 1

Fig. 12. Plots of log[�ln(1�Xt)] versus log t for the nonisothemalmelt crystallization of sPB/clay (2%) at different cooling rates.

2878 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

it has also been applied directly to describe the non-isothermal crystallization of polymers. Using Eq. (1)in double logarithmic form and plottinglog{�ln[1�Xt]} versus log t for each cooling rate,the curves are shown in Fig. 11. This non-linearityof the plots indicates that the mode of spheruliticnucleation and growth is more complex than thatin the isothermal crystallization process for sPB/organoclay (2%) nanocomposite and the Avramiexponent and crystallization mechanism dependson the temperature even in the early stages. All plots

shown in Fig. 12 exhibit three-regimes behavior,with slope of regime 2 being much larger than thoseregimes 1 and 3. The regimes 1 and 2 correspond toprimary crystallization, whereas regime 3, whichusually proceeds over much longer time, is attributedto the occurrence of secondary crystallization. Thisalso indicates that secondary crystallization, oftenoccurring under isothermal crystallization, alsooccurs under nonisothermal crystallization. If con-trasting primary crystallization under isothermalcrystallization with that under nonisothermal crys-tallization, the conditions under nonisothermal crys-tallization are more complex. The kinetic parametersobtained are summarized in Table 7. For primarycrystallization, the values of Avrami exponent np1

in the regime 1 range from 2.71 to 3.04, which indi-cate that, in all cases, crystallization is initiated by anathermal nucleation and proceeds with three-dimen-sional spherulite growth. This process occurs athigher temperatures and the results are in accor-dance with the primary crystallization mechanismunder isothermal crystallization (as indicated inTables 3 and 4). The values of Avrami exponentnp2 in regime 2 vary from 3.29 to 4.71, which indi-cates that as the crystallization time is prolongedand the crystallization temperature decreases, thespherulites become branched, and may lead to den-drite formation. Thus, a crystal in this regime growsin a higher dimension. At secondary crystallization,the values of ns range from 1.39 to 1.98, suggestingthat, in the later stage, the crystals impinge andcrowd; the growth geometry of the crystal is then

Table 7Kinetic parameters obtained from Avrami analysis of the nonisothermal melt-crystallization process for sPB/clay (2%)

Rate (�C/min) np1 logkp1 (min�np) np2 logkp2 (min�np) ns logks (min�np) Td(�C) td (min)

5 2.81 �0.72 3.29 �0.55 1.98 �0.35 143.6 1.277.5 2.80 �0.46 3.97 0.39 1.91 0.12 142.3 0.95

10 2.71 �0.36 4.45 1.01 1.96 0.42 141.8 0.8715 2.71 0.23 4.30 1.59 1.41 0.57 140.5 0.6320 3.04 0.53 4.71 2.58 1.39 0.85 139.6 0.56

0.0 0.2 0.4 0.6 0.8 1.0-320

-280

-240

-200

Effe

ctiv

e ac

itvat

ion

ener

gy, Δ

E(kJ

/mol

)

Relative crystallinity

Fig. 13. The activation energy of crystallization as a function ofrelative crystallinity for sPB/clay (2%) obtained from DSC data.

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2879

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

restricted and the crystals grow in lower dimensionin this regime.

Additionally, the onset temperature of the sec-ondary crystallization, Td, is lower than Tp (the tem-perature at which the crystallization rate reaches amaximum) and the onset time of secondary crystal-lization, td, is higher than tmax (the time at which thecrystallization rate reaches the maximum). This isalso in agreement with data from isothermal crystal-lization process. The rate constants in these threeregimes are dependent on the cooling rate andchange in the same relative way. The values of xp

decrease with increasing cooling rate. It wouldappear that a higher cooling rate is accompaniedby an increase in the mass fraction for secondarycrystallization. This may be explained as follows.An increase of cooling rate leads to decrease ofabsolute crystallinity during primary crystallization,and thus an increase in the fraction of amorphouscomponents between and within the spherulites.So the mass fraction for further crystallizationincreases with increasing cooling rate.

It should be pointed out that for the nonisother-mal crystallization process of sPB/organoclay (2%),the mechanism of crystallization is relative complex.The crystallization was divided into three regimes,primary crystallization includes different crystalliza-tion mechanism. The well-known Ozawa equation[25–29], the ‘series–parallel’ Avrami equation sug-gested for isothermal crystallization cannot satisfac-torily describe the nonisothermal crystallizationprocess of sPB/organoclay (2%). As nonisothemalcrystallization of polymer and polymer/organoclaynanocomposites has not received the same attentionas the isothermal crystallization. For improving thepredication of the parameters of dynamic solidifica-tion under real processing conditions, much atten-tion in theory and experiments should be paid tothe nonisothermal crystallization processes whichpossess different mechanisms of crystallization.

From nonisothermal crystallization experimentsof DSC. The activation energy is often estimatedby using the Kissinger method. Recently, Vyazov-

kin and Sbirrazzuoli [30–32] has suggested that theKissinger equation is no applicable to the noniso-thermal crystallization processes that occur on cool-ing and the differential isoconversional method ofFriedman [33] can be utilized to determine the acti-vation energy by

dX t

dt¼ A exp �DE

RT

� �f ½X t� ð8aÞ

or lndX t

dt

� �¼ ln½Af ðX tÞ� �

DERT

ð8bÞ

where Xt is the relative crystallinity, t is the time, A

is the pre-exponential factor, and f(Xt) is the crystal-lization model. At a fixed relative crystallinity, theslope of fitting linear on the plot of ln[dXt/dt] versus

1/T can be obtained to be DE/R. The values of theactivation energy DE (shown in Fig. 13) for sPB/organoclay (2%) nanocomposites are changed withthe increase of the relative crystallinity from�307.8 kJ/mol to �199 kJ/mol. In the literature,Ren et al. [7] have obtained the activation energyof a pristine sPB, ranging from �208 to �142 kJ/mol, based on Friedman method. We have also

2880 J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO

GY

obtained the activation energy of pristine sPB [6],ranging from �83 to �92 kJ/mol, based on Fried-man method.

6. Conclusions

A series of sPB/organoclay nanocomposites withdifferent clay loadings have been prepared by firstin situ vacuum distillation of toluene solutions andthen melt-pressed, followed by characterization byDSC, TGA, POM and XRD. The XRD shows thatexfoliated nanocomposites are formed in the major-ity at lower clay concentrations (less than 2%), athigher clay contents intercalated nanocompositesdominate. At the same time, the XRD indicates thatthe crystal structure of the sPB/organoclay nano-composites do not change, only the relative intensityvariation of the peaks corresponding to (010) and(200)/(1 10) crystal planes, respectively, wasobserved. The POM and DSC indicate that organo-clay layers can improve cooling crystallization tem-perature, crystallization rate and reducing thespherulite sizes of sPB. TGA shows that under argonflow the nanocomposites exhibit slight decrease ofthermal stability, while under oxygen flow the resis-tance of oxidation and thermal stability of sPB/organoclay nanocomposites were significantlyimproved relative to pristine sPB. The primary andsecondary crystallization for pristine sPB and sPB/organoclay (2%) nanocomposites were analyzedbased on different approaches. The nanocompositesexhibit smaller Avrami exponent and larger crystal-lization rate constant, with respect to pristine sPB.Primary crystallization under isothermal conditionsdisplays both athermal nucleation and three-dimen-sional spherulite growth and under nonisothermalprocesses, the mechanism of primary crystallizationbecomes very complex. Secondary crystallizationshows a lower-dimension crystal growth geometryfor both isothermal and nonisothermal conditions.The activation energy of crystallization of sPB andsPB/organoclay nanocomposites under isothermaland nonisothermal conditions were also calculatedbased on different approaches.

Acknowledgements

The authors express their acknowledgment to theNational Natural Scientific Fund of China (No.20274046) and thanks to Prof. S.A. Xu of East Chi-na University of Science and Technology for pro-viding the Cloisite 20A organoclay.

References

[1] Natta G, Corradini P. The structure of crystalline 1,2-polybutadiene and of other syndiotactic polymers. J PolymSci 1956;20:251–66.

[2] Mark JE. Polymer data handbook. New York: Oxforduniversity press; 1999. p. 318.

[3] Takeuchi Y. New thermoplastic resin 1,2-polybutatidene:fundamental properties and phot degradable characteristics.Jpn Plast Age 1972;10(7):12–7.

[4] Sasaki T, Sunago H, Hoshikawa T. Multiple meltingbehavior of syndiotactic 1,2-polybutadiene. Polym Eng Sci2003;43:629–38.

[5] Bertini F, Canetti M, Ricci G. Crystallization and meltingbehavior of 1,2-syndiotactic polybutadiene. J Appl PolymSci 2004;92:1680–7.

[6] Cai JL, Han Y, Yuan ZR, Cheng RS, Wang ZL, Jiang LS,et al. Crystallization behaviors of syndiotactic and atactic1,2-polybutadiene blends. Polym Int 2004;53:1127–37.

[7] Ren MQ, Chen QY, Song JB, Zhang HF, Sun X, Mo ZS,et al. Crystallization kinetics and melting behavior ofsyndiotactic 1,2-polybutadiene. J Polym Sci Part B: PolymPhys 2005;43:553–61.

[8] Napolitano R, Pirozzi B, Esposito S. Structural studies onsyndiotactic 1,2-poly(1,3-butadiene) by X-ray measurementsand molecular mechanics calculations. Macromol ChemPhys 2006;207:503–10.

[9] Varia RA, Ishii H, Giannelis EP. Synthesis and properties oftwo-dimensional nanostructures by direct intercalation ofpolymer melts in layered silicates. Chem Mater 1993;5:1694–6.

[10] Ray SS, Matiti P, Okamoto M. Polymer/layered silicatenanocomposites: a review form preparation to processing.Prog Polym Sci 2003;28:1539–641.

[11] Kojima Y, Usuki A, Kawasumi M, Okada A, Kurauchi T,Kamigaito O. One-pot synthesis of nylon 6-clay hybrid. JPolym Sci Polym Chem 1993;31:1755–8.

[12] Kickelbick G. Concepts for the incorporation of inorganicbuilding blocks into organic polymers on a nanoscale. ProgPolym Sci 2003;28:83–114.

[13] Wunderlich B. Macromolecular physics. New York: Aca-demic Press; 1976.

[14] Avrami M. Kinetics of phase change. I General theory. JChem Phys 1939;7:1103–12.

[15] Avrami M. Kinetics of phase change. II Transformation–time relations for random distribution of nuclei. J ChemPhys 1940;8:212–24.

[16] Myers RH. Classical and modern regression with applica-tions. Boston: Duxbury Press; 1986.

[17] Hillier HL. Modified Avrami equation for the bulk-crystal-lization kinetics of spherulitic polymers. J Polym Sci Part A1965;3:3067–78.

[18] Price FP. A phenomenological theory of spherulitc crystal-lization: primary and secondary crystallization process. JPolym Sci Part A 1965;3:3079–86.

[19] Velisaris CN, seferis JC. Crystallization kinetics of poly-etheretherketone (peek) matrices. Polym Eng Sci 1986;26:1574–81.

[20] Woo EM, Yau SN. Two-stage crystallization kineticsmodeling of a miscible blend system containing crystallizablepoly(butylene terephthalate). Polym Eng Sci 1998;38:583–589.

J. Cai et al. / European Polymer Journal 43 (2007) 2866–2881 2881

GY

[21] Lu XF, Hay JN. Isothermal crystallization kinetics andmelting behaviour of poly(ethylene terephthalate). Polymer2001;42:9423–31.

[22] Rybnika F. Secondary crystallization of polymers. J PolymSci 1960;44:517–22.

[23] Woo EM, Chen JM. Effects of solvent treatment oncrystallization kinetics of poly(p-phenylene sulfide). J PolymSci Polym Phys Edn 1995;33:1985–93.

[24] Kamal MR, Chu E. Isothermal and nonisothermal crystal-lization of polyethylene. Polym Eng Sci 1983;23:27–31.

[25] Cebe P, Hong SD. Crystallization behaviour of poly(ether–ether–ketone). Polymer 1986;27:1183–92.

[26] Ozawa T. Kinetics of nonisothermal crystallization. Polymer1971;12:150–8.

[27] Jeziorny A. Parameters characterizing the kinetics of thenonisothermal crystallization of poly(ethylene terephthalate)determined by d.s.c. Polymer 1978;19:1142–4.

[28] Privalko VP, Kawai T, Lipatoo YuS. Crystallization of fillednylon 6 III noniosthermal crystallization. Colloid Polym Sci1979;257:1042–8.

[29] Liu TX, Mo ZS, Wang SE, Zhang HF. Nonisothermal meltand cold crystallization kinetics of poly(aryl ether etherketone ketone). Polym Eng Sci 1997;37:568–75.

[30] Vyazovkin S. Is the Kissinger equation applicable to theprocesses that occur on cooling? Macromol Rapid Commun2002;23:771–5.

[31] Vyazovkin S, Sbirrazzuoli N. Isoconversional Analysis ofthe nonisothermal crystallization of a polymer melt. Mac-romol Rapid Commun 2002;23:766–70.

[32] Vyazovkin S, Sbirrazzuoli N. Isoconversional analysis ofcalorimetric data on nonisothermal crystallization of apolymer melt. J Phys Chem B 2003;107:882–8.

[33] Friedman H. Kinetics of thermal degradation of char-forming plastics from thermogravimetry: application to aphenolic plastic. J Polym Sci C 1965;6:183–95.

MA

CR

OM

OL

EC

UL

AR

NA

NO

TE

CH

NO

LO