transcrystallization of polypropylene on carbon fibres

ELSEVIER PII: SO032-3861(96)01O83-X

PolymerVol.38 No, 18,pp. 4715-4718,1997~ 1997ElsevierScienceLtd

Printed in Great Britain.All rights reserved0032-3861/97/$17.00+ 0.00

Transcrysta Hizationof polypropylene oncarbon fibres

Chi Wang* and C. R. LiuDepartment of Chemical Engineering, Yuan-Ze Institute of Technology, Aleili,Taoyuan Taiwan 320, R.O.C.(Received 30 May 1996; revised 15 October 1996)

Measurements have been made using a polarized optical microscope equipped with hot stages to investigatethe transcrystallization of polypropylene on carbon fibres. Both the nucleation rate and induction time atvarious crystallization temperatures were measured. It has been found that the induction time is inverselyproportional to the nucleation rate. Based on the theory of heterogeneous nucleation, the interracial freeenergy difference function Au of PP on carbon fibres was determined and compared to that in the bulkmatrix. From a thermodynamic point of view, transcrystallinity is likely to take place on carbon fibres due toa lower value of Au, 1.14ergcm–z, when compared with that in the bulk, 1.23erg cm-2. ~ 1997 ElsevierScience Ltd.

(Keywords:transcrystallinity;heterogeneousnucleation;inductiontime)

INTRODUCTION

It is well known that when fibres are incorporated insemicrystalline polymers, under appropriate conditions,a highly oriented layer is developed at the fibre/matrixinterface. This distinct morphology is called transcrystal-line layer (TCL) and is a consequence of high nucleatingability of the fibres, compared to that of the bulk matrix.In contrast, isotropic spherulites are normally found inthe bulk. Owing to its better mechanical properties,compared to the bulk matrixl’2, TCL has drawn muchattention recently in studyin fibre reinforced semicrys-

5talline polymer composites3- . There have been manytheories proposed to account for the formation of theTCL8-17. However, the mechanism for the origin of TCLis still uncertain.

In our previous article’g, the formation and morphol-ogy of the transcrystalline layers in PTFE fibre/PPcomposites have been extensively discussed. Moreover,good agreements have been found between inductiontime approach and nucleation rate approach, based on aheterogeneous nucleation theory, in determining theinterracial free energy difference function Au. Thetendency for transcrystallinity to develop is dependenton the relative magnitudes of Ao in the bulk and on thefibre, respectively. In the present paper, we report on aseries of further experiments on carbon fibre/PP compo-sites. The aim of this study is to assess the application ofinduction time approach to deduce Ars value to charac-terize the nucleating ability of fibres in composites.

EXPERIMENTAL

The isotactic polypropylene powder was kindly providedby Taiwan Polypropylene Co. The viscosity averagemolecular weight, M,, is 2.8 x 105. High modulus

*To whom correspondenceshould be addressed

carbon fibres (M40B) were supplied by Toray Industries,Inc. The fibre diameter is about 8 pm. To clean the fibresurface, the fibres were rinsed with acetone for 20 minand dried at the atmosphere. A scanning electronmicroscope (SEM) was used to examine the surface ofthe fibres.

A single fibre was sandwiched between two thinpolypropylene films. Prior to crystallization, the speci-mens were held at 200”C for 10min to erase the previousthermal effect. Then, the sample was quickly shifted toa well-controlled hot stage (Mettler, FP-82) where thepre-set crystallization temperature, TC,was maintained.Dry nitrogen was introduced to eliminate any possibledegradation in all cases.

Crystallization of polypropylene on fibres wasobserved with a polarized optical microscope (POM,Nikon MICROPHOT-FXA). Crystallization time wasrecorded after the samples were switched to the Mettlerhot stage. The number of nuclei was counted directlythrough the eyepiece of a microscope. The nucleationdensity on fibres was expressed by number of nuclei perunit length of fibre. The nucleation rate was calculatedfrom the slope of the plots of nucleation density vs.crystallization time. The induction time for nucleationwas determined from the intercept of the linear nuclea-tion density to the time axis as well. Details concerningthe measurements have been given elsewhere’8.

RESULTS AND DISCUSSION

The development of PP transcrystallinity on the carbonfibre is clearly shown in Figure Z. Due to the nucleatingability of fibres, many nuclei form on the surfacecontinuously and fill the available space, Figures la–c.Subsequent crystal growth is limited to the radialdirection (perpendicular to the fibre axis) after impinge-ment of neighboring crystals takes place, Figure Id. In

POLYMER Volume 38 Number 181997 4715

Transcrystallization of polypropylene on carbon fibres: C. Wang and C. R. Liu

contrast with the transcrystallinity of fibres, relativelyfew nuclei, which will develop the spherulites later, arefound in the bulk. According to the theory of nucleation,the rate of heterogeneous nucleation Zis given by’7)’9,

logz = logz~ –u“

2.303R(TC – Tm)

16m@T;*

–2.303kTC(ATAhff )2(1)

where 10is a constant, U* is the activation energy relatedto molecules to transport across the phase boundary, R isa gas constant, TCis crystallization temperature, Tm isthe temperature below which crystallization ceases, T;is the equilibrium melting temperature of the polymer,AT is the degree of supercooling (= T; – TJ, Ahf isthe heat of fusion per unit volume of the polymer andj“ is a correcting factor, being equal to 2TC/(TC+ 2%).Further details regarding these definitions are given inrefs 17and 19. Values of U*, Tm, and Ahf are taken fromliterature2~ ~to be 6.28 kJmol-l, 223 K, and 1.96 x109erg cm respectively. a and o. are the lateral andfold surface energies, respectively. To take account of theenergy change caused by the creation of new surface onthe top of foreign substrates, a quantity termed inter-racial free energy difference, Am, is incorporated.Consequently, the surface energy parameter OO,AO isdetermined from the slope by plotting log Z+U*/2.303R(TC – Tm) vs. l/TC(AT~)2. In this study,AC~c. is used to denote the interracial free energydifference for nucleation taking place in the TCL oncarbon fibres.

Figure 2 shows the variation of measured nucleationdensity (number of nuclei per fibre length) with crystal-lization time at different T,. The nucleation rate isdetermined from the linear slope and the induction timeis taken from the intercept of the time axis. It is evidentthat the nucleation rate decreases but the induction timeincreases when transcrystallization takes place at hightemperatures. A relationship between nucleation rateand induction time tihas been developed and usedrecently]7, as follows

Z(TC)ti(TC) = constant (2)

Thus, a plot of log(l/ti) + U*/2.303R(TC – Tm) VS.l/ TC(AT~)2also gives a straight line from which ao.Aacan be calculated as well. The dependence of nucleationrate and induction time on the crystallization tempera-ture is shown in Figure 3. Two sets of experiments havebeen carried out, depending on the crystallizationtemperature, to ensure the reliability and reproducibilityof the measured number of nuclei. For TCfrom 130 to134”C, POM with a magnification of cu. 400x was usedto observe the nucleation process. However, a lowermagnification (200x) was applied when crystallizationtakes place at high TC, 135 to 140°C. Thus, at least 30nuclei on carbon fibres were counted before a layer ofTCL was well developed. As seen in Figure 3, discon-tinuity at 135°C was evident for measurements ofnucleation rate and induction time due to the resolutionof eyes with different magnifications. However, parallellines are obtained for each set of experiments which inturn verifies the relation described by equation (2).According to equation (l), the slope is used to calculatethe energy parameter PU,AI-TTcLto be 835.6 erg3cm-c forPP transcrystallinity taking place on carbon fibres.

(a)

(b)

oc

(d)

Figure1 Nucleation process and crystal growth oftranscrystallinity atdifferent crystallization times (TC= 139”C,200x magnification, underpartially polarized light): (a) IOmin, (b) 20min, (c) 40min, (d) 60min

b’~ A\ A

A

:* %“O%VA:.$? v ■ m:Af$T ■

‘iiflf, , I I :

{

❑ n.mm ;>

T (“C)A Ac : 135A A : 136

A .*’H o : 137

~

7 : 138■ : 139

❑❑ : 140

0 3 6 9 12 15 18

time (s) XI O-B

Figure 2 Nucleation density of PP on fibre surface as a function oftime at different crystallization temperatures, TC

4716 POLYMER Volume 38 Number 181997


?●D

4“.wo

1 1

:82 slope =–2.30xl 08 Kg 21

H“

1

0 < Hslope =–2.29x10e KS

–1 I 1 1 1–10.8 1.0 1.2 1.4

l/TC(ATf )zx10-6

Figure 3 Comparison of the induction time and the nucleation rate methods for PP melt in TCL to determine value of aCT,kTCI (0: T. = 136–

14ti°C, 200x magnification; ❑l:T. = 130–134”C, 400x magnification)

The magnitude of Au~c~ is used to characterize thenucleating ability of carbon fibres. After deducing oaeAovalue from nucleation study, one has to estimate thevalue of Oae in order to determine AaTcL. However, thesurface energy parameter am. can be solely determinedfrom the study of crystal growth. At the initial stage oftranscrystallization, as shown in Figure Za, separatednuclei on the fibre surface can be seen. The radius of thegrowing front was measured until the entire fibre wascovered with PP crystals at which saturation of thenucleation density was reached. Thereafter, a layer oftranscrystallinity formed and was restricted to growperpendicularly to the fibre. Thus, the thickness of TCLwas measured instead. Figure 4 shows the variation ofcrystal size with crystallization time, measured in thismanner, for carbon fibre/PP composites at T. = 134”C.It is evident that both spherulitic growth and TCLgrowth have the same growth rate, G. The growth rate isdetermined from the slope to be 4.4nm S-l. Further-more, it has been shown in a previous reportlg that thespherulitic growth rate in the bulk is identical to thegrowth rate of TCL in a wide temperature range of 120–147”C. The presence of the fibres does not affect the PPcrystal growth rate. Hoffman’s regime theory20 issuccessfully employed in both TCL and spherulitic

18 A value of ~me is &terrnined from the sloPegrowth .by a plot of log G + U*/2.303R(TC – Tm) vs. l/ TCATJ.The deduced value of surface energy parameter ma,is 732.3 erg2 cm–4, regardless of different fibres beingused.

Value of Ao~cL was determined by dividing on. AoTcLwith m,. The calculated value of Aa~c~ for PP totranscrystallize on carbon fibre surface is 1.14erg cm-2.However, the deduced value of Ao~ for PP to crystal-lize in the bulk18 is 1.23ergcm-2. Because ACTTCLissmaller than AaB, nucleation of PP crystal is likely totake place on the carbon fibre surface rather than in thebulk. It is consistent with experimental findings. Moreover,

on carbon fibres is higher than that:e ;;;EOffi&;Yk o75ergcm-2 We conclude thatPTFE fibre possessm higher nucleating ability althoughits surface energy is lower.

Hata et al.2 have demonstrated that geometrical mor-phology of the surface of substrates plays an importantrole for transcrystallinity to develop. Figure 5 shows a

growth of TCL$$’

saturationof nuclei

growth

01 1 I 1 1 Io 500 1000 1500 2000 2500

time (s)

Figure 4 Radius of individual spherulite, at early stage, and TCLthickness, at later stage vs. time at T. = 134°C

Figure 5 Scanning electron micrograph of carbon fibre

SEM micrograph of carbon fibre. It is clear that somesmall ridges exist along the fibre axis. However, theroughness of the fibre surface is rather uniform, whencompared with that of the Kevlar fibres. The correlation,equation (2), between induction time and nucleation rate

POLYMER Volume 38 Number 181997 4717


which is valid for PTFE fibre/PP18 and the currentsystems is not applicable to Kevlar fibre/PP system dueto the non-uniform surface roughness of Kevlar fibres2].Moreover, the surface roughness may be critical tothe formation of the transcrystallinity because of theenhanced thermal-stresses (stress concentrations) at thepeaks or valleys, when specimens are cooled from hightemperatures to the crystallization temperatures. A

21 based on thermal-stress-induced crYstal-kinetic theory ,lization and surface roughness of fibres, is proposedto account for the origin of the transcrystallinity. Inaddition, the effect of TCL on the interracial strength ofcomposites is extensively investigated in thislaboratory22.

CONCLUSIONS

A polarized optical microscope has been used to followthe development of PP transcrystallinity on carbonfibres. Experimental results show primarily that theinduction time approach is successfully applied todeduce the value of aa,Ao~c~ from nucleation studies.Moreover, the formation of TCL and nucleating abilityof the fibres are associated mainly with the relative levelof interracial energy difference function at fibre surface,AoTc~, to that in the bulk, ArT~.

ACKNOWLEDGEMENTS

The financial support of this work by the NationalScience Council (NSC85-2815-C155 -01-017E) is greatlyappreciated.

REFERENCES

1.

2.

3.

4.

5.

6.7.8.9.10.11.12.13.

14.

15.

16.

17.18.

19.

20.

21.22.

Kewi, T. K., Schonhorn, H. and Frisch, H. L., J. Appl. Phys.,1967,38, 2512,Hata, T., Ohsaka, K., Yamada, T., Nakamae, K., Shibata, N.and Matsumoto, T., Proc. 16th Annual Symposium, AdhesionSociety, Williamsburg, VA, February, 1993, pp. 180-183.Varga, J. and Karger-Kocsis, J., J. Polym. Sci., Polym. Phys.Edn., 1996, 34, 657.Klein, N., Selivansky, D. and Marom, G., Polym. Compos.,1995, 16, 189.Avella, M., Volpe, G. D., Martuscelli, E. and Raimo, M.,Polym. Eng. Sci., 1992,32, 376.Felox, J. M. and Gatenholm, P., J. Mater. Sci., 1994,29, 3043.Chen, E. J. H. and Hsiao, B. S., Polym. Eng. Sci., 1992,32,66.Gray, D. G., J. Polym. Sci., Polym. Lett. Ed., 1974, 12, 645.Devaux, E. and Chabert, B., Polym. Commun., 1991, 15, 464.Keller, A. and Sawada, S., A4akromol. Chem., 1964,74, 190.Koutsky, J. A., Walton, A. G. and Baer, E., J. Polym. Sci., 1967,B5, 185.He, T, and Porter, R. S., J. Appl. Polym. Sci., 1988, 35, 1945.Thomason, J. L. and Van Rooyen, A. A., J. Mater. Sci., 1992,27, 889.Thomason, J. L. and Van Rooyen, A. A., J. Mater. Sci., 1992,27, 897.Campbell, D. and Qayyum, M. M., J. Polym. Sci., Polym. Phys.Edn., 1980, 18, 83.Chatterjee, A. M. and Price, F. P., J. Appl. Polym. Sci., 1975,13,2369.Ishida, H. and Bussi, P., Macromolecules, 1991, 24, 3569.Wang, C. and Hwang, L. M., J. Polym. Sci., Polym. Phys. Edn.,1996,34,47.Wunderlich, B., Macromolecular Physics, Vol. 2, Chap. 5. Aca-demic Press, New York, 1976.Hoffman, J. D., David, G. T. and Lauritzen, Jr, J. 1. Trealise onSolid-State Chemistry, Vol. 3, Chap. 7, ed. N. B. Hannay.Plenum Press, New York, 1976.Wang, C. and Liu, C. R. (manuscripts in preparation).Wang, C. and Hwang, L. M., J. Polym. Sci., Polym. Phys. Edn.,1996,34, 1435.

4718 POLYMER Volume 38 Number 181997

transcrystallization of polypropylene on carbon fibres

Documents