Polymer 40 (1999) 2303–2312

Thermal and mechanical properties of poly(l-lactic acid)–poly (ethylene-

co-vinyl acetate) blends 1
Jin-San Yoon a,*, Sung-Hwan Oh a, Mal-Nam Kim b, In-Joo Chin a, Young-Ha Kim c
a
School of Chemical Science and Engineering, Inha University, Inchon 402-751, South Korea
b
Department of Biology, Sangmyung University, Seoul 110-743, South Korea
c
Korea Institute of Science and Technology, Cheongrang, Seoul 130-650, South Korea

Received 15 August 1997; revised 19 May 1998; accepted 19 May 1998

Abstract

Poly(l-lactic acid)(PLLA) was blended with poly(ethylene-co-vinyl acetate)(EVA) and the miscibility of the blend was investigated by
DSC, polarizing optical microscopy and rheometry. The blend of PLLA with EVA70, that contained 70 wt% of vinyl acetate, was immiscible
because the T g and the spherulitic growth rate of the blend were nearly constant regardless of the change in the blend composition. On the
other hand, the T g, equilibrium melting temperature, the spherulitic growth rate during the isothermal crystallization of the PLLA–EVA85
blend were decreased with the increase in the EVA content. Moreover, the Flory–Huggins interaction parameter of the blend was found to be
negative, which clearly means the PLLA–EVA85 blend was miscible. The tensile strength and modulus of the PLLA–EVA85 blend were
dropped rapidly, followed by a more gradual decrease with the increase in the EVA85 content. Strain-at-break, however, was increased rather
slowly up to 70 wt% of EVA85 and then increased quite rapidly around 90 wt% of EVA85. 䉷 1999 Elsevier Science Ltd. All rights reserved.

Keywords: Poly(l-lactic acid); Poly(ethylene-co-vinyl acetate); Miscibility

1. Introduction ing 70 and 85 wt% of vinyl acetate unit. The themal and
mechanical properties of the blend were explored.
Poly(l-lactic acid)(PLLA) degrades biologically into
lactic acid, a product of the carbohydrate metabolism, and
its importance as a substitute for the non-degradable ther- 2. Experimental
moplastics has attracted a lot of attention in recent years.
Poly(lactic acid)(PLA) exists in l- and d-form, which are 2.1. Materials
optical isomers. PLA with large amount of l-form isomer is
highly crystalline. In general, the crystallinity and biode- Condensation polymerization of lactic acid with dipen-
gradability depend on the content of d-form isomer [1]. taerythritol(0.1 wt%) produced star-shaped PLLA. PLLA
Thanks to its biocompatibility PLLA is suitable for sutures was precipitated in methanol and dried before use.
[2,3], drug delivery systems [4,5], and implants for bone EVA70, that contained 70 wt% vinyl acetate, was purchased
fixation [6]. from Scientific Polymer Products and EVA 85, that con-
Since PLLA is very brittle at room temperature and tained 85 wt% vinyl acetate, was provided in latex form
hydrolyzes easily, its applicability has been limited [7–11]. by Taeyoung Chemical Co. The surfactant was removed
Several studies on blending or copolymers of PLLA have by the soxhlet extraction for 1 week in methanol. The
been reported in order to modify the properties and bio- chloroform soluble fraction of EVA85 was used for the
degradability [12–19]. blending. The composition of the copolymer was confirmed
Poly(vinyl acetate) has been shown to be miscible with by 1 H-n.m.r. Table 1 shows the molecular characteristics of
PLLA [16], and we have investigated the miscibility of the raw materials.
PLLA with poly(ethylene-co-vinyl acetate)(EVA) contain-
2.2. Blending
* Corresponding author.
1
Dedicated to Prof. Ick-Sam Noh on the occasion of his retirement from PLLA and EVA were dissolved in chloroform (3 wt%),
Inha University. and they were blended with stirring to form a uniform solu-
0032-3861/99/$ - see front matter 䉷 1999 Elsevier Science Ltd. All rights reserved.
PII: S0 03 2 -3 8 61 ( 98 ) 00 4 63 - 7
2304 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312

Table 1
Characteristics of polymers

a
Measured by g.p.c. in chloroform at 30⬚C. bMeasured by 1 H-n.m.r.

tion. Film samples were obtained by initially drying the 10⬚C min ¹1, holding it at 165⬚C for 1 min, cooling it to
solution at ambient temperature, followed by drying under room temperature at 10⬚C min ¹1, and then reheating it to
vacuum at 40⬚C. The films were compacted in a hot press at 165⬚C. The equilibrium melting temperature was estimated
150⬚C and 250 atm for 1 min. by heating the sample to 165⬚C at a rate of 10⬚C min ¹1,
holding it at 165⬚C for 1 min, quenching it to the pre-
determined crystallization temperature(T c) at a rate of
2.3. Analysis ¹200⬚C min ¹1, recrystallizing it at T c for 120 min, and
then reheating it at 10 ⬚C min ¹1.
The melting temperature was obtained by heating the E⬘, E⬙ and tand of the film (50 ⫻ 4 ⫻ 0.03 mm) were
sample from room temperature to 165⬚C at a rate of measured using a Rheovibron (DDV-II-C) at a fixed

Fig. 1. Tand curve of PLLA–EVA70 blends.

 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312 2305

Fig. 2. Tand curve of PLLA–EVA85 blends.

frequency of 11 Hz while heating the sample from ¹30⬚C at The microscopic heterogeniety of the blend was investigated
a heating rate of 10⬚C min ¹1. using a Rheometer (Physica MC-120). Cone-and-plate geo-
The spherulitic growth rate in the blend films was metry was used in the frequency range of 0.16–10 Hz.
determined using a polarizing microscope (Nikon OPTI- Mechanical properties of the film were measured with a ten-
PHOT2-POL). The sample was heated to 165⬚C for 1 min, sile testing machine (Lloyde Instrument, LR 5OK) at a cross-
and then quenched to the pre-determined T c, at a rate of head speed of 50 mm min ¹1 according to ASTM D882-90.
¹100⬚C min ¹1 before being observed with a CCD camera. The load cell used had a maximum range of 2.5 KN, the

Fig. 3. Glass transition temperatures of PLLA–EVA85 blend.Solid line: the Fox equation results.
2306 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312

Fig. 4. Crystallinity of the PLLA–EVA85 blends measured by DSC.

sample gauge length was 50 mm, and the sample width was independent of the blend composition. Thus, we may
10 mm. deduce PLLA and EVA70 are not compatible. On the
other hand, the PLLA–EVA85 blend shows a single glass
3. Results and discussion transition peak in Fig. 2, which is located between the T g of
PLLA and the T g of EVA 85. The transition peak of the
Fig. 1 shows tand as a function of temperature for the PLLA–EVA85 blend was also dependent upon the blend
PLLA–EVA70 blends, as determined by Rheovibron. composition.
Two separate glass transition peaks, each due to the respec- The glass transition temperature obtained from the Rheo-
tive constituent polymer, are clearly seen and they were vibron measurements is in general at a higher temperature

Fig. 5. Observed melting temperature (T m⬘) as a function of crystallization temperature (T c) for PLLA–EVA85 blends.
 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312 2307

Fig. 6. Equilibrium melting temperature of PLLA–EVA85 blends.

than that determined by DSC. The solid line in Fig. 3 is the where W is the weight fraction, and the subscript indicates
calculated T g, based on the following equation proposed by the constituent of the blend. Fig. 3 clearly shows that the
Fox [20]. agreement between the experimental T g of the PLLA–
1 W W EVA85 blend and the theoretical T g is as good as the
¼ PLLA þ EVA85 (1) agreement for the miscible PLLA–PVAc blend [16].
Tg Tg, PLLA Tg, EVA85
The samples for the T g determination by DSC and Rheo-

Fig. 7. Melting-point depression for PLLA–EVA85 blends and the Flory–Huggins equation results.
2308 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312

Fig. 8. Non-isothermal 2nd scan overall enthalpy of fusion based on the unit weight of PLLA in the blends (DHmⴱ ).

vibron were prepared by quenching at a rate of by the following equation.

¹200 ⬚C min ¹1 after melting at 165 ⬚C for 1 min, and by
DHmⴱ
pressing at 150 ⬚C, 250 atm followed by quenching. Xc ¼ (2)
Therefore, the two polymers are believed to be miscible DH o
in the melt state. where DH 0 is the heat of fusion per unit weight of the
The crystallinity of PLLA in the blend can be calculated perfectly crystalline PLLA, and DHmⴱ is the heat of fusion

Fig. 9. Enthalpy of fusion determined during the measurement of the equilibrium melting temperature based on the unit weight of PLLA in the blends as a
function of T c.
 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312 2309

Fig. 10. Radius of PLLA spherulites in the PLLA–EVA70 blends as a function of time at T c ¼ 105⬚C.

per unit weight of PLLA, that can be calculated from the blend was obtained by utilizing the heat of fusion of
heat of fusion of the blend (DH m) as follows the blend, which was determined by DSC, and was
DHm plotted in Fig. 4. As the content of EVA85 in the blend
DHmⴱ ¼ ⫻ 100 (3) was increased, the crystallinity of PLLA was drastically
PLLA content (wt%)
reduced.
Thus, the crystallinity of PLLA in the PLLA–EVA85 The equilibrium melting temperature, which can be

Fig. 11. Radius of PLLA spherulites in the PLLA–EVA85 blends as a function of time at T c ¼ 105⬚C.
2310 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312

Fig. 12. Radial growth rate of spherulite as a function of crystallization temperature for PLLA–EVA85 blends.

defined as the fusion temperature of the infinitely large temperature, and g is the ratio of the initial to the final
lamella, can be obtained by the following Hoffman– lamellar thicknesses. The value of g lies between 0 and 1.
Weeks equation [21]. In Fig. 5 the observed melting temperature was determined
by heating the blend specimen to 165⬚C at a rate of
Tc 1
Tm ⬘ ¼ þ (1 ¹ )Tmo (4) 10⬚C min ¹1. Each specimen was previously heated from
g g room temperature to 165⬚C at 10⬚C min ¹1, being held at
where T m⬘ is the observed melting temperature, T c is the 165⬚C for 1 min, and then crystallized for 120 min by
crystallization temperature, Tmo is the equilibrium melting quenching from 165⬚C at ¹200⬚C min ¹1 to the

Fig. 13. Plot of logG⬘ versus logG⬙ for PLLA–EVA85 (50/50) at various temperatures.
 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312 2311

predetermined crystallization temperature. The equilibrium fusion as a function of the blend composition obtained
melting temperature was then obtained by extrapolating the during the measurement of the equilibrium melting
melting temperature to T m ¼ T c. In Fig. 6 the equilibrium temperature. Contrary to the results in Fig. 8 the heat of
melting temperature was decreased with an increase in the fusion did not change much, as the EVA85 content and
EVA85 content in the blend, which indicates the constituent crystallization temperature were varied. Thus, the decrease
polymers in the blend were thermodynamically miscible. in the crystallinity (Fig. 4) and the heat of fusion (Fig. 8) of
According to the Flory–Higgins theory [22], [23], the PLLA crystallites with an increase in the EVA85 content
depression of the equilibrium melting temperature can be could be due to the crystallization rate of PLLA being very
descibed as follows slow and the crystallization being incomplete during the
    nonisothermal crystallization process in DSC. In fact, the
1 1 RV2 lnf2 1 1 2
¹ o¼¹ þ ¹ f þ x f enthalpy of fusion of the blend in the nonisothermal crystal-
Tm Tm DH o V1 m2 m2 m1 1 12 1
lization process was nearly constant and independent of the
(5) blend composition for the miscible blends of PHB–PEO
where x 12 is the interaction parameter between the [24], PHB–EVA85 [26] and PHB–poly(lactic acid-co-ethy-
constituents of the blend, V is the molar volume of the repeat lene glycol-co-adipate) [27], where the crystallization of
unit at the equilibrium melting temperature, f is the volume PHB was reasonably fast.
fraction of the constituent in the blend, m is the degree of The radius of PLLA spherulites of the PLLA–EVA70
polymerization, DH o is the heat of fusion of the perfect blend is plotted against the crystallization time in Fig. 10,
crystalline polymer and T m and Tmo are the equilibrium for a blend crystallized at 105⬚C. Straight lines were
melting temperature of the blend and the crystalline obtained for all the blend compositions and the slopes
constituent, respectively. Subscripts 1 and 2 represent the were almost identical irrespective of the EVA70 content
noncrystalline and the crystalline constituents, respectively. in the blend. Meanwhile, for the PLLA–EVA85 blend the
Eq. (5) can be rearranged as slope decreased with an increase in the EVA85 content, as
      shown in Fig. 11. In Fig. 12 the spherulitic growth rates
DH o V1 1 1 lnf2 1 1
¹ ¹ þ þ ¹ f ¼ f21 x12 between 90⬚C and 125⬚C are shown for different blend com-
RV2 Tm Tmo m2 m2 m1 1
positions. The growth rate of the spherulite was decreased as
(6) the EVA85 content in the blend was increased, and the
If x 12 is assumed to be independent of the blend composition growth rate of PLLA spherulite was the highest at 105⬚C,
and the melting temperature depression is assumed to be inde- regardless of the blend composition.
pendent of the morphological effects, a plot of the LHS of Eq. It is always necessary to investigate the presence of
(6) against f21 yields a straight line, and x 12 is calculable from microheterogeneity even though the blend shows a single T g
the slope. Fig. 7 was obtained using the appropriate numerical and the melting temperature depression. The existence of
values for the various parameters for the PLLA–EVA85 microheterogeneity can be examined by the plot of log G⬘
blend: (Tmo ¼ 436 K, V 1 ¼ 72.24 cm 3 mol ¹1, V 2 ¼ versus logG⬙ [28,29]. If such plot is dependent upon the tem-
59.5 cm 3 mol ¹1, m 1 ¼ 8357, m 2 ¼ 582, r 1 ¼ 1.07 g cm -3 r 2 perature or shows inflection points, the blend contains micro-
¼ 1.21 g cm -3 and DH o ¼ 1600 cal mol ¹1 [24], [25]). The heterogeneity. Fig. 13 is the plot of logG⬘ versus logG⬙ of the
interaction parameter between PLLA and EVA85 was thus PLLA–EVA85 blend (50/50). A temperature-independent
founded to be ¹ 0.043, indicating the PLLA–EVA85 blend straight line was obtained between 150⬚C and 170⬚C, which
formed a thermodynamically stable homogeneous solution in undoubtedly means the PLLA–EVA85 (50/50) blend is mis-
the melt state. cible without any microscopic inhomogeneity.
The enthalpy of fusion of PLLA in the PLLA–EVA85 PHB was also reported to show similar miscibility beha-
blend in Fig. 8 was obtained by DSC while the blend was vior toward EVA. PHB was immiscible with EVA70, while
scanned the second time. The enthalpy of fusion of PLLA it was miscible with EVA85 [26]. In general, the window of
crystallites in the blend was seen to decrease with an the solubility parameter for the miscible polymer pairs is
increase in the EVA85 content. Fig. 9 depicts the heat of rather narrow. However, PVAc formed miscible or

Table 2
Tensile properties of the PLLA–EVA85 blends

Sample Stress at break (kg mm ¹2) Tensile modulus (kg mm ¹2) Strain at break (%)

PLLA100 5.7 219 4.5

PLLA90 4.6 184 4.7
PLLA70 3.3 134 6.9
PLLA50 1.7 130 10.2
PLLA30 1.7 131 9.0
PLLA10 1.4 64 208.9
EVA85 1.4 62 244.9
2312 J.-S. Yoon et al./Polymer 40 (1999) 2303–2312

compatible polymer blends with many different polymers of [7] Miller RA, Brady JM, Cutright DE. J Biomed Mater Res 1977;11:711.
varying solubility parameter, including PLLA, PHB [32], [8] Vert M, Chabot F, Leyray J, Christel P. Makromol Chem Suppl
1981;5:30.
poly(methyl methacrylate) [33], poly(ethyl acrylate) [34] [9] Chu CC. J Appl Polym Sci 1981;26:1727.
and poly(butyl acrylate) [34]. The fact that EVA85 is mis- [10] Leenslag JW, Pennings AJ, Bos RM, Rozema FR, Boering G.
cible with both PHB and PLLA, whose hydrophilicities are Biomaterials 1987;8:311.
widely different, may result from the high vinyl acetate [11] Fukuzaki H, Yoshida M, Asano M, Kumakura M. Eur Polym J
content in EVA85. 1989;25 (10):1019.
[12] Reeve MS, McCarthy SP, Gross RA. Macromolecules 1993;26:888.
Table 2 shows the breaking strength, modulus, and strain- [13] Stevels WM, Bemard A, Van De Witte P, Dijkstra PJ, Feijen J. J Appl
at-break of the PLLA–EVA85 blend. The properties of Polym Sci 1996;62:1295.
PLLA alone were similar to those reported by Tsuji et al. [14] Koyama N, Doi Y. Macromolecules 1996;29:5843.
[30] and Hiljanen-Vainio et al. [31]. As the EVA85 content [15] Nakayama A, Kawasaki N, Arvanitoyannis I, Iyoda J, Yamamoto N.
was increased, breaking strength and modulus began to Polymer 1995;36:1295.
[16] Gajria AM, Davé V, Gross RA, McCarthy SP. Polymer 1996;37:437.
decrease sharply, which was followed by a rather slower [17] Li SM, Rashkov I, Espartero JL, Manolova N, Vert M.
decrease. On the other hand, strain-at-break was gradually Macromolecules 1996;29:57.
increased up to 70 wt% of EVA85 and then suddenly [18] Zhang L, Xiong C, Deng X. Polymer 1996;37:235.
increased at 90 wt% of EVA85. [19] Nolle AW. J Polym Sci 1950;5:1.
[20] Fox TG. Bull Am Phys Soc 1956;1:123.
[21] Hoffman JD, Weeks JJ. J Res Natl Bur Std 1962;66:13.
Acknowledgements [22] Flory PJ. Principles of polymer chemistry. Ithaca, NY: Cornell
University Press, 1953.
[23] Nishi T, Wang TT. Macromolecules 1975;8:909.
This work was supported by the Korea Science and Engi- [24] Avella M, Martuscelli E. Polymer 1988;29:1731.
neering Foundation (94-0900-07-01-3). [25] Barham PJ, Keller A, Otun EL, Holmes PA. J Mater Sci
1984;19:2781.
[26] Yoon JS, Oh SW, Kim MN. Polymer 1998;39:2474.
References [27] Yoon JS, Chang MC, Kim MN, Kang EJ, Kim C, Chin IJ. J Polym Sci,
Polym Phys 1996;34:2543.
[1] Zhang X, Goosen M, Wyss UP, Pichora D. J Macromol Sci, Rev [28] Han CD, Kim JK. Polymer 1993;34:2533.
Macromol Chem Phys 1993;C33 (1):81. [29] Chauang HK, Han CD. J Appl Polym Sci 1984;29:2205.
[2] Penning JP, Dijkstra H, Pennings AJ. Polymer 1993;34 (5):942. [30] Tsuji H, Ikada Y. Polymer 1995;36:2709.
[3] Vainionpaa S, Rokkanen P, Tormala P. Prog Polym Sci 1989;14:679. [31] Hiljanen-Vainio M, Varpomaa P, Seppälä J, Törmälä P. Macromol
[4] Heller J. Biomaterials 1980;1:51. Chem Phys 1996;197:1503.
[5] Jakanicz TM, Nash HA, Wise DL, Gregory JB. Contraception [32] Greco P, Martuscelli E. Polymer 1989;30:1475.
1973;8:227. [33] Song M, Long F. Eur Polym J 1991;27:983.
[6] Vert M, Christel P, Chabot F, Leyray J. In: Hastings GW, Ducheyne P. [34] Krause S. Polymer–polymer compatibility. In: Paul DR, Newman S,
Macromolecular biomaterials. Boca Raton, FL: CRC Press, editors. Polymer blends. New York: Academic Press, 1978 (and
1984:120–142. references therein).