Polymer 45 (2004) 7549–7561

www.elsevier.com/locate/polymer

Blends of propylene-ethylene and propylene-1-butene random

copolymers: I. Morphology and structure
Zbigniew Bartczaka, Valeria Chionob, Mariano Pracellab,*
a
Centre of Molecular and Macromolecular Studies, Polish Academy of Science, Sienkiewicza 112, Lodz 90-363, Poland
b
Institute for Composite and Biomedical Materials, IMCB-CNR, Section of Pisa, Via Diotisalvi 2, Pisa 56126, Italy
Received 17 February 2004; received in revised form 15 June 2004; accepted 16 July 2004
Available online 15 September 2004

Abstract
Samples of propylene-ethylene (EP) and propylene-(1-butene) (BP) random copolymers with various comonomer content (2–3.1 wt%
ethylene, 9.9 wt% 1-butene), were melt-mixed in Brabender internal mixer at various compositions (25/75, 50/50, 75/25). Films of
copolymers and blends, as well as of a homopolymer sample (iPP), obtained by compression moulding and with different thermal history
were characterized by optical and scanning electron microscopy (OM, SEM), small-angle light scattering (SALS), small- and wide angle
X-ray scattering (SAXS, WAXS) and differential scanning calorimetry (DSC). It was found that all copolymers and blends studied
crystallized exclusively in monoclinic a-modification forming spherulitic structure in a very broad undercooling range. The average size of
spherulites is smaller in the copolymer containing 1-butene as compared to those containing ethylene or to iPP homopolymer, due to
enhanced heterogeneous nucleation in BP copolymer. SEM microscopic observations demonstrated that EP and BP copolymers were
miscible at all examined compositions and form homogeneous blends. Structural and morphological analysis indicated that the comonomer
units are incorporated into growing crystallites in both EP and BP copolymers, while the non-crystallizing material is rejected out of the
crystallites. For small concentrations of comonomer some of non-crystallizing species are pushed ahead of the front of growing spherulite
into interspherulitic regions. For higher comonomer concentration these species are mostly trapped in intraspherulitic regions. Melting
behavior of copolymers reflects the incorporation of comonomer into crystalline phase: melting temperature and crystallinity degree decrease
in copolymers and blends as compared to plain iPP.
q 2004 Elsevier Ltd. All rights reserved.

Keywords: Polypropylene copolymers; Blends; Morphology

1. Introduction structure–property relationships of PP have been widely

studied. It was found that the introduction of ethylene co-
Polyolefins and their blends, especially those based on units into PP chains makes the crystallizable sequence short
isotactic polypropylene (PP), have been extensively inves- enough to induce the growth of g-form PP crystals instead
tigated in the recent years due to their useful properties and of conventional a-form crystals [3–5]. Laihonen et al. [4]
advantageous cost/performance ratio, which make them and Mezghani et al. [5] demonstrated that the amount of g-
available for large applications in various industrial sectors phase was proportional to the ethylene content and that
[1,2]. crystallization in the g-form was enhanced at low super-
The properties of polypropylene can be widely modified coolings and low cooling rate.
through changes of the crystallization behaviour in the The distribution of comonomer units with respect to
presence of comonomer units randomly distributed along crystalline phase is a key factor governing mechanical
the chain. The effect of type and concentration of a-olefin properties of a copolymer. However, for copolymers of
comonomers (ethylene, 1-butene, 1-octene, etc.) on the propylene with a-olefins the location of the comonomer
units in the solid state is still a subject of controversy. For
* Corresponding author. Tel.: C39-50-511-229; fax: C39-50-511-266 copolymers with ethylene Laihonen et al. [4] reported that
E-mail address: pracella@ing.unipi.it (M. Pracella). the average length of 3/1 helices of the copolymer was
0032-3861/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2004.07.055
7550 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561

shorter than that of homopolypropylene, and concluded that Polypropylene random copolymers, containing ethylene
inclusion of ethylene unit into the crystals occurs to some or 1-butene as co-units, are largely used in film packaging
extent. On contrary, Monasse and Haudin [6], Feng and Hay for their advantageous optical, mechanical and surface
[7], and Pérez et al. [8] on the basis of the decrease of both properties, improved processability and weldability as
crystallinity and temperature of melting with increasing compared to PP homopolymer. Blending of these copoly-
ethylene content draw an opposite conclusion, i.e. that mers offers an additional opportunity of formulation of
ethylene units are excluded from the crystals. Zimmermann materials whose properties can be suitably and precisely
[9] also postulated exclusion of ethylene units from the modulated by both blend composition and the microstruc-
crystals, which view based on the invariance of the heat of tural characteristics of the components (type and amount of
fusion when reduced to 100% crystallinity and on the comonomer), as well as the processing conditions. It is
decrease of the long period as well as lamellar thickness anticipated that the type as well as concentration of
with increasing ethylene content. The observed changes comonomer would influence the miscibility of blend
were mainly ascribed to the severe disturbance of helical components in the melt and the crystallization process,
regularity of PP by the ethylene units [10]. which in turn, should result in variation of the crystal
The influence of 1-butene comonomer on the overall superstructure and bulk properties.
crystallization rate and melting behaviour of polypropylene This paper is aimed at analysing the morphological
was examined by Crispino et al. [11]. The incorporation of characteristics of blends of propylene-(1-butene) copolymer
1-butene units into the polypropylene lattice was also with propylene-ethylene copolymers having different
reported [12–14]. This results in an increase of lattice ethylene content, crystallized from the melt at high under-
dimensions and in a reduction of crystallization and melting coolings, which are frequently used in production of films
temperature as well as enthalpy of fusion. According to for packaging. In a companion paper [18] a detailed study of
Hosoda et al. [14], 1-butene is easily incorporated in the the isothermal crystallization and melting behaviour of
crystalline phase, which gives rise to expansion of the unit these blends is reported.
cell, so that the amount of comonomer in the crystalline
phase is comparable with that in the amorphous phase. The
degree of inclusion of 1-butene in the crystals was found to 2. Experimental
be much higher than that of ethylene. This was attributed to
the similarity of the chain conformation of isotactic 2.1. Materials
polypropylene and polybutene-1 sequences, while ethylene
units simply interrupt the polypropylene 3/1 helices, which The materials studied were commercial as well as
consequences in shortening of the crystallizable sequence laboratory samples of isotactic propylene-ethylene (EP)
and significant exclusion of ethylene units into amorphous and propylene-(1-butene) (BP) random copolymers sup-
phase. plied by Basell Polyolefins, Italy. One of ethylene-
In recent years several new copolymers of propylene propylene copolymers (coded hereafter EP2) was a
with higher a-olefins have been synthesized and investi- commercial product containing about 3 wt% C2H4 while
gated due to introduction of new metallocene catalysts. the other ethylene-propylene and propylene-(1-butene)
Pérez et al. [8] reported that for fractions of a propylene/ copolymers (coded here as EPS and BPS, respectively)
1-hexene copolymer with different comonomer content the were experimental products with 2 wt% C2H4 and 9.9 wt%
glass transition temperature, temperature and enthalpy of C3H9, respectively. These materials were designed for
melting, and consequently the equilibrium melting tem- quality packaging applications, either as monolayer films or
perature of crystals, decrease with increasing comonomer as welding layers in coextruded multilayer structures and
content. Lovisi et al. [15] studied propylene/1-hexene and were formulated with slip and antiblock agents and
propylene/1-octene copolymers and found that thermo- antioxidants.
dynamic properties–such as crystallization enthalpy and A sample of isotactic polypropylene homopolymer
temperature, melting temperature, glass transition tempera- (Moplen X305, Basell Polyolefins, Italy) of MFIZ
ture–storage modulus and density, all decrease in a linear 9 g/10 min was also studied as a reference material. The
pattern with increasing comonomer content. It was also composition and molecular characteristics of the examined
shown that the longer the alkyl branch, the less comonomer copolymers are summarized in Table 1.
is necessary to decrease the crystallization ability of
copolymer chains and hence to make the material behaviour 2.2. Blend preparation
more rubbery [16].
Nitta et al. reported on the synthesis and characterization Binary blends of BPS and two types of ethylene
of novel block copolymers of isotactic polypropylene and copolymers, EPS and EP2, respectively, were prepared by
ethylene-propylene rubber which showed single glass simultaneous melt-mixing in a Brabender Plasticorder static
transition and large orientational effects of PP phase on mixer at a temperature of 190 8C, using a mixing time of
drawing [17]. 5 min and a rotation speed of 40 rpm. The blends prepared
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7551

Table 1
Molecular characteristic of studied polymers

Sample code Comonomer Comonomer content M.F.Ia (g/ Mw Mw/Mn mmmm (mol%)
10 min)
(wt%) (mol%)
EPS C2H4 2.0 3.0 9.5 216,000 3.5 89.3
EP2 C2H4 3.1 4.6 6.0 230,000 4.5 86.1
BPS C4H8 9.9 7.6 7.0 262,000 6.3 96.6
iPP – 0 0 9.0 350,000 7.5 (O95)
a
Measured at 230 8C and 2.16 kg.

had a BPS content of 25, 50 and 75 wt%, respectively. For (Linkam, model THMSE 600). In this experiment films
comparison the samples of plain iPP and plain copolymers obtained from pellets were sandwiched between a micro-
were also processed in the mixer under identical conditions. scope slide and a cover glass, heated up to 230 8C at a rate of
Films of blends and plain copolymers were obtained by 30 8C/min and kept at this temperature for 10 min to destroy
compression moulding of bulk samples in a Carver press at any trace of crystallinity; then the films were rapidly cooled
230 8C and subsequent quenching in liquid nitrogen to down at a rate of 100 8C/min to the desired crystallization
minimize any phase separation induced by crystallization temperature and allowed to crystallize under isothermal
process. conditions.
These films were used for preparation of samples
crystallized at high undercoolings in the temperature 2.4. Scanning electron microscopy
range of 60–90 8C. For this purpose the isotropic films
obtained previously were sandwiched between two con- Additional morphological study was performed with a
stantan foils, each 40 mm thick, together with an aluminum scanning electron microscope (SEM). The morphology of
spacer of 150 mm thickness. The sandwiched films were the surface of thin samples crystallized at high under-
melted in a laboratory press at 190 8C and 50 atm for 2 min coolings, as well as the internal bulk morphology of thicker
and then quickly transferred to the crystallization cell specimens crystallized by slow cooling from the melt were
consisting of two massive cylindrical aluminum blocks examined.
(approximately 5 kg each) equipped with electrical heaters In order to reveal the morphological details the speci-
and temperature sensors. The temperature of each block was mens were etched with a permanganic etchant [21,22] prior
controlled by an electronic controller with an accuracy to microscopic examination, according to the improved
higher than 1 8C. The block surfaces contacting with the procedure described in Ref. [23]. The reagent consisted of
sample were carefully polished in order to improve heat 10 vol. parts of concentrated sulphuric acid, 4 vol. part of
transfer between the sample and blocks. Other details of the orthophosphoric acid (85%), 1 vol. part of water and 1 wt%/
cell and procedures are given in Ref. [19]. Due to small heat vol. of potassium permanganate. Etching was performed at
capacity of thin sample, the large heat capacity of the cell room temperature for 2 h. Samples crystallized at high
and very good thermal contact between them the sample undercooling were etched without any prior preparation,
could be cooled down to the desired crystallization while in slowly crystallized samples the internal surface to
temperature very quickly. Measurements with thin thermo- be etched was exposed using an ultramicrotome equipped
couple embedded in the sample demonstrated that the with a freshly prepared glass knife. Etched specimens after
crystallization temperature was reached in few seconds after appropriate washing [24] and drying were mounted on the
transferring the molten sample into the cell. This time is microscopic stages, coated with a fine layer of gold (ca.
short enough to provide nearly isothermal conditions for the 20 nm thick) by ion sputtering device (Jeol JFC-1200) and
crystallization process in the temperature range applied then examined with a scanning electron microscope (Jeol
[19]. Constantan foil was chosen as an outer layer material JSM 5500LV).
in the sandwich since it does not induce any excessive
spherulite primary nucleation on the surface of polypropy- 2.5. Light scattering
lene sample being in contact with [20].
Samples crystallized at high undercoolings were inves-
2.3. Optical microscopy tigated by the small-angle light scattering technique (SALS)
in order to determine the average spherulite radius from the
The morphology of the samples of the blends crystallized Hv scattering pattern. A He–Ne laser (lZ632.8 nm) was
at high undercoolings was examined with an optical used to generate the scattering pattern. The 2-D Hv patterns
polarized light microscope (Leitz Ortholux II POL-BK). (obtained with crossed polars) were recorded using a high
For supplementary observations of the crystallization sensitivity video camera coupled to a computer equipped
behaviour the microscope was equipped with a hot stage with miro-VIDEO capture card. From Hv patterns the
7552 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561

scattering light intensity along the scattering angle q was from the position of maximum of corrected curve using the
measured at fixed polar angle mZ458 using the ImageQuant Bragg law.
software (Molecular Dynamics). From the angular position Wide-angle X-ray scattering (WAXS) analysis of the
of the maximum of scattering intensity, qmax, the fifth-order examined materials was performed at room temperature
average spherulite radius hR5i was calculated using the with DRON-2A diffractometer coupled to a source of
following formula [19,24]: filtered Cu Ka radiation (ceramic sealed-tube with line focus
powered by Philips PW3830 generator operating at 50 kV
4:09l and 30 mA). The scans were performed within the range of
hR5 i Z (1)
4p sinðqmax =2Þ 2qZ10–308 with scanning step of 0.01/0.028, in the
reflection geometry.
The number average spherulite radius hRi can be estimated
from the fifth-order average hR5i using the approximate
relation [19,25]: 3. Results and discussion
hRi zhR5 i=k (2)
3.1. Morphology
The value of coefficient k depends on the type of primary
nucleation and changes within the range from 1.25 (pure 3.1.1. Optical microscopy
instantaneous nucleation) to 1.33 (pure sporadic one) [19]. The morphology of the copolymers and their blends was
For most of practical cases however, where the fractions of examined for samples crystallized in the temperature range
instantaneous and sporadic nuclei are not known exactly, an from 60 to 132 8C by using microscopy and SALS
intermediate value of kZ1.3 can be a satisfactory techniques. At higher crystallization temperatures (above
approximation for estimation of hRi [25]. 110 8C) the growth of spherulites in samples of plain
copolymers and blends was observed in real time by optical
microscopy. Few exemplary optical micrographs of copo-
2.6. DSC analysis
lymers and blends, non-isothermally crystallized from the
melt, are shown in Fig. 1. These observations showed that
The melting behaviour of the isothermally crystallized
for all examined samples spherulites were formed on
samples was analyzed with a Perkin–Elmer PYRIS
cooling of the melt to temperatures in the range 100–
Diamond-DSC differential scanning calorimeter, calibrated
130 8C. In all cases the characteristic spherulitic mor-
with indium standard. The samples (5–10 mg) were heated
phology of isotactic polypropylene with a monoclinic
up to 230 8C at the rate of 10 8C/min, under nitrogen
crystal structure (negative birefringence) was observed.
atmosphere (1 bar). The melting temperature and enthalpy
The positive-birefringent spherulites of b-modification were
were calculated from the position of maximum and from the
sporadically observed only in samples of plain iPP. No
area of endothermic peaks, respectively. For the evaluation
change of spherulite appearance was detected with varying
of crystallinity the heat of fusion of 100% crystalline
the copolymer structure or blend composition. The copoly-
polypropylene was taken as DHmZ209 J/g [26].
mer components did not show any phase separation
phenomena during relative slow crystallization at moderate
2.7. Small- and wide angle x-ray scattering undercoolings.
In all observed samples most of the boundaries between
The lamellar structure of copolymers and blends samples impinged spherulites was shaped as straight lines, which is
was probed by 2-D small angle X-ray scattering (2-D typical for instantaneous heterogeneous nucleation [27]. In
SAXS) at room temperature. A Kiessig-type camera (1.1 m fact, morphological observations with scanning electron
long) was equipped with a tapered capillary collimator microscope, reported later in Section 3.1.3, revealed foreign
(XOS) combined with additional pinholes forming the inclusions frequently located in the spherulite centres (cf.
beam, and an imaging plate as a detector and recording Fig. 6b), that evidences the heterogeneous nature of
medium (Fuji). To reduce parasitic scattering the entire numerous primary nuclei. The number of spherulites per
beam-path was evacuated. The camera was coupled to a unit area varied with the type of copolymer and blend
Philips PW1830 X-ray generator (sealed-tube, fine point composition. Among pure copolymers, EP2 exhibited
Cu Ka filtered source) operating at 50 kV and 40 mA. The spherulites of the largest size at a given crystallization
time of collection of the pattern was usually around 4 h. temperature, indicating a reduced nucleation density with
Exposed imaging plates were read with PhosphorImager SI respect to other copolymers.
system (Molecular Dynamics). The spherulitic structure of samples crystallized at high
Long period (LP) was determined from the 1-D sections undercoolings (TcZ60–90 8C) could be observed only in
of 2-D pattern (sectioning made with the ImageQuant completely solidified specimens, due to very high crystal-
software). Background and Lorentz corrections were lization rate at such extreme conditions. Fig. 2 shows the
applied to the curves. Long period was calculated then optical micrographs of copolymer samples crystallized at
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7553

the characteristic 4-lobe shape observed in the Hv SALS

patterns of specimen clearly evidences the presence of
spherulites in all copolymer samples. Similar micrographs
and SALS patterns (not presented here) were obtained for
samples of blends crystallized at the same conditions.
Similarly to samples crystallized at TcR110 8C the
spherulitic structure formed at TcZ60–90 8C consisted of
spherulites of the a-form. Few b-form spherulites was
observed only in specimens of iPP crystallized at 80 and
90 8C (cf. Fig. 2a), while for iPP crystallized at lower
temperatures, as well as for copolymers and their blends no
b-form spherulites were observed.

3.1.2. SALS analysis

In order to determine the average radius of spherulites in
the studied samples the Hv SALS patterns of samples
crystallized at TcZ60-90 8C were recorded (Fig. 2); similar
patterns were obtained for the blends samples (not presented
here). Samples crystallized at temperatures above 90 8C did
not produce clear scattering pattern due to large size of
spherulites.
The polar angle respective to the scattering intensity
maxima was determined for every pattern and then the fifth-
order average spherulite radius hR5i was calculated using Eq.
(1) [19,24]. Note, that this high-order average favours large
spherulites, so that the SALS-determined average spherulite
radius is substantially higher (by 1.3 approximately [19,25])
than the number average spherulite radius, which can be
determined by other methods, e.g. microscopy. Fig. 3a
presents the fifth-order average spherulite radius hR5i as a
function of crystallization temperature, as determined for
samples of plain iPP and copolymers. Fig. 3b and c show the
dependencies determined for BPS/EPS and BPS/EP2
blends, respectively.
It can be seen that for all samples the average spherulite
size increases as Tc increases. The largest variation can be
observed between 60 and 80 8C. In this temperature range
the average spherulite radius increases more than twice,
from about 2–5 mm to approximately 5–12 mm, depending
on the copolymer type (cf. Fig. 3a). For a given Tc the
average radius hR5i measured for plain polymers changes in
the following order: RBPS!REP2!REPS!RiPP; i.e. the
average spherulite size decreases with increasing amount
as well as length of side branches in the copolymer chains.
In the case of blends, the variation of spherulite radius is
Fig. 1. Polarized light micrographs of thin film samples of (a) EP2, controlled by the blend composition. For both BPS/EPS and
(b) BPS/EP2 75:25 and (c) BPS/EP2 25:75 blends, taken during non- BPS/EP2 blends, at any Tc, the spherulite size decreases
isothermal crystallization (cooling rate 2 8C/min) in the range 100–115 8C. with increasing fraction of BPS copolymer in the blend (Fig.
3b and c).
the temperatures indicated. These micrographs are sup- Since the average spherulite size is determined by the
plemented with 2-D Hv SALS patterns (crossed polars). The number density of primary nuclei, the decrease of size with
micrographs show the spherulitic structure present in all decreasing temperature of crystallization, observed in all
copolymers, regardless the crystallization temperature. The samples studied, can be associated with a substantial
micrographs are somewhat diffuse since the spherulites increase of primary nucleation at temperatures below
grown at the reported range of temperature are substantially 80 8C. This is likely due to activation of the homogeneous
smaller than the specimen thickness (150 mm). However, nucleation mode in addition to already active heterogeneous
7554 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7555

the same temperature range. Comparing the experimental

curves with those calculated for several model cases of
primary nucleation, presented in Ref. [19], it appears that
the difference of spherulite radii between samples of iPP and
copolymers originated mainly from different density of
heterogeneous nucleation (being the lowest in iPP and
increasing with increasing comonomer fraction), while
homogeneous nucleation has been less affected. The
homogeneous nucleation rate likely decreased with increas-
ing comonomer content due to shortening of crystallizable
sequences within copolymer chain, which decreased the
probability of formation of stable nuclei and hence the
nucleation rate. Thus, the observed increase of hetero-
geneous nucleation density must be mainly related to higher
number of solid heterogeneities present in copolymer
samples as compared to PP homopolymer.
The plots of average spherulite radius hR5i of EPS/BPS
and EP2/BPS blends, presented in Fig. 3b and c show a
monotonic decrease of the radius with both crystallization
temperature and increasing content of BPS in the blends
within an envelope given by the curves of plain components.
These curves indicate that the changes of primary nuclea-
tion in blends simply follow the variation of their
composition.

3.1.3. SEM analysis

The morphology of spherulites was additionally exam-
ined by scanning electron microscopy (SEM). Observations
of the spherulitic structure on the surface of films crystal-
lized at high undercoolings were in agreement with
the SALS estimations of average spherulite sizes in the
samples. Unfortunately, it was difficult to examine the
structure in more details since the final morphology of that
surface was formed in contact with the metal foil in a
sandwich sample. The presence of the foil induces
modification of the sample morphology at the surface. As
a consequence, the SEM micrographs of that surface were
rather poor in details even after appropriate and careful
etching.
Finer details of the spherulitic structure were revealed by
Fig. 3. The dependence of fifth-order average spherulite radius hR5i on observations of the bulk morphology of samples crystallized
crystallization temperature for (a) plain copolymers, (b) BPS/EPS blends,
non-isothermally by slow cooling. The morphology was
and (c) BPS/EP2 blends.
analysed by cutting the material out with an ultramicrotome
followed by an appropriate etching, as described in Section
nucleation at temperatures between 70 and 80 8C. For 2. Figs. 4–6 present the representative SEM micrographs
isotactic polypropylene, as well as for its homogeneous of plain copolymers, BPS/EPS and BPS/EP2 blends,
blends with atactic polypropylene, a jump of primary respectively.
nucleation has been reported to occur at crystallization The micrograph of plain iPP (Fig. 4a) shows typical a-
temperatures above 70 8C, and was ascribed to activation of type spherulites bordering the b-spherulite (on the right-
the homogeneous nucleation mode [19]. The experimental hand side of the micrograph), which is occasionally found in
data presented here suggest the activation of homogeneous iPP. Due to cross-hatched arrangement of lamellae in a-type
primary nucleation occurring in both iPP and copolymers at spherulites [28] the appearance of this crystal modification

Fig. 2. Polarized light micrographs of samples of (a) iPP, (b) EPS, (c) EP2, and (d) BPS, crystallized at temperatures indicated. Every micrograph is
supplemented with the Hv SALS scattering pattern obtained for the respective sample.
7556 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561

Fig. 4. SEM micrographs of the etched internal planes of bulk specimens crystallized non-isothermally from the melt: (a) iPP, (b) BPS. (c) EPS, (d) EP2.

is much more compact than that of b-type spherulites, in growing spherulite into interspherulitic regions. Similarly,
which lamellae orientation and shape can be easily yet even stronger, features can be observed for the EP2
recognized. In a-spherulites, however, the lamellar mor- copolymer, with a higher concentration of ethylene co-units
phology is rather indistinct and only outlines of some (Fig. 4d). Both spherulite interior and interspherulitic
lamellae can be seen at higher magnification. In the interior boundary display a less compact structure than that of PP
of any spherulite several small circular pits are observed, homopolymer.
which presumably are artifacts resulting from over-etching Micrographs of propylene-(1-butene) copolymer (BPS)
(too long and/or too severe etching conditions can lead to an show relatively open and loose structure of spherulites with
excessive local etching in occasional places, e.g. around many visible pits and holes, distributed evenly in the
existing surface defects or inclusions of material of lower observed cross-section of spherulite (Fig. 4b). On the other
etching resistance; this would produce some artificial hand, the interspherulitic boundaries are more irregular and
features which were not necessarily related to the original much more diffuse as compared to PP and propylene-
structure of the sample [22]). The boundaries between ethylene copolymers, but there is no evidence of any over-
a-type spherulites in iPP are compact and filled with etching. These features probably result from relative large
interpenetrating lamellae of adjacent spherulites. amount of non-crystallizable species in this copolymer
The morphology of a-type spherulites in EPS copolymer (9.9 wt% 1-butene), which are most probably rejected
(Fig. 4c) is a little more open and loose than that of preferably into interlamellar layers and trapped here rather
spherulites of iPP. The main difference is in the appearance than pushed ahead by the growing front of the spherulite.
of the interspherulitic boundaries; the pits presumably Due to the large amount of non-crystallizing material the
produced by over-etching are concentrated along these front of growing spherulite becomes quite rough and
boundaries (where the concentration of non-crystalline irregular and the boundaries between adjacent spherulites
material is higher). This suggests some segregation of appear diffused and vague [29].
non-crystallizable molecules that are pushed by the front of Micrographs of the blends BPS/EPS and BPS/EP2 are
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7557

Fig. 6. SEM micrographs of the etched internal planes of bulk specimens of

(a) 25:75, (b) 50:50, (c) 75:25 BPS/EP2 blends crystallized non-
isothermally from the melt.
Fig. 5. SEM micrographs of the etched internal planes of bulk specimens of
(a) 25:75, (b) 50:50, (c) 75:25 BPS/EPS blends crystallized non-
isothermally from the melt. in the blend: the morphology of 25:75 BPS/EPS and
BPS/EP2 blends is close to that of plain EPS or EP2,
shown in Figs. 5 and 6, respectively. These micrographs respectively, while that of the 75:25 blends show the
reveal quite open internal structure of spherulites in features similar to these of plain BPS copolymer. In any
blends, resembling that found in plain copolymers. In of the blend no phase separation of components was
both sets of blends the morphology of spherulites detected suggesting that BPS is miscible with both EPS
changes with the blend composition. The observed and EP2 propylene-ethylene copolymers in the entire
changes correlate well with the concentration of BPS range of composition studied.
7558 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561

3.2. X-ray analysis diffraction curves were all characteristic of the a-mono-
clinic modification of polypropylene, in accordance with the
Fig. 7a presents diffractograms of the plain copolymers, microscopic observations which evidenced mainly the
while Fig. 7b and c show those obtained for BPS/EPS and occurrence of a-spherulites. The only exception was a
BPS/EP2 blends, respectively, all crystallized at TcZ90 8C. sample of iPP crystallized at TcZ90 8C in which a very
The diffractograms obtained for copolymers and blends small fraction of the hexagonal b-modification could be
crystallized at other temperatures do not differ from those detected (small peak centered around 2qZ16.18, which can
shown in Fig. 7, and therefore are not presented. The be attributed to the (300) reflection of the b phase).
Moreover, no trace of the g-modification was found in
any of copolymer or blend samples. The g-modification was
reported to appear in propylene-ethylene copolymers with
relatively high concentration of ethylene units, crystallized
at low undercoolings [3–5]. The samples studied here have a
relatively low concentration of ethylene (2–3 wt%), and
were crystallized at high undercoolings. However, samples
crystallized at moderate and low undercoolings (above
110 8C) also did not reveal any trace of g-modification.
An examination of position of the diffraction maxima
demonstrates that the diffraction peaks of copolymers are
slightly shifted towards lower diffraction angles as com-
pared to the reference iPP homopolymer. This can be easily
seen, especially for (110), (040) and (130) peaks, in the
diffraction curves of copolymers presented in Fig. 7a. The
shift of the diffraction peaks is small for EPS and EP2
copolymers (yet increasing with increasing concentration of
ethylene comonomer) while it is rather pronounced in the
case of BPS copolymer. Such shift can be accounted for by
the expansion of the crystal lattice in copolymers as
compared to plain iPP. In particular, the relatively largest
shift of the peak of (040) plane indicates that the expansion
of the lattice in [010] direction is stronger than in any other
[hk0] direction. An increase of the average size of the unit
cell indicates that the comonomer units had to be
incorporated in crystals at least to some extent. Similar
conclusions were already reported in Refs. [12–14] for
copolymers containing 1-butene. The X-ray data here
reported for propylene-ethylene copolymers show that the
unit cell increases its average size with increasing the
content of comonomer, which suggests that ethylene co-
units were also incorporated within crystals as they grew
[4]. However, the lower expansion of crystal lattice for EP
copolymers, as compared to BP copolymers, is to be
ascribed to the fact that methyl side groups disturb the size
of PP helix much less than the bulky ethyl branches.
The diffraction curves obtained for the blends of BPS
with EPS and EP2 (Fig. 7b and c) show the gradual
expansion of the lattice as the composition of the blends
change from EP-rich to BP-rich. This behaviour can support
the previous conclusion on miscibility of these copolymers.
The lamellar structure of the copolymers and blends was
probed with SAXS. Exemplary one-dimensional scattering
curves extracted from 2-D SAXS patterns are shown in Fig.
8. The long period values (LP), determined from the
Fig. 7. WAXS diffraction curves of (a) plain iPP and copolymers, background and Lorentz-corrected curves, are given in
crystallized at TcZ90 8C; (b) BPS/EPS and (c) BPS/EP2 blends at various Table 2. These data demonstrate that the presence of
composition, crystallized at TcZ90 8C. ethylene units in copolymer chains results in slight
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7559

reduction of the long period as compared to the homo- Table 2

polymer. Such a reduction was observed in the entire range Long period determined from SAXS data
of the crystallization temperature and probably reflects the Sample code Composition Tc (8C) LP (nm)
shortening of the crystallizable sequence due to the presence (wt/wt)
of ethylene co-units in the chain. Similar observations were IPP – 90 13.1
already reported for propylene-ethylene random copoly- EPS – 90 12.5
mers in Refs. [9,10]. On the other hand, the long period EP2 – 90 12.4
BPS – 90 12.8
determined for BPS copolymer practically does not differ
IPP – 60 10.9
from that of iPP (within the limits of experimental error). EPS – 60 10.2
This is in spite of a noticeable reduction of crystallinity EP2 – 60 10.5
degree of the propylene-(1-butene) copolymer with respect BPS – 60 11.5
to the homopolymer, as evaluated from DSC measurements BPS/EPS 50:50 90 12.3
BPS/EP2 50:50 90 12.2
(discussed in the next section).
For BPS/EPS and BPS/EP2 blends, the values of LP of
crystals are also reduced as compared to iPP and close to temperature. Experiments with different heating rates
those observed for EP copolymers, crystallized at the same demonstrated that the lower melting peak increases its
temperature. This reflects the fact that the crystallinity size with respect to the upper melting peak when the heating
degree of EPS and EP2 copolymers is higher than that of rate increases. This indicates that the latter peak arises from
BPS at the same crystallization conditions. Thus, in the recrystallization phenomena of less perfect and/or smaller
50/50 blend the relative amount of crystalline component crystals, since the samples were unable to form more perfect
formed from EPS (or EP2) is larger than that from BPS, and crystals when crystallized at high undercooling. Moreover,
consequently the long period of the blend is similar to that of for all copolymers and their blends the temperature of
plain EP copolymers. melting peaks decreases in linear fashion with decreasing
the overall content of PP units in the system (calculated on
3.3. Melting behaviour the basis of blend composition and the content of PP units in
its copolymeric components), as shown in Fig. 10. This
The melting behaviour of isothermally crystallized relationship does not include plain iPP for which melting
samples of plain copolymers and their blends is illustrated temperature is noticeably higher than for any copolymer.
by representative DSC curves in Fig. 9. The detailed The overall melting enthalpy of samples crystallized at
analysis of crystallization and melting behaviour of these TcZ60–90 8C is strictly influenced by the comonomer
polymers is reported in a companion paper [18]. Here we content and blend composition. This behaviour is shown for
note only that all plain homo- and copolymers as well as BPS/EPS blends in Fig. 11, as a function of EPS content. On
blends crystallized at Tc!90 8C show a double melting the other hand, if the same enthalpy data, supplemented with
peak, which is rather usual behaviour of iPP and its data for plain polymers, are plotted against the overall
copolymers crystallized below 100 8C [1,30,31]. The
temperature of both lower and upper melting peak
components changes linearly with the crystallization

Fig. 8. 1-D raw scattering profiles extracted from 2-D SAXS patterns of
plain iPP and copolymers samples, crystallized at Tc Z90 8C Fig. 9. DSC melting curves of BPS and EPS copolymers and 50:50
ðsZ 2 sinðQÞ=lÞ BPS/EPS blend, crystallized at various temperatures.
7560 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561

Fig. 10. Temperature dependence of the of the lower (TM1) and upper (TM2)
melting peaks on the overall content of PP units in plain copolymers and Fig. 12. Degree of crystallinity calculated from the enthalpy of melting for
blends crystallized at TcZ90 8C. Open symbols indicate upper melting copolymers and blends crystallized at TcZ60 and 90 8C. Symbols and solid
peak, while filled ones the lower peak: (>, %) plain iPP, (B, C) BPS/EPS lines indicate values calculated using DH8m of homopolymer (209 J/g);
blends and (,, -) BPS/EP2 blends. The shaded area marks samples of the broken lines were calculated on the basis of DH8m for plain copolymers
blends, the points outside of this refer to plain components. [18]. Open symbols refer to TcZ90 8C, while filled ones to TcZ60 8C: (>,
%) plain iPP, (B, C) BPS/EPS blends and (,, -) BPS/EP2 blends. The
content of PP units in the system, then a monotonic decrease shaded area marks the samples of the blends, the points outside of this refer
with decreasing content of PP units can be observed. This to plain components.

behaviour is illustrated in Fig. 12 where the melting

fusion. In the former hypothesis the temperature of melting
enthalpy was recalculated to overall crystallinity, assuming
of crystals should not differ too much from that of iPP,
a single value (209 J/g) of the heat of fusion of 100% which apparently is not the case here (cf. Fig. 10); on the
crystalline PP [26] for all samples of copolymers and blends other hand, the latter hypothesis predicts a suppression of
(solid lines). The crystallinity of the copolymers and blends, the melting temperature, which is consistent with our
evaluated in such a way, is noticeably lower than that of PP experimental results. This hypothesis is additionally sup-
homopolymer and even lower than it could be anticipated in ported by the diffraction data which suggest a reduction of
these systems on the basis of reduction of the amount of PP crystalline thickness and an expansion of crystalline lattice.
in the system, alone. Such a discrepancy can be accounted Moreover, the results of DSC analysis of samples
for by assuming either exclusion of an extra part of PP units isothermally crystallized at Tc’s above 110 8C indicated
from the crystallization process (due to shortening of PP that the melting behaviour of copolymers is consistent with
sequences in the copolymer chains), or alternatively, the incorporation model. On the basis of melting data, the
inclusion of some comonomer units in the crystals, resulting equilibrium melting enthalpy, DH8m, of copolymers was
in the formation of very imperfect crystals and in estimated (200.7 J/g for EPS and 197.1 J/g for EP2 and
consequent reduction of the temperature and enthalpy of BPS) [18]. These enthalpies were lower than that of plain
PP. Using these values of DH8m the variation of overall
crystallinity of copolymers and blends is represented by
broken lines in Fig. 12. It can be seen that even after such a
correction is applied, the crystallinity of copolymers and
their blends results to be depressed more than one could
expect on the basis of reduced amount of PP units in the
copolymer (such reduction should not exceed 4% for the
highest content of comonomer in the BPS sample).

4. Conclusions

The polypropylene-based copolymers with ethylene and

1-butene co-units, as well as the blends of these copolymers,
were found to crystallize in spherulitic fashion in a broad
Fig. 11. The dependence of melting enthalpy on composition of the blend temperature range, down to 60 8C. The primary nucleation
for BPS/EPS blends crystallized at TcZ60 and 90 8C. of spherulites was mostly the heterogeneous nucleation
 Z. Bartczak et al. / Polymer 45 (2004) 7549–7561 7561

above TcZ80 8C. When crystallization was carried out MIUR-CNR Agreement, PF.29, Project Chemistry (95/95),
below this temperature also homogeneous, sporadic nuclea- Research line ‘Synthesis and Characterization of New
tion became effective. The heterogeneous nucleation was Polyolefin Materials for Food Packaging’, and Polish
stronger in the copolymer with 1-butene than in copolymers Academy of Sciences. The authors are grateful to Basell
with ethylene and even stronger than in the reference iPP Polyolefins Italy for providing the polymer samples used in
homopolymer. This was probably due to larger number of this work and to Dr. Antonio Addeo for his helpful
heterogeneities dispersed in copolymers as compared to contribution. Thanks are also due to Dr. C. Sacchi from
plain iPP. ISM-CNR, Milano, Italy, for the NMR characterization
All copolymers and their blends, crystallized in the data.
temperature range 60–90 8C, formed exclusively crystallites
of the a-modification. No trace of either b or g forms were
detected. As demonstrated by X-ray diffraction the dimen-
sions of the unit cell of monoclinic a-crystals increased with References
an increase of the amount and the size of copolymer units in
the chain. This was accompanied by a decrease of the long [1] Karger-Kocsis J, editor. Polypropylene: structure, blends and
period and by a pronounced reduction of the enthalpy of composites. London: Chapman and Hall; 1995.
melting of the crystallites. All these effects indicated the [2] Utracki LA. Polymer alloys and blends. Munich: Hanser; 1989.
[3] Feng Y, Jin X, Hay JN. J Appl Polym Sci 1998;68:381–6.
incorporation of ethylene or 1-butene co-units within [4] Laihonen S, Gedde UW, Werner PE, Martinez-Salazar J. Polymer
growing crystallites. 1997;38:361–9.
SEM observations showed that the appearance of [5] Mezghani K, Phillips PJ. Polymer 1998;39:3735–44.
spherulites changed from quite compact structure of iPP [6] Monasse B, Haudin JM. Colloid Polym Sci 1988;266:679–87.
towards a more open and loose structure observed in [7] Feng Y, Hay JN. Polymer 1998;39:6589–96.
[8] Pérez E, Benavente R, Bello A, Perena JM, Zucchi D, Sacchi MC.
copolymers and their blends. In propylene-ethylene copo-
Polymer 1997;38:5411–8.
lymer with low ethylene content some fraction of non- [9] Zimmermann H. J Macromol Sci, Phys 1993;B32(2):141–61.
crystallizing species was rejected and pushed ahead by the [10] Starkweather Jr HW, Van-Catledge FA, Mac Donald RN. Macro-
front of growing spherulite into interspherulitic regions, molecules 1982;15:1600–4.
while in copolymers with higher ethylene or 1-butene [11] Crispino L, Martuscelli E, Pracella M. Makromol Chem 1980;181:
1747–55.
content that material remained in the intraspherulitic
[12] Cavallo P, Martuscelli E, Pracella M. Polymer 1976;18:42–8.
regions. Although the observation of the lamellar structure [13] Abiru T, Mizuno A, Weigand F. J Appl Polym Sci 1998;68:1493–501.
with SEM did not provide enough details, the micrographs [14] Hosoda S, Hori H, Yada K, Nakahara S, Tsuji M. Polymer 2002;43:
of copolymers and their blends suggested the occurrence of 7451–60.
cross-hatching morphology, similar to that characteristic for [15] Lovisi H, Tavares MIB, Da Silva NM, De Menezes SMC, De Santa
plain iPP. Maria LC, Coutinho FMB. Polymer 2001;42:9791–9.
[16] Arnold M, Bornemann S, Koller F, Menke TJ, Kressler J. Macromol
Since no phase separation was observed in any blend Chem Phys 1998;(12):2647–53.
sample it may be concluded that the copolymers studied are [17] Nitta KH, Kawada T, Prokhorov VV, Yamahiro M, Mori H,
miscible in any proportion. Copolymers and blends crystal- Terano M. J. Appl Polym Sci 1999;74:958–64.
lized at high supercoolings exhibited double melting peak, [18] Chiono V, Pracella M, (to be published).
where the upper peak originated from prior recrystallization [19] Bartczak Z, Galeski A. Polymer 1990;31:2027–38.
[20] Gadzinowska K, Piorkowska E. Polimery 2003;48:790–9.
of smaller/less perfect crystals on heating. Temperature of
[21] Olley RH, Hodge AM, Basset DC. J Polym Sci, Part B: Polym Phys
melting as well as crystallinity of copolymer and blends 1979;17:627–43.
showed a marked decrease with the total amount of [22] Bassett DC. In: Bassett DC, editor. Developements in crystalline
comonomer in the system. The crystallinity values resulted polymers-2. London: Elsevier; 1998. p. 67–114. Chapter 2.
to be depressed more than that expected on the basis of the [23] Shahin MM, Olley RH, Blisset MJ. J Polym Sci, Part B: Polym Phys
equilibrium melting enthalpy of copolymers. That behavior 1999;37:2279–86.
[24] Stein RS, Rhodes MB. J Appl Phys 1960;31:1873–84.
was interpreted again as a result of incorporation of [25] Prud’homme RE, Stein RS. J Polym Sci, Part B: Polym Phys 1973;11:
comonomer units within crystallites growing in copolymers 1683–701.
and their blends. [26] Bu HS, Cheng SZD, Wunderlich B. Makromol Chem, Rapid Commun
1988;9:75–7.
[27] Galeski A. J Polym Sci, Part B: Polym Phys 1981;19:721–30.
[28] Norton DR, Keller A. Polymer 1985;26:704–16.
Acknowledgements
[29] Padden FJ, Keith HD. J Appl Phys 1959;30:1479–84.
[30] Wunderlich B. Macromolecular Physics, vol. 3. New York: Academic
This work has been carried out with partial financial Press; 1980.
support of the Italian Ministry of University and Research, [31] Monasse B, Haudin JM. Colloid Polym Sci 1985;263:822–31.