eXPRESS Polymer Letters Vol.9, No.

3 (2015) 308–320
Available online at www.expresspolymlett.com
DOI: 10.3144/expresspolymlett.2015.28

Direct correlation between modulus and the crystalline

structure in isotactic polypropylene
A. Menyhárd1*, P. Suba3, Zs. László3, H. M. Fekete3, Á. O. Mester3, Zs. Horváth1, Gy. Vörös1,
J. Varga1, J. Móczó2
1
Laboratory of Plastics and Rubber Technology, Department of Physical Chemistry and Materials Science at Budapest
University of Technology and Economics, H-1111 Budapest, M$egyetem rkp. 3., Hungary
2
Hungarian Academy of Sciences, Chemical Research Centre, Institute of Environmental and Material Chemistry, H-1025
Budapest, Pusztaszeri út 59-67., Hungary
3
Tisza Chemical Works Ltd., H-3581, P.O. Box 20., Hungary

Received 9 October 2014; accepted in revised form 15 December 2014

Abstract. Mechanical properties and crystalline structure of isotactic polypropylene (iPP) types were studied using poly-
mers, which were polymerized differently in order to obtain diverse molecular architectures. The objective of this work is
to describe quantitative correlation between the crystalline structure and the elastic modulus in order to predict structures
with expectably advantageous properties. The molecular mass was measured by gel permeation chromatography (GPC)
and the regularity of molecular structure was investigated by Fourier transform infrared spectroscopy (FTIR) and stepwise
isothermal segregation technique (SIST). The chain regularity of the studied samples is varying in a wide range according
to the results of SIST and FTIR measurements. The crystalline structure was characterized by differential scanning
calorimetry (DSC) and wide angle X-ray scattering (WAXS). The tensile properties were determined by standardized ten-
sile tests. The results indicate clearly that the increased chain regularity is accompanied by a proportional advancement in
crystallinity and consequently proportionally larger stiffness. Moreover, the results of this work were compared to those
obtained on other previously produced iPP samples and it can be established that the correlations found during this work are
valid generally. An empirical model was developed also, which connects the stiffness to the structural parameters of iPP and
makes possible the design and prediction of materials with targeted molecular structure and properties.

Keywords: mechanical properties, modeling and simulation, crystalline structure, structure-property correlation, stiffness

t1. Introduction talline modifications (!-, "- and #-form) [2, 4–6],
iPP is one of the commodity polymers used in large which influences also its mechanical properties.
quantity nowadays [1]. Its advantageous price/per- The monoclinic !-modification is the thermody-
formance ratio is the key to the success of this poly- namically stable form, which is formed during the
mer. iPP is a crystalline polymer and its properties most frequently applied industrial practices [2].
are determined by its complex crystalline structure In the past decades considerable attention was paid
[1]. The basic structural units of iPP are the chain- to the study of those parameters, which influence the
folded fibrillar or lamellar primary crystallites, which mechanical properties of iPP. A short summary is
can be classified into supermolecular structures with given about these findings in the followings. Most of
diverse geometry (spherulites, cylindtrites, hedrites, the related works describe that the stiffness depends
quadrites, etc.) [2–4]. Moreover, iPP has three crys- on the crystalline structure developing during the

*
Corresponding author, e-mail: [email protected]
© BME-PT

308
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

processing of the product [1, 7–10], i.e. the higher Table 1. The iPP samples used during the recent work
the crystallinity the larger the stiffness. The presence Sample Donor type
MFR
[g/10 min]
of nucleating agents promotes the formation of spe-
iPP1 ED1 Propyl-trimethoxy-silane 2.8
cific crystalline structures and improves crystallinity, iPP2 ED1 Propyl-trimethoxy-silane 11.0
lamella thickness is also larger in the case of effec- iPP3 ED1 Propyl-trimethoxy-silane 1.2
tive nucleation [11–14]. Consequently, the stiffness iPP4 ED2 Cyclohexyl-methyl dimethoxy-silane 16.0
of iPP containing nucleating agent is larger than that iPP5 ED3 Dicyclopenthyl- dimethoxy-silane 3.0
of non-nucleated samples [15]. Moreover, some
special nucleating agents are selective to the "-form ent molecular architectures different external donors
of iPP [16, 17]. The "-modification has lower stiff- (ED) and the same Ziegler-Natta catalyst were
ness and significantly larger impact resistance com- applied during polymerization. The polymerization
pared to the !-form [16, 18]. Stern et al. [19] have was carried out in a small scale vessel in batch mode
pointed out that the molecular mass can influence at 70°C, 120 min and 250 L/min of stirring. The
stiffness as well. He found that the increasing amount of hydrogen feed was modified, the final
molecular mass results in decreased stiffness, MFR is given together with the ED types also in
because large molecular mass hinders the crystal- Table 1.
lization process of iPP with due to the kinetic effect. The polymer powders were homogenized with
However, we have to note that he has not paid atten- additives (1500 ppm Irganox B 215, 400 ppm glyc-
tion to the density and distribution of chain defects erol-monostearate, 400 ppm Ca-stearate (BASF,
in those iPP polymers, which were used during his Switzerland)) in a Brabender W 50 EHT internal
studies. Gahleitner et al. [20] found similar ten- mixer (Brabender GmbH, Germany) at 190°C,
dency, but the decreasing stiffness was explained by 50 rpm for 10 min and subsequently were compres-
the different spherulitic structure formed in the sion molded into 1 mm thick plates at 190°C for
polymers with different molecular masses. 5 min using a Fontijne SRA 100 laboratory compres-
The phenomenon that the molecular structure influ- sion molding machine (Fontijne, The Netherlands).
ences the overall crystallinity and consequently The platelets were cooled with water in the com-
mechanical properties is known, however, only few pression molding machine and the pressure was
work can be found according to our knowledge, kept constant during cooling. MFR values were
which are linking the molecular architecture to the measured after homogenization. DSC, WAXS, SIST,
crystalline structure and to the mechanical properties tensile, GPC measurements were carried out on sam-
quantitatively [21–27]. Pukánszky et al. [7] revealed ples cut from these plates. 100 %m thick films were
that crystallinity and lamella thickness determines the compression molded at 190°C from the samples for
stiffness of iPP, although the empirical correlation FTIR analysis. Cooling conditions used for films
provided by him is applicable only in a limited were the same as for the platelets.
crystallinity range. The aim of this work is to study
the stiffness of iPP samples with diverse molecular 2.2. Experimental techniques
architectures and to characterize the effect of molec- Melting and crystallization of the polymers was
ular architecture (isotacticity and chain regularity) studied by DSC technique using a Perkin Elmer
on the crystalline structure and consequently on the DSC7 apparatus (Perkin Elmer, United States) at a
tensile properties of iPP. In addition our goal is the heating and cooling rate of 10°C/min. The experi-
development of a theoretical model, which links the ments were carried out on samples with mass of 3–
crystalline structure directly to the mechanical prop- 5 mg under N2 atmosphere. The samples were heated
erties in the entire crystallinity range. up to 220°C and held there in order to eliminate the
thermal and mechanical prehistory. The crystallinity
2. Experimental of the samples was determined based on the melting
2.1. Materials and sample preparation curve recorded during first heating according to
Five iPP homopolymer grades produced by TVK Equation (1):
(Hungary) were used in the experiments. The speci- DHm
fication of iPP samples is included in Table 1. In X5 (1)
DHm0
order to produce polymers with significantly differ-

309
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

&Hm and &Hm0 are the experimentally recorded and the crystal density and &Hm0 is the equilibrium heat
equilibrium enthalpy of fusion. The value of &Hm0 of fusion of iPP. # is the lamella thickness. The val-
(146 J/g) has been taken from the work of Monasse ues of all constants appearing in Equation (2) are
and Haudin [28]. The SIST experiments were car- included in Table 2. Once the lamella thickness is
ried out on the same equipment between 160 and known, the regular iPP sequence length can be
100°C. After the thermal and mechanical prehistory obtained according to Equation (3):
of the samples had been eliminated at 220°C for
,
5 min the samples were cooled to 160°C at a cool- I53 (3)
c
ing rate of 80°C/min and held there for 3 hours. Sub-
sequently, the samples were cooled to next crystal- I is the length of the regular flawless isotactic PP
lization temperature (150°C) and held there for sequence and c is the height of the unit cell of !-iPP,
another 3 hours. Each temperature ramp took 3 hours which is 6.5 Å according to crystallographic sym-
and the step was 10°C. After the final crystalliza- metry data [33]. #/c has to be multiplied by 3 since
tion step at 100°C was completed the samples were iPP chains form 31 helices.
re-heated again and melting curves were recorded. Crystalline structure was studied by WAXS tech-
The SIST method assumes; if the polymer is crys- nique as well. Diffractograms were recorded using
tallized at high temperature, then only those fractions a Philips PW 1830/PW type equipment (Philips,
will be able to crystallize, wherein thick lamellas The Netherlands) with CuK! radiation at 40 kV and
can develop proportionally to the crystallization tem- 35 mA and X Pert PRO MPD type equipment (PAN-
perature. Of course the holding time should be long alytical, The Netherlands) with CuK! radiation at
enough, because the development of crystalline 40 kV and 30 mA. The crystallinity of the samples
phase at high temperatures is slow. Based on this was determined using the amorphous halo of iPP
idea the crystalline structure of the polymer can be taken form the work of Boger et al. [35]
fractioned by using appropriate temperature pro- The isotacticity of the samples were characterized
gram [29, 30]. We have to notice here that the accu- by FTIR. The spectra were recorded using a Matt-
rate experimental condition would be if the sample son Galaxy 3020 apparatus (Unicam, England) in
could be held for endless time at higher tempera- the wavelength range of 4000 and 400 cm–1. Deter-
tures. However, it can be established that if the hold- mination of isotacticity was carried out based on the
ing time is longer than 2 hours the loss of fractions work of Sundell et al. [36].
with long isotactic sequence is negligible. SIST can Mechanical properties were characterized by tensile
be used for characterization of distribution of stereo- testing using an Instron 5566 apparatus (Instron,
and regiodefects in iPP similarly to temperature rise Germany). Tensile modulus was determined at
elution fractionation (TREF) [31]. According to our 0.5 mm/min cross head speed and 115 mm gauge
earlier observation chain regularity estimated by length, while other tensile characteristics were meas-
calorimetry correlates strongly with properties, ured at 5 mm/min speed.
because they are determined by the crystalline struc- Molecular mass and its distribution were character-
ture [32]. After the samples were crystallized accord- ized by gel permeation chromatography (GPC) as
ing to the SIST method they were melted at a heat- well. GPC measurements were carried out using a
ing rate of 10°C/min. The partial amount of different PL-GPC 210 (Polymer Laboratories Ltd., United
crystalline fractions can be calculated by separating Kingdom) equipment according to ISO 16014 stan-
the melting peaks refer to each crystallization step. dard procedure. The samples were dissolved in
From the upper and lower temperature limit of each 1,2,4-trichloro-benzene (TCB) (1 mg/mL) at 160°C
fraction the characteristic range lamella thickness
can be calculated according to Gibbs-Thomson Table 2. All constants, which were used during the calcula-
tion of lamella thickness during evaluation of SIST
equation (Equation (2)):
experiments
2se Constant Value (dimension) Literature
Tm 5 Tm0 a 1 2 b (2)
rcDH0m , Tm0 481 K [2, 28]
!e 0.122 J/m2 [28]
where Tm0 is the equilibrium melting point, !e is the "c 936 kg/m3 [34]
free energy of folding surface of the lamella, "c is &Hm0 146 000 J/kg [2, 28]

310
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

for 2 hours and stabilized by 2,6-di-tert-butyl-4- Table 4. The ratio of IR bands at 998 and 973 cm–1
diethylphenol (BHT) (100 mg/L). Sample I998/I973
iPP1 0.81
iPP2 0.84
3. Results and discussion
iPP3 0.90
3.1. Molecular architecture iPP4 0.93
The molecular mass and polydispersity values of iPP5 0.97
the studied samples are presented in Table 3 and the
distribution of molecular mass is represented in to the isotacticity expressed in percentage. Based
Figure 1. The results indicate clearly that iPP3 and on IR results, ED2 and ED3 provide regular isotac-
iPP5 are so called bimodal polymers, because these tic chain structures. The I998/I973 value of iPP5 poly-
grades contain a fraction with larger molecular mass. merized using ED3 suggests that this polymer has
It appears as a shoulder on the GPC curves and it is exceptionally regular stereo-structure. The isotactic-
reflected on the larger average Mw and Pd of iPP3 ity of the samples is important, but I998/I973 value
and iPP5. does not provide any information either about the
The regularity of the polymer chains developed dur- amount of region-defects or about the distribution
ing polymerization using different EDs and hydro- of all chain defects.
gen feed is characterized by the isotacticity and by In order to characterize the distribution of chain
the regular iPP sequence length. The isotacticity of defects SIST measurements were carried out. The
the polymers was characterized by IR measure- results of these experiments are demonstrated in
ments, using the intensity ratio of bands at 998 and Figure 2. It is clear that the different ED systems are
973 cm–1, which ratio is proportional to isotacticity resulting in significantly different distribution of
[36]. The results are given in Table 4 indicating that chain defects.
the isotacticity of the samples varies in a wide According to Figure 2 lots of chain defects are form-
range. The closer the ratio of I998/I973 to 1 the larger ing during polymerization in the presence of ED1
the isotacticity, however, we have to mention that resulting in relatively low I (iPP1-iPP3). SIST dis-
the correlation is not linear so the ratio is not equal tribution curve of iPP3, which has the largest Mw
among iPP1-iPP3, indicate the longest isotactic chain
Table 3. The molecular mass data of selected materials
sequence in polymers polymerized using ED1. The
Mw Mn
Sample
[g/mol] [g/mol]
Pd chain regularity is significantly larger if ED2 or
iPP1 231 707 68 882 3.4 ED3 were applied and ED3 results in the most reg-
iPP2 207 973 52 559 4.0 ular structure, where the most frequent fraction is
iPP3 420 731 87 589 4.8 longer than 100 regular monomer units. The distri-
iPP4 222 370 58 993 3.8
iPP5 311 667 63 971 4.9

Figure 1. Distribution of molecular mass of the selected Figure 2. The distribution of length of flawless isotactic iPP
polymers sequences

311
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

bution curve reflects the chain regularity nicely,

however it cannot be correlated directly to the prop-
erties of the polymer, thus the weight average of
flawless isotactic sequence length (Iav) calculated
from the distribution curve is used for characteriza-
tion of chain regularity in the followings. If we know
the crystallinity (X) of the sample, which can be cal-
culated from the melting curve after the SIST treat-
ment and Mw, the average length of folding chain can
be obtained as well. The average number of repeat-
ing units (N) can be calculated according to Equa-
tion (4):
Mw
N5 (4) Figure 3. The correlation between sequence length and iso-
Mm
tacticity and its effect on crystallinity
where Mm is the molecular mass of the repeating
units. Consequently, the length of folding chain can
be expressed as Equation (5): 3.2. Crystalline structure of studied materials
N11 2 X2 The results of calorimetric experiments are summa-
F5 (5) rized in Table 5. The crystallinity of the samples
NX
varies in a wide range. The application of ED1 results
Iav in significantly lower crystallinity compared to ED2
The expression of N(1 –'X) gives the number of and ED3 due to the smaller chain regularity in its
repeating units, which are in the amorphous folding presence. The peak temperatures of melting (Tmp)
sequences and NX/Iav is the number parallel crys- of iPP1–3 are lower than that of iPP4 and iPP5, rep-
talline sequences. In the folded lamella the number resenting that thicker lamellas can form in samples
of parallel chain sequences are equal to the number with larger chain regularity during dynamic cooling
of folding of sequences (FS) or FS + 1. In our case at 10°C/min. The effect is pronounced in the case of
‘+ 1’ is negligible, thus F can be calculated accord- ED3, which may result in the most regular chain
ing Equation (5). In other words crystallinity can be structure.
expressed with the sequence lengths as well (Equa- The crystallinity of the samples was determined
tion (6)): from WAXS pattern as well and the results obtained
Iav by the two different methods are compared in Fig-
X5 (6) ure 4. Although the points presented in Figure 4
Iav 1 F
scatter, the tendency demonstrates clearly that crys-
The correlation between isotacticity and chain regu- tallinity determined by the two experimental tech-
larity and crystallinity is presented in Figure 3. Iav
increases with increasing of isotacticity as it was Table 5. Crystallization and melting characteristics of stud-
expected, however F decreases much steeper as Iav ied samples recorded after elimination of the pre-
history of the samples
increases. We have to remark here that Iav is also
!Hm X Tmp Tm Tcp
related to the lamella thickness. It is well discernible Sample
[J/g] [%] [°C] [°C] [°C]
that crystallinity increases steeply with the enhance- iPP1 69.2 46.7 161.8 166.5 106.1
ment in chain regularity. The tendencies reported iPP2 80.0 54.1 162.5 167.3 107.8
here are very important, because they demonstrate iPP3 84.1 56.8 163.2 169.2 110.1
iPP4 95.6 64.6 167.5 172.3 110.8
clearly that the crystallinity of the sample can change
iPP5 100.5 67.9 168.2 172.8 112.6
significantly, while the lamella thickness increases
&Hm is the enthalpy of fusion recorded during first heating run
slightly. Therefore, both key parameters – crys- Tmp and Tm are the peak and end temperatures of melting respec-
tallinity and lamella thickness [7] – must be consid- tively recorded during first heating run
ered during the characterization of change in crys- Tcp is the peak temperature of crystallization, which was recorded
during cooling run
talline structure.

312
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

niques is the same. X was obtained from the first As we mentioned earlier lamella thickness is the
melting trace in DSC measurements, which was second parameter, which should be considered inde-
recorded before the elimination of thermal and pendently, during characterization of crystalline
mechanical prehistories in order to represent the structure. Lamella thickness can be estimated from
complex structure formed during cooling as accu- the melting curves, because each point of the melt-
rately as possible. Accordingly, X determined from ing curves recorded by DSC can be attributed to a
WAXS patterns and DSC traces is characteristic for lamella thickness according to the well-known Gibbs-
the crystallinity of compression molded platelets. Thomson equation [13]. Romankievicz and Ste-
The equal values of X indicate that the &Hm0 deter- rzynski [37] reported a possible way for estimation
mined by Monasse and Haudin is a reliable constant of lamella thickness distribution curve directly from
and we are using this constant throughout all of our the calorimetric melting curve. Accordingly, the
calculations. We have to remark here that signifi- distribution function of lamella thickness can be
cantly different values of &Hm0 can be found in the obtained using Equation (7):
recent literature (see in the review of Varga [2]),
1 dM 1 dE 1Tm0 2 Tm 2 2rc
consequently the crystallinity calculated using dif- 5 (7)
M ~ d, M ~ dT ~ 2seTm0
ferent &Hm0 can be questionable. Figures 3 and 4
prove unambiguously that crystallinity increases Tm is the actual melting temperature (each point of
steeply with the enhancement of parameters referring the DSC curve), Tm0 is the equilibrium temperature
chain regularity (isotacticity and regular sequence of melting, !e (0.122 J/m2) is the free energy of the
length). Although, we have to remark here that crys- folded surface of a lamella, and # is the thickness of
tallinity obtained during SIST measurements (Fig- the lamella. M is the mass of the crystalline fraction.
ure 3) and that determined by DSC and WAXS on dE is the energy needed to melt the crystalline frac-
compression molded specimens differ considerably, tion with a mass of dM in the temperature range of
because of the different crystallization conditions. T + dT. The (1/M)·(dE/dT) term of Equation (7) can
More perfect crystalline structure with larger &Hm be measured directly by DSC, because it equals to
forms during isothermal crystallization in SIST meas- the recorded heat flow divided by the rate of heat-
urement at high temperatures. Data obtained by ing. The second term of the right side of the distri-
SIST, WAXS and DSC report similar tendency, thus bution function (Equation (7)) can be calculated
the effect of chain regularity is the same under numerically. As a consequence, the lamella distri-
dynamic cooling (practical conditions) and isother- bution curve can be plotted as a function of temper-
mal conditions. ature. The melting curve and the distribution func-
tion of iPP1 are demonstrated in Figure 5 in order to
illustrate the method. Figure 5a shows the melting
trace and Figure 5b is the distribution function.
The distribution function of lamella thickness is dif-
ficult to use for the characterization of the material
and it is impossible to compare with a modulus
data. Thus, the distribution function has to be char-
acterized with one quantity, which can be the thick-
ness at the peak as well as the weight average of
lamella thickness. We believe that an entire curve
cannot be represented accurately using any of its
points, therefore we decided using the weight aver-
age of lamella thickness function, which can be
deduced according to Equation (8):

Figure 4. The crystallinity of compression molded platelets 2 e f1, 2 ,d,

, 5 (8)
determined based on WAXS and DSC measure- e f1, 2 d,
ments

313
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

Figure 5. The melting curve recorded during first heating run of iPP1 (a) and the distribution function of lamella thick-
ness (b) obtained according to Equation (7)

–
Figure 6. Relationship between # value and (a) regular iPP sequence or (b) folding length

–
Figure 7. Correlation between # value and (a) crystallinity or (b) isotacticity

314
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

f(#) is the distribution function of lamella thickness Table 6. The results of tensile tests
–
and # is the lamella thickness. The obtained # is the Sample
E !y "y
[GPa] [MPa] [%]
weight average of the distribution function and its
iPP1 0.74±0.01 20.30±0.24 12.27±0.60
value is plotted against the parameters of chain reg- iPP2 0.93±0.02 24.15±0.40 11.71±0.36
ularity in Figure 6 and 7. iPP3 1.19±0.03 28.11±0.19 9.36±0.47
It is well discernable that regular sequence length iPP4 1.77±0.02 36.92±0.36 6.31±0.39
correlates strongly with lamella thickness. The iPP5 1.83±0.05 35.16±0.83 5.43±0.20

longer the regular iPP chain, the thicker lamellas

are formed in the polymer. The correlation seems to
be close to linear in spite of the scattering of the
points. F correlates strongly with lamella thickness
as well, but the relationship is not linear. The results
indicate unambiguously that high chain regularity
results in the formation of thick lamellas. In the
case of lower regularity lamella thickness decrease
considerably, but crystallinity can be still relatively
large. This tendency is represented by the non-lin-
ear relationship between lamella thickness and X
(Figure 7).

3.3. Tensile properties

The results of tensile test are presented in Table 6.
Mechanical properties of the studied samples differ Figure 8. Modulus as a function of stress at yield values of
considerably because of their significantly different studied samples
crystalline structure indicating that the conditions
of polymerization determine the properties of the strongly with the modulus. The more regular chain
product. In addition the stress at yield values corre- structure is accompanied by proportionally larger
late strongly with the modulus values, proving that modulus. Similar correlation can be found between
the tensile measurements are reliable. The modulus the modulus and isotacticity as well. Since these
values and their correlation with stress at yield are molecular parameters are influencing the tendency
represented in Figure 8. for crystallization they affect the crystalline struc-
Effect of chain regularity on modulus is given in ture and consequently the modulus according to the
Figure 9. As it was suspected the chain regularity same way (Figure 10).
determined by the SIST measurement correlates

Figure 9. The effect of Iav (a) and F (b) on modulus

315
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

3.4. Empirical modeling of stiffness using

structural parameters
Based on the earlier results [7, 38] and our experi-
mental data it can be concluded that stiffness depends
predominantly on crystallinity and lamella thickness.
Consequently, the following expression should be
solved (Equation (9)):
–
E = f(X, # ) (9)
E is the elastic moduli of iPP, X is the crystallinity
–
and # is the weight average of lamella thickness
function in Equation (9). We have to note that stiff-
ness changes from a minimum to a maximum value.
The minimum stiffness (Emin) is the stiffness of the
Figure 10. The modulus as a function of isotacticity amorphous iPP and it is in the range of 0.01–
0.02 GPa depending on the molecular mass [39].
The modulus is plotted as a function of lamella 0.01 GPa was used in our study as Emin value. The
thickness and crystallinity in Figure 11. It is clear maximum stiffness (Emax) is the stiffness of the per-
that modulus depends strongly on both parameters, fect iPP crystal, which value cannot be obtained
however the correlation seems not to be linear if the experimentally. Moreover, it depends on the unit cell
entire crystallinity range is demonstrated for exam- geometry of iPP, thus it should depend on the direc-
ple (Figure 11b). Although, a linear function can be tion as well. Emax can be calculated only theoretically
fitted to the plots from X ( 0.5 to 0.7, which covers from the propagation of longitudinal sound waves in
the range of iPP grades in the practice. That is why the material according to the suggestion of van Krev-
the linear correlation developed by Pukánszky and elen and Nijenhuis [40]. The propagation rate
coworkers [7, 38] provides accurate results. The depends on the crystallinity and experimental data
shape of the curve at lower lamella thickness and can be extrapolated linearly to a perfect crystal (X = 1
crystallinity can be speculated on (marked with in polymeric materials). The elastic modulus deter-
dashed line in Figure 11), but the shape of the corre- mines the vibration of the neighboring molecules,
lation at large lamella thickness and crystallinity is consequently the Equation'(10) can be written:
completely unknown. Therefore the modeling of
E = uL2 )cr (10)
stiffness in the upper region is an important ques-
tion of material design and development, even if the where E is the elastic modulus, uL is the propagation
modeling is based on empirical correlations. rate of the sound wave. The extrapolated value of uL

Figure 11. The modulus as a function of the two major parameters of crystalline structure – a) lamella thickness, b) crys-
tallinity

316
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

for iPP is 2650 m/s [40] and the crystal density is The correlation demonstrated in Figure 12 is in
936 kg/m3 (see in Table 2). Accordingly, the calcu- good agreement with our expectations, because
lated Emax is 6.6 GPa using the above data and Equa- lamella thickness is infinite at X = 1 and decrease to
tion (10). We have to note that this value is much zero at low crystallinity. Moreover, it demonstrates
smaller that the theoretical stiffness along the poly- clearly that the lamella thickness increases steeply at
mer chain obtained by Kunugi [41] and Sawatari and low crystallinity, which means that the formation of
Matuso [42] based on molecular dynamic calcula- thinner lamella than 5–8 nm is not probable even in
tions. The calculated Emax = 6.6 GPa is more realistic, the low crystallinity PP grades. Using Equation (3),
since it is valid without any orientation in the sample. we can recalculate that in average at least 20–30
Considering the abovementioned Emin and Emax data, regular monomer units are necessary to form crys-
an empirical function is suggested to describe the talline structure. The variables of a, b and c are iter-
stiffness in the entire crystallinity range (Equa- ative parameters without any physical meaning, but
tion (11)): using these values the number of independent vari-
12X a 1 b g ables in Equation (11) can be reduced to the crys-
a2a b ~a b b
X ,av tallinity (X) only and Equation (11) can be fitted to
E 5 Emin 1 1Emax 2 Emin 2e (11)
the experimental data shown in Figure 11b. The fit-
ted curve can be seen in Figure 13.
where E is the modulus, X is the crystallinity, #av is the Once the parameter set of $, % and & is determined
average lamella thickness and $, %, & are iterative the modulus of iPP samples can be predicted using
parameters characteristics for iPP. Equation (11) Equation (11) in the entire crystallinity range if X and
gives back Emin and Emax for completely amorphous #av are available. Figure 14 represents the calcu-
(X = 0 and #av = 0) and perfectly crystalline (X = 1 lated and measured modulus values of several
and #av = *) iPP respectively. We have to point out unpublished iPP samples studied earlier. Both ran-
that X and #av are handled as independent parameters dom copolymers and homopolymers were included
in Equation (11), despite the fact that larger X might in this prediction in order to check the validity of
be accompanied by proportionally thicker lamellas the proposed empirical equation. The results indicate
in iPP homopolymers (See in Figure 7a). The fol- clearly that the prediction is accurate and reliable.
lowing formula is used to describe the correlation Although, the results show some scattering, which
between #av and X presented in Figure 7a in the may originate from experimental errors and inaccu-
entire crystallinity range (Figure 12) (Equation (12)): racies, the predicted values are close to the measured
X c ones consequently Equation (11) can be used for
a b
12X predicting of modulus form crystalline parameters
,av 5 a 1 be (12)
obtained from calorimetric curves.

Figure 12. Fitting Equation (12) to lamella thickness – Figure 13. Fitting Equation (11) to modulus – crystallinity
crystallinity data data

317
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

ure, that thicker lamellas result in larger modulus at

the same level of crystallinity. The effect of improve-
ment in crystallinity can be seen in Figure 15b and
it is well discernible that the 3 GPa cannot be
exceeded if crystallinity is smaller than 70%. These
simulated results indicate that the desired mechani-
cal properties can be achieved if crystallinity is
around 75% and lamella thickness is larger than
30 nm. This is in good agreement with our earlier
results that 3 GPa of modulus can be reached by
carefully designed annealing experiments [38].

4. Conclusions
iPP grades with significantly different chain regu-
larities were studied in order to describe the effect
Figure 14. Correlation between the measured and calcu-
lated modulus values of molecular architecture on the crystalline struc-
ture and consequently the influence on stiffness.
The results support clearly the well-known fact that
3.5. Prediction of structure with desired the increased chain regularity results in larger crys-
properties tallinity and the formation of thicker lamellas and
The improvement of stiffness is a crucial issue in consequently proportionally larger modulus. More-
the industrial practice. Polypropylene grades have over, the deep analysis of chain structure revealed
stiffness in the range of 1.5–2.5 GPa, which is much that crystallinity and lamella thickness are two inde-
smaller than the calculated 6.6 GPa. Recently, the pendent parameters of crystalline structure despite
desired limit value to be exceeded is 3 GPa. Using of the fact that they show correlation in several
our simple equation the crystalline structure neces- cases. The results proved clearly that the modulus of
sary to fulfill this need can be predicted, if we sim- the iPP products depends strongly on the crys-
ulate the modulus using fixed crystallinity or lamella tallinity and lamella thickness, which correlation
thickness data. Figure 15 presents such simulation. was characterized quantitatively. The upper lower
The data demonstrated in Figure 15 indicate clearly boundary of modulus was determined and an empiri-
that 3 GPa is reachable. Figure 15a presents that cal correlation was developed, which links the mod-
lamella thickening has a limited effect. Only the ulus to the crystalline structure. A parameter set was
lamella thickness below 30 nm influences stiffness. obtained for iPP and using the suggested empirical
Although, it is also clearly demonstrated in this Fig- equation and these parameters, the modulus of

Figure 15. Modulus values predicted using (a) fixed lamella thickness or (b) fixed crystallinity values

318
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

injection molded iPP samples can be calculated [9] Kalay G., Bevis M. J.: Processing and physical prop-
from the crystallinity and lamella thickness data erty relationships in injection-molded isotactic poly-
propylene. 1. Mechanical properties. Journal of
obtained from calorimetric experiments. A crys-
Polymer Science Part B: Polymer Physics, 35, 241–
talline structure was predicted using the equation 263 (1997).
given this work for exceeding 3 GPa of elastic mod- DOI: 10.1002/(SICI)1099-0488(19970130)35:2<241::
ulus without any reinforcement. Lamella thickness AID-POLB5>3.0.CO;2-V
and crystallinity should be larger than 30 nm and 75% [10] Fujiyama M.: Higher order structure of injection-
according to the prediction. We have to note that molded polypropylene. in ‘Polypropylene: Structure,
blends and composites’ (ed.: Karger-Kocsis J.) Chap-
both parameters are influenced by chain regularity,
mann and Hall, London, Vol 1, 167–204 (1995).
which has to be increased significantly in order to [11] Mitsuishi K.: Polypropylene, nucleating agents. in ‘The
fulfill this goal. polymeric materials encyclopedya’ (ed.: Salamone J.
C.) CRC Press, Boca Raton, Vol 9, 6602–6609 (1996).
Acknowledgements [12] Binsbergen F. L.: Heterogeneous nucleation in the crys-
The Authors would like to express their gratitude to the tallization of polyolefins: Part 1. Chemical and physical
János Bolyai Research Scholarship of the Hungarian Acad- nature of nucleating agents. Polymer, 11, 253–267
emy of Sciences and for the financial support of the National (1970).
Scientific Research Fund of Hungary (OTKA Grant PD DOI: 10.1016/0032-3861(70)90036-4
109346) for the project on the structure-property correla- [13] Wunderlich B.: Crystal nucleation, growth, annealing.
tions of polymeric materials. Academic Press, London (1979).
[14] Bassett D. C.: Polymer morphology: A guide to macro-
molecular self-organization. Macromolecular Sym-
posia, 214, 5–16 (2004).
References
DOI: 10.1002/masy.200451002
[1] Moore E. P.: Polypropylene handbook: Polymeriza-
[15] Avella M., dell’Erba R., Martuscelli E., Ragosta G.:
tion, characterization, properties, processing, applica-
Influence of molecular mass, thermal treatment and
tions. Hanser-Gardner Publications, Cincinnati (1996).
nucleating agent on structure and fracture toughness of
[2] Varga J.: Crystallization, melting and supermolecular
isotactic polypropylene. Polymer, 34, 2951–2960
structure of isotactic polypropylene. in ‘Polypropylene:
(1993).
Structure, blends and composites’ (ed.: Karger-Kocsis
DOI: 10.1016/0032-3861(93)90620-P
J.) Chapmann and Hall, London, Vol 1, 56–115 (1995).
[16] Varga J.: "-modification of isotactic polypropylene:
[3] Varga J.: Supermolecular structure of isotactic poly-
preparation, structure, processing, properties, and appli-
propylene. Journal of Materials Science, 27, 2557–
cation. Journal of Macromolecular Science Part B:
2579 (1992).
Physics, 41, 1121–1171 (2002).
DOI: 10.1007/BF00540671
DOI: 10.1081/MB-120013089
[4] Padden F. J., Keith H. D.: Spherulitic crystallization in
[17] Menyhárd A., Varga J., Molnár G.: Comparison of dif-
polypropylene. Journal of Applied Physics, 30, 1479–
ferent -nucleators for isotactic polypropylene, charac-
1484 (1959).
terisation by DSC and temperature-modulated DSC
DOI: 10.1063/1.1734985
(TMDSC) measurements. Journal of Thermal Analysis
[5] Lotz B., Wittmann J. J., Lovinger A. J.: Structure and
and Calorimetry, 83, 625–630 (2006).
morphology of poly(propylenes): A molecular analysis.
DOI: 10.1007/s10973-005-7498-6
Polymer, 37, 4979–4992 (1996).
[18] Mai J. H., Zhang M. Q., Rong M. Z., Bárány T., Ruan
DOI: 10.1016/0032-3861(96)00370-9
W. H.: Crystallization behavior and mechanical prop-
[6] Phillips P., J., Mezghani K.: Polypropylene, isotactic
erties of nano-CaCO3/"-nucleated ethylene-propylene
(polymorphism). in ‘The polymeric materials ency-
random copolymer composites. Express Polymer Let-
clopedya’ (ed.: Salamone J. C.) CRC Press, Boca Raton,
ters, 6, 739–749 (2012).
Vol 9, 6637–6649 (1996).
DOI: 10.3144/expresspolymlett.2012.79
[7] Pukánszky B., Mudra I., Staniek P.: Relation of crys-
[19] Stern C., Frick A., Weickert G.: Relationship between
talline structure and mechanical properties of nucle-
the structure and mechanical properties of polypropy-
ated polypropylene. Journal of Vinyl and Additive
lene: effects of the molecular weight and shear-induced
Technology, 3, 53–57 (1997).
structure. Journal of Applied Polymer Science, 103,
DOI: 10.1002/vnl.10165
519–533 (2007).
[8] Alberola N., Fugier M., Petit D., Fillon B.: Tensile
DOI: 10.1002/app.24156
mechanical behaviour of quenched and annealed iso-
tactic polypropylene films over a wide range of strain
rates. Journal of Materials Science, 30, 860–868 (1995).
DOI: 10.1007/BF01178418

319
 Menyhárd et al. – eXPRESS Polymer Letters Vol.9, No.3 (2015) 308–320

[20] Gahleitner M., Berneitner K., Nei"l W., Paulik C., [32] Horváth Zs., Menyhárd A., Doshev P., Gahleitner M.,
Ratajski E.: Influence of molecular structure on crys- Varga J., Tranninger C., Pukánszky B.: Chain regular-
tallization behaviour and mechanical properties of ity of isotactic polypropylene determined by different
polypropylene. Polymer Testing, 14, 173–187 (1995). thermal fractionation methods. Journal of Thermal
DOI: 10.1016/0142-9418(95)93196-8 Analysis and Calorimetry, 118, 235–245 (2014).
[21] Halpin J. C., Kardos J. L.: Moduli of crystalline poly- DOI: 10.1007/s10973-014-3999-5
mers employing composite theory. Journal of Applied [33] Bai F., Li F., Calhoun B. H., Quirk R. P., Cheng S. Z.
Physics, 43, 2235–2241 (1972). D.: Physical constants of poly(propylene). in ‘Polymer
DOI: 10.1063/1.1661482 handbook’ (eds.: Brandrup J., Immergut E. H., Grulke
[22] McCullough R. L., Wu C. T., Seferis J. C., Linden- E. A.) Wiley, New Jersey, Vol 1, V/21–30 (1999).
meyer P. H.: Predictions of limiting mechanical per- [34] Clark E. J., Hoffman J. D.: Regime III crystallization
formance for anisotropic crystalline polymers. Poly- in polypropylene. Macromolecules, 17, 878–885 (1984).
mer Engineering and Science, 16, 371–387 (1976). DOI: 10.1021/ma00134a058
DOI: 10.1002/pen.760160517 [35] Boger A., Heise B., Troll C., Marti O., Rieger B.:
[23] Boyd R. H.: Prediction of polymer crystal structures Mechanical and temperature dependant properties,
and properties. Atomistic Modeling of Physical Prop- structure and phase transitions of elastic polypropy-
erties, 116, 1–25 (1994). lenes. European Polymer Journal, 43, 634–643 (2007).
DOI: 10.1007/BFb0080195 DOI: 10.1016/j.eurpolymj.2006.11.003
[24] Boyd R. H.: The mechanical moduli of lamellar semi- [36] Sundell T., Fagerholm H., Crozier H.: Isotacticity deter-
crystalline polymers. Journal of Polymer Science: Poly- mination of polypropylene using FT-Raman spec-
mer Physics Edition, 21, 493–504 (1983). troscopy. Polymer, 37, 3227–3231 (1996).
DOI: 10.1002/pol.1983.180210401 DOI: 10.1016/0032-3861(96)88466-7
[25] Crist B., Fisher C. J., Howard P. R.: Mechanical prop- [37] Romankiewicz A., Sterzynski T.: The lamellar distri-
erties of model polyethylenes: Tensile elastic modulus bution in isotactic polypropylene modified by nucle-
and yield stress. Macromolecules, 22, 1709–1718 ation and processing. Macromolecular Symposia, 180,
(1989). 241–256 (2002).
DOI: 10.1021/ma00194a035 DOI: 10.1002/1521-3900(200203)180:1<241::AID-
[26] Doyle M. J.: On the effect of crystallinity on the elastic MASY241>3.0.CO;2-9
properties of semicrystalline polyethylene. Polymer [38] Horváth Zs., Menyhárd A., Doshev P., Gahleitner M.,
Engineering and Science, 40, 330–335 (2000). Tranninger C., Kheirandish S., Varga J., Pukánszky B.:
DOI: 10.1002/pen.11166 Effect of molecular architecture on the crystalline struc-
[27] Fatahi S., Ajji A., Lafleur P. G.: Correlation between ture and stiffness of iPP homopolymers: Modeling
different microstructural parameters and tensile modu- based on annealing experiments. Journal of Applied
lus of various polyethylene blown films. Polymer Engi- Polymer Science, 130, 3365–3373 (2013).
neering and Science, 47, 1430–1440 (2007). DOI: 10.1002/app.39585
DOI: 10.1002/pen.20836 [39] Mouzakis D. E., Gahleitner M., Karger-Kocsis J.:
[28] Monasse B., Haudin J. M.: Growth transition and mor- Toughness assessment of elastomeric polypropylene
phology change in polypropylene. Colloid and Poly- (ELPP) by the essential work of the fracture method.
mer Science, 263, 822–831 (1985). Journal of Applied Polymer Science, 70, 873–881
DOI: 10.1007/BF01412960 (1998).
[29] Varga J., Menczel J., Solti A.: The melting of high- DOI: 10.1002/(sici)1097-4628(19981031)70:5<873::
pressure polyethylene subjected to stepwise heat treat- aid-app6>3.0.co;2-q
ment. Journal of Thermal Analysis, 17, 333–342 (1979). [40] van Krevelen D. V., Nijenhuis K. T.: Properties of
DOI: 10.1007/BF01914024 polymers. Elsevier, Amsterdam (2009).
[30] Varga J., Menczel J., Solti A.: Memory effect of low- [41] Kunugi T.: High-modulus and high-strength
density polyethylene crystallized in a stepwise man- polypropylene fibres and films. in ‘Polypropylene A-Z
ner. Journal of Thermal Analysis, 10, 433–440 (1976). reference’ (ed.: Karger-Kocsis J.) Kluwer, Dordrecht,
DOI: 10.1007/BF01909895 295–300 (1999).
[31] Garoff T., Virkkunen V., Jääskeläinen P., Vestberg T.: A [42] Sawatari C., Matsuo M.: Elastic modulus of isotactic
qualitative model for polymerisation of propylene with polypropylene in the crystal chain direction as meas-
a MgCl2-supported TiCl4 Ziegler–Natta catalyst. Euro- ured by X-ray diffraction. Macromolecules, 19, 2653–
pean Polymer Journal, 39, 1679–1685 (2003). 2656 (1986).
DOI: 10.1016/S0014-3057(03)00069-7 DOI: 10.1021/ma00164a036

320