Nucleating Agents for High-Density Polyethylene—A Review

Karl M. Seven,1 Jeffrey M. Cogen,1 James F. Gilchrist2

1
The Dow Chemical Company, Collegeville, Pennsylvania
2
Department of Chemical and Biomolecular Engineering, Lehigh University, Bethlehem, Pennsylvania

A review of nucleating agent (NA) types and their effect on undergoes a phase change upon cooling where chains orient and
crystallization in high-density polyethylene (HDPE) is pre- align at the molecular scale into a periodic lattice. During crys-
sented. The focus is on methods to improve the physical tal growth, polymer chains orient around the initial nucleation
properties of HDPE due to its widespread use in commer- site, or nucleus, in a 3D pattern most often forming a spherical
cial applications and high volume use in typical industrial
crystal cluster called a spherulite. At early stages of crystal
processes including extrusion, injection molding, and blow
molding. The crystallization process in HDPE significantly growth, spherulites grow individually until they encounter other
affects its final optical, mechanical, and thermal properties. spherulites or the material interface, each defining a crystalline
The addition of NAs affects the physical properties of domain. Complete crystallization occurs when the crystalline
HDPE by controlling the crystallization from the melt state. and noncrystalline domains become space filling. The final
Specific NAs improve properties such as clarity, cycle material is characterized by the size and shape of these domains,
time, and modulus. NAs are more widely developed for
polypropylene (PP) than HDPE as its slower crystallization often defined by the initial nucleation density and growth rate
rate allows greater control in achieving property improve- [9].
ments. While certain NAs are effective in improving prop- There are three classes of nucleation: (1) homogeneous
erty characteristics in HDPE, greater control over the nucleation or spontaneous nucleation; (2) self-nucleation; and
crystallization process would achieve further improve- (3) heterogeneous nucleation [9]. Homogeneous nucleation
ments in specific properties. Research has progressed in occurs in pure polymers due to the influence of supercooling
identifying effective NAs for HDPE, though the magnitude
of the effects remains lower than those generally observed also known as undercooling, which is the difference between
in PP. Inorganic and organic NAs are reviewed with an the melt temperature (Tm) and the onset crystallization tempera-
emphasis on the mechanisms by which they function. Fun- ture (T0) [10]. Self-nucleation is a result of partially melted
damentals of polymer crystallization and modeling kinetics polymer that provides structural heterogeneity, with the crystal-
during both isothermal and nonisothermal studies provide line portion acting as a nucleus. Heterogeneous nucleation
the necessary framework for characterizing the effects of involves the addition of a foreign material that provides a seed
a NA in HDPE. Finally, the interactions between HDPE, NA,
and industrial processing conditions as related to practical upon which crystal growth can occur.
applications are discussed. POLYM. ENG. SCI., 56:541–554, Nucleation and crystal morphology are affected by the addi-
2016. V
C 2016 Society of Plastics Engineers tion of a nucleating agent (NA) that promotes heterogeneous
nucleation. A NA is a dispersed material that acts as a local
nucleus that is added to the polymer. As the ultimate physical
properties of HDPE are strongly affected by crystal nucleation,
INTRODUCTION growth, and final crystal morphology, the addition of NA pro-
In industrial applications, polymers are melted in processing vides a means to control the crystallization process and enable
steps such as extrusion and molding and the size, dimension, improvements in many physical properties [9, 11]. Thus, the
and distribution of crystalline domains, or crystallites developed interfacial properties of NAs often determine their effectiveness
upon cooling determine the final properties of the material. Spe- in seeding crystallization.
cifically in high-density polyethylene (HDPE) with a density of The addition of effective NAs in most polymers increases the
0.941–0.970 g/cc, often used in films, bottles, pipes, and tubes, rate of crystallization and the crystallization temperature, Tc.
optical, mechanical, thermal, and chemical properties are signifi- The increase in rate of crystallization and in Tc depends on the
cantly affected by the crystallization process [1, 2]. Directed polymer/NA combination. However, there are limitations to the
modification of the crystalline morphology during solidification current effectiveness of NAs. Generally, NAs in HDPE cause a
from the melt state in HDPE can alter a wide range of physical 18–38 increase in Tc, a smaller crystallite size, and no substantial
properties such as water vapor transmission rate [3], optic clarity change in the fraction of the material that crystallizes. In con-
[3], shrinkage [4, 5], and cycle time in extrusion and molding
trast, nucleated PP for which crystallization behavior is more
[6, 7]. As these properties are directly related to the crystalline
widely studied, increases in Tc trend up to 268 higher than for
morphology of the polymer, directed modification to control the
neat P [12, 13]. A beneficial result from the addition of effec-
crystallization of HDPE can lead to significant improvements in
tive NAs is reduced cycle times in polymer processing such as
targeted physical properties.
extrusion or molding. Fabricated parts solidify faster, increasing
Crystallization occurs in two main stages known as nuclea-
tion and crystal growth [8, 9] During nucleation the polymer the rate of production. Another benefit is greater transparency or
clarity in HDPE as NAs reduce crystal sizes to a range smaller
than the wavelength of visible light to reduce light scattering
Correspondence to: K.M. Seven; e-mail: [email protected]
DOI 10.1002/pen.24278
[3]. Also, reductions in postinjection mold and postextrusion
Published online in Wiley Online Library (wileyonlinelibrary.com). shrinkage values are achieved by the addition of a suitable NA
C 2016 Society of Plastics Engineers
V due to a more uniform distribution of crystal sizes [14, 15].

POLYMER ENGINEERING AND SCIENCE—2016

 FIG. 1. Schematic representation of spherulite growth during initial stages (A) and final stages (B) of crystalliza-
tion. (Reproduced from Ref. 20, with permission from Elsevier).

Researchers comparing the effect of NAs on HDPE to their by noncrystallized amorphous regions [19]. Crystallization rates
effect on polypropylene (PP) have concluded that nucleation in can be observed by measuring the growth of spherulites as a
HDPE is more difficult to control than in PP due to HDPE’s function of time using either optical microscopy or transmission
faster crystallization rate. PP’s lower crystallization rate allows electron microscopy (TEM) of thin sections. The isothermal
the NA more time to increase the nucleation density [16]. Rela- radial growth of spherulites is typically linear as a function of
tive to HDPE, the effects of NAs on PP crystallization have time. Optical microscopy techniques show that spherulites con-
been widely researched and improved physical properties are tinue to grow until they impinge on adjacent spherulites as
well documented. The many effective NAs commercialized for shown in the second picture in Fig. 1 [20], creating a space-
PP are successfully used industrially to improve processing filling polycrystalline matrix [21].
times, as well as the material properties of PP, which include Many beneficial physical properties such as higher modulus,
optical clarity, warpage/shrinkage control, and modulus [17]. higher tensile strength, and higher end-use temperatures of semi-
Compared to PP, relatively few commercially developed NAs crystalline polymers such as HDPE can be attributed to crystal
are marketed for use in HDPE. Nevertheless, studies show that size distribution and degree of crystallinity, when compared to
various organic and inorganic compounds are effective as NAs similar polymers that are mostly or fully amorphous. In contrast
in HDPE. Effective NAs have the potential to achieve the same to fully amorphous polymers, the crystalline regions in semi-
property improvements in nucleated HDPE as in PP though gen- crystalline polymers have a distinct melting temperature typi-
erally, to a lesser extent. cally measured by differential scanning calorimetry (DSC).
This article reviews the effects of NAs on the crystal formation Also above the glass transition temperature (Tg), and below
process in HDPE and the resulting crystalline morphology. Critical the melting point, the modulus of a semicrystalline polymer is
polymer processing factors, such as cooling conditions, that control higher than that of a fully amorphous polymer, which is
and affect the crystallization process and the resulting physical explained by the reinforcing effect of crystals in the polymer or
properties in HDPE are discussed. Flow-induced crystallization locking in of the polymer structure caused by physical crosslink-
effects are briefly discussed but are not the focus of this article. ing of the amorphous regions by the crystallites or spherulites
The techniques and mathematical models used to analyze [10]. The crystallization process in polymers can only occur in
HDPE crystallization kinetics are reviewed followed by a survey the range of temperatures between the Tg and Tm of the poly-
of the NAs investigated. Finally, this review summarizes the mer. The plot of temperature under isothermal conditions versus
properties needed for designing effective NAs and potential new rate of crystallization takes the same basic shape shown in Fig.
directions for material properties and processing conditions for 2 for all semicrystalline polymers including HDPE. The charac-
achieving higher degrees of control of crystallization and nucle- teristic shape of the curve in Fig. 2 is caused by the slowing of
ation in HDPE. crystal growth with increasing viscosity at temperatures close to
the Tg and by the reduced thermodynamic driving force as the
FUNDAMENTALS OF HDPE CRYSTALLIZATION melting point Tm of the polymer is approached. The rate of crys-
Crystallization in polymers leads to local alignment of the tallization then peaks (Tp) at some temperature between Tg and
molecular chains into thin lamellae. Lamellae are formed from Tm. The crystallization rates at both the Tg and Tm points are
polymer chains that are folded back and forth in a regular pat- theoretically zero [10].
tern and the term lamella is often used interchangeably with Crystallization in HDPE and in most semicrystalline poly-
crystallite [18]. The lamellae combine to form spherulites, mers develops over time while cooling from the melt state. The
which are spherically shaped clusters of crystallinity that grow general form of the curve is sigmoid shaped and is shown in
radially from each nucleation point during crystallization. Fig. 3 [10]. An initial induction time is required for the forma-
Within the spherulite are layers of oriented lamellae separated tion of spherulitic nuclei from the melt state. During primary

542 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

FIG. 2. Rate of crystallization versus temperature for semicrystalline poly-
mers. (Reproduced from Ref. 10, with permission from Elsevier). [Color fig-
ure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

crystallization spherulites slowly form and then the crystalliza-

tion process increases rapidly during which time the spherulites
grow radially outward. Primary crystallization includes nuclea-
tion, crystal growth, and spherulitic organizing processes. Sec- FIG. 4. Degree of crystallinity versus time for linear PE. (Reprinted with
ondary crystallization occurs when crystallization slows due to permission from E. Ergoz et al. [1] Copyright 1972 American Chemical
the spherulites impinging on each other at which point the Society).
majority of the transformation is complete. During secondary
crystallization, additional polymer chains may continue to move
onto existing crystals and the crystallization may continue to nonisothermal conditions with high cooling rates (>50008C/
increase although at a much slower rate. Secondary crystalliza- min) may distort the curve. The start of the crystallization or
tion is typically more pronounced at higher cooling rates as onset time of crystallization is T0 in Fig. 3. The first portion of
there may be inadequate time for complete spherulite develop- this curve before the sharp increase is often referred to as the
ment during the primary crystallization phase and therefore fur- induction time and is the time needed for the formation of
ther crystallization is shifted to the secondary crystallization nuclei. Te is the time to reach complete crystallization. The crys-
phase [22]. Secondary crystallization can lead to additional tallization half-time, T1/2, is the time to reach 50% of the maxi-
shrinkage in finished molded or extruded parts in particular at mum attainable crystallinity [21]. Generally, most polymers
temperatures greater than the Tg [19]. including HDPE do not achieve complete crystallinity due to
The general development of crystalline structure, which is their semicrystalline nature. Figure 3 shows the normalized frac-
described as crystalline fraction or degree of crystallinity as a tion of polymer crystallized versus time [23].
function of time, is shown in Fig. 3. Both isothermally and non- Polymer chains start out in a random coiled state and then
isothermally crystallized polymers cooled at a constant rate undergo considerable reconfiguration to arrange into an ordered
roughly follow the shape shown in Fig. 3. However, extreme crystalline structure. The typical extent of crystallization for
HDPE spans a broad range, roughly from 54 to 83%, depending
on factors such as polymer branching, and the measurement
method used to determine the value. The reconfiguration of the
polymer chains into the crystalline structure is always obstructed
to a certain extent by chain entanglements, polymer branches,
and/or side groups. The formation of the crystalline structure in
HDPE causes closer packing of polymer chains into the spheru-
lite/crystal lattice structure, which leads to a change in density.
Therefore, the change in density of the crystalline HDPE solid
is proportional to the crystalline fraction [19].
As a general example of this phenomenon in polyethylene,
Fig. 4 shows the actual measured degree of crystallinity devel-
opment versus time for linear PE [1]. Generally, as Tc decreases
from Tm a rapid increase in crystallization rate results as seen in
Fig. 4. Also, the induction time becomes progressively longer as
FIG. 3. Generalized curve for the degree of crystallinity versus time.
Tc increases. Independent of Tc, the curves show the same basic
(Reproduced from Ref. 20, with permission from Elsevier). [Color figure sigmoidal shape. Over time the same degree of crystallinity is
can be viewed in the online issue, which is available at wileyonlinelibrary. achieved regardless of the crystallization temperature as long as
com.] it is above Tg and below Tm [1, 23].

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 543

 FIG. 5. Degree of crystallinity versus time for HDPE (A) and corresponding double logarithmic Avrami plot (B).
(Reproduced from Ref. 26, with permission from John Wiley and Sons).

Both kinetic and thermodynamic elements promote the crystal- the crystallinity at time t, and Xc(t1) is the crystallinity at the
lization process. From a kinetic standpoint increased cooling or end of the crystallization process [26]. This equation enables a
“undercooling” (Tm – Tc) promotes both the nucleation and crys- convenient analysis of crystallization based on calorimetry
tallization process. From the thermodynamic standpoint, the crys- experiments using a DSC. The overall crystallinity is the per-
tal structure that forms is in a lower energy state than the melt centage of the measured heat of fusion compared to the theoreti-
state. A NA present in HDPE decreases the degree of undercooling cal heat of fusion of the pure polymer [9]. Two additional
needed to start the crystallization process and also leads to smaller techniques for measuring relative crystallinity include density
spherulite sizes. The ultimate degree of crystallization achieved and wide-angle X-ray scattering (WAXS). The density method
immediately after cooling to ambient temperature depends on the is calculated based on tabulated density values for the amor-
rate of crystallization and cooling rate [24]. As cooling rate phous and crystalline phases along with the measured density of
increases especially for very rapid cooling conditions greater than the polymer. In X-ray diffraction the crystalline regions in the
10008C/min, a lower degree of crystallization is achieved. For polymer produce sharp diffraction peaks whereas the amorphous
instance, polymers subjected to rapid cooling in a cold mold in regions produce broad halos. The degree of crystallinity can be
injection molding processes do not have adequate time to develop estimated by integrating the relative intensities of the peaks and
their full crystallinity. However. in the case of very rapid cooling halos [9, 27].
to ambient temperature, the polymer will eventually reach the The Avrami equation (2), first published in 1939 and still
same level of crystallinity if maintained at temperatures between used successfully today, describes the isothermal crystallization
Tg and Tm; however, the process may take several weeks or kinetics of many semicrystalline polymers including HDPE.
months at ambient temperature which is not a practical time scale Several examples of HDPE blends analyzed by the Avrami
[1, 19]. Cooling rates also have a significant effect on the time to method are included in this article [28–30]. The equation
start the crystallization process as typically observed in nonisother- describes how solids transform from one state of matter or phase
mal crystallization studies. The general trend seen in DSC cooling to another at constant temperature and can also describe some
curves is that crystallization starts at higher temperatures when the types of chemical reactions [25].
cooling rate is lower because there is more time at higher tempera-
tures for the nuclei to start the nucleation process and begin crys- XðtÞ5 12exp ð2Ktn Þ ; (2)
tallization. Conversely, higher cooling rates delay nucleation
activity as there is insufficient time to initiate the crystallization where X(t) is the time-dependent relative crystallinity, n is the
process at higher temperatures and therefore the nucleation occurs Avrami exponent, K is the Avrami rate constant involving
predominantly at lower temperatures. nucleation and growth parameters, and t is the crystallization
time. The Avrami exponent n depends on the type of nucleation
ISOTHERMAL CRYSTALLIZATION KINETICS AND THE and growth process. Generally the values of n are between 1
AVRAMI EQUATION and 4 [28–30].
The determination of the relative crystallinity X(t) as a func- The Avrami equation takes the following assumptions into
tion of time at a constant temperature can be used to analyze account: (1) that the nucleation process occurs randomly in space;
isothermal crystallization kinetics in polyethylene. The relative (2) that time dependence of nucleation is either zero or first order
crystallinity at time t is defined in Eq. 1 and has a maximum as described below; and (3) crystals have a specific geometric
value of 1 and minimum value of 0 [25]. form. Both assumptions 2 and 3 are described by the n value [20].
The value of the Avrami exponent n is composed of two
Ð t dH
Xc ðtÞ 0 dt dt DHt terms N and C where n 5 N 1 C [31]. The Avrami equation
XðtÞ5 Ð
5 1 dH
5 ; (1) includes both instantaneous nucleation where nuclei develop at
Xc ðt1 Þ 0 dt dt DH1
once on cooling the polymer to the crystallization temperature
where Ht is the heat generated through time t, H1 is the total called zero-order where N 5 0 and first-order sporadic nuclea-
heat generated to the end of the crystallization process, Xc(t) is tion where the number of nuclei formed increases linearly with

544 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

time where N 5 1. However, in many cases the nucleation pro- Nakamura et al. [10] published a detailed study of the noniso-
cess falls somewhere between the completely instantaneous and thermal crystallization kinetics of neat HDPE. This study utilized
completely sporadic cases and the result is a noninteger Avrami incident X-rays scattered from an HDPE film sample inside a spe-
index [31]. The typical value of n for polyethylene is 2.6–4.0 cial sample holder with heating stage and thermocouple to monitor
[20]. The value of n often decreases as the crystallization pro- changes in crystalline diffraction peaks versus temperature during
ceeds [32]. The geometric shape of the crystals also influences n, cooling. The degree of crystallinity X(t) is calculated by comparing
where C is the dimension of the growth. For 3D spheroidal the diffraction peak intensities at time t to the completely amor-
growth, 2D planar discoidal growth, and 1D rod growth, C 5 3, phous and completely crystalline states. These in situ experimental
2, and 1, respectively. It is common for n 5 4 where nucleation is results obtained for nonisothermal crystallization agreed with the
first order and the growth dimension is spherical. theoretical predictions calculated from the Nakamura equation.
Equation 2 is typically rearranged in a linear form and a dou- The Nakamura equation assumes instantaneous nucleation [37, 38].
ble logarithm applied: As derived, the Avrami equation applies only to isothermal
crystallization. Jeziorny [39] modified the Avrami equation to
log½2logð1 – XðtÞÞ 5 n log t 1 log K: (3) describe the kinetics of nonisothermal crystallization. A cor-
rected form of the rate term (K) used in the Avrami equation
A typical example of an isothermal relative crystallinity curve assuming a constant cooling/heating rate was proposed. The rel-
for a HDPE sample is shown in Fig. 5A. The sigmoid-type curves ative crystallinity at a set cooling rate (R) is a function of the
typically shift to the right with increasing isothermal crystalliza- crystallization temperature (T) and rewriting Eq. 1: [40]
tion temperature and the crystallization rate slows down [26].
If Eq. 3 adequately describes the isothermal crystal kinetics Ð Tc dH
T dT
of the polymer then the plot of log[2log(1 – X(t))] and log t is XðtÞ5 Ð T01 dT ; (4)
dH
a straight line and the values of n and K can be obtained from T0 dT dT
its slope and intercept, respectively. An example plot is shown
in Fig. 5B for HDPE based on the data shown in Fig. 5A [26]. where X(T) is the relative crystallinity as a function of crystalli-
For the Avrami equation to successfully describe the kinetics in zation temperature. The crystallization onset temperature is T0
the system the data points in this plot must be highly linear and Tc and T1 represent the crystallization temperature at time t
within a specific conversion range with a very high correlation and after the completion of the crystallization process. The crys-
coefficient [31]. The Avrami equation adequately described the tallization time variable t can be obtained using:
primary crystallization kinetics in the isothermal study of the
barium sulfate NA system. However, the authors commented T0 2T1
t5 ; (5)
that the plots tend to deviate from linear behavior in the later R
stages of crystallization due to the effect of secondary crystalli- where R is the cooling rate in 8C/min. The rate of nonisothermal
zation. A similar nonlinear behavior occurred in the Avrami crystallization depends on R, and K can be corrected to obtain
kinetic studies of UHMWPE as NA [33, 34]. the primary rate constant Kr:
In some cases the data may fit a straight line over a smaller range
of crystallization conversions often around the 50% area. In this log Kr 5 log K=R: (6)
case, some improvements in the data fit may be achieved by assign-
ing a noninteger value of n because samples may not have the same Jeziorny theory was used to model the nonisothermal crystalliza-
consistent geometry of rod, disc, or sphere throughout their volume tion kinetics in both SiO2 and TiO2 NA studies. An acceptable linear
and a combination of zero- and first-order nucleation may be pres- fit of the data was achieved with this model in both systems except
ent. Grenier and Homme [35] published a review of the limiting fac- at the very end stages during secondary crystallization [33, 34].
tors for applying the Avrami equation. Below, examples of Ozawa modified the Avrami equation for nonisothermal crys-
crystallization kinetics analysis using the Avrami equation are tallization assuming that the polymer was heated or the polymer
included in the Nucleating Agent Developments section. melt was cooled at a constant rate. According to the Ozawa
theory, the Avrami equation becomes Eq. 7, which is called the
NONISOTHERMAL CRYSTALLIZATION KINETICS Ozawa equation. The Ozawa equation is based on both tempera-
Often research on polymer crystallization is conducted under ture and cooling rate R:
idealized isothermal conditions, which greatly simplifies the  
KðTÞ
mathematical analysis but fails to account for the varying cool- XðtÞ512exp 2 m ; (7)
ing rates and temperatures typically encountered within the R
polymer in applications. However, the study of crystallization in where K(T) is the cooling or heating temperature function, m is
an environment with continuously changing temperatures is of the Ozawa exponent which depends on the dimensions of the
greater practical importance as industrial processes generally crystal growth and order of nucleation.
occur under nonisothermal conditions [10]. Taking the double-logarithmic form:
One of two models presented by Ziabicki et al. [36, 81]
describes the nonisothermal crystallization of polymers as a log½2 logð12 Xt Þ 5 log K ðT Þ – m log R: (8)
sequence of isothermal steps. The Ziabicki equation is a series
expansion of the Avrami equation thus having limited effective- Measuring a crystallization process at different cooling rates
ness because it is restricted to temperature ranges where experi- and plotting log[2log(1 2 Xt)] against log R at a given tempera-
ments measure isothermal crystallization [36]. ture, straight lines should be obtained if the Ozawa method is

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 545

valid. The parameters m and K (T) can be determined from the where G is the universal gas constant, Tp is the peak crystalliza-
slope and intercept, respectively. tion temperature at maximum crystallization rate, and R is the
Eder and Wlochowicz [41] showed that the Ozawa method cooling rate, dT/dt. This equation allows comparison between
did not produce straight lines in the log[2log(1 2 Xt)] versus log experiments conducted at various cooling rates [42]. Other
R plot for HDPE indicating that this model was not adequate, examples of this type of activation energy calculation are pre-
which was also confirmed by Liu et al. [42]. Likewise, attempts sented in the Nucleating Agent Developments section.
to utilize the Ozawa method in the SiO2 and CaCO3 NA systems Another method to estimate DE utilizes the Friedman
showed unacceptable nonlinear behavior in both the neat HDPE equation.
and composites during both primary and secondary crystallization  
phases [34, 43]. In all cases, this discrepancy was attributed to dXt DE
ln 5constant2 :
factors neglected in the Ozawa theory including secondary crys- dt RT
tallization and dependence of lamellar thickness on the Tc. The
Ozawa theory also does not include variability of n with Tc. Slow The instantaneous crystallization rate as a function of time is
or secondary crystallization in PE can sometimes be greater than dXt. The constant is an empirical pre-exponential factor. Plots of
40% of the total crystallinity [41]. dXt/dt versus 1/T yield a straight line with slope equal to –DE/R
To achieve a better description of the crystallization behav- [33, 43].
ior, Mo and coworkers [42] proposed combining Eqs. 2 (Avrami
equation) and [7] (Ozawa equation). The combination provides HOMOGENEOUS NUCLEATION AND HDPE MOLECULAR
a relationship between cooling rate R and crystallization temper- WEIGHT FRACTIONS
ature T. Before considering the effects of NAs in HDPE it is useful
to discuss the most basic form of crystallization in HDPE which
log ðK Þ 1 n log t 5 log K ðT Þ – m log R (9) involves homogeneous nucleation. The average molecular
weight (Mw) and molecular weight distributions of HDPE influ-
can be rewritten as ence its nucleation and crystallization. Generally, an increase in
Mw results in higher viscosity slowing the process by which the
log R 5 log FðT Þ – a log t (10) chains move into an ordered crystalline structure during cooling,
The parameter F(T) is the heating or cooling rate: FðT Þ5½KðTÞ= and thus lowering the crystallization rate.
Kt1=m and a is the ratio of the Avrami exponent n to the Ozawa For instance, studies conducted on PP copolymers with a
exponent m or a 5 n/m. Mo’s method was successful in describ- range of melt flow index (MFI) showed that the crystallization
ing the nonisothermal kinetics at both lower and higher degrees rate and level of crystallinity increased with decreasing Mw or
of crystallinity (X > 0.8) for TiO2, SiO2, and CaCO3 and overall viscosity, or increased MFI. Similar trends would be expected
a better linear fit was obtained in comparison to the Jeziorny for HDPE. Physical properties of PP, such as crush resistance
and Avrami methods [33, 34, 43]. and shrinkage, were improved with increased MFI polymers.
The improved crush resistance was attributed to increased levels
of crystallinity in the higher MFI polymers [48].
ACTIVATION ENERGY
Chiu and Fu [49] studied the effect of various HDPE Mw frac-
One way to interpret the mechanism of heterogeneous nucle- tions on the crystallization kinetics and related crystal morphol-
ation is the active NA reduces the interfacial free energy ogy by utilizing DSC, TEM, X-ray scattering, and other
between crystalline polymer and the NA constituent as com- techniques. Results confirmed that Mw affects the crystallization
pared to the polymer–polymer crystalline-amorphous interfacial rate, Avrami exponent, and crystal morphology. Generally, the
energy. The activation energy or DE is the energy to transport lower Mw samples showed higher crystallization rates under iden-
molecular chains from the molten state to the crystal growth tical Tc and conditions. As expected, based on crystal growth
surface [44]. The reduced interfacial free energy decreases the theory the crystallization process was faster for the samples crys-
activation energy for nucleation, thus lowering the amount of tallized at lower Tc (higher DT) regardless of Mw. The authors
cooling required. As temperature is lowered, the drive to crystal- hypothesized that the nucleation mechanism changes from spon-
lization is favorable until the increase in polymer viscosity taneous to sporadic nucleation as Tc increases. This conclusion is
inhibits the chain alignment, making the crystallization process supported by polarized optical microscopy (POM) images taken
more difficult at low temperatures [45]. of low and high Mw samples at differing Tc values [49].
Binsbergen’s study of polyolefin crystallization supports this POM micrographs of the low Mw samples crystallized from
theory, showing a greater tendency of the HDPE crystallite to the melt state to room temperature showed considerably differ-
adsorb the foreign material rather than the liquid (melt) phase ent morphologies as compared to the high Mw samples with
because of reduced interfacial energy thereby lowering the acti- Mw > 5800. The samples with Mw > 5800 had a higher density
vation energy required for nucleation [46]. of nuclei which led to smaller spherulites as compared to spher-
To estimate the activation energy in nonisothermal systems, ulite sizes in the lower Mw samples [49].
Kissinger [47] relates the maximum rate of extent of conversion
to activation energy, DE, while cooling [34]. NUCLEATING AGENT DEVELOPMENTS
Classes and Characteristics of NAs
d½logðTR2 Þ DE
p
52 ; (11) While there is no method for a priori prediction of the per-
dðT1p Þ G formance of a NA for a particular polymer, previous studies led

546 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

 TABLE 1. Summary of effective nucleating agents for HDPE. effectiveness of the NA. Finally, the NA must be well dispersed
within the polymer melt to maximize its effectiveness [16, 46].
Category Nucleating agent Reference Examples of these effects are discussed further in the next
Nanoscale/ Calcium carbonate 43, 51
sections.
inorganic Titanium dioxide 33
fillers Barium sulfate 26 Effect of NA Concentration and Crystallization Temperature
Silicon dioxide 34
Expanded graphite 52, 53 Studies of NAs in HDPE frequently address whether the con-
POSS 25, 54 centration of the NA impacts performance, as measured by the
Multiwall carbon 55–57 increase in Tc and crystallization rate indicating a more effective
nanotubes (MWNT) NA [12, 68, 69]. For the NAs listed in Table 1, the underlying
Montmorillonite clay 32
studies that address concentration include the NA ranges from
Vermiculite 58
nanocomposite 0.5 to 5 wt% while the other studies use a fixed amount of the
mineral NA that is generally within that range. The NAs in Table 1 all
Talc 14 show an increase in Tc and crystallization rate at all NA concen-
Haloysite 59 trations studied in HDPE. Those studies of NAs in HDPE that
nanotubular
include the effect of concentration show that the performance of
clay
Organic Ultra high-molecular- 4
the NA is concentration dependent. For example, studies of NAs
fillers/ weight PE listed in Table 1, including UHMWPE [4], CaCO3 [43], TiO2
additives Sisal fibers 60 [33], SiO2 [34], BaSO4 [26], POSS [25], graphite [52], MWNT
High-modulus 61 [55], and ionomer-treated sisal fibers [60], correlate performance
PE fiber with the amount of the NA in the HDPE. It is therefore pre-
1,2-Cyclohexanedicarboxylic 62, 63
sumed that concentration-dependent behavior is the norm rather
acid, calcium salt:zinc stearatea
Anthracene 64 than the exception. Of course, there must be limitations to this
Pottasium hydrogen phthalate 46, 64 trend as product properties increasingly depend on the compos-
Benzoic acid type compounds 12, 65 ite nature of these materials at higher concentrations, causing
Sodium benzoate type compounds 12 the additional amount of NA to have diminishing returns at high
Disodium bicyclo[2.2.1]heptane-2,3- 66 loading.
dicarboxylateb
1,3:2,4-bis(3,4-dimethylbenzylidene) 66
In some studies the concentration of NA necessary to achieve
sorbitolc the largest increase in Tc and crystallization rate was deter-
Zinc monoglycerolated 67 mined. For example, Song et al. [4] showed with plots of crys-
tallization time versus Xt that the crystallization rates of the
POSS, polyhedral oligomeric silsesquioxanes.
a composites vary with the concentration of UHMWPE. The sam-
For example, HyperformTMHPN-20E.
b
For example, HyperformTM HPN-68L. ple with 3 wt% UHMWPE had both the highest crystallization
c
For example, MilladTM 3988. rate and the shortest time to reach maximum relative crystallin-
d
For example, IrgastabTM 287. ity. As the level of UHMWPE was gradually increased from 0.5
to 3 wt% the crystallization rate continued to increase as shown
researchers to identify common characteristics of the more in Fig. 6, with the crystallization rate highest for 3 wt%. How-
effective NAs [12, 50]. Prior to crystallization these NAs are ever, as the 3 wt% concentration is approached the increases in
typically insoluble in the melt with a higher Tm than the poly- crystallization rate become less and less significant. Based on
mer. These NAs are often crystalline materials with preferably these results the authors conclude that above 3% no significant
similar crystalline structure to the polymer. The complementary increases in crystallization rate would be expected [4].
registry enables epitaxial crystallization on the NA surface [12]. Similarly, Minkova and Mihailov studied the nonisothermal
Effective NAs frequently have both polar and nonpolar attrib- crystallization kinetics of blends of HDPE with UHMWPE as a
utes, such as salts of benzoic acid which have nonpolar benzene function of blend composition, demonstrating effectiveness of
rings that align in parallel layers or rows that interface with the 30 wt% of UHMWPE; both HDPE and UHMWPE crystalliza-
polymer chains. These salts of benzoic acid can form an alter- tion rates were found to be higher than their corresponding neat
nating polar and nonpolar layered structure, with the nonpolar systems [70, 71].
portion in close contact with the polymer melt. NAs for HDPE A maximum crystallization rate was determined in an iso-
are divided into inorganic and organic types as shown in Table thermal study of POSS, which included a concentration range of
1. The list of NAs in Table 1 is not exhaustive but does include 5–10 wt% of POSS in HDPE. The maximum crystallization rate
the most commonly researched and reported NAs proven effec- was determined to be 1 wt% as determined by measuring T1/2
tive for HDPE. Whether inorganic or organic, the effectiveness values by DSC [25]. However, a different concentration depend-
of the NA is determined by measuring the Tc on cooling in a ence of POSS versus Tc was found in a separate study under
nucleated HDPE sample. A higher Tc value indicates faster crys- nonisothermal conditions, which showed that only a sample
tallization and, therefore, a more effective NA. Each NA listed with 10 wt% POSS had significant increases in Tc. This high-
in Table 1 includes one or more of the common characteristics lights issues in comparing various systems and techniques, in
stated above. Other properties such as geometrical form, opti- particular the range of results possible when studying crystalli-
mum particle size, and surface morphology can increase the zation under isothermal and nonisothermal conditions [54].

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 547

 FIG. 6. Plot of relative crystallinity of UHMWPE/HDPE composites at (A) 121.58C and (B) 122.08C isothermal
crystallization temperatures. (Reproduced from Ref. 4, with permission from Elsevier).

Reduced Degree of Crystallinity sisal fiber composites showed higher Xc as reflected in higher
In most studies, a reduction in overall crystallinity as a func- DHm [60]. In some studies involving POSS [25] and BaSO4
tion of increasing NA concentration was reported as compared [26], crystallinity measurements were not included and therefore
to neat HDPE. Typically although with a few exceptions, the no conclusions could be determined regarding this effect for
crystallinity decreases as the NA concentration increases as those NAs.
compared to the neat HDPE. The reduction in crystallinity is The effect of MWNTs in one study showed only slight dif-
explained by the competing effect of reduced polymer chain ferences in degree of crystallinity that were not considered sig-
mobility in the presence of a NA. The particles can impede or nificant by the authors [55]. However, in another study
confine the mobility of polymer chains making it more difficult conducted under different conditions the MWNTs showed a
for the chains to arrange themselves into the ordered crystal decrease in crystallinity with increasing MWNT concentration
structure. Generally, the free volume of the polymer chains is [72]. The concentration effect of UHMWPE as NA on degree of
reduced as the NA content increases, which leads to decreased crystallinity was too small to be considered significant. The
polymer chain motion. The confined motion of the chains can UHMWPE samples with 0.5, 2, and 3 wt% levels showed about
cause a slight reduction in crystallization rate particularly as the 1% reduction in crystallinity as compared to the pure HDPE [4].
loading of the NA is increased above the optimum level for NA
effectiveness [50]. For example, this effect was reported to The Effect of NAs on Crystal Morphology and Structure
occur with CaCO3 and the reduced crystallinity was confirmed The focus here is mostly on the crystal morphology which
by DSC and wide angle X-ray diffraction (WAXD) curves. The refers to the external form or shape of the crystal in the macro-
crystallinity of the simple blend of HDPE with 5 wt% CaCO3 scopic scale, whereas crystal structure refers to the molecular or
showed a reduction in crystallinity as compared to the neat atomic level of order within the crystal [8]. Crystallization
HDPE. However, further reductions in crystallinity were kinetic studies as indicated by n often show that the addition of
obtained in a sample containing a maleic anhydride PE (manPE) NAs does not change the crystalline morphology of HDPE. For
compatibilizing agent with a HDPE/manPE system blended also example, in SiO2 [34] and TiO2 [33], n did not change signifi-
with 5 wt% CaCO3. The authors hypothesized that the manPE cantly in the nucleated versus pure HDPE, indicating that the
compatibilizer leads to increased interaction between the HDPE geometry or mechanism of HDPE crystallization did not change
chains and CaCO3, which increases the confinement of the poly- [33]. Contrary to this result sisal fibers showed significantly
mer chains [43]. Similar reductions in crystallinity with increas- higher n of about 3 in composite samples as compared to the
ing NA amount were determined in studies with TiO2 [33], pure HDPE which showed an n value of about 2 [60].
expanded graphite (EG) [52], and MWNTs [72]. WAXD is another tool used to characterize changes in crys-
In contrast, the percent crystallinity of the HDPE portion of talline structures in HDPE upon addition of NA. For example,
an HDPE/sisal fiber composite increased significantly as the WAXD of the TiO2 [33], CaCO3 [43], carbon nanotube (CNT)
amount of sisal fiber in the HDPE increased. However, the [56], and BaSO4 [26] NA studies showed no change in crystal
crystallization enhancement may be different at the fiber/poly- structure. However, the lamellar thicknesses can change, which
mer interface as compared to the bulk of the matrix. This result is related to the size of the particular crystal morphology rather
was contrary to typical studies of NAs in HDPE in which a than structure. In some studies, WAXD was also used to mea-
decrease in crystallinity of the HDPE fraction is measured with sure lamellar thickness. For example, WAXD showed that the
increasing filler content [60]. lamellar thickness increased in both TiO2 and BaSO4 composite
The change in crystallinity due to the presence of NAs corre- samples. Increased lamellar thickness in nucleated composites
lates to changes in DHm, which is the enthalpy change during corresponds to higher enthalpy change or DHm as compared to
melting. These changes are reflected in Xc or DHm both meas- the pure HDPE at the same cooling rates [33, 43].
ured by DSC. Studies of NAs such as CaCO3 [51], SiO2 [34], Lamellar thickness typically increases in nucleated samples
graphite [52], and TiO2 [33] showed lower Xc results, while as compared to the pure HDPE. Some studies include results of

548 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

 FIG. 7. Crystal morphology of neat HDPE (A) and HDPE/UHMWPE (B) composites after 50 min at 1298C. The
white bar is 50 mm. (Reproduced from Ref. 4, with permission from Elsevier). [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]

Tm values of nucleated versus non-nucleated samples. For exam- portional to T0 – Tc. For example, the spherulite growth rate
ple, BaSO4, UHMWPE, and CNTs as NAs showed higher Tm increased in an HDPE/CaCO3 sample compared to the neat HDPE,
than pure HDPE, which indicated that the NA caused thicker which corresponds to an increased rate of crystallization [51].
crystals or more perfect crystals to form as compared to the
pure HDPE [26, 56]. Contrary to this result both graphite and The Effect of NA Particle Size and Particle Surface
SiO2 showed a 28 and 58 decrease in Tm, respectively, compared Nanosized particles can in principle produce higher effi-
to pure HDPE, indicating increased crystal imperfections due to ciency of nucleation as lower concentrations provide a higher
the presence of the NA [34]. The Tm values of CaCO3 [43] and number of available nucleating sites due to the increase in sur-
MWNT [55] showed no significant changes in Tm versus the face area to volume ratio for decreasing particle size [73]. Nano-
neat HDPE. particles are different from typical reinforcing agents used in
polymers in that the nanoparticle is capable of interacting with
Spherulite Size and Number the polymer on a molecular level due to its small size and high
Early studies in the 1970s utilized a microscopy technique to surface area. Nanoscale NAs are added at low concentrations,
evaluate the nucleating effect of roughly 2000 compounds on typically 1–3 wt%. However, under certain conditions and con-
polyolefins. While the testing focused on PP, substances active centrations, a reduction in mobility of the molecular chains can
in PP were often active in other polyolefins such as HDPE. The occur which reduces crystal growth rate and NA effectiveness
method involved observing the size and number of spherulites [25, 43]. For example, studies involving both EG and untreated
in the crystallizing mixtures. The results showed that a good graphite (UG) have shown that particle surface area has an
NA reduced the spherulite diameters by 1/10 to 1/5 as compared effect on the NA performance independent of surface chemistry
to the spherulite diameters in the non-nucleated polymer. Bins- [52, 53]. At 10 wt%, a larger increase in Tc occurred with the
bergen [46] showed that the number of active seeds is strongly EG than with the UG as compared to the neat HDPE. The
temperature dependent. authors concluded that the higher surface-to-volume ratio of the
The modern microscopy technique used to study the crystalli- EG as compared to UG in the HDPE matrix led to a higher
zation process utilizes POM and a hot stage to allow real-time number of nucleation sites. The higher surface-to-volume ratio
capture of the crystallization process at selected temperatures [4]. may yield better interaction between crystalline HDPE and
For example, POM was used to study the crystallization process graphite [52].
of neat HDPE and 3 wt% UHMWPE composites at 1298C at Typically inorganic NA studies describe changes that are
selected time intervals as shown in Fig. 7. Figure 7 shows the made to the NA particle surfaces. The surface coating or lack of
POM results after 50 min of crystallization time at 1298C. The surface coating on NAs such as CaCO3 affects the crystalliza-
POM results provided further confirmation of the much faster tion of HDPE. In one study a precipitated CaCO3 with no ste-
crystallization rate of the composite sample with 3 wt% aric acid surface treatment acted as a NA whereas stearic acid-
UHMWPE showing more nuclei versus neat HDPE at a given treated CaCO3 did not [74]. The SiO2 study utilized a silane
time interval. These results showed that the spherulites in the coupling agent to modify the particle surface; however, the
UHMWPE composite were much smaller than those in the neat study did not focus on how the surface changes affected the NA
HDPE [4]. The POM technique was also utilized in the TiO2 and performance [34]. In another study, BaSO4 particles were
CaCO3 studies and showed similar increases in the number of treated with sodium stearate as this treatment is widely used to
nuclei and smaller spherulite sizes and these results are presumed improve the dispersion and toughness in the matrix as compared
to occur to some extent with all effective NAs [33, 43]. to untreated particles [26]. Without systematic study, it was pre-
Spherulite growth rates typically increase in samples with NA sumed that the treated BaSO4 particles would function better as
as compared to the pure polymer and growth rate is inversely pro- NAs as compared to untreated particles. Sisal fibers were treated

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 549

with an ionomer coupling agent in order to improve fiber–
matrix bond strength and improve the reinforcement effects of
the sisal fibers in the matrix. The NA effectiveness of these par-
ticles was determined; however, the study did not focus on how
the changes to the particle surface affected the NA performance
[60].

The Effect of NA Dispersion

Effective dispersion of the NA in HDPE is critical to maxi-
mizing its effectiveness [16, 46]. Huang et al. [43] studied noni-
sothermal crystallization of neat HDPE, HDPE/CaCO3, HDPE/
maleic anhydride (man), and HDPE/manPE/CaCO3 composites.
The manPE is a maleic anhydride-modified material which acts
as a compatibilizing additive for the CaCO3 in the HDPE. DSC
results showed that the HDPE/manPE/CaCO3 sample had the
highest T0 and Tp values of all the samples. The authors FIG. 8. Nonisothermal crystallization activation energy of neat HDPE and
hypothesize that the increased compatibility caused by the HDPE/TiO2 composites. (Reproduced from Ref. 33, with permission from
manPE produced better overall CaCO3 particle dispersion in Springer). [Color figure can be viewed in the online issue, which is available
at wileyonlinelibrary.com.]
HDPE, which led to an increased number of effective nucleating
particles [43].
X-ray diffraction and TEM studies with POSS as a NA Friedman equation, is shown in Fig. 8 as a function of Xt. The
showed that crystalline aggregates start to form at concentrations HDPE/TiO2 composites all show lower activation energy at any
above 1wt%. These aggregates did not effectively promote the given Xt, demonstrating that TiO2 promotes the crystallization
crystallization process or nucleation activity, resulting in con- of HDPE. The HDPE/anatase TiO2 blend has the lowest activa-
stant n values at >1 wt% POSS. The authors showed that only tion energy. The results shown in Fig. 8 suggest that the HDPE
well-dispersed POSS affects the crystallization mechanism and becomes more difficult to crystallize toward the end of the pro-
the rate of crystallization of the HDPE matrix and acts as a NA cess as activation energy tends to level out at roughly 60–90%
[25]. Xt. The CaCO3 NA shows similar trends in activation energy
CNTs are commonly known to be difficult to disperse and [33, 43].
studies have shown that only well-dispersed CNTs can function
as effective NAs [56]. Vega et al. [55] studied nanocomposites PRACTICAL APPLICATIONS
of HDPE melt mixed with an in situ polymerized HDPE/multi- Much of the published information on use of NAs in HDPE
wall carbon nanotube (MWNT) masterbatch. The nanocomposite is found in the patent literature, rather than textbooks and jour-
was studied under isothermal conditions containing 0.52wt% nals. The patent literature describes practical industrial applica-
MWNT. The in situ technique was utilized to prevent CNT tions of NAs in HDPE in which the addition of certain NAs
aggregation and optimize dispersion in the final composite and improved film clarity and haze, increased modulus, improved
results from simple melt blended systems may differ signifi- barrier properties, reduced shrinkage, warping, and sag in
cantly. Trujillo et al. [56] showed that simple melt blended extruded and molded parts, and reduced polymer processing
incorporation of MWNTs into HDPE had poor dispersion and times in molding and extrusion operations. Several representa-
did not form crystals with higher Tm and the crystal lamellae tive examples are detailed below.
were thinner as compared to the in situ produced composites. UHMWPE can prevent sag in large pipe extrusion processes
He et al. [72] completed a similar study but used a solution using PE-100 bimodal HDPE resin as well as improve produc-
method to crystallize the HDPE onto MWNTs to overcome the tion cycle time by decreasing the solidification or crystallization
poor dispersion problem. The method involved cooling and crys- time [4]. Nonuniform cooling of the inside versus outside surfa-
tallizing dissolved HDPE from hot p-xylene with a controlled ces of pipes during extrusion leads to internal stresses. In nor-
amount of dispersed MWNTs [72]. mal pipe production water is used to rapidly cool the inside and
outside of the pipe which causes differing cooling gradients
Activation Energy within the pipe wall. This uneven cooling causes stresses within
Most studies with a few exceptions confirm that NAs cause a the wall of the pipe due to slower crystallization and nonuni-
reduction in activation energy necessary for nucleation. For form shrinkage in those regions. It is commonly known in the
example, with CaCO3 the activation energies, estimated using pipe industry that higher crystallization rates are favorable for
the Friedman equation, were lower as compared to the neat eliminating or reducing the sag problem with PE pipe resins. A
HDPE at the same relative crystallinity. The composite with high crystallization rate leads to fast hardening of the molten
only CaCO3, as mentioned above promoted more spherulite polymers and reduces the residual stresses caused by differing
nucleation on the CaCO3 surface than the CaCO3/HDPE blend shrinkages in the surfaces versus core regions. For example, the
with maleic anhydride compatibilizer [43]. This effect is also crystallization rate of a HDPE/3wt% UHMWPE blend increased
evident in TiO2 used as NA, comparing anatase and rutile significantly as indicated by T1/2 values decreasing from 13.4
phases of TiO2. The plot of crystallization activation energy of min in the pure HDPE to 5.7 min in the blend. The higher crys-
neat HDPE and HDPE/TiO2 composites, as determined by the tallization rate achieved by the addition of UHMWPE led to a

550 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

reduction in pipe sag and faster pipe hardening. Similar benefits HPN-20E for HDPE (calcium hexahydrophthalate) primarily to
may be achieved in other types of industrial extrusion processes reduce cycle time and improve blow molding performance.
where sag or residual stresses occur [4, 75, 76]. HPN-20E is a high aspect ratio rod-shaped organic salt with an
HDPE film that was nucleated with a 0.2 wt% 1,3:2,4- average particle width of 2.8 lm (see also NA Table 1). The
bis(3,4-dimethylbenzylidene sorbitol NA was found to have unique shape of this NA enables its particles to align in the
enhanced clarity. The crystallization temperature increased from polymer flow direction especially in melt processes with nontur-
109.18C in the virgin HDPE resin to 109.88C in the nucleated bulent laminar flow. HPN-20E has been used as a NA in a wide
resin indicating its effectiveness as a NA. The resulting clarity variety of polymer melt processes including extrusion, blown
values improved significantly from 70.4% in the virgin resin to film, injection molding, and extrusion blow molding and is typi-
94.9% in the nucleated resin. Haze values dropped from 99.1% cally used in packaging applications [3].
in the virgin resin to 88.7% in the nucleated resin. It is pre- As with all effective NAs, HPN-20E causes an increased
sumed that the smaller spherulites that result from the NA peak Tc when present in HDPE. Milliken reported a Tc of
account for these optical changes [51]. 121.48C with the HPN-20E/HDPE blend as compared to
The addition of NA can affect the flexural modulus of 118.58C for the neat HDPE sample. Studies showed the
HDPE. HDPE containing three NAs, 0.05 wt% disodium bicycle increased Tc value enabled substantial reduction in cycle times
[2.2.1]hepatane-2,3-dicarboxylate, 0.45 wt% impact PP copoly- (10–20%) in the injection molding of caps and closures [3].
mer, and 0.5 wt% CaCO3, was injection molded into bars and Studies were also conducted on injection blow-molded bot-
the resulting flexural modulus was measured at differing storage tles using HDPE nucleated with HPN-20E. The nucleated
times at room temperature after injection molding. The flexural HDPE showed more consistent processing and improved dimen-
modulus increased in the nucleated bars at all storage times. For sional stability as indicated by an improved neck ovality ratio
example, after 10 h the flexural modulus went from 610 MPa in enabling approximately 10% cycle time reduction [6, 7]. Like-
the virgin resin to 655 MPa in the nucleated sample. This wise, a 15% improvement in water vapor and oxygen barrier
increase in modulus could be a product of the resulting spheru- properties is achieved in a substantially linear nucleated HDPE/
lites size or the degree of crystallization of these injection- HPN-20 film as compared to neat film of the same HDPE resin.
molded parts. However, no explanation was provided to explain The most dramatic improvement in barrier properties applies
the reason for the increased modulus [66]. only to a specific linear HDPE resin with minimal long chain
Likewise, extrusion blow molding studies were carried out in branching (LCB) and a narrow molecular weight distribution
order to measure cycle time improvements in 350 ml HDPE [67]. In one example, 0.1 wt% HPN-20E is melt blended into
bottle production. The minimum cycle time was recorded in the linear HDPE and converted into a 1.25 mil thick film on a
each case, which was defined as the shortest cycle time that blown film line. A control film is also produced under the same
repeatedly yielded bottles without defects. Each tested sample processing conditions without the NA. The improvements in
involved reducing the cycle time by 0.5 s and then evaluating both water vapor and oxygen barrier were 46 and 50%. Similar
the quality of the produced bottle. These studies showed that the improvements in these barrier properties were also achieved
cycle time could be reduced by 25% or from about 12 to 9 s in with zinc monoglycerolate (IrgastabTM 287) as NA. The combi-
a sample nucleated with 1000 ppm of the disodium bicyclo nation of mica and HPN-20E in HDPE showed synergistic
[2.2.1] heptane-2.3-dicarboxylate [66]. This enhancement is improvements in oxygen barrier properties. The improvements
most likely a result of NA increasing the rate of crystallization. were better than the oxygen barrier values obtained with each of
In patent application US2012/0101209 halloysite nanotubular the components used alone. The patent application hypothesizes
clay (HNT) is claimed to be an effective NA in HDPE at 0.1–1 that the HPN-20 does not provide an improved physical barrier
wt% loading resulting in both reduced postextrusion shrinkage to oxygen but rather causes improved orientation of the crystal-
and improved clarity The preferred HNT has a median particle line lamellae which then leads to a more effective barrier to
size of 0.01 to 20 lm (d50) and is melt mixed into the HDPE. oxygen. The mica component may provide a physical barrier
Only a single cooling rate of 2.58C/min was evaluated. The Tc and increase the orientation of the crystalline lamellae for
was increased by about 1.98C versus the neat HDPE, corre- improved barrier properties [77].
sponding to an increased crystallization rate. The 1.98C increase This improved cycle time is expected in any type of extru-
is considered quite significant and would be greater at higher sion process, including blown film extrusion. Dotson [78]
cooling rates. The patent application states that this NA can pro- reported a drop in frost line height in LLDPE in a blown film
vide a similar benefit in many types of polymer processes process which was attributed to the faster crystallization kinetics
including extrusion, injection molding, spinning, and clear film in a nucleated LLDPE sample with Hyperform HPN-20E. Simi-
production. In film extrusion the faster crystallization rate pro- lar improvements would be expected in HDPE blown film. The
duces smaller and less developed spherulites, which improves drop in frost line height can allow for increases in extruder
film clarity. The increased rate of crystallization also yields throughput or line speed. Line speed increases of 5–20% were
reduced postextrusion shrinkage by reducing secondary crystalli- seen in one particular trial but these results would also depend
zation processes that can cause further volume reduction and on other factors such as machine configuration and process con-
shrinkage in the solid state. A significant reduction in injection ditions [6, 7]. DSC studies showed that the crystallization tem-
molding cycle time for samples with the HNT was observed, perature was increased by about 58 in LLDPE with HPN-20E as
including a 24% reduction in mold cycle time with the HDPE compared to the 38 in a HDPE/HPN-20E sample. Both increases
sample containing 1 wt% HNT [59]. are highly significant. Nucleated film thickness variation may
Over the last decade, Milliken & Company developed and also be improved due to more uniform crystallization of the
commercialized an effective NA product called HyperformTM polymer with NA [78].

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 551

 A combination of NAs can reduce both the cycle time and how to achieve better control over it by selecting or creating
coefficient of friction with a synergistic combination of 1 wt% NAs.
talc as a NA and 0.1 wt% behanamide as slip agent melt mixed To this end a deeper understanding of the affects of NA molec-
into bimodal HDPE. The preferred talc had a mean particle size ular structure and physical affects that occur during the crystalliza-
(d50) of 1–10 lm. Increased Tc values were measured with the tion process are needed. Although not yet at the level of control
combination of talc and behanamide as compared to the Tc val- that is reached with NAs in PP, progress has been made in identi-
ues with the talc or behenamide alone. The patent application fying effective NAs for HDPE and achieving a range of property
example shows that the neat resin has a Tc of 117.58C versus improvements as was seen in many examples in this article.
118.4 and 118.88C for the HDPE/1 wt% talc and HDPE/0.1 Optimizing these properties leads to reduced spherulite sizes
wt% behenamide samples, respectively. However, the combina- and an increased Tc. It is not yet known whether spherulite sizes
tion of 1 wt% talc plus 0.1 wt% behenamide has a much higher can be further reduced or whether further increases in Tc can be
Tc of 120.28C. The coefficient of friction also improved with the achieved to approach levels of improvement similar to those
HDPE/talc/behenamide sample as compared to the neat HDPE, reached with nucleated PP.
and the samples with talc and behenamide alone [79]. Gaps in understanding the optimum molecular structure for
Notably it has been shown that a NA can have a significant achieving optimal NA effectiveness in HDPE also need to be
negative effect, particularly on the barrier properties. Utilizing a resolved. Only general principles are known about how molecu-
NA called IrgastabTM NA 11 [methylene-bis-(4,6-di-tert-butyl- lar structure and characteristics impact NA efficiency as out-
phenyl) phosphate sodium salt] caused an 11% reduction in water lined in this article, including solubility/insolubility, polarity,
vapor transmission and a 23% reduction in oxygen transmission and dispersion within HDPE. These characteristics relate to
versus the control sample. Also comparative examples of nonlin- molecular structure in a more qualitative sense only. Discover-
ear HDPE and HPN-20E with LCB > 0.5 and MFR > 65, both of ing the optimum molecular characteristics for nucleating HDPE
which are outside of the claimed ranges, showed relatively little could lead to a custom-tailored NA molecule. An ultimate goal
improvement in both water and oxygen barrier properties. The in this field would be to understand the crystallization process
reason for the specificity of this NA was not discussed. such that NAs could be engineered specifically to optimize NA
melt dispersion, nucleation, crystallization, and final crystalliza-
Processing and Melt Flow Effects tion fraction using a cost-effective method.
In polymer processing, flow of material in the melt state can
REFERENCES
also play a major role in the crystallization of semicrystalline
polymers such as HDPE. Therefore, the specific thermal and 1. E. Ergoz, J.G. Fatou, and L. Mandelkern, Macromolecules, 5,
rheological conditions imparted onto the polymer during proc- 147 (1972).
essing, which include shear stresses, elongational effects, and 2. K.H. Buschow, R.W. Cahn, M.C. Flemings, B. Ilschner, E.J.
molecular orientation, can be used to control the final properties Kramer, S. Mahajan, Encyclopedia of Materials—Science and
of the HDPE. Flow-induced crystallization is beyond the scope Technology, Volumes 1-11: Polyethylene: High-Density, Elsev-
of this article and is only briefly mentioned here. ier, Amsterdam, 7178 (2001).
In the melt state HDPE polymer chains are easily deformed, 3. N. Farmer, Trends in Packaging of Food Beverages and other
elongating and orienting themselves in the direction of polymer Fast-Moving Consumer Goods (FMCG)—Markets Materials
flow. This polymer chain alignment and stretching affects the and Technologies, Woodhead Publishing, Oxford, UK, 85
kinetics, thermodynamics, and final crystal morphology of (2013).
HDPE. Generally, crystallization processes that occur during 4. S. Song, P. Wu, M. Ye, J. Feng, and Y. Yang, Polymer, 49,
polymer flow show increased Tc, nucleation, and crystallization 2964 (2008).
rates. The behavior of NAs during deformation is further com- 5. K. Nezhad, H. Orang and M. Motallebi, “The Effects of Adding
plicated by particle–particle interactions mediated by hydrody- Nano-Calcium Carbonate Particles on the Mechanical and
namics and viscoelasticity. Control of final properties of HDPE Shrinkage Characteristics and Molding Process Consistency of
may be achieved by improved understanding of structure–prop- PP/nano-CaCO3 Nanocomposite,” in Polypropylene, F. Dogan,
erty relationships that result from variations in processing type Ed., InTech, Rijeka, Croatia (2012).
and conditions [80]. 6. W. Van de Velde, “Newest Developments in the Nucleation of
Polyethylene and Its Effects on Physical Properties Perform-
CONCLUSION AND FUTURE OUTLOOK ance,” in SPE—International Polyolefins Conference, Houston,
The crystallization process in HDPE was shown to signifi- March 1 (2011).
cantly affect its optical, mechanical, and thermal properties. The 7. M. Tolinski, Additives for Polyolefins: Getting the Most out of
amount of crystallization and the type of crystals formed are con- Polypropylene, Polyethylene, and TPO, William Andrew,
trolled by thermal history, cooling rate, and NA additives. Owing Oxford, UK, 166 (2009).
to widespread use of HDPE in many applications there is contin- 8. B. Wunderlich, Thermal Analysis of Polymeric Materials,
ued interest in gaining a deeper understanding of the crystalliza- Springer, Berlin, Heidelberg, 262 (2005).
tion process. Greater control over the crystallization process is 9. A. Peacock and A. Calhoun, Polymer Chemistry—Properties
needed in order to achieve further improvements in specific prop- and Applications, Hanser Publishers, Munich, 125 (2006).
erties. However, it will first be necessary to overcome significant 10. M.L. Di Lorenzo and C. Silvestre, Prog. Polym. Sci., 24, 917
gaps in current understanding of the crystallization process and (1999).

552 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen

11. P.E. Malacari, Effect of Both Talc Fineness and Talc Loading 37. K. Nakamura, K. Katayama, and T. Amano, J. Appl. Polym.
on Heterogeneous Nucleation of Block Copolymer Polypropyl- Sci., 17, 1031 (1973).
ene, SPE Antec, Orlando, FL, 2264 (2010). 38. J.A. Martins and J.J.C. Pinto, Polymer, 41, 6875 (2000).
12. J.P. Mercier, Polym. Eng. Sci., 30, 270 (1990). 39. A. Jeziorny, Polymer, 19, 1142 (1978).
13. M. Dong, K. Xu, H. Zou, and L.Y. Zhang, “Excellent Proper- 40. I.A. Hussein, J. Appl. Polym. Sci., 107, 2802 (2008).
ties of a Novel Composite b-Nucleating Agent for Isotactic
R 41. M. Eder and A. Wlochowicz, Polymer, 24, 1593 (1983).
Polypropylene,” ANTECV 2013—Proceedings of the Technical
Conference & Exhibition, Society of Plastics Engineers, Cincin- 42. T. Liu, Z. Mo, S. Wang, and H. Zhang, Polym. Eng. Sci., 37,
nati, OH, 2180 (2013). 568 (1997).
14. V. Cornelia and P. Mihaela, Practical Guide to Polyethylene, 43. J.W. Huang, Y.L. Wen, C.C. Kang, W.J. Tseng, and M. Yeh,
Smithers Rapra Technology, 94 (2005). Polym. Eng. Sci., 48, 1268 (2008).
15. D. Kang, Proceedings of the 58th International Wire and Cable 44. G.M. Benedikt, Metallocene Technology in Commercial Appli-
Symposium, 210 (2008). cations—34.4.3 Activation Energy for Non-Isothermal Melt
16. K. Hoffman, G. Huber, and D. Mader, Macomol. Symp., 176, Crystallization by Kissinger Method, William Andrew Publish-
83 (2001). ing/Plastics Design Library, Norwich, NY (1999).
17. C. Maier and T. Calafut, Polypropylene—The Definitive User’s 45. L.A. Utracki, Clay-Containing Polymeric Nanocomposites, Vol-
Guide & Databook, William Andrew Publishing Plastics Design umes 1-2: 3.4.2 Fundamentals of Crystallisation, Smithers
Library, CRC Press, Norwich, NY, 34 (1998). Rapra Technology, NY, 396 (2004).
18. A. Peacock, Handbook of Polyethylene Structures, Properties 46. F.L. Binsbergen, Polymer, 11, 253 (1970).
and Applications, CRC Press, NY, 101 (2000). 47. H. Kissinger, J. Res. Natl. Bureau Stand., 57, 217 (1956).
19. J.M. Fischer, Handbook of Molded Part Shrinkage and Warp- 48. B. Risch, ANTEC, Plastics: Plastics on My Mind, Vol. 1: Proc-
age, Elsevier, William Andrew Publishing Plastics Design essing Influence of Nucleating Agent and Melt Flow Index on
Library, Norwich, NY, 26 (2013). the Impact Polypropylene Fiber Optic Buffer Tubes, Society of
20. K.H. Bushow, “Bulk Crystallization,” in Encyclopedia of Plastics Engineers, Atlanta, GA, 1016 (1998).
Materials—Science and Technology, Polymer Crystallization: 49. F.C. Chiu and Q. Fu, ANTEC, 3, Society of Plastics Engineers,
General Concepts of Theory and Experiments, Vol. 1-11, Elsev- New York (1999).
ier, Amsterdam, 7247 (2001). 50. Y.P. Khanna, Macromolecules, 26, 3639 (1993).
21. L.H. Sperling, Introduction to Physical Polymer Science, 2nd 51. S. Sahebian, S.M. Zebarjad, J.V. Khaki, and S.A. Sajjadi, J.
ed., Wiley, New York, 232 (1992). Mater. Process. Technol., 209, 1310 (2009).
22. P. Zou, S. Tang, Z. Fu, and H. Xiong, Int. J. Therm. Sci., 48, 52. W. Zheng, X. Lu, and S.C. Wong, J. Appl. Polym. Sci., 91,
837 (2009). 2781 (2004).
23. M. Muthukumar, “Nucleation in Polymer Crystallization,” in
53. W. Zheng, S.C. Wong, and H.J. Sue, Polymer, 43, 6767 (2002).
Advances in Chemical Physics, S.A. Rice, Ed., John Wiley &
Sons, Hoboken, NJ, 128, 1 (2004). 54. M. Joshi and B.S. Butola, Polymer, 45, 4953 (2004).
24. M.E. Brown and P.K. Gallagher, Handbook of Thermal Analysis 55. J.F. Vega, J.M. Salazar, M. Trujillo, M.L. Arnal, A.J. Miller, S.
and Calorimetry—Recent Advances, Techniques and Applica- Bredeau, and Ph. Dubois, Macromolecules, 42, 4719 (2009).
tions, Volume 5: 13.4.1 Crystallization, Elsevier, Amsterdam, 56. M. Trujillo, M.L. Arnal, A.J. Muller, E. Laredo, and S.T.
514 (2008). Bredeau, Macromolecules, 40, 6268 (2007).
25. M. Joshi and B.S. Butola, J. Appl. Polym. Sci., 105, 978 57. M. Trujillo, M.L. Arnal, and A.J. Muller, Macromolecules, 41,
(2007). 2087 (2008).
26. X. Chen, J. Shi, L. Wang, H. Shi, Y. Liu, and L. Wang, Polym. 58. S.C. Tjong and S.P. Tjong, J. Polym. Sci. Part B: Polym. Phys.,
Compos., 32, 177 (2010). 43, 253 (2005).
27. T. Kanai and G. Campbell, Film Process, Hanser Publishers, 59. Y.P. Khanna, A. Dharia, and A.M. Zeitoun, U.S. Patent
Vol. 167 (1999). 0,101,209 (2012).
28. M. Avrami, J. Chem. Phys., 7, 1103 (1939). 60. A. Choudhury, Mater. Sci. Eng., 491, 492 (2008).
29. M. Avrami, J. Chem. Phys., 8, 212 (1940). 61. H. Ishida and P. Bussi, Macromolecules, 24, 3569 (1991).
30. M. Avrami, J. Chem. Phys., 9, 177 (1941). 62. M. Horrocks, C. Kerscher, and D. Dotson, A Novel Nucleating
31. A.T. Lorenzo, M.L. Arnal, J. Albuerne, and A. Muller, Polym. Agent for Polyethylene, The Plastics And Rubber Institute of
Test., 26, 222 (2007). Singapore, Tanglin, Singapore, 29 (2008).
32. G. Liang, J. Xu, and W. Xu, J. Appl. Polym. Sci., 91, 3054 63. L.M. Sherman, “Polyolefin Additives: Pushing the Limits in
(2004). Films & Molded Parts,” in Plastics Technology, Plastics Tech-
33. S. Wang and J. Zhang, Therm. Anal. Calorim., 115, 63 (2014). nology, Cincinnati, OH (2014).
34. Q. Jiasheng and H. Pingsheng, J. Mater. Sci., 38, 2299 (2003). 64. J.C. Wittmann and B. Lotz, J. Polym. Sci.: Polym. Phys. Ed.,
35. D. Grenier and R.E.P. Homme, J. Poly. Sci. Part B, 18, 1655 19, 1837 (1981).
(1980). 65. P. Wijga, U.S. Patent 3,207,735 (Shell Oil Company) (1960).
36. C. Silvestre and S. Cimmino, Crystallization, Morphology and 66. R. Hanssen, M.B. Barker, N.A. Mehl, W.S. Wolters, S.M.
Melting in Polymer Blends, Smithers Rapra Technology, New Bernhardt, and A.T. Radwanski, U.S. Patent 7786203B2 (Milli-
York, 439 (2002). ken and Company) (2010).

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—2016 553

67. J.S. Borke, D.C. McFaddin, and S.M. Imfeld, U.S. Patent 75. P.J. Deslauriers, M.P. McDaniel, D.C. Rohlfing, R.K.
0,225,742 (Equistar Chemicals) (2013). Krishnswamy, S.J. Secora, E.A. Benham, P.L. Maeger, A.R.
68. H.N. Beck and H.D. Ledbetter, J. Appl. Polym. Sci., 9, 2131 Wolfe, A.M. Sukhadia, and B.B. Beaulieu, Polym. Eng. Sci.,
(1965). 45, 1203 (2005).
76. E. Zopf, A. Arellano, S. Tyagi, A, Tynys, J. Braun, and M.
69. H.N. Beck, J. Appl. Polym. Sci., 11, 673 (1967).
Gahleitner, WO2013060736 (Borealis AG) (2013).
70. L. Minkova and M. Mihailov, Colloid Polym. Sci., 267, 577
77. J.R. De Oliveira, A, Rangel, B. Mano, C. DeLemos, and F.
(1989).
Decarvhalho, WO2014/091309 (Braskem S.A.) (2014).
71. L. Minkova and M. Mihailov, Colloid Polym. Sci., 265, 1 78. D.L. Dotson, “A Novel Nucleating Agent for Polyethylene,” in
(1987). 2007 Proceedings of the 2007 PLACE Conference, St. Louis,
72. L. He, Q. Xu, R. Song, and C. Hua, Polym. Compos., 31, 913 MO (2011).
(2010). 79. K. Bhawna, M. Luc, and B. Laurent, WO2012/000958 (2012).
73. M. Gahleitner, C. Grein, S. Kheirandish, and J. Wolfschwenger, 80. B. Vergnes, M. Vincent, and J. Haudin, Design of Extrusion
Intern. Polym. Process., XXVI, 1 (2011). Forming Tools, Smithers Rapra Technology, Shawbury, UK, 55
74. A. Lazzeri, S.M. Zebarjad, M. Pracella, K. Cavalier, and R. (2012).
Rosa, Polymer, 46, 827 (2005). 81. A. Ziabicki, Appl. Polym. Symp., 6, 1 (1967).

554 POLYMER ENGINEERING AND SCIENCE—2016 DOI 10.1002/pen