Polymer 44 (2003) 2473–2480

www.elsevier.com/locate/polymer

Postponing sharkskin of metallocene polyethylenes at low temperatures:

the effect of molecular parameters
A. Santamarı́aa,*, M. Fernándeza, E. Sanzb, P. Lafuenteb, A. Muñoz-Escalonac
a
Department of Polymer Science and Technology, University of the Basque Country, P.O. Box 1072, 20080-E San Sebastián, Spain
b
Centro de Technologı́a, REPSOL-YPF, Ctra, N-V, Km. 18, 28931 Móstdes, Madrid, Spain
c
GIDEM (Grupo de Investigación y Desarrollo en Macromoléculas), Instituto de Estructura de la materia/CSIC, Serrano, 113bis-119, 28006 Madrid, Spain

Abstract
The effect of temperature on extrusion rheometry of single site metallocene-catalyzed polyethylenes and polyethylene copolymers is
investigated. Samples of molecular weight, Mw ; ranging from 90,000 to 330,000 and short-chain branching degree (SCB) from 0 to 21.2
CH3/1000C, as well as samples with a small amount of long-chain branching, are analyzed. It is observed that all the samples display a low
temperature region, limited by induced crystallization and gross melt fracture, in which smooth extrudates are produced at shear rates similar
to those of industrial extrusion. A characteristic temperature of this region, Ts ; is defined as the highest temperature at which sharkskin
disappears. Clear symptoms of non-slip conditions at the capillary wall, are detected in this low temperature region. We assume that the
necessary slip – stick conditions to produce sharkskin, would only be produced at shear rates above those involved in gross melt fracture. The
analysis of the effect of the molecular parameters, leads to the conclusion that only SCB has a direct effect on Ts : A linear correlation between
Ts and SCB level is established, showing the decrease of the former as the latter is increased. Considering the wide spectrum of the molecular
characteristics of our samples, we claim that decreasing temperature is a sound route to postpone sharkskin of any polyethylene.
q 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Metallocene polyethylene; Sharkskin instability; Short-chain branching

1. Introduction of the molecules, which increases cohesiveness and reduces

the tendency to sharkskin. On the other hand, from 1990 to
The quality of a polymer extrudate depends on the flow 1996 Keller et al. [2–9] published a series of papers, on what
defects during polymer melt extrusion, a subject which has they defined as a ‘temperature window of extrudability’ of
received considerable attention in the literature, in particular polyethylenes. There is a sharp minimum in applied pressure,
as to what ‘melt fracture’ and ‘sharkskin’ instabilities of so flow resistance is reduced, within a narrow temperature
polyethylenes are. range of 150–152 8C. Under these window conditions, there
The most basic strategy to eliminate or postpone flow are no other extrudate irregularities but gross melt fracture is
instabilities in polymers, is to rise the die temperature. observed. The research carried out by Keller et al. led them to
However, the opposite approach, that is a temperature state that this window effect can be traced to the formation of a
decrease has been suggested by some researchers from time low viscosity phase, identified with the mobile hexagonal
to time since the seventies. A 1972 British Patent (32559172), phase: This arises through flow-induced alignment of
commented by F.N. Cogswell in a paper in 1977 [1], molecules adsorbed to the capillary. The hypothesis developed
constitutes one of the first references of decreasing tempera- by these authors has been used to discuss experimental results,
ture as a way to eliminate instabilities. Applying cooling to the such as those of Pudjijanto and Denn [10] who found the so
die lip, to lower the temperature to just above the melting called ‘stable island’, where stable flow with smooth
point, eliminates sharkskin in polyethylene, polymethyl extrudates is achieved at 139–146 8C for linear low density
methacrylate, acetal copolymer and polyvinyl chloride. polyethylenes (LLDPE). In their turn, Wang and Drda [11]
According to this author, intense stretching flow, combined refer to the work of Keller et al. as they propose a flow-induced
with the low temperature of the surface, induces some ordering mesophase, to justify what they consider an anomaly in the
variation of the critical stress for the slip–stick transition of
* Corresponding author. Tel.: þ 34-9-3018184; fax: þ 34-9-3212236. polyethylenes in the range 160–200 8C. More recently, Pérez
E-mail address: popsaiba@sq.ehu.es (A. Santamarı́a). González et al. [12] analyzed capillary flow behavior of a
0032-3861/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0032-3861(03)00048-X
2474 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480

Fig. 1. Vanishing of severe sharkskin instability observed at T ¼ 175 8C for sample mPE6 (Table 1), as temperature is increased or decreased.

metallocene-catalyzed LLDPE with hexene as the comono- weights, ranging from Mw ¼ 92; 000 to Mw ¼ 335; 000;
mer (melting temperature, Tm ¼ 91 8C), over a wide range of different degree of SCB, ranging from SCB ¼ 0 CH3/1000C
temperatures. Interestingly enough, they find that unstable to SCB ¼ 21.2 CH3/1000C and polydispersity index of
spurt flow is avoided and a remarkable improvement on the approximately Mw =Mn ¼ 2:
extrudate appearance is obtained at a much lower temperature, Molecular weight distribution curves were determined
112–116 8C, than the aforementioned cases. The results with a Waters 150 gel permeation chromatograph, coupled
obtained with LLDPE, as well as the shift of the minimum in with differential refractive index and a 150R model
extrusion pressure to a lower temperature ðT ¼ 145 8CÞ viscosimeter detector.
observed by Waddon and Keller [2] for moderately branched A Brucker DPX-300 NMR Spectrometer operating at
PE, indicate that there is a certain effect of the molecular 75 MHz was used to measure the level of short chain
architecture on the phenomenon. branching.
In recent years, we have been working with a large series A Perking – Elmer Differential Scanning Calorimeter
of well characterized single-site metallocene-catalyzed (DSC-7) was used to determine the melting point of the
polyethylenes and polyethylene copolymers. We have samples at a heating rate of 5 8C/min.
analyzed systematically the effect of temperature on
extrusion rheometry results, arriving at the conclusion that
all the samples display a low temperature region for smooth 2.2. Rheological measurements
extrudability, whatever their molecular parameters are. In
this paper representative results are presented, analyzing the
effect of molecular weight, short-chain branching degree Capillary extrusion experiments were performed in a
(SCB) and small amounts of long-chain branching (LCB), Göttfert Rheograph 2002 rheometer using a flat entry
on the temperature range of this region. capillary tungsten die of 1 mm diameter and a L=D ratio of
30.
The surface of the sharkskin distorted extrudates was
2. Experimental part analyze by scanning electron microscopy with a Hitachi
2100 microscope.
2.1. Materials The viscoelastic functions were determined in a ARES
Rheometric Scientific rheometer operating in the oscillatory
The materials under study are metallocene-catalyzed shear mode, using parallel plates geometry with d ¼ 25 mm;
polyethylenes and copolymers of different molecular in the frequency range from 1022 to 10 Hz.

Table 1
Limit temperature for smooth extrudate, Ts ; (see text) of four completely linear polyethylenes (SCB ¼ 0 CH3/1000C) of different molecular weight Mw ; and
two polyethylene copolymers of practically the same SCB level and different molecular weight

Material Mw Mw =Mn SCB (CH3/1000C) Ts (8C)

mPE6 118,000 2.15 0 150

mPE7 158,000 2.16 0 149
mPE8 214,900 2.41 0 151
mPE9 195,000 2.0 0 145
mPE3 128,100 1.78 8.6 145
mPE10 174,500 3.28 8 145
 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480 2475

A detailed sketch of the evolution of shear stress and

extrudate characteristics with time, at different temperatures
and shear rates, is given in Fig. 3. At the higher temperatures
considered, 150 and 160 8C, four different surface features
can be observed, as the shear rate is increased: smooth,
sharkskin, slip – stick and melt fracture. An unusual
phenomenon is noticed at 150 8C, with respect to 160 8C:
smooth extrudates are obtained at high shear rates, in the
range between slip –stick and melt fracture. This result has
been reported in the literature for conventional polyethyl-
enes, and resembles the trait described by Keller et al. [2 – 9]
Fig. 2. Variation of the shear stress, taken at a constant shear rate g_21 ¼ in their ‘temperature window’ research. However, the
36:1 s21 ; as a function of temperature for sample mPE4 (Table 2). The very clear reduction of flow resistance (a dramatic
deviation at low temperatures is associated to shear induced crystallization pressure decrease, after slip – stick oscillations) reported
(see text).
by these authors, is not observed in our metallocene-
catalyzed polyethylenes. We note that at 140 8C sharkskin
3. Results and discussion instability does not appear, so that smooth extrudates can
be obtained at shear rates typical of industrial extrusion.
3.1. Description of the temperature limit for smooth This leads us to define a limit temperature for smooth
extrudates extrudate, Ts ; as the highest temperature (of the low
temperature range) at which sharkskin disappears.
Our experience with metallocene catalyzed polyethyl- Decreasing temperature to T ¼ 130 8C in Fig. 3, confirms
enes allows to assert that decreasing the temperature appears the stable flow and the absence of sharkskin over a wide
to be a suitable option to eliminate sharkskin. A representa- range of shear rates; but favors the crystallization process,
tive example is shown in Fig. 1. Starting from the very associated with the shear stress increase over time, at the
severe sharkskin at 175 8C, smooth extrudates are obtained highest shear rate.
increasing the temperature to 210 8C, but also lowering it to Gathering all the results relative to the surface charac-
145 8C. teristics of the samples, as well as the induced crystal-
Industrial polymer extrusion involves shear rates typi- lization data, temperature – shear rate maps for the surface
cally above 100 s21, which constitutes an important factor morphology of the extrudates can be drawn. These maps,
in the context of this work, because it is well known that two of which are shown in Fig. 4, allow the extrusion
shear rate can induce crystallization at temperatures slightly conditions to be defined for: (a) stable flow with smooth
above melting [13]. Therefore, attempts to eliminate extrudates, (b) sharkskin, (c) slip –stick, (d) gross melt
sharkskin by lowering the temperature, but maintaining a fracture and (e) induced crystallization. It is important to
moderate or high shear rate, can lead to a crystallization note that in the slip –stick region, shear rates are not actually
process. Clear evidence of crystallization during extrusion constant because the observed pressure oscillation leads the
at low temperatures can be seen in Fig. 2. Experiments melt to be compressed to different extents [14]. The most
performed at constant shear rate, show that at a certain interesting result is that a region of stable flow and smooth
temperature the viscosity increases significantly, deviating morphology conditions, limited at high shear rates by melt
from the Arrhenius-like viscosity – temperature dependence: fracture and at low temperatures by crystallization, is
  obtained at low temperatures and high shear rates. Maps like
E
h0 ðTÞ ¼ A exp a ð1Þ those of Fig. 4 have been obtained for all the analyzed
RT
samples, confirming that in all the cases, shear rates similar
where Ea is the activation energy of flow. to those of industrial processing can be used to obtain
Table 2
Ts and Tm data of polyethylene copolymers of very similar molecular weight, but different degrees of short chain branching. Data of Ref. [12] are also included

Material Mw Mw =Mn SCB Ts Tm

(CH3/1000C) (8C) (8C)

mPE1 128,900 1.06 4.87 153 120.0

mPE2 124,400 1.17 5.73 151 120.3
mPE3 128,100 1.78 8.58 145 117.3
mPE4 123,000 2.3 10.99 141 112.6
mPE5 115,800 2.14 21.23 117 95
EXACT Ref. [12] 116 91
2476 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480

Fig. 3. The evolution of shear stress and extrudate surface characteristics with time, at different temperatures and shear rates for sample mPE3 (Table 2): (a)
T ¼ 130 8C; (b) T ¼ 140 8C; (c) T ¼ 150 8C and (d) T ¼ 160 8C:

smooth, good quality, extrudates at low temperatures. This such a noticeable pressure decrease when smooth extrudates
clearly opens a route to facilitate the extrusion of are obtained in the low temperature region.
metallocene-catalyzed polyethylenes that are known to The upper border of the smooth zone of low temperatures
present flow instabilities at very low shear rates. The results which extends to high shear rates (Fig. 4) is marked by Ts ;
outlined in our surface morphology maps, show evident the highest temperature at which sharkskin disappears. This
similarities with low temperature anomalies found in analysis can also be considered from the point of view of the
conventional polyethylenes, such as the aforementioned effect of temperature on the critical shear stress value, ss ;
temperature window of extrudability of Keller et al. [2 – 9] for the inception of sharkskin, which in our case implies the
and the stable island (where the extrudates become smooth) compilation of the shear stress values following the smooth/
discovered by Pudjijanto and Denn. Contrary to our results sharkskin border of surface maps (Fig. 4). A plot of ss
of Fig. 4, the alleged stable island lies exclusively inside the versus temperature is given for one of the samples in Fig. 5,
slip –stick region, neither limiting with crystallization nor which also includes ss =T versus T plots to analyze the
with melt fracture. On the other hand, ‘window’ experi- linearity of the dependence. The variation is linear
ments would correspond to an ascending vertical line ( ss =T ¼ constant) only in the range 160 – 190 8C. A
(increasing temperature at an apparent constant shear rate) minimum is denoted at 145 8C, giving rise to an asymptotic
which, starting from the smooth region, crosses it and increase in both ss and ss =T as the temperature is decreased
reaches slip – stick and sharkskin regions. In the analysis from 145 to 118 8C. The asymptote of ss drawn in the
carried out by Keller et al., the applied pressure falls figure, marks the highest temperature (of the low tempera-
substantially at a certain temperature. A pressure drop is ture region) at which sharkskin is not detected, designed as
also noticed in the stable island of Pudjijanto and Denn. Ts ; also defined in Fig. 4. It is seen that at a temperature of
However, in the set of experiments performed with our 117 8C, which corresponds to Ts defined in both Figs. 4(b)
metallocene-catalyzed polyethylenes, we have not found and 5, the progressive increase of shear stress does not lead
 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480 2477

Fig. 5. ss (full symbols) and ss =T (empty symbols) versus temperature plots

for the sample of Fig. 4(b). The flow induced crystallization temperature is
determined using plots like those of Figs. 2 and 3. The data at low
temperatures are fitted to an asymptotic function. In region I (in bright),
sharkskin is forbidden and in region II (in dark) flow is blocked because
crystallization.

Considering that slip – stick is an interfacial phenomenon,

they use the expression derived by Brochard and de Gennes
[16]:

sC ¼ nKT=D ð2Þ

where n is the density of adsorbed chains and D the

entanglement distance, to assume that sC should scale
linearly with temperature. This is experimentally confirmed
over decreasing temperatures in the range 260 – 200 8C with
their HDPEs, and at lower temperature ranges (typically
from 190 to 160 8C) in our polyethylene samples, as can be
seen in Fig. 5. However, at lower temperatures, decreasing
from 200 to 160 8C, Wang and Drda [11] report an
anomalous behavior, consisting of the non-linearity of the
Fig. 4. Temperature–shear rate maps for the surface morphology of the correlation between sC and T (sC =T is lowered as T is
extrudates, obtained gathering all the results relatives to the surface decreased). This is also observed, in the range 190 – 140 8C,
characteristics of the samples, as well as induced crystallization data. (a) with a metallocene-catalyzed LLDPE investigated by Pérez
and (b) correspond, respectively to sample mPE5 (Table 2) and sample et al. [12], and in the range 160– 145 8C (see Fig. 5), for the
mPE4 (Table 2) The borders are drawn on the basis of the experimental
results marked by circles. The arrow labels the characteristic temperature,
inception of sharkskin ss ; with our metallocene-catalyzed
Ts ; defined in the text. The crosses (instead of circles) of two of the PE samples. Considering that at low temperatures the
experimental data in (a), indicate that smooth samples are obtained not density of adsorbed chains, n; is only expected to increase as
withstanding pressure oscillation is observed. A comment on the reliability temperature decreases, Wang and Drda argue that the
of data in the slip– stick region is made in the text. decrease of sC =T with temperature, should be explained by
assuming an increase in the distance between entangle-
to sharkskin, even though gross melt fracture is detected at
ments, D: They suggest that the formation of a flow-induced
high shear rates. Therefore, in Fig. 5 we define a stable flow ordered phase, near the PE/wall interface, may be the origin
zone, delimited by Ts and the crystallization temperature, in of this anomaly, since chains would disentangle easier due
which sharkskin is forbidden. This leads us to estimate that to local orientation. Therefore, the research of Keller et al.
within this temperature range, the tensile stress at the on the formation of a phase through flow-induced alignment
entrance of the die necessary to gross melt fracture [15], is of presumably adsorbed molecules, which gives rise to the
developed at a lower shear rate than the critical shear rate temperature window of extrudability, supports the observed
for sharkskin. All the metallocene-catalyzed PEs studied decrease of sC =T as temperature is lowered. But the rapid
showed the same general behavior, although the limiting increase of both, ss and ss =T; as temperature is decreased
temperatures vary according to the characteristics of each from 145 to 118 8C (Fig. 5), cannot be explained solely by
sample. In particular, the effect of the molecular parameters the formation of a mesophase. Actually, considering Eq. (2)
on Ts is analyzed below. the increase of sC =T implies the reduction of D and/or an
Wang and Drda [11] analyze the variation of the critical increase in the density of adsorbed chains, n: The process
stress value of the slip – stick, sC ; with temperature. may imply incipient crystallization close to the wall. Our
2478 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480

Fig. 6. Correlation between the highest temperature at which sharkskin Fig. 7. Variation of Ts as SCB level is increased. The molecular parameters
disappears, Ts ; and melting temperature, Tm : Symbol O represents the result of the samples are given in Table 2.
of a commercial metallocene catalyzed LLDPE, taken from reference [12].
The molecular parameters of the samples are given in Table 2.
window’, have been reported for high molecular weight
guess is that stretching at the exit (which combined with (2.5 £ 105 –106) and polydisperse conventional polyethyl-
adhesion conditions at the wall gives rise to sharkskin) may enes. Within the framework of the explanation offered for the
promote local crystallization at these low temperatures. In temperature window, Waddon and Keller interpret the effect
any case, the evolution of the shear stress for inception of of SCB in terms of the –CH3 groups lowering the temperature
sharkskin, ss ; with temperature, indicates that adhesion is for a mobile hexagonal mesophase.
enhanced at low temperatures. In the range of smooth stable The effect of short chain branching on Ts is depicted in
extrudates, between Ts and crystallization temperature, the Table 2, where samples of very similar molecular weight
extrudate swell of the samples increases with shear rate and and polydispersity, but different levels of SCB, are
lies typically within the range from 1.2 and 1.4: this is a considered. Since it is known that chain defects, in the
symptom of non-slip conditions at the capillary wall. form of alkyl branches, lower the melting temperature [18],
Therefore, according to the results of Figs. 4 and 5, the in Table 2 we have also included melting values, Tm ; as
necessary slip – stick conditions to produce sharkskin, would determined by DSC. The correlation between the melting
only be produced at shear rates above those involved in temperature and the highest temperature (of the low
gross melt fracture. To the best of our knowledge, such temperature range) at which sharkskin disappears, Ts ; is
behavior has never been reported in the literature. Only the shown in Fig. 6. Also shown in this figure is the equivalent
result reported by Ramamurthy for highly branched LDPE of Ts (defined in this case as the lowest temperature for total
[17] resembles in some sense to our finding: this author elimination of surface defects) of a commercial metallocene
observes that the shear rate for the onset of gross melt catalyzed LLDPE, taken from Pérez et al. The very
fracture is extremely close to that of sharkskin. significant decrease of Ts as the SCB level is increased,
can be observed in Fig. 7, which also shows the similar
effect of short branches on Tm : Unfortunately, the result of
3.2. Effect of molecular architecture on Ts the LLDPE used by Pérez et al., shown in Table 2 and Fig. 6
can not be plotted in this figure, because its degree of
The following molecular parameters of all the samples branching is not known. The correlation between Ts and
studied in this work, are considered: weight average SCB level is given by the equation:
molecular weight, Mw ; which ranges from 92,000 to
Ts ¼ 164 2 2:2 SCBðCH3 =1000CÞ ð3Þ
335,000; polydispersity index, Mw =Mn (where Mn is the
number average molecular weight), always close to 2; and This equation provides a clear guide to find the temperature
short chain branching level, SCB, which ranges from 0 to at which smooth extrudates can be obtained, on the basis of
21.2 CH3/1000C. The presence of small amounts (typically the level of short branches, that is to say the presence of
less than one branch per molecule) of long chain branches is hexane comonomer units.
also considered. Our experimental results cover all the singular
Our results show that there is no effect of the molecular temperatures reported in the literature for smooth extrud-
weight on Ts ; over the range of Mw investigated. As can be seen ability: temperature window of extrudability; stable island
in Table 1, for the same or similar SCB level, the same value of of Pudjijanto and Denn and the ‘easy flow’ of Pérez et al.
Ts is observed for different molecular weights. It is also noticed The window effect has been extensively investigated by
that Ts decreases when short branches are incorporated in the Keller et al. from 1990 to 1996, but with a reduced number
chain. A substantial effect of SCB [2,5,6], and a practically of linear, polydisperse, polyethylene samples of Mw ranging
insignificant effect of molecular weight [5,6], on the from 130,000 to 1,000,000 and one PE with a methyl branch
temperature location of pressure minimum of the ‘extrusion density of 4.9 CH3/1000 carbons. The specific temperatures,
 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480 2479

Fig. 8. Complex viscosity as a function of frequency for two samples

Fig. 9. Ts versus SCB level for samples which are very slightly long-chain
defined respectively, in Tables 2 and 3 (mPE11: full symbols and mPE3:
empty symbols). The two samples have very similar molecular weight, branched. The line correspond to Eq. (3), obtained for samples completely
polydispersity and SCB level, but sample mPE11 possesses a very low free of LCB (see Fig. 7). The samples can be identified by their SCB level
amount of LCB. given also in Table 3.

defined for these polymers at the minimum in the pressure – properties are particularly remarkable at low frequencies and
temperature trace, are T ¼ 150 to 152 8C for linear PE s and shear rates [20], but classical methods such as size exclusion
T ¼ 145 8C for the short branched polymer. These data fit chromatography (SEC) and NMR are not useful to evaluate
well with our Ts values, as can be seen in Tables 1 and 2. On this low level of LCB [21,22]. Therefore, in the case of
the other hand, the stable island, where the extrudates metallocene-catalyzed PEs, one is often obliged to refer to
become smooth, in the slip – stick region of a LLDPE of samples ‘suspected of possessing small amounts of LCB’.
Mw ¼ 114; 000; lies between T ¼ 139 and 146 8C. These This is shown in Fig. 8, where two samples of very similar Mw ;
critical temperatures are lower than those reported by Keller polydispersity index and SCB, give, however, completely
et al., which is consistent with the high SCB level of the different dynamic viscoelastic results. The extremely high
LLDPEs. The result shown by Pudjijanto and Denn is viscosities at low frequencies, observed for one of the samples,
compatible with the effect of the SCB on Ts ; shown in Fig. 7. denotes that repetition is delayed [23]. This is considered a
This sensitivity of the transition temperature to changes in symptom of the presence of a very small amount of LCB. In
the level of SCB, also explains the smooth extrudability Table 3 three samples of unknown LCB level are presented,
found by Pérez González et al. for a metallocene LLDPE at together with two samples of low LCB, as determined by SEC.
a very low temperature, T ¼ 116 8C: This result has been A copolymer of ethylene and propylene produced by
discussed earlier, together with the results of Fig. 6. Brookhart’s type single-site catalysts, with a dynamic visco-
In comparison with Ziegler–Natta catalysts, single-site elastic response reflecting extremely long relaxation times
metallocenes produce a better control of molecular architec- [24], is also included in Table 3. In spite of the complex
ture, leading to narrower molecular weight and short chain architecture attributable to these samples, smooth extrudates
branch distribution. However, the production of small can be obtained by decreasing the temperature to the Ts values
amounts of LCB (0.1 to 1 long branch per molecule) during indicated in Table 3. This is particularly relevant in the case of
the metallocene-catalyzed polymerization process is difficult Brookhart’s type copolymer, which has been shown to exhibit
to control, because it depends on several factors such as sharkskin and the hydrodynamic phenomenon known as
the polymer and monomer concentrations, temperature, ‘flow-split’ [24], at high temperatures and very low shear rates.
comonomer, hydrogen concentration, etc. [19]. The effects Since these samples posses a well defined degree of SCB, in
that small amounts of LCB produce on the rheological Fig. 9 the Ts results as compared with those obtained with

Table 3
Ts data of slightly long chain branched samples. Different molecular weights and SCB are considered. Sample BPE is a copolymer of ethylene and propylene
produced by Brookhart’s type single-site catalysts

Material Mw Mw =Mn SCB LCB Ts

(CH3/1000C) (CH3/1000C) (8C)

a
mPE11 129,900 2.62 9.98 145– 150
a
mPE12 211,866 3.14 4.26 160
BPE 335,000 2.3 7.4 – 150
mPE13 108,000 2.16 7.27 0.137 145
MPE14 92,600 2.24 6.06 0.132 145
a
Samples suspected of containing small amounts of LCB.
2480 A. Santamarı́a et al. / Polymer 44 (2003) 2473–2480

Eq. (2) are shown. Although clear deviations are noticeable, it studied, we claim that decreasing temperature is a sound route
is seen that the influence of SCB on Ts is also followed by these to postpone processing instabilities of any polyethylene.
rather peculiar samples. Small amounts of LCB, which clearly
alter the rheological response of polyethylenes (see Fig. 8),
have practically no effect on the limiting temperature, Ts ; References
to obtain stable flow at low temperatures. Needless to say,
we do not try to make a generalization on the effect of LCB, [1] Cogswell FN. J Non Newtonian Fluid Mech 1977;2:37–47.
since samples with high levels of LCB are not considered. [2] Waddon AJ, Keller A. J Polym Sci: Part B, Polym Phys 1990;28:
However, the results obtained with samples of such different 1063–73.
characteristics can be considered as prove of the ‘universality’ [3] Narh KA, Keller A. Polymer 1991;32:2512 –6.
[4] Waddon AJ, Keller A. J Polym Sci: Part B, Polym Phys 1992;30:
of lowering temperature as a route to produce smooth
923 –30.
extrudates in polyethylenes. [5] Kolnaar JWH, Keller A. Polymer 1994;35:3863–74.
[6] Kolnaar JWH, Keller A. Polymer 1995;36:821–36.
[7] Van Bilsen HMM, Fisher H, Kolnaar JWH, Keller A. Macro-
4. Conclusions molecules 1995;28:8523–7.
[8] Kolnaar JWH, Keller A, Seifert S, Zschunke C, Zachmann HG.
Polymer 1995;36:3969–74.
A large series of single-site metallocene-catalyzed poly- [9] Kolnaar JW, Keller A. J Non-Newtonian Fluid Mech 1996;67:
ethylenes of different molecular architectures, has been 213 –40.
investigated. It is observed that all the samples show a low [10] Pudjijanto S, Denn MM. J Rheol 1994;38:1735 –44.
temperature region, limited by induced crystallization and [11] Wang SQ, Drda PA. Macromolecules 1996;29:4115–9.
gross melt fracture, where smooth extrudates (free of surface [12] Pérez González J, Vargas L, Pavlı́nek V, Hausnerová B, Sáha P.
J Rheol 2000;44:441–51.
defects), are produced at shear rates similar to those of [13] Southern JH, Porter RS. J Appl Polym Sci 1970;14:2305– 17.
industrial extrusion. A characteristic temperature of this [14] Cogswell FN. J Rheol 1994;39:809–10. Letter to the Editor.
region, Ts ; is defined as the highest temperature at which [15] Kim S, Dealy JM. Polym Engng Sci 2002;42:495–503.
sharkskin disappears. The results indicate that good polymer/ [16] Brochard F, de Gennes PG. Langmuir 1992;8:3033 –7.
wall adhesion (non-slip conditions) is associated with this low [17] Ramamurthy AV. J Rheol 1986;30:337 –57.
[18] Martı́nez Salazar J, Sánchez M, Baltá FJ. Colloid Sci 1987;265:
temperature stable region. Therefore, sharkskin and slip–stick
239 –45.
instabilities are avoided, because the shear stress necessary for [19] Hamielec AE, Soares JB. Prog Polym Sci 1996;21:651.
slip is very high, above that involved in melt fracture. The [20] Vega JF, Fernández M, Santamarı́a A, Muñoz-Escalona A, Lafuente P.
analysis of the effect of molecular weight, SCB level and the Macromol Chem Phys 1999;200:2257–68.
presence of small amounts of LCB, leads to the conclusion that [21] Nakajima N. Appl Rheol 1999;May/June:116– 125.
[22] Janzen J, Colby RH. J Mol Struct 1999;569:485 –6.
only SCB has a significant effect on Ts : A linear correlation
[23] Doi M, Edwards SF. The theory of polymer dynamics. Oxford:
between Ts and SCB level is established, showing the decrease Clarendon Press; 1986.
of the former as the latter is increased. Considering the wide [24] Fernández M, Santamarı́a A, Muñoz-Escalona A, Méndez L. J Rheol
spectrum of the molecular characteristics of the samples 2001;45:595–602.