eXPRESS Polymer Letters Vol.3, No.

10 (2009) 639–649
Available online at www.expresspolymlett.com
DOI: 10.3144/expresspolymlett.2009.80

Comparison of the influence of Cu micro- and

nano-particles on the thermal properties of
polyethylene/Cu composites
J. A. Molefi1, A. S. Luyt1*, I. Krupa2

1Department of Chemistry, University of the Free State (Qwaqwa Campus), Private Bag X13, Phuthaditjhaba, 9866,
South Africa
2Polymer Institute, Slovak Academy of Sciences, 842 36 Bratislava, Slovak Republic

Received 19 May 2009; accepted in revised form 8 August 2009

Abstract. Polyethylene (LDPE, LLDPE and HDPE) composites with different copper (micro- and nano-sized particles)
contents were prepared by melt mixing and compression moulding. The melting and crystallization behaviour of the differ-
ent composites was analysed using a differential scanning calorimeter (DSC), and the thermal stability in a thermogravi-
metric analyser (TGA). The thermal conductivities of the samples were also determined. The DSC results show that the Cu
micro- and nano-particles influence the crystallization behaviour of the polyethylenes in different ways. The extent to
which the copper particles influence the crystallization behaviour of the polyethylenes also depends on the respective mor-
phologies of the different polyethylenes. The TGA results show an observable influence of both the presence of copper and
the sizes of the copper particles on the thermal stabilities of the polymers. Thermal conductivities increased with increasing
Cu content, but there was little difference between the thermal conductivities of the samples containing Cu micro- and
nano-particles.

Keywords: polymer composites, polyethylene/Cu, crystallization behaviour, thermal stability

1. Introduction [1–3]. One study compared composites that con-

This paper forms part of a comprehensive study on tained respectively micro- and nano-sized particles
the use of copper micro- and nano-particles to [3]. The authors showed that incorporation of
improve thermal conductivity in phase change micro- or nano-sized Cu particles into LDPE
materials for thermal energy storage based on poly- reduced the melting temperature, increased the
ethylene/wax blends. This paper reports on the crystallization temperature, and lowered the degree
influence of Cu micro- and nano-particles on the of crystallinity of the matrix of the composites.
thermal properties of polyethylene/Cu composites Similar effects were observed for the melting peak
for three different types of polyethylene, and temperatures, melting enthalpies and crystallinities
changes in these properties are explained in terms of LDPE and LLDPE filled with copper particles
of morphology differences between the different [1]. It was found that the degree of crystallinity of
polyethylenes. the matrix of the nanocomposites was lower than
The thermal properties of polyethylenes filled with that of the microcomposites [3]. In another study
copper particles were investigated by a few groups two different copper particle sizes (both in the

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

© BME-PT

639
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

micron range) were used as fillers in a PP matrix improve the thermal conductivity of, for example,
[2]. This study showed an insignificant change in HDPE. The study was based on HDPE filled with
matrix crystallinity. The influence of graphite, silver-coated polyamide (PA). It was found that
another possible filler to improve the thermal con- very low silver content, if deposited on the PA par-
ductivity of phase change materials, on the change ticles, significantly improved the thermal conduc-
in the degree of crystallinity of HDPE was studied tivity of the composites. Boudenne et al. [2] inves-
by Krupa et al. [4]. A slight decrease in the melting tigated the thermal conductivity of polypropylene
temperature of the composites was observed, which filled with two different sizes (23 and 230 μm) of
suggested that the filler reduced the lamellar thick- copper particles. A non-linear increase in the ther-
ness of the crystallites. The same behaviour was mal conductivity was observed with an increase in
observed in another study [5], where both HDPE the volume fraction of both fillers. At a given filler
and PS were filled with 60 wt% of graphite. Silver- concentration, a higher thermal conductivity was
coated polyamide (PA6) particles were also used as observed in the composite filled with the smaller
filler for HDPE [6]. In this case a slight decrease in Cu particles. This difference was more significant
the degree of crystallinity with an increase in silver- at higher filler concentrations.
coated PA6 particle content was observed.
Generally, the literature focused on the thermal
2. Experimental
conductivities of metal containing polymers. How-
ever, a few papers are devoted to the thermal stabil- In this work, LDPE and LLDPE were supplied in
ity of polyethylene filled with metals [1, 3, 7]. pellet form by Sasol Polymers. LDPE has an MFI
When LDPE/Cu nano- and micro-composites were of 7.0 g/10 min (ASTM D-1238), a melting point
compared [3], the thermal decomposition tempera- of 106°C, a molecular weight of 96 000 g·mol–1,
tures of the nano-composites increased sharply at and a density of 0.918 g·cm–3 and LLDPE has an
low Cu nano-particle contents, and reached an opti- MFI of 1.0 g/10 min (ASTM D-1238), a molecular
mum value at 2 wt% Cu, after which it decreased weight of 191 600 g·mol–1, a melting point of
with further increase in nano-Cu content. In the 124°C, and a density of 0.924 g·cm–3. HDPE was
case of the LDPE/Cu micro-composites, the opti- supplied in pellet form by DOW Chemicals. It has
mum thermal stability was reached only after addi- an MFI of 8 g/10 min (ASTM D-1238), a molecu-
tion of 6 wt% Cu, but higher Cu contents did not lar weight of 168 000 g·mol–1, a melting point of
further reduce the thermal stability. A study by 130°C, and a density of 0.954 g·cm–3. Soft paraffin
Luyt et al. [1] showed that the thermal stability of wax (M3 wax) was supplied in powder form by
LDPE/Cu micro-composites increased, whereas the Sasol Wax. It has an average molar mass of
thermal stability of LLDPE/Cu micro-composites 440 g·mol–1 and a carbon distribution between C15
decreased, with increasing Cu content. Yang et al. and C78. Its density is 0.90 g·cm–3 and it has a
[8] studied the thermal stability of aligned vapour melting point range around 40–60°C. Merck
grown carbon nano-fibre reinforced polyethylene Chemicals in South Africa supplied the copper
samples drawn at different take-up speeds, and all powder, which was used as one of the conducting
these composites showed higher thermal stability fillers. It has a melting point of 1083°C and a den-
than pure HDPE. sity of 8.96 g·cm–3, and the particles sizes were less
Polymers are materials with very low thermal than 38 μm determined by using a laboratory test
conductivity values that roughly vary from sieve with a pore sizes of 38 μm. The copper nano-
0.15 W·m–1·K–1 for amorphous polymers to particles were supplied by Lawrence Packaging
0.5 W·m–1·K–1 for highly crystalline polymers such Supply Corp., Moonachie, New Jersey, USA, Lot #
as HDPE [9]. Several inorganic materials, graphite R402 and the particle size was 50 nm.
or metallic powders, are frequently used as ther- All the samples were prepared by mixing the com-
mally conductive fillers [10, 11]. Silver, copper and ponents in a Brabender Plastograph 50 ml internal
alumina particles are the most commonly used mixer at 160°C and a speed of 70 rpm for 15 min.
fillers due to their high inherent thermal conductiv- After the mixing, the samples were melt pressed at
ity [5, 10–12]. Krupa et al. [13] demonstrated that 100 bar and 160°C for 15 min.
metal-coated particles can be effectively used to

640
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

DSC analyses were done in a Perkin-Elmer Pyris-1 3. Results and discussion

differential scanning calorimeter under flowing The morphologies of the composites were studied
nitrogen (flow rate 20 ml·min–1). The instrument by SEM analysis of the fracture surfaces of the
was computer controlled and calculations were
done using Pyris software. It was calibrated using
the onset temperatures of melting of indium and
zinc standards, as well as the melting enthalpy of
indium. Samples (5–10 mg) were sealed in alu-
minium pans and heated from –40 to 160°C at a
heating rate of 10°C min–1, and cooled at the same
rate. For the second scan, the samples were heated
and cooled under the same conditions. The peak
temperatures of melting and crystallization, as well
as melting and crystallization enthalpies were
determined from the second scan. All DSC meas-
urements were repeated three times for each sam-
ple. The melting and crystallization temperatures,
as well as enthalpies, are reported as average values
with standard deviations.
TGA analysis was carried out on a Perkin-Elmer
TGA7 thermogravimetric analyser. Samples rang-
ing between 5 and 10 mg were heated from 30 to
650°C at a heating rate of 20°C·min–1 under flow-
ing nitrogen (flow rate 20 ml·min–1). The instru-
ment was computer controlled and calculations
were done using Pyris software.
A Shimadzu model ZU SSX-550 Superscan scan-
ning electron microscope was used for the SEM
analyses. All the samples were frozen in liquid
nitrogen, then fractured by simply breaking the
specimens in an appropriate size to fit in the speci-
men chamber, and then mounted onto the holder.
The surfaces of the samples were coated with gold
by an electro-deposition method to impart electrical
conduction before recording the SEM micrographs.
This was done to prevent the accumulation of static
electric charge on the specimen during electron
irradiation.
The thermal conductivity was measured using a
multi-purpose apparatus ISOMET, (Applied Preci-
sion, Bratislava, Slovakia) for non-steady measure-
ments of thermal properties. The thermal conduc-
tivity values were calculated automatically from a
time dependence of the thermal flow in the mate-
rial. Samples were melt-pressed into spherical
(7 cm diameter) 1 mm thick sheets in a hot-melt
press at 160°C for 5 min at a pressure of 100 bar. Figure 1. SEM images of (a) a 90/10 v/v LDPE/Cu micro-
Thereafter, measurements were made at 25±2°C composite (b) a 99/1 v/v LDPE/Cu nano-com-
with a flat probe. posite, and (c) a 95/5 v/v LDPE/Cu nano-com-
posite

641
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

composites. Figure 1a shows a fair dispersion of

micro-particles, with no obvious agglomeration. As
can be seen in Figures 1b and 1c, the Cu nano-par-
ticles are fairly well dispersed, but also form
agglomerates. To optimally improve the conductiv-
ity of the composites, it is important for the parti-
cles to form a percolation network within the
polymer matrices. No such network formation was
observed in the SEM images of the investigated
samples. In Figure 1a there seems to be voids
around the Cu particles, which indicate that the
interaction between the polymer and the Cu parti-
cles was weak. The lack of adhesion between the
polymer and filler indicates poor interfacial interac- Figure 2. DSC heating curves of the pure polyethylenes
tion. As the content of Cu particles increased, the
agglomeration of the particles also increased. in line with the known crystallinities of the respec-
Figure 2 shows the DSC heating curves for pure tive polyethylenes.
LDPE, LLDPE, and HDPE. The curves show Table 1 shows the DSC results of the different
endotherms with melting peak temperatures at 107, polyethylenes and their Cu micro-composites. For
128 and 135°C for LDPE, LLDPE and HDPE. Pure the LDPE/Cu micro-composites the peak tempera-
HDPE has a relative high enthalpy value of tures of melting did not significantly change and
149.3 J·g–1, while LLDPE and LDPE have are within experimental error. From this it is clear
enthalpies of 86.9 and 75.4 J·g–1. These values are that the presence of up to 20 vol% of Cu micro-par-

Table 1. DSC results for polyethylene/Cu micro-composites

v/v Tp,m [°C] ΔHmobs [J·g–1] ΔHmcalc [J·g–1] Tp,c [°C] ΔHcobs [J·g–1]
LDPE/Cu
100/0 106.8 ± 1.5 75.4 ± 6.2 – 091.7 ± 0.6 –63.7 ± 7.6
99/1 107.1 ± 1.4 061.5 ± 6.3 68.8 ± 5.6 091.5 ± 1.0 –57.0 ± 6.9
97/3 106.6 ± 1.7 51.3 ± 8.3 58.1 ± 4.7 091.4 ± 0.7 –41.2 ± 4.1
95/5 106.3 ± 1.2 39.9 ± 2.6 49.8 ± 4.1 091.8 ± 1.1 –43.5 ± 4.9
90/10 106.4 ± 1.1 23.4 ± 1.8 36.2 ± 6.4 091.8 ± 0.9 –27.6 ± 3.2
85/15 105.8 ± 0.8 18.6 ± 1.6 27.8 ± 4.6 092.8 ± 0.9 –21.3 ± 3.9
80/20 106.0 ± 0.9 14.4 ± 0.4 21.9 ± 3.4 092.2 ± 0.8 –16.0 ± 3.1
75/25 105.5 ± 1.2 10.4 ± 2.7 17.9 ± 1.5 092.9 ± 1.1 –12.4 ± 2.2
LLDPE/Cu
100/0 127.7 ± 2.1 86.9 ± 1.0 – 109.6 ± 0.9 –50.1 ± 0.8
99/1 127.9 ± 1.2 69.4 ± 4.5 79.3 ± 0.9 110.2 ± 1.0 –45.2 ± 2.0
97/3 126.8 ± 1.3 58.8 ± 4.5 66.6 ± 0.8 110.0 ± 1.3 –37.7 ± 4.1
95/5 126.2 ± 0.5 45.3 ± 9.7 57.4 ± 0.6 110.3 ± 1.3 –32.7 ± 2.5
90/10 126.5 ± 1.3 30.9 ± 7.0 41.7 ± 0.5 110.9 ± 1.7 –22.2 ± 3.4
85/15 125.4 ± 1.4 19.6 ± 4.3 31.9 ± 0.4 111.6 ± 1.2 –17.1 ± 4.5
80/20 125.5 ± 1.4 15.2 ± 3.4 25.3 ± 0.3 111.0 ± 1.3 –14.3 ± 3.1
75/25 125.4 ± 1.6 10.2 ± 2.5 20.6 ± 0.3 112.5 ± 1.3 –10.4 ± 4.2
HDPE/Cu
100/0 134.7 ± 0.5 149.3 ± 11.7 – 113.9 ± 1.1 –134.8 ± 2.6
99/1 134.1 ± 1.3 138.1 ± 3.1 136.2 ± 10.6 114.3 ± 0.1 –132.1 ± 2.1
97/3 133.6 ± 0.1 110.4 ± 7.3 115.0 ± 9.0 115.0 ± 0.9 –110.4 ± 0.4
95/5 132.6 ± 0.3 101.8 ± 0.8 098.5 ± 7.6 114.7 ± 0.6 –97.0 ± 2.3
90/10 131.6 ± 0.2 067.4 ± 6.7 071.7 ± 5.6 115.3 ± 0.3 –72.0 ± 1.6
85/15 131.4 ± 0.2 052.3 ± 8.0 054.8 ± 4.3 115.7 ± 0.2 –53.3 ± 0.3
80/20 130.8 ± 0.5 041.4 ± 1.6 043.4 ± 3.1 116.0 ± 0.4 –43.9 ± 0.3
75/25 130.2 ± 0.2 027.2 ± 9.8 035.2 ± 2.8 116.3 ± 0.6 –34.3 ± 0.4
Tp,m, ΔHmobs, ΔHmcalc, Tp,c, ΔHcobs are melting peak temperature, observed melting enthalpy, calculated melting enthalpy, crystallization
peak temperature, crystallization enthalpy

642
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

ticles in LDPE had little influence on the crystallite in the presence of Cu micro-particles. Since HDPE
sizes. However, if the enthalpy values in Table 1 is much more crystalline, the amorphous regions
are compared, it is clear that the Cu micro-particles are much smaller and it is possible that crystalliza-
reduce the LDPE chain mobility which gives rise to tion will start at the Cu surfaces and that Cu parti-
a lower crystallinity. When the measured melting cles will act as nucleation points for HDPE crystal-
enthalpy values (ΔHmobs) are compared with the cal- lization. This explains the slight lowering in HDPE
culated values (ΔHmcalc), it can be seen that the melting temperature (indicating the presence of
LDPE in the composites had significantly lower smaller crystallites). This is consistent with a study
enthalpy (or crystallinity) values than would be done by Xia et al. [3]. When they added 13 wt% of
expected if it was assumed that the Cu micro-parti- Cu micro-particles into an LDPE matrix, the melt-
cles had no influence on the LDPE crystallization ing peak temperatures of pure LDPE and the
behaviour. The ΔHmcalc values were calculated from LDPE/Cu micro-composites were the same within
the melting enthalpy of pure LDPE and the weight experimental error. Pure LDPE had a melting
fractions of LDPE in the respective composites. enthalpy of 88 J·g–1 which reduced to 76 J·g–1 when
A slight decrease in the melting temperature of 13 wt% of Cu micro-particles was introduced, and
LLDPE with an increase in Cu content is observed the crystallization temperature increased for the
for the LLDPE/Cu micro-composites (Table 1). It same composition. The increased crystallization
seems as if the presence of Cu slightly reduces the temperature was explained as being due to the pres-
LLDPE chain mobility which gives rise to smaller ence of the Cu micro-particles, which acted as
crystallites and a lower crystallinity. In the case of nucleating agents. However, the Cu particles at the
the HDPE/Cu micro-composites, a more observ- same time seemed to have immobilized the LDPE
able decrease in the melting temperature of the chains, giving rise to reduced crystallinity, as can
composites with an increase in Cu content is be seen from the smaller melting enthalpy values.
observed. It seems as if the influence of Cu micro- Table 2 shows the DSC results of the different
particles on the crystallite size of polyethylene polyethylenes and their Cu nano-composites. For
becomes more significant with decreasing branch- the LDPE/Cu nano-composites the peak tempera-
ing and increasing crystallinity of the polymer. tures of melting did not significantly change and
Another reason may be that the presence of highly are within experimental error. From this it is clear
thermally conductive Cu causes a faster heat distri- that the presence of up to 5 vol% of Cu nano-parti-
bution through the polymer, which will result in a cles in LDPE had a negligible influence on the
lower melting temperature. Table 1 shows that the crystallite sizes. On the other hand, if the enthalpy
ΔHmobs and ΔHmcalc values are almost the same for values in Table 2 are compared, it is clear that the
all the HDPE/Cu micro-composites, and therefore presence of the Cu nano-particles restricts the
the presence and amount of Cu had very little influ- LDPE chain mobility which leads to a lower crys-
ence on the HDPE crystallization behaviour. tallinity. When the observed melting enthalpy val-
The observations described in the previous para- ues (ΔHmobs) are compared with the calculated val-
graphs may be explained if it is accepted that the ues (ΔHmcalc), it can be seen that the LDPE in the
Cu particles will either reduce polymer chain composites had significantly lower enthalpy (or
mobility (giving rise to reduced crystallinity) or act crystallinity) values than would be expected if it
as nucleation points for polymer crystallization was assumed that the Cu nano-particles had no
(giving rise to increased crystallinity). It seems as if influence on the LDPE crystallization behaviour.
the chain structure (and the resultant degree of crys- Similarly, the copper nano-particles in the LLDPE/
tallinity) of the polymer determines which of the Cu and HDPE/Cu nano-composites had no signifi-
two effects will be dominant. Since LDPE and cant influence on the melting peak temperatures of
LLDPE are branched, and therefore have signifi- the respective polyethylenes. The observed melting
cantly lower crystallinities than HDPE, which con- enthalpies of these polyethylene composites were
firms that the Cu particles are located in the amor- also lower than the calculated values. However, the
phous regions of the polymers. Chain immobiliza- differences between the observed and calculated
tion will then be the more dominant effect, and values seem to decrease with increasing crys-
therefore these polymers show reduced crystallinity tallinity from LDPE to LLDPE to HDPE. Since

643
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

Table 2. DSC results for polyethylene/Cu nano-composites

v/v Tp,m [°C] ΔHmobs [J·g–1] ΔHmcalc [J·g–1] Tp,c [°C] ΔHcobs [J·g–1]
LDPE/Cu
100/0 106.8 ± 1.5 75.4 ± 6.2 – 91.7 ± 0.6 –63.7 ± 7.6
99/1 105.9 ± 0.3 49.4 ± 2.3 68.8 ± 5.6 90.8 ± 0.1 –54.3 ± 1.9
97/3 105.7 ± 0.1 38.8 ± 0.9 58.1 ± 4.7 91.7 ± 0.3 –48.3 ± 0.8
95/5 105.3 ± 0.1 35.5 ± 1.9 49.8 ± 4.1 91.9 ± 0.1 –41.5 ± 0.6
LLDPE/Cu
100/0 126.7 ± 2.1 86.9 ± 1.0 – 109.6 ± 0.9 –50.1 ± 0.8
99/1 126.9 ± 0.4 62.6 ± 1.4 79.3 ± 0.9 109.3 ± 0.1 –47.0 ± 4.0
97/3 125.6 ± 0.2 56.6 ± 2.3 66.6 ± 0.8 110.7 ± 0.1 –38.3 ± 2.1
95/5 126.0 ± 0.1 43.9 ± 0.7 57.4 ± 0.6 111.7 ± 0.1 –31.0 ± 2.7
HDPE/Cu
100/0 134.7 ± 0.5 149.3 ± 9.7 – 113.9 ± 1.1 –134.8 ± 2.6
99/1 133.2 ± 0.2 132.1 ± 2.5 136.2 ± 10.6 113.8 ± 0.2 –126.2 ± 2.2
97/3 133.1 ± 0.1 106.8 ± 1.5 115.0 ± 9.0 113.9 ± 0.3 –109.5 ± 3.2
95/5 134.4 ± 1.0 82.9 ± 0.8 098.5 ± 7.6 114.7 ± 0.3 –90.0 ± 1.6
Tp,m, ΔHmobs, ΔHmcalc, ΔHcobs Tp,c, are melting peak temperature, observed melting enthalpy, calculated melting enthalpy, and crystalliza-
tion peak temperature

LDPE and LLDPE are branched, and therefore ites, indicating that Cu nano-particles had a
have lower crystallinities than HDPE, the Cu parti- stronger influence on the motion and nucleation of
cles are situated in the amorphous regions of these the polymer chain segments. This behaviour indi-
polymers. Chain immobilization will then be the cates that the Cu nano-particles restricted the poly-
more dominant effect. Since HDPE is much more mer chain mobility to a larger extent than the Cu
crystalline, the amorphous regions are much micro-particles, probably because the Cu micro-
smaller and therefore the decrease in total crys- particles counteract this effect through their ability
tallinity is less significant. to act as nucleating agents that promote the crystal-
Figure 3 shows the experimental melting enthalpies lization process of LDPE.
as function of Cu content for LDPE/Cu nano- and It seems as if the influence of Cu micro-particles on
micro-composites. It can be seen that the experi- the crystallite size of polyethylene becomes more
mental melting enthalpies of the LDPE/Cu micro- significant with decreasing branching and increas-
particles are higher than those the of LDPE/Cu ing crystallinity of the polymer. Tracz et al. [14]
nano-composites at the same volume percentages found that the nucleating and ordering effects of
of copper. This clearly shows that the crystallinities fillers on polyethylene crystallization depend to a
of the LDPE/Cu nano-composites are observably larger extent on the nano-structure of its surface.
lower than those of the LDPE/Cu micro-compos- Nano-particles do not have any nucleating effect,
but they act as obstacles for the crystallization of
polyethylene. Even though some nano-particle
agglomerates may nucleate PE crystallization, the
overall influence of the nano-particles on the crys-
tallization of LDPE can be considered to be a retar-
dant effect. Therefore, the degree of crystallinity of
LDPE in the nano-composites is lower than that of
pure LDPE. Micro-particles can nucleate PE crys-
tallization, because they can provide sufficiently
large flat domains. However, the sizes of the micro-
particles are comparable to the spherulite size of
LDPE, and thus the micro-particles also hinder
spherulitic growth, so that the degree of crys-
tallinity of LDPE in the presence of micro-particles
Figure 3. Comparison of ΔHobs values for the melting of is not significantly higher than that of the pure
LDPE in LDPE/Cu micro- and nano-composites LDPE. Huang et al. [15] found that the addition of

644
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

2 and 4% of Al nano-particles can either facilitate

or hinder the crystallization of LDPE, depending on
the dispersion of the nano-particles in the polymer.
The well-dispersed Al nano-particles did not have a
nucleating effect and mainly acted as obstacles, but
the agglomerates of Al nano-particles acted as
nucleating agents and slightly improved the crystal-
lization. The micro-particles had a nucleating effect
and facilitated the crystallization process. These
observations and explanations are in line with the
results of our study, although in our case the pres-
ence of Cu micro-particles was found to reduce the
total LDPE crystallinity, but to a smaller extent
than the Cu nano-particles. In LDPE/Al micro- and Figure 5. Comparison of ΔHobs values for the melting of
nano-composites and in LDPE/TiO2 nano-compos- HDPE in HDPE/Cu micro- and nano-composites
ites the melting behaviour indicated that the lamel-
lar thickness distribution of both nano- and tion is that the difference between the ΔH values of
micro-composites did not significantly change the respective micro- and nano-composites seems
compared to that of pure LDPE [14, 15]. to decrease with increasing filler content in the case
The measured enthalpy values of the LLDPE/Cu of LDPE and LLDPE, while it seems to increase in
and HDPE/Cu nano-composites are also lower than the case of HDPE. It is difficult to give an exact
those of the micro-composites, but the differences explanation for this observation, because a number
are less significant (Figures 4 and 5). The reason of factors may determine the influence of inorganic
for this is probably the higher crystallinities and particles on the crystallization behaviour of poly-
lower extent of branching in these polymers, that ethylene. Amongst these are the sizes and size dis-
result in a smaller influence of these particles on the tributions of the filler particles, the dispersion of
polymer chain mobility. The uncertainties in the the particles in the polymer matrix, the effective
ΔH values are relatively high in the case of the surface areas of the particles, the crystallinities and
microcomposites. The reason for this is probably a crystallization mechanisms of the respective poly-
bigger size distribution of the micro-particles com- mers, and the extent of interaction between the
pared to the nano-particles. Unfortunately we did polymer and the filler particles. Depending on these
not statistically determine the size distribution of factors, the filler particles will immobilize the poly-
the micro-particles. Another interesting observa- mer chains (reducing crystallinity) and/or act as
nucleation centres (increasing crystallinity).
Although high-temperature thermal stability of
these composites does not directly influence their
applications, it is interesting to see how the thermal
stabilities of the polymers are influenced by the
presence of copper filler, and whether these obser-
vations can be explained by the same morphologi-
cal features used to explain the influence of copper
particles on the crystallization behaviour of the PE
matrices. TGA studies were also used to establish
the dispersion efficiency of the copper particles in
the matrix. The TGA data of the PE/Cu micro-com-
posites are summarized in Table 3. There is a non-
linear increase in thermal stability with an increase
Figure 4. Comparison of ΔHobs valuesfor the melting of in Cu micro-particle content up to 20 vol%. This is
LLDPE in LLDPE/Cu micro- and nano-compos- probably due to the inability of the polymer chains
ites
to move freely because of the presence of Cu

645
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

Table 3. Temperatures of 10% degradation and DTG peak the LLDPE crystals (especially at higher Cu con-
temperatures, as well as wt% of residue, of PE/Cu tents). The higher heat capacity and thermal con-
micro-composites ductivity of copper then causes more effective heat
v/v
T10 Tpeak wt% wt% transfer to the LLDPE crystals. In the case of
[°C] [°C] residue Cu in sample HDPE there is no clear trend for the influence of Cu
LDPE/Cu
100/0 443.3 485.5 – –
content on its thermal stability and it seems that,
99/1 449.7 487.6 07.3 08.8 within experimental error, the Cu micro-particles
95/5 454.3 489.9 32.6 33.0 did not significantly influence its thermal stability.
90/10 459.7 485.2 51.7 52.0 This is probably the result of a balance between
80/20 461.2 480.2 69.9 69.0
LLDPE/Cu
thermal stability enhancement and thermal degra-
100/0 454.1 492.9 – – dation initiation effects by the copper particles. The
99/1 468.5 501.8 09.7 08.9 higher thermal conductivity of the metallic Cu par-
95/5 461.4 491.7 34.0 33.1 ticles may override its stiffening effect. Since
90/10 470.4 488.9 52.0 52.2
80/20 470.4 485.4 70.3 69.5
HDPE is highly crystalline and inherently less ther-
HDPE/Cu mally conductive, the transfer of heat through the
100/0 467.0 500.1 – – entire sample is delayed. When the Cu content is
99/1 463.6 496.3 08.8 08.5 above 10 vol%, there may be a strong interconnec-
95/5 469.2 497.3 32.3 33.0
90/10 455.2 485.7 50.4 51.0
tion between the Cu particles, resulting in increased
80/20 458.4 471.4 69.4 69.7 thermal stability. This may also be attributed to the
hindered polymer chain mobility enforced by the
Cu micro-particles.
micro-particles, which will retard the movement of The TGA results of the PE/Cu nano-composites are
free radical chains and inhibit the degradation summarized in Table 4. The thermal stability of
process. Another reason may be that the diffusion LDPE and LLDPE generally increases with increas-
of the volatile degradation products may be ing Cu nano-particle content, but for LDPE a
retarded by the presence of the copper particles, decrease is observed for the highest Cu content of
which will lead to these products only coming off at 5 vol%. This decrease in thermal stability is proba-
higher temperatures. In other studies [3, 16] it was bly due to the presence of Cu nano-particle agglom-
also found that an increase in conductive filler con- erates that cause an increase in energy propagation.
tent enhances the thermal stability of LDPE. The formation of these agglomerates was observed
Krump et al. [16] found that the thermal stability of in the SEM results (Figure 1). The delayed thermal
LDPE/Cu micro-composites increased with an degradation for both LDPE/Cu and LLDPE/Cu
increase in Cu micro-particle content. Xia et al. [3]
showed that the optimum thermal stability was Table 4. Temperatures of 10% degradation and DTG peak
reached after addition of 6 wt% Cu, but at higher temperatures, as well as wt% of residue, of PE/Cu
Cu contents the thermal stability remained con- nano-composites
stant. Omastová et al. [17] showed that carbon T10 Tpeak wt% wt%
v/v
black (CB) and modified CB in an LDPE matrix [°C] [°C] residue Cu in sample
improved the thermal stability of the filled compos- LDPE/Cu
100/0 443.3 485.5 – –
ites, with modified CB having a more significant 99/1 456.0 480.7 08.4 08.8
effect. 97/3 458.0 491.4 22.0 23.1
The thermal stability of LLDPE improves with an 95/5 440.2 480.9 33.4 33.0
increase in Cu micro-particles up to 1–3 vol% Cu, LLDPE/Cu
100/0 454.1 492.9 – –
and the temperature at maximum decomposition 99/1 441.8 491.7 10.5 08.9
rate increases for 1 vol% Cu, after which it 97/3 464.6 495.9 22.5 23.0
decreases (Table 3). This is in line with the obser- 95/5 468.1 495.6 33.4 33.1
vations by Xia et al. [3], but differs from the obser- HDPE/Cu
100/0 467.0 500.1 – –
vations on the LDPE/Cu composites, probably 99/1 409.7 460.0 09.7 08.9
because of the higher crystallinity of LLDPE which 97/3 437.3 484.1 26.3 23.0
results in more intimate contact between the Cu and 95/5 460.9 486.3 31.2 33.0

646
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

nano-composites is due to the restricted chain and where λm, λc, λa are the thermal conductivities of
volatile degradation products mobility imposed by the polymer and its crystalline and amorphous
the Cu nano-particles. In all of these polyethylenes phases, respectively, and wc, wa are the weight frac-
filled with Cu particles, the shift towards higher tions of the crystalline and amorphous phases of the
temperature is in the order of their polymer chain polymer.
regularity. There is an increase in residue with Generally, most inorganic fillers have much higher
increasing Cu content for all the investigated sam- thermal conductivity than polymers, and therefore
ples, and these values are in line with the initial their incorporation into the polymers leads to an
wt% Cu mixed into the samples (Tables 3 and 4). increase in thermal conductivity of the composites
This indicates that the presence of Cu particles did [5, 19]. Unfortunately, the prediction of thermal
not cause any char formation. conductivity of filled polymers is very difficult and
The thermal conductivities of the PEs (LDPE, depends on the geometry and orientation of the
LLDPE and HDPE) filled with copper micro-parti- filler particles in the matrix, the concentration of
cles are presented in Figure 6. A non-linear the filler, and the ratio between the thermal conduc-
increase in thermal conductivities is observed with tivity of the filler and the thermal conductivity of
an increase in copper content. This is foreseeable, the matrix. Based on these factors, many different
because the filler has a significantly higher thermal models have already been developed, but none of
conductivity than the polymeric matrices. It can be them has general validity, since most of the models
seen that thermal conductivities of the HDPE/Cu are derived for regularly shaped particles, flakes or
micro-composites are higher than those of both the fibres that have a uniform size distribution. In this
LLDPE/Cu and LDPE/Cu micro-composites. This study, the copper particles were not regular.
is due to the higher degree of crystallinity of the The presence of copper nano-particles in the poly-
HDPE matrix. In the case of semi-crystalline poly- mer matrices also increases the thermal conductiv-
mers, an increase in thermal conductivity with an ity of the nano-composites over that of the pure
increase in crystalline content was observed as a polymers (Figure 7). The copper nano-particles
consequence of the better transport of heat in the also improve the thermal conductivity of the com-
crystalline phase [4]. The simple relation between posites. HDPE filled with nano-particles shows the
thermal conductivity of semi-crystalline polymers highest thermal conductivity, similar to the micro-
and the weight fraction of the crystalline phase is composites. The thermal conductivities of the
expressed in terms of the additive rule [18], given nano-composites were in the same as those of the
by Equation (1): micro-composites, for the same volume fractions of
λ m = λ c wc + λ a w a copper filler.
(1)

Figure 6. Thermal conductivity of the polyethylene com- Figure 7. Thermal conductivity of the polyethylene com-
posites as function of Cu micro-particle content posites as function of Cu nano-particle content

647
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

4. Conclusions References
This paper discussed the morphology and thermal [1] Luyt A. S., Molefi J. A., Krump H.: Thermal, mechan-
ical and electrical properties of copper powder filled
properties of the composites of three different poly-
low-density and linear low-density polyethylene com-
ethylenes (LDPE, LLDPE and HDPE) with Cu posites. Polymer Degradation and Stability, 91, 1629–
micro- and nano-particles. No previous paper 1636 (2006).
reports such a comparative investigation. The Cu DOI: 10.1016/j.polymdegradstab.2005.09.014
particles in the PE/Cu micro-composites were [2] Boudenne A., Ibos I., Fois M., Majestè J. C., Gehn E.:
Electrical and thermal behaviour of polypropylene
fairly well dispersed in the polymer matrix, but as
filled with copper particles. Composites Part A:
the filler content increased, voids around the Cu Applied Science and Manufacturing, 36, 1545–1554
particles were observed indicating poor interaction (2005).
between the polymer and the Cu micro-particles. DOI: 10.1016/j.compositesa.2005.02.005
The lack of adhesion between the polymer and [3] Xia X., Cai S., Xie C.: Preparation, structure and ther-
filler indicates poor interfacial interaction. The Cu mal stability of Cu/LDPE nanocomposites. Materials
Chemistry and Physics, 95, 122–129 (2006).
nano-particles were well dispersed in the polymer
DOI: 10.1016/j.matchemphys.2005.05.010
matrix, except that they tended to form agglomer- [4] Krupa I., Novák I., Chodák I.: Electrically and ther-
ates. Incorporation of both Cu micro- and nano-par- mally conductive polyethylene/graphite composites
ticles into the different polyethylenes reduced the and their mechanical properties. Synthetic Metals,
crystallinities of the different polyethylenes. Gener- 145, 245–252 (2004).
ally the nano-composites had a more significant DOI: 10.1016/j.synthmet.2004.05.007
[5] Krupa I., Chodák I.: Physical properties of thermo-
influence because of the larger total surface area of plastic/graphite composites. European Polymer Jour-
these particles. Even though the nano-particles nal, 37, 2159–2168 (2001).
tended to form agglomerates, their influence was DOI: 10.1016/S0014-3057(01)00115-X
still more significant than that of the micro-parti- [6] Krupa I., Miková G., Novák I., Janigová I., Nógellová
cles. The micro-particles generally had a nucleating Z., Lednický F., Prokeš J.: Electrically conductive
composites of polyethylene filled with polyamide par-
effect and facilitated the crystallization process,
ticles coated with silver. European Polymer Journal,
while the well dispersed nano-particles hindered 43, 2401–2413 (2007).
the crystallization of polyethylene. Since linear DOI: 10.1016/j.eurpolymj.2007.03.033
chain polyethylenes are more difficult to decom- [7] Murty M. V. S., Grulke E. A., Bhattacharyya D.:
pose than the branched ones, the pure LDPE Influence of the metallic on thermal degradation and
degraded at much lower temperatures compared to liquefaction of high density polyethylene (HDPE).
Polymer Degradation and Stability, 61, 421–430
LLDPE and HDPE. Generally the presence of Cu (1998).
particles in the polyethylenes showed an increase in DOI: 10.1016/S0141-3910(97)00228-0
thermal stability with increasing Cu content, indi- [8] Yang S., Taha-Tijerina J., Serrato-Diaz V., Hernandez
cating that the Cu particles either immobilized the K., Lozano K.: Dynamic mechanical and thermal
free radical chains or inhibited the diffusion of the analysis of aligned vapor grown carbon nanofiber
reinforced polyethylene. Composites Part B: Engi-
volatile degradation products. The Cu nano-parti-
neering, 38, 228–235 (2007).
cles had a more significant influence than the DOI: 10.1016/j.compositesb.2006.04.003
micro-particles, and the influence of both types of [9] Bandrup J., Immergut E., Grulke E.: Polymer hand-
particles became less significant with increasing book. Wiley, New York (1999).
polymer crystallinity. Thermal conductivities [10] Sundstrom D. W., Lee Y-D.: Thermal conductivity of
increased with increasing Cu content, but there was polymers filled with particulate solids. Journal of
Applied Polymer Science, 16, 3159–3167 (1972).
little difference between the thermal conductivities
DOI: 10.1002/app.1972.070161210
of the samples containing Cu micro- and nano-par- [11] Bujard P., Munk K., Kuehnlein G.: Thermal conduc-
ticles. tivity. Technomic Publishing, Lancaster, (1994).
[12] Boudenne A., Ibos L., Géhin E., Fois M., Majesté J.
C.: Anomalous behavior of thermal conductivity and
Acknowledgements diffusivity in polymeric materials filled with metallic
The National Research Foundation of South Africa (GUN particles. Journal of Materials Science, 16, 4163–4164
62693) and the University of the Free State are acknowl- (2005).
edged for financial support of this research. DOI: 10.1007/s10853-005-3818-2

648
 Molefi et al. – eXPRESS Polymer Letters Vol.3, No.10 (2009) 639–649

[13] Krupa I., Boudenne A., Ibos L.: Thermophysical prop- [16] Krump I., Luyt A. S., Molefi J. A.: Changes in free
erties of polyethylene filled with metal coated surface energy as an indicator of polymer blend misci-
polyamide particles. European Polymer Journal, 43, bility. Materials Letters, 59, 517–519 (2005).
2443–2452 (2007). DOI:10.1016/j.matlet.2004.10.039
DOI: 10.1016/j.eurpolymj.2007.03.032 [17] Omastová M., Podhradská S., Prokeš J., Janigová I.,
[14] Tracz A., Kucinska I., Wostek-Wojciechowska D., Stejskal J.: Thermal ageing of conducting polymeric
Jeszka J. K.: The influence of micro- and nano-parti- composites. Polymer Degradation and Stability, 82,
cles on model atomically flat surfaces on crystalliza- 251–256 (2003).
tion of polyethylene. European Polymer Journal, 41, DOI: 10.1016/S0141-3910(03)00218-0
501–509 (2005). [18] Park J. W., Oh S. C., Lee H. P., Kim T., Yoo K. O.: A
DOI: 10.1016/j.eurpolymj.2004.09.019 kinetic analysis of thermal degradation of polymers
[15] Huang X., Ke Q., Kim C., Zhong H., Wei P., Wang using a dynamic method. Polymer Degradation and
G., Liu F., Jiang P.: Nonisothermal crystallization Stability, 67, 535–540 (2000).
behaviour and nucleating of LDPE/Al nano- and DOI: 10.1016/S0141-3910(99)00155-X
micro-composites. Polymer Engineering and Science, [19] Rusu M., Sofian N., Rusu D.: Mechanical and thermal
47, 1052–1061 (2007). properties of zinc powder filled high density polyeth-
DOI: 10.1002/pen.20784 ylene composites. Polymer Testing, 20, 409–417
(2001).

649