Comparative thermoanalytical investigation of the cross-linking behaviour of three different paraffin waxes in the presence of potassium persulphate

Article abstract of DOI:10.1016/s0040-6031(99)00178-1

The cross-linking of three different paraffin waxes - a hard oxidised wax, a medium wax, and a hard wax with narrow molar mass distribution - in the presence of potassium persulphate (PPS) was investigated thermoanalytically. DSC and TG curves for untreated waxes and waxes treated with different amounts of PPS were obtained and the observations were compared with gel content data obtained through gravimetric analysis. Although initial indications were that cross-linking did occur, reheat curves of wax/PPS mixtures as well as gel content data showed that no cross-linking occurred. It was, however, found that PPS most probably decomposed into KHSO₄ and that this decomposition gave rise to functionalisation of the waxes.

Full text of DOI:10.1016/s0040-6031(99)00178-1

Thermochimica Acta 339 (1999) 55±67
Comparative thermoanalytical investigation of the cross-linking behaviour of
three different paraf®n waxes in the presence of potassium persulphate
a,*
A.S. Luyt , K. Ishripersadh , D.N. Timm
b
b
a
Department of Chemistry, University of the North (Qwa-Qwa), Private Bag X13, Phuthaditjhaba, 9866 South Africa
b
Department of Chemistry, ML Sultan Technikon, PO Box 1334, Durban, 4000 South Africa
Received 11 January 1999; accepted 21 June 1999
Abstract
The cross-linking of three different paraf®n waxes ± a hard oxidised wax, a medium wax, and a hard wax with narrow molar
mass distribution ± in the presence of potassium persulphate (PPS) was investigated thermoanalytically. DSC and TG curves
for untreated waxes and waxes treated with different amounts of PPS were obtained and the observations were compared with
gel content data obtained through gravimetric analysis. Although initial indications were that cross-linking did occur, reheat
curves of wax/PPS mixtures as well as gel content data showed that no cross-linking occurred. It was, however, found that PPS
most probably decomposed into KHSO and that this decomposition gave rise to functionalisation of the waxes. # 1999
4
Published by Elsevier Science B.V. All rights reserved.
Keywords: Wax; Cross-linking; Differential scanning calorimetry; Thermogravimetry
1
. Introduction
hydrogen peroxide. In recent (unpublished) experi-
ments it was also used as initiator in the curing of
polymer membranes in an aqueous medium.
Waxes, being long-chain hydrocarbons, may occur
naturally as animal, vegetable and mineral wax or may
be synthesised [1,4]. Most of these waxes can be cross-
linked, but very little is known on the cross-linkability
of the different waxes and the in¯uence of cross-
linking on their physical and mechanical properties.
Potassium persulphate (PPS), also known as potas-
sium peroxodisulphate, is principally used as an initia-
tor for ole®n polymerisation in aqueous systems [2]. It
is also used for etching and curing of grouts for soil
stabilisation. The dry solid decomposes thermally to
oxygen and potassium pyrosulphate [3] and hydro-
lyses in acid to potassium peroxymonosulphate and
Luyt et al. [3] investigated the possible curing of a
hard paraf®n wax in the presence of PPS thermoana-
lytically. They tried to establish the cross-linkability
and extent of cross-linking through FTIR analyses and
wax solubility. Although these results positively indi-
cated that cross-linking may have occurred in the
presence of PPS, the results of this study will prove
beyond any doubt that PPS does not initiate wax cross-
linking.
In this study, we investigated the cross-linkability of
three different waxes in the presence of PPS. Differ-
ential scanning calorimetry (DSC) and thermogravi-
metric analysis (TGA) were used to study possible
cross-linking processes, while gravimetric analysis of
*
Corresponding author. Fax: +27-58-713-0152
0040-6031/99/$ ± see front matter # 1999 Published by Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 9 9 ) 0 0 1 7 8 - 1

5
6
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
gel content was used to establish whether cross-link-
ing actually occurred.
with a narrow carbon number distribution ranging
from C ±C [5,6]. It is chemically inert and is free
38
93
of aromatics. Table 3 shows some of its properties.
PPS, obtained from BDH Laboratory Supplies in
Poole, England is supposed to decompose at a tem-
perature below 1008C [7]. The grade used was BDH
AnalaR.
2
. Materials and methods
The materials used were three types of paraf®n
waxes (waxes 1±3) supplied by Sch uÈ mann-Sasol in
South Africa. Wax 1 is a typical hard oxidised Fischer-
Tropsch wax that exhibits some properties of micro-
crystalline waxes [5,6]. This wax consists of long-
chain acids, esters, alcohols, ketones and some
unreacted paraf®ns (alkanes). It is also an emulsi®able
wax which has good solvent±retention properties that
allows it to be applied in polishes and coatings. Some
of its characteristics are listed in Table 1. Wax 2 is a
paraf®nic Fischer-Tropsch wax with an average mole-
cular formula of C H [5,6]. This medium melting
The differential scanning calorimeter was a Mettler
DSC20 with a Mettler TC11 thermal analysis proces-
sor. The thermogravimetric analyser was a Mettler
TG50 thermobalance connected to a Mettler TC10A
thermal analysis processor. For both DSC and TG
�
1
analyses the heating rate was 108C min and the
�
1
nitrogen ¯ow through the cell was 85 ml min
The average sample size was 10±15 mg.
.
The amount of gel produced as a result of cross-
linking was determined by extraction in a mixture of
toluene and water, since uncured wax will dissolve in
2
3 48
point wax has chains that are about 84% unbranched
and is used mostly in candle moulding and polishes.
Some characteristics of this wax are seen in Table 2.
Wax 3 is a high molecular mass Fischer-Tropsch wax
toluene and unreacted PPS/KHSO will dissolve in
4
water. Wax samples, initially mixed with, respectively,
10%, 30% and 50% of curing agent, were cured
thermally from 308C to 2158C in the DSC in a nitrogen
Table 1
Properties of wax 1
Property
Value
Average molecular mass
Boiling point
660
2398C
Congealing point
Softening point
808C minimum
1008C
Off-white waxy beads
Physical appearance
Reactivity data
Reactions: reacts at high temperatures with oxidising agents
Thermal decomposition: occurs at Æ2508C forming hydrocarbon gases
Air and water: no reaction
Table 2
Properties of wax 2
Property
Value
Average molecular mass
Congealing point
440
65±708C
Off-white waxy beads
Physical appearance
Reactivity data
Reactions: reacts at high temperatures with oxidising agents
Thermal decomposition: occurs at Æ2508C forming hydrocarbon gases
Air and water: no reaction

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
57
Table 3
Properties of wax 3
Property
Value
Average molecular mass
Congealing point
1300
102±1088C
Hard white waxy beads
Physical appearance
Reactivity data
Reactions: reacts at high temperatures with oxidising agents
Thermal decomposition: occurs at Æ2508C forming hydrocarbon gases
Air and water: no reaction
atmosphere. The cured samples were weighed and
inserted into a pre-weighed wire mesh (38 mm). The
samples were inserted in a Soxhlet apparatus and
heated for 6 h. The extracted samples were removed
from the Soxhlet apparatus and vacuum dried over-
night before weighing. The gel content was calculated
as follows [8]:
where W1 is the mass of the sample, W2 mass of the
mesh pouch, sample and pan after extraction, and W3
is the mass of the mesh pouch, sample and pan before
extraction
Step 2 : %Gel ˆ 100 � %Extract:
3
. Results and discussion
Step 1 : %Extract
ˆ
ꢀmass loss during extraction=sample mass†
The thermal events observed during the heating of
pure potassium persulphate (PPS) are seen in Fig. 1.
 100 ˆ ‰ꢀW3 � W2†=W1†Š  100;
Fig. 1. DSC and TG curves for potassium persulphate (PPS).

5
8
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
The DSC curve shows an immediate change in heat
capacity at the start temperature, followed by a weak
endotherm at about 1408C and a strong exotherm
between 2008C and 2508C. The immediate reaction
occurring at about 308C and the weak endotherm are
reproducible, but cannot be explained because of a
lack of available information on the thermal decom-
position mechanism of PPS. The exotherm between
however, changes the melting characteristics of the
wax, as seen in the shapes of the wax melting
endotherms and the positions of their maxima. In
all the three treated samples the melting endotherms
form sharper peaks with maxima at about 1008C. For
these samples, the exotherms between 1858C and
2408C are similar to that observed for pure, unmixed
PPS (Fig. 1), but the peak maxima are at somewhat
lower temperatures. The decomposition exotherms
also increase in size with increasing PPS content,
but this decomposition of PPS does not give rise to
cross-linking of the wax, as will be seen later in the
discussion.
Reheating the PPS treated wax samples showed an
interesting phenomenon (Figs. 3±5). There are slight
changes in the shapes of the wax melting endotherms
and the positions of their maxima, but the total areas
remained approximately the same than those of the
wax melting endotherms in the ®rst heat. Two
endotherms, the ®rst considerably weaker than the
second, are also observed at about 1808C and 2008C.
The only explanation for this observation was that the
peaks represented phase changes and/or melting of the
decomposition product(s) of PPS. To verify whether
2
PPS into K SO , which is the decomposition product
008C and 2508C probably is the decomposition of
2
4
according to literature [3]. The decomposition tem-
perature is, however, much higher than the acceptable
literature value [7]. Thermogravimetric analysis of
PPS shows a small mass loss starting at about
2
008C which is the same than the start temperature
of the decomposition exotherm in the DSC curve. It
was, however, not possible to draw a correlation
between the observed mass loss and that expected
for the decomposition reaction mentioned above.
Fig. 2 represents the DSC curves of the untreated
wax and 10%, 30% and 50% m/m mixtures of potas-
sium persulphate (PPS) with the wax. Melting of the
untreated wax is seen as a broad endotherm in the
temperature range 50±1008C. The presence of PPS,
Fig. 2. DSC curves of untreated wax 1 as well as samples mixed with different amounts of PPS.

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
59
Fig. 3. Heat and reheat curves of 90/10 m/m wax 1/PPS.
Fig. 4. Heat and reheat curves of 70/30 m/m wax 1/PPS.

6
0
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
Fig. 5. Heat and reheat curves of 50/50 m/m wax 1/PPS.
4
Fig. 6. Comparative DSC curves for KHSO and reheat of a wax/PPS mixture.

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
61
this was true, DSC analyses were done on pure
samples of different possible PPS decomposition pro-
ducts. Comparison of the two DSC curves in Fig. 6
of the wax in the treated samples have different shapes
to that of the untreated sample, indicating that the
added PPS in¯uenced the structure and melting beha-
viour of the wax. The peak intensities also decrease
with increasing amount of PPS as a result of the
smaller wax fractions in the respective samples.
Exotherms, similar to the decomposition exotherm
observed for pure PPS (Fig. 1), can be seen in the
temperature range 200±2308C. The increase in peak
intensity relates to the increased PPS fractions in the
samples. There is also a slight shift to lower tempera-
tures with increased amount of PPS. We shall also
show later that, although cross-linking was initially
thought to occur, the decomposition of PPS did not
initiate wax cross-linking.
proves that KHSO is, within reasonable doubt, the
4
decomposition product of PPS when heated in a wax
mixture. None of the other compounds gave DSC
curves that looked nearly the same than the reheat
curves discussed above. A possible explanation for
this phenomenon is that the decomposing PPS extracts
hydrogen from the wax chain, and in the process,
functionalises the wax with sulphur and oxygen con-
taining groups. FTIR spectra, not presented here, show
the development of vibration bands in the 1000±
�
1
1
250 cm region.
The TG curves of the untreated wax and 10%, 30%
and 50% m/m mixtures of PPS with the wax are seen
in Fig. 7. Here the untreated wax shows only a small
mass loss towards the end of the curve and gradually
increasing mass losses, starting at approximately
1508C, are observed for the mixtures containing
increasing amounts of PPS.
DSC thermal events of untreated and PPS treated
wax 2 are seen in Fig. 8. Untreated wax 2 melts in the
temperature range 40±808C. The melting endotherms
Comparison of the heating and reheating DSC
curves of treated waxes are shown in Figs. 9±11.
The melting endotherm of the wax in the sample
did not decrease in size after heating it in the presence
of PPS, indicating the absence of any cross-linking in
this process. The two higher temperature endotherms
in the reheat curves indicate, as discussed earlier, that
PPS probably decomposed to KHSO , functionalising
4
the wax in the process.
Fig. 7. TG curves of untreated wax 1 as well as samples mixed with different amounts of PPS.

6
2
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
Fig. 8. DSC curves of untreated wax 2 as well as samples mixed with different amounts of PPS.
Fig. 9. Heat and reheat curves of 90/10 m/m wax 2/PPS.

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
63
Fig. 10. Heat and reheat curves of 70/30 m/m wax 2/PPS.
Fig. 11. Heat and reheat curves of 50/50 m/m wax 2/PPS.

6
4
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
Fig. 12. TG curves of untreated wax 2 as well as samples mixed with different amounts of PPS
Thermogravimetric analysis of treated and untreated
Comparison of the reheat curves with the initial heat
curves (Figs. 14±16) shows essentially no decrease in
size for the wax melting endotherms on reheating the
samples. There are, however, slight changes in the
shapes and positions of these endotherms compared to
the endotherms during ®rst heat. This again points to
theabsenceofanycross-linkingreactioninthepresence
ofPPS,butfunctionalisationofthewaxprobablycaused
the change in shapes and positions of the wax melting
endotherms. As was the case with the previous two
waxes, the reheat curves display two weak endotherms
between1808Cand2108C,typicalofthephasetransition
and melting endotherms for KHSO4.
The TGA curves (Fig. 17) follow essentially the
same pattern observed for the previous waxes. The
untreated wax shows an initial slight increase in mass,
which may be attributed to moisture absorption. It did,
however, show no signi®cant mass loss up to the
maximum temperature of analysis. The treated sam-
ples show small mass losses, starting at about 2008C.
The extent of mass loss for the respective samples
increases with increasing amounts of PPS mixed into
the samples.
samples of wax 2 can be seen in Fig. 12. The untreated
as well as treated samples show equivalent amounts
of mass loss, indicating that the mass loss in this case
is primarily due to wax degradation. Since we know
that PPS decomposes to a certain extent, the small dif-
ference in total mass loss at 2308C is probably due to
evaporation of volatile PPS decomposition products.
The DSC curves in Fig. 13 show wax melting
endotherms between 958C and 1258C for all the
samples. Contrary to the DSC analyses previously
discussed, the melting peaks for the untreated and
PPS treated samples have a similar shape, indicating
that the history built into the initial wax structure was
not disturbed by the presence of PPS. The respective
peak sizes are also related to the fraction of wax
present in each sample. The exotherms between
2
008C and 2308C for the treated samples are for the
decomposition of PPS and their respective heights are
directly proportional to the amount of PPS initially
mixed into the sample. As in the case of wax 2, these
exotherms also shift to a slightly lower temperature
with increasing PPS : wax ratio.

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
65
Fig. 13. DSC curves of untreated wax 3 as well as samples mixed with different amounts of PPS.
Fig. 14. Heat and reheat curves of 90/10 m/m wax 3/PPS.

6
6
A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
Fig. 15. Heat and reheat curves of 70/30 m/m wax 3/PPS.
Fig. 16. Heat and reheat curves of 50/50 m/m wax 3/PPS.

A.S. Luyt et al. / Thermochimica Acta 339 (1999) 55±67
67
Fig. 17. TG curves of untreated wax 3 as well as samples mixed with different amounts of PPS.
The gel contents of the different samples were
determined and the values are summarised in Table 4.
For all the three waxes the gel content values for
untreated and treated samples were constant and low
within experimental error. This is a clear indication
that no wax cross-linking occurred in the presence of
potassium persulphate.
PPS. We did, however, establish that heating of PPS in
the presence of a wax gave rise to the formation of
KHSO and that, in the process, functionalisation of
4
the wax took place.
Acknowledgements
ML Sultan Technikon, the University of the North
(Qwa-Qwa), and the Foundation for Research Devel-
opment in South Africa are thanked for their ®nancial
assistance.
4
. Conclusions
None of the investigated waxes cross-linked in the
presence of potassium persulphate, as was evident
from the DSC reheat curves of the treated samples
and the lack of any gel formation in the presence of
References
[
1] R.Z. Hamilton, Chemistry, Molecular Biology and Functions,
st ed., Dundee, 1995.
Table 4
1
Gel content values
[
[
2] Encyclopedia of Chemical Technology, 3rd ed., vol. 18.
3] A.S. Luyt, N.S. Nhlapo, H.C.M. Vosloo, Comparison of two
curing agents in thermal cross-linking. I. Hard paraffin wax, J.
Appl. Polym. Sci., accepted for publication.
Sample %Gel content
Untreated
wax
Wax ‡ 10% Wax ‡ 30% Wax ‡ 50%
PPS
PPS
PPS
[4] G.C. Gerrans, H.R. Corbett, Spectrum 28 (3) (1990).
[
[
5] Sasolchem Technical Bulletin 1993.
Wax 1
Wax 2
Wax 3
4
12
1
6
12
1
6
14
3
6
14
3
6] R.C. Weast (Ed.), CRC Handbook of Chemistry and Physics,
58th ed., CRC Press, FL, 1977.
[
7] Annual Book of ASTM Standards, 1990, pp. 132±138.