1 Carbon Hydrogen Polymers - Smithers Rapra Producing Alkanes Dienes C H 2 H 2 CC 2 (1.1) 1 Carbon Hydrogen Polymers. 2 Thermal Stability of Polymers ... A mechanism [1] involving

1

1.1 Polyethylene

Thermal degradation of polymers usually proceeds via a number of possible mechanisms, which can generally be grouped into three classes: (a) random scission, (b) depolymerisation, and (c) side group elimination.

1.1.1 Random Scission

Random scission results from the production of free radicals along the backbone of the polymer, which causes the macromolecule to be fragmented into smaller molecules of varying chain lengths. On chromatographic analysis these fragments reveal a repeating series of oligomers frequently differing in chain length by the number of carbons in the original monomer:

C C C CC C C +

C C C C C C C C

H3C CH3

H3C

Alkanes

Producing

Alkanes

Dienes

C H2

H2 C H2C C (1.1)

1 Carbon Hydrogen Polymers

2

Thermal Stability of Polymers

Polyolefins generally degrade through a random scission mechanism, and polyethylene (PE) is a good example of this behaviour. When a free radical is formed along the chain of PE, chain scission occurs, producing a molecule with an unsaturated end and another with a terminal free radical. This free radical may abstract a hydrogen from a neighbouring carbon, producing a saturated end and a new radical, or it may combine with another free radical to form an alkane. Multiple cleavages produce molecules small enough to be volatile, with double bonds at both ends, one end, or neither end. Since the scission is random, molecules are made with a wide variety of chain lengths. These appear in the pyrogram as a series of triplet peaks. Each triplet consists of an alkane, an alkene, and a diene of a specific chain length. The hydrocarbons in each triplet have one more carbon than the molecules in the triplet that eluted just prior to it.

The chromatogram resulting from the pyrolysis of PE at 750 °C shows oligomers containing up to 30 carbons.

It is interesting to compare results obtainable by gel permeation chromatography (GPC) of PE, polypropylene (PP), and an ethylene–propylene copolymer. The pyrolysis products were hydrogenated at 200 °C by passing through a small hydrogenation section containing 0.75% platinum on 30/50 mesh aluminium oxide. The hydrogenated pyrolysis products were then separated on squalane on a fireback column, and the separated compounds detected by a katharometer. Under these experimental conditions only alkanes up to C9 could be detected.

It can be seen that major differences occur in the products of thermal degradation that are obtained for these three similar polymers. PE produces major amounts of normal C2 to C8 alkanes and minor amounts of 2-methyl and 3-methyl compounds such as isopentane and 3-methylpentane, indicative of short-chain branching on the polymer backbone. For PP, branched alkanes predominate, these peaks occurring in regular patterns, e.g., 2-methyl, 3-ethyl, and 2,4-dimethylpentane and 2,4-dimethylheptane, which are almost absent in the PE pyrolysate. Minor components obtained from PP are normal paraffins present in decreasing amounts up to n-hexane. This is to be contrasted with the pyrogram of PE, where n-alkanes predominate. The ethylene–propylene copolymer, as might be expected, produces both normal and branched alkanes. The concentrations of 2,4-dimethylpentane and 2,4-dimethylheptane are lower than those that occur in PP.

A mechanism [1] involving two equilibrium stages, i.e., the formation of free radicals and abstraction of hydrogen atoms by these radicals, has been suggested for the thermal degradation of polyalkenes. The initiation of polyalkene degradation consists of the cleavage of the carbon–carbon bond and the macromolecules to form free radicals:

3

Carbon Hydrogen Polymers

~CH2–CH2–CH2–CH2~ ~CH2–CH2 + CH2–CH2~ (1.2)

Such cleavage occurs at the weakest bonds (the tertiary carbon atom in groups containing other atoms, and so on).

The scission of the chain, with the elimination of small amounts (1%) of monomer, occurs at its free-radical ends, a process which has been termed ‘unzipping’:

~CH2–CH2–CH2–CH2 ~CH2–CH2 + CH2=CH2 (1.3)

The chain transfer, in which the free radical abstracts an hydrogen atom, may proceed both intra- and inter-molecularly. In this case both saturated and unsaturated groups are formed at the chain ends together with a new free radical:

~CH2–CH2 + CH2~CH2–CH2–CH2–CH2~ ~CH2–CH3 +~CH2–CH2–CH2=CH2 or ~CH2–CH–CH2–CH2~ (1.4)

Chain scission occurs because of the recombination of two free radicals with the formation of one linear or branched chain and, in some cases, of a crosslinked polymer. As is seen from the reaction schemes given, free radicals formed during polyalkene degradation are involved in two competing reactions: free radical transfer of an hydrogen atom and chain scission to form monomer. The number of hydrogen atoms in the chain determines which reaction will predominate. Since PE is the polyalkene most saturated with hydrogen atoms, the chain transfer reaction predominates. The low yield of monomer during the thermal degradation of PE may thus be associated with this mechanism.

If a fraction of the hydrogen atoms in the polyalkene chain are replaced by methyl or other small groups, then the hydrogen atom transfer process becomes difficult; this leads to the formation of free radicals which continue polymer degradation to produce monomer. For example, during the thermal degradation of polyisobutylene, the cleavage of a fraction of the carbon–carbon bonds causes the formation of free radicals which promote chain scission to produce monomer (in up to 18% yield):

~CH2–C(CH3)2–CH2 ~CH2+C(CH3)2=CH2 (1.5)

4


The thermal decomposition of polyalkenes may be presented as described next. In contrast to low molecular mass substances, for example alkanes, macromolecules do not behave as a single kinetic unit: some of their elements may acquire greater amounts of energy, while the others may acquire less. At the same time all three types of motion of these elements of the macromolecule are limited by their being chemically bound to the remainder of the macromolecule with its large dimensions. The resultant fluctuation in tensions leads to the cleavage of chemical bonds occurring in different parts of the macromolecule. If, as in PE, there are sufficient quantities of mobile hydrogen atoms in the chain, then such a macromolecular scission is accompanied by the abstraction of an hydrogen atom from the carbon atom nearest to the site of scission. However, when the hydrogen atom content is low, as in polyisobutylene and propylene, the occurrence of chain scission is not accompanied by the transfer of hydrogen atoms. Instead, the free radicals formed continue the process of chain cleavage to form monomer.

1.1.2 Depolymerisation

Depolymerisation is a free radical mechanism in which the polymer essentially reverts to a monomer or monomers. Unlike random scission, which produces fragments of a variety of chain lengths, depolymerisation generates a simple chromatogram consisting of large peaks for the monomers from which the polymer or copolymer was produced:

(~CH2Ph – CH2Ph~)n → nPhCH = CH2

1.1.3 Side Group Elimination

This is usually a two-stage process in which the polymer chain is first stripped of atoms or molecules attached to the backbone of the polymers, leaving an unsaturated chain. This polyene then undergoes further reactions, including scission, aromatisation, and char formation.

It is known that, in the absence of oxygen, PE is thermally stable. The thermogravimetric analysis (TGA) of this polymer shows that decomposition of this polymer begins at about 280 °C, and when the temperature is near to 350 °C, its thermal degradation proceeds rapidly with the elimination of considerable quantities of volatile materials. The half-life temperature (i.e., that leading to 50% weight loss on heating for 40–45

5


minutes) of PE is 406 °C. Only 1% of monomer is formed during the thermal degradation of PE, which indicates the absence of the chain depolymerisation reaction of this polymer. Since all carbon–carbon bonds in PE (except for those sited at the chain ends, at branching sites and at other side-groups) possess the same strength, then the probability of their degradation on heating is the same. Thus, the pattern of thermal degradation of PE macromolecules is random.

It has been found that, on the thermal degradation of comparatively low molecular mass PE under vacuum or in a nitrogen atmosphere under normal pressure, the volatile products of degradation consist of molecular fragments of PE macromolecules. The viscosity-average molecular mass of the residue falls sharply on a 10 hours exposure above 315 °C (Figure 1.1). Subsequent heating leads to a smoother decrease in the molecular mass of the polymer. The degradation of PE in a nitrogen atmosphere results in growth of its unsaturation with the extent of its decomposition; this may be associated with the disproportionation of more low molecular mass macroradicals at their end-groups or with the transfer of hydrogen atoms from two neighbouring carbon atoms of the main chain.

0.06[n0][n]

W(%)

0.04

0.02

05 10 15

Figure 1.1 Variation in relative intrinsic viscosity of the residue from polymethylene decomposition (W is the degree of conversion: at 375 °C;

at 480 °C; X at 390 °C; at 400 °C; the solid curve is that calculated. Source: Author’s own files

Analysis of the degradation products of PE produced under vacuum and under atmospheres of nitrogen and helium at different temperatures made it possible to establish that the higher the decomposition temperature, the lower the molecular mass

6


of the decomposition products, and the larger the fraction of gaseous compounds. Thus, the fraction of ethylene is 0% at 500 °C, 5.5% at 800 °C and 26.4% at 1,200 °C. This may be explained by the fact that, at temperatures above 800 °C, both competing reactions proceed, i.e., the cleavage of chains into comparatively high molecular mass fragments with the transfer of hydrogen atoms, and the cleavage of free macroradicals via the chain mechanism to form monomer. The rate and significance of the latter process increase at higher temperatures.

Studies on the thermal degradation of PE samples with different molecular masses in the isothermal regime at different temperatures have shown that the kinetic curves have linear plots up to 70% weight loss (Figure 1.2), which point to a zero-order reaction. The activation energy of thermal degradation increases with the molecular mass of the polymer from 192.3 kJ/mol (molecular mass 11,000) up to 276.3 kJ/mol (molecular mass 23,000) [2].

604

3

2

1

20

0 2.4 (a) 7.2

W(%)

45

321

100

20

0 2.4 (b) 7.2 t(103s)

W(%)

45

32

1

60

20

0 2.4 (c) 7.2 9.6 t(103s)

W(%)

Figure 1.2 Kinetic curves of the thermal degradation of polyethylene with molecular mass (a) 23,000, (b) 16,000 and (c) 11,000 at different temperatures: (a) 1 – 405 °C, 2 – 393 °C, 3 – 412 °C, 4 – 436 °C; (b) 1 –317 °C, 2 – 376 °C, 3 – 388 °C, 4 – 403 °C, 5 – 402 °C, (c) 1 –374 °C, 2 – 393 °C, 3 – 410 °C, 4 – 420

°C, 5 – 396 °C. Source: Author’s own files

7


A comparison of the decomposition processes of three PE samples (commercial PE having a molecular mass of 20,000, unbranched high molecular mass polymethylene and strongly branched PE) differing in structure and molecular mass has enabled the characterisation of substantial differences in their thermal degradation.

The dependence of the degradation rate upon the extent of degradation was calculated from weight loss curves of commercial PE (Figure 1.3). It was found that in its initial stage (up to 10–40%) the rate of the process is very high, but then it is sharply reduced and transforms into a virtually linear dependence, which is extrapolated to zero at 100% decomposition. Such a kinetic pattern derives from two main factors: (i) the polydisperse nature of the polymer, whose low molecular mass fractions are easily removed in the initial degradation stages, and (ii) the presence of weak bonds in the main polymer chain caused by the presence of hydroperoxide and other groups. The weak links lead to the splitting of carbon–carbon bonds and other reactions in the initial stages of the process. These particular factors diminish in significance over the course of degradation, and the rate of the process decreases sharply.

4 5

32

1

0.9

0.7

0.5

0.3

0.10

20 40

v/(%min–1)

W(%)Figure 1.3 Dependence of the rate of thermal degradation (υ) of commercial PE with molecular mass 20,000 on the degree of decomposition at temperatures: 1 -

392 °C, 2 – 387 °C, 3 – 382 °C, 4 – 377 °C, 5 – 372 °C. Source: Author’s own files

Studies on the rate of thermal degradation of unbranched high molecular mass PE (polymethylene) have shown that the process follows almost ideal first-order kinetics. The rate of thermal degradation is directly proportional to the temperature of the process. The dependence of the rate of elimination of volatiles on their quantity

8


shows complex behaviour. The rate curves pass through a maximum of degrees of decomposition from 2% to 10%, and then become gradually straight, finally falling to zero at 100% decomposition of the polymer.

For the thermal degradation of branched PE, as in the case of low molecular mass commercial PE, high rates of decomposition are observed in the initial stages of the process. After 15–30% decomposition of the polymer, the rate falls sharply.

Anderson and Freeman [2] found that in the thermal decomposition of high-pressure PE (low-density polyethylene; LDPE) under vacuum there were three stages of decomposition. In the first stage, up to 3% conversion, zero-order kinetics was followed and the value of the energy of activation (E) was about 48 kcal/mole; in the second stage, 3–15% conversion, the order was still zero and the value of E was found to be about 6 kcal/mole (Figure 1.4). From 15–35% conversion, the reaction appeared to involve the transition from zero-order to first-order kinetics, and above 35% conversion, the kinetics were first-order and the value of E was found to be about 67 kcal/mole (Figure 1.5). The initial stage was attributed to the degradation of branched chains of short length since these should require less energy for bond rupture. The second stage was attributed to end chain cleavage, and the final stage, to the random rupture of carbon–carbon bonds, which requires an energy similar to that found for this stage.

10

5.0

1.0

0.5

0.1

1.44 1.48 1.52 1.56

(1)

(2)

Rea

ctio

n ra

te (d

W/d

t), m

g/m

in

1.600.05

103/T(°K)–1

Figure 1.4 Temperature dependency plot of the low temperature thermal degradation of polyethylene in vacuum: (1) up to 3% degradation, (2) from 3 to

15% degradation. Reproduced with permission from D.A. Anderson and E.S. Freeman, Journal of Polymer Science, 1961, 54, 253, ©1961, Wiley [2]

9


2095%

35%

10

105 15 20 25

–10

0

∆ lo

g dW

/dt ×

102

∆ log Wr × 102

Figure 1.5 Kinetics of the thermal degradation of polyethylene in vacuum. Reproduced with permission from D.A. Anderson and E.S. Freeman, Journal of

Polymer Science, 1961, 54, 253, ©1961, Wiley [2]

1.1.3.1 Differential Thermal Analysis

Irgashi and Kambe [3] also studied the thermal degradation of polyethylene and used dynamic thermal analysis (DTA) as well as TGA techniques. The experiments were carried out in both air and nitrogen. The PE studied were, two low-density samples. By means of DTA, the crystallinities of the high-pressure samples were found to be 33% and 36% while those for the low-pressure samples were 64% and 77%, and the melting points of the latter samples were higher than those of the former.

Figure 1.6 shows DTA curves for a high-density PE (HDPE) in air and in nitrogen. The initial large peaks represent the melting of the sample while the two final peaks denote thermal decomposition. The upper curve also shows two small exothermic peaks, which presumably are due to oxidation. From such curves it would appear that thermal decomposition for high and low-pressure PE occurs in one stage. The weight losses of the two low-density samples began at about 360 °C, which is about 20 °C higher than for the high-pressure samples. These results were attributed to the branching present in the high-pressure samples. A branched PE is presumably more thermally unstable than a linear one since thermal decomposition may be initiated at a branched point with a tertiary hydrogen atom. Isothermal experiments [4] support these findings.

10


100

Diff

eren

tial t

empe

ratu

re (e

ndot

herm

ic)

0 200 300 400 500Temperature, °C

1 ° C

Figure 1.6 Differential thermal analysis curves for high-density Hisex 5000. Upper curve, in air; lower curve, in a nitrogen atmosphere. Reproduced with permission

from S. Irgashi and H. Kambe, Polymer Preprints, 1964, 5, 333. © 1964, American Chemical Society [3]

In Figure 1.7 are shown weight losses for a paraffin in comparison with an HDPE. The weight loss of the paraffin began at 180 °C and that of the polymer at 380 °C. The degradation products were trapped and gave melting points, for paraffin and the polymer of 55.0–55.5 °C and 51.0–67.0 °C, respectively. The melting point of the trapped products for paraffin corresponds well with the original one, indicating vaporisation of the paraffin, while the wide melting point range of the polymer degradation products indicates the presence of a large number of hydrocarbons. On the basis of isothermal methods, Madorsky [5] has found that in the degradation of PE, a whole spectrum of chain fragments was obtained, each containing from one to fifty or more carbons, depending upon the pyrolysis temperature. These results would appear to support the view that random chain scission is involved in the degradation of PE. However, it may be interesting to note that on the basis of isothermal methods, others [6] have reported that in the degradation of PE, a large amount of ethylene is formed at higher temperatures than they used (about 600 °C). The apparent activation energy for the degradation in nitrogen was 14 kcal/mol for both LDPE and HDPE. However, some workers obtained reaction orders of zero and some of one (compare the results reported by Madorsky [5] with those obtained by Jellinek [7]). This seems to be dependent on details in the method of measurement and clearly needs to be resolved.

11


2000

1.0

2.0

3.0

4.0

5.0

300 400 500Temperature, °C

Wei

ght l

oss,

%/ ° C

Figure 1.7 Thermogravimetric analysis curves for paraffin and high-density polyethylene in a nitrogen atmosphere: ( ) paraffin; (O) high-density polyethylene.

Reproduced with permission from S. Irgashi and H. Kambe, Polymer Preprints, 1964, 5, 333. © 1964, American Chemical Society [3]

1.1.3.2 Differential Scanning Calorimetry

Camacho and Karlsson [8] have examined the thermal stability of recycled HDPE, PP and their blends in their discussion of environmental concerns and producers’ liability concerning the dispersal, collection and recycling of these polymers.

These studies included a comparison of results obtained by differential scanning calorimetry (DSC), TGA and chemiluminescence.

All these techniques demonstrate that the blend and its components alone undergo substantial degradation after the first or second extrusion. Accordingly, upgrade of the resin through the addition of a re-stabilising system becomes necessary to avoid their premature failure.

The thermogravimetric (TG) measurements on the reprocessed blend were shifted towards higher temperatures than the simulated curve indicating that the blends,

12


despite the multiple extrusions are more stable than expected in nitrogen. The HDPE has a stabilising effect on the PP.

DSC and TG can be used to determine the thermal/oxidative stability of PP, PE and their blends. The oxidation induction time (OIT) and the oxidation temperature (Tox) provide relatively, rapid information about the total amount of effective antioxidants in the reprocessed resin, which is important to establish the need for re-stabilisation or upgrade of the resins.

Figure 1.8 shows the DSC thermograms of the multi-extruded resins. According to Figure 1.8a and Figure 1.8b the DSC traces of PP and HDPE remained practically unaltered after the first and second extrusion, whereas further reprocessing induced changes in the peaks’ shape probably due to chain scission in PE and PP. For example, the thermogram of HDPE after the second extrusion pass exhibits a bimodal melting peak that might be attributed to the presence of species with a lower molecular weight (Mw) than the original material formed as a result of severe chain scission and not able to co-crystallise. Even though bimodality was not observed in the re-processed PP an increase in the low molar mass tail after each extrusion was noticed.

110 120

PP × 6

PP × 5

PP × 4

PP × 3PP × 2

PP × 1PP × 0

(a)

130 140 150 160 170

13


110 120 130 140 150 °C1009080

PE × 6

2 PE × 5

PE × 4

PE × 3

PE × 2

PE × 1

(b)

PE × 0

PE

PP

110 120 130 140 150 160 170 °C10090

(c)

Figure 1.8 DSC thermograms of multi-extruded: (a) polypropylene, (b) polyethylene and (c) PP/HDPE (20/80 blend). Reproduced with permission from W. Camacho and S. Karlsson, Polymer Degradation and Stability, 2002, 78, 385.

© 2002, Elsevier [8]

14


Figure 1.9a and Figure 1.9b display the TG curves of PP after the corresponding extrusion pass in nitrogen and oxygen, respectively. The differences are clearer in the experiments under nitrogen, where a systematic drop in the degradation temperature is observed after every extrusion pass. This is due to the chain scission of PP during processing. Under oxygen, Tox drops sharply for samples extruded up to four times and remains almost constant for the samples processed further. These results are in good agreement with the measurements obtained with OIT.

105

90(a)

(I)

75

60

Mas

s (%

)

45

30

15

300 350

PP ×0PP ×1PP ×2PP ×3PP ×4PP ×5PP ×6

400Temperature (°C)

450 5000

105

90(b)

(I)

75

60M

ass (

%)

45

30

15

200 250



350 4000

105

90(a)

(II)

75

60

Mas

s (%

)

45

30

15

400



5000

105

90(b)

(II)

75

60

Mas

s (%

)

45

30

15

150



450 6000

Figure 1.9 Thermogravimetric traces of multi-extruded polyolefins under different atmospheres: (a) nitrogen and (b) oxygen, I – PP, II – PE. Reproduced with

permission from W. Camacho and S. Karlsson, Polymer Degradation and Stability, 2002, 78, 385. © 2002, Elsevier [8]

Hussain and co-workers [9] also investigated the thermomechanical degradation of LDPE during the conditioning of samples in a batch blender using a variety of techniques including DSC, high-performance chromatography, GPC, nuclear magnetic resonance and dynamic viscosity measurements.

15


DSC was discussed as it was not expected to show any significant shifts in glass transition temperature or melting temperature, which are relatively insensitive to moderate changes in Mw resulting from polymer degradation. However, melt viscosity measurements are highly sensitive to Mw changes and these workers used them as a tool for monitoring polymer degradation.

The objective of this study was to make sure that degradation of PE was prevented during the conditioning process. Different techniques were used to examine the stability of PE in the melt blender. Small-strain dynamic oscillatory measurements of viscoelastic properties (η´) in a mechanical spectrometer as well as Mw and molecular weight distribution from GPC analysis were used to assess the stability of samples of linear low-density polyethylene (LLDPE) and LDPE in the melt blender.

The study included samples with and without additional antioxidants - results were compared to the properties of the ‘as-received’ samples. The results of using the different techniques can be integrated to explain: (a) the modifications that can occur due to the melt blending of PE, (b) their relation to the polymer chemistry, and (c) the possible means for detection and prevention of degradation.

If degradation is to take place, then Mw will either increase (chain build-up) or decrease (chain breakdown) and the polydispersity will be broadened. Four measurements were carried out for each of the six samples. Results are shown in Table 1.1 and indicate that there is a decrease in Mw accompanying the higher polymer conditioning temperatures. The Mw and polydispersity of the four different measurements were averaged and the standard deviation, given in parentheses, was calculated for each case.

Table 1.1 GPC characterisation of S216 (LDPE) and S229 (LLDPE)

Blender conditions MW (SD) to PD (SD) to

S229: none 105,313 (1285) - 3.57 (0.07) -

S229: Tconditioning = 190 °C 95,794 (1750) 8.770 3.32 (0.10) 4.096


S216: none 99,464 (902) - 6.45 (0.36) -

S216: Tconditioning = 190 °C 100,255(1818) 0.78 5.98 (0.14) 2.434


SD = Standard deviation Reproduced with permission from I.A. Hussain, K. Ho, S.K. Goyal, E. Karbashewski and M.C. Williams, Polymer Degradation and Stability, 2000, 68, 381. © 2000, Elsevier [9]

16


1.1.3.3 Other Techniques

Wang and co-workers [10] used Fourier-transform infrared spectroscopy, pyrolysis gas chromatography–mass spectrometry (Py-GC-MS) and TGA in their studies of thermal degradation of magnesium hydroxide (MH) and red phosphorus flame retarded LDPE composites. They calculated the apparent activation energy (Ea) of the degradation of LLDPE, and LLDPE/magnesium hydroxide and LLDPE/magnesium – red phosphorus composites in nitrogen atmosphere using the Kissinger and Flynn–Wall methods based on TG data. The Ea values obtained from both methods agree well. The peak temperature (Tk) obtained from the derivative TG curves and Ea values of the degradation of the PE and its composites are listed in Table 1.2.

Table 1.2 Apparent activation energy (Ea) obtained with the Kissinger method

Sample Heating rate (°C/min)

Tk (°C) Degradation state

Ea (kJ/mol)

LLDPE 10 492 Whole process 247

15 500

20 509

40 517

LLDPE/50% magnesium hydroxide

10 496 2nd 245

15 503

20 509

40 523

LLDPE/40% MH/10% red phosphorus

10 507 2nd 234

15 514

20 523

40 536

Reproduced with permission from Z. Wang, G. Wu, Y. Hu, Y. Ding, K. Hu and W. Fan, Polymer Degradation and Stability, 2002, 77, 427. © 2002, Elsevier [10]

17


The Py-GC-MS results show that the addition of magnesium hydroxide and phosphorus to LLDPE resin leads to great changes of the ratio of heavy compounds and light compounds. The LLDPE/40% magnesium hydroxide/10% red phosphorus composite produces heavier compounds than the LLDPE/50% magnesium hydroxide composite.

El-Goudy and El Shansbury [11] used a combination of x-ray fluorescence spectroscopy, TGA and DSC in their studies of the effects of radiation on PE.

1.2 Polypropylene and Polyisobutylene

Since on the PP macrochain every second carbon atom is tertiary and in polyisobutylene (PIB) it is quaternary, then the strength of the carbon–carbon bonds falls going from PE, through PP to PIB.

This is clearly confirmed by the data on the thermal degradation of propylene and PIB under vacuum (Table 1.3). Comparison of the number of volatile products released at corresponding temperatures has shown that PIB is less thermally stable than PP; thus the PP half-life temperature is 387 °C, while that of PIB is 344 °C.

Table 1.3 Thermal degradation of PP and PIB (0.5 h under vacuum)

PP PIB

Decomposition temperature (°C)

Quantity of volatiles (%)

Decomposition temperature (°C)

Quantity of volatiles (%)

328 8.2 288 2.8

374 28.6 336 35.7

380 41.5 353 79.3

384 46.1 376 99.2

393 63.2 390 99.7

395 70.8 425 100.0

400 86.8 500 99.4

410 96.4 800 96.6

800 100.00

Source: Author’s own files

18


Mass spectroscopic analysis of the volatile fractions released by PP during its half-life at 380–410 °C has shown that butane, butane, hexane, hexene, pentane, pentene and propylene are the main products of decomposition.

In PP, the tertiary hydrogen atom is more reactive, undergoing scission more readily than the secondary hydrogen atom. Accordingly, bond cleavage in the polymer chain occurs mainly with the transfer of an hydrogen atom:

CH2–CH –CH2–CH –CH2~ CH2– C(CH3)=CH2+CH2–CH2~

CH3 CH3 CH3

(1.6)

With PP, the cleavage of carbon–carbon bonds with the formation of free radicals and the subsequent elimination of monomer molecules is more frequent than for PE.

The presence of two methyl groups at every second carbon atom in the PIB chain causes steric difficulties for cleavage followed by the transfer of an hydrogen atom. This results in a considerable increase in the number of chain breaks with the formation of free radicals, which, on decomposition by a chain mechanism, produce monomer:

~C(CH3)2– CH2– C(CH3)2– CH2– C(CH3)2– CH2~ C(CH3)2

–CH2– C(CH3)2 + CH2– C(CH3)2– CH2~ (1.7)

~C(CH3)2 –CH2–C(CH3)2 ~C(CH3)2 + CH2=C(CH3)2 (1.8)

Simultaneously a significant number of C–C scissions occur with the transfer of hydrogen atoms, leading to the formation of compounds with saturated and unsaturated groups at the chain ends.

Studies on the rate of thermal degradation of PP and PIB have shown that the reaction order determined over the linear sections of the kinetic curves is 1, and the activation energies for the degradation of PP and PIB are 242.4 and 204.8 kJ/mol, respectively.

19


The results of studies on the influence of molecular mass and molecular mass distribution of PIB on the kinetics of its thermal degradation are of interest because of the effect of chemical structure on the thermal stability of the polymer. Several high and low molecular mass fractions and non-fractionated samples of PIB with high and low molecular masses have been used in these studies. It has been found that the molecular mass of PIB sharply decreases from about two million to about 25,000 in the initial period (10% of weight loss) of polymer degradation under vacuum at 300 °C. Thereafter the decrease in molecular mass of the polymer decelerates.

In their initial stages of decomposition, the pattern of the dependence of the rate on the degree of decomposition differs for low molecular and high molecular mass samples of PIB (Figure 1.10). While a drastic decrease in the degradation rate is typical of low molecular mass fractions (Figure 1.10, curves 1–3), the occurrence of maxima in the rate curves within the same range of polymer degradation (10–20%) is typical of high molecular mass samples of PIB.

0.161

23 4

5

10 30 50 W(%)

v/(% min–1)

6

0.12

0.08

0.04

0

Figure 1.10 Dependence of the rate of thermal degradation on the degree of decomposition of polyisobutylene with different molecular masses: 1 – 23,400,

2 – 40,000, 3 – 440,000, 4 – 100,000, 5 – 198,000 and 6 – 700,000. Source: Author’s own files

The high values of the initial rate of degradation of low molecular mass PIB are probably associated with the presence of very short chains, leading to the formation of monomer which volatilises.

At later stages of the process, the decomposition curves (Figure 1.10) reveal that the initial values of the molecular mass of the polymer and its distribution have no critical

20


influence on the rate of thermal degradation. This is explained by the fact that even in the initial stages of decomposition, the molecular mass of even high molecular mass PIB is sharply reduced through cleavages of the carbon–carbon bonds of the main chain to form low molecular mass (about 25,000) polymer and monomer.

Gomez-Elvira and co-workers [12] studied thermal stability at 90 °C to 140 °C of predominantly isotactic metallocene catalysed PP samples ranging from 3,000 to 41,000 molecular weight by chemiluminescent analysis and compared results with those obtained for predominantly isotactic Ziegler-Natta PP (ZNiPP) of increasing molecular weight. The structural differences of these two types of polymer is sufficiently different to expect very different thermal stability behaviour between them.

Gomez-Elvira and co-workers [12] presented oxidation kinetics at 90 °C, 100 °C, 120 °C and 140 °C for metallocene polymerised PP. The oxidation was carried out on powder. The chemiluminescence curves were recorded all along the induction period and the auto acceleration stage, and stopped well after auto acceleration had finished. An example of the kinetic curves obtained is shown in Figure 1.11, together with the way in which the induction period has been estimated. For these samples a bimodal distribution was always found, no matter what the molecular weight and the isotacticity were (Figure 1.12). As in the case of ZNiPP, the low temperature component appears about 15 °C before the main melting peak, the only difference being its lower relative intensity in the higher number average molecular polymer weight (Mn = 41,000).

This work revealed that the microstructural control of tactic errors, provided by metallocene catalysis, ensured that the configurational microstructure is a parameter which can lead to improve the iPP thermal stability over 100 °C.

Brambilla and co-workers [13] studied the outdoor degradation of iPP plates by means of positron annihilation lifetime spectroscopy (PALS), absorption infrared spectroscopy, DSC and density measurements. Infrared spectra reveal the presence of oxygenated species in the exposed polymer induced by external agents. Results from thermal and density analysis suggest an increase of crystallinity of the sample with exposure time. Positron data strengthened such a conclusion, showing a reduction of the amorphous zones as monitored by the corresponding decrease of positronium (Ps) formation. Furthermore, an estimation of the average sizes of the free volume holes and of the defects in the crystalline regions was obtained. PALS is a relatively simple technique which can probe the properties of the free volume holes in a non-destructive way. It is based on the fact that some of the positrons injected into the material under investigation are trapped in regions with reduced electron density where they may form a bound positron–electron positronium system.

21


200000

150000

100000

50000

00 1000 2000

induction time

t (s)

l cl (c

ps)

3000

Figure 1.11 An example of the chemiluminescence experiment showing the determination of the induction time. The curve corresponds to a sample of

metallocene catalysed PP Mn = 19,000, oxidised in O2 at 140 °C. Reproduced with permission from J.M. Gomez-Elvira, P. Tiemblo, M. Elvira, L. Matisova-Rychla

and J. Rychly, Polymer Degradation and Stability, 2004, 85, 873. © 2004, Elsevier [12]

131°

15 °CT (°C)

120 140

First scan

137° 143° 144°

Figure 1.12 Normalised first DSC scans of M-iPP samples: 5 (– - –), 4 (- - - ), 3 (——) and 2 (– – –). Reproduced with permission from J.M. Gomez-Elvira, P.

Tiemblo, M. Elvira, L. Matisova-Rychla and J. Rychly, Polymer Degradation and Stability, 2004, 85, 873. © 2004, Elsevier [12]

22


The outdoor thermal and oxidative degradation of 20 to 50 µm sheets of iPP has in previous work been shown to begin at 700 hours exposure and lead to extreme sample brittleness after 2,500 hours exposure.

An intensity parameter known as I4 determined by PALS correlates well with PP crystallinity. This intensity I4 goes to zero for a 100% crystalline polymer. The results from these various techniques are in agreement with each other. The outdoor degradation process takes place after about 2,520 hours of exposure with the formation of oxygenated species. The increase of density with weathering is well correlated to the increase of melting enthalpy, which can be interpreted as an increased crystallinity in the exposed samples. PALS data allow for a microscopic interpretation of such a conclusion, since the decrease of the amorphous regions manifests itself through a decrease Ps formation, but the characteristic sizes of free volume holes do not display significant variations during weathering.

Various workers [14, 15] have evaluated the simultaneous macroscopic and molecular reinforcement of PP with glass fibres and polymer liquid crystals (PLC) in which interlayers of PP and polyethylene terephthalate/0.6 p-hydroxybenzoic acid, PLC blends are introduced between the co-woven PP and glass fibre plies. Analysis conducted using dynamic mechanical thermal analysis, DSC, and TGA indicate that low fractions of PLC material in the composite have the dual benefits of increased rigidity and toughness. The crystallisation kinetics are influenced by the dual nucleating effects of both reinforcing agents. A decrease in crystallinity is associated with increasing PLC presence. Degradation kinetics of the composites demonstrate only one kinetic step (see Figure 1.13) in contrast to the interlayers alone.

The influence of temperature on the degradation kinetics is markedly different for the composites in comparison to the interlayers – as the TGA results in Figure 1.13 indicate. The single slope associated with the composites indicates that the degradation mechanism has a single kinetic rate. Complete degradation of the matrix occurs at 420 °C. In contrast, the interlayers have multiple degradation steps, with complete degradation occurring only at 680 °C. This indicates that an inter-phase due to the glass fibre sizing and the matrix serves to accelerate degradation once initiated.

It has been reported that PP can form crystalline inclusion complexes with β-cyclodexins (βCD). Li and co-workers [16] investigated the thermal stability of such complexes. Complex formation improved the thermostability of βCD.

23


015B

100

80

80 160 240 320 400

Temperature (degrees C)

% w

eigh

t los

s

480 560 640

60

40

20

0

025B

035B

050B

100B

015C000C

025C

035C

100C

Figure 1.13 Thermogravimetric analysis of interlayers and composites. Source: Author’s own files

1.3 Polystyrene and Copolymers

1.3.1 Polystyrenes

The most probable mechanism of the thermal degradation of polystyrene (PS) is as follows. In PS of any molecular mass above 40,000–60,000, the initial stage of the thermal degradation 5–10% weight loss) is marked by a sharp decrease in the molecular mass to 40,000–60,000, and at these values the process becomes stabilised. Values of the ‘stable’ molecular mass are virtually independent of both its initial value and the molecular mass distribution of the polymer.

The drastic decrease in the molecular mass of PS in the initial stage of the process is evidently associated with the cleavage of weak bonds in the polymer chain. This

24


continues until these breakages reach an equilibrium at the expense of the chain decomposition of lower molecular mass chains yield monomer.

Monomer formation occurs when free macroradicals decompose via a chain mechanism without hydrogen atom transfer. The half-life temperature for PS is 360 °C. The thermal degradation products consist mainly of monomer, dimer and trimer.

The thermal degradation of PS may proceed in the following ways:

• Decompositionwithouthydrogen–atomtransferandwithformationofmonomericproducts:

~CH–CH2–CH–CH2–CH–CH2–CH–CH2~

C6H5 C6H5 C6H5 C6H5 (1.9)

~CH –CH2–CH –CH2 + CH –CH2–CH –CH2~

C6H5 C6H5 C6H5 C6H5 (1.10)

~CH –CH2–CH –CH2 ~CH – CH2 + CH=CH2

C6H5 C6H5 C6H5 C6H5 (1.11)

~CH –CH2–CH –CH2–CH – CH2–CH –CH2–CH –CH2~

~CH –CH2–CH –CH2–CH2 + CH2=C –CH2–CH –CH2~

C6H5 C6H5 C6H5 C6H5 C6H5

C6H5 C6H5 C6H5 C6H5 C6H5 (1.12)

~CH2–CH –CH2 + ~CH –CH2–CH –CH2–CH –CH2~

CH2–CH –CH3 + ~ CH –CH2–CH + CH2=C –CH2~

C6H5 C6H5 C6H5 C6H5

C6H5 C6H5 C6H5 C6H5 (1.13)

25


• Theformationofchainfragmentsofasizegreaterthanthatofmonomer(dimer,trimer) occurs according to the following scheme:

~CH2–CH –CH2–CH –CH2–CH –CH2–CH –

~CH2–C=CH2 + CH2–CH2–C=CH2 +

CH2–CH2–C=CH2 + CH2 –CH2~

~CH2–CH –CH2–CH –CH2~

C6H5 C6H5 C6H5 C6H5

C6H5

C6H5

C6H5 C6H5 C6H5

C6H5 C6H5

C6H5

(1.14)

The formation of trimer and tetramer proceeds by an analogous mechanism.

There are also oligomeric compounds with an average molecular mass of about 2,000 produced on the thermal decomposition of PS. The ratio between the oligomeric and monomeric fractions is in the range 1.4–1.6.

If the degradation of PS occurs under vaccuum at temperatures up to about 530 °C, then the monomer is the main product of its degradation. However, increasing the temperature to 830–1,230 °C leads to a decrease in the monomer yield and release of ethylene, acetylene and benzene in considerable quantities. Benzene may be formed either as a result of transfer of the hydrogen atom located at the neighbouring carbon atom of the chain to the phenyl group with concomitant elimination from the main chain, or as a result of the phenyl group being eliminated as a free radical which abstracts an hydrogen atom from neighbouring chains, thus instigating the subsequent thermal decomposition of the main polymer chain; ethylene and acetylene may form because of the extensive decomposition of both the monomer and the main chain.

Quite extensive and systematic studies have been carried out concerning the rate of thermal degradation of PS samples having different molecular masses and molecular mass distributions.

Opinions differ as to whether the thermal degradation of PS follows a zero-order or first-order mechanism.

26


Figure 1.14 shows that a fall in temperature leads to the appearance of a plateau in the curves, and at the lower temperature (320 °C) the part of the curve corresponding to 30–55% weight loss is almost the straight line parallel to the abscissa i.e., zero-order. The formation of the plateau in these curves is explained by the mechanism of thermal degradation of PS. Random breaks in the main chain in the initial stages of the degradation lead to the formation of three types of groups: terminal, saturated and free radical. Most volatile substances are formed by the cleavage of monomer (and a certain quantity of dimers) via a chain mechanism from chain ends existing as free radicals. These products make up 60% of the total amount of the volatile materials.

1.1v(% min–1)

7

65

43

21

0.9

0.7

0.5

0.3

0.1

10 30 50 70 W(%)0

Figure 1.14 Dependence of the rate of thermal degradation of polystyrene (molecular mass – 230,000) on its decomposition temperature; 1 – 318 °C, 2 – 323 °C, 3 – 328 °C, 4 – 333 °C, 5 – 338 °C, 6 – 343 °C, 7 – 448 °C. Source: Author’s

own files

The random cleavage of chains and the scission of monomer molecules from free-radical ends of the chain take place at different rates. The total rate of weight loss for polymers is mainly determined by the number of chains with free radicals at their ends, which are always present during the thermal degradation of polymers. After the initial stage of the process, the rate of random scission sharply decreases, the molecular mass of the polymer becoming stabilised. At rather low decomposition temperatures the rate of formation of new terminal groups equates with the rate of disappearance of such groups because of the complete decomposition of fragments. Therefore, a plateau appears in the curves from which the reaction constant is determined and, consequently, the apparent order of the reaction is zero. At higher

27


pyrolysis temperatures, the equilibrium between these two reactions exists over a narrow range of decomposition and for a shorter period, and therefore maxima are observed in the rate curves (Figure 1.14).

At higher temperatures the degradation of the polymer may also proceed by a first-order process. The activation energy of the thermal degradation of PS of different molecular masses calculated from the reaction rate constants is 230 kJ/mol.

The replacement of the hydrogen atom in the α-position by a methyl group (α-methylstyrene) leads to a substantial effect on the thermal properties of poly(α-methylstyrene). Thus, whilst about 40% of monomer is produced during the thermal degradation of polystyrene under vacuum at temperatures of 230–530 °C, poly(α-methylstyrene) yields up to 95–100% of monomer under analogous conditions. This is due to the presence of the quarternary carbon atom which weakens the neighbouring C–C bond. The presence of phenyl and methyl groups in the α-position virtually blocks the transfer of an hydrogen atom during pyrolysis at temperatures up to 530 °C. Fragments terminating in radical sites, formed as a result of chain cleavage, easily decompose to monomer via a chain mechanism. The half-life temperature of poly(α-methylstyrene) is 290 °C [17].

If the pyrolysis of poly(α-methylstyrene) is conducted at around 830–1,230 °C, then considerable quantities of fractions with a molecular mass exceeding that of the monomer are formed. This is a result of the volatilisation of chain fragments produced during pyrolysis of the polymer from the high temperature zone before chain decay to monomer can take place. At temperatures of 830–1230 °C a definite quantity of small molecules such as acetylene, benzene, ethylene, hydrogen and methane are formed as products of the secondary, more extensive decomposition of monomer.

The rate of degradation of poly(α-methylstyrene) increases linearly over time up to about 80% decomposition, after which it is sharply reduced. This linear dependence implies the reaction is of zero-order. The activation energy of the thermal degradation of poly(α-methylstyrene) is 187.3 kJ/mol.

Other workers state that the thermal degradation of poly(α-methylstyrene) is undoubtedly a first-order reaction with an activation energy of 271.7 kJ/mol. The first-order kinetics are explained in terms of a mechanism of chain decay of the polymer proceeding randomly, which leads to the formation of free radicals. Then follows a rapid chain decay of chains having terminal free radicals to form monomer molecules. Since the rate of random cleavage of the chain is essentially lower than that of chain decay, the former predetermines the reaction rate as a whole, which therefore displays first-order behaviour.

28


A kinetic plot of the thermal decomposition of PS with a molecular weight of 360,000 under vacuum is shown in Figure 1.15. The initial points, representing about 10% reaction up to 370 °C, do not fall on a straight line and the temperature dependency plot of this low temperature stage indicate zero-order kinetics. In this case an E of 46 kcal/mole is obtained. This supports isothermal evidence that the initial low temperature degradation stage involves the splitting off of styrene monomer. Also, Madorsky [5] has indicated that in the isothermal degradation of PS, in the absence of air, a zero-order reaction is involved at the lower reaction temperatures used.

6

4

2

2

15%

95%

1 3∆Log Wr × 101

∆Log

(dW/dt)

× 10

1

4 5

0

−2

−4

Figure 1.15 Kinetics of the thermal degradation of polystyrene in vacuum Δ(1/T) = 1.0 x 10-5/K. Source: Author’s own files

At higher temperatures (over 370 °C and 15–95% conversion), values of E and n were found to be 60 ± 5 kcal/mole and unity, respectively. Isothermal experiments [7] have indicated that at the higher temperatures used, the value of n was no longer zero, and an E of 55 kcal/mole was reported. Thus, it appears that there are two degradation mechanisms for PS degradation, one predominant at lower temperatures and one at higher temperatures.

Udseth and Friedman [18] carried out a mass spectroscopy (MS) study of PS with a molecular weight of 21,000. The PS was evaporated from a probe filament heated at 1000 °C/s under both electron ionisation (EI) and methane and argon chemical

29


ionisation conditions. Smooth rhenium ribbon surface direct insertion probes were used with both the EI and thermal degradation chemical ionisation sources. Under EI conditions extensive fragmentation and depolymerisation were observed but oligomers up to C8H8 were detected. Under chemical ionisation conditions oligomers up to (C8H8)27 were detected and spectra that approximately reproduced the oligomers’ distribution were obtained. These workers also measured quantitatively temperature dependencies of rates of desorption and activation energies of evaporation rates for many of the ionic species.

Evaporation studies were carried out for both a chemical ionisation and an EI ion source and in each case the products were extracted and determined by a computer-controlled quadrupole MS.

Figure 1.16 shows EI mass spectra. Numbers above the selected ions indicate the temperature at which counting rates of 550 counts/ms were observed. The numbers inside the parentheses give the number of the styrene monomer units and the respective masses of the ion. Temperatures of desorption were arbitrarily defined in terms of 550 counts/ms which corresponds to approximately two-thirds of the maximum intensity of the relatively low abundance ions in Figure 1.16a curve II. The EI data in Figure 1.16a are divided into two molecular weight regions which overlap at the trimer ion at m/z = 370. The solid lines in the figure are used to indicate ions that consist of styrene monomer units and the butyl initiator, with masses given by the relationship m = 104n + 58. These ions are designated as the A oligomer sequence. Ions with values of n ranging from 1 to 11 are found in the EI mass spectrum. The A sequence constitutes the only ions observed above m/z = 780. The dashed lines represent ions with masses that do not fit the A sequence, produced by thermal surface reactions and/or EI decompositions. Sequential mass distributions can also be found in the dashed line spectra. For example, there is a set of ions with masses 291, 395, 499, 603, and 707, which differ by the mass of a styrene monomer unit.

Temperatures of desorption shown in Table 1.4 show a gradual increase for A oligomer ions with increasing molecular weight. The observation of lower molecular weight ions at lower temperatures and during earlier stages in the desorption is inconsistent with the hypothesis that these ions are EI fragmentation products. Temperatures of desorption establish the lower molecular weight ions as primarily products of thermal depolymerisation reactions. Low molecular weight fragment ions not part of the A sequence with masses at 325 and 351 are included in Table 1.4.

30


10,000

Curve 1

352

(3; 3

70)35

3 (2

; 266

)(2; 1

62)

63365

0

(a)

1,000

100

10

200 300Mass (amu)

Rel

ativ

e in

tens

ity

4000

100

50

500 1,00Mass (amu)

Rel

ativ

e in

tens

ity

Curve 2

352

(3; 3

70)

362

(4; 4

74)

388

(5; 5

78)

410

(6; 6

82)

425

(7; 7

86)

448

(8; 8

90)

465

(9; 9

94)

478

(10;

109

8)

505

(11;

120

2)(a)

0

Figure 1.16 Electroionisation and chemical ionisation spectra of polystyrene (a) Partial EI spectrum of polystyrene. The solid lines are the trimer spectrum and the numbers in parenthesis give the first number of monomer units present and second the assigned monoisotropic mass. The numbers in parenthesis give the evaporation temperature in kelvin. The dashed lines are fragment sequence peaks. Reproduced with permission from H.R. Udseth, Analytical Chemistry, 1981, 53, 29. © 1981,

American Chemical Society [18]

31


Table 1.4 Evaporation temperatures and activation energies for some of the peaks in the EI spectrum

Mass (amu) Activation energy (kcal/mol*)

Temperature (°C)

266 51 80

325 17 377

351 17 360

370 41 79

474 36 89

578 27 115

682 26 137

786 27 152

890 29 175

994 21 192

1098 18 205

1202 18 232

*: Estimated uncertainty ± 3 kcal/mol amu: Atomic mass unit Reproduced with permission from H.R. Udseth and L. Friedman, Analytical Chemistry, 1981, 53, 29. © 1981, American Chemical Society [18]

These ions, indicated by dashed lines in Figure 1.16a appear late in the desorption process and at high enough temperatures to be either products of pyrolysis or EI decompositions of the highest molecular weight A sequence oligomers detected. With the exception of these fragment ions, the correlation of increasing mass and temperature of desorption of the A sequence is clearly shown in Table 1.4.

The second column in Table 1.4 gives activation energies for the rate-limiting processes responsible for the generation of the respective ions in the EI spectrum. These activation energies are calculated from slopes of linear plots of the logarithm of relative ion intensity versus the reciprocal of the absolute temperature.

Activation energies in Table 1.4 for A sequence oligomers are higher for the n = 2 and n = 3 oligomers, relatively constant for n = 4–8, and lowest for n = 9–11.

Various other workers have applied MS to PS stability studies [19–21].

32


Zuev and co-workers [22, 23] have carried out several studies on the application of Py-GC-MS to PS stability studies covering poly(4-n alkyl styrenes) [23] and a range of para substituted PS containing ether electron donating groups (NH2, CH3) or electron attracting groups (NO2, Cl, Br) [22]. In their investigation of the thermal degradation of poly(4-n-alkyl styrenes) containing alky groups from hexyl to decyl. Zuev and co-workers [23] used Py-GC-MS at 600 °C to analyse 4-n alkyl styrene monomers (toluene and so on and substituted styrene), alkanes, alkenes, aromatic light products and heavy products, such as dimers.

They found that thermal degradation of poly(4-n-alkyl styrenes) followed mainly a free radical depolymerisation mechanism. The main product is a monomer similar to unsubstituted PS, i.e., 59% to 92% monomer from poly(4-n alkyl styrenes) ranging from Mw 136,500–737,000 and Mn = 37,000–99,000. The amounts of this monomer decrease with increasing length of alkyl sidechain from hexyl to decyl. This behaviour is connected with the stability of monomer under isothermal pyrolysis conditions at 600 °C.

Zuev and co-workers [23] conclude that the thermodestruction of poly(4-n-alkyl styrenes) is mainly similar to the thermodestruction of PS itself, i.e., formation of monomers. The process is not complicated by the formation of any cyclical polyaromatic structure.

In further work, Zuev and co-workers [22] consider the thermal degradation of para substituted PS containing electron donating (NH2, CH3) and electron attracting (NO2, Cl, Br), i.e., CH2OHC6H4 substituents under isothermal heating conditions using Py-GC-MS. They also used TGA under dynamic conditions.

The pyrolysis of these substituted PS gives monomer as the main product for all polymers (60–80 wt%). The other distribution of products supports the view that the thermodestruction of these polymers starts from a random chain scission. The main process for all substituted PS is depolymerisation, similar to unsubstituted PS. In the case of para-substituted PS a good linear dependence was found between Tmax on the thermogravimetry curve and the Hammett constants of the substituents (Tmax = 403.5 – 67.487 σx). The results indicate that the Hammett relationship can describe quantitatively the trends in Tmax and thus thermostability of substituted PS and that thermostability of these polymers depends only on the electronic nature of substituents and their possibility to stabilise macroradicals forming on chain scission.

The aims of this work were four-fold:

1. To establish the temperature ranges in which thermal transformations of these polymers take place.

33


2. To compare the thermal stability of PS and substituted PS and to determine the order of thermal stability in the series.

3. To carry out initial investigations of the mechanism of thermal transformation of these polymers.

4. To establish new relationships which make it possible to develop general concepts of the mechanism of thermal degradations of polymers.

The light pyrolysis products obtained for each substituted PS are shown in Table 1.5.

Table 1.5 The light products in thermodestruction of SPS-X (wt%)

Product NH2 CH3 Cl Br NO2

Benzene Trace Trace Trace Trace 0.03

Toluene 3.95 Trace Trace Trace 0.82

Styrene 2.90 0.17 Trace Trace 0.47

Substituted toluene

2.69 2.58 1.26 5.56 0.57

Substituted benzene

Trace 0.32 1.28 1.08

α-Methylstyrene 1.44 - -

Substituted α-methylstyrene

3.09 0.71 0.73 2.91 1.07

Monomer 56.40 78.29 79.93 68.91 81.78

Dimer 4.13 5.28 2.34

Substituted p-ethylbenzene

5.47 1.85 2.67

Reproduced with permission from: V.V. Zuev, F. Bertini and G. Audisio, Polymer Degradation and Stability, 2001, 71, 213. © 2001, Elsevier [22]

Similar analyses were performed for heavy products of amino, methyl and nitro substituted PS. The dimer CH2 = CAr – CH2–CH2Ar predominated in all cases – 5.3% of pyrolysate for methyl polystyrene, 4.1% for amino polystyrene and 1.5% for nitro polystyrene.

34


Roland and co-workers [24] carried out studies of the thermal stability of nitroxide capped PS using TGA and Py-GC-MS.

Other studies of the degradation of PS [25–27] include an investigation of photolytic degradation [27].

1.3.2 Polystyrene Copolymers

1.3.2.1 Styrene Acrylonitrile

Igarashi and Kambe [28] have studied, by TGA methods, the effects of molecular weight and composition on the thermal stability of a styrene–acrylonitrile copolymer in nitrogen. Two series of copolymers were used.

From TGA curves for a series of copolymers of varying molecular weights it was shown that at about 270 °C, a marked weight loss begins for each copolymer and the temperature corresponding to the maximum rate is almost identical for each sample. The weight loss of each copolymer is completed at about 390 °C. Thus, it appears that molecular weight has little effect on the thermal stability of the styrene–acrylonitrile copolymers studied when the acrylonitrile content is maintained constant for each sample.

For another series of copolymers of varying acrylonitrile composition it can be observed that the copolymers decompose in one stage over a temperature range of about 360–400 °C and that the position of a peak shifted to lower temperatures and the residue of degradation increased as the acrylonitrile content was increased (Figure 1.17). Polyacrylonitrile homopolymer is much more thermally stable than the styrene homopolymer. Thus, at about 390 °C, the PS polymer degrades completely and leaves no residue, whereas at the same temperature, the polyacrylonitrile polymer leaves a residue of about 96% of the original sample weight. This explains why the residue and degradation rate of the copolymer increases and decreases, respectively, as the acrylonitrile content is increased. The thermal stability of polyacrylonitrile may be attributed to intramolecular cyclisation of the CN groups [29, 30]. However, it must be noted, at this point, that a correlation between thermal stability and acrylonitrile content may not necessarily be obtained if the average block length of the acrylonitrile in the copolymer does not increase much with increasing acrylonitrile content. Thus, as the acrylonitrile content was increased from 24 to 35 wt%, the average block length (based on theoretical calculations) only increased from 1.04 to 1.19 and the decomposition residue increased from about 5% to 7%. However, when the

35


acrylonitrile content was raised to 57%, the average acrylonitrile block length rose to a value of 2.64 and the decomposition residue increased to 25%.

4

3

2

1

0 0

20

40

60

Residual w

eight, %

80

100

250 300 350Temperature, °C

Wei

ght l

oss,

%/°

C

400

Figure 1.17 Thermogravimetric analysis curves for styrene-acrylonitrile copolymers of different content of acrylonitrile: ( ) 11.1 wt%, ( ) 24.0 wt%, ( ) 35 wt%, ( ) 57 wt%. Reproduced with permission from S. Irgashi and H. Kambe, Die

Makromolekulare Chemie, 1964, 79, 180. © 1964, Wiley [28]

1.3.2.2 Styrene–divinylbenzene

Two studies have been conducted on this copolymer [31, 32].

Nagagawa and co-workers [32] investigated the thermal degradation behaviour of anionically and radically polymerised styrene divinylbenzene (DVB) copolymer.

36


Investigations by Py-GC showed that these two types of copolymer differed in network structure as depicted in Figure 1.18.

C C C C C C C C C C C C C C C C C~ ~



C C C C C C C C C ~

C C C C C C C C

C

C C C C C C C C~ ~

C C C C C C C C C C C C C C C C~ ~

a

b

Figure 1.18 Expected network structures for (a) anionically copolymerised and (b) radically copolymerised St-DVB copolymers. Reproduced with permission from H. Nakagawa, Y. Matsushita and S. Tsuge, Polymer, 1987, 28, 1512. © 1987,

Elsevier [32]

37


Typical TGA weight loss curves of the anionically copolymerised gels and the corresponding radically copolymerised ones are shown in Figure 1.19. The fact that the weight loss curves generally shift to the higher temperature with increasing DVB content suggests that the overall thermal stability of the copolymers increases together with the DVB content. Although the amount of residue also increases with the rise of the DVB content, there are no significant differences between the weight loss curves of the anionically and corresponding radically copolymerised gels for the samples with lower DVB contents. However, the initial weight loss for the radically copolymerised sample occurs at a considerably higher temperature than that of the corresponding anionically copolymerised sample. From this result, it is apparent that the overall thermal stability of the radically copolymerised gels with the higher DVB content is higher than that for the corresponding anionically copolymerised ones since the randomly distributed network structures in the radically copolymerised gels contribute to the higher thermal stability than the locally distributed ones.

100

R–1A–1

R–4

R–3

A–4

A–3

50

200 300 400Temperature (°C)

Wei

ght l

oss (

%)

500 600 700

Figure 1.19 TGA weight loss curves of anionically and radically copolymerised St-DVB copolymers. R = radically copolymerised, R–1 = 89.1% styrene,

R–4 = 37.7% styrene, A = anionically copolymerised, A–1 = 89% styrene, A–4 = 35.1% styrene. Reproduced with permission from H. Nakagawa,

Y. Matsushita and S. Tsuge, Polymer, 1987, 28, 1512. © 1987, Elsevier [32]

1.3.2.3 Styrene-Isoprene (Kraton 1107)

Factor analysis has been used for extracting information from mass spectra recorded in a TG-MS analysis of the styrene-isoprene mixture. Statheropoulos and co-workers

38


[33] used principal component analysis and a special diagram, the contour variance diagram (ContVarDia), for performing the factor analysis. The method was applied for studying the thermal decomposition of Kraton 1107 copolymer. Py-GC-MS was used for identification of the pyrolysis products of Kraton 1107 (see Figure 1.20) [33]. The application of factor analysis resulted in the determination of the main thermal decomposition steps and the prediction of the mass spectrum corresponding to each step. Those mass spectra were either pure spectra corresponding to the main evolved gases or average spectra corresponding to multiple gases evolved in one decomposition step. The advantages and the limitations of the chemometric approach were discussed. The substances identified are listed in Table 1.6.

1.14

2.46

4.11

4.87

6.34

6.98

8.58

x106 (a)

(b)

(d)

(c)

(e)(g)

(f)

7

8

6

Abu

ndan

ce(a

rbitr

ary

units

)

5

4

3

2

1

05.00Time (min) –> 5.00 5.00 5.00 5.00 5.00 5.00

Figure 1.20 Py-GC-MS chromatogram of Kraton 1107 copolymer (a) isoprene, (b) methyl-benzene, (c) isomers of dimethyl benzene, (d)styrene, (e) 1,4-dimethyl-

4-ethenyl-cyclohexene, (f) isopropenyl-benzene and (g) 1-methyl-4-(1-methylethenyl)-cyclohexene. Reproduced with permission from M. Statheropoulos,

K. Mikedi, N. Tzamtzis and A. Pappa, Analytica Chimica Acta, 2002, 461, 215. © 2002, Elsevier [32]

The first decomposition step of Kraton 1107 seems to consist of four overlapped decomposition steps. In the first step, 1-methyl-4-(1-methylethenyl)-cyclohexene and 1,4-dimethyl-4-ethenyl-cyclohexene start to evolve. In the second step, the evolution of isoprene monomer is taking place. In the third step, a mixture of alkyl-benzene, cycloalcadienes, cycloalkenes and saturated polycyclic hydrocarbons evolves whilst

39


the evolution of styrene monomer (fourth step) is starting later when the pyrolysis of polyisoprene has already reached its maximum.

Table 1.6 Substances identified by the electronic library (Wiley) in the Py-GC-MS analysis of Kraton 1107 with the eight most abundant masses and their

relative intensities according to the Eight Peak Index (Mass Spectrometry Data Center)

Retention time (min)

Substance Eight most abundant masses (relative intensities)

1.14 Isoprene (2-methyl-1,3-butadiene) 67, 53, 68, 39, 27, 40, 41, 42 (100, 86, 83, 71, 48, 45, 40, 22)

2.46 Methyl-benzene 91, 92, 65, 39, 63, 51, 45, 89 (100, 60, 12, 9, 6, 6, 5, 3)

4.11 Isomers of dimethyl-benzene -

4.87 Styrene (ethenylbenzene) 104, 103, 78, 51, 77, 105, 50, 52 (100, 45, 32, 21, 17, 9, 7, 7)

6.34 1,4-Dimethyl-4-ethenyl-cyclohexene 68, 67, 93, 107, 136, 121, 53, 41 (100, 35, 29, 22, 17, 15, 15, 15)

6.98 α-Methylstyrene (isopropenyl-benzene)

118, 117, 103, 78, 51, 77, 39, 91 (100, 64, 58, 37, 29, 26, 25, 22)

8.58 1-Methyl-4-(1-methylethenyl)-cyclohexene

68, 93, 67, 39, 41, 27, 53, 79 (100, 47, 42, 33, 32, 28, 23, 22)

Reproduced with permission from M. Statheropoulos, K. Mikedi, N. Tzamtzis and A. Pappa, Analytica Chimica Acta, 2002, 461, 215. ©2002, Elsevier [33]

1.3.2.4 Miscellaneous Copolymers

Thermal copolymer studies include work on styrene–ethylene [34–36], styrene–vinyl cyclohexane [37] and propene–pentene [38].

40


1.4 Carbocyclic Polymers

Among these compounds containing anthracene, benzene, naphthalene and other aromatic compounds are of particular interest.

The backbone of polycyclic molecules may involve either only rings (polyphenylenes) or additionally, aliphatic groups, e.g., [poly(p-xylylene)].

The carbocyclic polymers include condensed systems (polycyclopentadiene or polybutadiene) [39] and also polyphenyl. Studies on the thermal degradation of these two polymers have shown that they have almost the same thermal stability (the temperature at which 50% of the polymer weight is lost (T50) of polyphenyl is 432 °C, while T50 of poly(p-xylylene) is 430 °C) and they decompose completely over the temperature range 410–470 °C. The MS analysis of volatile products of the degradation of polyphenyl demonstrates the presence of toluene (5.9% mass), benzene (1.4% mass) and xylene (0.1% mass). The thermal degradation of poly(p-xylylene) results in the liberation of such products as xylene (2.83% mass), toluene (0.29% mass), methylethylbenzene (0.28% mass), methylstyrene (0.14% mass) and benzene (0.06% mass). The absence of monomer, i.e., p-xylene, in the volatile products is probably associated with its high reactivity and ability to undergo polymerisation on leaving the heating zone.

The methylene bonds are the weakest in the poly(p-xylene) macrochain. The cleavage of such bonds causes the formation of two macroradicals, which should decompose via a chain mechanism to yield monomer:

~CH2– –CH2–CH2– –CH2~

~CH2– –CH2+ CH2– –CH2~

CH2 CH2, etc. (1.15)

The activation energy of the thermal degradation of polyphenyl and poly(p-xylylene) as determined by their maximum decomposition rates are 209 and 305.1 kJ/mol, respectively.

41


It has been established [39] that polyacenaphthylene which is produced on polymerisation of the monomer (125 °C) begins to decompose at temperatures above 297 °C to form the initial monomer:

~ CH–CH2 ~

(1.16)

Polyarylenequinones produced by the action of p-benzoquinone on different bis-dinitrogenated aromatic diamines are stable in an inert atmosphere up to 600 °C and in air to 450 °C. The copolymer of anthracene and styrene which at 305 °C loses only 5.5% of its initial mass after four hours is also a thermally resistant polymer [39].

The previous data demonstrate that the introduction of benzene rings into the vinyl polymer chain increases its thermal stability. The nature of the bridging groups, which connect the aryl groups in polymers also plays a key role in their thermal stability. The thermal stability of polymers containing various bridging groups between the aromatic rings decreases as follows: CO > CH2 > O > O2 [16].

References

1. S.S. Pavlova, I.V. Zhuraleva and Y.I. Tolchinskii, Thermal Analysis of Organic and High Molecular Weight Compounds, Khimiya, Moscow, Russia, 1983. [In Russian].

2. D.A. Anderson and E.S. Freeman, Journal of Polymer Science, 1961, 54, 253.

3. S. Igarashi and H. Kambe, Bulletin of the Chemical Society of Japan, 1964, 37, 176.

4. L.A. Wall and S. Straus, Journal of Polymer Science, 1960, 44, 313.

5. S.L. Madorsky, SPE Journal, 1962, 18, 1482.

6. W.J. Bailey and C.L. Liotta, Polymer Preprints, 1964, 5, 333.

7. H.H.G. Jellinek, Journal of Polymer Science, 1949, 4, 13.

8. W. Camacho and S. Karlsson, Polymer Degradation and Stability, 2002, 78, 385.

42


9. I.A. Hussain, K. Ho, S.K. Goyal, E. Karbashewski and M.C. Williams, Polymer Degradation and Stability, 2000, 68, 381.

10. Z. Wang, G. Wu, Y. Hu, Y. Ding, K. Hu and W. Fan, Polymer Degradation and Stability, 2002, 77, 427.

11. E.H. El-Gendy and I.A. El-Shanshoury, Journal of Applied Polymer Science, 2004, 92, 3710.

12. J.M. Gomez-Elvira, P. Tiemblo, M. Elvira, L. Matisova-Rychla and J. Rychly, Polymer Degradation and Stability, 2004, 85, 873.

13. L. Brambilla, G. Consolati, R. Gallo, F. Quasso and F. Severini, Polymer, 2003, 44, 1041.

14. C.S. Own, D. Seader, N.A. D’Souza and W. Brostow, Polymer Composites, 1988, 19, 107.

15. Q. Yuan, W. Jiang, L. An and R.K.Y Li, Polymers for Advanced Technologies, 2004, 15, 409.

16. J-Y. Li, D-Y. Yana, Y. Mub, Y-F. Yaoc and Q. Chenc, Chinese Journal of Polymer Science, 2003, 21, 347.

17. V.V. Korshak, Chemical Structure and Temperature Characterisation of Polymers, Nauka, Moscow, 1970. [In Russian]

18. H.R. Udseth and L. Friedman, Analytical Chemistry, 1981, 53, 29.

19. S.G. Coloff and N.E. Vanderb0rgh, Analytical Chemistry, 1973, 45, 1507.

20. S. Foti, A. Liguori, P. Maravigna and G. Montaudo, Analytical Chemistry, 1982, 54, 674.

21. T.H. Risby, J.A. Yergey and J.J. Scocca, Analytical Chemistry, 1982, 54, 2228.

22. V.V. Zuev, F. Bertini and G. Audisio, Polymer Degradation and Stability, 2001, 71, 213.

23. V.V. Zuev, F. Bertini and G. Audisio, Polymer Degradation and Stability, 2000, 69, 169.

24. A.I. Roland, M. Stenzel and G. Schmidt-Naake, Die Angewandte Makromolekulare Chemie, 1998, 259, 69.

43


25. A. Basch and M. Lewin, Journal of Polymer Science: Polymer Chemistry Edition, 1973, 11, 3095.

26. M.M. Krauze, J. Trzeszczynski and M. Dzieciol, Popular Plastics and Packaging, 2004, 49, 100.

27. N. Grassie and N.A. Weir, Journal of Applied Polymer Science, 1965, 9, 975.

28. S. Igarashi and H. Kambe, Die Makromolekulare Chemie, 1964, 79, 180.

29. S. Straus and S.L. Madorsky, Journal of Research of the National Bureau of Standards, 1958, 61, 77.

30. N. Grassie and I.C. McNeill, Journal of Polymer Science, 1958, 27, 207.

31. P.E. Tulupov and O.N. Karpov, Zhurnal Fizicheskoi Khimii, 1973, 47, 1420.

32. A. Nakagawa, M. Matsushita and S. Tsuge, Polymer, 1987, 28, 1512.

33. M. Statheropoulos, K. Mikedi, N. Tzamtzis and A. Pappa, Analytica Chimica Acta, 2002, 461, 215.

34. T-L. Chang and T.E. Mead, Analytical Chemistry, 1971, 43, 534.

35. B.V. Kokta, J.L. Valade and W.N. Martin, Journal of Applied Polymer Science, 1973, 17, 1.

36. M.D. Judd and A.C. Norris, Journal of Thermal Analysis and Calorimetry, 1973, 5, 179.

37. V.G. Zaikin, R.G. Mardanov, V.I. Kleiner, B.A. Krentsel and N.A. Plate, Journal of Analytical Applied Pyrolysis, 1990, 17, 291.

38. G. Costa, P. Stagnaro, V. Trefiletta, M.C. Sacchi, F. Forlini G.C. Alfonso, I. Tincul and U.M. Wahner, Macromolecular Chemistry and Physics, 2004, 205, 383.

39. V.V. Korshak, Thermostable Polymers, Nauka, Moscow, Russia, 1969. [In Russian]

44


Documents

1 Carbon Hydrogen Polymers - Smithers Rapra Producing Alkanes Dienes C H 2 H 2 CC 2 (1.1) 1 Carbon Hydrogen Polymers. 2 Thermal Stability of Polymers ... A mechanism [1] involving