BAJAJ PLAST
Plast UV stabilizer
Overview
Overviewing
UV Stabilizer
Introduction
Polymer photo-oxidation is the result of the combined action of light and oxygen. The action of sunlight in the presence of the air’s oxygen is by far the most important example. As a rule, thermal oxidation is superimposed on photo-oxidation.
The most visible result of these phenomena is the deterioration of the appearance of materials. At the same time however, mechanical and physico-chemical properties are also altered.
Because many synthetic resins suffer from photo-oxidation, the investigation of ways to inhibit or, at least retard this type of degradation has been a major effort of the industry. As a result of these investigations, specialty chemicals called light stabilizers or UV stabilizers were developed, which interfere with the physical and chemical processes of light-induced polymer degradation.
The most important UV stabilizer classes are 2-hydroxybenzophenones, 2-hydroxyphenyl-benzotriazoles, organic nickel compounds, and sterically hindered amines (HALS). Salicylates, cinnamate derivatives, resorcinol monobenzoates, oxanilides, and p-hydroxy-benzoates are also used, although to a lesser extent. The hindered amines are the latest development in this field. In numerous applications, they show considerable superiority over the other light stabilizer classes available [1, 2, 3].
The protection of plastics from the effects of light may also be achieved through the addition of carbon black [4] and other pigments [5]. Nevertheless, in this chapter, only organic and organometallic compounds producing only slight discoloration or no discolora-tion at all in plastics are considered typical light stabilizers. Carbon black and other pigments are discussed mainly as colorants. However, the interaction of carbon black and other pigments with classical light stabilizers is also treated in this chapter.
The data in Table 2.1 show the economic importance of light stabilizers in different areas of the world [6]. Table 2.2 shows the importance of the main light stabilizer types in these areas [6]. It can be seen that the volume of HALS is approximately the same as that of the benzotriazole and benzophenone UV absorbers combined. The relative amounts of UV stabilizers consumed in major thermoplastics are presented in Table 2.3. In 1996, almost 75% of the light stabilizers produced were used with polyolefins. Among polyolefins, the amount used with PP surpasses the amount used with all the different PE types combined.
Photodegradation of Synthetic Polymers
Ultra-Violet Spectrum of Sunlight
Radiation from the sun reaching the outer layers of the earth’s atmosphere shows a continuous energy spectrum in the wavelength region between 0.7 and approximately 3000mm. On passing through the atmosphere, part of the long wavelength radiation is absorbed by water vapor and carbon dioxide. Finally, only the short wavelength part of the infra-red radiation reaches the earth’s surface.
The short wavelength UV radiation below 175 nm is absorbed by oxygen in the atmospheric layers more than 100km above the surface. The radiation between 175 and 290 is absorbed by the ozone layer of the stratosphere, which begins at about 15km above sea level for mean latitudes and has a maximum density between 25 and 30km. It is the remaining UV part of sunlight, i.e., radiation between 290 and 400 nm, that initiates the degradation of the plastics on outdoor weathering.
In addition to the absorption of part of the UV radiation by ozone, the scattering of sunlight on interaction with air molecules and aerosol particles (water droplets, dust) in the atmosphere has to be considered. As a consequence, the radiation responsible for the aging of plastics, i.e., the radiation reaching the surface of the earth, is composed of direct sunlight (“solar radiation”) and scattered light (“sky radiation”). The scattered light can be calculated according to Rayleigh’s law. If I_0 is the intensity of the incident light and I is the total intensity of the light scattered in all the directions.
n addition to UV radiation, changes in temperature and humidity with climatic zone have to be considered. In fact, the relative importance of thermal oxidative degradation, which is always superimposed onto photo-oxidation, increases with temperature and can become dominant. Water may affect the course of degradation by extractive or hydrolytic processes as well as by participating in photochemical reactions involving some pigments, such as titanium dioxide.
In connection with the optical properties of the atmosphere, different emissions resulting from human activities should be taken into account. On the one hand, photochemical reactions lead to the formation of various oxidants (smog), especially ozone, in the troposphere [16]. On the other hand, they may even disrupt the ozone level of the stratosphere and as a consequence, lead to an increase in the short wavelength UV radiation reaching the earth [17, 18, 19]. The potential role of stratospheric ozone depletion on polymer photodegradation is a growing concern [20, 21]. The direct role of various air pollutants in initiating photo-oxidation is examined in detail later in this chapter.
Physico-Chemical Processes Occurring on Light Absorption
Incident light is either reflected from the surface of a polymer or scattered or absorbed in the bulk of the polymer. According to the first law of photochemistry (Grotthus – Draper), only the part of the light effectively absorbed leads to photochemical transformations, i.e., to degradation. The second law of photochemistry (Stark-Einstein), completed this explanation by stating that the absorption of light by a molecule is a one-quantum process, so that the sum of the primary process quantum yields,P hi, must be unity.
The absorption of light by synthetic resins is first of all related to their structure. It is well known that saturated hydrocarbons do not absorb light with wavelengths greater than 250 text nm. In the presence of double bonds (chromophores), absorption is shifted to longer wavelengths. This is particularly pronounced for double bonds between carbon atoms and heteroatoms; for example, carbonyl compounds absorb already in the wavelength region above 290 text nm.
Light of wavelengths above 290nm may, nevertheless, induce the degradation of polymers such as polyolefins, which, because of their structures do not absorb such wavelengths directly. This is caused by traces of impurities resulting from manufacturing or by structural irregularities often present in technical polymers, e.g., catalyst residues or oxidation products. The last may show pronounced absorption in the UV range and lead to photo-chemical transformations. Moreover, in semi-crystalline polymers, because of light scattering by the crystalline particles [22, 23], the light path is considerably increased in comparison with amorphous materials. Thus, even at low concentrations of chromophoric groups, appreciable quantities of energy may be absorbed.
Polymer degradation is wavelength dependent. Apart form the fact that the different UV wavelengths in sunlight are not equally effective, they may even produce different types of values, in comparison with those observed in the photochemistry of small molecules, result from the fact that in plastics the energy initially locally absorbed can be distributed in nearby domains. Moreover, in solid polymers, the cage effect very often induces the recombination of the broken chain [31]. The same effects cause considerable reduction of the free radical formation by photolysis of ketones in PE.degradation. Thus, the wavelength sensitivity of plastics is important for an understanding of the factors that determine weathering.
The activation spectrum of a polymer represents its relative sensitivities to the wavelengths of a specific radiation source based on measurement of a single criterion of degradation. The emission characteristics of the light source, i.e., the wavelengths and their intensities, determine to a large extent the effect of radiation on a material. The activation spectrum is specific to the type of source because it determines the response of the material not corrected for differences in the intensity at different wavelengths [24]. Normalization of the data to equal intensities is generally not valid because degradation does not vary linearly with light intensity. Typical activation spectra for PE film obtained with an unfiltered xenon arc are presented in Fig. 2.3 [24]. The important role of the test criterion is quite obvious. The activa-tion spectra maxima are at 310 and 340,nm for yellowing and carbonyl buildup, respectively.
The spectral sensitivity of LDPE and PP with respect to photo-oxidation is compared in Fig. 2.4 [25]. The data have been normalized to an energy of 1,-2. It can be seen that for wavelengths below 330,nm PP is more sensitive than LDPE, whereas at wavelengths above 330,nm, LDPE is more sensitive. The activation spectra and/or the spectral sensitivity of polymers other than polyolefins have also been determined. Polyamide, poly- (phenylene sulfide), and polycarbonate among others have been examined [25, 26, 27]. Recently the wavelength sensitivity of PE has been examined on both artificial and natural weathering [28]. It has been found that xenon arcs and sunlight give similar results. Moreover, the spectral region 300-350 caused the greatest degradation at the beginning of artificial exposure. However, after 68,text, exposure in the region of 335-360, was found to be the most effective in this respect [28].
From an energetic point of view, the sensitivity of some polymers towards light is not astonishing. In Table 2.5, it can be seen that the energies of the photons found in the ultraviolet region between 290 and 400,nm range from 419 to 097\,{kJ}\,{Einstein}^{-1} (1,{Einstein} = 1,{mole} photons = 6.10^23, photons). They are significantly higher than the bond energies of some typical bonds found in synthetic resins. However, the energy must at first be absorbed by a group and then transferred to the bond to be scissioned.
After absorption of photons by plastics, the functional groups involved move to a higher energy level (excited state) [29, 30]. This is shown schematically in Fig. 2.5. Through spin inversion, singlet states may convert to triplet states (intersystem crossing). The excited states are able to lose the energy initially received in several processes (fluorescence, phosphorescence, and radiationless decay). An additional possibility is the energy transfer in which the energy of the excited state (singlet, triplet) is transferred to a suitable acceptor molecule. Considering the time scale of all these processes, triplet states are relatively long-lived. Therefore, they may become chemically active and thus, play an important role in polymer photodegradation.
The ratio of the polymer molecules reacting chemically to the number of photons absorbed is designated as quantum yield [29]. For some plastics the quantum yields for chain scissions [29] are between 10^{-2} and 10^{-3}. This means that among 100 to 100,000 molecules that in fact have absorbed light, only one polymer molecule really degrades. These quantum yield values, in comparison with those observed in the photochemistry of small molecules, result from the fact that in plastics the energy initially locally absorbed can be distributed in nearby domains. Moreover, in solid polymers, the cage effect very often induces the recombination of the broken chain [31]. The same effects cause considerable reduction of the free radical formation by photolysis of ketones in PE.
Photo-oxidation Scheme
In polymers exposed to light, free radicals are formed as a consequence of the excitation of absorbing functional groups in the polymer. This is a function of the energy of the light and of the structure of the polymer molecule. The recombination of the radicals that escape the primary cage is very slow because they have to find each other through diffusion in the polymer matrix. In the presence of oxygen, the radicals simultaneously oxidize (photo-oxidation). Thus, it is often difficult to distinguish the pure photochemical processes from the superimposed thermal processes (oxidation). Some of the fundamental mechanisms involved have been elucidated [3, 29] only recently through systematic studies involving technical plastics.
The mechanism developed initially for the thermal oxidation of rubber [33, 34, 35] can be applied to other substrates and also to photo-oxidation. Scheme 2.1 illustrates a possible reaction sequence. Some of the reactions involved are still a matter of discussion. Recently, the role of air pollutants in initiation of polymer degradation has been stressed [36-40]. The additional reactions involved are shown in Scheme 2.2.
Photo-oxidation of Polyolefins
Photodegradation of polyolefins has been attributed mainly to the presence of catalyst residues, hydroperoxide groups, and carbonyl groups introduced during polymer manufacturing , processing , or storage [41]. All these species absorb UV light above 290, nm to varying degrees and can participate in numerous photochemical reactions. Table 2.7 summarizes data about some PP chromophores [41-44].
The high initiating capability of hydroperoxides and probably also of peroxides essentially results from the high quantum yields (approximately 1) with which they decompose into free radicals (see Reactions 4a, 4b, and 5 in Scheme 2.1).
Carbonyl groups seem to have only a small initiating efficiency [32, 41]. The free radicals formed initially in the Norrish type I reaction (see Reactions 21a and 21b in Scheme 2.3) can recombine obviously before any further reaction. The products from the Norrish type II reaction (see Reaction 22 in Scheme 2.3) are inefficient photo initiators. It has been found that there is a considerable difference between aldehydes and ketones with respect to photolysis in polymers such as PE [32]. Thus, aldehydes are photolyzed mainly according With respect to initiation of photo-oxidation by carbonyl groups, alpha, beta-unsaturated carbonyl groups have been given special attention. However, their importance in photoinitiation of polyolefins is still a matter of discussion. The first proposals [45, 46] have found some acceptance [47, 48] but no unequivocal experimental evidence. As a matter of fact, some results lead to the conclusion that conjugated carbonyl compounds are photostabilizers [49, 50]. This effect has been attributed to their ability to undergo photoisomerization in a non-radical process essentially fluorescing ,unsaturated carbonyl groups, contribute to the photo-oxidation induction period, i.e., that they exert a stabilizing influence. This finding has been substantiated by other experiments [52]. In fact, hexane extraction of the polymer, which removes reduction of the induction period of PP photo-oxidation. On the other hand, addition of unsaturated groups with the structure of alpha, beta-unsaturated carbonyl compounds, leads to a hexane extracts increases the induction time markedly. In the last mentioned experiment, the stabilizing effect of the double bonds dominates the sensitizing effect of the additional atactic material introduced with the hexane extract. This is in contradiction with older compounds. results [53] that had pointed to an extraction of sensitizing polynuclear aromatic compound.
The results of the experiments involving bromination can be considered dubious because the absence of the induction time may well be due to initiation by photolysis of the brominated compounds. Isomerization reactions such as Reaction 23 (Scheme 2.4) cannot proceed with all types of , unsaturated ketones likely to be present in polyolefins. They are likely, at least in part, for ketone groups in position to vinylidene (in PE) or ethylidene (in PP) groups. However, ketone groups vicinal to vinyl or vinylidene groups do not show such a reaction. Moreover, terminal , unsaturated groups can be considered as more initiating than similar groups in the chain because on photolysis, according to the Norrish type I process, one of the radicals formed has a low molecular mass. Then it can diffuse more easily so that recombination is less likely than when both radicals formed have a high molecular mass.
The discussion above shows that there is no conclusive evidence either for an initiating role or for a stabilizing role of $\alpha, \beta$-unsaturated carbonyl compounds. In fact, both aspects may show up, their relative importance depending on the polymer and external conditions, such as UV light and temperature. However, effective stabilization could be expected only if the compounds were present at levels much higher than those normally encountered.
Some work [54] shows carbonyl compounds to be involved in the formation of singlet oxygen through energy transfer from their excited triplet state to oxygen in the ground state ((see Reaction 24 in Scheme 2.5). The lifetime of singlet oxygen in polymers is long enough to allow its reaction with double bonds, leading to the formation of allylic hydroperoxides (see Reaction 25 in Scheme 2.5). The possible contribution of allylic hydroperoxides to the photodegradation of polyolefins has been discussed [53, 55].
The formation of singlet oxygen may also be induced by polynuclear aromatics (naphthalene, anthracene) found in PP [56] and by quinones resulting from the destruction of antioxidants and dyestuffs [57].
In connection with the initiation reaction, charge-transfer (CT) complexes between oxygen and PP [58], HDPE [59], and poly-4-methylpentene [60] have been considered. These complexes can absorb light of longer wavelengths than the polymer as such and thus, may become photochemically active (see Table 2.7).
Several results [51, 52, 61] seem to indicate that in the absence of hydroperoxides, charge-transfer complexes between oxygen and PP play a decisive role in the initiation of hydroperoxide formation and therefore, in the initiation of photo-oxidation.
Initiation by CT complexes has been the object of much controversy lately. At first, it was pointed out that a reaction such as Reaction 26 (Scheme 2.6) cannot account for initiation because the hydroperoxides formed in PE are poor initiators of photo-oxidation [63]. Another reaction was proposed, instead, involving a six-member transition state and resulting in the formation of trans-vinylene groups and hydrogen peroxide (Reaction 27 in Scheme 2.6). The latter can decompose directly into hydroxy radicals if it possesses an excess of energy (Reaction 28). Alternatively, it can also yield free radicals on subsequent photolysis (Reaction 30 in Scheme 2.6) [62]. The possible contribution of CT complexes to initiation of photo-oxidation is still a matter of discussion [64, 65]. Considerations of the energy involved with a CT state led to an amendment of the first proposal [66-71]. Instead of CT complexes, corresponding exciplex formation polymer-oxygen has been envisaged. Nevertheless, the reactions and their products were still the same as those outlined in Scheme 2.6. More recently, the potential role of air pollutants in initiation of photo-oxidation has been examined [36-40].
The finding that ozone in ambient air has a determining role in initiating thermal oxidation of polymers such as PP and PE was of prime importance for understanding possible new mechanisms for photo-oxidation. However, initiation of photo-oxidation is more complicated and air pollutants other than ozone may also be involved, such as nitrogen dioxide and sulfur dioxide [36-40]. Comparing the mean monthly concentrations of the main air pollutants leads to the interesting finding that nitric oxide and sulfur dioxide vary considerably during the year, as does the concentration of ozone. Moreover, the concen-trations of nitric oxide and sulfur dioxide are lowest in summer when the concentration of ozone is highest. It is only the mean concentration of nitric dioxide that remains relatively constant through the year. Therefore, initiation of photo-oxidation mainly by photolysis of nitrogen dioxide would be in agreement with the observation of a rather constant rate of photo-oxidation in artificial exposure devices such as the Weather-Ometer Ci 65. The considerable acceleration of degradation under exceptionally high concentrations of ozone could then be attributed to a complete change in mechanism. One such possibility is the passage from bulk initiation to surface initiation [39].
Photolysis of ozone yields a very reactive oxygen atom that can initiate an oxidative chain reaction as shown in Scheme 2.2. The same is valid for photolysis of nitrogen dioxide, also shown in Scheme 2.2. As a matter of fact, over the year, the mean concentrations of nitrogen dioxide in air are significantly higher than the corresponding concentrations of ozone.
Photolysis of ozone is a very effective process. It occurs not only with UV but also with visible light, up to $590\,\text{nm}$, with a quantum yield of 1. The situation is similar for nitric dioxide; photolysis occurs up to about $370\,\text{nm}$ with a quantum yield of 1 and with decreasing quantum yields up to $420\,\text{nm}$. The subsequent reactions of the ground state oxygen atoms are typical hydrogen abstraction reactions as shown in Scheme 2.2. These reactions are limited by recombination of oxygen atoms with oxygen molecules; recombination of ozone with nitric oxide to yield oxygen and nitrogen dioxide has also been included in the limiting reactions in Scheme 2.2. It can also be seen as a neutral reaction, because if one active initiator molecule, ozone, disappears, another one, nitric dioxide, is generated.1
The reaction of peroxy radicals with nitric oxide shown in Scheme 2.2 may have a determining role. It involves the normal chain carriers of polymer oxidation, i.e., the perox2y
radicals, and nitric oxide present as an air pollutant and resulting also from the photolysis of nitrogen dioxide [36-40].
In this way, a new potential initiator of photo-oxidation is produced, without interrupting the normal oxidation chain. In fact, the peroxy radical is transformed into a more reactive alkoxy radical. Through this catalytic cycle, the initiation reaction can itself be a chain reaction, superimposed on the polymer chain oxidation and, at the same time, entangled with it.
The initiation by air pollutants just discussed most likely involves the surface layers of the polymer. Nevertheless, some reactions, especially those based on the catalytic cycle, can be very important for bulk oxidation [36-40]. This could explain the results of the photo-oxidation of LDPE films that show primarily surface oxidation but significant contributions from bulk oxidation as well, as reported previously [72, 73].
It is thought that initiation by air pollutants is only one aspect of initiation of photo-oxidation. Initiation by catalyst residues (Ti) and direct initiation by HALS are additional modes. Furthermore, as already discussed, in PP, initiation by photolysis of tertiary hydroperoxides dominates after an initial period if the HALS concentration is not too high. Moreover, photolysis of hydroperoxides may also be important in PE in the presence of significant amounts of transition metals [70, 71].
The carbon radicals formed in the photochemical primary processes can react very fast with oxygen (Reaction 2 in Scheme 2.1). The alkyl peroxy radicals RO_2bullet formed this way play a key role as chain carriers because they abstract mainly hydrogen from the polymer and thus produce the polymer radicals necessary for the propagation of the oxidation (chain propagation, Reaction 3 in Scheme 2.1). In this respect, it should be noted that the reactivity of a specific substrate is not only a function of its structure, i.e., the ease of hydrogen abstraction. It is also determined by the solubility and the diffusion rate of oxygen. In semi-crystalline plastics, the density of the crystalline phase is usually higher than that of the amorphous phase and almost impermeable to oxygen. Hence, oxidation occurs exclusively in the amorphous material. Poly(4-methyl-1-pentene (PMP)) constitutes an exception in this respect: the densities of the crystalline and amorphous phases are almost identical at 60C; at room temperature, the density of the amorphous phase is even higher than that of the crystalline phase [74]. It has been found that isotactic PMP is more susceptible to oxidation than atactic PMP [75, 76]. This has been attributed to the fact that in isotactic PMP crystals, oxidation takes place preferentially at the chain folds because the tension resulting from the chain configuration reduces the oxidation activation energy [75] and also because the crystal dimensions are such that the crystalline regions are permeable to oxygen. For Poly- (1-butene) too, the activation energy of the oxygen reaction has been found to be lower for the isotactic than for the atactic polymer.
Photo-oxidation of PP
In PP, isolated hydroperoxide groups and sequences of hydroperoxide groups are formed in intermolecular (Reaction 31 in Scheme 2.7) and intramolecular (Reaction sequence 32 in Scheme 2.7) oxidation steps [78]. Therefore, after a short exposure period, during which the chromophores initially present become photochemically active, appreciable quantities of hydroperoxides may be formed (Reactions 31 and 32). The subsequent degradation is determined by the decomposition reactions of these hydroperoxides.
The photolysis of hydroperoxides is generally considered to yield macro alkoxy radicals and hydroxyl radicals according to Reaction 33 (Scheme 2.7). The macro alkoxy radicals can abstract a hydrogen atom from the polymer (Reaction 34). They can also undergo unimolecular decomposition reactions resulting in polymer chain scission (Reaction 35a). The scission, according to Reactions 35a, has been considered the main reaction responsible for PP degradation [43]. Reaction 35b (Scheme 2.7) also involves scission, but only elimination of a methyl radical occurs, without scission of the main chain.
The importance of the scission process in polyolefin degradation has been questioned [79]. At first, it was believed that the alkoxy radicals and hydroxyl radicals formed on photolysis of hydroperoxides react with ketones to form carboxylic acids and esters. That
initial proposal has been corrected [80] and it was then assumed that a complex between a hydroperoxide group and a ketone was photolyzed. The last reaction yields carboxylic acids and other groups without the production of free hydroxyl radicals. However, the formation of carboxylic acids can also be explained by other reactions. The newly commercialized metallocene PP [81] shows photo-oxidative behavior comparable to that of conventional PP [82].
Although many aspects of PP photo-oxidation seem well established, research on the phenomena is continuing on a large scale [83-95]. The physical-mechanical aspects of photo-oxidation are given due attention.
Photo-oxidation of PE
PE is less sensitive to photo-oxidation than PP. Nevertheless, the combination of UV light and air induces its degradation. The general photo-oxidation scheme (Scheme 2.1) applies also to PE. However, PE photo-oxidation seems more complex because this polyolefin comes in several forms. HDPE can be manufactured by two main processes involving Phillips- or Ziegler-type catalysts. The new technologies bring even more variations. For example, high pressure, low density PE (LDPE) does not involve metal catalysts but is manufactured either in a continuous process in tubular reactors or in a batch process in autoclaves. Linear low density PE (LLDPE) is, in fact, a copolymer of ethylene with various olefins such as 1-butene, 1-hexene, 1-octene, 4-methyl-1-pentene. Because of the correspondingly high number of tertiary hydrogen atoms in LLDPE, it is expected that this polyolefin has similar reactivities with PE. The newly developed metallocene PEs (m-PE) are similar to LLDPE but do not involve Ziegler-type catalysts (Ti) in manufacturing [96]. However, there are also catalyst residues remaining in the polymer. In addition to short and long chain branching, unsaturated groups are also structural irregularities present in many PE types. They are mainly vinyl and vinylidene groups and a few vinylene groups.
The products stemming from PE photo-oxidation that have been identified are vinyl-alkenes, acids, ketones, gamma-lactones, and esters [97]. The main differences from PP photo-alkenes, acids, ketones, gamma-lactones, and esters [97]. The main differences from PP photo-oxidation products are significantly higher amounts of vinyl groups and acids and lower amounts of esters. It is noteworthy that, so far, aldehydes have not been identified unambiguously as PE photo-oxidation products. This is the consequence of either a very low rate of formation or a rapid rate of oxidation to other products, essentially acids.
The most fundamental difference between PE homopolymers and PP is the behavior of the hydroperoxides towards photolysis. Hydroperoxides do not accumulate in PE on photo-oxidation, whereas in PP, they do. These results are in contrast with those observed on thermal oxidation: hydroperoxides accumulate in both PP and PE. The hydroperoxides accumulated in PE thermal oxidation have been used to study their photolysis. It has been found that the hydroperoxide level drops rapidly upon UV exposure [98-100]. It has been concluded that during the processing of LDPE, hydroperoxidation occurs in position to vinylidene groups [99]. The allylic hydroperoxides formed have been found to initiate the oxidation of PE photolytically [99]. In contrast to the preceding results, it has been shown that hydroperoxides formed on thermal oxidation of LDPE at do not have a photo initiating effect
impossibility of such a cage reaction for the radicals resulting from the homolytic splitting of the corresponding hydroperoxide. This interpretation has been challenged [105–114]. New mechanisms of hydroperoxide decomposition have been proposed to explain the photolysis of hydroperoxides formed on thermal oxidation of LDPE. The main photolysis products, i.e., ketones, were at first considered reaction products of a bimolecular reaction involving the hydroperoxide and a neighboring chain segment. This proposal has been amended recently [115]. Now the formation of ketones is attributed essentially to the intramolecular Reaction 36 in Scheme 2.8. In fact, quantum chemical calculations show Reaction 36 to be much more favorable energetically than the intermolecular reaction [115].
There is another photolytically induced intramolecular hydroperoxide decomposition of some importance with PE, as shown in Scheme 2.8 (Reaction 37). This reaction seems to be much less important than the intramolecular Reaction 36 yielding the ketone in the chain. However, it involves main chain scission and, as a consequence, affects directly the
1. Intramolecular photolysis of secondary hydroperoxides
2. Intramolecular photolysis of tertiary hydroperoxides
mechanical properties of the polymer. The same is valid for the intramolecular photolytic decomposition of tertiary hydroperoxides shown in Reaction 38 in Scheme 2.8. Although Reaction 38 in Scheme 2.8 is shown for PP it can easily be adapted to the corresponding reaction of PE hydroperoxides.
The photoinductive ability of tertiary hydroperoxides in PP and analogous polymers such as PS is attributed to the impossibility of a reaction analogous to Reaction 36 in Scheme 2.8. However, there would still be an interaction of the hydroperoxide group with a hydrogen atom of the polymer chain. Then, the hydroperoxide would decompose without intermediate formation of a hydroxyl radical according to Reaction 4b in Scheme 2.1.
So far, most studies concerning PE photo-oxidation deal with high pressure LDPE and HDPE of the Phillips or Ziegler type. Recently, copolymers and blends have drawn attention.
Copolymers of ethylene with vinyl acetate (EVA) were the first PE copolymers to gain industrial importance. Some results show EVA copolymers with vinyl acetate contents less than $10\%$ to be at least as resistant to photo-oxidation as PE homopolymers [116]. For EVAs with vinylacetate content greater than $40\%$, it has been found that photo-oxidation has similar effects to those in vinyl acetate homopolymers [117]. Thus, the higher the vinyl acetate content, the more the copolymers resemble polyvinyl acetate, which is much more susceptible to photo-oxidative degradation than PE.
LLDPE, as already mentioned, is produced by a Ziegler type polymerization of ethylene with various $\alpha$-olefins. In contrast to high pressure LDPE, which has both short and long chain branches as well as unsaturated groups, LLDPE shows only short chain branching and unsaturation. It has been found that LLDPE (1-octene comonomer) is significantly less sensitive to photo-oxidation than LDPE with comparable density and molecular mass [118]. On the other hand, it is found that the mechanisms of photo-oxidation observed with 1-octene-LLDPE are very similar to those occurring in high pressure LDPE having a low content of vinylidene unsaturation [119]. Moreover, large variations in polymer density caused by variations in the number of branching points did not yield a significant effect on the photo-oxidation rate [119]. Comparison of blends of LDPE and LLDPE to unblended polymers does not show any effect of composition if carbonyl development is considered [120]. However, the deterioration of mechanical properties, such as tensile strength and elongation, proceeds more rapidly for LLDPE and blends containing high amounts of the elongation, proceeds more rapidly for LLDPE and blends containing high amounts of the latter. This result is attributed to branching reactions occurring during photo-oxidation [120]. It may also be the result of, at least in part, the fact that the UV light of an unfiltered xenon arc was used for the exposures.
Recently the photo-oxidation of PE-PP blends has been examined [121]. It has been found that with increasing amounts of PP in PE, the photo-oxidation behavior of the blends changes gradually from that of pure PE to that of pure isotactic PP. In particular, the rate of hydroperoxidation of the PE-PP blends increases with the PP content [121].
Photo-oxidation of crosslinked PE has been given fair attention [122–125]. It has been found that photo-oxidation of peroxide crosslinked PE proceeds at rates comparable to
hose of LDPE that is not crosslinked [122]. Obviously, the amount of branching and tertiary hydrogen atoms resulting from crosslinking are not sufficient to induce an increase in the photo-oxidation rate. Silane crosslinked PE behaves similarly to peroxide crosslinked PE [123].
The rate of photo-oxidation of PE crosslinked bygamma-radiation is found to be higher than for LDPE that has not been irradiated [124, 125]. As is for PP, research on photo-oxidation of PE is continuing on a large scale with special emphasis on physico-chemical aspects.
