Limits of UV-light acceleration on the photooxidation of low-density polyethylene

https://doi.org/10.1016/j.polymdegradstab.2020.109443Get rights and content

Highlights

  • Photooxidation rate of LDPE does not increase above a certain value of irradiance.

  • Increasing temperature can accelerate photooxidation even over this limit value.

  • The Schwartzchild coefficient for LDPE is always well below one.

Abstract

Lifetime prediction of polymers requires accelerated ageing based on an increase in UV light intensity and temperature. However, the representativeness of artificial ageing is still an issue. The relevance of highly accelerated UV-induced weathering raises the question of reciprocity failure depending on the polymer. The influences of irradiance and temperature on the photooxidation kinetics of polyethylene were studied to determine the extent to which increasing light intensity leads to accelerated photooxidation of polyethylene. Polyethylene films were subjected to accelerated photooxidative experiments at various UV light intensities and different temperatures, and the amounts of carbonylated photoproducts were monitored by IR spectroscopy. The results show reciprocity failure for polyethylene photooxidation and the limits of acceleration by UV light.

Introduction

The issue of lifetime prediction of polymeric materials has been a major concern for decades. Laboratory studies based on artificial ageing are required for proper determination of the long-term durability of polymers. Despite the long-term effort of various laboratories and industrial R&D centres, the reliable and accurate prediction of polymer lifetime is still an issue. Acceleration of UV light-induced degradation of polymers requires devices which allow working at intensities much higher than the solar light intensity. Experiments for predicting the environmental durability are based on accelerated artificial ageing methods, in which the acceleration is usually obtained by increasing the intensity of the light impinging on the surface and increasing the temperature of the exposed samples. Testing in these accelerated conditions can provide useful information in shorter time periods, but the stresses (light, heat,…) and their intensity have to be chosen accurately to avoid unrealistic degradation modes and failure mechanisms not observed in “true” environmental conditions [1,2]. There are many sources of misleading and erroneous results. Understanding the sources of errors is mandatory for developing relevant testing methods [3,4,5,6,7].

The issue of the representativeness of accelerated artificial ageing has been the object of many papers published in the scientific literature. Accurate prediction of the durability of polymers and data indicating that the material is capable of fulfilling its expected service requirement safely are of crucial importance.

Studies focusing on the influence of temperature on thermooxidative degradation have gained considerable attention for years [8]. Thermal degradation will not be considered in this article. It is often reported that the chemical processes governing thermal oxidation follow Arrhenius behaviour, although there is evidence that some polymers show curvature in the Arrhenius plot of oxidation rates [9]. It must also be considered that increasing the temperature of the samples during the experiments can cause heterogeneous degradation due to diffusion-limited oxidation. This issue is also well documented and has been reported in a large number of scientific papers [10], [11].

In the case of the reactions produced via light exposure, the situation is more complex. The accelerated ageing methods use commercial devices for accelerated weathering or homemade constructed devices. The acceleration of ageing is usually obtained by increasing the intensity of the UV light and by simultaneously increasing the temperature of the exposed samples. Once again, the main question is the representativeness of the accelerated ageing experiments for various reasons.

The mechanism of photooxidation can be wavelength-dependent. Most papers agree with the fact that light sources emitting short wavelengths, which are not present in sunlight, have to be rejected. The difference between the harshness of short and long wavelengths in the UV and visible domains of the solar radiation spectrum is well understood and documented in the literature [12], [13], [14]. In addition to the influence of the light distribution on the mechanism, it must also be recalled that the rate of photooxidation can be wavelength-dependent and thus may vary with the spectral distribution of the irradiation.

Increasing the intensity of UV light may be the cause of errors when predicting the lifetime. Light sources with excessively high intensities (lasers) should be avoided, since this kind of source may produce biphotonic processes, triplet-triplet annihilation, or excessive radicals recombinations, producing results similar to those obtained when using sources with short wavelengths. This causes ageing mechanisms not observed in natural conditions The light intensities used in conventional accelerated weathering devices are, however, well below the level which would cause a significant amount of these processes associated with pulsed lasers. Previous results published some years ago unambiguously showed that the photooxidative rate depends on the light intensity [15], [16], [17]. However, it remains a fact that the effect of irradiance, even in acceptable ranges of light intensity, is still the object of severe debates. Currently, there is a growing interest regarding this issue [18], [19], [20], [21], [22] because of the technical, economic and environmental problems arising from the degradation of polymers.

Special attention must be paid to ensure that there is no change between the mechanisms of the accelerated tests and outdoor exposure. Nonlinear oxidative degradation would be expected if diffusion of oxygen into the sample, additive migration, or other thermally controlled reactions become the rate-limiting step. Although this is a well-documented area, the effect of irradiance, even if not producing heterogeneous oxidation and remaining in acceptable ranges of wavelengths and light intensity, is not understood well and is still the object of severe debates.

In the presence of light and air, light induces photochemical reactions, which can be described by a classical chain oxidation mechanism. The global photooxidation processes depend upon both the effective irradiance and the temperature. Photooxidation is thermally activated, with activation energies on the order of a few tens of kJ/mol. Increasing the temperature allows for accelerating the photooxidation processes, and it also allows maintaining a relevant balance between the photochemical processes and thermodegradation. This is important in the case of polymers susceptible to parallel modification provoked by light or by temperature [23,24,25].

The effect of irradiance on the rate of a photochemical process can be characterized by the reciprocity. A material obeys reciprocity if the degradation is a function of the total radiant energy and not a function of the rate at which the energy is applied. The reciprocity law can be described as:It=constant,withI=irradianceandt=exposuretime

The rates of photochemical processes as a function of light intensity usually follow the Schwarzschild law [26]:k=AIpwhere k is the reaction rate, A is a proportionality constant, I is intensity (or irradiance), and p is the Schwarzschild coefficient, which is an experimentally derived number. This may also be written as:Ipt=constant

When p =1, Schwarzschild's law becomes the reciprocity law. For most (unstabilized, stabilized or pigmented) polymeric systems, the literature reports that the p-coefficient ranges between 0.5 and 1 [18]. However, it has to be considered that most reciprocity experiments reported were conducted with commercial, stabilized or pigmented materials and, more importantly, most of the studies on the effect of irradiance and/or temperature reported in the literature focus on properties such as colour change and yellowing, gloss loss, crystallinity, modulus, elongation at break, and mechanical properties [20,27,28]. These properties are interesting from a practical point of view, but the changes in these properties are not a direct consequence of UV light degradation and in most cases involve several processes, some of which are not light-dependent. A unique and clear relationship between the degradation of the polymers properties and the chemistry which is involved is still an issue [29]. This makes the understanding of degradation more complex and prevents researchers from proposing general concepts.

Therefore we focused on the influence of the extent of the applied stresses (light, temperature) on the most primary detectable events, which are the chemical reactions involved in the degradation of the properties.

Reciprocity failure occurs when the coefficient of proportionality changes with light intensity. This is critical in lifetime prediction because light is usually applied at a high and constant irradiance during testing, while in general, the usage conditions are lower irradiance and cycles of light exposure. As recalled above, in most cases, the light intensities used in conventional accelerated weathering devices are well below the level which would cause a significant amount of undesired processes.

In the present study, we focused on the photodegradation of polyethylene, a widely used polymer whose degradation mechanisms are well known. This paper reports a study of the influence of light intensity on the rate of photooxidation of polyethylene. Chemical modification provoked by UV light was characterized by infrared spectrometry measurements. We also present results on the photodegradation of bisphenol A polycarbonate (PC) and poly(ethyleneterephthalate) (PET). The long-term objective is to define the limits and accuracy of accelerated weathering of polyethylene and to discuss the relevance of “highly” accelerated UV weathering tools.

Section snippets

Materials

Linear low-density polyethylene (LLDPE) was supplied by SABIC® (LLDPE 324CE, density = 924 kg.m−3, Mw = 73,000 g.mol−1) (description in reference [30]).

PC was purchased from Aldrich (Mw = 64,000 g.mol−1).

PET films were kindly supplied by Toray Film Europe (Miribel-France) in the form of thin, uniform films with thicknesses of 125 microns.

Preparation of polymer films

LLDPE films with different thicknesses (from 50 to 130 microns) were obtained by compression-moulding of LLDPE pellets at 140° and 200 bar.

PC films (100

Results and discussion

Infrared spectroscopy has proven to be a useful and is frequently used analytical technique for monitoring the oxidation process of polyethylene [31]. Fig. 1 illustrates the changes to the LLDPE spectra, which occurred during the photooxidation, in the range of the carbonyl absorption.

The changes in the spectrum indicate that ketones (1718 cm−1) are formed during the initial steps, and carboxylic acids (1713 cm−1), esters (1735 cm−1) and lactones (1780 cm−1) are formed during secondary

Conclusions

Accelerating degradation has been desired and explored for years, but it must be considered that the Schwarzschild coefficient p may vary from one given polymer to another, which is well known; however, it may also vary for a given polymer depending on the stability of the material brought by stabilizers. In the case of unstabilized polyethylene, no acceleration of the oxidation can be obtained by increasing the irradiance above a certain level not considered as high. Reciprocity failure with a

CRediT authorship contribution statement

Sandrine Therias: Writing - review & editing. Géraldine Rapp: Investigation. Claire Masson: Resources. Jean-Luc Gardette: Writing - original draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The financial support of the Agence Nationale de la Recherche through the program LabCom POPBA is greatly appreciated.

References (43)

Cited by (30)

View all citing articles on Scopus
View full text