Socioeconomic Data and Applications Center
Environmental Effects of Ozone Depletion 1998 Assessment

Wavelength Sensitivity of Photodegradation

Assessing the damage to materials from exposure to UV-enhanced sunlight requires an understanding of their spectral sensitivity. Spectral sensitivity data for polymers are typically generated using a source of monochromatic radiation or a white light source such as a filtered xenon source (one whose spectral irradiance distribution is designed to closely approximate that of terrestrial sunlight at unit air mass). With experiments based on exposure of materials to monochromatic radiation, the effectiveness Á (l ) in units of damage per incident photon (defined as the ratio of the measured change in the property of interest D P to the number of incident photons), is obtained for several irradiation wavelength. In most instances a linear relationship exists between the logarithm of the effectiveness of damage and the wavelength of exposure, with higher damage per incident photon obtained at the shorter UV wavelengths. The gradient B of a plot of the natural logarithm of Á (l ) versus l is a measure of the monochromatic wavelength sensitivity. The Table 7.2 lists the values of B that quantifies the wavelength sensitivity of polymeric materials reported in the literature. All data in the table are based on laboratory exposure experiments using monochromatic radiation. The spectral sensitivity of a material assessed by ths approach depends upon the type of damage process investigated.

When using a white-light source, a series of cut-on filters is used to isolate different bands of the source spectrum (Andrady 1997). The increment of radiation between two adjacent filters in the series depends on their transmission spectra T(l ) and the spectral irradiance distribution of the source H(l ).

Increment = H (l ) ( Ti (l ) - Ti+1(l ) )

Fig. 7.3. Summary of activation spectra for common polymers exposed to solar-simulated radiation (filtered xenon source radiation) published in the literature showing the wavelength interval in which most photodamage was obtained.

a) All data reported were generated using cut-on filter technique in combination with a filtered xenon source. The wavelength ranges shown are the intervals within which maximum photodamage was obtained. b) Abbreviations for plastics. ECO - {ethylene-carbon monoxide (1%) } copolymer: PC - Bisphenol A polycarbonate sheet: PVC - poly(vinyl chloride). c) 1 and 2. Newsprint paper made of mechanical pulp from Southern Pine. Photodamage measured is the increase in yellowness index and the optical brightness of the paper (Andrady et al., 1991). 3. Data estimated from the publication (Launer, 1965). 4. Rigid PVC sample similar to that used in vinyl siding applications except that no opacifier (titanium dioxide) was used. The absence of the opacifier allowed photodamage to be obtained within a reasonable duration of laboratory exposure. The reference source includes data for samples formulated with opacifier as well. (Andrady et al., 1989). 5. Data is for yellowing of bisphenol A polycarbonate sheet. The study also found crosslinking damage at wavelengths below 315 nm (Pyrde, 1985). 6 Commercial bisphenol A polycarbonate containing a light absorber as a photostabilizer was used to obtain these data using natural sunlight as a source of radiation (Andrady et al., 1991). 7 Data for expanded extruded polytyrene sheets designed to photodegrade faster than regular resin (Andrady unpublished data) 8. Data for polypropylene films (Zhang et al., 1996). 9. Injection molded polypropylene (Andrady unpublished data) 10-13. Data on polyethylene films (Andrady, 1996). 14. Nylon fibers (Hu, 1998).

Typically, the change in a property of interest (D Damage), before and after exposure of identical samples exposed behind filters i and i+1 is obtain from the experiment. This is plotted as a function of the bandpass ( Ti (l ) - Ti+1(l )) at half bandwidth for each pair of filters used. The resulting bar diagram has been referred to as an activation spectrum in the materials research literature. Wavelength sensitivity studies using a xenon source (that provides radiation spectrally similar to solar radiation, Hirt et al., 1967) are particularly valuable in assessing the potential effects of enhanced UV-B radiation in sunlight (Fig. 7.3). A qualitative estimate of the relative significance of various ultraviolet wavelengths in the terrestrial solar spectrum in causing specific photodamage in polymers can be discerned from the published activation spectra. Figure 7.4 shows a compilation of these regions from published data on various polymers. With activation spectra for most polymers, it is the less efficient UV-A wavelengths rather than the UV-B wavelengths that yield the largest relative amount of photodamage. This is expected because of the relatively low UV-B content in the solar spectrum. However, in activation spectra for yellowing of PVC (rigid formulations), polycarbonate, and polyamide, the maximum damage is obtained in a spectral region that includes UV B wavelengths as well. It is clear from the figure that most of the polymers will undergo considerable degradation when exposed to the UV-B and/or UV-A spectral regions, suggesting that any increase in the solar UV will result in an increase in damage.

Fig. 7.4. Cumulative wavelength sensitivity curve for change in extensibility of low-density polyethylene film exposed to a filtered xenon source for 469 hours at ambient temperature (Hu, 1997).

The change in property of interest may also be plotted as a function of the 10 percent transmittance value of filters to yield a cumulative spectral sensitivity curve. Data on wavelength sensitivity of polyethylene film samples xposed to filtered xenon-source radiation is shown in Figure 7.4. The data shows that wavelengths shorter than about 400 nm affect the mechanical properties of the film. Using a full white light spectrum as opposed to narrow near-monochromatic bands of radiation in wavelength sensitivity experiment has the advantage that synergism (or antagonism) at different wavelengths can contribute to the results.

The spectral sensitivity data from the two experimental approaches (using near monochromatic radiation and using a white light source) are interconvertible provided several assumptions are made. i) the law of reciprocity applies ; ii) effects of different wavelengths are additive ( the lack of synergistic or antagonistic effects); and iii) the light absorbance characteristics of the polymer are not time-dependent within the durations of exposure. For even the common polymers not enough data is often available to validate these assumptions. However, in the case of poly(vinyl chloride), PVC, and mechanical pulp materials, the activation spectra have been used to successfully derive monochromatic spectral sensitivity curves in the UV-B and UV-A wavelength range (Andrady, 1997).

For more information about CIESIN and our activities contact CIESIN User Services. E-mail: ciesin.info@ciesin.org. Tel.: (914) 365-8988