Exposure study on UV-induced degradation of PTFE and ceramic optical diffusers

Abstract

We report on a study of the ultraviolet (UV)-induced degradation on optical grade polytetrafluoroethylene (PTFE) and ceramic diffuser samples. Long-term UV exposure may significantly alter the reflectance and lead to an error in the calibration of optical instruments. A large integrating sphere was used to irradiate the samples for 334.7 days at an irradiance level of $194.9{\text{W/m}}^2$. Samples were qualified and measured for reflectance factor, bidirectional reflectance distribution function, and fluorescence, before and after the exposure, and at 12-week intervals during the exposure. This study revealed significant differences between the aging behavior of ceramic and PTFE samples.

1. INTRODUCTION

Applications of imaging that depend on radiometrically derived data products are growing. The applications are expanding beyond the visible spectral region into the ultraviolet (UV) and infrared (IR). These spectral regions have lagged the visible spectral region due to fewer choices of detectors and optical components. UV imaging has seen recent growth in applications and increased product availability. These applications include UV imaging for astronomy, photolithography, material inspection, forensics, chemical detection, and skin reflectance. This increased demand in the UV will likely drive an increase in demand for reliable optical references that are stable in this spectral region.

Calibration of radiometric instruments requires stable, high-quality standards. One factor that could potentially change the characteristics of calibration standards is UV-induced degradation. For example, in onboard space calibrations, UV exposure from the sun can result in errors in the calibration of space-flight instruments.

Several studies have been conducted, attempting to relate UV exposure to changes in the optical properties of materials used for optical diffusers. A study investigating exposure to low levels of UV radiation, such as those found in typical laboratory conditions, found minimal evidence of aging [1]. Möller et al. measured the change in directional-hemispherical reflectance for Spectralon following low-level exposure to irradiation sources at 254 and 365 nm [2]. Although they observed changes that were <1%, the authors recommended that Spectralon only be used for the short-term transfer or comparison of reflectance scales.

In contrast, studies investigating exposure to high levels of vacuum UV and UV radiation, such as those experienced by onboard space-flight instruments, have found varying levels of degradation.

Georgiev et al. exposed polytetrafluorethylene (PTFE) samples to vacuum UV radiation [3]. They found that the change in directional-hemispherical reflectance was significant, with the greatest changes occurring at the shortest wavelengths. Georgiev et al. also investigated the change in bidirectional reflectance and found that magnitude of the change was dependent on the measurement geometry. In contrast, Leland and Arecchi found that most of the degradation in reflectance of Spectralon is due to UV exposure (200–380 nm), and that vacuum UV exposure contributes very little to the optical degradation [4]. Furthermore, the authors estimated a degradation rate at 400 nm of less than 0.05% per hour of exposure to extra-atmospheric UV/vacuum UV radiation. Bruegge et al. exposed Spectralon to UV radiation and observed noticeable decreases in directional-hemispherical reflectance for wavelengths less than 800 nm [5]. The greatest decreases occurred for the shortest wavelengths. Gibbs et al. studied the influence of UV irradiation by mercury and xenon lamps on the directional-hemispherical reflectance of barium sulfate and Spectralon [6]. They found that all materials changed following UV irradiation, although changes for Spectralon were minimal when it was irradiated with only the xenon lamp.

Sun and Wang reported that the solar diffuser used in the Visible Infrared Imaging Radiometer Suite onboard the Suomi National Polar-orbiting Partnership satellite degraded during 2 years of service by 28.5% at 412 nm and 1.2% at 935 nm [7]. Furthermore, they noted that the degradation was exponential before becoming stable in late 2013. Xiong et al. examined data from Moderate Resolution Imaging Spectroradiometers instruments operating onboard the Terra and Aqua spacecraft and found that the solar diffusers had degraded by 48% and 19%, respectively, at 410 nm [8]. Degradation at 940 nm was less, 2.3% and 0.6%, respectively. Smaller changes, 1%–2%, were observed for Spectralon diffusers aboard the Multi-angle Imaging Spectroradiometer and Medium Resolution Imaging Spectrometer, respectively [9,10]. Li et al. described measurement of spatially varying bidirectional reflectance distribution function (BRDF) using an imaging spectrometer [11].

The objective of this study is to assess UV exposure effects on materials commonly used as reflectance standards using UV irradiation generated by the National Institute of Standards and Technology (NIST) facility Simulated Photodegradation via High Energy Radiation Exposure (SPHERE) [12]. The facility consists of a 2 m integrating sphere that provides a source of intense ultraviolet radiation from 290 to 400 nm. It is designed to uniformly illuminate samples with a high level of irradiance that mimics terrestrial UV solar exposure at an accelerated rate. To this end, sintered PTFE and ceramic samples were exposed to the UV irradiation generated by SPHERE for four 12-week periods beginning in 2014, a total exposure time equivalent to 334.7 days. For each sample, the directional-hemispherical reflectance, BRDF, and fluorescence were measured before the UV exposure began and after the UV exposure ended a year later. The reflectance of each sample was also measured every 12 weeks during the exposure period. Results show significant changes in the reflectance properties of the ceramic samples, whereas the reflectance properties of the PTFE samples were unchanged for most wavelengths.

2. DESCRIPTION OF SAMPLES

Two types of materials were investigated in this study. The first is sintered PTFE, which is a commonly used material for reflectance standards. The second is ceramic specially formulated for use as an optical diffuser. The ceramic used in this study is 99% alumina (Al₂O₃) and 1% of a proprietary silicate glass, which is used to bind the Al₂O₃ particles during kiln firing. This ceramic has a matte finish that resists particulate contamination due to its neutral electrostatic charge. Both materials have high spectral reflectance values (greater than 90%) across much of the reflected solar region (250–2500 nm).

Each set was composed of four samples of the same material. Three of the samples from each set were exposed to UV irradiation, while one from each set was designated as a control sample and was not exposed. All eight samples were 50 mm × 50 mm. The PTFE samples were 10 mm thick, and the ceramic samples were 8 mm thick.

The ceramic is known to have small but measurable fluorescence response, centered at approximately 690 nm, likely due to chromium ion impurities [13]. Based on exploratory studies at NIST, the contribution of fluorescence to the total light reflected is greatest with UV excitation and decreases to a negligible amount by 650 nm. (The fluorescence contribution appears to be somewhat sample dependent and is approximately 1% at 300 nm and 0.1%–0.3% at 500 nm.)

3. SAMPLE CHARACTERIZATIONS

Prior to UV exposure, the optical properties of the materials were investigated. The directional-hemispherical reflectance factor (DHRF), bidirectional reflectance factor (BRF), and fluorescence were measured for all eight samples used in the study.

A. DHRF

For this study, the directional-hemispherical reflectance measurements were acquired using a commercial spectrophotometer fitted with an integrating sphere accessory. The integrating sphere is lined with sintered PTFE and has a diameter of 150 mm. The spectrophotometer’s sample beam passes through an oval entrance port (15 mm × 25 mm) on the integrating sphere to the sample port (25 mm diameter), where it illuminates the sample at an angle of incidence of 8° from the sample normal. The size of the illumination spot on the sample is 14 mm × 6 mm. Because of the geometry of this measurement, the measurand is referred to as 8°/h spectral reflectance factor.

The illumination source was a deuterium lamp for wavelengths shorter than 319 nm and a quartz-tungsten-halogen (QTH) lamp for longer wavelengths. The detector was a photomultiplier tube for wavelengths shorter than 860 nm and an indium-arsenide-gallium detector for longer wavelengths. For all measurements, the incident beam was depolarized using a 30 mm depolarizing element. The spectral bandwidth was restricted to 3 nm over the spectral region of 250–860 nm but could vary, with a maximum of 20 nm, over the 860–2500 nm spectral region.

Measurement of the 8°/h spectral reflectance factors of each sample was achieved by relative comparison to a reference standard and required three separate spectral scans. First, the sample was placed on the sample port. Then, the reference standard was placed on the sample port. Finally, the sample port was left empty to collect the dark signals. The reference standard was pressed PTFE. Its values for spectral reflectance factor were established using the absolute method of Van den Akker [14].

The DHRF of the sample, $R_x$, at a wavelength, $\lambda$, was determined by the following measurement equation:

$$R_x\left(\lambda \right)=\frac{S\left(\lambda \right)-S_d\left(\lambda \right)}{S_s\left(\lambda \right)-S_d\left(\lambda \right)}\cdot R_s\left(\lambda \right),$$ (1)

where $R_s$ is the 8°/h spectral reflectance factor of the reference standard, $S$ is the signal from a scan of the sample, $S_s$ is the signal from a scan of the standard, and $S_d$ is the dark signal (empty sample port).

The DHRFs of the eight samples used in this study were measured before UV exposure from 250 to 2500 nm by 10 nm, and the DHRFs of two samples are plotted in Fig. 1. The expanded uncertainty $\left(k=2\right)$ for these measurements is 0.005, which includes contributions from the reference standard, wavelength, and repeatability. The reflectance factors of PTFE and ceramic are spectrally flat across most of the measured spectral region. In the UV, the reflectance factor for PTFE decreases modestly with wavelength, whereas the reflectance factor for ceramic begins to decrease rapidly for wavelengths shorter than 600 nm. Spectral features occur for PTFE at wavelengths longer than 1900 nm. The spectral features for ceramic at these longer wavelengths are moderate compared to PTFE.

Fig. 1.

DHRF of two samples before UV exposure as a function of wavelength. The expanded uncertainty $\left(k=2\right)$ for these measurements is 0.005.

B. BRF

Measurements of BRDF were acquired using the NIST Spectral Tri-function Automated Reference Reflectometer (STARR) [15]. This home-built instrument serves as the national reference for spectral bidirectional reflectance measurements. It consists of a source, a monochromator to provide wavelength resolution, a goniometer for varying sample and detector orientations, and a detector.

The source is a QTH lamp, which is focused through an order-sorting filter and a shutter onto the entrance slit of a single-grating monochromator. The beam exits the monochromator and is collimated by an off-axis parabolic mirror. It then passes through a Glan–Taylor polarizer before passing through the sample goniometer. The beam is collected by the detector aperture and focused by a lens onto an ultraviolet-enhanced silicon photodiode. The goniometer enables the acquisition of both reflected and incident signals. A schematic of STARR is available in Ref [14].

Measurement of the BRDF of each sample required two separate measurements: spectral scan of the flux incident on the sample and spectral scan of the reflected flux at various detection angles. The BRDF, $f_r$, was determined by

$$f_r=\frac{S_r}{S_i}\cdot \frac{R_i}{R_r}\cdot \frac{D^2}{A_r\cdot \cos {\theta }_r},$$ (2)

where $S_r$ and $S_i$ are the net signals (with dark signals subtracted) in the reflected and incident directions, respectively; $R_i$ and $R_i$ are the detector responsivities to the incident and reflected fluxes, respectively; $D$ is the distance between the detector aperture and the sample; $A_r$ is the area of detector aperture; and ${\text{θ}}_r$ is the detection angle with respect to the sample normal. The BRF is then calculated by

$$R=\pi f_r.$$ (3)

The BRFs of the eight samples used in this study were measured before UV exposure at 400, 700, and 1100 nm at 0° incident angle and viewing angles from −60° to 60° by 15° (not including the specular angle at 0°) about the sample normal. Figure 2 plots the BRFs for two samples at 700 nm. The values for both PTFE and ceramic are symmetric about the 0° viewing angle. Consistent with the spectrally flat trend observed for DHRF values, the BRF values for PTFE at 400 and 1100 nm are similar in magnitude to those values at 700 nm. Similarly, where a significant decrease in DHRF values for wavelengths less than 600 nm is observed for the ceramic sample, the BRF values for the ceramic sample at 400 nm are significantly smaller (by 0.03–0.04) than the BRF values at 700 and 1100 nm. The expanded uncertainty $\left(k=2\right)$ for these measurements is 0.005, which includes contributions from detection angle, wavelength, sample alignment, and repeatability.

Fig. 2.

BRF of two samples at 700 nm before aging as a function of viewing angle. The expanded uncertainty $\left(k=2\right)$ for these measurements is 0.005.

C. Fluorescence

Bidirectional fluorescence measurements were obtained by illuminating the sample with an optical parametric oscillator and measuring the reflected light with a stray-light corrected spectrometer [16]. Each sample was illuminated at an incident angle of 0° for excitation wavelengths from 310 to 660 nm in 50 nm intervals. The fiber probe of the spectrometer viewed the reflected light at a detection angle of 45° and measured its spectrum distribution from 300 to 1100 nm.

The percent bidirectional fluorescence %Fl (PBF) at excitation wavelength ${\lambda }_{\text{ex}}$ was calculated by the following equation:

$$\%\text{Fl}\left({\lambda }_{\text{ex}}\right)=\left(100\%\right)\times \frac{{\int }_{\lambda 1}^{\lambda f}R\left(\lambda \right)I\left(\lambda \right)\text{d}\lambda -{\int }_{\lambda 1}^{\lambda 2}R\left(\lambda \right)I\left(\lambda \right)\text{d}\lambda }{{\int }_{\lambda 1}^{\lambda f}R\left(\lambda \right)I\left(\lambda \right)\text{d}\lambda },$$ (4)

where $I$ is the bidirectional fluorescence intensity as a function of the wavelength $\lambda$; $R$ is the spectral responsivity of a silicon detector; ${\lambda }_1$ and ${\lambda }_2$ are the starting and ending wavelengths (within ±15 nm of the excitation wavelength) for the peak of the bidirectional fluorescence intensity; and ${\lambda }_f$ is the cutoff maximum wavelength (792 nm) for the wings to be considered in the fluorescence calculations.

The PBF observed for sintered PTFE from 300 to 1100 nm is low, approximately 0.1% for UV excitation wavelengths and decreasing to 0.01% for increasing excitation wavelengths. The PBF for the ceramic samples for the same spectral distribution is greater, approximately 1% for UV excitation wavelengths and decreasing to 0.04% for increasing excitation wavelengths. The expanded uncertainty $\left(k=2\right)$ for these measurements is 0.03, which includes contributions from residual stray light and spectrometer calibration.

4. UV EXPOSURE CONDITIONS AND PROCEDURES

All eight samples were exposed and characterized according to a set measurement schedule. The UV source in the SPHERE is described, as are the environmental conditions.

A. Sphere Facility

The NIST SPHERE is a 2 m integrating sphere used for accelerated weathering [12]. The SPHERE system provides intense, uniform UV light as well as precise temperature, relative humidity, and wavelength (via filters) exposure. There are six lamps (Heraeus Noblelight America, LLC) that provide UV light; each lamp is capable of producing 6000 W. The lamps and SPHERE are cooled by a heating, ventilation, and air conditioning unit, with the conditioned space separated from the environmental chambers by a quartz window. Each bulb is located inside a reflector assembly that minimizes the visible and IR radiation of the UV source. A WG295 (SCHOTT Germany) glass window is installed in the optical path between the UV lamp system and the integrating sphere to remove radiation below 290 nm. The exterior shell of the sphere is aluminum and the interior surface is lined with Spectralon, PTFE that is diffusely reflective in the UV regions. The environmental chambers allow precisely and independently controlled relative humidity (0%–75%), temperature (30°C–85°C), and UV-visible irradiance within the sample exposure space. Using bandpass and/or neutral-density filters, wavelength and intensity dependency can be explored. There are 28 such chambers attached to the SPHERE, each with the capability of exposing 17 or more specimens. There are also four chambers equipped with an apparatus that allows the application of mechanical strain during exposure. A schematic of SPHERE is available in Ref. [11].

The spectrum and flux of the UV intensity emitted from sphere exit ports were measured with a UV-visible spectrometer (Instrument Systems CAS 140CT) equipped with a cosine collection sphere [12] at specified intervals as needed. The UV irradiance at the interior of the SPHERE wall was also monitored constantly during the exposure. The output flux and uniformity of the SPHERE at each port are not only affected by changes (aging) in the lamps themselves (replaced every 6 months) but also in the transmittance or reflectance of any material in the optical path before the experimental specimen. Variation between chambers will depend on the optical components installed. There is an UV intensity gradient of approximately 10% (from center of the port across the surface area at which samples are loaded) for each port of a chamber. At the output of the SPHERE, a compound parabolic concentrator produced a collimated and highly uniform UV flux of approximately $173{\text{W/m}}^2$ in the 295–400 nm wavelength range onto the samples. Figure 3 shows a comparison of the SPHERE output to a terrestrial reference solar spectrum, ASTM G173-03 [17]. An extra-terrestrial reference solar spectrum, ASTM E490-00 [18], is not shown. As shown, the SPHERE output is similar to the standard up to 340 nm. Between 340 and 440 nm, the intensity of the SPHERE is greater. The lamp output was chosen because it was the best match to natural sunlight available with this plasma lamp technology. However, the SPHERE UV intensity is much higher than natural sunlight. In addition, the acceleration factor is also dependent on environmental conditions (temperature and humidity) and the specimen’s chemistry and chromophores.

Fig. 3.

Comparison of SPHERE output and ASTM solar spectrum. Sum irradiance for the range from 275 to 450 nm is $194.9{\text{W/m}}^2$ for the SPHERE and $46.2{\text{W/m}}^2$ for the ASTM G173 spectrum, respectively.

B. Measurement Schedule

Before the UV exposure schedule commenced (at $173{\text{W/m}}^2$) in the SPHERE, preliminary diagnostics were performed on all eight samples for DHRF, BRF, and PBF. Two samples were chosen as the control samples, which were never exposed to irradiation in the SPHERE. The other six samples were exposed to UV irradiation in the SPHERE for four 12-week periods separated by 1-week intervals. During the 1-week intervals (at ${\tau }_1$, ${\tau }_2$, ${\tau }_3$, ${\tau }_4$), all eight samples were characterized for DHRF. At the end of the fourth period (${\tau }_4$), the DHRF, BRF, and PBF for eight samples were measured again. Adjusting for SPHERE shutdown times, each of the exposed samples were irradiated for a total of 334.7 days or an equivalent of 3.87 solar years (which was calculated as the product of the number of years in the SPHERE multiplied by the ratio of the SPHERE irradiance to the ASTM G173 reference solar spectrum irradiance between 275 and 400 nm).

C. Environmental Conditions

For the duration of the UV exposure, the relative humidity was 0%, except for a spike to 3.6% relative humidity for about half of the first 12-week period. The temperature in the sample compartment was 27.4°C ± 1.1°C.

5. RESULTS AND ANALYSIS

For all eight samples, the DHRF, BRF, and PBF were calculated and reported.

A. Directional-Hemispherical Reflectance

At each characterization time (${\tau }_1$, ${\tau }_2$, ${\tau }_3$, ${\tau }_4$), the DHRF was measured, and the normalized difference with respect to the DHRF measured at ${\tau }_0$ was calculated as follows:

$$\Delta R_{{\tau }_x}\left(\lambda \right)=\frac{R_{{\tau }_x}\left(\lambda \right)-R_{{\tau }_0}\left(\lambda \right)}{R_{{\tau }_0}\left(\lambda \right)}.$$ (5)

Figure 4 shows the normalized difference in DHRF of the control PTFE sample. The figure also includes bounds denoting the scale variability, which quantifies how much the scale can be expected to change between measurements. The scale variability is evaluated as the combined (root-sum-square) uncertainty of the repeatability of the instrument, wavelength uncertainty, and variability of the reference standard. (Black lines depict the expected scale variability in Fig. 4 and, subsequently, in Figs. 5–12.) For most of the spectral region, there is no change in DHRF for the control sample. However, a decrease in DHRF is observed for wavelengths shorter than 350 nm over time. Figure 5 shows the normalized difference in DHRF for one of the UV-exposed PTFE samples. Like the control sample, there is no change in DHRF for most of the spectral region. Yet, for the UV-exposed PTFE sample, an increase in DHRF is generally observed for wavelengths shorter than 300 nm, with the exception of the final measurement, which is within the scale variability. This trend was observed for all UV-exposed PTFE samples.

Fig. 4.

Normalized difference (%) in DHRF for the control PTFE sample at each characterization time as a function of wavelength.

Fig. 5.

Normalized difference (%) in DHRF for a UV-exposed PTFE sample at each characterization time as a function of wavelength.

Fig. 12.

Normalized difference (%) in BRF at ${\tau }_4$ of all ceramic samples at 700 nm as a function of viewing angle.

Figure 6 shows the normalized difference in DHRF of the control ceramic sample. Except for a few values near 360 and 1930 nm, the DHRF values of the control sample do not change significantly over time. However, that is not true for the UV-exposed ceramic samples. Figure 7 shows the percent change in DHRF for one of the UV-exposed ceramic samples. A dramatic decrease in DHRF values in the UV spectral region occurs following the first exposure period (${\tau }_1$). This change appears to stabilize to approximately −5% at 350 nm during further exposure periods (${\tau }_2$, ${\tau }_3$, ${\tau }_4$). This trend was observed for all UV-exposed ceramic samples.

Fig. 6.

Normalized difference (%) in DHRF for the control ceramic sample at each characterization time as a function of wavelength.

Fig. 7.

Normalized difference (%) in DHRF for a UV-exposed ceramic sample at each characterization time as a function of wavelength.

Figures 8 and 9 show the normalized difference in DHRF for all PTFE and ceramic samples, respectively, at 400 nm as a function of the characterization period. As observed in Figs. 4–7, there is no change in DHRF at 400 nm for any of the PTFE samples whether exposed to UV or not. In contrast, the UV-exposed ceramic samples exhibit a significant decrease in DHRF during the first exposure period. This decrease remains stable despite further UV exposure.

Fig. 8.

Normalized difference (%) in DHRF at 400 nm for PTFE samples as a function of characterization time.

Fig. 9.

Normalized difference (%) in DHRF at 400 nm for ceramic samples as a function of characterization time.

B. BRF

The normalized difference with respect to the BRF measured at ${\tau }_0$ was calculated. Figures 10 and 11 show the results at 400 nm for all PTFE and ceramic samples, respectively. Like the DHRF results, the BRF values for PTFE samples remain stable whether they are exposed to UV or not. However, the BRF values of the exposed ceramic samples decrease significantly compared to the control sample. This decrease in BRF values for exposed ceramic samples is not observed at longer wavelengths, as seen in Fig. 12. Although the data indicate that there is a slight asymmetry in the BRF values, the decrease in BRF values at 400 nm observed for the exposed ceramic samples is generally consistent, approximately 3.4%, over the range of viewing angles measured.

Fig. 10.

Normalized difference (%) in BRF at ${\tau }_4$ of all PTFE samples at 400 nm as a function of viewing angle.

Fig. 11.

Normalized difference (%) in BRF at ${\tau }_4$ of all ceramic samples at 400 nm as a function of viewing angle.

C. Fluorescence

The PBF at each excitation wavelength was calculated for each of the exposed PTFE and ceramic samples. The difference between mean PBF for PTFE samples before (${\tau }_0$) and after (${\tau }_4$) UV exposure is shown in Fig. 13 along with the corresponding values for the ceramic samples. Black lines depict the expected uncertainty. These results indicate that the fluorescence did not change for either sample type following UV exposure. Thus, fluorescence cannot explain the UV-induced change in reflectance observed for the ceramic samples.

Fig. 13.

Difference in PBF for PTFE and ceramic samples at eight excitation wavelengths before and after UV exposure.

6. SUMMARY AND CONCLUSIONS

Two commonly used materials for reflectance standards, PTFE and ceramic, were exposed to UV irradiation generated by the NIST SPHERE facility for an equivalent of 4.35 solar years. Significant changes in the reflectance properties of the ceramic samples were observed, whereas the reflectance properties of the PTFE samples were unchanged for most wavelengths.

Directional-hemispherical reflectance (8°/h) measurements were acquired before, during, and after UV irradiation of the samples. For wavelengths greater than 350 nm, neither the control nor exposed PTFE samples showed any change in their reflectance properties. However, the control and exposed PTFE samples exhibit opposite trends for wavelengths less than 350 nm, with the DHRF of the control sample decreasing over time, and the DHRF of the exposed samples generally increasing with continued UV irradiation. These changes could be due to chemical, metallurgical, or topographical changes in the sample surface. The decrease in DHRF of the PTFE control sample at wavelengths less than 350 nm is difficult to interpret. An internal check standard was measured prior to each set of measurements as a quality control measure, and the results indicate that the instrument was operating as expected, and the reproducibility of the standard’s reflectance values were acceptable. Given that this sample had limited exposure to light during the year-long study, one possible explanation is that the changes in the PTFE control sample are related to its storage conditions during the study. Whereas the exposed samples were almost always in the environmental chambers on SPHERE, the control sample was stored in its original packaging. Another possible explanation is related to the measurement uncertainty of the spectrophotometer in the UV. The efficiency of the integrating sphere decreases for wavelengths below 350 nm due to the decreasing reflectance of the sintered PTFE lining the sphere. Although we have attempted to account for these changing properties in the scale variability (as described in Section 5.A), further characterization of the sphere and its contribution to the uncertainty in the UV may be warranted.

Bidirectional reflectance measurements were acquired before and after UV irradiation for 0° incident angle at 400, 700, and 1100 nm. Consistent with the results for directional-hemispherical reflectance, the exposed ceramic samples showed significant decrease in reflectance at 400 nm. An average decrease of 3.4% is observed for all viewing angles. The bidirectional reflectance for exposed PTFE at all wavelengths and exposed ceramic at 700 and 1100 nm does not change.

The decreases in directional-hemispherical and bidirectional reflectance observed for the exposed ceramic occur within the first 12 weeks of UV exposure and remain stable for the remaining 36 weeks of exposure. An investigation of all samples before and after UV exposure find no discernible changes in fluorescence. Additionally, visual inspection of the samples reveals no change in the appearance of surfaces.

As described in Section 1, previous studies have investigated changes to the reflectance properties induced by UV exposure for a variety of commonly used diffuse reflectance materials. Exposure levels ranged from low, such as those experienced during typical laboratory and measurement conditions, to high, such as those experienced by onboard space-flight instruments. The results presented here represent an intermediate scenario, namely, long-term, terrestrial UV solar exposure.