SDN 0003.26  -  Report on IR Reflectance of Anodized Samples

1.       Introduction

This report describes near-IR reflectance measurements of a number of anodized aluminum samples to determine, with some degree of control, the degree to which various anodizing processes produce "black" coatings in the infrared.

2.      Samples and Testing

Seven samples were prepared, all 2 inch square 6061 Al 0.25 inch thick.  All samples were manually polished with Simichrome to give a reasonably specular surface.  Since the IR reflectance attachments used with our laboratory spectrographs do not distinguish between absorbed and scattered light, the polished sample was intended to minimize scattering.  Four of the samples were black anodized with various processes, one was clear "Alodined" (also known as Chem-Film; iridite is a similar process) and at least one was an unprocessed polished piece.  To ensure a fair test, the samples were numbered, but not identified as to the anodizing process, so they will be referred to by sample number in this report (however, the experimenter was able to determine that samples 1, 2, 3, and 5 were black anodized).

The samples were measured on the Lambda-9 near infrared lab spectrometer over the range 600 - 2500 nm and on the Beckman 4260 spectrometer over the range 2500 - 15000 nm.  Both runs utilized a single-reflectance attachment, although the actual mechanisms were different for the two spectrographs, leading to some discontinuity at the overlap wavelength.  A silver mirror (r ~ 0.99) was used as the baseline reference for both instruments.

3.      Results

The results are plotted in Figures 1 - 3.  Figure 1 plots only the 600 - 2500 nm region obtained with the Lambda-9 spectrometer, since this region is highly compressed in the other plots.

The three polished samples (4, 6, and 7) are all rather similar in their behavior.  The sharp falloff in reflectance short of 2 microns is almost certainly an effect of the surface quality, and it is not possible to ascribe the differences between the samples to anything other than the inevitable spread in the surface quality resulting from manual polishing.  Sample 4 (which had the Chem-Film treatment) does drop more quickly than the others short of 1 micron (this can be seen more clearly in Figure 1).  This could be attributed to the Chem-film treatment, although such a conclusion would be tentative at best.  Note that this sample is intermediate in properties between the two untreated samples, so the Chem-Film is clearly not changing things very much.

The black anodized samples fall into two general areas.  Samples 1 and 3 (which were hard black anodized) had significantly lower reflectance at short wavelengths than 2 and 5 (which were sulfuric anodized).  In fairness, it must be noted that these samples had a much more matte appearance, so it is possible that the lower measured reflectance results in part from higher scattering.  All four samples reflect about 15% of the light at 15 microns, a wavelength at which we should be measuring only the absorbing properties of the anodizing.

The sharp discontinuities at 2.5 microns may result from the different experimental setup on either side of this wavelength; the effect is smaller for the shinier samples 2 and 5 (and is undetectable for the highly polished samples).  The Beckman may have a slightly larger acceptance angle for small angle scattered light.  The deep absorption bands at 2.6 and 6 microns are real and may represent H₂O within the anodizing.  The deep absorption band between 8 and 12 microns is consistent with Al₂O₃ absorption.  Even a thin (0.25 micron) layer will absorb significantly at an incidence angle of 45°, although the absorption should be very small for normal incidence.  The reflectance attachment in the Beckman operates at an incidence angle of 20°, which may be sufficient to produce the large absorption in the 10 micron region.  Interestingly, the apparent noise in the 4 - 6 micron region in samples 2 and 5 is actually due to well-developed interference fringes, probably within the anodize layer itself.

The black sulfuric anodized samples 2 and 5 actually appear visually darker than samples 1 and 3 on the unpolished back surface.  Keeping in mind the inability of the experimental setup to distinguish absorption from scattering, we measured the apparent reflectance of all four black anodized samples on the back side at 2.5, 2.8, and 4.5 microns.  The idea was that the back unpolished surface might result in similar scattering properties for all of the samples, with the difference being due to spectral behavior of the anodizing itself.  Since the intent is to not bead blast or intentionally roughen the surfaces of anodized parts in GNIRS, the back surfaces of the samples are similar to those which will be in the instrument.

The machining marks on the back surfaces give a "grain" which resulted in a relative difference in the reflectance of a few percent depending on the orientation.  The values in the table are the average of the two orientations.  The polished surface values are presented for comparison.

The relative back surface reflectances at 4.5 microns are different for the hard black and black sulfuric samples, although the contrast is less.  Within the absorption band at 2.8 microns, the back surface reflectivity was identical for all samples.

4.      Conclusions

The polished samples (4, 6, and 7) differ only slightly, within the expectations of variations resulting from the manual polishing of the surface.  The Chem-film treatment of sample 4 may be evident in the slightly lower reflectance short of 1 micron, but this is a tentative conclusion at best.

The black anodized samples reflect varying amounts (maximum 5 - 20 % in the region short of 5 microns).  The reflectance correlates with the visual appearance of the samples, so it is possible that the sample preparation is a factor.

The back (machined but not polished) surface reflectivity showed less variation from sample to sample, although that of the black sulfuric samples was still somewhat higher than that of the hard black samples.  The back surfaces of the hard black anodized samples appeared slightly lighter to the eye in diffuse illumination, reinforcing the conclusion that there is more scattering from these samples.

It is clear that hard black anodize results in less specular reflection from the surface, but it appears that this is largely because the surface (with a thicker layer of anodizing) is rougher. The measurements are consistent with total reflection plus scattering for both surface treatments being similar (if there is some influence of added roughness on the back surface measurements), or with the hard black anodize being slightly better.

For applications in GNIRS where we are specifically trying to control stray light – baffles, in other words – we have chosen to use black paint (Aeroglaze Z306: see SDN 003.24). The black anodize treatment is therefore intended to mainly control light that is already diffuse, and thus where reflection and scattering are roughly equivalent. For these applications, the results imply that either treatment will work about as well.

 Dick Joyce

10 October 2000

revised 13 October 2000

Figure 1.  Reflectance spectra of the samples over the range 600 - 2500 nm.  The peak near 800 nm is probably real, although the raw reflectance data have a large discontinuity at this wavelength, where a grating change occurs within the spectrometer.  Of the polished samples, #4 was Chem-film treated.

Figure 2.  Reflectance of the three polished samples over the entire 600 - 12000 nm range.  Sample 4 has been Chem-film treated.

Figure 3.  Reflectance of the four black anodized samples over the range 600 - 15000 nm.  Samples 1 and 3 were hard black anodized, 2 and 5 were black sulfuric anodized.  The apparent noise in the spectra of samples 2 and 5 in the 4000 - 7000 nm range is actually due to interference fringes, presumably in the anodize layer.

If you have any questions or suggestions regarding this website, please contact Melissa Bowersock.               

National Optical Astronomy Observatories, 950 North Cherry Avenue, P.O. Box 26732, Tucson, Arizona 85726,
Phone: (520) 318-8000, Fax: (520) 318-8360