Long Term Study of Aluminum Coatings for Astronomical Mirrors

Join OMIcoatings

Subscribe to
OMIcoatings

There are several types of aluminum coatings available for astronomical mirrors. These mirrors are used and stored in a variety of environments, which is expected to impact the longevity of aluminum reflective coatings. Environmental factors impacting this longevity include, but are not limited to, humidity, pollutants in the atmosphere and the storage environment.

Environmental factors often vary as a function of climate and storage practices. Examples of general climate types include; Desert Southwest (mostly dry, yearly monsoon), Mid-West (warm, humid summer, cold winter), Gulf Coast (hot, humid summer, mild humid winter). Some telescopes reside in permanent installations such as a dome or roll-off roof observatory. Some are stored in a trailer, garage, basement, or air conditioned house.

This experiment seeks to determine the general rate of deterioration of each of several coating types in various environments. Because of the diversity of environments, it is not expected that a hard rule for longevity can be established. However, it is our hope that the data obtained from various climates will allow us to arrive at a fairly representative life expectancy and to provide additional information on the effect of a particular climate. We also hope to gauge how much the life expectancy should be expected to differ from the average due to a particular environment.

Hypotheses

• Mirror coatings located in dry climates will last longer than coatings in humid climates
• Mirror coatings located in areas with low pollution will last longer than coatings in high pollution areas
• Mirror coatings located in non-salt environments will last longer than coatings in high salt environments

Assumptions

• Glass-type and polishing will be uniform enough to not affect the longevity of coatings
• Processes used for coating will be uniform within coating type

Aluminum Mirror Coatings

The reflective coatings most commonly used on astronomical mirrors feature a base layer of pure aluminum deposited by evaporation/condensation in a vacuum chamber. We will study three types of aluminum coatings: bare aluminum, protected aluminum and enhanced aluminum.  Each of the three coatings will be studied with and without the inclusion of Ion Beam Assisted Deposition (IBAD), for a total of six experimental conditions.

Bare Aluminum

The simplest mirror coating, bare aluminum, is often used on large mirrors at professional observatories that have the facilities to coat mirrors on site. Bare aluminum offers a known spectral reflectance, which is useful in spectroscopic studies. The downside of bare aluminum is that it is delicate and prone to rapid oxidation. Oxidation of the aluminum coating significantly reduces the reflectivity of the mirror. Bare aluminum coatings must be stripped and reapplied frequently to maintain good reflectivity.

Protected Aluminum

Because of its tendency to oxidize and the resultant loss of reflectivity, the aluminum layer is often over-coated with a protective layer of silicon dioxide (SiO2). The SiO2 layer slows the oxidation process by protecting the aluminum from moisture and other atmospheric contaminants. In addition, the abrasion resistance of the SiO2 layer allows for periodic cleaning of the mirror surface to remove dust and other contaminants. A clean mirror provides higher reflectivity than a dirty mirror, therefore the ability to clean the surface as needed is a significant factor in obtaining maximum reflectivity from an aluminum coating.

Enhanced Aluminum

An enhanced aluminum coating provides the same protective benefits and cleaning advantages as a protected aluminum coating. It provides the added benefit of a significant increase in the reflectivity of the mirror across a broad spectrum. The simplest enhanced aluminum coatings consist of two layers on top of the aluminum. The first layer is a low index of refraction material such as SiO2. The second layer is a high index material such as tantalum pentoxide (Ta2O5). The constructive interference of light in these two layers provides the reflectivity enhancement.

Ion Beam Assisted Deposition (IBAD)

A recent advance in the quality of reflective coatings has come in the form of the ion mill. The ion mill provides a beam of high energy gas molecules that are employed during the coating process to clean the mirror surface prior to deposition. This ion cleaning process promotes adhesion of the aluminum to the substrate material. The ion beam also provides extra energy during deposition to compact and densify the coating materials as they are deposited. Finally, the ion beam is used to drive oxidization reactions to completion during deposition, resulting in greater purity in the oxide layer (SiO2 or Ta2O5) and a consistent index of refraction. It is speculated that increased layer density achieved with IBAD will provide better protection to the underlying aluminum layer. This speculation is based on the idea that a less porous coating should reduce the rate of oxidation of the aluminum. In addition, a more dense coating will absorb less moisture. The absorption of moisture in coating layers causes a shift in refractive index and increases absorption of light, both of which affect the reflectivity and longevity of the coating.

The Specific Aluminum Coatings in the Study

We will produce samples of each of the above listed coatings with and without IBAD, using the following production methods:

IBAD-Bare Aluminum: Prior to loading in the coating machine the mirror will be cleaned with soap and water. It will then be rinsed with deionized water. After the mirror is loaded in the chamber and the chamber is evacuated the mirror will receive two ion beam cleanings (scrubs) prior to deposition of the aluminum layer. The gas used in the ion scrub is oxygen (O2). The ion scrubs will remove adsorbed moisture from the mirror surface as well as residual organics not removed during the pre-cleaning process mentioned above. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited to a thickness of 120 nm at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the IBAD-Bare Aluminum coating. The back of each sample will be engraved with the identifier IBAD-Al and a serial number.

Non-IBAD-Bare Aluminum: The mirror will receive only the above mentioned pre-cleaning process that is applied before the mirror is loaded in the coating chamber. The ion beam scrub will not be applied prior to deposition of the aluminum layer. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited to a thickness of 120 nm at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the Non-IBAD-Bare Aluminum coating. The back of each sample will be engraved with the identifier N-IBAD-Al and a serial number.

IBAD-Protected Aluminum: The mirror will receive the pre-clean and two ion mill scrubs prior to deposition of the aluminum layer. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. Subsequently, one protective layer of SiO2 will be applied to 1/2-wave thickness. The ion beam will be employed during deposition of the SiO2 layer. The background pressure during deposition of the SiO2 layer will be 1E-4 of pure O2. The O2 background and O2 ion beam will produce a highly pure and dense layer of SiO2. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the IBAD-Protected Aluminum coating. The back of each sample will be engraved with the identifier IBAD-PAL and a serial number.

Non-IBAD-Protected Aluminum: The mirror will receive only the pre-cleaning process that is applied before the mirror is loaded in the coating chamber. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. Subsequently, one protective layer of SiO2 will be applied to 1/2-wave thickness at a rate of 2 nm per second. The ion beam will not be employed during deposition of the SiO2 layer. The SiO2 will be applied at a pressure of 5E-6 Torr or lower. There will be no background of O2 as this gas is provided by the ion beam which remains off during the non-IBAD process. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the non-IBAD-Protected Aluminum coating. The back of each sample will be engraved with the identifier N-IBAD-PAL and a serial number.

IBAD-Enhanced Aluminum: The mirror will receive the pre-clean and two ion mill scrubs prior to deposition of the aluminum layer. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. Subsequently, an enhancing layer of SiO2 will be applied to 1/4-wave thickness at a rate of 2 nm per second. The ion beam will be employed during deposition of the SiO2 layer. The background pressure during deposition of the SiO2 layer will be 1E-4 Torr of pure O2. A final layer of Ta2O5 will be applied to ¼-wave thickness at a rate of 3 nm per second. The ion beam will be employed during deposition of the Ta2O5 layer. The background pressure during deposition of the Ta2O5 layer will be 1E-4 Torr of pure O2. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the IBAD-Enhanced Aluminum coating. The back of each sample will be engraved with the identifier IBAD-EAL and a serial number.

Non-IBAD-Enhanced Aluminum: The mirror will receive only the pre-cleaning process that is applied before the mirror is loaded in the coating chamber. The ion beam will not be employed during deposition of the aluminum layer. The aluminum will be deposited at a rate of 10 nm per second at a pressure of 5E-6 Torr or lower. Subsequently, an enhancing layer of SiO2 will be applied to 1/4-wave thickness at a rate of 2 nm per second. The ion beam will not be employed during deposition of the SiO2 layer. The SiO2 will be applied at a pressure of 5E-6 Torr or lower. There will be no background of O2. A final layer of Ta2O5 will be applied to ¼-wave thickness at a rate of 3 nm per second. The Ta2O5 will be applied at a pressure of 5E-6 Torr or lower. The ion beam will be not employed during deposition of the Ta2O5 layer. There will be no background of O2. One coating run of 100 sample slides of dimension 1 inch x 3 inches will be produced with the non-IBAD-Enhanced Aluminum coating. The back of each sample will be engraved with the identifier N-IBAD-EAL and a serial number.

Samples from each of the above coating runs will be tested for abrasion resistance (rub test) and adhesion (peel test) utilizing Mil-spec test methods. The rub test is performed by applying twenty rubbing strokes across the mirror surface with one pound of weight on a lint-free cotton cloth with an area the size of an eraser head. Both IBAD and non-IBAD bare aluminum coatings are expected to show abrasion marks as a result of the rub test. All protected and enhanced coating types are expected to pass the rub test (no abrasion marks). If a protected or enhanced aluminum sample fails the rub test the entire batch of that particular type will be recoated. The peel test is performed by applying cellophane tape to the mirror surface then quickly peeling it off. If the coating peels off with the tape it fails the test. All six coating types are expected to pass the peel test. If a sample fails the peel test the entire batch of that particular type will be re-coated.

The Study

We will assemble kits containing one of each of the coating samples listed above (six samples per kit). The sample kits will be distributed to study participants. Our goal is to have the samples live with the participant’s telescope, experiencing the same environmental conditions that the telescope mirror experiences. For example, when the telescope is covered or in storage then the coating sample kit should be covered and stored with the telescope. When the telescope is in use with its mirror exposed then the samples in the kit should be exposed. If the participant cleans their telescope mirror they should also clean the mirrors in their kit. Study participants will receive explicit instructions on mirror cleaning and other aspects the study.

The entire sample kit will be retuned to us periodically for inspection and reflectivity measurement. We will take reflectivity measurements of the samples prior to and after in-house cleaning. This will allow us to determine any loss of reflectivity due to age and environmental conditions for each coating sample. It will also allow us to determine the loss of reflectivity due to dust and other contaminants on the mirror and make some recommendations as to cleaning frequency.

As data is periodically gathered, it will be analyzed. The primary method of data collection will be the measurement of the mirrors reflectivity vs. wavelength using a calibrated spectral reflectometer. The results will be made available to the study participants and to the public. Availability of information will come through our web site and a Yahoo group established for ongoing discussion purposes.

The Results

1. As data comes in, we hope to answer the following general questions:
2. How does the reflectivity of each coating decrease with time?
3. How do environmental conditions affect the reflectivity of each coating?
4. How can one tell when it is time to recoat a mirror?
5. How does the reflectivity of each coating change with cleaning, both longitudinally and as a function of pre-cleaning vs. post-cleaning in a single instance?
6. Does glass type make a difference in longevity?

Experimental Challenges

Dewing

A telescope mirror has considerably more thermal mass than the coating samples. In the typical observing situation the temperature drops during the night. As a result, the mirror is almost always warmer than the ambient air. This helps to prevent the formation of dew on the telescope mirror.  The small mirror coating samples, due to their low thermal mass, will track ambient air temperatures more closely. Therefore, the samples are more likely to form dew. This increased exposure to moisture is expected to cause the aluminum coating on the samples to deteriorate faster than the coating on the telescope mirror. It is not known to what extent this will skew the results, however it can be expected that in a given environment, the type of aluminum coating on the telescope mirror should last at least as long as the same type of aluminum coating on an experiment sample.

Participation

This is a long term study that could run five years or longer. We will rely on the initial participation of up to 100 participants. It is expected that some participants will drop out or not send in their samples for evaluation in a timely fashion. However, because of the level of interest and participation of amateur astronomers it is expected that the drop out rate will be low and that a statistically significant number of participants will stay with the study long enough to obtain meaningful results.

Abnormal Mirror Handling

Participants may drop out of the sample pool as a result of problems with handling their kit mirrors. For instance, the impact of rodents, pets, or children damaging the mirrors will undermine the interpretability of the results.

Location/Storage Change

Some participants may move or change storage practices during the time of the study, thus exposing their kit to more than one set of environmental conditions. This will likely confound the results. These participants will have to be tracked separately so as to be parsed out at time of analyses.

Mirror Cleaning

Mirror cleaning can impact the longevity of coatings. As such, the results of the study are generalizable only to coatings that undergo the same time intervals between cleanings. The study also assumes that mirrors are cleaned correctly. Participants will receive explicit cleaning instructions.

Statistics

A longitudinal study with multiple data points per participant requires stats that necessitate large sample sizes. We may end up better served looking at descriptive statistics and drawing tentative hypotheses, rather than trying to make causal inferences.

Time Between Testing/Shipping

Because of differences in geography and participants having their own lives, there will likely be some variability both in the time between testing and how mirrors will be shipped. The directions will need to be quite explicit about time and shipping process. Because participants are likely to forget exact timelines, having a reminder e-mail, etc. would be helpful.

Invitation to Participate in the Long Term Mirror Coating Study

Readers are invited to participate in the long term coating study. If you are interested please join the Yahoo OMIcoatings discussion group at

http://tech.groups.yahoo.com/group/OMIcoatings

Participants are not required to have an OMI coating on their telescope mirror as the samples to be studied will be provided in the kit supplied by OMI.