The TAO project

The report of performance evaluation test of the mirror coating system for TAO

What is a mirror coating system?

An astronomical telescope collects light from astronomical objects and focuses it onto a sensor. In other words, it acts as a light collector to reveal various physical properties of the objects. Since light from the objects is extremely faint, it is important to collect it as efficiently (without losses) as possible. Therefore, mirrors and lenses used to collect light must have high reflectivity and transmittance. Especially in the case of reflecting telescopes such as the TAO telescope, reflectivity decreases over time due to contamination or corrosion of the mirror surface caused by long-term observation. Then, it is necessary to periodically remove the old reflective coating and apply a new one. The mirror coating system plays this role.

The mirror coating system primarily consists of a vacuum chamber large enough to accommodate the entire mirror, a vacuum pump to maintain a high vacuum inside the chamber, an ion bombardment device to clean the mirror surface with ionized gas, a lot of filaments pre-impregnated with a metal film, and a control unit to manage these components. In the case of the system of the TAO telescope, a distinctive feature is that the primary mirror cell is "sandwiched" within the upper and lower parts of the chamber. In other words, the mirror cell itself serves as part of the vacuum chamber.

▲Layout Diagram of the mirror coating plant for the TAO 6.5-meter telescope in the operations building

Overview of mirror coating test

The various components of the mirror coating system were manufactured both domestically and overseas. The main coating chamber passed structural inspections-including checks on its shape and functionality as a vacuum vessel-and was then transported to Yokohama where the site of the comprehensive coating test. In addition, vacuum pumps, control units, and other equipment were gradually delivered to Yokohama upon completion. In the first half of 2020, these components were installed, and piping and wiring connections were made based on the original design. Subsequently, startup and basic operational tests were conducted on each piece of equipment, followed by comprehensive testing of the system combining all components. (Please refer to this news.)

This article will describe the details of results of the test. The test items are as follows: (1) measurement of attainable vacuum level as a vacuum chamber and needed time to reach it, (2) selecting the fill gas required for ion bombardment-performed immediately before firing after mirror cleaning to remove the oxide film from the mirror surface and ensure molecular-level cleaning-and searching the optimal current and voltage parameters, (3) determining the optimal parameters-such as the applied current and voltage, timing and duration of firing (metal evaporation), and so on.

▲Photograph of the comprehensive test

Vacuum test

It is known about evaporation that the higher the vacuum, the better the film adhesion and the higher the reflectance. To achieve a high vacuum sufficiently in a short time, a hybrid system consisting of four rotary pumps, four turbo molecular pumps, and seven cryopumps is used. It has been confirmed that a vacuum level suitable for ion bombardment can be reached within one hour, and a vacuum level suitable for firing can be achieved in less than half a day.

▲Results of vacuum test: the time variation of vacuum level. (see Takahashi et al.2020) A graph showing the time elapsed since the start of the vacuum pumping process and the pressure (vacuum level) inside the chamber. After reaching a vacuum level suitable for ion bombardment (<1 hour), the vacuum level temporarily rises (40 minutes to 2 hours) due to the ion bombardment process. It can be seen that firing started after approximately 6 hours. Note that in case the primary mirror is installed, the upper and lower parts of the chamber are separated into high-vacuum and low-vacuum zones by a vacuum isolation cover and an inflatable seal.

Ion-bombard test

For the mirror coating test, a sample stage modeling the curvature of the primary mirror was installed in the lower chamber, and film status was monitored at 108 positions using a total of 27 sample glass plates (arranged along two axes perpendicular to each other and within a 90-degree range to account for symmetry). After coating, the adhesion strength of the films was evaluated via a tape test, and the film thickness and reflectance of the sample mirrors were measured. One of the steps in the vapor deposition process is called ion bombardment. In this process, the gas sealed inside the chamber is ionized, and the ions are then struck against the mirror surface by a voltage difference applied to the electrodes, thereby removing microscopic contaminants and oxides from the mirror's surface. If this process is not performed correctly, the metal film will not adhere firmly to the mirror and may peel off after deposition. In this ion bombardment test, we were finally able to obtain bombardment parameters that passed the tape test at all sampling locations. Finally oxygen is the gas selected for this use.


▲(left) Sample glasses installed in the chamber (right) A view inside the chamber during ion bombardment. You can see the oxygen ionizing and glowing with a whitish light.

Mirror coating test

The TAO coating system uses four power sources to control 147 filaments. In order to efficiently and uniformly evaporate the aluminum impregnated in the filament, the current (voltage) is applied in three parts: pre-heating, first ignition, and second ignition. Results of multiple deposition tests showed that the thin portions of the film are at least 100 nm thick, and observations at infrared wavelengths can be performed without problems (some portions are slightly thicker than specifications, though). It was also found that uniformity of the film could be achieved by adjusting the amount of impregnation of the filament to different values for the inner and outer circumference.

▲During firing process: the chamber is filled with dilute aluminum vapor.

▲Distribution of coated film thickness of sample mirrors (27 sample mirrors at 108 positions). We found that we could get the expected results. (see Takahashi et al. 2000)

▲Result of mirror reflectance: sufficiently meets the specification values (more than 90% at 570 nm and over 900 nm)

Following this test, the vapor deposition plant was disassembled, packed, and shipped to Chile. After a voyage of about one month, it arrived safely in Chile and is currently stored at the 5,000-meter base of Co. Chajnantor and in a warehouse in Calama. Once the operations building is completed, we will finally proceed with its installation at the summit, along with the other cleaning equipment that has been developed. After that, the 6.5-meter primary mirror will finally be applied coating, in order to collect light from object in the distant universe.