1   Summary

• This report is first a description to achieve clean CCD cryostats and to avoid therefore contamination on CCD detectors.
• Second some cryostat materials are proposed.
• Third a cleaning procedure is suggested which has to be applied to all parts before integration.
• Fourth a baking procedure is described which will be done after all integrations and tests and before shipping the cryostat to a consortium or to the telescope and which can be done periodically in the observatory, if a problem occurs.

2   Introduction

Contamination on CCD detectors first is showing up as a loss of transparency in the UV and blue wavelengths bands. Later layers are visible for the naked eye up to very ugly greasy deposits which are stable up to + 35 °C inside the vacuum.

3   Recommended materials inside vacuum CCD cryostats

The best material is stainless steel (see   list of materials). It is a physical law, that under atmospheric conditions all surfaces are covered more or less with gas, which consists mainly of water molecules. Water molecules cause no critical contamination like oil, but they are hard to pump.

One should avoid anodised surfaces and plastic materials if possible.

Nevertheless for the temperature range envisaged for a dewar, the choice of material is not too difficult.

Some materials should be avoided when designing a CCD tank, especially specific out-gasing heaters and plastic shrinking sleeves ( at soldering points ).

One should not hesitate to test suspicious materials before using it at the cryostat tank and head, especially if one find signs of contamination near suspicious materials. Tests are possible in a very clean cryostat monitoring CCD contamination over 1 - 4 weeks or with a mass spectrometer combined with a gas chromatograph.

All used self-gluing tapes should be avoided from the design.

Sorption pumps should be cooled to at least - 190 °C.

The temperature control of the CCD should be programmed in a way that during warming up the CCD is heated. The CCD should never be the coldest point inside the cryostat.

The use of fibre glass parts was investigated and it seems to produce no significant contamination.

Only ultra high vacuum grease from Balzers should be used if necessary, which has a partial pressure of less than 10^-12 mBar at 20 °C.

What is most important is a low absorbency surface ( not anodised ) and the ability to remove low vapour pressure contaminants from that surface. A raw machined surface inevitably includes "rolled over" surface features which trap machining oil etc. This is best removed by de-greasing followed by acid etching, to give an open surface. Provided this surface is kept clean ( no finger grease ) it will not be a significant out-gassing source.

Also electropolishing reduces the out-gassing surface. Electropolishing has to be done before our described cleaning procedure.

4   Investigation of cryostat materials

1. The use of a good mass spectrometer in combination with a gas chromatograph to be able to analyse individual new materials, but also the the out-gassing of the whole cryostat and especially the grease on a contaminated CCD.
2. Complete disassembly of one cryostat and measuring ( out-gassing and mass spectral analysis ) of the individual cryostat parts.

5   Cleaning of parts before integration of cryostat

The cleaning of the components should be done first by hand with a tissue and acetone or alcohol, then in a ultrasonic bath with acetone.
( Some people propose water with a detergent. )

A second washing should be in alcohol. Also distilled water is under discussion, because it is chemically the best solvent.

The same should be done with the electronic boards after soldering and all the wiring.

Afterwards the components are baked in a vacuum oven up to appropriate temperatures depending from the materials ( at least to 60 °C ). The parts are stored in sealed plastic foils.

(See   treatment of cryostat components.)

6   Handling of cleaned parts

(See   list of needed equipment.)

An exception could be the installation of the CCD detector into the cryostat head. In this case safety has priority above cleanliness. Everything is integrated under cleanroom condition class 100.

7   Baking of complete integrated system before delivery

To achieve the cleanest results we do the following:
( Point 1 and 2 are not essential and can be left out. )

1. Wrapping of the Cryostat with ISOPAD 220 Volt heater bands ( approx. minimum of 160 Watt and 2 meters long ). The bands are fixed very tight with special heat resistive strings from ISOPAD. The cryostat should be thermally isolated from a stand as much as possible in order to allow proper baking.
2. On the cryostat 2 cm near a heater strap a PT100 sensor is placed in order to control the baking temperature. For this an ISOPAD baking controller is used, which is adjusted to + 60 °C. A second security measure is the power limitation. If a heating strap with more than 150 W is used, it is recommended a reduced voltage with a transformer and a simple power meter to check this. It is worth to invest these 100 Euro more in order not to burn a CCD for 150.000.- Euro!
3. The cold-plate of the cryostat is heated ( if possible ) to + 50 °C and the CCD to + 50 °C. For this heater we need an additional thermal fuse installed nearby inside the cryostat tank like in the case of the CCD detector assembly plate. The sorption pump heater should be adjusted to + 100 °C in the case of a Bath Cryostat ( if the adsorbent is zeolite, 100 °C are not sufficient, 200 - 350 °C are needed ) and to + 60 °C in the case of a CF-Cryostat. This happens, because the sorption pump of a CF-Cryostat is thermally to strong connected to the cold-plate. This should be improved very soon, because otherwise a proper regeneration of the sorption pump in CF-Cryostats is not possible.
4. After half an hour pumping is started.
5. This baking and pumping process should have a duration of 3 - 5 days.
6. This baking process should be properly logged. Parameters like date, duration, temperatures versus time, location, vacuum pressure versus time and used materials are interesting.

8   Periodic baking at the telescope site

If the cryostat was properly cleaned as above described and some bad materials have been replaced a re-baking and -pumping of every 12 months should be sufficient in the observatory.

Back to   "Ultraclean CCD project"

---

Send comments to   <odt@eso.org> Last update: Mar 24, 2004 Written by   Sebastian Deiries, webdesign by   EVI