Visit to the United States of America - October 1998

Willie Koorts

What prompted this trip?

It seems like ever since we got involved with CCD's, we were plagued by contamination problems, resulting in the in-house phrase: "CCD is a four letter word"! A typical contamination occurrence mostly acts like a snake striking at any time without you even knowing where it came from.

When I first got involved in CCD work, preparing TEK7 back in 1994, it happened again with the chip developing green droplets after releasing the vacuum for the first time. In wanting to gain a better understanding of the problem, I started reading every book on vacuum technology available from our library, made numerous phone calls and sent e-mails to experts locally and around the world and even attended a vacuum course in Johannesburg. (By the way, it was very interesting to see how many other people had similar skeletons in their cupboards!) Although we gained more insight into the problem and adapted our cryostat preparation, cleaning and vacuum pumping procedures accordingly, the problem would still occur from time to time for no apparent reason. At some stage it seemed that the more we clean, the more contamination problems we got!

When we finally had success with the spectrograph CCD in 1996, we thought we then had it under control by starting out with a new cryostat, a cleaning regime that can cope with the cutting fluids used on the machined parts made by our workshop and complete oil-free pumping. Great was our disappointment when we had exactly the same problems back with the very next system we prepared, while using the same exact strict preparation procedures!

It was at this stage when the applications for overseas trips for the next year had to be submitted and Geoff had no hesitation in asking for a trip to the US for me to visit other groups, involved in the same thing. By this time we also had approval for building a dedicated CCD development laboratory and we were already looking at a replacement for our aging Merlin CCD controller.


The trip was approved and in planning it, I tried to fit in as many different CCD development groups as possible, trying to cover the full spectrum of CCD developers. My final itinerary then looked as follows:

  1. Infrared Laboratories, Tucson, Arizona:
  2. They are the suppliers of most of the Infra-Red dewars used on our telescopes at present as well as the two new cryostats most recently bought for imaging CCD development. Apart from manufacturing cryostats in all shaped and sizes, they have also been involved in building complete camera systems, particularly in the Infra Red, based on the SDSU (San Diego State University) Leach controller. Optical design is also part of their portfolio.

  3. Spectral Instruments, Tucson, Arizona:
  4. Spectral makes complete camera systems, aimed mainly at the scientific community, with quite a few systems supplied to the astronomical community before. They specialize in the development and manufacture of CCD based analytical instrumentation and detector systems. Their spectrophotometers and cameras are of a very high standard, on par with what professional institutes like us are building. Cooling options range from Liquid Nitrogen (LN2) and Closed Cycle to Peltier with systems built around almost any chip - I saw 2k TEKs and 4k Lorals.

  5. Apogee Instruments, Tucson, Arizona:
  6. Apogee builds a range of very affordable, compact, medium performance, high speed cameras, aimed at the lower end of the professional market or where high-speed cameras are required, but have a very impressive list of previous costumers, including astronomical groups and NASA. They claim to get the best performance out of Kodak chips, to the point where Kodak now make use of them to test their new chips. All systems are 2-stage Peltier cooled only.

  7. Steward Observatory CCD Laboratory, University of Arizona, Tucson - Michael Lesser:
  8. Apart from very specialised wafer testing, mounting and coating, they also run up CCD systems for use "on the mountain" like all other technical astronomical support groups. Seeing their wafer processing labs were very impressive and I gleaned a lot of useful information from looking in detail at their cleanrooms. Like us, they also build up systems by adapting standard InfraRed Lab. cryostats for smaller CCDs and they have long since standardised on the SDSU CCD controller. Very impressive, was seeing the big mosaics and Mike also showed me the world's largest chip made so far.

  9. NOAO (National Optical Astronomy Observatories), University of Arizona, Tucson:
  10. NOAO constitutes of CTIO (Cerro Tololo Inter-American Observatory), KPNO (Kitt Peak National Observatory), NSO (National Solar Observatory) and SCOPE (NOAO's SCience OPErations Division). Here I was able to learn from the master, Bill Ditsler, a vacuum engineer who specialises in CCD and cryogenic systems as well as cleanrooms. Bill Ditsler is capable of preparing a CCD system that will hold its vacuum well enough to last a year on only one pump-down, cooled or uncooled! He is the only person who I know of with a clean record as far as contamination is concerned. His lab is not even that impressive, but he gets these results by knowing exactly what he is doing and keeping total control of every step along the way. To copy what he is doing would also involve acquiring an electro-polishing plant. Nobody else I visited went to such lengths.

  11. Kitt Peak National Observatory, Tucson, Arizona:
  12. Bill arranged for me to take a trip up the mountain and I was delighted to be shown around the observatory, including seeing the giant 4-m telescope and the modern 3.5-m WIYN. What struck me as interesting, was the very similar set of technical problems and ways of solving them that we are facing in the day to day running of our observatory.

  13. Apache Point Observatory, Sunspot, New Mexico:
  14. Visiting APO on my way to McDonald Observatory initially came about because of a personal interest in a little research project into the 1882 Transit of Venus, observed from my birthtown and hometowns respectively. What was most impressive here is the Sloan Digital Sky Survey, a 2.5-m, 3 degree field Telescope, dedicated to one particular task. It uses two main instruments namely a camera consisting of 30 Tektronix/SITe 2048 x 2048 CCDs arranged in a 5 by 6 array and a 640-optical fibre fed spectrograph. The other big telescope on site is a 3.5-m ARC (Astrophysical Research Consortium) telescope. Of particular interest here was the fact that everything was new and modern and by looking at the layout of their detector lab, I noticed a number of ideal requirements that should be included in our new proposed CCD lab.

    In the little observatory town of Sunspot where we stayed (down NM State Highway 6563 - the wavelength of H-alpha!) something not as new, was looking in at the National Solar Observatory of Sacramento Peak. This setup with its impressive 100m, 250-ton rotating vacuum tower solar telescope, was originally set up there by the US Air Force in 1969.

  15. McDonald Observatory, Texas :
  16. Here I first had a tour of the "old part" of the observatory but the main attraction was seeing the massive Hobby-Eberly Telescope (HET) in action. I spent some time with the technical operations crew, responsible for the running the telescope which was still in its engineering phase. A lot of useful technical information was exchanged, some of which was already found relevant, even in this early stages of SALT.

  17. Department of Astronomy , University of Texas at Austin, Texas:
  18. My final stop was to visit Phillip MacQueen at the support group for the McDonald Observatory. They run a very active instrument development group, doing a lot of their own design. Seeing their well laid out lab and discussing their cryostat preparation procedures as well as their in-house built CCD controller was very fruitful.

Weing up

Looking back and summarising the techniques, procedures, etc. used by such a vast array of divergent CCD system developers is very interesting. What struck me was how different groups, sometimes just across the street from each other or in the same town, are trying to solve similar problems in complete isolation. In a lot of cases they converge on the same answer, but sometimes not. Comparing their ways with ours, made me realise that we were not doing anything grossly wrong and that the secret must lie in some minute bit of detail that we keep overlooking. This is the advantage of visiting in person above exchanging hundreds of e-mails, which payed off in the end as will be seen later.

It became clear to me that, except for Bill Ditsler, everyone at some stage had some form of contamination problem and most people managed to solve theirs by "starting to do things right", by applying sound vacuum principles. This was exactly what we did too, but our problems stayed - or at least seemed to come and go! I started looking seriously at the differences and similarities between the different groups and what we have been doing.

Comparing differences

It was interesting to find that in some areas, complete opposite schools of thought exist. For example, there is one school that states that vacuum systems using o-rings are vacuum limited by the permutation of the o-rings since "they leak like sieves" and that it is useless trying to pump a good vacuum since the vacuum would soon after taking the system off the pump, have deteriorated. They would therefore not bother to try and pump a good vacuum, but rely on the sorption pump to maintain the vacuum while the system is cold. The opposite school says that o-rings make near-perfect seals and by getting the system outgassing is under control, the vacuum should stay good for a long time. The sorption pump thus merely acts by taking care of the remainder of the outgassing that might still exist. We have always been of the latter opinion.

Another area where differences were noted, was in the choice of sorption material to be used. Some swear by activated charcoal, mainly because of ease of regeneration where others prefer zeolite. It seems generally agreed that charcoal gives up most (if not all) of the gasses pumped while cold, as soon as it warms up back to room temperature, while zeolite needs to be regenerated at elevated temperatures. One group even used both charcoal and zeolite - the charcoal being on the cold surface becomes the getter while the zeolite only acts as a desiccant for the water vapour that gets released by the charcoal on warm-up. A whole range of regeneration temperatures for zeolite are used by different groups ranging from as low as 120C up to 360C. Although experimenting with charcoal recently, we have always used zeolite, regenerated at 280C under vacuum.


There seems to be general agreement that Viton o-rings are the best and that systems should be assembled "dry", i.e. no vacuum grease should be used on o-rings. Everyone avoids using heatsink compounds for thermal contact. This all still agreed with the ways we adopted so far. By comparing the materials used for manufacturing of the parts used inside the vacuum, no great differences between us and them were found. Except for Bill Distler, who use LN2 trapped oil-diffusion vacuum pumps (what we used to do), everyone else use oil-less turbo-molecular vacuum pumps - one of the most expensive options we explored too.

By comparing cleaning procedures, everybody uses gloves to handle and assemble cleaned cryostat parts - same as us. Comparing their recipe for cleaning, a great variety was found, as can be expected, but the main difference between theirs and ours was that people generally finish off with a solvent wash as their final stage. We could never get an except able finish in this way and therefore always followed our final solvent wash with a laboratory soap and water wash and a hot distilled water rinse. This is still an except able practice according to some vacuum books.

Something that we never did seriously but which soon became very notice able where ever I went, was that everybody vacuum-baked parts and/or complete systems after cleaning. This ranged from just wrapping complete systems in heating tape to proper vacuum-baking ovens. Since our very first attempt at baking turned into a disaster (the CCD temperature control path became unstable) we became reluctant to try much more although we subsequently experimented again, using a small fan heater on the completed Spectrograph CCD, the system that never showed contamination! Could this be the missing link? Still, what is causing the contamination then and where is it coming from?

Problem solved?

With the last system where we had this trouble, some contamination formed on the cryostat window as a very slight condensation. We removed and kept this window and after getting back, I managed to find someone who could analyse such a small quantity on a mass spectrometer for us. The immediate result yielded some answers but also more questions - the good news was that we now knew what it was and how to get rid of it but the bad news was that we still did not know its origin. The substance was identified as a plasticiser - Di-octyl phthalate (C24 H38 O4) to be exact! To get rid of it, is by vacuum baking - the same process used by the mass spectrometer to analyse it - but we would still like to avoid it forming in the first place!

To find information on this plasticiser proved difficult and finally after an extensive WWW search, I came across the Environmental Health Centre's Chemical Backgrounder pages (URL below) which described it as a colourless, oily liquid - insoluble in water and soluble in most organic solvents. It is used in the manufacturing of vinyl gloves used for medical examinations and surgery and also vacuum pump oil!

A few years ago, before we changed to oil free pumps (for this very reason), I would have immediately blamed it on the vacuum pump oil - which we did and probably was responsible for some early cases of contamination. Eliminating this option, immediately shifted the blame to the vinyl gloves we changed to lately. Looking back at old invoices during the time we prepared the Spectrograph CCD revealed that we were using Latex gloves then. Adding all the clues, I then realized that the problem arises as we take parts in and out of solvents, using the vinyl gloves. The plasticizer then gets dissolved in the solvent, contaminating it and the parts that follow. Since it is insoluble in water, the subsequent water wash does not get rid of it. This also explained why we could never get an acceptable result with a final solvent wash and why everyone who did vacuum baking could solve their problems (without even realising how).

Remembering back, watching Bill Ditsler prepare a cryostat, although wearing gloves, he was actually handling the parts using stainless steel tongs all the time. This tiny bit of information was never conveyed in all the numerous e-mails we exchanged over the years, probably since it seemed so insignificant!

Impressions of America

Personally, being my first trip ever, just the trip to a foreign country in itself was an experience. I took some leave to explore a little of the country while there. After all, what is any US trip without seeing Disney, Hollywood and Universal Studios (highly recommended!)? A trip through the breathtaking views of the sandstone formations of Southern Utah and later the Grand Canyon, spending one night in the lights of Las Vegas on the way, was well worth it. Seeing the "Bone Yards" with row upon row of de-commissioned aircraft and the vast range of rocket and space equipment on display at the Pima, White Sands and Alamogordo museums, demonstrated the economic strength of the country to me.

What I found amazing was the Halloween-hype that takes hold of the country before this big event - I still wonder what those dedicated Halloween shops do for the rest of the year? To an outsider it all seemed a little ridiculous. The extent to which the "drive-trough" culture had developed in America also seemed phenomenal - apart from the more understandable drive-through take-away restaurants, a drive-through pharmacy seemed less odd after coming across a drive-through lumber store! My boy made an interesting observation - after driving through all the shops, conveniently doing all your shopping from your car, when you need to fill up your tank, the Americans are finally forced out of their cars, having to pump their own petrol ... I mean, gas. Therefore, come to South Africa where we have all drive-through gas stations!

Web resources for organisations mentioned in the text.

Apache Point Observatory ;

Apogee Instruments ;

Environmental Health Centre ;

Infrared Laboratories ;

Kitt Peak National Observatory ;

McDonald Observatory ;

National Solar Observatory, Sacramento Peak ;

NOAO (National Optical Astronomy Observatories) ;

Spectral Instruments ;

Steward Observatory CCD Laboratory ;

University of Texas at Austin, Astronomy Department ;

1882 Transit of Venus from Wellington ;