UV FLOODING RECIPEES (fwd)

[Ralph Florentin]

Dear Olaf,
Here is some of our experiences:
Best wishes Ralph.


> 
> 
> 
> Dear CCD folks, 
> 
> Based on our recent experience with the characterisation of Mike Lesser's 
> devices the following EMail was compiled. It is also triggered upon the 
> occasion of some recent EMails by John Geary and Mike Lesser as well 
> as the Danish CCD group. I would like to get your comments and point of view 
> to better compare the UV flooding results and methods of the different 
> organisations and to reach the best AND MOST REPRODUCIBLE device behaviour, 
> also simplifying its maintenance in a REMOTE observatory.
> 
> (The flooding works, but seems to depend on at least some other issuses 
> we did not compare in detail so far, so they are probably not applied similar 
> and therefore not really reproducible between different sites.)
> 
> 
> As a compilation of the different groups' results I therefore try to list 
> first those points which seem to be CLEAR by now 
> 
> !!! IF ANY OBJECTIONS PLEASE LET ME KNOW !!! :
> 
> 1.) Oxygen and air filling of the dewar to 1 atmosphere do NOT cause any 
>     difference in the CCD QE behaviour
> 2.) The filling with oxygen can be done, but complicates the maintenance (also
>     in terms of potential explosion danger, should the pump not be 100% oil 
>     free). 
>     (Even more complicated would be the UV flooding directly at the 
>     telescope in terms of maintenance, should a routine or accidental 
>     warm-up occur)
> 3.) The use of normal ambient air (humidity between 30 and 50%) does not seem 
>     to result in very good reproducibility
> 4.) Therefore the best method seems to be the usage of dry air which can be 
>     obtained in bottles (and is therefore better reproducible in terms of 
>     its humidity)

I agree. Thats is what we use in Copenhagen. Ralph.

> 5.) The lamp type used for flooding (Zn / Cd) results in a negligible 
>     difference in the CCD QE 
> 
> NOW TO THE REAL QUESTIONS AND THINGS WE SUSPECT TO BE DIFFERENT AND GIVING AN 
> EFFECT ON THE QE BEHAVIOUR :
> 
> A.) Baking out of the dewar seems to matter to prevent a contamination of 
>     the dewar with gasses not necessarily contaminating the CCD surface 
>     visible, but influencing the chemical UV flooding reaction ?
> B.) What is the window material in use at the different sites using these 
>     CCDs ? (It could be that the window material in use at site A caused by 
>     its different transmission causes a different UV flooding time, compared 
>     to the set-up at site B) 

The window should ideally be of fused silica - for two reasons:
1.: It must transmit well 220 nm (dominant emission lines for Cd lamp)
2.: Fused silica has the lowest natural radioactivity. Ralph.

> C.) Which is the time needed to let the device sit in air or in oxygen BEFORE 
>     applying the UV flooding and does it affect the QE data ?

No time required (as far as I understand). Bake out carefully while pumping.
Then let the air in and start the UV flooding. Ralph.

> D.) What is the time duration of the applied UV lamp  
>       i.) heating up ?

15 minutes for the 5W Cadmium lamp. Ralph.

>       ii.) flooding of the CCD ?

15-20 minutes. Timing does not appear to be critical. Ralph.

> E.) What is the temperature of the CCD during the flooding process ? 

Room temp. Ralph.

> F.) What is the temperature of the CCD from which on it is storing 
>     (descending temperature) the UV flooding (0 degrees ?) ?

I do not understand the question. Ralph.

> G.) What is the time needed for evacuating the dewar to a given vacuum 
>     pressure range (e.g. 10exp-5 / 10 exp-6 ) after the UV flooding ? 

We reach 10e-5 mbar in a couple of minutes. Then fill with LN2. Ralph.

> H.) What is the cool - down rate / profile of the CCD actually in effect ? 
> I.) Is the UV flooding lamp left on during pumping / cool down of the device 

No, we do not do that. John Geary got bad results w.r.t. QE by leaving
the lamp on during pumping. Ralph.

>     and
>       i.) is it necessary ?
>       ii.) does it have an effect on higher dark current generation ?
> J.) What is the physical/chemical process requiring the ambience of oxygen ?
> K.) Flat fielding of the device is affected by :
>       i.) the duration of UV flooding ? 
>       ii.) the distance of the lamp ?
>       iii.) the history of a flooding applied before ?
> L.) To which extent did you measure the long term stability of the UV 
>     flooding ?

We have not had the thinned CCDs long enough to be able to test the long
term stability. But I should certainly like to know if others have done
such tests. Mike Lesser may know. Ralph.

> M.) What is the physical / chemical reaction with hydrogen (as it erases the 
>     UV flooding)  very rapidly and is there any coincidence with ambient 
>     HUMID air ?
> 
> THE IDEAL CASE WOULD BE THAT WE WOULD HAVE A POOL OF CCD COOL DOWN PROFILES 
> WITH MARKED UV FLOODING / PUMPING TIME POINTS / TIME DURATIONS TO MORE EASILY 
> COMPARE....
> 
> Some points may also seem trivial and were listed already, nevertheless I 
> think it is useful to compare.
> Any information is appreciated but would be most helpful if directly 
> referenced to the above given numbers / letters.
> 
> 
> With best wishes 
> 
> 
> Olaf Iwert
> 
> ESO 
> Optical Detector Group
> Karl Schwarzschildstr. 2 
> 85748 Garching near Munich
> Germany
> 
> Telephone and voice Mail : +49 / 89 / 320 06 353
> Fax                      : +49 / 89 / 320 23  62
> EMail direct             : oiwert at eso.org
> CCD Mailinglist          : ccd at eso.org 
> 
> 
---- I think it is extremely valuable, Olaf, that you try to pool
everybody's experience. By the way we will mount our first thinned
Loral 2kx2k three edge buttable CCD in the Danish Faint Object
Spectrograph and Camera, DFOSC a week from now. I do hope that there
will be a cylinder of dry and oil free atmospheric air available at
La Silla!
	Finally one word of warning: One certain way to ruin the
AR coating of a thinned CCD is to let non-dry air 
into the cryostat before the CCD has reached room temperature. 
The water vapour will settle on all cold surfaces - including the
CCD - as little ice crystals. A new coating costs a lot of dollars!
Happy New Year to everybody,
Ralph.