I can suggest that you put the contents in the equivalent of a Knudsen diffusion cell container. Put a pin-hole on one side of a "box". Face your sample perpendicular to the hole in the container. Heat the box to cause evaporation from the hole only.

With some care in alignment, this can eliminate or dramatically minimize the background contamination. In other words, this is a case where an ounce of prevention is worth a pound of cure. Also, with some design elegance and experimentation, this can provide you with a way to calibrate the dose level according to the power supplied to the box and/or the temperature of the box.

I propose these options based on work that I did a few decades ago to deposit potassium on to a single crystal surface in a UHV chamber.

I can otherwise only speculate (perhaps also with some light humor) as to the danger and the possibility to remove the contaminate later.

--> Post a Sign on the Escalab ... Danger: Do not lick your fingers after touching the inside of this UHV chamber as it likely contains Cs contamination.

--> Removal: I might suggest to bake out in a low background of oxygen to convert the Cs on the chamber walls to a (harmless and lower vapor pressure) oxide.

ps ... You will NOT want the Cs to float around in the chamber during deposition and/or bake out and end up on the detector, as that will change the work function of the analyzer and throw off your XPS spectra alignments.

Dear Saulius Kaciulis,

I have never tried Cs deposition but Cd was the disaster for the vacuum chamber.

The Cd vapour was able to propagate everywhere coming through the different stages of the vacuum system. It was able to come through the diffusion pump and rotary pump because after this incident I was found Cd deposited everywhere. The first problem appear when Cd deposited on vacuum head element (thermocouple vacuum gauges), it will not show proper value anymore and every head has to be replaced. Be careful, especially if you have an expensive equipment in the same chamber. Such a chamber it's not the best way to test deposition of materials.

It's better to have separated chamber with no analytics inside.

Kurt J. Lesker suggestion is to use Quartz crucible for Cs deposition.

https://www.lesker.com/newweb/deposition_materials/materialdepositionchart.cfm?pgid=0#c

Best regards

Yevhen

Thanks for the answers! However, my plan is a little bit different: the deposition will be done in the prep chamber, therefore the analysis chamber will not be contaminated.

Also the Cs source is different: it is  SAES alkali metal dispenser which contains only 3 - 20 mg of alkali metals. More information on this dispenser:

https://www.saesgetters.com/sites/default/files/alkamax%20application%20note_0.pdf

Maybe I must reformulate my question:

How the Cs is reacting with stainless steel? Does it make some stable alloy?

What could happen with turbo and rotary pumps, when a small amount of Cs will reach them? Do I need the Fomblin oil in the rotary pumps?

As it was already indicated by Yevhen, the vacuum meters (in my case a Pirani and ion gauges) can be seriously contaminated, but maybe it is enough to switch them off during the deposition?

I have used Cs for activation of superlattices, in particular for permanent cesiation, during few month. Did not have any problems with ion gauges, pumps and chamber walls. Only thing I observe that after long time (5-6 month) Cs was setteled at the holder.Need to be cleaned after that.