Well, that partially will depend on the (model of) TMP that you're using and the quality of the roughing vacuum. There is already something you can do at this stage to reduce the hydrogen partial pressure (by choosing a TMP with good compression for hydrogen).

And yes, using getter pumps will improve the vacuum and in particularly can get the hydrogen partial pressure down. Whether SAES is a good choice I don't actually know (I have only used their alkali metal evaporators). In any case, getter pumps working with evaporation of films are what you may want to call "periodical pumps". This means that typically while you generate a fresh gettering surface, you temporarily have bad vacuum and hydrogen partial pressure (because the evaporant gets hot and - depending on conditions also the area you deposit on gets hot).

In some experiments where we needed very good vacuum this was because we experimented with rare earth metals (to fabricate ultrathin films). Whenever possible we installed the evaporator such that we could chose whether to evaporate onto the sample or the UHV chamber walls. So, prior to actually producing the sample we first produced a thin film on the chamber. Waiting for some time this improved hydrogen partial pressure quite a bit. This used much less power and intrduced much less heat than with the more usual titanium sublimation pumps (serving the same purpose, i.e. chemical gettering) which we had in other installations. Under these circumstances also, nitrogen cooling was not required.

The larger the area you can deposit on, the more effect you get (and the more evaporant you will use...).

Well, that partially will depend on the (model of) TMP that you're using and the quality of the roughing vacuum. There is already something you can do at this stage to reduce the hydrogen partial pressure (by choosing a TMP with good compression for hydrogen).

And yes, using getter pumps will improve the vacuum and in particularly can get the hydrogen partial pressure down. Whether SAES is a good choice I don't actually know (I have only used their alkali metal evaporators). In any case, getter pumps working with evaporation of films are what you may want to call "periodical pumps". This means that typically while you generate a fresh gettering surface, you temporarily have bad vacuum and hydrogen partial pressure (because the evaporant gets hot and - depending on conditions also the area you deposit on gets hot).

In some experiments where we needed very good vacuum this was because we experimented with rare earth metals (to fabricate ultrathin films). Whenever possible we installed the evaporator such that we could chose whether to evaporate onto the sample or the UHV chamber walls. So, prior to actually producing the sample we first produced a thin film on the chamber. Waiting for some time this improved hydrogen partial pressure quite a bit. This used much less power and intrduced much less heat than with the more usual titanium sublimation pumps (serving the same purpose, i.e. chemical gettering) which we had in other installations. Under these circumstances also, nitrogen cooling was not required.

The larger the area you can deposit on, the more effect you get (and the more evaporant you will use...).

As was pointed out by Kai, it's really a question of what's your chamber, how you treat it etc.

Another important question is how 'big' is your TMP. This may be an equally important question as whether to use a TMP. By 'big' I mean does it work faster than the leakage and outgassing during operation of the chamber? The difference between 1E-10 and 5E-11 could also be the position and condition of your ion gauge (near the pump vs near the load lock).

On our MBE chamber we get to the low 1E-10 Torrs with just one big TMP on the growth chamber. To be fair, we also deposit Ti in there frequently so the whole chamber acts as a TSP (but then again, we also introduce oxygen as well). I don't have hands-on experience with TSP, but to get from low -10s to mid -11s should not be too hard using the right TSP or ion-getter pump.

Needless to say that the application/operation of the system also dictates the choice of pumps, e.g. you don't want IGPs where you frequently introduce oxygen, you don't want cryopumps where vibrations are an issue, you don't want IGPs where magnetic fields are important, and so on...

You can greatly reduce (virtually eliminate) hydrogen back-streaming by backing the turbo pump with a small diffusion pump. In our system we use a Seiko Seiki STP-400 turbo pump backed by a 50l/s diffusion pump which allow us easily achieve ultimate pressure well below 1E-10 Torr.

Kai Fauth Lior Kornblum

Thank you for your valuable suggestions.

The purposed UHV system will consist of large sized SAES non-evaporable getter pump(3500L/s) running simultaneously with Pfieffer high compression TMP(700 L/s). Do you have any suggestion on good quality TMP and roughing pumps?

Our experimental chamber is large in size possibly more than 100K cm^2 of surface area and nearly 1.5 m^3 volume. It will be 316L stainless steel and electropolished.

Cryo pump was ruled out because of vibration and ion pumps due to its strong magnetic field.

Alexei V Ermakov Thank you for your valuable suggestions. Because of the oil contamination, I think diffusion pump wont work well.

Basu Ram Lamichhane

I have no experience with the numbers. Their relevance can vary if you deposit stuff in the chamber - coated walls can act as sponges and take longer to outgas etc. As a general rule of thumb, I always prefer to pay a bit extra (when I can afford it) and get a slightly stronger TMP than what the specs say. The performance of the chamber and pumping system will not be as good in 5 years as they are on day 1, so if you have this redundant pumping power it can always help.

Lior

What kind of ion gauge did you use to measure 10-10 torr pressure? Lior Kornblum

It also depend on what kind of ion species, energy will be running through your chamber and what kind of reaction products will be measured.

Ruitian Zhang , we use a standard Bayard-Alpert Ionization Gauge, like this one:

https://www.mksinst.com/f/274-nude-bayard-alpert-ionization-vacuum-gauge

While your second comment is correct, we don't really care about the precise absolute values of the pressure as long as our process works consistently and well (also, the exact reading significantly depends on the position of the gauge). We also have an RGA instrument on our chamber, in case we want to rule out contamination etc.