J Provine jprovine at
Wed Oct 20 08:45:54 PDT 2010

i just ran into mary in the hallway, and she suggested that instead of
swapping the NH3, it would be preferred from a safety standpoint to swap out
one of the precursors.  to me the perferred precursor choice is the W source
1) it is the least used
2) it has the lowest vapor pressure

the added effort for removing W vs NH3 is pretty minimal but non-zero.  it
1) more time for cooling the source and jacket a bit for removal
2) an additional particle filter/RFO (since the one on the ammonia line
would still be intact)

that's it - otherwise it is like any other precursor swap.
additionally, if it wasn't clear from the previous email this isn't
something any of the users would be doing, instead i would do the changes
both before and after experiment.

as before, advice and assistance in moving this project forward is welcome.

On Wed, Oct 20, 2010 at 6:39 AM, J Provine <jprovine at> wrote:

> john et al,
> any further thoughts on this project?
> j
> On Fri, Oct 15, 2010 at 5:52 PM, J Provine <jprovine at> wrote:
>> john,
>> all lines on the system are currently being used.  it doesn't have to be
>> the ammonia bottle.  it could be one of the organo-metallic precursors (h2o
>> is no good because it is to be used for in situ etching of the GeOx that is
>> formed).
>> i chose the ammonia bottle because it has two valves for double shut off
>> and already has the needle valve and doesn't utilize a heater jacket.
>> so changing to another line (other than h2o) is fine, but it just trades
>> ammonia for an organo-metallic precursor.
>> j
>> On Fri, Oct 15, 2010 at 5:34 PM, John Shott <shott at> wrote:
>>>  J:
>>> Swapping ammonia bottles on a per-run basis makes me somewhat nervous.
>>>  Am I correct in thinking that, at the moment, the Savannah is not using all
>>> of its metallorganic "slots"?  Rather than breaking into the ammonia line,
>>> is there a reason that one of the other channels that has a currently-unused
>>> ALD valve, can't be used for the oxygen injection needle valve?  In that
>>> way, there would be no hardware changes required between runs, the ammonia
>>> line doesn't need to be disturbed at all, and selection of oxygen vs.
>>> ammonia is controlled by the recipe.
>>> What am I missing?
>>> Thanks,
>>> John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the specmat mailing list