Re: Revisions to the high-Z impurity correction table

[Tom Banks <tbanks@socrates.berkeley.edu>]

> 2) You assume that in our run8 clean fill conditions we had mainly impurity
>     captures on Nitrogen. This was however a controversial issue, since
>     the Chemists analysed about 10 ppb Nitrogen - while the capture yield
>     suggested rather some 100 ppb.
>
>     In runs 9 and 10, using a humidity sensor,  we concluded so far that at
>     least half, but very likelly 90% of the observed impurities come from
>     humidity (Oxygen).
>     I do not see a difference to run 8, since in all these runs the Nitrogen
>     was efficiently cleaned by the CHUPS system.
>
>    Therefore, would it not be more appropriate to take from table 3
>    the Oxygen value (0.800) rather than Nitrogen (0.823), or at least a mix?

Dear Claude,

thanks for taking the time to carefully read through the memo.  Concerning
your question about the impurity composition of the Run8 clean fill: You
are absolutely correct that the final impurity correction will include a
mix of nitrogen and (most likely) oxygen contributions; see my August 1
worklog page

  http://weak0.physics.berkeley.edu/weakint/research/muons/private/tbanks_dir/TeleConf/2006Aug1/2006Aug1.html

for a preliminary treatment of this next stage in the correction
procedure.  However, the bulk of the Run8 clean fill captures were due to
nitrogen, from the period when CHUPS was cleaning the pressure vessel.
See, e.g., the period from 0--200 hours in attachment 2 in

  http://kaon.physics.berkeley.edu:8080/run8/1112

When you say that 90% of the impurities come from humidity, I think you
are referring to equilibrium conditions *after* CHUPS has cleaned out the
system as much as possible.  In Run8, such equilibrium conditions were
achieved for only a portion of the clean fill running period.  When
Bernhard performed his Run8 correction estimates last May,

  http://weak0.physics.berkeley.edu/weakint/research/muons/private/bernhard/analysis_10May06.pdf
  http://weak0.physics.berkeley.edu/weakint/research/muons/private/bernhard/highz-analysis-run8-10may06.html

he conservatively assumed that 90% of the Run8 clean fill capture yield
was due to nitrogen, and that 10% was due to oxygen.  It remains for me to
corroborate these numbers.

It is worth noting that, as you mentioned, the (delta_r/Y) numbers are
pretty much the same for all of the relevant high-Z elements.  Of greater
concern is the detection efficiency for different elements: we can
calibrate the efficiency for nitrogen, but oxygen is at present an
unknown.  Since the detection efficiency appears in the denominator of the
correction formula, it has the potential to dramatically blow up the
error.

Regards,
Tom