Larry
___________________________________________________________________
Dear Folks,
I had a very long and helpful conversation with Ulrich Becker
(MIT LEP-L3) today. They are using Ar/CO2/Isobutane 86/10/4 for the
L3 vertex chambers. They chose this mixture to give the same v_d as
60/40 Ar/ethane (which they were prohibited from using because of
safety concerns). They have done extensive drift velocity and Lorentz
angle tests of chamber gases (see
http://marie.mit.edu/~uchida/drift.html). They fixed the isobutane
fraction at the maximum permitted by CERN, then they adjusted the CO2
fraction to replicate the desired drift velocity. They found that
86/14 Ar/CO2 had the same velocity but was not stable enough (gain
stability and plateau width). Ref: 'The forward muon detector of L3,'
NIM 383, pp342-366 1996
CO2 and He are both good for reducing Lorentz angles
(desirable for varied B field operation). The Ar/ethane 60/40 had a
Lorentz angle at B=0.5T of 20 degrees, the 86/10/4 Ar/CO2/isobutane
had a Lorentz angle of 15 degrees.
We discussed aging (a subject on which he is not an expert).
His benchmark figure for worrying about aging is 0.01 C/cm!!!!!!!!!
He suggested CF4 as an antiaging gas. In certain gas mixtures the CF3
radical etches away deposits on wires. It is a very fast (100 mu/ns),
strongly quenching, not electronegative gas. It is also a freon and
thus very expensive and hard to get. There are some discrepancies in
the aging results depending on exact mixtures and impurities. Kadyk
is the expert on this. I do not plan to investigate this too much.
CO2: it is controversial with respect to aging. It reacts
badly with certain contaminants (Becker mentioned H2O specifically).
CO can form which can and will form deposits on wires. If you have
very high quality and purity CO2, then there is almost no aging. The
LEP L3 group has a special purifier and a quadrupole mass spectrometer
to analyze contaminants. They also recirculate 95% of the gas and
flow one chamber volume every 6-7 hours. They keep the N2 level below
500 ppm to avoid changing drift velocities (NIM A351 (1994) 583-584
and errata NIM A361 (1995) 401).
He has a low opinion of isopropyl alcohol. He thinks it has a
very small helping effect. It provides some quenching of sparks.
He has offered to provide calculations of drift velocity and
Lorentz angle for any gas mixture we want to send him.
I am now tracking down some of his other references and will
thoroughly explore his web site.
Sincerely,