Transcript New tests with new gas mixture (A. Paoloni)

Study of gas mixture containing SF6 for the
OPERA RPCs
A. Paoloni, A. Mengucci (LNF)
The present gas mixture for streamer mode operation
Ar/TFE/i-but = 48/48/4
• Argon : high charge multiplication at low voltages
• Iso-butane : quenching, but limited by flammability requirements
• Tetrafluoroethane : electronegative gas, good quenching power
• ZTFE = 50 : high primary ionisation  better intrinsic efficiency
and time resolution
Unfortunately….
TFE + discharges in gas
HF production, aggressive to electrodes surface
Detector damaging
A possible solution
Decrease TFE and increase Ar, with fixed i-butane quantity:
1. The operating voltage decreases, and so does the streamer
charge
2. The loss of quenching power can be compensated with the
addition of little quantities ( <1% ) of SF6, a strongly
electronegative gas already used for RPCs working in
avalanche mode
3. Lower charge released in the gas + less TFE  much less HF
 less chamber damaging
4. Moreover the gas mixture is less expensive
Experimental Set-up
• M1, M2, M3 trigger RPCs (50*50) cm²,
flushed with the standard gas mixture
• RPC1, (60*70) cm², under test with
different gas mixtures
• RPC1 current read out with the scope on
a resistor placed in the HV circuit
• RPC signals are read out with a single
pad covering the entire surface
(Threshold=45 mV)
• Signal shape: RC > streamer duration 
peak=Q/C; t=RC (R=50 W)
• Charge-per-count=current/rate
• Prompt charge measured integrating the
average pad signal acquired with a
scope
Decreasing TFE without SF6
4% iso-butane fixed
Lower operating voltage, but higher
Currents !!!
Very low ohmic currents  current only due to discharges in the gas
The effect of SF6
Baseline gas
Ar/TFE/i-but=76/20/4
• The operating voltage is shifted depending on the SF6 quantity
• The current is strongly reduced
• The effect seems to be independent from the SF6 quantity
• Also the prompt charge is reduced
Decreasing TFE with SF6
4% iso-butane and 1% SF6 fixed
From the point of view of the current, the chamber behaviour seems to
be independent from the TFE concentration in a wide range…..
Final comparison (I)
Ar/TFE/I-but=48/48/4
Ar/TFE/I-but=76/20/4 + 0.5% SF6
The operating point is about 2 kV lower
Final comparison (II)
Ar/TFE/I-but=48/48/4
Ar/TFE/I-but=76/20/4 + 0.5% SF6
No loss of efficiency with half charge-per-count and prompt charge
Strip signals study
• Cosmic rays are triggered with two scintillators
• RPC strips are terminated on their characteristic impedance only
on one side; on the other side they are terminated on 110 W (to
simulate OPERA read-out)
• In order to measure efficiencies, the signals are taken from the
110 W side and sent to a discriminator with 50 W input impedance
Threshold scan
200 V inside
plateau
Ar/TFE/I-but=48/48/4
150 V inside
plateau
Threshold=50 mV
To be scaled by ~3
because of the discr.
Rin (for OPERA
read-out comparison)
Ar/TFE/I-but=76/20/4
+ 0.5% SF6
320 V inside
plateau
Conclusions
• HF produced by discharges in gases containing TFE can
damage electrodes surface
• TFE concentration can be reduced compensating the loss
of quenching power by adding little quantities (<1%) of
SF6, a strongly electronegative gas
• With a gas mixture Ar/TFE/i-but=76/20/4 + 0.5% SF6
the operating voltage is ~2 kV lower than with the
standard mixture (Ar/TFE/i-but=48/48/4)
• No loss of efficiency is observed and both the current
and the prompt charge are about one half with respect to
the standard mixture
• More to come: time resolution and afterpulses studies