Test of the cleaning gas mixtures Seog Oh

Test of the cleaning gas mixtures Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Purpose · Find out the safe operating limit of the glass wire-joint in mixtures with CF 4 in case silicon cleaning is necessary. · Run parameters u Upper chamber s 14 glass wire-joints – 5 were pulled out on July 25. The rest was pulled out on Sept 27. – June/July/August : Ar-CO 2 -CF 4 (70 -14 -16) : 16% mixture – September : Ar-CO 2(70 -30) u Lower chamber s 11 glass wire-joints – 4 were pulled out on July 25. The rest was pulled out on Sept 27 – Ar-CO 2 -CF 4 (70 -22 - 8) : 8% mixture u u u Moisture level : 700 ppm on average Gain s June/July/August - ~30, 000 -40, 000 s September ~ 70, 000 -80, 000 Current draw s Average : 10 n. A/cm – 30 n. A/cm s Around wire-joints : 30 n. A-100 n. A/cm Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Test module 2 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Test setup Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% Results · Failure (do not hold HV) occurred when the integrated charge was 0. 2 -0. 3 C/cm. u 0. 22, 0. 19, 0. 26, 0. 21, 0. 27, 0. 34 C/cm · Tension loss proceeded the failure. · All glass beads exhibited heavy etching reasonably proportional to the integrated charge. · Despite the etching, there was no evidence of silicon deposit downstream (SEM & Gain measurement). · This test ended at the end of August. At the time, there were two surviving wire-joints (#30, #31). The gas mixture was changed to Ar-CO 2 (70 -30) to study the residual effect of CF 4 (more later). Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Failed wire-joints (~0. 3 C/cm)16% Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% · Wire #25. 0. 27 C/cm. Broken (8/25) Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire-joints in 16% · Note that there is no Silicon peak. Si is replaced by F. Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire surface scan (16%) · Left : “normal”. Right : somewhat smooth Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% · Left : “normal”. Right : somewhat smoother Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire · Downstream of a wire-joint. There is no silicon deposit Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire #26 (0. 27 C/cm). Broken 8/18 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

8% Results · One failure occurred when the integrated charge was ~0. 45 C/cm. · Tension loss happened as early as ~0. 3 C/cm. · Etching on the failed joint was also evident. However compare to the failed ones in 16%, the etching was less. · There was also no evidence of silicon deposit downstream (SEM & Gain measurement). Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Failed wire-joint (~0. 45 C/cm)8% · Wire #12 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

8% (#10) · Wire #10. 0. 31 C/cm. Lost 22 grams of tension Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire-joint (8%) · There is still some silicon peak Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire #12 (0. 45 C/cm) Broken 9/14 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Residual effect of CF 4 · The 16% mixture was changed to Ar-CO 2 (70 -30) on September 3. · The gas gain : ~70, 000 (from ~30, 000) · Current Draw : ~17 n. A/cm (=> ~50 n. A around the wire -joints) · Charge accumulation u u The integrated charge at the time of the switch : ~0. 15 C/cm The integrated charge since the switch : ~0. 1 C/cm · Results u u The degree of etching was consistent with ~0. 15 C/cm rather than ~0. 25 C/cm. No gain change was observed after the switch Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

0. 15 C/cm in 16%+0. 1 C in Ar-CO 2 · Wire #30, 31 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% -> 0% · Wire #30 and #31 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire-joint #30 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% -> 0% · 4 cm downstream of wire-joint Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

16% -> 0% · 20 cm downstream of wire-joint. The chunks are gold. Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Wire #31 Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Summary · Integrated charge u u u From the 16% and 8% test, it seems that the failure integrated charge is roughly proportional to the CF 4 fraction when the moisture level of ~700 ppm. s 16% : fail at 0. 2 -0. 3 C/cm s 8% : fail at 0. 4 -0. 5 C/cm Including the earlier measurement (CF 4 ~20%, H 2 O ~ 20003000 ppm, integrated charge at the failure ~ 0. 1 C/cm), it seems that the failure also roughly proportional to the H 2 O level. There is no silicon deposit or gain change. · There is no gross residual effect when the gas is switched to non CF 4 mixture. Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN

Proposal · Test 4% mixture – (may not be necessary) · Test the effectiveness of the cleaning mixtures u u u Produce silicon deposit with Ar-CO 2 (70 -30) <=> gain drop of ~10% s Bubble through silicon oil? s Use contaminated straws? Switch the gas to one of the cleaning mixtures. Determine the integrated charge necessary to restore the gain. Parameters s CF 4 fraction s Moisture level s Gain s Irradiation level Seog Oh/ Duke University/Oct, 2002 TRT/ ATLAS week at CERN