physics0303059/att.tex
1: \documentclass{elsart}
2: 
3: \usepackage{graphicx}
4: \usepackage{amssymb}
5: 
6: 
7: \newcommand{\tmq}{$t_{\langle Q\rangle}$}
8: \def\NIMA#1#2#3{Nucl. Instr. Meth. Phys. Res. A {\bf #1}\ (#2)\ #3}
9: \def\NIM#1#2#3{Nucl. Instr. Meth. {\bf #1}\ (#2)\ #3}
10: \def\IEEE#1#2#3{IEEE Trans. Nucl. Sci. vol. {\bf #1}\ (#2)\ #3}
11: \def\PPNP#1#2#3{Prog. Part. Nucl. Phys. {\bf #1}\ (#2)\ #3}
12: 
13: \begin{document}
14: 
15: \begin{frontmatter}
16: 
17: \title{Pulse height measurements and electron attachment in drift chambers 
18: operated with Xe,CO$_2$ mixtures}
19: 
20: \author[gsi]{A.~Andronic\thanksref{info}},
21: \author[hei]{H.~Appelsh\"auser}, 
22: \author[gsi]{C.~Blume}, 
23: \author[gsi]{P.~Braun-Munzinger}, 
24: \author[mue]{D.~Bucher}, 
25: \author[gsi]{O.~Busch}, 
26: \author[gsi]{A.~Castillo Ramirez}, 
27: \author[buc]{V.~C\u at\u anescu}, 
28: \author[buc]{M.~Ciobanu}, 
29: \author[gsi]{H.~Daues}, 
30: \author[gsi]{A.~Devismes}, 
31: \author[hei]{D.~Emschermann}, 
32: \author[dub]{O.~Fateev},
33: \author[gsi]{C.~Garabatos}, 
34: \author[hei]{N.~Herrmann}, 
35: \author[gsi]{M.~Ivanov}, 
36: \author[hei]{T.~Mahmoud}, 
37: \author[mue]{T.~Peitzmann}, 
38: \author[hei]{V.~Petracek}, 
39: \author[buc]{M.~Petrovici},
40: \author[mue]{K.~Reygers}, 
41: \author[gsi]{H.~Sann}, 
42: \author[mue]{R.~Santo},
43: \author[hei]{R.~Schicker},  
44: \author[gsi]{S.~Sedykh}, 
45: \author[dub]{S.~Shimansky},
46: \author[gsi]{R.S.~Simon}, 
47: \author[dub]{L.~Smykov},
48: \author[hei]{H.K.~Soltveit}, 
49: \author[hei]{J.~Stachel}, 
50: \author[gsi]{H.~Stelzer}, 
51: \author[gsi]{G.~Tsiledakis}, 
52: \author[hei]{B.~Vulpescu}, 
53: \author[gsi]{J.P.~Wessels}, 
54: \author[hei]{B.~Windelband},
55: \author[mue]{O.~Winkelmann},
56: \author[hei]{C.~Xu},
57: \author[mue]{O.~Zaudtke},
58: \author[dub]{Yu.~Zanevsky},
59: \author[mue]{V.~Yurevich}
60: 
61: \address[gsi]{Gesellschaft f\"ur Schwerionenforschung, Darmstadt, Germany}
62: \address[hei]{Physikaliches Institut der Universit\"at Heidelberg, Germany}  
63: \address[mue]{Institut f\"ur Kernphysik, Universit\"at M\"unster, Germany}  
64: \address[buc]{NIPNE Bucharest, Romania}  
65: \address[dub]{JINR Dubna, Russia}  
66: 
67: {for the ALICE Collaboration}
68: 
69: \thanks[info]{Corresponding author: GSI, Planckstr. 1, 64291 Darmstadt,
70: Germany; Email:~A.Andronic@gsi.de; Phone: +49 615971 2769; 
71: Fax: +49 615971 2989}
72: 
73: \begin{abstract}
74: We present pulse height measurements in drift chambers operated with Xe,CO$_2$ 
75: gas mixtures. 
76: We investigate the attachment of primary electrons on oxygen and SF$_6$ 
77: contaminants in the detection gas.
78: The measurements are compared with simulations of properties of drifting 
79: electrons.
80: We present two methods to check the gas quality: gas chromatography and
81: $^{55}$Fe pulse height measurements using monitor detectors.
82: \end{abstract}
83: 
84: \begin{keyword}
85: drift chambers \sep Xe,CO$_2$ mixtures 
86: \sep electron attachment
87: \sep gas chromatography
88: 
89: \PACS 29.40.Cs   
90: \sep 29.40.Gx   
91: \end{keyword}
92: \end{frontmatter}
93: 
94: \section{Introduction} \label{aa:intro}
95: 
96: The ALICE Transition Radiation Detector (TRD) \cite{aa:tdr} has to provide 
97: both electron identification and particle tracking. To achieve this, 
98: accurate pulse height measurement in drift chambers operated with 
99: Xe,CO$_2$(15\%) gas mixture over the drift time of the order of 2~$\mu$s 
100: (spanning 3~cm of drift length) is a necessary requirement.
101: For such precision measurements, it is of particular
102: importance not to loose charge by electron attachment, i.e. the absorption of 
103: drifting electrons by electronegative molecules present in the detector gas 
104: as contaminants.
105: The large volume (28~m$^3$) of the ALICE TRD and the high cost of xenon make the 
106: above arguments very serious for the operation of the final detector.
107: 
108: Attachment is a well studied and generally understood phenomenon, both
109: fundamentally \cite{aa:chris2,aa:chris} and concerning its practical implications
110: for gas drift chambers \cite{aa:huk}.
111: For electron energies relevant to gaseous detectors (energies below a few eV), 
112: attachment occurs mainly via two mechanisms: resonance capture and dissociative 
113: capture. Resonance capture, also called Bloch-Bradbury mechanism \cite{aa:bb}, 
114: has the largest cross-section. It can be written as: 
115: \begin{equation} 
116: I + e^- \rightarrow I^{-*} \label{aa:eq1}
117: \end{equation}
118: \begin{equation} 
119: I^{-*} + S \rightarrow I^- + S^* . \label{aa:eq2}
120: \end{equation}
121: $I$ denotes the impurity and $S$ is a third body stabilizer, which 
122: in case of gas detectors, is usually the quencher. 
123: The star ($*$) denotes a vibrationally excited state. Besides decaying by the 
124: resonant energy transfer (\ref{aa:eq2}), $I^{-*}$ could also decay by electron 
125: emission (autodetachment), in which case there is no signal loss. 
126: The rate of process (\ref{aa:eq2}), and thus the magnitude of attachment,
127: depends on the concentration of $S$ and on the lifetime of the 
128: excited state $I^{-*}$.
129: Also, as a result of different vibrational levels available for the energy
130: transfer (\ref{aa:eq2}), the attachment depends on the type of quencher 
131: \cite{aa:huk,aa:wen}.
132: An excess of electron attachment with respect to the Bloch-Bradbury mechanism
133: has been identified and assigned to van der Waals complexes \cite{aa:kok}.
134: 
135: The most common electronegative molecule is O$_2$,
136: present in gaseous detectors as a residual contaminant in the gas supply or from 
137: the atmosphere due to imperfect tightness of the system.
138: Attachment on O$_2$ has been extensively studied for Ar-based mixtures in ranges 
139: of parameters relevant for drift chambers \cite{aa:huk,aa:wen}.
140: Another common contaminant in gas detectors is H$_2$O, usually outgassed by 
141: assembly materials, and often appreciated as an ageing-limiting agent 
142: \cite{aa:kad}.
143: It has been found that, for certain Ar-based mixtures, attachment on H$_2$O 
144: alone is negligible, but a few hundred ppm of H$_2$O can double the attachment 
145: coefficient on O$_2$ \cite{aa:huk}.
146: 
147: In one of our early measurements with TRD prototype chambers operated 
148: with Xe,CH$_4$(10\%) we observed electron attachment under very low contamination
149: levels of O$_2$ and H$_2$O. 
150: As that particular gas supply was exhausted during the measurements, we were not 
151: able to analyze it and so not able to attribute the attachment to a defined 
152: impurity.
153: Subsequent measurements (performed with a new supply of Xe) proved to be free 
154: of visible attachment.
155: However, recent 
156: observation of strong attachment, this time with the mixture Xe,CO$_2$(15\%) 
157: forced us to investigate the issue in more detail.
158: We have been able to identify the impurity responsible for attachment:
159: sulphur hexafluoride, SF$_6$, which was found to be present at the ppm level in 
160: a xenon supply.
161: This heavy gas, well known for its excellent insulating properties,
162: has an extremely large cross section for electron attachment at low electron 
163: energies.
164: We note that pulse height distributions as a function of drift time were 
165: measured before in drift chambers with Xe-based mixtures, also related to TRDs
166: \cite{aa:wat,aa:zeus,aa:hol,aa:d0}. 
167: A decrease of the average pulse height as a function of drift time was observed
168: in all these cases and it was attributed to electron attachment \cite{aa:d0},
169: but not quantitatively understood.
170: 
171: Here we report on measurements performed during prototype tests of the
172: ALICE TRD \cite{aa:tdr}.
173: Drift chambers operated with Xe,CO$_2$ mixtures are investigated.
174: The experimental setup and method of data analysis are described in the next 
175: section. We then present measurements under clean conditions (no attachment).
176: The following sections contain our measurements of attachment on oxygen and
177: on SF$_6$. We compare our results with simulations of charge transport
178: for various gas mixtures.  
179: We present gas quality checks employing gas chromatograph analyses and 
180: $^{55}$Fe measurements using especially-built monitoring detectors.
181: 
182: 
183: \section{Experimental setup} \label{aa:meth}
184: 
185: Most of the results are obtained using prototype drift chambers (DC) with 
186: a construction similar to that anticipated for the final ALICE TRD
187: \cite{aa:tdr}, but with a smaller active area (25$\times$32~cm$^2$).
188: In Fig.~\ref{aa:prin} we present a schematic view of the detector.
189: The DC has a drift region of 30~mm and an amplification region of 7~mm.
190: The anode wires (W-Au, 20~$\mu$m diameter) have a pitch of 5~mm.
191: For the cathode wires (Cu-Be, 75~$\mu$m diameter) we use a pitch of 2.5~mm.
192: We read out the signal on a segmented cathode plane.
193: The pads (of 6~cm$^2$ each) have either chevron \cite{aa:yu} or rectangular
194: shape.
195: The entrance window (25~$\mu$m aluminized Kapton) simultaneously serves  
196: as gas barrier and as drift electrode.
197: 
198: \begin{figure}[hbt]
199: \centering\includegraphics[width=.6\textwidth]{att1.eps}
200: \caption{Schematics of the drift chamber.}
201: \label{aa:prin} 
202: \end{figure} 
203: 
204: A charge-sensitive preamplifier/shaper (PASA) was especially designed and built 
205: (with discrete components) for the purpose of prototype tests.
206: It has a gain of 2~mV/fC and noise of about 1800 electrons r.m.s. 
207: The FWHM of the output pulse is about 100~ns for an input step function. 
208: For the read-out of the DC we use an 8-bit non-linear Flash ADC (FADC) system 
209: with 100~MHz sampling frequency, 0.6~V voltage swing and adjustable baseline.
210: The FADC sampling was rebinned in the off-line analysis in order to resemble
211: the performance of the final detector \cite{aa:tdr}.
212: The data acquisition (DAQ) is based on a VME event builder and was developed 
213: at GSI Darmstadt \cite{aa:mbs}.
214: As the beam diameter is of the order of a few cm, we usually limit the readout 
215: of the DC to 8 pads. This also minimizes data transfer on the VSB bus connecting 
216: the FADC and the event builder.
217: 
218: The measurements were carried out at beam momenta of 1~GeV/c at GSI 
219: Darmstadt \cite{aa:gsipi} and 3~GeV/c at the CERN PS \cite{aa:cernpi}.
220: The beams were mixtures of electrons and negative pions.
221: For the present analysis we have selected clean samples of pions using 
222: coincident upper thresholds on a Cherenkov detector and on a lead glass 
223: calorimeter (see ref. \cite{aa:andr} for details).
224: To minimize the effect of space charge on the pulse height measurement, 
225: which occurs for tracks at normal incidence to the anode wires (for which all 
226: charge collection takes place at a narrow spot on the anode wire), we adjusted 
227: the angle of incidence of the beam to about 15$^\circ$ with respect to the 
228: normal incidence to the anode wires. 
229: A particle trajectory through the detector is sketched in Fig.~\ref{aa:prin}.
230: 
231: The standard gas mixture for our detectors is Xe,CO$_2$(15\%) at 1 mbar above
232: atmospheric pressure.
233: The continuous flow of gas through the detectors is either vented out 
234: or recirculated via our gas system.
235: The detectors are usually operated at gas gains around 8000.
236: Our standard supply of xenon is from Messer-Griesheim \cite{aa:messer} and 
237: proved to provide a very good detector performance. 
238: We have also used xenon from Linde \cite{aa:linde}, which provided a
239: strikingly poor signal at first examination.
240: SF$_6$ was detected early on with gas chromatography techniques in 
241: this xenon supply, and its concentration was measured to be of 
242: the order of 1~ppm.
243: Both xenon supplies were used in beam measurements reported on below.
244: 
245: \section{Measurements under clean conditions} \label{aa:clean}
246: 
247: For these measurements we use SF$_6$-free xenon. 
248: The oxygen content in the gas was continuously monitored and kept below 
249: 10~ppm using a flow of 2-3~liters of fresh gas per hour into two chambers
250: of about 9 liters total volume. The water content was about 150~ppm.
251: 
252: Distributions of average pulse height, $\langle PH \rangle$, as a function 
253: of drift time for different drift voltages are shown in Fig.~\ref{aa:ph1}
254: for pions of 1~GeV/c momentum. 
255: The detector gas is our standard mixture, Xe,CO$_2$(15\%).
256: The time zero has been arbitrarily shifted by about 0.3~$\mu$s to
257: have a measurement of the baseline and of noise.
258: Similar distributions have been measured for the mixture Xe,CO$_2$(20\%).
259: 
260: \begin{figure}[hbt]
261: \centering\includegraphics[width=.6\textwidth]{att2.eps}
262: \caption{The average pulse height as function of drift time for different
263: drift fields for the mixture Xe,CO$_2$(15\%).}
264: \label{aa:ph1} 
265: \end{figure} 
266: 
267: The primary electrons from ionization energy loss of pions drift towards
268: the anode wire, where they are amplified. 
269: The signal (charge induced on the pads) is determined mainly by the 
270: slow-moving ions, producing long tails in the PASA output.
271: The overlap of these tails, convoluted with the response of the preamplifier, 
272: results in a slightly rising average pulse height as a function of the drift 
273: time, as seen in Fig.~\ref{aa:ph1}.
274: The peak at short drift times 
275: originates from the primary clusters generated in the amplification region, 
276: where the ionization from both sides of the anode wires contributes to the 
277: same time interval.
278: Note that, for the present conditions, lower values of drift field imply 
279: smaller drift velocity (see Fig.~\ref{aa:vdrift2} in section \ref{aa:calc}), 
280: leading to a stretching of the signal over longer drift times.
281: Our measurements established for the first time \cite{aa:andr} the expected 
282: time evolution of the signal in drift chambers of the type studied here.
283: 
284: 
285: \section{Attachment on oxygen} \label{aa:oxygen}
286: 
287: Again, for these measurements we use clean (SF$_6$-free) xenon. 
288: In Fig.~\ref{aa:ph2} we show the average pulse height distributions as
289: a function of drift time for different values of the oxygen content in the
290: range of a few hundred ppm.
291: A decrease of the signal as a function of drift time is seen when the 
292: concentration of oxygen increases.
293: This is a clear indication of electron attachment. 
294: Notice that the signal in the amplification region, where electrons drift 
295: very little, is affected to a much smaller extent. 
296: 
297: \begin{figure}[hbt]
298: \centering\includegraphics[width=.6\textwidth]{att3.eps}
299: \caption{Average pulse height as function of drift time for different
300: values of the oxygen concentration in Xe,CO$_2$(15\%).}
301: \label{aa:ph2} 
302: \end{figure} 
303: 
304: In case of attachment, the number of electrons (and the corresponding measured
305: pulse height) decreases exponentially as a function of drift time, $t$:
306: \begin{equation} 
307: N(t)=N(0) \cdot \mathrm{e}^{-A\cdot t} ,
308: \end{equation} 
309: where $A$ is the attachment rate \cite{aa:huk}.
310: At a given gas pressure, $p$ (which is the atmospheric 
311: pressure in our case), the attachment rate depends linearly on the 
312: concentration (partial pressure) of the impurity responsible for attachment, 
313: $p_I$, and can be factorized \cite{aa:huk} as:
314: \begin{equation} 
315: A=p \cdot p_I \cdot C_{I},
316: \end{equation} 
317: where $C_{I}$ is the attachment coefficient of the impurity $I$.
318: 
319: From the above measurements of pulse height distributions as a function of drift 
320: time we deduce, for the present value of the drift field of 0.77 kV/cm, 
321: an attachment coefficient on O$_2$, $C_{O_2}$=400~bar$^{-2}\mu$s$^{-1}$.
322: This value is very similar to values measured for Ar,CO$_2$ mixtures with
323: comparable CO$_2$ content \cite{aa:wen} and more than an order or magnitude 
324: larger than values measured for Ar,CH$_4$ mixtures \cite{aa:huk}.
325: Given the short drift time in our detectors, attachment on oxygen does not
326: impose any severe constraint on the tightness of the drift chambers for the
327: final detector. 
328: 
329: \section{Attachment on SF$_6$} \label{aa:sf6}
330: 
331: All the measurements presented below have been carried out with pions of 
332: 3~GeV/c momentum, using Linde xenon with 1.1~ppm SF$_6$ contamination.
333: For these measurements the O$_2$ and H$_2$O contamination was 150~ppm 
334: and 400~ppm, respectively.
335: These rather large values arise because, due to the contamination of the xenon, 
336: the gas was vented and not recirculated (and, as a consequence, not filtered),
337: as usual, through our gas system.
338: 
339: \begin{figure}[hbt]
340: \centering\includegraphics[width=.6\textwidth]{att4.eps}
341: \caption{Average pulse height as a function of drift time for different
342: values of the drift field for the mixture Xe,CO$_2$(15\%).}
343: \label{aa:ph3} 
344: \end{figure} 
345: 
346: In Fig.~\ref{aa:ph3} we show the average pulse height distributions as
347: a function of drift time for different values of the drift field
348: for the standard gas mixture, Xe,CO$_2$(15\%).
349: Beyond the decrease of the pulse height due to longer drift times for lower
350: drift fields, there is obviously a dramatic loss of signal due to electron 
351: attachment. 
352: The relative loss of signal is most pronounced between the two highest values 
353: of the drift field, for which the variation of the drift velocity is very small.
354: At drift fields higher than 1~kV/cm, the plateau of the average pulse 
355: height in the drift region starts to recover.
356: This trend is explained by the fact that the energy of the drifting 
357: electrons increases with the electric field and, therefore, the attachment 
358: coefficient of the SF$_6$-polluted mixture decreases (see next section).
359: Similar behavior of the attachment as a function of drift field is known
360: for Ar-based mixtures \cite{aa:huk}.
361: 
362: \begin{figure}[hbt]
363: \centering\includegraphics[width=.56\textwidth]{att5.eps}
364: \caption{As Fig.~\ref{aa:ph3}, but for the mixture Xe,CO$_2$(10\%).}
365: \label{aa:ph4} 
366: \end{figure} 
367: 
368: \begin{figure}[hbt]
369: \centering\includegraphics[width=.56\textwidth]{att6.eps}
370: \caption{As Fig.~\ref{aa:ph3}, but for the mixture Xe,CO$_2$(5\%).}
371: \label{aa:ph5} 
372: \end{figure} 
373: 
374: The effect of different CO$_2$ concentrations on the attachment efficiency 
375: of  the contaminated xenon has also been investigated. 
376: Fig.~\ref{aa:ph4} and Fig.~\ref{aa:ph5} 
377: show the average pulse height distributions for 10\% and 5\% CO$_2$, 
378: respectively.
379: The gas gains were not identical for the three concentrations of CO$_2$,
380: so the corresponding distributions can only be compared on a relative 
381: basis.
382: Notice that, for the same drift field, the extension of the signal in time
383: is different for different concentrations of CO$_2$ because of different
384: drift velocities (see Fig.~\ref{aa:vdrift2} in the next section).
385: The signal loss due to electron attachment decreases for lower 
386: CO$_2$ concentrations.
387: For the Xe,CO$_2$(5\%) mixture, the pulse height distribution almost 
388: completely recovers at the highest electric field studied here.
389: As we discuss in the next section, the variation of attachment as a function 
390: of quencher concentration is due to the dependence of the average energy of
391: drifting electrons on the CO$_2$ content.
392: 
393: 
394: \section{Comparison to calculations} \label{aa:calc}
395: 
396: To understand the measurements presented above, we have performed calculations 
397: using the packages GARFIELD  \cite{aa:garf}, MAGBOLTZ \cite{aa:magb} and 
398: HEED \cite{aa:heed}.
399: In Fig.~\ref{aa:vdrift2} we present the calculated drift velocities for 
400: 5\%, 10\% and 15\% of CO$_2$ admixture in Xe. 
401: The dotted vertical lines mark the values of the electric field used for the 
402: measurements of attachment on SF$_6$. The solid vertical line segments indicate 
403: the electric fields used for measurements under clean conditions 
404: (Fig.~\ref{aa:ph1} in section \ref{aa:clean}).
405: All these values are in a region where the drift velocity has a strong 
406: dependence on the drift field, a trend reflected in the measurements presented 
407: above.
408: 
409: \begin{figure}[hbt]
410: \centering\includegraphics[width=.62\textwidth]{att7.eps}
411: \caption{Dependence of the drift velocity on the drift field for 5\%, 
412: 10\% and 15\% CO$_2$ content in xenon, as calculated using GARFIELD/MAGBOLTZ 
413: \cite{aa:garf,aa:magb}.}
414: \label{aa:vdrift2} 
415: \end{figure} 
416: 
417: 
418: \begin{figure}[hbt]
419: \centering\includegraphics[width=.6\textwidth]{att8.eps}
420: \caption{Calculated average pulse height as a function of drift time for 
421: different values of the drift field for the mixture Xe,CO$_2$(15\%),
422: SF$_6$-free.} \label{aa:phclean} 
423: \end{figure} 
424: 
425: To compare how well the measured pulse height distributions can be reproduced by 
426: simulations, we have calculated the detector signals using the GARFIELD package,
427: under the exact experimental conditions. 
428: The preamplifier response has been included in these calculations.
429: The time dependence of the average pulse height for different electric fields 
430: for the clean Xe,CO$_2$(15\%) mixture is presented in Fig.~\ref{aa:phclean}.
431: For an easier comparison with the measurements (Fig.~\ref{aa:ph1} in section 
432: \ref{aa:clean}), we introduced a time shift of 0.4~$\mu$s for these 
433: distributions. The calculations reproduce the measured signals reasonably well, 
434: although not in all details.
435: The slope of signal increase in the drift region as a function of drift time
436: is larger for the calculations. This may be an indication that some residual
437: attachment is still present for the measured data, possibly as a result of 
438: oxygen, water and undetectectable amounts of other contaminants.
439: The slightly asymmetric gaussian preamplifier response affects this slope 
440: very little. Already before folding the preamplifier response the calculated 
441: signals show a larger slope than the measured ones.
442: For the lower values of the drift field the calculations are in disagreement with
443: the measurement concerning the time extension of the signal. This discrepancy,
444: reaching 14\% for the field value of 0.72~kV/cm, may reflect a different field
445: dependence of the drift velocity in measurements and calculations.
446: We note that a good agreement was found between calculations and measurements
447: in other Xe-based mixtures \cite{aa:kunst,aa:becker}.
448: Since the bow of the entrance window (which is also the drift electrode) due 
449: to gas overpressure introduces an obvious uncertainty in the present 
450: measurements, it is too early to assess the above discrepacy quantitatively. 
451: Precision measurements of the drift velocity for the standard TRD gas mixture 
452: Xe,CO$_2$(15\%) are in progress.
453: 
454: \begin{figure}[hbt]
455: \centering\includegraphics[width=.6\textwidth]{att9.eps}
456: \caption{Calculated average pulse height as a function of drift time for 
457: different values of the drift field for the mixture Xe,CO$_2$(15\%),
458: with 0.94~ppm of SF$_6$ contamination.}
459: \label{aa:phsf6} 
460: \end{figure} 
461: 
462: Figure~\ref{aa:phsf6} shows the time dependence of the average pulse height
463: for different electric fields for the Xe,CO$_2$(15\%) mixture with 0.94~ppm 
464: of SF$_6$, corresponding to the SF$_6$ fraction in the measurements.
465: For a direct comparison with the measurements, the time shift is 0.3~$\mu$s 
466: in this case.
467: The measured attachment (Fig.~\ref{aa:ph3} in the previous section) is 
468: reproduced almost quantitatively, except for the smoother reduction of 
469: attachment towards higher field values seen in the measurements.
470: Not reproduced is the measured stronger variation of the signal in the
471: amplification region.
472: 
473: \begin{figure}[hbt]
474: \centering\includegraphics[width=.6\textwidth]{att10.eps}
475: \caption{The energy dependence of cross section of electron attachment on SF$_6$.
476: Recommended values from ref. \cite{aa:chris} are plotted.}
477: \label{aa:att-sigma} 
478: \end{figure}
479: 
480: We turn now to a more detailed investigation of the attachment on SF$_6$.
481: In particular, the measured drift field dependence of the attachment (nicely 
482: reproduced by simulations) is expected to reflect the characteristics 
483: of the attachment cross section as a function of electron energy.
484: Indeed, the cross section of electron attachment on SF$_6$ is large and has a 
485: strong energy dependence, as shown in in Fig.~\ref{aa:att-sigma} \cite{aa:chris}.
486: Resonance capture, which leads to the formation of SF$_6^-$, is the most 
487: important mechanism, with cross sections up to 8$\cdot$10$^{-17}$~m$^2$ for 
488: near-zero electron energies.
489: The cross section of dissociative capture, resulting in a free fluorine atom
490: and SF$_5^-$, peaks at the value of 4$\cdot$10$^{-20}$~m$^2$ for electron 
491: energy of about 0.3~eV.  
492: These characteristics of attachment cross sections of SF$_6$ are quite 
493: different compared to the O$_2$ case \cite{aa:chris2}.
494: For example, in case of O$_2$, the cross section for dissociative capture peaks
495: at about 10$^{-22}$~m$^2$ for electron energy of 6.5~eV \cite{aa:chris2}.
496: 
497: The energy spectra of the drifting electrons for various electric fields are 
498: shown in Fig.~\ref{aa:endis} for the mixture Xe,CO$_2$(15\%).
499: These distributions are computed with the simulation program Imonte 
500: \cite{aa:magb}.
501: A significant high-energy component (energies above 1~eV) is present towards
502: higher field values.
503: The average energy of drifting electrons (also computed with Imonte), 
504: $\bar{\varepsilon}$, is plotted in Fig.~\ref{aa:emean} as a function of drift 
505: field for the three concentrations of CO$_2$ in Xe. 
506: The average energy increases strongly with the electric field and is larger
507: for smaller CO$_2$ concentrations.
508: The dotted vertical lines mark the values of the drift field used for the 
509: measurements.
510: As the attachment cross section decreases as a function of the electron energy, 
511: the trends seen in Fig.~\ref{aa:emean} explain both the observed dependence of 
512: attachment on drift field and on CO$_2$ concentration mixture.
513: 
514: \begin{figure}[hbt]
515: \begin{tabular}{lr}\begin{minipage}{.48\textwidth}
516: \centering\includegraphics[width=1.0\textwidth,height=1.08\textwidth]{att11.eps}
517: \caption{Energy distribution of drifting electrons for different values of 
518: the drift field for the mixture Xe,CO$_2$(15\%).}
519: \label{aa:endis} 
520: \end{minipage} & \begin{minipage}{.48\textwidth}
521: \centering\includegraphics[width=1.0\textwidth,height=1.08\textwidth]{att12.eps}
522: \caption{Average energy of drifting electrons as a function of the drift field 
523: for different concentrations of CO$_2$ in Xe.}
524: \label{aa:emean} 
525: \end{minipage} \end{tabular}
526: \end{figure} 
527: 
528: The concentration of CO$_2$ may influence the rate of the resonant energy
529: transfer in the second step of the Bloch-Bradbury process (\ref{aa:eq2}),
530: but this seems  not to be the case for attachment on SF$_6$.
531: The measured pulse height distributions for 15\%, 10\% and 5\% CO$_2$ 
532: (Figs.~\ref{aa:ph3} to \ref{aa:ph5} in the previous section) for the field 
533: values of 1.36, 1.07 and 0.79 kV/cm (for which the corresponding average 
534: energies are similar, see Fig.~\ref{aa:emean}), respectively, look very similar.
535: Although we cannot make a quantitative statement, apparently all the influence 
536: of CO$_2$ concentration on the attachment stems from the average energy of 
537: the drifting electrons.
538: In contrast, a dependence of oxygen attachment on quencher content 
539: (beyond electron energy contribution) has been measured for 10\% and 20\% CH$_4$ 
540: in Ar \cite{aa:huk}.
541: This difference between SF$_6$ and O$_2$ attachment is probably the result of 
542: different lifetimes of the corresponding excited states of the negative ions.
543: Indeed, the autodetachment lifetime is larger than 1~$\mu$s for SF$_6^{*-}$ 
544: \cite{aa:chris2}, comparable with the drift time in our detectors, 
545: whereas for O$_2^{*-}$ it is about 10$^{-4}$~$\mu$s \cite{aa:kok}.
546: 
547: \section{Methods for checking the gas quality} \label{aa:check}
548: 
549: Using xenon supplies as SF$_6$-free as possible is an important requirement. 
550: In the following we describe our monitoring procedures, i.e. gas chromatography
551: and $^{55}$Fe pulse height measurements using dedicated monitor detectors.
552: 
553: Gas chromatography allows the detection and quantification of traces of 
554: pollutants in a gas, by separating the different species contained in the 
555: sampled gas in a chromatograph column, and by detecting them in a suitable 
556: detector placed downstream of the column. The separated effluents give rise to 
557: characteristic peaks in a time diagram which can be identified and quantified 
558: after proper calibration of the device.
559: Since SF$_6$ has a high electron capture cross section, an Electron 
560: Capture Device (ECD) is a suitable detector. An ECD consists of a cavity 
561: through which a so-called make-up gas (nitrogen) flows. 
562: Electrons from a beta source ($^{63}$Ni, maximum energy 66 keV) partly 
563: ionise the nitrogen gas. The total current produced is collected 
564: by an electrode. 
565: If an electronegative substance flows through the cavity at a given time, 
566: the missing collected charge is converted into a
567: peak in the corresponding chromatogram. The area under the peak is proportional 
568: to the amount of electrons captured.
569: 
570: \begin{figure}[hbt]
571: \centering\includegraphics[width=.68\textwidth]{att13.eps}
572: \caption{Chromatograms of syringe-injected samples of three xenon supplies: 
573: SF$_6$-free (Messer) and with SF$_6$ contaminations of 0.3 ppm (Linde2) and 
574: 1.1 ppm (Linde1).}
575: \label{aa:chro} 
576: \end{figure} 
577: 
578: An ECD has been connected to our gas chromatograph \cite{aa:finn} 
579: in order to analyse our different supplies of xenon. 
580: Concentrations as low as 1~ppb are detectable.
581: The device has been calibrated for SF$_6$ by putting small amounts of this 
582: gas into a glass container filled with helium. Special care has to be 
583: taken with the injection syringes since they get temporarily contaminated 
584: when  exposed to high concentrations of SF$_6$.
585: In addition to the two xenon supplies (Messer and Linde1) used for the beam 
586: measurements reported above (section \ref{aa:sf6}), we have also investigated 
587: a more recent supply from Linde (Linde2).
588: The resulting chromatograms for the two kinds of xenon from Linde and the xenon 
589: from Messer-Griesheim are shown in Fig.~\ref{aa:chro}. 
590: The calibration yields 1.1 and 0.35 ppm SF$_6$ for the Linde1 and Linde2 gas 
591: samples, respectively, with an error of about 15\%. The xenon from 
592: Messer-Griesheim showed 1.5 ppb SF$_6$, most probably coming from the 
593: contaminated syringe.
594: The injection method also leads to inevitable air contamination, as revealed by 
595: the oxygen peak (corresponding to 400 ppm) in the chromatograms.
596: 
597: \begin{figure}[hbt]
598: \centering\includegraphics[width=.57\textwidth]{att14.eps}
599: \caption{Comparison of pulse height distributions of $^{55}$Fe in 
600: Xe,CO$_2$(15\%) for three supplies of xenon.}
601: \label{aa:fe} 
602: \end{figure} 
603: 
604: Gas chromatography is a simple, accurate and economic 
605: technique, but the chromatograph itself is a rather
606: expensive device. However, for most detector applications, one does not
607: necessarily need chromatography
608: in order to assess the gas quality. Measurements with a $^{55}$Fe source on 
609: standard detectors are sensitive enough to reveal possible problems due to
610: attachment.
611: In our case, in order to minimize Xe consumption for such tests, we have 
612: built small monitor detectors for the special purpose of checking 
613: the gas quality using $^{55}$Fe pulse height measurements.
614: These monitor detectors have the same electrical field configuration as the 
615: drift chambers used for the beam measurements described above.
616: Their small volume of about half a liter minimizes the loss of xenon gas. 
617: A collimated $^{55}$Fe source is placed in front of the entrance window. 
618: As the X-rays are absorbed preferentially at the beginning of the drift 
619: region, the cluster of primary electrons drifts in most cases 3~cm and is 
620: subject to attachment over this distance.
621: For these measurements the anode voltage has been tuned for a gas gain of about 
622: 10$^4$. 
623: Here we have used a different preamplifier, with a gain of 6~mV/fC and 
624: noise of about 1000 electrons r.m.s.
625: As a result of a low flow through the detector, the contamination with O$_2$ 
626: and H$_2$O was 70 and 400~ppm, respectively.
627: 
628: Figure~\ref{aa:fe} shows pulse height spectra for the three supplies of xenon 
629: discussed above.
630: The value of the drift field (which corresponds roughly to the 
631: anticipated operational point of the final TRD in ALICE) is 0.71 kV/cm.
632: Compared to clean Xe (Messer), the 0.35 ppm SF$_6$ contamination in Linde2 
633: Xe leads to a pulse height distribution with a much smaller value for the 
634: main peak. 
635: In addition, a clear tail towards larger pulse height is seen, originating from
636: absorption of X-rays deeper into the drift region, and thereby subject to
637: less attachment loss.
638: In case of Linde1 (1.1~ppm SF$_6$) the $^{55}$Fe signal is completely
639: lost. The spectrum recorded is the result of X-rays absorbed in the vicinity 
640: of the anode and of cosmic-ray background. 
641: Notice that the number of counts is much smaller in this case
642: (for a comparable acquisition time).
643: Notice also that, for the Messer and Linde2 cases, the escape peak of Xe 
644: (at 1.76 keV, compared to 5.96~keV full energy of $^{55}$Fe)
645: is clearly visible 
646: (in case of Linde2, the escape peak is partially cut by the threshold).
647: 
648: \begin{figure}[hbt]
649: \centering\includegraphics[width=.55\textwidth]{att15.eps}
650: \caption{Centroid (upper panel) and energy resolution (lower panel) 
651: of pulse height distributions of $^{55}$Fe spectra in Xe,CO$_2$(15\%) 
652: as a function of drift field for two supplies of xenon.}
653: \label{aa:ed} 
654: \end{figure} 
655: 
656: When possible, we performed gaussian fits of the main peak (also plotted in 
657: Fig.~\ref{aa:fe}) and extracted the centroid and energy resolution. 
658: Both quantities are influenced by attachment.
659: In Fig.~\ref{aa:ed} we show the dependence of the amplitude of the main peak
660: and its FWHM on the drift field for Messer and Linde2 supplies. 
661: The clean gas shows the expected variation of amplitude as a function 
662: of drift field arising from gain increase due to the transparency of the cathode 
663: wire grid. 
664: For comparison we include a measurement with an Ar,CH$_4$(10\%) mixture 
665: (P10). Compared to this, at low fields, even for clean Xe, small 
666: deviations are seen, consistent with attachment on O$_2$; taking the
667: P10 signal as reference, 
668: we deduce an attachment coefficient of 506 bar$^{-2}\mu$s$^{-1}$ for the 
669: field value of 0.72 kV/cm, in reasonable agreement to the value determined
670: from beam tests (section~\ref{aa:oxygen}).
671: For the Linde2 case this dependence of the signal on electric field is much 
672: stronger, as a result of the attachment.
673: This is confirmed by the energy resolution (lower panel in Fig.~\ref{aa:ed}), 
674: which improves dramatically when going (as a function of drift field) from 
675: strong to weak attachment.
676: 
677: \begin{figure}[hbt]
678: \centering\includegraphics[width=.62\textwidth]{att16.eps}
679: \caption{Attachment coefficient on SF$_6$ in Xe,CO$_2$(15\%) as a function of 
680: drift field.}
681: \label{aa:atcoef} 
682: \end{figure} 
683: 
684: We have determined the attachment coefficient on SF$_6$ (under our experimental 
685: conditions mentioned above) using the ratio of the $^{55}$Fe amplitudes for the 
686: Messer and Linde2 cases. The drift time is extracted using the drift velocities 
687: calculated with GARFIELD and assuming that the absorption of X-rays takes place 
688: right at the entrance window.
689: The results are presented in Fig.~\ref{aa:atcoef} as a function of drift field.
690: As expected, the attachment coefficient on SF$_6$ is large, of the order of 
691: 10$^6$ bar$^{-2}\mu$s$^{-1}$.
692: In addition, it has a pronounced dependence on the drift field, in line with our
693: arguments presented in section \ref{aa:calc}.
694: If we consider the absorption length of 5.9~keV X-rays in the Xe,CO$_2$(15\%) 
695: mixture, which is about 3 mm, the attachment coefficient would increase by 10\%.
696: For comparison, we include in this plot the results obtained from the beam 
697: mesurements, taking as a reference the time distribution of signals measured 
698: in the clean case and normalized in the amplification region. 
699: Taking into account the uncertainties, in particular coming from the 
700: normalization for the beam measurements, the extracted values for the attachment
701: coefficient are in reasonable agreement.
702:  
703: 
704: \section{Summary} \label{aa:conc}
705: 
706: We have performed measurements of pulse height distributions in drift chambers
707: operated with Xe,CO$_2$ mixtures.
708: After studying the general behavior of these distribution under clean 
709: conditions, we have investigated the role of oxygen and SF$_6$ contamination 
710: of the detection gas. 
711: A small signal loss due to attachment is seen for O$_2$ impurities 
712: up to a few hundred ppm. 
713: In case of SF$_6$, a contamination even at the level below 1~ppm 
714: produces a dramatic loss of signal over our drift length of about 3~cm.
715: Attachment on SF$_6$ is studied here for the first time 
716: concerning its practical implications for gas detectors.
717: As the SF$_6$ was found accidentally in some xenon supplies, it is important to
718: have a careful monitoring of the SF$_6$ contamination when precision measurements
719: are performed using Xe-based gas mixtures in drift chambers.
720: We have used ECD gas chromatography analysis to detect and quantify small traces 
721: of SF$_6$.
722: We have shown that measurements of $^{55}$Fe signals in monitor detectors are 
723: very sensitive to SF$_6$ contamination, thus allowing an inexpensive in situ 
724: check of the gas quality.
725: 
726: \section*{Acknowledgments}
727: We acknowledge the skills and dedication of A. Radu and J. Hehner in building
728: our detectors.
729: We are indebted to S. Ilie and C. Jeanpetit for the first analysis of our gas.
730: We appreciate the advice from R. Veenhof concerning the GARFIELD calculations.
731: We acknowledge P. Szymanski for help during the measurements at CERN.
732: 
733: \begin{thebibliography}{00}
734: 
735: 
736: \bibitem{aa:tdr} {\em ALICE TRD Technical Design Report}, CERN/LHCC 
737: 2001-021, October 2001;
738: {\it http://www.gsi.de/$\sim$alice/trdtdr}.
739: \bibitem{aa:chris2} L.G. Christophorou, Atomic and Molecular Radiation 
740: Physics, Wiley-Interscience, 1971 (Chapter 6).
741: \bibitem{aa:chris} L.G. Christophorou and J.K. Olthoff, 
742: J. Phys. Chem. Ref. Data 29 (2000) 267.
743: \bibitem{aa:huk} M. Huk, P. Igo-Kemenes and A. Wagner, \NIMA{267}{1988}{107}.
744: \bibitem{aa:bb} F. Bloch and N.E. Bradbury, Phys. Rev. 48 (1935) 689.
745: \bibitem{aa:wen} S. Wenig, R. Bock, H.G. Fischer, R. Sendelbach and R. Stock,
746: GSI Scientific Report 1993, 94-01, p. 264 (1994).
747: \bibitem{aa:kok} Y. Kokaku, Y. Hatano, H. Shimamori and R.W. Fessenden, 
748: J. Chem. Phys. 71, (1979) 4883.
749: \bibitem{aa:kad} J.A. Kadyk, \NIMA{300}{1991}{436}.
750: \bibitem{aa:wat} Y. Watase et al., \NIMA{248}{1986}{379}.
751: \bibitem{aa:zeus} R.D. Appuhn, K. Heinloth, E. Lange, R. Oedingen and 
752: A. Schl\"osser, \NIMA{263}{1988}{309}.
753: \bibitem{aa:hol} M. Holder and H. Suhr, \NIMA{263}{1988}{319}.
754: \bibitem{aa:d0} J.-F. Detoeuf et al., \NIMA{265}{1988}{157}.
755: \bibitem{aa:yu} B. Yu et al., \IEEE{38}{1991}{454}.
756: \bibitem{aa:mbs} H.G. Essel and N. Kurz, \IEEE{47}{2000}{337}.
757: \bibitem{aa:gsipi} J. Diaz et al., \NIMA{478}{2002}{511}.
758: \bibitem{aa:cernpi} CERN PS, {\it http://psdoc.web.cern.ch/PSdoc/acc/pscomplex.html}.
759: \bibitem{aa:andr} A. Andronic et al., \IEEE{48}{2001}{1259} [nucl-ex/0102017].
760: \bibitem{aa:messer} Messer-Griesheim, {\it http://www.messer.de/}.
761: \bibitem{aa:linde} Linde AG, {\it http://www.linde.de/}.
762: \bibitem{aa:garf} R. Veenhof, \NIMA{419}{1998}{726};
763: {\it http://consult.cern.ch/writeup/garfield/}.
764: \bibitem{aa:magb} S.F. Biagi, \NIMA{421}{1999}{234}.
765: \bibitem{aa:heed} I. Smirnov, HEED, an ionization loss simulation program, 1997.
766: \bibitem{aa:kunst} T. Kunst, B. G\"otz and B. Schmidt, \NIMA{324}{1993}{127};
767: B. Schmidt, private communication.
768: \bibitem{aa:becker} U. Becker et al., \NIMA{421}{1999}{54}.
769: \bibitem{aa:finn} Thermo Finnigan, {\it http://www.thermo.com/}.
770: 
771: \end{thebibliography}
772: 
773: \end{document}
774: 
775: