Particle Identification with a Fine Sampling Ionization

Introduction charge collection) of Q 104. The drift field was chosen to give an electron drift velocity of 1 cm/usec. %1/5 the saturated velocity in this gas mixture. The volume of uniform field (adequate for longitudinal drift measurenents of dE/dx) extends to within +3 mm of the sense plane and +l mm of the cathode planes. Hence, in principle, 80% of the track length through a chamber of this geometry is usable for ionization loss measurements. In a recent series of tests1 it has been shown that the sensitivity for charged particle identification via the relativistic rise in ionization loss can be enhanced by using longitudinal drift and fast Alectronics to effect very small sampling intervals. With the chamber geometry illustrated in Fig. la (i.e., operation in the "time expansion" mode) electrons from individual ionizing collisions drift sequentially onto the anode wire with a time structure determined by the drift velocity in the gas. After fast pulse shaping the resultant waveform is digitized with a flash-encoding ADC at a frequency of 1OCMHz. For the tests described in Ref. 1 the smallest sampling interval was l/4 mm, and the sensitivity of mass (velocity) discrimination was studied for 1 meter of track length with sampling intervals ranging from l/4 mmfsample to 16 mm/ sample, the upper and of this range being typical of conventional ionization sampling devices. The results of these initial tests are summarized in Fig. 2 as a function of the sample size. The result achieved with 1 meter of conventional sampling (ISIS-l, Ref. 2) is shown for comparison. For large sample sizes the results are comparable to those of traditional techniques, and a significant improvement is achieved when the sampling frequency is increased by roughly an order of magnitude. These results were gotten with a small test chamber in which meter-long tracks were simulated by piecing together many traversals of a single drift gap. We report here on the status of further studies to see whether this improvement in sensitivity can be duplicated with a chamber of practical dimensions and many readout channels.