Definition of Project

One of the important features of PHENIX experiment is capability of particle identification (PID). High resolution time-of flight counter (TOF), ring-imaging Cerenkov counter (RICH) and Pb-Sc. electromagnetic calorimeter (EMCal) with timing measurement capability provides PID as shown in upper two rows of Fig.1. TOF is limited to moderately low $p_T$, while at high $p_T$ RICH provides pion identification. Clearly, within the present TOF & RICH configuration has significant hole of PID.

New features found in the first RHIC data has brought people's attention to an importance of particle identification in the region above 2 GeV/c. The addition of an aerogel Cerenkov counter with an index of refraction of  1.01 will provide PID of pion, kaon, and proton separation seemlessly upto $p_T$ of almost 10 GeV/c as illustrated in Fig.1.

図 1: Hadron PID of PHENIX and Aerogel Cerenkov Counter. See section 4 for the detail.

Aerogel Cerenkov counter with limited acceptance will be installed on the West Arm. Geometrical size of the entire counter will be $\sim 200 \times 200 \times \lesssim 45 {\rm cm}^3$ and is segmented to $\sim 200$ cells. According to current prototype design, each cell has $14 \times 14 \times 10 {\rm cm}^3$ of aerogel (refractive index of $\sim 1.015$) followed by an integration cube viewed by two 3'' PMT's (left of Fig.2).

図 2: Left; Current prototype with integration cube. Right; Stack of cells.

Test beam experiments done at KEK in May 2002 shows that positional uniformity with this configuration is found to be the best among other configurations. To elliminate dead space, cells will be stacked reversing the direction of the cell in every second row as shown in right of Fig.2. With this configuration, all the sensitive volume are kept in one plane and there is no dead space in between, which gives us uniform responses. Disadvantage of this configuration is that there is cells to which particles enters in reverse direction. According to our test beam at KEK, when a particle enters from the integration cube to the aerogel, number of p.e. is smaller by $10 \sim 20 \%$, which is not significant. We can obtain $\gtrsim 15$ p.e. by adjusting the thickness of the aerogel. We also have backup configurations which will be compared at next test beam experiment at KEK in Dec. 2002.