Electron identification

 All results reported here were obtained with beams of 20 GeV pions and electrons. Electron identification using transition radiation makes use of the difference in the energy deposited in the straws by electrons and by charged pions, which are the main background source to electrons at the LHC. For electrons, the tail above 5-7 keV is dominated by transition radiation hits; for pions it is mostly due to $\delta$ -rays. The pion rejection is calculated by counting the number of high threshold hits on reconstructed tracks for pions and electrons.

  

Figure 5.18: The distribution in the number of high-threshold hits (above 6 keV) on reconstructed tracks from 20 GeV pions and electrons in a magnetic field of 0.8 T.

  

Figure 5.19: Distribution of the number of high-threshold hits for pions and electrons in a magnetic field of 0.8 T. The crosses represent the data and the histograms a fit of the data to a binomial distribution.

The distribution of high-threshold hits on the reconstructed tracks is expected to follow a binomial law. This is illustrated in fig. 5.19, where a fit has been performed to the electron and pion data of fig. 5.18. The distributions follow the binomial law across several orders of magnitude and do not display any poorly understood tails.

By requiring more than a certain number of high-threshold hits along the track the efficiency for misidentifying pions as electrons is measured as a function of the electron efficiency. In fig. 5.20 is shown the result of applying this procedure to the distributions from fig. 5.18, which were obtained with data taken in a 0.8 T field and with a high-threshold discriminator setting of 6 keV. For an electron efficiency of 90%, the measured pion efficiency is around 1.2%, corresponding to a rejection factor of 80 against pions.

  

Figure 5.20: Pion versus electron efficiency, obtained as a function of the number of high-threshold hits, as shown in fig:HighThreshold, required on reconstructed tracks.

  

Figure 5.21: The pion efficiency as a function of the discriminator threshold for a fixed electron efficiency of 90%, and for 20 GeV pions and electrons in a 0.8 T magnetic field.

The high-threshold setting for the TRDA chip can be varied and the pion rejection depends to some extent on the value chosen. In fig. 5.21 the pion efficiency for a fixed electron efficiency of 90% is shown as a function of the setting of the high-threshold discriminator. The electron identification performance is seen to be quite stable for thresholds between 5 and 7 keV. Further results reported in this paper for pion efficiencies are given for a threshold of 6 keV and an electron efficiency of 90%.