10 - H1 - Desy

66 Prompt photon selection
∆η < 0.4 and azimuthal angle ∆φ < 0.4 causing the estimated signal selection efficiency
drop by 3% on average.
Events
3
10
4
10
a. b.
Events
3
10
2
10
2
10
10
10
1
0 0.2 0.4 0.6 0.8 1
∆η
0 0.5 1 1.5 2 2.5 3
∆φ
Figure 5.3: The difference between photon candidate cluster and the closest enabled BT
in pseudorapidity ∆η (a) and azimuthal angle ∆φ (b). The vertical lines indicate the cut
values.
5.2.2 Veto on charged particles
Electrons and positrons shower in calorimeter in a very similar way to photons. Contrary
to them though they leave signal in all tracking detectors. A tracking veto is thus
commonly applied for the photon selection.
In this analysis the main detector providing the veto condition is the CIP. A sensitive
area in the CIP is defined for each photon candidate and each CIP layer. A cone of four
neighbouring pads in both z and φ dimensions around the photon candidate direction is
checked for a signal in each of the CIP layers. Candidates with at least one active pad
in at least two layers are vetoed as possible charged particles. The efficiency of the veto
condition has been studied using both single electrons MC and data and has been found
to be above 99%.
Electrons scattered under a wide angle are the main source of charged particle background.
Due to the Q 2 dependence of the cross sections, they are expected to appear mainly in the
backward region of the calorimeter and in this area the CIP veto has been strengthened
by the CJC tracking condition. All clusters in the polar angle range θ γ > 45 ◦ and with
distance of closest approach to the nearest CJC track extrapolated to the LAr surface
DCA track < 15 cm are removed from the sample of photon candidates.
5.2.3 Neutral hadrons background
The main background to the analysis comes from neutral hadrons decaying into multiple
photon final states. The great majority of it (87% as estimated in [66]) comes from
neutral pions decaying into two photons usually merged into one electromagnetic cluster.
Though definite removal of the neutral hadron background is not experimentally possible,