Precios unitarios aproximados por combinación de material de pavimento propuesto:

tum efficiency from the literature

The effective CsI QE has been largely investigated by several groups, in particular, in the context of the RD26 collaboration [110], as the goal of this collaboration was to develop large gaseous photon detectors for particle

identification employing CsI photocathodes. A protocol for CsI deposition that maximises the effective QE and allows to obtain reproducible results is one of the achievements of RD26 collaboration. The CsI is evaporated by Joule effect in vacuum (10−7 _{to 10}−6 _{Torr), at rates between 2 to 10 nm/s}

and at 50 to 60 ◦_{C. The substrates are accurately cleaned and out-gassed}

at 60 ◦_{C under vacuum before the evaporation. The crucibles, closed by}

shutters, are kept at evaporation temperature for a few minutes before start- ing the coating. The same shutters are closed when the desired thickness of the CsI film (typically 600 nm) is reached. After evaporation, the layer is kept for several hours under vacuum at 50-60 ◦_{C. Besides, the morphology}

of the CsI coating depends on the substrate. The morphological informa- tion is obtained by scanning Electron Emission (SEM), Scanning Tunnelling Microscope (STM), by Electron Microscopy for Chemical Analysis (ESCA) and Photo-Emission Electron Microscopy. The structural information is ob- tained from the X-rays diffraction measurements. The outcome is that the CsI layers exhibiting high effective QE are characterised by a structure of contiguous micro-crystals, which is obtained using well polished substrates, while the CsI deposited on “rough” printed circuit boards exhibits lower QE and a very inhomogeneous texture. Moreover it was observed (X-rays diffraction spectra) that the CsI evaporated on pure Cu substrates and on gold-covered copper exhibits peaks of pure Cs and pure iodine instead of CsI peaks. In fact, during the evaporation, Cu promotes the dissociation of the CsI or prevents its formation from the vapour phase. To prevent this phenomenon, a chemical deposition of a nickel layer covered by a film of gold is applied on the Cu substrate. Figure 8.1 summarises the CsI effective QE versus wavelength in vacuum obtained from a large set of measurements performed within the RD26 collaboration.

About QE in vacuum, it is important to remark that it depends also on the electric field and it reaches a plateau at about 100 V/cm, as shown in Fig. 8.2.

About the backscattering, the photoelectrons return to the photocath- ode because of the elastic scattering with the gas molecules. The large mass difference between the elastically scattered electrons and gas molecules de-

8.2 EFFECTIVE CsI QE FROM LITERATURE 121

Figure 8.1: The CsI effective QE from RD26 measurements: reference curve, namely maximum obtained using a stainless steel substrate (dotted line); band of values obtained using a PCB substrate without Ni coating (black band) and with Ni coating (white band) [110].

Figure 8.2: Measurement of the photo-current versus drift field in vacuum and in some gasses at atmospheric pressure [97].

termines a wide range of scattering angles, with a small energy transfer per collision. For energies below ionisation threshold, the backscattering effect is therefore more pronounced, in particular in nobles gasses, where the elastic cross section largely exceeds the inelastic one, because of the lack of vibra- tional and rotational levels.

At high electric field the energy of photoelectrons becomes high enough to excite and to ionise the molecules of the gas and the probability for elas- tic scattering decreases due to the opening of new, inelastic channels in the electron-molecule collisions.

The inelastic channels don’t lead to backscattering, since the photoelec- trons losses their energy without a substantial change of its direction of mo- tion. When the probability of the elastic collisions drops down to zero, the QE in gasses should reach the vacuum value. Recently Monte Carlo simula- tions have been performed and compared with the experimental data in order to investigate better the photoelectron backscattering effects in several gas mixtures. In particular, Neon mixtures with quenchers (Figs. 8.3 and 8.4) have been considered [107]. Figure 8.5 summarises a large set of experimental data concerning the extraction efficiency normalised to the vacuum measure- ment. The extraction efficiency for CH4 and CF4 is larger than 0.7 times the

value in the vacuum already at 1 kV/cm thanks to large vibrational cross- section as extensively reported in literature. The extraction efficiency of the Neon with 10 % of methane is around 0.6 times the value in the vacuum for 1 kV/cm at atmospheric pressure [107].

In [111], a CsI QE enhancement has been also observed in correspondence of breakdown events, when the detector is in multiplication mode or under high light flux. A concentration of positive ions on the CsI nonconductive surface may become large enough to create a high electric field within the CsI film. This could be responsible for the increase of the photoelectron ex- traction probability. Positive ions deposited on such regions can be gradually neutralised by charge transfer processes, for example as a charge exchange reaction with gas molecules. The neutralisation process can depend on the nature of the gas. Most likely, this phenomenon is at the base of the long recovery time after a discharge in gaseous detectors with CsI photocathodes (Sec. 2.2).

8.2 EFFECTIVE CsI QE FROM LITERATURE 123

Figure 8.3: Photoelectron ex- traction efficiency measurements (symbols) compared with the Monte Carlo results (full and dot- ted curves) as a function of the reduced field (bottom scale) and electric field at 1 atm (top scale) for Neon/methane based gas mix- tures [107].

Figure 8.4: Photoelectron ex- traction efficiency measurements (symbols) compared with the Monte Carlo results (curves) as a function of the concentration η of methane in the Ne/CH4 mixture

for a set of reduced electric field,

E p [107].

Figure 8.5: The extraction efficiency normalised to the vacuum extraction efficiency as a function of the reduced electric field (bottom scale) and the electric field (top scale) for several gasses and gas mixtures, experimental data. The curves joining the data points are guidelines [107].