3.4 Discussion

Typically the secondary electron yield of the investigated surfaces was lower than one at energies below 400eV per incident atom. No surface with outstanding yields was found. One of the most surprising findings was the decrease of the secondary electron yield first observed on MgO coated surfaces. Such a decay would render the surface useless for NPD style detectors. The decrease could be explained by the extended exposure of one specific spot of the surface to the beam causing a change in the surface configuration. The beam might sputter particles away from the surface (especially water). The surface would be cleaned over time and the yield would change - lowered in our case. However by exposing the surfaces to UV radiation or by heating, the surfaces recovered by some extent. UV exposure also would clean the surface by cracking water molecules on the surface. The time constants of the decay did not depend whether the surface was exposed the first time to the beam after the pump down of the chamber or whether the surface was exposed to UV radiation or heated (Figure 3.12). An extended time in the chamber (24h) at moderate vacuum conditions (10mbar) without exposure to the beam did not change the secondary electron yield. Any decay stopped during this time. The presence of insulating materials near the surface seems to be related to the decay: AlO and to a lesser extent MgO (Figure 3.12). This might be due to surface charging, at least on a very localized basis despite the thin insulating layers (typically less than 150nm.) Samples with MgF and Al top-layers did show a similar behavior (Figure 3.11) but probably the aluminum top layer on these samples was oxidized when exposed to air and thus be insulating.

Metallic samples or samples with a thick metallic overlayer did not show a time dependent secondary electron yield also suggesting that the decrease might be related to the surface conductivity.