We have a molecular beam set-up for collisions between alkali atoms and halogen molecules or atoms The primary beam of the alkali atoms is monitored by two surface-ionization detectors, before and after crossing the secondary beam. (photo)
The surface ionization detector consists of a 0.1mm hot iridium wire. The iridium wires were kept at 1325 K and periodically flashed in oxygen at 10-5 torr.
We use Ir wire, because the work function of this element is quite high: about 5.4 eV .
Nevertheless, we sometimes had problems detecting Na and Li, especially in the low-energy range. We found that we could overcome these problems by a procedure, similar to that used by Touw and Trischka. We treated the wire at a working temperature of 1325K in an oxygen bath of torr for 5 minutes. This treatment guarantees a stable high work function of the detecting wire for at least one day. which is essential for the detection of sodium and lithium.
It was checked that the efficiency of the detector was insensitive for small temperature changes around 1325K. For that purpose, the relative detection efficiency was measured as a function of the wire temperature and the incident energy for the three different alkali beams. This was done by keeping one of the detectors at constant temperature and varying the temperature of the other one. The temperatures at which the detection efficiency is no longer dependent on temperature was determined to be 875 K for potassium, 945 K for sodium and 1095 K for lithium. The setting of the temperature was well above these values, in order to ensure reproducible detection. The temperature of the wire was measured by a pyrometer and corrected for the emissivity .
This, however, does not guarantee an energy-independent efficiency. Reflection of atoms on the surface is possible at high energies and hence the detection efficiency may be lower. Such an energy dependence was found by different authors for potassium atoms above 3eV. , , .
All the measurements presented here should be corrected by a factor taking this effect into account. But, as the detection efficiency varies smoothly with the energy and because the energy range is not very large, it does not substantially affect the measurements.
The beam monitor placed before the interaction region intercepts 1/5 of the beam.