Trapping molecules using photostop

Abstract

Successful magnetic trapping of both SH at a density of 2.4 x 10^5 cm^-3 and atomic oxygen at a density of 3000 +/- 900 cm^-3 has been demonstrated with respective trapping lifetimes of 40 +/- 10 ms and 82 +/- 3 ms inside a 0.41 T deep, anti-Helmholtz configuration magnetic trap.

To achieve trapping of these species, a stable precursor was cooled by supersonic expansion in a molecular beam and photodissociated in such a way as to match and oppose a photofragments recoil speed to the precursors molecular beam speed. This process is called photostop.

Velocity-map images have been taken of photostop of atomic oxygen from photodissociation of NO2 in a molecular beam. They show that atomic oxygen can be photostopped from a variety of NO photodissociation co-fragments with a suitably broad molecular-beam velocity-distribution. Although photostop is simple and cheap to use, the trapping lifetimes and densities that have been achieved in this thesis are not competitive compared to other methods of trapping similar
molecular species. The main problem that presents universally in photostop experiments is due to the large number of fast molecules from the supersonic expansion that pass through the trap after the photostop event. Only a small fraction of the molecules in the molecular beam are stopped and the remaining ones cause significant trap loss through high energy elastic collisions with trapped
species.

Work has been carried out to develop the cavity-enhanced laser-induced-fluorescence (CELIF) technique to be used more widely in the cold molecules field. The technique was employed unsuccessfully to measure photostopped SH radicals but was used elsewhere to measure cold SD radicals
after Stark deceleration. In addition, the technique has been initially developed for use inside magnetic traps.

Clusters of NO.Ar(n) have been studied using [1+1'] REMPI to assess the viability of a new soft-ionisation technique for molecular clusters of atmospheric interest. This work did not show conclusive evidence that [1+1'] REMPI would reduce cluster fragmentation and displayed some interesting but poorly understood data.