A source of translationally cold molecular beams

Currently the fields studying or using molecules with low kinetic energies are experiencing an unprecedented growth. Astronomers and chemists are interested in chemical reactions taking place at temperatures below or around 20 K, spectroscopists could make very precise measurements on slow molecules and molecular physicists could chart the potential energy surfaces more accurately. And the list continues. All of these experiments need slow molecules, with kinetic energies from around 10 cm-1 down to 0. Several designs of cold sources have already been made. The most interesting ones are presented.;This work describes the design and the testing of a cold source based on the collisional cooling technique: the molecules of interest are cooled well below their freezing point by a precooled buffer gas. This way condensation is avoided. The source is a copper cell cooled to 4.2 K by an external liquid helium bath. The cell is filled with cold buffer gas (helium). The molecules of choice (ammonia) are injected through a narrow tube in the middle of the cell. The cold molecules leave the cell through a 1 millimeter hole. Two versions of pulsing techniques have been employed: a shutter blade which covers the source hole and opens it only for short moments, and a chopper that modulates the beam further downstream. Both produced pulse lengths around 1 millisecond.;The source is tested in an experiment in which the emerging molecules are focused and detected. Time of flight technique is used to measure the kinetic energies. Two detectors have been employed: a microwave cavity to analyze the state of the molecules in the beam, and a mass spectrometer to measure the number density of the particles.;The molecules coming out of the source hole are formed into a beam by an electrostatic quadrupole state selector. The quantum mechanical aspects and the elements of electrodynamics involved in the focusing are described. A computer simulation program is presented, which helped calculate the trajectories of the molecules from source to the final detector. The simulations revealed interesting artifacts the focuser can make: if a short state selector is used, the speed distribution of the beam arriving at the detector is strongly modulated by a "focuser transfer function". The essence of the phenomenon is caused by the inability of the focuser to focus all speed at the same time: a few velocity groups are well focused, while many others are defocused. The transfer functions gained from the simulations were used to correctly interpret the detector signal.;The design of the microwave cavity (Fabry-Perot) and its driving electronics is presented. Due to unavoidable noises present in the circuit, the use of this detector was limited to the preliminary experiments.;A simple, commercial mass spectrometer was used. To improve its characteristics, a few modifications have been made. With the improved version, only the water background (an interfering signal at mass 17 amu) was disturbing the signal. A liquid nitrogen trap has been used to minimize the background.;The experimental measurements revealed several undesired effects accompanying the production of the slow beam: cluster formation, snow accumulation inside the cell and very high sensitivity to the balance of the two gas pressures. Clusters made up the majority of the emitted particles. Switching from the shutter-blade pulsing (closed cell) to the chopper modulation (open cell) improved the ratio of the useful molecules in the beam significantly. The snow accumulation limited the useful experimenting time to around 20 minutes. After this the accumulated ammonia snow either perturbed the signal significantly or obstructed the source hole completely. A very high sensitivity to the ratio of the buffer gas pressure to the ammonia injection rate has been noticed. A 30% deviation from the optimum value can change the speed distribution and the amount of emitted molecules completely. The stabilization of the ballast pressures was necessary.;The beam obtained in the optimal regime of the source cell is presented and analyzed. The characteristic values of most interest in a future collision experiment are the following: the speed distribution of the beam at the detector has a height of the order of 105 molecules per unit speed interval and per pulse; the useful speed interval is from 55 m/s to 500 m/s.