Different Configurations of the Even-Lavie Valve

The results discussed above were obtained for a nozzle orifice of 100 µm followed by a conically shaped exit channel with an opening angle of 110°. However, the front gasket between plunger and nozzle orifice had an orifice of only 60 µm diameter thus limiting the orifice. The same experimental series were also recorded with a former configuration of the valve differing in the opening diameter of the front gasket (100 µm), the recoil spring and the plunger. The recoil spring had a lower force constant and the plunger was made of a weaker magnetic alloy and had a slightly smaller curvature of its polished sealing surface. Changing from the 100 µm-setup to the 60 µm-setup reduced the gas load by one order of magnitude and lowered the heat intake by the operating valve. Both contributes to the ability to work at higher repetition rates.

The Rayleigh scattering signal showed qualitatively the same dependencies as described above. Though, the signal was more sensitive on changes of stagnation conditions and the repetition rate. The higher sensitivity is presumably due to the somewhat higher optimum driving current corresponding to a capacitor voltage of ≈ 23 V compared to

≈19 V for the 60µm-setup. This lead to a larger heat intake limiting the repetition rate and thus no reasonable Rayleigh signal was observed already at 30 Hz. Despite of the larger amount of expanded gas, the peak intensities at low repetition rates were found

to be equal for the two setups.

This was also found for the peak intensity of the LIF signal of Pc doped into the droplet beam for repetition rates up to 50 Hz. At higher rates the signal generated with the 100µm-setup drops faster due to the larger heat intake and gas load and no signal could be detected at 200 Hz. The higher gas load prevents cooling in the expansion efficient enough to form droplets at elevated pressures in the source chamber due to the increased repetition rates. Differences in the performance of the two setups with respect to the LIF signal are the driving current and especially the different dependence of the LIF signal on the stagnation temperature. The optimum driving current for LIF signal corresponded to a capacitor voltage of ≈19 V and is thus much lower than for the Rayleigh signal, whereas the same current was optimum for both kinds of signal with the 60µm-setup.

Time profiles for different stagnation conditions recorded via LIF using the 100 µm-setup are shown fig. 3.23 (a). For each stagnation pressure the signal intensity shows no significant dependence on the nozzle temperature up to a certain temperature above which it starts to disappear and thus contrasts the behavior found for the 60µm-setup.

The maximum temperature raises with increasing stagnation pressure as found for the 60µm-setup.

The line shape studies at the origin of Pc doped into the droplet beam generated with the 100µm-setup are shown in fig. 3.23 (b) for a wide range of stagnation conditions. As discussed for the other configuration of the valve, changes in the stagnation conditions hardly affect the line shape depending on the droplet size distribution. Changing the conditions in a way that larger droplets are expected to be formed only leads to a slight blue shift of the peak. Fig. 3.23 (c) shows time profiles of the doped beam recorded under conditions optimum for Rayleigh-signal with the dye laser on and off resonance and also a profile recorded with the Ar⁺ laser without a pick-up cell inside. As for the 60 µm-setup, the series of data presented reveal a bimodal droplet size distribution. Depending on the stagnation conditions, either large droplets causing Rayleigh scattering or smaller droplets which can be doped are formed more efficiently. The size of the dopable droplets can be varied only within a very small range compared to the continuous beam even upon drastically changing the stagnation conditions, e.g. from 20 bar, 20 K to 80 bar, 7 K. At the same time, the droplet size distribution is hardly affected and does not reveal a log-normal distribution.

To conclude, the droplet beam produced with the two setups showed minor differences.

The 60 µm-setup provides with the same peak intensity for Rayleigh and LIF signal at low repetition rates while the heat intake and gas load is decreased and enables to work reasonable with higher repetition rates up to 500 Hz. Next to this two well investigated configurations of the valve preliminary experiments were also made with other configurations. They correspond to the 100 µm-setup described above but with different exit channels. Instead of the conical shaped exit channel a trumpet shaped exit

Fig. 3.23: Left: Time profiles of the doped helium droplet pulse operating at 10 Hz with stagnation pressure and nozzle temperature as indicated, while the gas flux was kept constant (a). Right: Excitation spectra at the electronic origin of single Pc in helium droplets recorded for different expansion conditions as indicated shifted by 15088.92 cm⁻¹ .(b) Time profiles of the doped droplet beam while operating under ideal conditions for Rayleigh signal with the dye laser once in (red) and off (black) with the resonance of single Pc. Also shown is the scaled time profile recorded with the Ar+-laser under identical conditions without the pick-up oven inside (green).(c) All data were recorded with the setup including the 100 µm gasket.

channel was used, though no significant differences in the performance were observed.

Another configuration was a conical shaped channel with a diameter of 250 µm and an opening angle of 40°. Rayleigh and LIF signals could only be observed for repetition rates up to 20 Hz and were not more intense than with the setups discussed above.

The larger amount of gas prevents a cooling in the expansion efficient enough to form droplets at higher repetition rates. Further, the smaller opening angle provides a more gradual cooling, i.e. the expanded gas is cooled within longer distances and reaches the temperatures required for droplet formation at distances where the density of gas is already too low to enable cluster formation requiring three body collisions. [Eve]