Cargando…

Influence of the cusp field on the plasma parameters of the Linac4 H− ion source

When the H$^−$ ion source of CERN’s Linac4 is operated in volume mode, a maximum of the extracted current is obtained at varying RF power. The power required for this maximum and its absolute value is strongly influenced by the cusp magnets installed at the source for electron confinement: without m...

Descripción completa

Detalles Bibliográficos
Autores principales: Briefi, S, Mattei, S, Lettry, J, Fantz, U
Lenguaje:eng
Publicado: 2017
Materias:
Acceso en línea:https://dx.doi.org/10.1063/1.4995736
http://cds.cern.ch/record/2319809
Descripción
Sumario:When the H$^−$ ion source of CERN’s Linac4 is operated in volume mode, a maximum of the extracted current is obtained at varying RF power. The power required for this maximum and its absolute value is strongly influenced by the cusp magnets installed at the source for electron confinement: without magnets, 15−20 mA are typically obtained at 20 kW whereas with magnets a factor of two more power is needed and 25−30 mA are achieved. In order to access the reasons behind the peaked performance with varying RF power and for determining the influence of the cusp field on the discharge, optical emission spectroscopy (OES) measurements of the atomic Balmer series and of the molecular Fulcher transition have been carried out. In all investigated cases, the gas temperature of the discharge has been virtually equal to the ambient temperature as the short discharge pulse length of 500 $\mu$s is not long enough for considerable heavy particle heating. When no cusp magnets are installed, the plasma parameters evaluated with the collisional radiative models Yacora H and Yacora H$_2$ show a minimum in the electron temperature of 3.25 eV and a maximum in the electron density of $4 \times 10^{19} m^{−3}$ and also in the vibrational excitation of the hydrogen molecule at 20 kW. Assessing the relevant production and destruction processes demonstrates that the H$^−$ yield is maximal at this point thereby explaining the optimum ion source performance. When the cusp magnets are applied, the same general trends are observed but the required RF power is a factor of two higher. The OES results indicate an optimum performance around 30 kW whereas the highest H$^−$ current is actually achieved around 40 kW. Furthermore, a higher H$^−$ yield is indicated without cusp magnets but a better ion source performance is observed with magnets. These differences can most likely be attributed to changing gradients in the plasma parameters which are not accessible by OES. Nevertheless, the obtained plasma parameters can be used as benchmark for RF coupling codes simulating the Linac4 ion source.