CERN Accelerating science

Scheme of the \SI{70}{\meter} long KATRIN beamline: a) rear section, b) windowless gaseous tritium source, c) transport section, d) spectrometer section with aircoil, e) detector section.

Schematic of the Rydberg induced background mechanism. Highly excited Rydberg atoms can be created via $\alpha$-decays in the surface and accompanying sputtering processes. If they are ionized, e.g. by thermal radiation, low-energy electrons are generated. The whole process is enhanced by the implantation of daughter isotopes which in turn initiate additional sputtering. To simulate background mechanisms such as the Rydberg mechanism, a \radium{223} source was installed at the spectrometer surface level. The wall of the vacuum vessel is at high voltage. The inner electrode (IE) is up to \SI{200}{\volt} more negative in electric potential.

Installation of the \radium{223}-source on the top side of the KATRIN Main Spectrometer. The source was mounted inside the vacuum chamber at the tip of the magnetically coupled stainless steel rod of the linear motion device, which can be lowered into the spectrometer.

: ~volume magnetic field configuration : ~surface magnetic field configuration

: Measurement configuration: volume \protect\subref{fig:volume} and surface \protect\subref{fig:surface} magnetic field configuration enable the investigation of volume and surface-induced events. The colored lines show the magnetic field lines reaching the silicon detector. Each line corresponds to a boundary between the ring-wise segmentation of the detector. : Caption not extracted

: ~volume magnetic field configuration : ~surface magnetic field configuration

: The \thorium{228}-induced rate in volume magnetic field setting versus time after closing the valve to the \thorium{228} source \protect\subref{fig:BGafter228ThVolume}. A fit of equation \ref{eq:thoriumFitExp} to the measurement data was performed. The observed half-life of \SI{10.66 +- 0.16}{\hour} matches the literature value \SI{10.64 +- 0.01}{\hour} \cite{Browne2005} for \lead{212}. The background offset is \SI{0.763 +- 0.005}{\cps}. \thorium{218}-induced rate in surface magnetic field setting versus time after closing the valve to \thorium{228} source \protect\subref{fig:BGafter228ThSurface}. A fit of equation \ref{eq:thoriumFitBateman} to the measurement data was performed. The background offset is \SI{2891 +- 3}{\cps}. : Caption not extracted

Normalized distributions of events for each detector ring. The distribution was normalized by dividing by the sum of all events and multiplying by the number of detector rings.

Correlation between the background rate from the spectrometer surface and the spectrometer volume. The Pearson correlation coefficient is 0.999.

: ~spectrometer surface : ~spectrometer volume

: The measured background rates after the withdrawal of the \radium{223} source in the \protect\subref{fig:ra_surface} surface and \protect\subref{fig:ra_volume} volume magnetic field configurations, see \cref{fig:fieldconfig}. Both decay exponentially. A fit of \cref{eq:thoriumFitExp} to the measurement data is performed. The fitted half-life is in agreement with the expected one of \lead{211}, $t_{1/2}=$~\SI{36}{\minute}\,=\,\SI{0.6}{\hour}. The fitted constant offsets are \SI{1432 \pm 3}{\cps} and \SI{0.87 \pm 0.02}{\cps} for the surface and volume magnetic field configuration, respectively. : Caption not extracted

Background rate relative to zero inner electrode potential (the actual voltage reading slightly deviates for the zero setpoints) for the normal Main Spectrometer background and the $\alpha$ source induced background.