The CERN Resonant Weakly Interacting Sub-eV Particle Search (CROWS)

The subject of this thesis is the design, implementation and first results of the ``CERN Resonant WISP Search'' (CROWS) experiment, which probes the existence of Weakly Interacting Sub-eV Particles (WISPs) using microwave techniques. Axion Like Particles and Hidden Sector Photons are two well motivated members of the WISP family. Their existence could reveal the composition of cold dark matter in the universe and explain a large number of astrophysical phenomena. Particularly, the discovery of an axion would solve a long standing issue in the standard model, known as the ``strong CP problem''. Despite their strong theoretical motivation, the hypothetical particles have not been observed in any experiment so far. One way to probe the existence of WISPs is to exploit their interaction with photons in a ``light shining through the wall'' experiment. A laser beam is guided through a strong magnetic field in the ``emitting region'' of the experiment. This provides photons, which can convert into hypothetical Axion Like Particles or other WISPs. The photons are blocked by a wall but the WISPs practically do not interact with matter and can hence propagate to the ``detection region''. They can reconvert into photons, which subsequently can be detected. A hypothetical WISP would reveal itself as light, apparently shining through the opaque wall. The effect is not explained by classical physics and persists, even with flawless electromagnetic shielding. These kind of experiments have already been carried out at several institutes in the optical regime. To improve sensitivity, it has been suggested to implement Fabry-Pérot resonators on the emitting and detecting side of the experiment. The CROWS experiment is based on the same principles, but operates in the microwave regime at a frequency of several GHz, replacing the laser by a signal generator and the optical detector by a sensitive receiver. CROWS exploits the fact, that microwave cavity resonators are technologically easier to implement than optical Fabry-Pérot ones. On the other hand, electromagnetic shielding between the fields in the two cavities is one of the primary concerns %most critical points to achieve a sufficient detection sensitivity. An attenuation factor in the order of 300~dB is required to reject direct leakage below the detection threshold. The shielding has been organized in several, demountable shells, providing $\approx$~100~dB of attenuation each. This allowed each shell to be characterized by conventional measurement techniques. Moreover, a test signal was emitted within the shielding to ensure that its performance did not deteriorate during the experimental run e.g., due to vibration or corrosion. Two shielding enclosures have been built, one for the detecting cavity, which was placed within a 3~T magnet and one for the components of the signal receiver, which was placed outside the magnet. Transmission of analog signals between the two shielding enclosures has been accomplished by optical transceivers, which had to be customized to operate in the strong magnetic field environment. A custom made double heterodyne signal receiver and a commercial signal analyzer have been utilized to record a 20~kHz or 2~kHz wide slice of the signal spectrum, where a WISP signal would have been expected. The power spectrum of the recorded time-trace has been evaluated, which corresponds to narrowband filtering. For a 29~h measurement run, resolution bandwidths of $< 10~\mu$Hz have been achieved. This allows to detect signals with less than -210~dBm of power, corresponding to a flux of $\approx 0.5$~photons/s for a 3~GHz signal. Both receivers were optimized for minimum frequency drifts by phase locking all oscillators to a common reference clock. No HSPs or ALPs were observed in the most sensitive measurement-runs of the CROWS experiment. For HSPs with a mass of $m_{\gamma'} = 10.8~\mu$eV, a limit was set on the coupling constant of $\chi < 4.1 \cdot 10^{-9}$, which allowed us to explore a previously unexplored region in the parameter space. The result corresponds to an improvement in sensitivity over the previous exclusion limit by a factor of $\approx 7$. For ALPs with a mass of $m_a = 7.2~\mu$eV we set a limit on their coupling constant of $g < 4.5 \cdot 10^{-8}$~GeV$^{-1}$. In a small mass range, this is more sensitive than other purely laboratory based experiments (namely laser LSW of the first generation like ALPS-1) but less sensitive than extraterrestrial experiments like CAST. This is the first time ALPs have been probed by a microwave based LSW experiment.