First results of the CERN Resonant Weakly Interacting sub-eV Particle Search (CROWS)

The CERN Resonant Weakly Interacting sub-eV Particle Search probes the existence of weakly interacting sub-eV particles like axions or hidden sector photons. It is based on the principle of an optical light shining through the wall experiment, adapted to microwaves. Critical aspects of the experiment are electromagnetic shielding, design and operation of low loss cavity resonators, and the detection of weak sinusoidal microwave signals. Lower bounds are set on the coupling constant $g=4.5\times {10}^{-8}{\mathrm{GeV}}^{-1}$ for axionlike particles with a mass of $m_a=7.2\mu \mathrm{eV}$. For hidden sector photons, lower bounds are set for the coupling constant $\chi =4.1\times {10}^{-9}$ at a mass of $m_{{\gamma }^{\prime }}=10.8\mu \mathrm{eV}$. For the latter we are probing a previously unexplored region in the parameter space.

Figure 1

Schematic of a microwave LSW experiment.

Figure 2

Inside dimensions of the cavity in [mm]. The coupling loop can be seen on top. The electric field configuration of two modes is shown on a transverse cutting plane. (a) ${\mathrm{TM}}_{010}$ mode for the ALP search, (b) ${\mathrm{TE}}_{011}$ mode for the HSP search.

Figure 3

Tuning range of several modes in the cavity, measured with a VNA in reflection. The depth of the tuning screw $d$ has been increased by 1 mm for each measurement. Black crosses indicate complementary simulation results.

Figure 4

Result of an $S_{11}$ measurement of the detecting cavity with a VNA, immediately before the ALPs run. The cavity parameters have been extracted by means of a curve fitting procedure, adapted from 21.

Figure 5

Noise power density $N_0$ at the coupling port of the detecting cavity, indicating its resonant frequency in relation to $f_{\mathrm{sys}}$. The measurement has been done immediately before and after the 25 h ALPs run in June 2013.

Figure 6

(a): Emitting cavity absorbed RF power (defined as $P_{\mathrm{EM}}=P_{\mathrm{inc}}-P_{\mathrm{refl}}$). (b): Physical temperature of the detecting cavity. (c): Noise power density from the detecting cavity. Data were recorded during the ALPs run in June 2013.

Figure 7

Photo of the emitting cavity (1) and shielding enclosure (2) containing the identical detecting cavity. For the ALP search, both parts were placed in the bore of a solenoid magnet with the same arrangement as shown in the picture.

Figure 8

Detailed block diagram of the experimental setup.

Figure 9

The components of the RF front end, unmounted from the cavity shielding box. From left to right: Analog optical link, low noise amplifier and bandpass filter.

Figure 10

Response of three spectral bins without zero padding. The amplitude of a signal falling between the bins can be attenuated by a factor of 0.64 ($-3.92\mathrm{dB}$).

Figure 11

Monte Carlo simulation of the fluctuation of one spectral bin governed by background noise (N) and another bin with an additional sinusoidal signal ($\mathrm{S}+\mathrm{N}$). The analytical PDFs show good agreement with the histogram of the Monte Carlo data. Furthermore, an exemplary detection threshold $P_{th}$ and the resulting error probabilities $P{R}_{\alpha }$ and $P{R}_{\beta }$ have been indicated. Note that the power levels and probabilities are arbitrary and do not correspond to a measurement run.

Figure 12

Resulting power spectrum from the ALPs run in June 2013. (a) Overview over the full recorded span. (b) Zoom on $f_{\mathrm{test}}$, where the test tone signal is clearly visible. (c) Zoom on $f_{\mathrm{sys}}$, where no WISP signal is visible. The green shaded areas mark the frequency range where a signal would be expected. Blue dots indicate the spectral bins calculated by a direct DFT operation; the grey lines show the underlying continuous spectrum, approximated by zero padding the time domain data. All three plots share the same Y axis.

Figure 13

CROWS: exclusion limits for ALPs for the measurement run in June 2013 in a 3 T magnet. Confidence level: 95%. Laser LSW: exclusion limits from ALPS-1 [14], the most sensitive optical LSW experiment to date. CAST: helioscope observing the Sun to search for solar ALPs [36]. More details on the other experiments can be found in 10.

Figure 14

CROWS: exclusion limits for HSPs for the measurement run in September 2013. Confidence level: 95%. UWA LSW [18] and ADMX LSW [37] are similar microwave-based LSW experiments. More details on the other experiments can be found in 10.