Constraints on the polarization angle oscillations of the Crab Nebula with the Simons Array and its applications to the search for axion-like particles

TL;DR

The paper tests the stability of Tau A’s polarization angle at 90 GHz with the Simons Array (PB-2a), extending PB24’s time-resolved approach to search for polarization oscillations and their possible ALP origin. By employing detailed TOD demodulation, per-detector calibration, and null tests, it finds no globally significant oscillation and delivers a median 95% upper bound on the oscillation amplitude of $A<0.12^{\circ}$ over $f$ in the range $3.39\ { m yr^{-1}}$ to $1.50\ { m day^{-1}}$. The absence of a signal is translated into competitive bounds on the ALP-photon coupling $g_{a\gamma\gamma}$ in the mass window $4.4\times10^{-22}$ to $7.2\times10^{-20}$ eV, assuming Tau A’s polarization rotation is exclusively due to an ALP field. The results also provide consistency checks against the PB24 hints and lay groundwork for stronger limits with upcoming 150 GHz data and extended observing seasons. Overall, the work strengthens Tau A as a polarization calibrator and tightens ALP constraints from astrophysical polarization measurements.

Abstract

We present a search for polarization oscillation of the Crab Nebula, also known as Tau A, at millimeter wavelengths using observations with the Simons Array, the successor experiment to POLARBEAR. We follow up on previous work by POLARBEAR using 90 GHz band data of the 2023 observing season of the Simons Array to evaluate the variability of Tau A's polarization angle. Tau A is widely used as a polarization angle calibration source in millimeter-wave astronomy, and thus it is necessary to validate the stability. Additionally, an interesting application of the time-resolved polarimetry of Tau A is to search for axion-like particles (ALPs). We do not detect a global signal across the frequencies considered in this analysis and place a median 95% upper bound of polarization oscillation amplitude $A<0.12^{\circ}$ over oscillation frequencies from 3.39 year$^{-1}$ to 1.50 day$^{-1}$. This constrains the ALP-photon coupling at a median 95% upper bound of $g_{aγγ}< 3.84\times 10^{-12}\times\left(m_a/10^{-21}\,\mathrm{eV}\right)$ in the mass range from $4.4\times10^{-22}$ to $7.2\times10^{-20}$ eV, assuming the ALP constitutes all of dark matter, its field is a stochastic Gaussian field, and it is the sole source of Tau A's polarization angle oscillation. Additionally, we do not detect signal at the frequencies where 2.5$σ$ hints were previously reported by POLARBEAR, but we do not exclude these signals at the 95% confidence level.

Figures (8)

• Figure 1: Cutaway view of the PB-2a receiver cryostat, with key elements labeled.
• Figure 2: Season co-add polarization intensity map of Tau A. The white lines represent the orientation of the polarization angle in each pixel. The red dashed circle covers the integration area used to estimate the polarization angle for individual observations and to evaluate systematic errors. The FWHM of the PB-2a beam is shown for comparison.
• Figure 3: Transfer function $F_f$ (bottom) and loss $1-F_f$ (top) for the frequency bins, described by Eq. \ref{['eq:frequencies']}. The transfer function is averaged over all the phases of oscillations.
• Figure 4: Top left: the measured polarization angles of Tau A during the 2023 observing season. Top right: The histogram of the measured polarization angles scaled by their assumed uncertainties. Bottom: Amplitude of the LSSA for frequencies less than 0.5 days$^{-1}$ as well as the approximate amplitudes of various levels of local and global significance. The two PB24 frequencies are called out with vertical lines.
• Figure 5: Observed polarization oscillation amplitudes and 95% upper limits. The MLEs for the amplitude at the two PB24 frequencies are 0.04$^\circ$ and 0.08$^\circ$, respectively. The $95\%$ upper limits at the two PB24 frequencies are 0.13$^\circ$ and 0.15$^\circ$, respectively.