Searching for Axion-like particle Dark Matter with Time-domain Polarization: Constraints from a protoplanetary disk

Abstract

Axion-like particles (ALPs) can induce a birefringence effect that rotates the polarization angle of light, offering a probe of ultralight dark matter. We analyze archival near-infrared polarimetric data of the protoplanetary disk (PPD) around HD 163296. Whereas previous studies considered only single-epoch snapshots, we perform a consistent multi-epoch time-series analysis, extracting the polarization angle and its uncertainty from the polarized images. The resulting six-epoch time series is consistent with a constant polarization angle within the measurement uncertainties, while being sensitive to timescales of $\sim 170-400$ days. The typical polarization angle uncertainties are $1.6$--$6.4$ degrees, partly driven by multiple scattering in the optically thick disk, which broadens the intrinsic polarization angle distribution and introduces additional dispersion in the representative polarization angle. Based on these data, we derive the first upper limits on the ALP-photon coupling from PPD polarization variability, $g_{aγ} \lesssim 7.5 \times 10^{-12} (m_a / 10^{-22}\,{\rm eV})\,{\rm GeV}^{-1}$. Furthermore, we forecast that achieving a polarization angle uncertainty of $σ\sim 0.1$ degrees would enable world-leading sensitivity to ALP-induced birefringence.

Figures (5)

• Figure 1: Schematic illustration of the polarization geometry in a protoplanetary disk (PPD) and the expected effect of ALP-induced birefringence. The central panel shows the intrinsic azimuthal polarization pattern produced by scattered light on the disk surface. Polarization is analyzed using local Stokes parameters $Q_\phi$ and $U_\phi$ defined in the azimuthal frame; the left and right panels illustrate the corresponding polarization orientations. In the presence of ALP birefringence, the polarization plane is rotated by an angle $\chi$, which oscillates in time, producing a time-varying polarization angle as described by Equation \ref{['eq:theta_final']}.
• Figure 2: The mean value of the polarization angle of HD 163296 as a function of days since 2021 April 6. Error bars represent the $1\sigma$ statistical uncertainties derived from the Stokes $Q_\phi$ and $U_\phi$ maps.
• Figure 3: $95\%$ CL upper limit on the axion–photon coupling constant derived from the HD 163296 dataset. The accessible frequency range corresponds to oscillation periods between $\sim$100 and 400 days, or ALP masses $m_a \sim 10^{-22}$–$10^{-21}$ eV. The purple dash-dotted line and dashed line denote the optimistic and conservative upper bounds from PPD polarization reported by 2019PhRvL.122s1101F, respectively. The blue, green, red and light blue shaded regions are excluded by laboratory (CAST 2017NatPh..13..584C) and astrophysical observations (SN 1987A 2015JCAP...02..006P, quasar polarization 2012JCAP...07..041P and Pulsar Timing Array 2024arXiv241202229X). The black dashed lines indicate the projected sensitivities of forthcoming experiments (ALPS-II 2013JInst...8.9001B and IAXO 2014JInst...9.5002A), while the brown long-dashed and dotted lines represent tentative lower limits inferred from astrophysical observations (soft X-ray excess 2014JCAP...09..026A and $\gamma$-ray transparency 2013PhRvD..87c5027M).
• Figure 4: Required signal-to-noise ratio per measurement ($S/N$) as a function of the number of independent data points $N$ to achieve different target precisions in the polarization angle: $\sigma_\chi$ = $0.1^\circ$ (green), $0.01^\circ$ (red) and $0.05^\circ$ (blue). The shaded regions indicate the parameter space where the corresponding precision is attainable. The figure illustrates the trade-off between observational depth ($S/N$) and cadence ($N$), and shows that sub-degree to sub-millidegree polarization angle precision is feasible with a sufficiently large number of high-quality observations.
• Figure 5: Forecasted 95% confidence-level (CL) sensitivities on the axion--photon coupling $g_{a\gamma}$ achievable with long-term monitoring of protoplanetary disks. The orange and blue dotted lines correspond to expected limits for polarization angle precisions of 0.1$^\circ$ and 0.01$^\circ$, respectively. Other shaded regions and reference limits are the same as in Fig \ref{['fig:hd163296_limit']}.