No evidence of polarization in the $11.3\,μ$m PAH emission line by independent analyses

Abstract

Polycyclic aromatic hydrocarbons (PAHs) are commonly used as proxies for star formation, molecular gas content, and other interstellar medium (ISM) properties in our Galaxy and other galaxies. Given their abundance and brightness, polarization measurements of PAH features could, in principle, provide a probe of the ISM magnetic field and intrinsic PAH properties; however, the diagnostic power of PAH polarization remains to be established. Previous studies reported that the $11.3\,μ$m PAH emission line in the northwestern nebula of the Herbig Be star MWC 1080 was polarized at $1.9\pm0.2$%. This level of polarization was explained via the paramagnetic relaxation process, which may allow the characterization of magnetic fields in the ISM. Using the same observations, here, we re-analyzed the $8-13\,μ$m spectro-polarimetric observations taken with CanariCam on the 10.4-m Gran Telescopio CANARIAS (GTC), and we measure a polarization of $0.5\pm0.6$% within $11.3\pm0.2\,μ$m, consistent with an unpolarized source, $0.6\pm0.2$% (instrumental polarization). We reproduce the previously reported polarized PAH emission line if the polarization fraction spectrum is oversubtracted by a constant instrumental polarization and the polarization uncertainties, which is inconsistent with the fundamentals of polarimetric data analysis. Thus, the published $8-13\,μ$m spectro-polarimetric data taken with CanariCam/GTC provide no statistical evidence for a polarized $11.3\,μ$m PAH emission line, in agreement with current dust models.

Figures (5)

• Figure 1: Calibration. The polarization efficiency (top), instrumental polarization (second row), and polarization (third row) and PA (bottom) of the polarized source, AFGL 2591, at the original resolution (black), $6$ spectral pixels ($0.12$$\mu$m) binned (blue) are shown. For comparison, the polarization measurements using CanariCam narrow-band imaging filters (red circles) and the spectro-polarimetric observations (green circles) from Smith2000 are displayed.
• Figure 2: Spectro-polarimetric observations of MWC 1080 NW. The Stokes I (top-left), $q$ (top-right), and $u$ (bottom-right), the polarization fraction (bottom-left) spectra (black line), and their $1\sigma$ uncertainties (gray region) are shown. The instrumental polarization (red line; as in Fig. \ref{['fig:fig1']}) is displayed in the same panel as the polarization fraction spectrum. The Stokes I has been smoothed using a $0.06\,\mu$m boxcar, while the Stokes $q$ and $u$ are downsampled to $0.12\,\mu$m and then smoothed with a $0.36\,\mu$m boxcar. The $11.3\,\mu$m PAH emission (blue solid line), the aperture width of $0.4\,\mu$m (blue region), and the approximated extent of the telluric $0_{3}$ band (yellow region) are shown. The values of the Stokes $q$ and $u$ and the polarization fraction and their uncertainties, within the $11.3\pm0.2\,\mu$m wavelength range (blue region), are displayed.
• Figure 3: Continuum subtracted spectro-polarimetic observations of MWC 1080 NW. Same display as Fig. \ref{['fig:fig2']} but for the continuum-subtracted spectro-polarimetric observations.
• Figure 4: Comparison with and reproduction of Zhang2017 results. Our Stokes I, $q$, and $u$ (black lines) are the same as in Fig. \ref{['fig:fig1']}. Our polarized spectrum, without debias ($P_{\rm{b}}$), were subtracted by a constant instrumental polarization of $0.6$%, and the uncertainty of the polarization fraction, $\sigma_{p}$, such as $P=P_{\rm{b}} - P_{\rm{inst}} - \sigma_{\rm{p}}$ (black solid line). The absolute values of this polarization spectra are display as solid dashed lines. The digitalized spectra of Zhang2017 are shown as green solid lines. The polarization fraction spectra estimated using the Stokes $q$ and $u$ is displayed as dashed green lines.
• Figure 5: Continuum subtraction of the o-rays and e-rays spectra. The o-rays and e-rays spectra, the continuum fitting lines (dashed line), and the continuum-subtracted spectra are shown.