2022ApJ...929...23H

Using new large-area maps of the cold neutral medium (CNM) fraction, f_CNM, we investigate the relationship between the CNM, the abundance of polycyclic aromatic hydrocarbons (PAHs), and the anomalous microwave emission (AME). We first present our f_CNM map based on full-sky HI4PI data, using a convolutional neural network to convert the spectroscopic H I data to f_CNM. We demonstrate that f_CNM is strongly correlated with the fraction of dust in PAHs as estimated from mid- and far-infrared dust emission. In contrast, we find no correlation between f_CNM and the amount of AME per dust emission and no to weakly negative correlation between f_CNM and the AME peak frequency. These results suggest PAHs preferentially reside in cold, relatively dense gas, perhaps owing to enhanced destruction in more diffuse media. The lack of positive correlation between f_CNM and AME peak frequency is in tension with expectations from theoretical models positing different spectral energy distributions of AME in the cold versus warm neutral medium. We suggest that different PAH abundances and emission physics in different interstellar environments may explain the weaker-than-expected correlation between 12 µm PAH emission and AME even if PAHs are the AME carriers.