Duration and properties of the embedded phase of star formation in 37 nearby galaxies from PHANGS-JWST

TL;DR

The paper addresses how long star formation remains embedded in dust within GMCs across a diverse set of 37 nearby galaxies. It employs a statistically rigorous tuning-fork framework (Heisenberg) to convert spatial decorrelations among CO, 21 μm, and Hα tracers into absolute timescales, using t_{CO} as a metallicity-aware reference. The main findings show a median total 21 μm emission time of t_{21 μm} ≈ 3.9 Myr, an embedded-feedback phase t_{fb,21 μm} ≈ 3.4 Myr, and a short obscured phase t_{obscured} ≈ 0.8 Myr (≤4 Myr in all galaxies, with ≤1 Myr in 28/37), indicating rapid dust clearing and a dominant role for pre-supernova feedback. The embedded phase durations correlate with GMC properties (mass and velocity dispersion) and, marginally, with morphology and metallicity, implying environmental dependence of cloud dispersal and dust heating; diffuse 21 μm emission remains substantial (~62%), reflecting a mix of compact and diffuse dust heating. These results bolster a picture where embedded star formation is brief and heavily influenced by local GMC conditions and galactic environment, with JWST enabling the first large-sample statistical view and offering insights relevant for high-redshift GMC populations and feedback processes.

Abstract

Light reprocessed by dust grains emitting in the infrared allows the study of the physics at play in dusty, embedded regions, where ultraviolet and optical wavelengths are attenuated. Infrared telescopes such as JWST have made it possible to study the earliest feedback phases, when stars are shielded by cocoons of gas and dust. This phase is crucial for unravelling the effects of feedback from young stars, leading to their emergence and the dispersal of their host molecular clouds. Here we show that the transition from the embedded to the exposed phase of star formation is short (< 4 Myr) and sometimes almost absent (< 1 Myr), across a sample of 37 nearby star-forming galaxies, covering a wide range of morphologies from massive barred spirals to irregular dwarfs. The short duration of the dust-clearing timescales suggests a predominant role of pre-supernova feedback mechanisms in revealing newborn stars, confirming previous results on smaller samples and allowing, for the first time, a statistical analysis of their dependencies. We find that the timescales associated with mid-infrared emission at 21 μm, tracing a dust-embedded feedback phase, are controlled by a complex interplay between giant molecular cloud properties (masses and velocity dispersions) and galaxy morphology. We report relatively longer durations of the embedded phase of star formation in barred spiral galaxies, while this phase is significantly reduced in low-mass irregular dwarf galaxies. We discuss tentative trends with gas-phase metallicity, which may favor faster cloud dispersal at low metallicities.

Figures (17)

• Figure 1: Composite image of the 37 galaxies drawn from the PHANGS surveys, ordered following their Hubble morphological T-type from the LEDA database, from massive spirals to low-mass irregular galaxies. All the maps have been convolved to the resolution of the ALMA/CO(2-1) maps, corresponding to the minimal aperture size reported in Table \ref{['table_input_param']}. The white masks show regions that are excluded from our analysis (H$\alpha$ and CO bright peaks identified in Kim_2022_environmental_dep as well as artifacts and diffraction spikes in any of the three tracers). The green circles show the brightest peaks in 21 $\mu$m, which are masked in our analysis, as described in Section \ref{['subsect_data_homogenization']}. The number of masked peaks and corresponding surface brightness thresholds are reported in the Table \ref{['table_input_param']}.
• Figure 2: Measured deviation of the gas-to-stellar flux ratio with respect to the galactic average as a function of the aperture sizes for each galaxy obtained by contrasting CO emission as a gas tracer respectively with H$\alpha$ (black triangles) and 21 $\mu$m (blue circles) as an SFR tracer. The positive deviations correspond to measurements focusing on gas peaks (traced by CO), while the negative deviations were obtained by focusing on stellar peaks (traced respectively by H$\alpha$ or 21 $\mu$m). For each data point, we also show the effective 1$\sigma$ error, after the covariance between data points is taken into account. The horizontal plain line corresponding to a deviation of zero in log represents the galactic average. The dotted gray lines connecting the measurements correspond to a polynomial fit of the tuning-fork branches, following Kruijssen_tf_2018. The arrows indicate the typical separation length, $\lambda$, at which the two tracers decorrelate. The two last panels show the histograms of $\lambda$ and inferred feedback timescales ($t_{\rm fb}$, defined in Section \ref{['subsec_measuring_timescales']}), derived for the whole sample using either H$\alpha$ or 21 $\mu$m as a proxy for SFR, as well as the median and 1$\sigma$ standard deviations associated with these distributions.
• Figure 3: Schematic evolutionary timeline of a star-forming region from the molecular cloud assembly until it becomes invisible in any of the three tracers we consider. In this study, we focus on the embedded feedback phase, during which stars are embedded within their host GMC. This phase is composed of an exposed phase, during which CO, 21$\mu$m, and H$\alpha$ emit simultaneously, and of an obscured phase, during which H$\alpha$ emission is not detected. After the GMC dispersal, we measured an isolated phase of emission for both SFR tracers before they fade away. For most galaxies, 21$\mu$m fades faster than H$\alpha$ (upper arrow), but a few galaxies display a longer 21$\mu$m emission (bottom arrow).
• Figure 4: Typical evolutionary timescales in our sample evolving from an inert CO-emitting phase to an embedded star-forming phase traced with 21 $\mu$m and eventually to an exposed star-forming phase traced with H$\alpha$ emission. The feedback timescale $t_{\rm fb,~21~\mu m}$ is an upper limit in 25 out of 37 galaxies, which are flagged with $U$. The upper and lower errorbars associated with each phase of the evolutionary cycle are shown as thin lines, with the same color. Galaxies are ordered by their Hubble morphological type.
• Figure 5: Feedback timescale measured based on 21 $\mu$m emission versus feedback timescale measured based on H$\alpha$. The color bar shows the Hubble morphological type from the HyperLEDA database Paturel_hyperleda1_2003Paturel_hyperleda2_2003Makarov_HyperLEDA_2014. The dotted line shows the 1:1 relation. The obscured feedback phase traced by 21 $\mu$m ($t_{\rm 21~\mu m}$ - $t_{\rm H\alpha}$) is less than 4 Myr in all the galaxies from our sample, as shown by the plain black line, and less than 1 Myr for 28 out of 37 galaxies, as shown by the shaded area. We display the Spearman correlation coefficient as well as the log p-value of the data, excluding upper limits.