The Calibration of Short Wavelength Polycyclic Aromatic Hydrocarbon Emission as Star Formation Rate Indicators with JWST

TL;DR

This paper establishes JWST-based calibrations linking short-wavelength PAH emission (3.3 and 7.7 μm) to star formation rate on the fundamental 40 pc scale around emerging young star clusters in four nearby galaxies, revealing a consistently sub-linear relation with α ≈ 0.8 and a strong metal dependence. The authors develop rigorous continuum subtraction, eYSC selection, and photometry methods, and they quantify how IMF sampling, PAH destruction, and aging influence the PAH–SFR relation, providing explicit calibrations for Σ_SFR as a function of ΣL_{3.3} and ΣL_{7.7} (and their in-band variants). They find substantial PAH deficits and increased scatter in metal-poor environments (e.g., NGC 4449), and they report that the 3.3/7.7 μm ratio rises at lower metallicity, plausibly due to a smaller PAH size distribution. A key implication is that a large fraction of PAH emission in typical local galaxies is excited by older stellar populations rather than recent star formation, complicating the use of PAHs as global SFR tracers but still enabling reliable, high-resolution SFR indicators with careful context, especially at high redshift where the 3.3 μm feature remains accessible with JWST. The work thus advances our understanding of PAH heating and turnover across metallicities and supports nuanced interpretations of PAH-based SFR metrics in the JWST era.

Abstract

We use JWST/NIRCam and MIRI imaging acquired by the Feedback in Emerging extrAgalactic Star clusTers (FEAST) program along with archival HST imaging to map ionized gas (Pa$α$, Br$α$, and H$α$) and Polycyclic Aromatic Hydrocarbon (PAH) emission (3.3 and 7.7 $μ$m) across a sample of four nearby galaxies (NGC 5194, 5236, 628, and 4449). These maps are utilized to calibrate the PAH features as star formation rate (SFR) indicators in 40 pc size regions around massive emerging young star clusters (eYSCs). We find a tight, sub-linear (power-law exponent, $α{\,}{\sim}{\,}0.8$) relation between the PAH luminosities (3.3 and 7.7 $μ$m) and SFR (extinction corrected Pa$α$) in near solar metallicity environments. PAH destruction in more intense ionizing environments and/or variations in the age of our sources may drive the deviation from a linear relation. In the metal-poor environment of NGC 4449 (${\sim}$1/3 Z$_{\odot}$), we see substantial deficits in the PAH feature strengths at fixed SFR and significantly higher scatter in the PAH-SFR relations. We determine that the 3.3/7.7 $μ$m PAH luminosity ratio increases towards lower metallicity environments. This is interpreted as a result of a shift in the size distribution towards smaller PAHs at lower metallicities, possibly due to inhibited grain growth. Focusing on the regions in NGC 4449, we observe a decreasing 3.3/7.7 $μ$m ratio towards higher SFR, which could indicate that small PAHs are preferentially destroyed relative to larger PAHs in significantly sub-solar metallicity conditions. We estimate that ${\sim}$2/3 of the PAH emission in typical local star-forming galaxies is excited by older stars and unrelated to recent ($<$10 Myr) star formation.

Figures (12)

• Figure 1: Spectrum of a 2 Myr old, 10$^{5}$ M$_{\odot}$ eYSC powering an HII region and PDR from the MAPPINGS III derived models of 2008ApJS..176..438G (black line). The JWST/NIRCam and MIRI filter throughputs (colored curves) are overlaid on the model to outline the spectral regions sampled by our data. The F187N and F405N target the Pa$\alpha$ and Br$\alpha$ hydrogen recombination lines, respectively. The F335M and F770W target the 3.3 and 7.7 $\mu$m PAH emission features, respectively. All other filters target continuum emission. Note that the MIRI/F1000W filter is only available in one of our targets, NGC 628.
• Figure 2: Top panel: A three-color composite image showing the full Br$\alpha$ (red channel), 3.3 $\mu$m PAH (green channel), and Pa$\alpha$ (blue channel) emission maps (at the F444W resolution) for NGC 5194 (M51). Overlaid on top of the image are the 20 pc ($\sim$0.55$"$) radius apertures (red circles) used to measure the photometry of the regions around each identified eYSC--I source (cospatial Pa$\alpha$, Br$\alpha$, and 3.3 $\mu$m PAH emission peaks). The white bar shows the 500 pc scale. Bottom panel: A zoom-in corresponding to the star-forming spiral arm region shown by the white box in the top panel.
• Figure 3: Histograms showing the distribution of the nebular oxygen abundance or 12+log(O/H) (top left), the color excess or E(B$-$V) (top right), the equivalent width (EW) of Pa$\alpha$ (bottom left), and the extinction corrected Pa$\alpha$ luminosity surface density (bottom right) for our measurements around eYSC--I sources in each of the four galaxy targets, shown by the various colors. The bins are matched across the targets. The oxygen abundance is derived from the observed radial gradient in combination with the galactocentric radius of each source. E(B$-$V) is derived from the observed H$\alpha$/Pa$\alpha$ luminosity ratio. The vertical colored lines show the median values for each galaxy. The lower numbers of sources in the bottom left panel result from the fact that the Pa$\alpha$ EW cannot be reliably estimated for all of our sources.
• Figure 4: Left panel: The extinction corrected 3.3 $\mu$m PAH luminosity surface density as a function of SFR surface density (corrected Pa$\alpha$). The data points show the measurements around eYSC--I sources for each galaxy and are color-coded by the nebular oxygen abundance (gas-phase metallicity), derived from the observed radial gradients. The various black lines show the best-fit relations for each galaxy determined from a Bayesian linear regression using the LINMIX package, while the shaded regions show the 1$\sigma$ confidence intervals. The figure caption gives the total number of sources (N), the Spearman correlation coefficient ($\rho$), and the values of the best-fit slope ($\alpha$) and y-intercept ($b$) and their 1$\sigma$ uncertainties determined from the Bayesian regression for each galaxy. Right panel: Same as the left, but instead for the 7.7 $\mu$m PAH luminosity surface density.
• Figure 5: The extinction corrected 3.3 $\mu$m PAH (left panels) or 7.7 $\mu$m PAH (right panels) luminosity surface density as a function of SFR surface density with an additional luminosity cut. The cut has been applied to limit the effects from stochastic sampling of the stellar IMF. Regions with H$\alpha$ luminosity below that expected from a 4 Myr old cluster with a stellar mass of 3000 M$_{\odot}$ are removed, based on Starburst99 models with Z=0.02 and the Padova AGB evolutionary tracks. The best-fit relations (black lines) are shown either for each galaxy separately (top panels) or in 2 bins (high/low) of metallicity (bottom panels). The three high metallicity spirals (NGC 5194, 5236, and 628) constitute the high bin, while the lower metallicity dwarf (NGC 4449) constitutes the low bin. Note that the luminosity cutoff is the same for all galaxies, however, the high inclination angle (I=68$^{\circ}$) of NGC 4449 makes it appear lower in the space of inclination corrected surface density. See Figure \ref{['fig:f4']} for a more complete description.