High star cluster formation efficiency in the strongly lensed Sunburst Lyman-continuum galaxy at z=2.37

Abstract

We investigate the strongly lensed (µ ×10 − 100) Lyman continuum (LyC) galaxy, dubbed Sunburst, at z = 2.37, taking advantage of a new accurate model of the lens. A characterization of the intrinsic (delensed) properties of the system yields a size of 3 sq.kpc, a luminosity M UV = −20.3, and a stellar mass M 10 9 M ; 16% of the ultraviolet light is located in a 3 Myr old gravitationally-bound young massive star cluster (YMC) with an effective radius of ∼ 8 pc (corresponding to 1 milli-arcsec without lensing) and a dynamical mass of ∼ 10 7 M (similar to the stellar mass), from which LyC radiation is detected (λ < 912Å). The star formation rate and stellar mass surface densities for the YMC are Log 10 (Σ SFR [M yr −1 kpc −2 ]) 3.7 and Log 10 (Σ M [M pc −2 ]) 4.1, with sSFR > 330 Gyr −1 , consistent with the values observed in local young massive star clusters. The inferred outflowing gas velocity (> 300 km s −1 ) exceeds the escape velocity of the cluster. The resulting relative escape fraction of the ionizing radiation emerging from the entire galaxy is higher than 6-12%, whilst it is 46 − 93% if inferred from the YMC multiple line of sights. At least 12 additional unresolved starforming knots with radii spanning the interval 3 − 20 pc (the majority of them likely gravitationally-bound star clusters) are identified in the galaxy. A significant fraction (40-60%) of the ultraviolet light of the entire galaxy is located in such bound star clusters. Adopting a formation timescale of the star clusters of 20 Myr we derive a cluster formation efficiency Γ 30%. The star formation rate surface density of the Sunburst galaxy (Log 10 (Σ SFR ) = 0.5 +3 −0.2 ) is consistent with the high inferred Γ, as observed in local galaxies experiencing extreme gas physical conditions. Overall, the presence of a bursty event (i.e., the 3 Myr old YMC with large sSFR) significantly influences the morphology (nucleation), photometry (photometric jumps) and spectroscopic output (nebular emission) of the entire galaxy. Without lensing magnification, the YMC would be associated to an unresolved 0.5 kpc−size star-forming clump. The de-lensed LyC and UV magnitude m 1600 (at 1600Å) of the YMC are 30.6 and 26.9, whilst the entire galaxy has m 1600 24.8. The Sunburst galaxy shows a relatively large rest-frame equivalent width of EW rest (Hβ+ [O iii]λλ4959, 5007) 450Å with the YMC contributing to ∼ 30% (having a local EW rest 1100Å) and ∼ 1% of the total stellar mass. If O-type (ionizing) stars are mainly forged in star clusters, then such engines were the key ionizing agents during reionization and the increasing occurrence of high equivalent width lines (Hβ+[Oiii]) observed at z > 6.5 might be an indirect signature of a high frequency of forming massive star clusters (or high Γ) at reionization. Future facilities, which will perform at few tens milli-arcsec resolution (e.g., VLT/MAVIS or ELT), will probe bound clusters on moderately magnified (µ < 5 − 10) galaxies across cosmic epochs up to reionization.

Exploring the physical properties of lensed star-forming clumps at $2\lesssim z \lesssim6$

“…Three of those complexes show signatures of Wolf-Rayet stars, implying young ages of ∼ 5 Myr (Bastian et al 2006). Similar properties are observed in the Sunburst 5.1 knot presented in detail in Vanzella et al (2021a) and in some of our clumps shown in Figure D.1 of Vanzella et al (2021b). Other scenarios, such as the fragmentation of gas-rich discs, are potentially responsible for the observed high SFR among high redshift clumpy structures (Noguchi 1999;Dekel et al 2009).…”

“…Indeed, recent observations of strongly lensed systems at 𝑧 ∼ 1 − 3 reveal the existence of clumps with sizes of ∼ 100 pc (Livermore et al 2012(Livermore et al , 2015Cava et al 2018), down to ∼ 10 pc (Rigby et al 2017) in some extremely magnified cases. Our previous studies (Vanzella et al 2017b(Vanzella et al , 2019(Vanzella et al , 2020(Vanzella et al , 2021aCalura et al 2021) have further extended this work to spectroscopically confirmed compact stellar structures, with sizes down to 10 pc, which are interpreted as young massive clusters and/or globular cluster precursors at 𝑧 ∼ 3 − 6 (see also Bouwens et al 2021). This demonstrates the importance of gravitational lensing to explore compact stellar structures considered as building blocks of high-redshift starforming galaxies.…”

“…We consider samples where SFRs was estimated through SED fitting (as for our sample) or from H𝛼, using the prescription by Kennicutt (1988). Our sample and the Sunburst clumps (Vanzella et al 2021a), despite being at 𝑧 ∼ 2 − 6, have sizes comparable to local HII regions and GCs (Jones et al 2010;Swinbank et al 2009;Livermore et al 2012;Johnson et al 2017), but they have SFRs ∼ 300 times higher. They also have SFRs ∼ 100 times higher than clumps from Wuyts et al (2014) and Livermore et al (2015), which are part of a sample of lensed systems at 𝑧 > 2.…”

“…In Figure 9, we show the location of our sample of individual clumps in the SFR-M ★ plane and compare it with the other published samples. Thanks to lensing amplification, our sample probes the low-SFR and low-M ★ region, down to M ★ ∼ 10 5 M , where only a few other lensed sources from the literature are located (e.g., the Sunburst clumps at 𝑧 = 2.37, Vanzella et al 2021a).…”

Stellar populations and star formation histories of the most extreme [OIII] emitters at $z=1.3-3.7$

“…When EELGs are in the current burst phase, the massive stars are likely being formed in very young star clusters as demonstrated by spatially-resolved observations of a few strongly lensed galaxies at high redshift (e.g. Vanzella et al 2019Vanzella et al , 2021.…”