Probing the ISM of Heiiλ1640 emitters at z = 2–4 via MUSE

Abstract Heiiλ1640 emission in the absence of other metal lines is the most sought-after emission line to detect and characterize metal free stellar populations. However, even recent stellar population models with sophisticated treatment of stellar evolution also lack sufficient He+ ionising photons to reproduce observed He 0.1em ii fluxes. We use VLT/MUSE GTO observations to compile a catalogue of 15 z ∼ 2–4 He ii λ1640 emitters from ∼10–30 hour pointings. We show that both He ii λ1640 detections and non-detections occupy similar distribution in UV absolute magnitudes. Rest-UV emission line analysis of our sample shows that the emission lines of our He ii λ1640 emitters are driven by star-formation in solar to moderately sub-solar (∼1/20th) metallicity conditions. However, we find that even after considering effects from binary stars, we are unable to reproduce the He ii λ1640 equivalent widths. Alternative mechanisms are necessary to compensate for the missing He+ ionising photons.


Introduction
Several works have tried to obtain observational signatures for the existence of metalfree (pop-III) stars in the early Universe via current ground and space based telescopes (e.g., Cassata et al. 2013;Sobral et al. 2015). The failure to obtain observational evidence for these first generation of stars has been attributed to reasons such as the short lifetime (∼ 1Myr) of pop-III systems, photometric calibration, presence of active galactic nuclei (AGN), pristine cold mode gas accretion to galaxies, and limited understanding of high-redshift stellar populations and the inter-stellar-medium (ISM). All these effects have contributed in varying degrees to the complexity of detecting and identifying pop-III host systems (Fardal et al. 2001;Yang et al. 2006;Sobral et al. 2015;Agarwal et al. 2016;Bowler et al. 2017;Matthee et al. 2017;Shibuya et al. 2017;Sobral et al. 2018).
With large samples of high-z galaxies, candidates for galaxies containing a significant population of pop-III stars can be selected due to the presence of strong Lyα and He ii emission lines in the absence of other prominent emission features. These lines can be interpreted as existence of pristine metal poor stellar populations (Tumlinson et al. 2003;Raiter et al. 2010;Inoue et al. 2011;Sobral et al. 2015). This interpretation is however challenging in the face of other processes that can produce He + ionising photons (E> 54.4 eV, λ < 228Å). Therefore, to make compelling constraints of stellar populations in the presence of strong He ii emission and link with pop-III hosts, a comprehensive understanding of He ii emission mechanisms is required. 235 236 T. Nanayakkara, J. Brinchmann & The MUSE Collaboration Stellar populations in a variety of ages and physical/chemical conditions undergo various mechanisms that may contribute to He ii emission, such as young O/B type stars (e.g., Shirazi & Brinchmann 2012), hydrogen-stripped massive evolved Wolf-Rayet stars (e.g., Shirazi & Brinchmann 2012), very massive low-Z WNh stars (hydrogen rich WN stars; Gräfener & Vink 2015), post-asymptomatic giant branch stars (e.g., Binette et al. 1994), X-ray binary stars (e.g., Casares et al. 2017), radiative shocks (e.g., Izotov et al. 2012), and AGN (e.g., Shirazi & Brinchmann 2012). Binary interactions and stellar rotation prolong the lifetime of young O/B stars extending the total amount of He + photons present at a given star-formation history (e.g., Eldridge et al. 2017;Götberg et al. 2017). Even with a variety of such mechanisms, we still lack E> 54.4 eV photons in stellar population models to produce observed He iiλ1640 line profiles consistently with other rest-UV emission lines (e.g., Shirazi & Brinchmann 2012;Senchyna et al. 2017;Berg et al. 2018).

Data & Analysis
The new generation of sensitive multiplexed optical instruments in 8-10m class telescopes such as the The Multi Unit Spectroscopic Explorer (MUSE; Bacon et al. 2010) has enabled astronomers to obtain spatially-resolved spectroscopy of galaxies throughout cosmic time in unprecedented numbers (e.g., Inami et al. 2017). Here, we present a sample of 15 He iiλ1640 detections (including 3 AGN) obtained from deep ∼ 10 − 30 hour pointings as a part of multiple MUSE guaranteed time observation programs (Bacon et al. 2015Epinat et al. 2018;Marino et al. 2018) between z = 1.93 − 4.67. The Universe at z ∼ 2 − 4 was reaching the peak of the cosmic star-formation rate density (Madau & Dickinson 2014), with galaxies in a diverse range of physical and chemical properties (e.g., Kacprzak et al. 2016;Kewley et al. 2016;Steidel et al. 2016;Nanayakkara et al. 2016Nanayakkara et al. , 2017Strom et al. 2017). With MUSE we are able to obtain rest-UV spectroscopy of young, low-metallicity, highly star-forming systems which may give rise to a diverse range of exotic phenomena capable of producing high-energy ionizing photons.
We remove AGN from our sample and use multiple emission line diagnostics from Gutkin et al. (2016) and Xiao et al. (2018) to explore the ISM conditions of the He iiλ1640 emission in the non-active galaxy sample. Additional details on sample selection are given in Nanayakkara et al., (submitted). In Figure 1 (left panel), we show the C iii]/He iiλ1640 vs O iii]/He iiλ1640 line ratio diagrams for single-star stellar population models from Gutkin et al. (2016). Our values agree with literature data of high-z sources (Patrício et al. 2016;Berg et al. 2018) and have considerably lower metallicities compared to z = 0 sources from Senchyna et al. (2017). In this line ratio diagram, our galaxies occupy a region, that can be described by star-forming galaxies with solar to ∼ 1/20th solar metallicities. We note that, when effects of binary stellar evolution are considered, the line-ratio diagnostics become more degenerate (also see Xiao et al. 2018). However, the line-ratios of our He iiλ1640 emitters fall within the range powered by star-formation.
The main discrepancy between model and data arise only once line EWs are compared. As shown in the right panel of Figure 1, Xiao et al. (2018) binary models are able to reproduce the observed Ciii] EWs but lacks sufficient He + ionising photons to reproduce the observed He iiλ1640 EWs, which is expected to be driven by the lack of photons below λ < 228Å in BPASS models (e.g., Berg et al. 2018). By matching observed Ciii] luminosities to model C iii] luminosities, we find that only extreme sub-solar metallicities (∼ 1/200th) are able to accurately predict the observed He + ionising photons in BPASS models, which is strongly in contrast with predictions from line-ratio diagnostics.

Conclusions & Future directions
We used deep optical spectroscopy from MUSE to obtain a sample of He iiλ1640 detections at z ∼ 2 − 4 to study their ISM conditions using state-of-the-art stellar-population/ photo-ionization models. Emission line ratios of our He iiλ1640 emitters could mostly be explained by Z∼ 0.05 − 1.0 Z photo-ionisation models, but, even the BPASS binary models lack sufficient ionising photons to re-produce observed He iiλ1640 EWs. In order to reproduce the observed He iiλ1640 luminosities, BPASS stellar population models with extreme sub-solar metallicities (∼ 1/200th) are required. Such low metallicities are in contradiction with our line-ratio diagnostics and stellar populations models can suffer large uncertainties due to lack of empirical calibrations in this regime. Extra contribution to the number of ionizing photons from X-Ray binaries, sub-dominant AGN, or effects related to stellar rotations at high metallicities might be necessary to alleviate the tension between models and observations. In addition, top-heavy initial-mass-functions in star-forming galaxies (e.g., see Nanayakkara et al. 2017), will contribute to higher levels of ionising photons, which could increase the He + ionising photon budget.
Future deep surveys such as the MUSE extreme deep field survey (PI R. Bacon), a single 160 hour pointing by MUSE, along with several deep pointings in Hubble frontier field parallels (PI L. Wisotzki) will provide extremely high signal-to-noise rest-UV spectra at z = 2 − 4 to perform spectro-photometric analysis by simultaneous combination of nebular emission features with weaker ISM and photospheric emission and absorption features. By constraining the stellar population properties to finer detail within this epoch, we will be able to make accurate predictions for future surveys in the era of James Webb Space Telescope. Given that individual detections of pop-III stars will be unlikely until proposed future space telescopes such as LUVOIR, we should push the current instruments to their maximum potential to constrain the stellar population properties of galaxies leading to the buildup of the peak of the cosmic star-formation rate density.