Periodic Reporting for period 2 - NEFERTITI (NEar FiEld cosmology: Re-Tracing Invisible TImes)
Okres sprawozdawczy: 2020-11-01 do 2022-04-30
The first stars profoundly affected the unfolding of our Universe. Their mass distribution is unknown, but it controls the injection of energy, momentum and newly created heavy elements (metals) into the gas, affecting subsequent star-formation and the build-up of the first galaxies. Despite extraordinary progress in theoretical modelling and observational techniques, very little is known about the properties of the first stars and of the first galaxies hosting them.
A direct exploration of their formation time is a tremendous challenge. This chapter of the cosmic history was written between 13 and 13.5 Gyr ago, an epoch that the James Webb Space Telescope (JWST) will explore. But even its superb sensitive eyes will be blind to faint dwarf galaxies: the nurseries of the first stars, and the building-blocks of present-day galaxies.
In the Local Group, ultra-faint dwarf galaxies (UFDs) represent the most common galaxy population. These small systems comprise stars that are over 13 Gyr old and that likely built-up the Galactic stellar halo. Such early Universe survivors can be observed individually to retrace the chemical evolution and star-formation history of the gas during those “invisible” times.
First stars of different masses produce unique nucleosynthetic products, which are dispersed into the gas and preserved in the photosphere of ancient second-generation stars. The chemical abundance pattern of second-generation stars can be measured with high-resolution spectroscopy and interpreted with models to uncover the properties of the first stars. Second-generation stars can be found in UFDs and in the Galactic halo but they are extremely rare. Yet, in our current epoch of wide and deep Local surveys the number of identified ultra-faint dwarf galaxies will constantly increase and that of Galactic halo stars will rise by orders of magnitude.
The project NEFERTITI aims at fully exploiting this unprecedented data flow to catch the local stellar fossils and figure out the properties of the first stars and galaxies. In particular, it is expected to make a major step-forward in our understanding of the properties of the first stars and galaxies by constraining the mass distribution of the first stars and uncovering the physical processes driving the build-up of the first star-forming systems, i.e. mini-haloes.
To this end, the NEFERTITI project adopts a novel approach that integrates theoretical and observational research in a unique way. This is reflected into the composition of the NEFERTITI Team, which is composed by a mixture of theoreticians and observers: the PI (Salvadori: theory), the long-term researcher (RTD-A, Skuladottir: obs), two post-docs (Koutsouridou: theory; Aguado: obs), and three PhD students (Rossi, Gelli, Vanni) working on models/simulations to interpret observations and make predictions.
A direct exploration of their formation time is a tremendous challenge. This chapter of the cosmic history was written between 13 and 13.5 Gyr ago, an epoch that the James Webb Space Telescope (JWST) will explore. But even its superb sensitive eyes will be blind to faint dwarf galaxies: the nurseries of the first stars, and the building-blocks of present-day galaxies.
In the Local Group, ultra-faint dwarf galaxies (UFDs) represent the most common galaxy population. These small systems comprise stars that are over 13 Gyr old and that likely built-up the Galactic stellar halo. Such early Universe survivors can be observed individually to retrace the chemical evolution and star-formation history of the gas during those “invisible” times.
First stars of different masses produce unique nucleosynthetic products, which are dispersed into the gas and preserved in the photosphere of ancient second-generation stars. The chemical abundance pattern of second-generation stars can be measured with high-resolution spectroscopy and interpreted with models to uncover the properties of the first stars. Second-generation stars can be found in UFDs and in the Galactic halo but they are extremely rare. Yet, in our current epoch of wide and deep Local surveys the number of identified ultra-faint dwarf galaxies will constantly increase and that of Galactic halo stars will rise by orders of magnitude.
The project NEFERTITI aims at fully exploiting this unprecedented data flow to catch the local stellar fossils and figure out the properties of the first stars and galaxies. In particular, it is expected to make a major step-forward in our understanding of the properties of the first stars and galaxies by constraining the mass distribution of the first stars and uncovering the physical processes driving the build-up of the first star-forming systems, i.e. mini-haloes.
To this end, the NEFERTITI project adopts a novel approach that integrates theoretical and observational research in a unique way. This is reflected into the composition of the NEFERTITI Team, which is composed by a mixture of theoreticians and observers: the PI (Salvadori: theory), the long-term researcher (RTD-A, Skuladottir: obs), two post-docs (Koutsouridou: theory; Aguado: obs), and three PhD students (Rossi, Gelli, Vanni) working on models/simulations to interpret observations and make predictions.
The existence of very massive metal-free stars, M* = (140-260) M, has never been proved. Yet, these elusive stars should exist according to state-of-the-art numerical simulation, and they are expected to be the key sources of early metal-enrichment, ionizing photons, and primordial stellar black holes. Probing the existence of these stars is thus fundamental not only for Cosmology but also for galaxy formation. A possible way to make this major step-forward is by catching the chemical signatures that these elusive stars left in their descendants, i.e. in long-lived low-mass stars formed out of their ``ashes”. Indeed, very massive first stars are predicted to end their life as energetic Pair Instability Supernovae (PISN), which spread out in the surrounding Interstellar Medium (ISM) their peculiar chemical products.
By investigating the free parameter space of the problem, we studied how does the peculiar chemical abundance pattern of an ISM polluted by massive first stars vary when subsequent generations of “normal” stars contaminate it. In fact, normal (Pop II) stars are predicted to form early on after the first stellar generations and thus they can washed-out their key chemical signatures. But when and how this happens? We found that, independent of the choice of the free parameters, an ISM imprinted by the heavy elements from massive first stars at a >50% level it is most likely deficient of some key chemical species: Copper and Zinc. Further, these stars predominantly have [Fe/H] ~ 2.
Looking for the absence of these “killing elements” in large stellar samples might then be an effective method to identify the direct descendants of massive first stars. By exploiting literature data, we found once such a Zinc-poor star at [Fe/H] ~ −2, i.e. BD +80 245. Our analysis and its comparison with the developed model show that this star can be the smoking gun of a PISN explosion (Salvadori et al. 2019).
To probe and constrain the unknown mass distribution of the first stars we need to catch many massive first star descendants. The new strategy developed in SS et al. (2019) paves the way to find more of these rare objects by looking for stars with [(Zn, Cu)/Fe] < 0. Our predictions show that the descendants of massive first stars should be predominantly found in the Galactic bulge, or in relatively massive ancient dwarf galaxy, such as Sculptor and Fornax, which can retain the chemical products of such energetic explosions. The ESO/4MOST large observational program 4DWARFS (512 000 fibre hours) has been recently awarded (Dec. 2021) to our Team (PI: Skuladottir, 33 co-Is including SS, Gelli, Rossi) and it will allow us to search for these objects. In the upcoming years we will then be able to finally constrain the high-mass end of the mass distribution of the first stars. But what about the low-mass end?
Sophisticated but time limited 3D numerical simulations found that fragmentation process is possible in the primordial ISM, leading to the possible formation of low-mass, M* < 0.8 M, zero-metallicity stars. If these stars really existed then they should be found today, since the lifetime of sub-Solar mass stars is longer than the age of the Universe. We can thus exploit the persisting non-detection of zero-metallicity stars to put limits on the minimum mass of the first stars (Mmin). Unfortunately, the stellar sample required to be observed in the Galactic halo or Bulge order to limit Mmin at > 68% confidence level is extremely huge.
In Rossi, SS, Skuladottir (2021) we probed that UFDs are key systems to limit the mass distribution of the first stars at the lowest mass-end. We develop a novel model to follow the formation and evolution of the UFD galaxy Bootes I and we simulated its Color Magnitude Diagram (CDM) as observed today. By performing a statistical comparison between the number of long-lived zero metallicity stars predicted for different first-star mass distribution and those observed so far, we have been able to put tight constraints on Mmin and the characteristic (Mch) mass of the first stars. We found Mmin > 0.8 Msun and Mch > 1 Msun at 99% of the confidence level. Furthermore, we demonstrated that by exploiting next generation telescopes and instruments, such as MOSAIC on ELT we will be able to exclude at the 68% confidence level the existence of low-mass primordial stars, i.e. Mmin > 0.8 M independent on the first-star mass distribution. This result is thus extremely important. But how can be sure that UFD are low-mass mini-halos from the early Universe?
In Gallart et al. (including SS) 2021 we studied the star formation history (SFH) of UFD galaxy Eridanus II by exploiting deep CMD diagrams obtained with new HST observations. Our work demonstrates that the stellar population of Eridanus II is > 13 Gyr old, and that it was formed during a very short period, lasting < 500 Myr, despite a low star-formation rate ~ 0,001 Msun/yr. This result strongly supports the idea that UFDs are associated to low-mass mini-halos.
In conclusion, our work so far has found two key results to constrain the mass distribution of the first stars: we can exploit low-mass UFDs to infer limits on the minimum and characteristic mass, while massive dwarf galaxies can allow us to set its maximum mass and shape. But is the mass distribution of the first stars the only key physical quantity to be investigated?
In Skuladottir, SS, Amarsi et al. (2021) we presented the discovery of AS0039, a chemically peculiar star in the Sculptor dwarf spheroidal galaxy. AS0039 is the most iron-poor star, which has been observe outside of the Milky Way (MW). Contrary to most MW stars at this iron abundance, AS0039 is not enhanced in carbon and it lacks α-element uniformity, in stark contrast with the near solar ratios observed in C-normal stars within the Milky Way halo. Its unique abundance pattern indicates that AS0039 formed out of material that was predominantly enriched by a ~20 M zero-metallicity progenitor star with an unusually high explosion energy E = 10 × 1051 erg. AS0039 represents some of the first observational evidence for zero-metallicity hypernovae. This result is thus fundamental, since it opens a new important window that should be explored studying the properties of the first stars: the energy distribution function.
Finally, with our Team we have been working on the prospects of observing dwarf galaxy satellites with JWST, which has been successfully launched on the 25th of December 2021. We found that planned deep surveys of JWST will be able to catch "for free" the dwarf galaxy satelllites of high-z Lyman Break galaxies (Gelli, SS et al. ApJL 2021).
By investigating the free parameter space of the problem, we studied how does the peculiar chemical abundance pattern of an ISM polluted by massive first stars vary when subsequent generations of “normal” stars contaminate it. In fact, normal (Pop II) stars are predicted to form early on after the first stellar generations and thus they can washed-out their key chemical signatures. But when and how this happens? We found that, independent of the choice of the free parameters, an ISM imprinted by the heavy elements from massive first stars at a >50% level it is most likely deficient of some key chemical species: Copper and Zinc. Further, these stars predominantly have [Fe/H] ~ 2.
Looking for the absence of these “killing elements” in large stellar samples might then be an effective method to identify the direct descendants of massive first stars. By exploiting literature data, we found once such a Zinc-poor star at [Fe/H] ~ −2, i.e. BD +80 245. Our analysis and its comparison with the developed model show that this star can be the smoking gun of a PISN explosion (Salvadori et al. 2019).
To probe and constrain the unknown mass distribution of the first stars we need to catch many massive first star descendants. The new strategy developed in SS et al. (2019) paves the way to find more of these rare objects by looking for stars with [(Zn, Cu)/Fe] < 0. Our predictions show that the descendants of massive first stars should be predominantly found in the Galactic bulge, or in relatively massive ancient dwarf galaxy, such as Sculptor and Fornax, which can retain the chemical products of such energetic explosions. The ESO/4MOST large observational program 4DWARFS (512 000 fibre hours) has been recently awarded (Dec. 2021) to our Team (PI: Skuladottir, 33 co-Is including SS, Gelli, Rossi) and it will allow us to search for these objects. In the upcoming years we will then be able to finally constrain the high-mass end of the mass distribution of the first stars. But what about the low-mass end?
Sophisticated but time limited 3D numerical simulations found that fragmentation process is possible in the primordial ISM, leading to the possible formation of low-mass, M* < 0.8 M, zero-metallicity stars. If these stars really existed then they should be found today, since the lifetime of sub-Solar mass stars is longer than the age of the Universe. We can thus exploit the persisting non-detection of zero-metallicity stars to put limits on the minimum mass of the first stars (Mmin). Unfortunately, the stellar sample required to be observed in the Galactic halo or Bulge order to limit Mmin at > 68% confidence level is extremely huge.
In Rossi, SS, Skuladottir (2021) we probed that UFDs are key systems to limit the mass distribution of the first stars at the lowest mass-end. We develop a novel model to follow the formation and evolution of the UFD galaxy Bootes I and we simulated its Color Magnitude Diagram (CDM) as observed today. By performing a statistical comparison between the number of long-lived zero metallicity stars predicted for different first-star mass distribution and those observed so far, we have been able to put tight constraints on Mmin and the characteristic (Mch) mass of the first stars. We found Mmin > 0.8 Msun and Mch > 1 Msun at 99% of the confidence level. Furthermore, we demonstrated that by exploiting next generation telescopes and instruments, such as MOSAIC on ELT we will be able to exclude at the 68% confidence level the existence of low-mass primordial stars, i.e. Mmin > 0.8 M independent on the first-star mass distribution. This result is thus extremely important. But how can be sure that UFD are low-mass mini-halos from the early Universe?
In Gallart et al. (including SS) 2021 we studied the star formation history (SFH) of UFD galaxy Eridanus II by exploiting deep CMD diagrams obtained with new HST observations. Our work demonstrates that the stellar population of Eridanus II is > 13 Gyr old, and that it was formed during a very short period, lasting < 500 Myr, despite a low star-formation rate ~ 0,001 Msun/yr. This result strongly supports the idea that UFDs are associated to low-mass mini-halos.
In conclusion, our work so far has found two key results to constrain the mass distribution of the first stars: we can exploit low-mass UFDs to infer limits on the minimum and characteristic mass, while massive dwarf galaxies can allow us to set its maximum mass and shape. But is the mass distribution of the first stars the only key physical quantity to be investigated?
In Skuladottir, SS, Amarsi et al. (2021) we presented the discovery of AS0039, a chemically peculiar star in the Sculptor dwarf spheroidal galaxy. AS0039 is the most iron-poor star, which has been observe outside of the Milky Way (MW). Contrary to most MW stars at this iron abundance, AS0039 is not enhanced in carbon and it lacks α-element uniformity, in stark contrast with the near solar ratios observed in C-normal stars within the Milky Way halo. Its unique abundance pattern indicates that AS0039 formed out of material that was predominantly enriched by a ~20 M zero-metallicity progenitor star with an unusually high explosion energy E = 10 × 1051 erg. AS0039 represents some of the first observational evidence for zero-metallicity hypernovae. This result is thus fundamental, since it opens a new important window that should be explored studying the properties of the first stars: the energy distribution function.
Finally, with our Team we have been working on the prospects of observing dwarf galaxy satellites with JWST, which has been successfully launched on the 25th of December 2021. We found that planned deep surveys of JWST will be able to catch "for free" the dwarf galaxy satelllites of high-z Lyman Break galaxies (Gelli, SS et al. ApJL 2021).
The discovery of the star AS0039, likely enriched by zero-metallicity hypernovae, forced us to develop novel methodologies in order to study the properties of long-lived ancient stars in the Local Group not as a function of the mass distribution of the first stars but of their energy distributions. By the end of the action, thanks to the synergy between models and observations and the recently awarded massive 4MOST survey focusing of dwarf galaxies, 4DWARFS (PI: Skuladottir) we really expect to be able to limit not only the mass distribution of the first stars but also their energy distribution function.