Population III (Pop III) stars, the first stellar generation formed from metal-free primordial gas, are predicted to have been born in high-redshift minihaloes. If any formed with masses below roughly 0.8 solar masses, they would still shine today and could in principle be found in the Milky Way and its satellites, providing a direct fossil record of the first stars. Despite dedicated searches over decades, no genuinely metal-free star has ever been observed. This problem asks whether any low-mass metal-free stars formed in the early Universe survive to the present day; how many and where (spatial distribution and kinematics) they should be found; and whether they can be observationally distinguished from metal-poor second-generation stars, given that accretion of interstellar material may pollute their surfaces. Theoretical predictions of survivor counts still span orders of magnitude, simulations have not converged on whether the primordial initial mass function reached sub-solar masses, and the statistical interpretation of every non-detection remains limited by uncertain search selection functions and by unresolved metal-free wind, corona, and accretion physics. As of the September 2025 review by Glover and Klessen, the existence question itself remains open, and the related fronts of survivor numbers, kinematics, and identifiability remain unsettled in the cited specialist literature.

Background

Population III (Pop III) stars are the hypothetical first generation of stars, assembled from primordial gas composed only of hydrogen, helium, and trace lithium left over from the Big Bang. They are thought to have formed at redshifts of order 10–30 in minihaloes - small dark-matter haloes of roughly a million solar masses in which gas cools solely through molecular hydrogen, since no dust or metal lines exist to radiate away heat. Because primordial gas cools less efficiently than present-day interstellar gas, early theory predicted that Pop III stars were very massive (hundreds of solar masses) and short-lived. The mass distribution with which stars form is called the initial mass function (IMF); only stars lighter than about 0.8 solar masses have main-sequence lifetimes exceeding the age of the Universe and can survive to the present day.

Whether the primordial IMF extended to such low masses is unsettled. Radiation-hydrodynamics simulations by Hosokawa et al. (2011) found that ultraviolet feedback from a growing protostar evaporates its accretion disk once the star reaches roughly 43 solar masses, suggesting relatively few low-mass objects, while other simulations of the same era found that primordial accretion disks fragment prolifically, producing low-mass protostars that can be ejected from their birth clusters before growing massive. The outcome is numerically fragile: Prole et al. (2022) performed the first resolution study of primordial star formation with sink mergers enabled (sink particles are point masses that stand in for unresolved stars in simulations) and found that the number of fragments keeps increasing with numerical resolution without convergence, that the IMF shifts toward lower masses at higher resolution, and that in their highest-resolution runs a fraction 0.21 of the fragments light enough to survive to today were ejected from their clusters - a channel that would let low-mass metal-free stars escape the pollution of later supernovae.

Searching for these survivors is the goal of near-field stellar archaeology, the inference of earlier stellar generations from elemental abundances measured in present-day stars. Metallicity is conventionally written [Fe/H][\mathrm{Fe}/\mathrm{H}], the logarithm of the iron-to-hydrogen ratio relative to solar; stars with [Fe/H]<−3[\mathrm{Fe}/\mathrm{H}]<-3 are called extremely metal-poor (EMP) and those with [Fe/H]<−4[\mathrm{Fe}/\mathrm{H}]<-4 ultra metal-poor (UMP). The record holders, such as SMSS J031300.36-670839.3 (Keller et al. 2014, iron undetected) and SMSS J160540.18-144323.1 (Nordlander et al. 2019, [Fe/H]=−6.2[\mathrm{Fe}/\mathrm{H}]=-6.2), all show detectable metals, so they are interpreted as second-generation stars enriched by one or few Pop III supernovae, not as first-generation stars themselves. Komiya, Suda, and Fujimoto (2015) proposed the complementary view that the most iron-deficient stars could be Pop III survivors whose surfaces were polluted after birth: using a hierarchical galaxy-formation chemical-evolution model, they found that accretion of iron-group elements from the interstellar medium onto genuinely metal-free low-mass stars reproduces the observed lowest-metallicity tail, which would mean some known UMP stars are disguised first stars.

The non-detection of metal-free stars is itself informative. Magg, Klessen, Glover, and Li (2019) translated the community's cumulative EMP/UMP discoveries (532 EMP and 42 UMP stars known at the time) into a non-detection probability for model-predicted survivor populations, treating each detected EMP star as a random draw from the combined EMP-plus-pristine pool; they excluded survivor counts of 1800–5200 (the prediction of the semi-analytic model they built on Caterpillar merger trees) at 99.9 percent confidence, allowing at most about 1650 survivors at 98 percent confidence. Their own caveats define the present frontier: the total size of the EMP/UMP haystack is uncertain, the metal-poor searches are targeted rather than blind, and stars that escaped their birth haloes and were accreted onto the Milky Way later may follow different orbits that their model cannot treat. Parallel work sharpened the pollution physics in both directions - Tanaka et al. (2017) and Washinoue and Suzuki argue that magnetic winds and hot, dense, metal-free coronae build a magnetosphere that blocks interstellar accretion and keeps surfaces pristine, while Tanikawa et al. (2018) and Kirihara et al. (2019) show that collisions with interstellar objects like 'Oumuamua can by themselves raise the surface iron abundance of a 0.7–0.8 solar-mass survivor to [Fe/H]≈−6[\mathrm{Fe}/\mathrm{H}]\approx -6 to −2-2, making some known metal-poor stars candidate polluted first stars. Whether surviving Pop III stars are therefore observationally indistinguishable from genuine second-generation (Pop II) stars remains genuinely undecided. As summarized in the September 2025 review by Glover and Klessen, no truly metal-free star has been found, current data rule out a present-day-like primordial IMF, and either the minimum Pop III mass exceeds 0.8 solar masses or the low-mass primordial IMF is much flatter than today's - a disjunction that the field cannot yet resolve.

Problem Statement

Do any low-mass, metal-free (Population III) stars formed in the early Universe survive to the present day, and can they be detected in the Milky Way or its satellites? The existence question itself is resolved affirmatively by the discovery and spectroscopic confirmation of a single genuinely metal-free star (no metals above Big Bang nucleosynthesis composition). On the negative side, the constraint currently available comes from translating survey non-detections into bounds on survivor counts, which - as in the analysis of Magg et al. (2019) - applies to survivors that have preserved their pristine surface composition since formation and cannot by itself exclude an arbitrarily small surviving population; the negative answer is established by a demonstration, theoretical or from simulation, that the primordial initial mass function and ejection channel cannot produce low-mass stars that survive to the present day. As summarized in the September 2025 review by Glover and Klessen, the non-detections so far rule out a present-day-like primordial IMF and leave open whether the minimum Pop III mass exceeds 0.8 solar masses or whether the low-mass end of the primordial IMF is much flatter than today's. The question therefore bears on the four linked points below, with (i) governing whether a survivor population exists and (ii)-(iv) giving the survivor count, locations, and identifiability that follow from that answer: (i) whether the primordial initial mass function extended below roughly 0.8 solar masses and whether such fragments were produced and ejected from their birth clusters frequently enough to escape enrichment and survive - currently an unconverged simulation question; (ii) the predicted present-day number of survivors, which across published models ranges from zero to about 10^5; (iii) their expected present-day locations and kinematics, including whether survivors that escaped their birth haloes and were later accreted onto the Galaxy follow distinct orbits, and whether low-mass Milky Way satellites or the inner halo are the most promising hunting grounds; and (iv) whether survivors that accreted metals from the interstellar medium or from interstellar objects - the scenario in which the most iron-deficient known stars are disguised first stars - can be observationally distinguished from genuine second-generation extremely metal-poor stars, given that competing calculations find stellar winds and hot metal-free coronae either block such pollution entirely or allow pollution up to [Fe/H]≈−2[\mathrm{Fe}/\mathrm{H}]\approx -2.

References

  1. Magg M., Klessen R. S., Glover S. C. O., Li H. (2019), 'Observational constraints on the survival of pristine stars', Monthly Notices of the Royal Astronomical Society 487, 486-490, DOI 10.1093/mnras/stz1210, arXiv:1903.08661, https://arxiv.org/abs/1903.08661
  2. Prole L. R., Clark P. C., Klessen R. S., Glover S. C. O. (2022), 'Fragmentation-induced starvation in Population III star formation: a resolution study', Monthly Notices of the Royal Astronomical Society 510, 4019-4030, DOI 10.1093/mnras/stab3697, arXiv:2112.10800, https://arxiv.org/abs/2112.10800
  3. Komiya Y., Suda T., Fujimoto M. Y. (2015), 'The most iron-deficient stars as the polluted Population III stars', The Astrophysical Journal Letters 808, L47, DOI 10.1088/2041-8205/808/2/L47, arXiv:1507.01664, https://arxiv.org/abs/1507.01664
  4. Glover S. C. O., Klessen R. S. (2025), 'The first stars', Encyclopedia of Astrophysics, Volume 2 (2026), pp. 211-229, Elsevier, DOI 10.1016/B978-0-443-21439-4.00076-6, arXiv:2509.14799, https://arxiv.org/abs/2509.14799
  5. Da Costa G. S., Bessell M. S., Mackey A. D., Nordlander T., Asplund M., Casey A. R., Frebel A., Lind K., Marino A. F., Murphy S. J., Norris J. E., Schmidt B. P., Yong D. (2019), 'The SkyMapper DR1.1 search for extremely metal-poor stars', Monthly Notices of the Royal Astronomical Society 489, 5900-5918, DOI 10.1093/mnras/stz2550, arXiv:1909.06227, https://arxiv.org/abs/1909.06227
  6. Nordlander T., Bessell M. S., Da Costa G. S., Mackey A. D., Asplund M., Casey A. R., Chiti A., Ezzeddine R., Frebel A., Lind K., Marino A. F., Murphy S. J., Norris J. E., Schmidt B. P., Yong D. (2019), 'The lowest detected stellar Fe abundance: the halo star SMSS J160540.18-144323.1', Monthly Notices of the Royal Astronomical Society: Letters 488, L109-L113, DOI 10.1093/mnrasl/slz109, arXiv:1904.07471, https://arxiv.org/abs/1904.07471
  7. Chandra V., Schlaufman K. C. (2021), 'Searching for Low-mass Population III Stars Disguised as White Dwarfs', The Astronomical Journal 161, 197, DOI 10.3847/1538-3881/abe535, arXiv:2102.05058, https://arxiv.org/abs/2102.05058
  8. Magg M., Hartwig T., Agarwal B., Frebel A., Glover S. C. O., Griffen B. F., Klessen R. S. (2017), 'Predicting the locations of possible long-lived low-mass first stars: importance of satellite dwarf galaxies', Monthly Notices of the Royal Astronomical Society 473, 5308-5323, DOI 10.1093/mnras/stx2729, arXiv:1706.07054, https://arxiv.org/abs/1706.07054
  9. Ishiyama T., Sudo K., Yokoi S., Hasegawa K., Tominaga N., Susa H. (2016), 'Where are the Low-Mass Population III Stars?', The Astrophysical Journal 826, 9, DOI 10.3847/0004-637X/826/1/9, arXiv:1602.00465, https://arxiv.org/abs/1602.00465
  10. Hartwig T., Bromm V., Klessen R. S., Glover S. C. O. (2015), 'Constraining the primordial initial mass function with stellar archaeology', Monthly Notices of the Royal Astronomical Society 447, 3892-3908, DOI 10.1093/mnras/stu2740, arXiv:1411.1238, https://arxiv.org/abs/1411.1238
  11. Kirihara T., Tanikawa A., Ishiyama T. (2019), 'Effect of interstellar objects on metallicity of low-mass first stars formed in a cosmological model', Monthly Notices of the Royal Astronomical Society 486, 5917-5923, DOI 10.1093/mnras/stz1277, arXiv:1905.02974, https://arxiv.org/abs/1905.02974
  12. Tanikawa A., Suzuki T. K., Doi Y. (2018), 'Metal pollution of low-mass Population III stars through accretion of interstellar objects like 'Oumuamua', Publications of the Astronomical Society of Japan 70, 80, DOI 10.1093/pasj/psy075, arXiv:1804.08200, https://arxiv.org/abs/1804.08200
  13. Tanaka S. J., Chiaki G., Tominaga N., Susa H. (2017), 'Blocking metal accretion onto Population III stars by stellar wind', The Astrophysical Journal 844, 137, DOI 10.3847/1538-4357/aa7e2c, arXiv:1707.00989, https://arxiv.org/abs/1707.00989
  14. Washinoue H., Suzuki T. K. (2019), 'Coronae of Zero/Low-metal, Low-mass Stars', The Astrophysical Journal 885, 164, DOI 10.3847/1538-4357/ab48ec, arXiv:1903.10148, https://arxiv.org/abs/1903.10148
  15. Washinoue H., Suzuki T. K. (2021), 'Coronal properties of low-mass Population III stars and the radiative feedback in the early Universe', Monthly Notices of the Royal Astronomical Society 506, 1284-1294, DOI 10.1093/mnras/stab1809, arXiv:2106.12740, https://arxiv.org/abs/2106.12740
  16. Tanikawa A., Chiaki G., Kinugawa T., Suwa Y., Tominaga N. (2022), 'Can Population III stars be major origins of both merging binary black holes and extremely metal poor stars?', Publications of the Astronomical Society of Japan 74, 521-532, DOI 10.1093/pasj/psac010, arXiv:2202.00230, https://arxiv.org/abs/2202.00230
  17. Keller S. C., Bessell M. S., Frebel A., Casey A. R., Asplund M., Jacobson H. R., Lind K., Norris J. E., Yong D., Heger A., Magic Z., Da Costa G. S., Schmidt B. P., Tisserand P. (2014), 'A single low-energy, iron-poor supernova as the source of metals in the star SMSS J031300.36-670839.3', Nature 506, 463, DOI 10.1038/nature12990, https://doi.org/10.1038/nature12990
  18. Hosokawa T., Omukai K., Yoshida N., Yorke H. W. (2011), 'Protostellar Feedback Halts the Growth of the First Stars in the Universe', Science 334, 1250, DOI 10.1126/science.1207433, arXiv:1111.3649, https://arxiv.org/abs/1111.3649
  19. Komiya Y., Suda T., Fujimoto M. Y. (2013), 'Current signatures and search for Pop. III stars in the Local Universe', arXiv:1312.5069 (preprint), https://arxiv.org/abs/1312.5069
  20. Johnson J. L. (2015), 'The chemical signature of surviving Population III stars in the Milky Way', Monthly Notices of the Royal Astronomical Society 453, 2772-2779, DOI 10.1093/mnras/stv1815, arXiv:1411.4189, https://arxiv.org/abs/1411.4189
  21. Rossi M., Salvadori S., Skuladottir A., Vanni I., Koutsouridou I. (2025), 'Hidden Population III Descendants in Ultrafaint Dwarf Galaxies', The Astrophysical Journal 987, 121, DOI 10.3847/1538-4357/add5e9, arXiv:2406.12960, https://arxiv.org/abs/2406.12960
  22. Magg M., Hartwig T., Chen L.-H., Tarumi Y. (2022), 'A-SLOTH: Ancient Stars and Local Observables by Tracing Halos', Journal of Open Source Software 7(74), 4417, DOI 10.21105/joss.04417, arXiv:2209.07339, https://arxiv.org/abs/2209.07339