Spectroscopic Supermassive Dark Star candidates

Abstract

Dark Stars (DSs), i.e., early stars composed almost entirely of hydrogen and helium but powered by Dark Matter (DM), could form in zero metallicity clouds located close to the center of high redshift DM halos. In 2023, three of us identified (in a PNAS work) the first three photometric DS candidates: JADES-GS-z11-0, JADES-GS-z12-0, and JADES-GS-z13-0. We report here our results of a follow-up analysis based on available NIRSpec JWST data. We find that JADES-GS-z11-0 and JADES-GS-z13-0 are spectroscopically consistent with a DS interpretation. Moreover, we find two additional spectroscopic DS candidates: JADES-GS-z14-0 and JADES-GS-z14-1, with the former being the second most distant luminous object ever observed. We furthermore identify, in the spectrum of JADES-GS-z14-0, a tentative feature ([Formula: see text]) indicative of the smoking gun signature of DSs: the He II [Formula: see text]1640 absorption line. In view of ALMA's recent identification of a probable O III nebular emission line in the spectrum of JADES-GS-z14-0, the simple interpretation of this object as an isolated DS is unlikely. If both spectral features survive follow-up observations, it would imply a DS embedded in a metal rich environment, requiring theoretical refinements of the formation of evolution of DSs, which in previous studies were assumed to form in isolation, without any companions.

Keywords: Dark Matter; James Webb Space Telescope; cosmology; high-z galaxies; stars.

Fig. 1.

(A-D) The first four Supermassive Dark Star spectral candidates. The data (blue line) and uncertainty band (shaded gray) plotted against our best fit DS models (orange line). The red dashed line represents where He II$\lambda 1640$ absorption feature, expected only for DSs, might be observed. The normalized residuals ($\frac{F_{\mathit{simul}}-F_{\mathit{measured}}}{{\sigma }_{\mathit{measured}}}$) displayed in the lower panels of each of the SEDs show that our Supermassive DS models lie consistently within $1-\sigma$ of the NIRSpec data for each of the four objects considered. The vertical drop in the fluxes represents the Lyman break, as expected for $z\gtrsim 6$ objects due to absorption by neutral H along the line of sight (69). The vast majority of the other “features” in the observed spectra are actually due to noise. In the title of each plot we display the values assumed for the gravitational lensing factor ($\mu$) and the best fit values for $z_{\mathit{spec}}$ (along with uncertainties).

Fig. 2.

(A-D) Corner plots representing 2D and 1D marginalized posteriors for the free parameters of our Supermassive DS spectral fits (as described in Methods). Note that for JADES-GS-z14-1 (lower right hand plot), since it is not resolved, we only need stellar mass ($M$) and redshift ($z$), whereas for the other three objects we include the number density of hydrogen ($n_H$) in the nebular cloud surrounding the DS. We assumed SMDSs are powered by 100 GeV WIMPs adiabatically contracted (AC) inside the star.

Fig. 3.

Top panels (A-C): Corner plots representing 2D and 1D marginalized posteriors for the free parameters (Sérsic index and angular size in arcseconds) of our H ionization bound nebulae powered by Supermassive DS models for JADES-GS-z11-0, JADES-GS-z13-0, and JADES-GS-z14-0. For more details on the set up of the Monte Carlo simulation, see Methods. Bottom panels (D-F): Radial profiles though F200W NIRCam band for each object. Data are represented by black dots, whereas our best fit model is shown in red. The blue line (and dots) represent the Point Spread Function (PSF) for F200W.

Fig. 4.

Equivalent width of the potential He II absorption feature in JADES-GS-z14-0. The normalized spectral flux is plotted in black, the Voigt profile fit to the feature is in pink, and the line center is shown in the dashed blue line. The uncertainty in the flux is shown in gray. The equivalent width is calculated based on the Voigt profile fit and found to be 3.26 $\mathrm{\AA }$.