The High-Redshift Lensed Quasar That Wasn't—and the One That Still Might Be

Two faint red sources once looked like rare prizes: gravitationally lensed quasars from the first billion years of cosmic time. If real, they would offer magnified views of the early universe and help test how often quasar light is magnified—and how lensing can bias samples of the brightest early quasars. A new multi-wavelength study, posted as a preprint on arXiv, follows both candidates with JWST, Hubble, Gemini, and the Large Binocular Telescope. The result is a split verdict. One candidate is not what it seemed. The other remains stubbornly ambiguous.

A Break in the Spectrum That Broke the Wrong Way

J0109-5424 was selected as a z6 lensed quasar candidate because its optical spectrum showed a sharp drop at about 8600 Å, resembling the Lyman break. That break is the edge where intergalactic hydrogen absorbs almost all quasar light at wavelengths shorter than the Lyman-alpha line. For a quasar at z6, the break lands in optical filters and is a standard calling card of cosmic dawn.

But JWST changed the picture. Using the NIRSpec instrument in its integral-field mode, which splits light from a small patch of sky into a spectrum, the team detected hydrogen Paschen emission lines at a redshift of z=2.07. The sharp break was not a Lyman break at all. It was a blend of magnesium and iron broad absorption lines, created by gas flowing outward from the quasar. The object is a FeLoBAL quasar: an iron low-ionization broad absorption line quasar, a class known for reddening its own light and hiding strong emission lines.

That reclassification matters. The object is not a beacon from the first billion years but an intermediate-redshift quasar whose spectrum mimicked a far more distant one. From the Paschen-beta emission line, the team estimates a black hole mass of about 3.9 × 10^8 solar masses and a bolometric luminosity—its total energy output across all wavelengths—of 7.7 × 10^45 erg per second. Its Eddington ratio is 0.16, meaning it is accreting at roughly 16 percent of the theoretical maximum set by radiation pressure. The team notes it is among the highest-redshift FeLoBAL quasars known. The paper adds that the revised redshift alone does not entirely exclude the possibility that the z~2 quasar is still lensed; any foreground lens would lie at z<2.07 and might be detectable in optical imaging.

A Compact Neighbor, but No Lens Galaxy

P170+20 is a different puzzle. It was flagged as a possible z~6.4 lensed quasar because its spectrum showed excess light blueward of the Lyman-alpha break. A foreground galaxy could bend and magnify the quasar, while also contributing its own light at those wavelengths.

Hubble's ACS camera imaged the field in the F555W filter, which probes below the quasar's Lyman limit—the short-wavelength edge where hydrogen absorption becomes nearly complete—where the quasar itself should be dark. The image revealed a compact source only 0.09 ± 0.04 arcsec from the quasar. The team modeled and subtracted the point-spread function, the blurring pattern of a point source. No extended emission remained. That is hard to square with a foreground galaxy acting as a lens, because such a galaxy should usually be resolved.

Spectra from Gemini North and South and the Large Binocular Telescope tell a mixed story. The combined optical and near-infrared spectrum is consistent with a quasar plus an M-dwarf star. But the interpretation is not comfortable: a chance alignment between an ultracool dwarf and a quasar within 0.1 arcsec is unlikely. The team estimates about a 0.01 percent chance for any one quasar; the probability that at least one of the more than 480 known quasars at z≳5.5 has such an alignment rises to about 4 percent. A compact elliptical galaxy at z~1.04 can explain part of the break, but it fits the near-infrared spectrum poorly.

The Proximity Zone That Is Too Small—and What Could Settle It

One clue still pulls toward lensing. The quasar's proximity zone, the bubble of ionized hydrogen it carves around itself, appears unusually small. The team cannot measure it reliably because of foreground contamination, but in the Gemini spectrum the Lyman-alpha break is sharp enough to potentially suggest a zone of about 1 megaparsec or less. Given the quasar's reported brightness, that is about three times smaller than expected.

There are two ways out. The quasar could be extremely young, with its bright phase starting only 10^3 to 10^4 years ago. Or it could be intrinsically fainter and magnified by a gravitational lens by a factor of 15 or more. That second option keeps the lensing scenario alive.

The study does not claim a final answer for P170+20. A dedicated analysis of Hubble imaging could search for quasar light split into multiple lensed components. Additional epochs could measure proper motion. If the companion is an M dwarf in the Milky Way halo, it might shift by about 0.2 arcsec in roughly nine years, according to the team's estimate. Until then, the lensing possibility cannot be conclusively ruled out.

High-redshift quasar surveys often assume that the light we see is the light the quasar actually emits. Gravitational lensing breaks that assumption. Magnification can inflate black hole mass estimates and bias the bright end of the quasar luminosity function, the census of how many bright quasars exist at a given cosmic time. Only one gravitationally lensed quasar at z>6 is confirmed, J0439+1634 at z=6.51, with a magnification of about 50.

That makes impostors important. A FeLoBAL quasar can masquerade as a Lyman-break source. An M-dwarf can mimic a quasar's red colors. The new work shows how multi-wavelength follow-up can unmask one candidate and leave another hanging in the balance. For now, one suspected cosmic dawn quasar belongs to a later era, and the other remains a flicker at the edge of what current telescopes can resolve.