XMM-VID1-2075: Early Slow Rotator at z=3.45

• The paper shows that XMM-VID1-2075 exhibits dispersion-dominated stellar kinematics (σ ≈ 387 km/s), confirming its classification as a slow rotator at z=3.449.
• Methodologies such as Monte Carlo SSP modeling and JWST/NIRSpec IFU observations reveal a compact structure, high Sérsic index, and minimal ordered rotation.
• The findings imply that stellar mass assembly, quenching, and angular momentum loss occurred within the first 2 Gyr, marking early slow-rotator formation.

XMM-VID1-2075 is an ultra-massive, quiescent galaxy at $z = 3.449$, observed when the Universe was about $1.8$ Gyr old. It is the first galaxy beyond $z \approx 2$ whose stellar kinematics unambiguously classify it as a slow rotator. Its combination of compact structure, an evolved absorption-line spectrum, minimal ordered rotation, high velocity dispersion, and faint low-surface-brightness asymmetries indicates that the transformation from rotational support to dispersion support can occur within the first two billion years of cosmic time (Forrest et al., 14 Aug 2025).

1. Identification and basic astrophysical properties

XMM-VID1-2075 was selected from the VISTA Deep Extragalactic Observations (VIDEO) survey for its brightness ($m_K = 20.80$ AB), high photometric redshift ($z_{\rm phot} > 3$), and red SED. Spectroscopic confirmation and initial mass and star-formation constraints came from the MAGAZ3NE survey with Keck/MOSFIRE. A Monte Carlo approach using multiple stellar population libraries yields a median systemic redshift of $\tilde{z} = 3.44898 \pm 0.00057$ (stat.) $^{+0.00039}_{-0.00258}$ (syst.), while the E-MILES best-fit value adopted for the kinematic analysis is $z_E = 3.44927$. These values are consistent with previous Keck/MOSFIRE and NIRES measurements (Forrest et al., 14 Aug 2025).

The galaxy has a stellar mass of $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$. Its spectrum is dominated by absorption features, including the Balmer series and the metal lines Ca H&K, G-band, Mg b, and Na D, with only very weak emission. The absence of detectable [O III] $5007$ Å and the prior MOSFIRE constraint $1.8$0 are consistent with a quenched system. Deep ALMA Band 7 and MeerKAT $1.8$1 GHz non-detections further suggest little dust-obscured star formation and no strong radio AGN. At the same time, the detection of [Ne III] $1.8$2 and line ratios such as [N II] $1.8$3/H$1.8$4 in the central spaxels indicate a weak AGN; a heavily obscured AGN remains possible.

Property	Reported value
Redshift	$1.8$5; $1.8$6 (stat.) $1.8$7 (syst.)
Stellar mass	$1.8$8
Star formation	$1.8$9
Quiescence indicators	Absorption-dominated spectrum; very weak emission; no [O III] detection

These properties place XMM-VID1-2075 among the most massive known quiescent galaxies at high redshift. A plausible implication is that substantial stellar mass assembly, enrichment, and quenching had already occurred by $z \approx 2$0.

2. Observational basis and data processing

The decisive observations were obtained with JWST/NIRSpec IFU on 2024-08-02 under GO-2913 (PI Forrest), using the G235M grating and F170LP filter. The wavelength coverage is $z \approx 2$1 at $z \approx 2$2 ($z \approx 2$3). The total integration time was $z \approx 2$4 h, using a 4-point dither and NRSIRS2 readout. The IFU field of view is $z \approx 2$5 with $z \approx 2$6 spaxels, corresponding to about $z \approx 2$7 pc per spaxel at $z \approx 2$8. The central spaxel reaches $z \approx 2$9 per wavelength element, equivalent to about $m_K = 20.80$0 in the rest frame (Forrest et al., 14 Aug 2025).

The data were reduced with JWST pipeline v1.14.0 (jwst_1256.pmap), including detector sensitivity corrections, ramp-fitting, flat-fielding, flagged MSA shutters, outlier detection, 3D cube building, and 1D extraction. Additional custom processing masked contaminating flux and applied stricter cosmic-ray and hot-pixel rejection. The background spectrum was derived from twenty off-source spaxels, smoothed with a Savitzky–Golay filter of window length $m_K = 20.80$1 and polynomial order $m_K = 20.80$2, and then subtracted. The reduction was designed to preserve the very weak nebular emission.

Complementary imaging came from JWST/NIRCam in F150W, F200W, F356W, and F444W under GO-3567 (PI Valentino), with approximately $m_K = 20.80$3 h per band and $m_K = 20.80$4–$m_K = 20.80$5 pixel scale. The NIRCam data reach $m_K = 20.80$6 in F150W and $m_K = 20.80$7 in the other bands. Because the NIRSpec IFU spatial sampling is coarser than that of NIRCam, morphological parameters from the IFU are modestly blurred. Kinematic measurements were therefore corrected for PSF and beam smearing following Harborne et al. (2020) and d’Eugenio et al. (2024).

3. Structure and morphology

XMM-VID1-2075 is very compact, approximately round in appearance, and characterized by a high Sérsic index. From the collapsed NIRSpec/IFU cube, the half-light semi-major axis is $m_K = 20.80$8 kpc and the ellipticity is $m_K = 20.80$9, implying a circularized half-light radius of $z_{\rm phot} > 3$0 kpc and an axis ratio $z_{\rm phot} > 3$1 (Forrest et al., 14 Aug 2025).

Independent NIRCam Sérsic modeling yields consistent structural parameters. Single-Sérsic fits in F150W, F200W, F356W, and F444W give $z_{\rm phot} > 3$2 kpc, $z_{\rm phot} > 3$3, and $z_{\rm phot} > 3$4, with ranges $z_{\rm phot} > 3$5, $z_{\rm phot} > 3$6–$z_{\rm phot} > 3$7 kpc, and $z_{\rm phot} > 3$8–$z_{\rm phot} > 3$9. Two-component and point-source-augmented models do not improve the fit relative to a single Sérsic profile.

Residual images after subtraction of the Sérsic model, especially in F444W, reveal extended faint asymmetric structure on the north-east side. The flux ratio between the central body and this feature is $\tilde{z} = 3.44898 \pm 0.00057$0, consistent with either a minor merger or tidal debris from a past event. The IFU-collapse RGB image likewise shows the half-light ellipse together with faint asymmetric extensions. This suggests a disturbed outer morphology superposed on a compact, centrally concentrated stellar body.

The morphology is therefore not that of an undisturbed rotationally supported disk. The combination of high central concentration and faint asymmetric residuals is consistent with a system that has experienced an interaction capable of modifying its angular-momentum structure.

4. Stellar kinematics and slow-rotator classification

The stellar kinematics were extracted with pPXF from Voronoi-binned spectra. The binning targeted $\tilde{z} = 3.44898 \pm 0.00057$1 per bin, and the measured $\tilde{z} = 3.44898 \pm 0.00057$2 is reported to be insensitive to reasonable changes in the binning scheme (Forrest et al., 14 Aug 2025). The line-of-sight velocity distribution was fit using multiple stellar population libraries—E-MILES, FSPS, GALAXEV, and XSL—while accounting for the instrumental resolution and adopting the E-MILES-based systemic redshift.

The galaxy is dispersion dominated. The stellar velocity dispersion is about $\tilde{z} = 3.44898 \pm 0.00057$3 in the center, and within $\tilde{z} = 3.44898 \pm 0.00057$4 the measured value is $\tilde{z} = 3.44898 \pm 0.00057$5, consistent with the earlier ground-based estimate $\tilde{z} = 3.44898 \pm 0.00057$6. By contrast, all spatial bins have $\tilde{z} = 3.44898 \pm 0.00057$7, with a median absolute deviation $\tilde{z} = 3.44898 \pm 0.00057$8. The kinematic maps show no large-scale velocity gradient, and the ratio map has $\tilde{z} = 3.44898 \pm 0.00057$9 everywhere. There is also no rise of $^{+0.00039}_{-0.00258}$0 toward the outskirts.

The spin parameter is defined as

$^{+0.00039}_{-0.00258}$1

and, in the discrete-bin form used for the measurement within $^{+0.00039}_{-0.00258}$2,

$^{+0.00039}_{-0.00258}$3

summing over bins with $^{+0.00039}_{-0.00258}$4. Before PSF correction, the measured value is $^{+0.00039}_{-0.00258}$5; after PSF correction, $^{+0.00039}_{-0.00258}$6.

This value places XMM-VID1-2075 in the slow-rotator regime of the $^{+0.00039}_{-0.00258}$7–$^{+0.00039}_{-0.00258}$8 plane, below empirical fast–slow boundaries such as those of Emsellem et al. (2011) and van de Sande et al. (2021). The paper also evaluates whether a face-on fast rotator could mimic the observed low $^{+0.00039}_{-0.00258}$9. Using the IFU axis ratio with an oblate disk-like intrinsic shape of $z_E = 3.44927$0,

$z_E = 3.44927$1

the inferred inclination is $z_E = 3.44927$2. At this inclination, substantial rotation would still be detectable. Combined with the morphology ($z_E = 3.44927$3) and the lack of skewness–$z_E = 3.44927$4 correlations, a face-on fast-rotator interpretation is disfavored.

5. Spectral diagnostics, dynamical mass, and physical interpretation

The rest-frame optical spectrum is dominated by stellar absorption features rather than nebular emission. Prominent Balmer absorption and metal lines imply an evolved, metal-enriched stellar population. The stellar-continuum modeling employed multiple SSP libraries through pPXF, but no single age or metallicity solution is reported. Dust attenuation is also not tightly constrained. In the emission-line budget, very weak H$z_E = 3.44927$5 infilling, [N II] $z_E = 3.44927$6, [S II], and [Ne III] $z_E = 3.44927$7 are detected in the central regions, while [O III] $z_E = 3.44927$8 is not detected; the [O II] doublet lies just blueward of the observed wavelength range (Forrest et al., 14 Aug 2025).

Central bins show [N II]/H$z_E = 3.44927$9 ratios suggestive of weak AGN or LINER-like excitation, although other LINER diagnostics such as [O I] are not significantly detected. The paper notes that high-redshift degeneracies between metallicity, ionization parameter, and excitation complicate AGN versus star-formation discrimination. Nonetheless, the absence of star-formation signatures and the overall absorption-dominated spectrum support the interpretation of a quiescent galaxy with weak central nuclear activity.

A simple virial estimator for early-type galaxies,

$$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$0

with $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$1, $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$2 kpc, $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$3, and canonical $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$4, gives $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$5. The paper emphasizes that this value is illustrative and depends on $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$6, anisotropy, structure, any residual rotation, and beam-smearing corrections, but it is consistent with the stellar mass estimate.

The physical interpretation advanced in the paper is that XMM-VID1-2075 has already undergone the angular-momentum loss and structural transformation associated with the formation of massive slow rotators. The faint asymmetric outer structure suggests interaction-driven evolution, while the high central dispersion and negligible ordered rotation indicate a pressure-supported state. This suggests that quenching and kinematic transformation were already well advanced by $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$7.

6. Comparisons, uncertainties, and significance for galaxy evolution

Within the same JWST program, other massive quiescent galaxies at similar redshift show markedly different kinematics. XMM-VID3-1120 and XMM-VID3-2457, both at $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$8 and with stellar masses $$M_\star = 3.3^{+0.1}_{-0.3} \times 10^{11}\ M_\odot$$9 and $5007$0, exhibit clear rotational features, including velocity amplitudes up to $5007$1 and higher spin parameters. More generally, most massive high-redshift quiescent galaxies with spatially resolved spectroscopy have been fast rotators. Against that backdrop, XMM-VID1-2075 is exceptional: prior to this result, no slow rotator had been confirmed from stellar kinematics beyond $5007$2 (Forrest et al., 14 Aug 2025).

The paper discusses several viable formation pathways. Simulations permit cumulative minor mergers, counter-rotating major mergers, or isotropic gas inflow that removes rotation before or during the starburst phase and triggers feedback and quenching. The observed brightness contrast of $5007$3 between the main body and the asymmetric feature is consistent with either a minor merger or tidal debris from a past event, and therefore does not uniquely determine the mechanism. Environment also remains unconstrained: local slow rotators often inhabit dense environments, but no local overdensity or cluster association is reported for this source.

The robustness of the slow-rotator classification is addressed through several tests. Redshift estimates from different SSP libraries are mutually consistent, and the fitted $5007$4 is minimized at the adopted redshift. The kinematic conclusion does not hinge on a single template library. The measured $5007$5 is corrected for PSF and beam smearing and is insensitive to reasonable choices of aperture and binning strategy. At the same time, limitations remain: the spectral resolution of $5007$6 and modest signal-to-noise in the outer bins limit precise deconvolution of beam smearing and prevent measurement of higher-order kinematic moments with high precision. Detailed stellar population gradients and dust attenuation are likewise not tightly constrained.

The broader significance is that XMM-VID1-2075 demonstrates the feasibility of forming a compact, ultra-massive, dispersion-supported slow rotator within $5007$7 Gyr of the Big Bang. In the $5007$8–$5007$9 plane it occupies the slow-rotator regime, in clear contrast to the predominantly fast-rotating massive quiescent galaxies previously observed at $1.8$00. This suggests that the mechanisms that remove angular momentum and produce early-type, pressure-supported systems can operate much earlier than previously confirmed observationally.