Inside-Out Quenching in Galaxies

• Inside-out quenching is a spatially resolved process where galaxy centers experience early star formation suppression relative to the outer disks.
• Observational and simulation studies use resolved sSFR profiles, stellar age gradients, and gas diagnostics to identify the central depletion of star-forming activity.
• Proposed mechanisms such as SMBH feedback, wet compaction, and environmental starvation offer complementary insights into the physical drivers and timescales of this phenomenon.

Inside-out quenching is a spatially resolved mode of galaxy quenching in which star formation is suppressed first in the central kiloparsecs and only later at larger radii, so that the nucleus or bulge becomes more strongly quenched than the outer disk. In practice, the phenomenon is identified through centrally depressed specific star-formation-rate profiles, outward-increasing stellar age profiles, central deficits of cold gas, or centrally concentrated maps of quenched spaxels. The concept is now used across simulations, IFU surveys, slitless spectroscopy, molecular-gas imaging, and JWST spectroscopy, but the inferred prevalence and timescale depend strongly on the tracer and on how “quenching” is operationalized (Nelson et al., 2021, Tacchella et al., 2015, Lin et al., 2019, Ho et al., 31 May 2026).

1. Definitions and diagnostic framework

A common starting point is the resolved specific star-formation rate, $\mathrm{sSFR} = \mathrm{SFR}/M_*$, measured as a function of radius. In this language, inside-out quenching is the emergence of a centrally depressed $\mathrm{sSFR}(r)$ profile relative to the outer disk, often expressed as an outward-rising profile or as a central-to-disk $\mathrm{sSFR}$ ratio below unity. In studies anchored to the star-forming main sequence, the relevant control parameter is frequently the main-sequence offset $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$, so that galaxies below the main sequence can be compared at fixed mass and epoch (Nelson et al., 2021).

Other formulations use surface-density profiles rather than integrated annular quantities. At $z \approx 2.2$, inside-out quenching is diagnosed jointly from $\Sigma_\star(r)$, $\Sigma_{\rm SFR}(r)$, and $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$, with the key signature being inner $\mathrm{sSFR}$ values orders of magnitude lower than in the outer disk at fixed total stellar mass (Tacchella et al., 2015). In local IFU work, quenching is often mapped non-parametrically through the quenched fraction

$F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$

and the concentration of quenched area $\mathrm{sSFR}(r)$0, so that centrally concentrated quenched spaxels are classified as inside-out and peripherally concentrated quenched spaxels as outside-in (Lin et al., 2019).

A distinct but related class of diagnostics follows stellar populations rather than instantaneous star formation. Negative age gradients, positive mass-to-light gradients, and color gradients can all encode earlier central shutdown. At cosmic noon, JWST/NIRSpec measurements define radial gradients as two-point slopes in units of per $\mathrm{sSFR}(r)$1 for $\mathrm{sSFR}(r)$2, [Fe/H], [Mg/H], and [Mg/Fe], and interpret negative age gradients as evidence that cores formed and quenched earlier than outskirts (Cheng et al., 15 Sep 2025). The choice of tracer is consequential: in MaNGA DR17, sSFR, $\mathrm{sSFR}(r)$3, post-starburst selection, and LI(N)ER emission probe different timescales and therefore produce different apparent quenching geometries (Ho et al., 31 May 2026).

Diagnostic family	Operational signature	Representative studies
Resolved sSFR, EW(H$\mathrm{sSFR}(r)$4), $\mathrm{sSFR}(r)$5	Central sSFR depression or outward-rising $\mathrm{sSFR}(r)$6	(Nelson et al., 2021, Tacchella et al., 2015)
Non-parametric spaxel metrics	High $\mathrm{sSFR}(r)$7 at fixed $\mathrm{sSFR}(r)$8 for centrally concentrated quenched areas	(Lin et al., 2019, Ho et al., 31 May 2026)
Gas-based diagnostics	Low central $\mathrm{sSFR}(r)$9 or short central $\mathrm{sSFR}$0	(Spilker et al., 2019)
FIR density diagnostics	Low [O III] $\mathrm{sSFR}$1m from feedback-lowered central density	(Inoue et al., 2021)
Archaeological gradients	Negative age gradients; evolving $\mathrm{sSFR}$2 and color gradients	(Avila-Reese et al., 2023, Cheng et al., 15 Sep 2025)

2. Observational evidence across cosmic time

At high redshift, one of the clearest resolved demonstrations came from star-forming galaxies at $\mathrm{sSFR}$3, where the most massive systems show disk-like $\mathrm{sSFR}$4, increasingly bulge-like $\mathrm{sSFR}$5, and strongly outward-rising $\mathrm{sSFR}$6. Within $\mathrm{sSFR}$7, $\mathrm{sSFR}$8 declines from approximately $\mathrm{sSFR}$9 at $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$0 to approximately $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$1 at $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$2. A forward-evolution model then yields quenching timescales of much shorter than $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$3 in the inner $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$4 and approximately $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$5–$$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$6 in the outskirts, implying a radially propagating quenching front (Tacchella et al., 2015).

At $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$7, a direct comparison between 3D-HST and TNG50 established that massive sub-main-sequence galaxies with $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$8 show strong central $$\Delta \mathrm{MS} = \log_{10}(\mathrm{SFR}) - \log_{10}(\mathrm{SFR}_{\mathrm{MS}(M_*,z)})$$9 suppression in both data and simulation. The depression extends across the inner few kiloparsecs; the central $z \approx 2.2$0 in these systems is substantially reduced relative to the outer disk, and the suppression amplitude agrees between TNG50 and 3D-HST to within $z \approx 2.2$1 dex. By contrast, lower-mass galaxies generally have flat or gently rising $z \approx 2.2$2 profiles, more consistent with self-similar growth than with strong central suppression (Nelson et al., 2021).

Environmental studies at similar epochs reinforce the mass dependence. In the Spiderweb protocluster at $z \approx 2.2$3, galaxies with $z \approx 2.2$4 show central $z \approx 2.2$5 suppressed by approximately an order of magnitude relative to the outskirts, whereas lower-mass star-forming galaxies have nearly flat $z \approx 2.2$6 profiles. The same dataset finds an anti-correlation between central star-formation activity and Sérsic index, so that bulge-dominated systems have the strongest central suppression (Laishram et al., 21 Dec 2025).

Local surveys add two further dimensions. First, MaNGA analyses based on LI(N)ER-selected quenched spaxels find that inside-out quenching is more common than outside-in quenching in all environments and becomes more prevalent with both stellar mass and halo mass; for example, 68% of galaxies with $z \approx 2.2$7 are classified as inside-out, compared with 5% outside-in (Lin et al., 2019). Second, a statistically controlled SDSS study shows that low-mass satellites in clusters quench in a more inside-out pattern than isolated controls matched in mass, total sSFR, and fiber coverage, with the effect strongest below the main sequence and in the cores of massive clusters (Wang, 2022).

3. Physical mechanisms proposed for inside-out quenching

A major mechanism is centrally coupled SMBH feedback. In TNG50, inside-out quenching in massive sub-main-sequence galaxies is attributed to the low-accretion SMBH feedback mode, in which energy is injected kinetically in time-pulsed events regulated by the Eddington ratio $z \approx 2.2$8. The energetics follow $z \approx 2.2$9. This feedback acts locally on gas surrounding the black hole on kiloparsec scales, driving turbulence, heating, and ejective outflows that evacuate the inner kiloparsecs and depress central star formation. The contrast with the original Illustris implementation is explicit: Illustris injected low-accretion feedback non-locally via hot bubbles at $\Sigma_\star(r)$0–$\Sigma_\star(r)$1 and did not reproduce the observed central suppression (Nelson et al., 2021).

A second framework is wet compaction followed by depletion. In VELA zoom-in simulations, galaxies pass through a blue-nugget phase in which central gas cusps and central SFR peaks form after rapid dissipative inflow. Once inflow into the inner kiloparsec drops below $\Sigma_\star(r)$2, the center becomes gas-poor, the stellar core saturates, and a star-forming ring emerges at radii of approximately $\Sigma_\star(r)$3–$\Sigma_\star(r)$4. The resulting $\Sigma_\star(r)$5 rises outward and the outer disk quenches roughly $\Sigma_\star(r)$6–$\Sigma_\star(r)$7 after the center (Tacchella et al., 2015). THESAN-ZOOM generalizes this into a recurrent high-redshift cycle of central compaction, starburst above the main sequence, inside-out quenching, and size regrowth through spatially extended star formation (McClymont et al., 6 Mar 2025).

Semi-analytic work shows that AGN wind bubbles can also generate inside-out quenching without requiring immediate galaxy-wide gas removal. In a flickering-AGN model, the injected energy is of order the gas binding energy, so outflows remain mostly within $\Sigma_\star(r)$8 during active phases and then coast outward. The external pressure of the bubble enhances star formation more strongly in outer disks than in centers, while central gas is consumed and depleted more rapidly, producing an inside-out pattern in gas exhaustion (Zubovas et al., 2016).

Not all inside-out quenching is attributed to AGN alone. Observational studies at $\Sigma_\star(r)$9 and in the Spiderweb protocluster also discuss bulge growth, morphological quenching, and compaction as complementary processes, with AGN-driven outflows serving as one possible accelerator of central suppression rather than the sole cause (Tacchella et al., 2015, Laishram et al., 21 Dec 2025). In nearby NGC 1371, the combination of a stellar bar, a central $\Sigma_{\rm SFR}(r)$0 HI hole, no detected molecular gas in the inner region, and low-power AGN jets with $\Sigma_{\rm SFR}(r)$1 radio bubbles supports a mixed picture in which bar-driven depletion and AGN heating or removal of gas both contribute to inside-out quenching (Veronese et al., 23 Sep 2025).

Environmental mechanisms can also yield an inside-out signature. In low-mass cluster satellites, the proposed pathway is starvation: stripping of the hot gaseous halo suppresses low-angular-momentum accretion and radial inflows into the center, so the inner disk exhausts its gas first even if the outer star-forming disk is not yet strongly truncated (Wang, 2022). This suggests that “inside-out” describes a spatial pattern, not a unique physical channel.

4. Gas, stellar populations, and structural consequences

The most direct fuel-based evidence comes from high-resolution CO imaging. In the compact star-forming galaxy COSMOS27289 at $\Sigma_{\rm SFR}(r)$2, both the molecular gas fraction and the depletion time are lower in the central $\Sigma_{\rm SFR}(r)$3–$\Sigma_{\rm SFR}(r)$4 than at radii of approximately $\Sigma_{\rm SFR}(r)$5–$\Sigma_{\rm SFR}(r)$6. Quantitatively, $\Sigma_{\rm SFR}(r)$7 rises from about $\Sigma_{\rm SFR}(r)$8 in the center to about $\Sigma_{\rm SFR}(r)$9–$$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$0 at larger radii, while $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$1 rises from about $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$2–$$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$3 to about $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$4–$$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$5. This implies that the center will exhaust its molecular reservoir first and provides direct molecular-gas evidence for imminent inside-out quenching (Spilker et al., 2019).

A complementary approach uses FIR fine-structure line ratios as density tracers. Using TNG and Illustris outputs, one study proposes the [O III] $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$6m/$$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$7m ratio as a probe of BH-driven inside-out quenching, because the ratio is sensitive to local electron density. In IllustrisTNG, where feedback abruptly lowers central gas density when black holes reach a critical regime, the ratio drops sharply around $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$8 and around $$\mathrm{sSFR}(r)=\Sigma_{\rm SFR}(r)/\Sigma_\star(r)$$9 in integrated apertures. In Illustris, where the low-accretion mode is not as tightly coupled to central gas evacuation, no such break appears (Inoue et al., 2021).

Resolved stellar-population gradients provide an archaeological record of the same process. In JWST-SUSPENSE spectroscopy of quiescent galaxies at $\mathrm{sSFR}$0, the measured gradients are negative in age, positive in [Mg/H] and [Mg/Fe], and flat in [Fe/H]. The negative age gradients indicate older cores and are interpreted as inside-out quenching, while the Mg-deficient cores are taken to suggest rapid gas expulsion as the central quenching mechanism. The flat [Fe/H] and positive [Mg/Fe] gradients remain puzzling in this scenario, and the authors explicitly note that the abundance pattern is not fully captured by a simple central-first shutdown picture (Cheng et al., 15 Sep 2025).

At low redshift, fossil-record analyses of classical ellipticals recover a different but related signature: negative $\mathrm{sSFR}$1 and color gradients that steepen and then flatten in time, driven mainly by inside-out quenching rather than by substantial outer mass growth within approximately $\mathrm{sSFR}$2–$\mathrm{sSFR}$3. The constancy of the inner-versus-outer stellar mass surface density contrast argues against strong late structural growth over the radii probed, while positive $\mathrm{sSFR}$4 indicates earlier central shutdown (Avila-Reese et al., 2023).

Inside-out quenching also alters photometric decomposition. If the central disk has been quenched, extrapolating an outer exponential disk inward overestimates the true central disk brightness. This “Dio” effect causes oversubtraction of disk light underneath the bulge and therefore underestimation of bulge luminosity. Evolutionary models and IFU-based age slicing indicate that the resulting deficit can exceed about $\mathrm{sSFR}$5 mag in the optical and can strongly bias bulge-to-total ratios and bulge–SMBH scaling relations (Papaderos et al., 2021).

5. Methodological dependence and interpretive disputes

A central issue in the literature is that inside-out quenching is not inferred identically by all methods. In radial SED fitting of 1,440 galaxies at $\mathrm{sSFR}$6, traditional assumptions—a fixed dust law and uniform priors—produce steep quiescent $\mathrm{sSFR}$7 gradients already at $\mathrm{sSFR}$8. When physically motivated priors are adopted instead, the interpretation changes: from $\mathrm{sSFR}$9 to $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$0, green-valley and quiescent galaxies have similar $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$1 slopes, implying quenching at all radii simultaneously; only from $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$2 to $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$3 do the profiles steepen, implying later inside-out quenching (Vega et al., 10 Jan 2025).

Tracer choice has a similarly strong effect in MaNGA DR17. Using the same $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$4–$F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$5 framework on approximately 10,220 galaxies, the sSFR criterion yields comparable inside-out and outside-in fractions, $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$6 and LI(N)ER strongly favor inside-out patterns, and post-starburst spaxels preferentially occupy a distinct, often outside-in-like region of the plane. These differences reflect the fact that sSFR, $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$7, PSB selection, and LI(N)ER emission probe complementary timescales of star-formation suppression rather than a single universal quenching geometry (Ho et al., 31 May 2026).

Not all negative gradients imply quenching at all. In local galaxies, color-gradient modeling has shown that the observed $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$8 and $F_q = N_{\mathrm{quenched}}/N_{\mathrm{all}}$9 distributions can be reproduced either by inside-out growth or by inside-out quenching at the level of gradient amplitude, but under the assumptions of a constant galaxy formation rate and a common SFH, the distribution of gradients is better matched by inside-out growth: the inner region has a shorter e-folding timescale in the growth phase rather than an earlier onset of rapid quenching (Lian et al., 2017). Likewise, in the Milky Way inner disk, the quenching episode at lookback time $\mathrm{sSFR}(r)$00 is interpreted not as a whole-disk inside-out wave but as a bar-linked decline in star-formation efficiency inside the outer Lindblad resonance, with no evidence for inside-out thick-disk formation (Haywood et al., 2018).

These disagreements do not necessarily negate inside-out quenching. They indicate that the phrase can refer to at least three distinct inferences: instantaneous central SFR suppression, long-timescale stellar-population gradients, or structural consequences of central shutdown. A plausible implication is that different methods are sampling different phases of the same transformation sequence.

6. Comparative synthesis and future directions

Taken together, the current literature supports inside-out quenching as a major, but not exclusive, mode of galaxy shutdown. It is most clearly expressed in massive systems below the star-forming main sequence, in galaxies with high central stellar densities or high Sérsic indices, and in quiescent or quenching populations whose centers are older, gas-poorer, or more strongly suppressed than their outskirts (Nelson et al., 2021, Tacchella et al., 2015, Laishram et al., 21 Dec 2025). Simulation-based classifications at $\mathrm{sSFR}(r)$01 suggest that inside-out quenched galaxies are more likely to be field primaries, can take up to approximately $\mathrm{sSFR}(r)$02 to quench, and are morphologically separable from outside-in systems through the joint evolution of the concentration and size of the star-forming component, whereas outside-in quenched galaxies are more often satellites and typically quench on approximately $\mathrm{sSFR}(r)$03 timescales (Lawlor-Forsyth et al., 28 Feb 2026).

At the same time, the topic remains method-sensitive. Some datasets imply simultaneous suppression at all radii over part of cosmic time, some emphasize inside-out growth rather than inside-out quenching, and some show that outside-in suppression becomes common in low-mass satellites or in particular transitional tracers (Vega et al., 10 Jan 2025, Lian et al., 2017, Ho et al., 31 May 2026). The most robust generalization is therefore conditional: inside-out quenching is a spatial pattern that can emerge from several mechanisms—SMBH feedback, compaction followed by depletion, bulge-driven stabilization, starvation, or bar-driven gas redistribution—and its observational prominence depends on whether one is measuring recent star formation, gas supply, ionized-gas diagnostics, or the archaeological imprint in stellar populations.

The immediate empirical agenda follows directly from these tensions. High-resolution IFU and JWST/NIRSpec observations can test the radial extent of suppression where TNG50 is slightly high at $\mathrm{sSFR}(r)$04–$\mathrm{sSFR}(r)$05, and can connect age and abundance gradients to resolved gas kinematics (Nelson et al., 2021, Cheng et al., 15 Sep 2025). ALMA and NOEMA can search for the blue-nugget cusps and post-compaction rings predicted by zoom-in simulations, while low-$\mathrm{sSFR}(r)$06 CO and FIR line ratios can test whether central density drops accompany quenching onset (Tacchella et al., 2015, Inoue et al., 2021). Spatially resolved studies with JWST, ELT, and Euclid can also quantify the Dio bias in quenched central disks and its consequences for bulge demographics (Papaderos et al., 2021). In that sense, inside-out quenching has become not merely a descriptive label for radial SFR decline, but a cross-wavelength framework for linking feedback, gas cycling, stellar-population gradients, and structural evolution in galaxy formation.