Mass–Metallicity Relation (MZR)

• The mass–metallicity relation is a systematic correlation between a galaxy's stellar mass and its metallicity, observed from dwarf galaxies to massive systems.
• Observations using SDSS, CALIFA, and JWST demonstrate that the relation evolves over cosmic time with varying slopes and normalization shifts.
• Theoretical frameworks connect the MZR to star formation, feedback efficiency, gas inflow dilution, and metal recycling, providing key insights into galaxy evolution.

The mass–metallicity relation (MZR) describes the systematic correlation between a galaxy's stellar mass ($M_*$) and its metallicity, typically measured by gas-phase oxygen abundance (12 + log(O/H)) or stellar [Fe/H]. The MZR is observed across cosmic time and serves as a fundamental diagnostic of galaxy evolution, encapsulating the interplay between star formation, gas inflow, feedback, and metal retention. It is observed from dwarf galaxies ($M_* \sim 10^4\,M_\odot$) to the most massive systems ($M_* \gtrsim 10^{11}\,M_\odot$), and persists out to high redshift ($z > 6$), with both its normalization and slope evolving systematically over cosmic history.

1. Observational Characterization and Evolution

The MZR’s canonical form is an increasing function: more massive galaxies possess higher gas-phase and stellar metallicities. Observations using SDSS, CALIFA, and JWST data demonstrate this relation among both local and high-redshift galaxies, with systematic shifts in the zero point and, at low masses, variations in the slope.

• In the local universe ($z \sim 0$), the MZR follows a form such as:

$$12 + \log(\mathrm{O/H}) \sim Z_0 + \gamma[\log(M_*) - \log(M_0)]$$

where $Z_0$ is a saturation metallicity (e.g., $Z_0 \simeq 8.98$), $\gamma$ is the low-mass slope (typically $0.2$–$0.7$), and $M_0$ is a characteristic mass [$1910.08689$].

• At higher redshifts (e.g., $z \simeq 2.3$), the MZR is offset such that metallicities at fixed $M_*$ are a factor of three to five lower than locally [$1406.6069$], and the high-mass “flattening” remains evident in both observations and simulations [$2211.01382$, $1904.02721$].
• JWST observations now extend the MZR to $z \sim 10$ and $M_* \sim 10^6\,M_\odot$, revealing continuity in the MZR trend and a downward shift in normalization by $\sim0.5$ dex compared to the local MZR, with observable scatter $\sigma_Z \sim 0.3$ dex [$2408.00061$], broadly aligned with results from CROC and SERRA simulations [$2210.16750$, $2408.00061$].

2. Physical Origins and Theoretical Frameworks

The MZR reflects a balance between metal production (via star formation), metal loss (via galactic winds), inflowing metal-poor gas, and the ability of galaxies to retain or recycle metals.

• Feedback Efficiency: Energy-driven and momentum-driven winds are more effective at expelling metals from shallow potential wells in low-mass galaxies. At higher $M_*$ (above a “characteristic mass,” $M_* \sim 10^{9.5}\,M_\odot$), feedback becomes less efficient, allowing galaxies to retain more metals, steepening the MZR slope and eventually saturating it at high mass [$1904.02721$].
• Gas Accretion and Dilution: Metal-poor inflows dilute the ISM, especially prominent in low-mass and high-redshift galaxies with high specific SFRs. The role of gas inflow is evident in the anti-correlation between metallicity and SFR at fixed $M_*$ [$1406.6069$, $1910.08689$, $2211.01382$].
• Metal Recycling: Simulations that explicitly treat metal-rich outflows and recycling (e.g., FIRE, IllustrisTNG, SERRA) show that metals ejected from galaxies are frequently re-accreted, thus mitigating the degree of metal loss and flattening in the MZR [$1504.02097$, $1711.05261$, $2408.00061$].
• Analytic and Regulator Models: The “regulator” model family ties equilibrium ISM metallicity to the balance of yield $y$, gas inflow rate $\dot{M}_\mathrm{in}$, outflow mass loading $\lambda$, and star formation efficiency $\epsilon$:

$$Z_\mathrm{eq} = Z_0 + \frac{y}{1 + \lambda/(1-R) + \epsilon^{-1}[(1+\beta-b)\mathrm{SFR}/M_* - 0.15]}$$

(Lilly et al. 2013, $1406.6069$). These models reproduce the observed shape and evolution of the MZR, with its normalization set primarily by gas fraction and inflow rate (through dilution), and its slope by the mass (or potential well) dependence of feedback efficiency [$1711.05261$, $2408.00061$].

3. Secondary Dependencies: The Fundamental Metallicity Relation (FMR)

At fixed stellar mass, galaxies with higher SFR (or, equivalently, higher gas fraction) tend to have lower metallicity—an anti-correlation known as the FMR. This three-parameter relation is frequently parametrized as:

$\mu_\alpha = \log M_* - \alpha \log \mathrm{SFR}$

with $\alpha$ (empirically) $\sim0.32-0.66$, where $12 + \log(\mathrm{O/H})$ shows minimized scatter when plotted at fixed $\mu_\alpha$ [$1406.6069$, $1910.08689$, $1810.08928$, $2107.00672$].

• The physical origin of the FMR is gas infall: elevated SFR is often triggered by recent inflow of metal-poor gas, which both increases the SFR (by refueling star formation) and lowers the ISM metallicity through dilution [$2101.11021$, $2211.01382$, $2309.07919$].
• The invariance (or otherwise) of the FMR with redshift is debated. The physically-motivated regulator model matches high-redshift observations if regulator parameters are unchanged [$1406.6069$], while strictly empirical parameterizations tend to break down when extrapolated to the SFRs typical of high-$z$ galaxies.
• Recent studies point to a modulation of the strength of SFR–metallicity anti-correlation by stellar population age: the anti-correlation is most pronounced in galaxies with low Dn(4000) (young stellar ages) and essentially disappears or saturates for older/massive galaxies [$2205.01203$].

4. Characteristic Mass Scales, Slope Transitions, and Universality

The MZR is not a pure power law across all masses and epochs.

• Breaks in Slope: A transition at $M_* \sim 10^{9} - 10^{9.5}\,M_\odot$ is commonly observed, above which the slope increases and eventually flattens at the high-mass end [$1904.02721$, $2211.01382$, $2107.04067$]. This transition is interpreted as a shift from feedback-dominated ("metal-loss") to retention-dominated ("metal-retention") regimes.
• Low-mass Regime: Dwarf galaxy MZRs, observed via JWST at $z \sim 2-3$ as well as RR Lyrae metallicities at $z \gtrsim 3$, exhibit a shallow slope ($\alpha\sim0.1-0.2$) until the characteristic mass, steepening above it [$2503.04925$, $2211.01382$]. Theoretical models and semianalytic modeling confirm this double power-law behavior and the near-universality of the MZR's shape across a very large range in $M_*$ and host halo mass [$1903.06054$, $1905.13374$].
• Universality: Simulations and semianalytic models find that the MZR's power-law slope at low/intermediate mass is typically $0.2-0.4$, with minor variations due to environment, halo mass, or satellite vs. central status, as long as sufficient averaging is performed [$1903.06054$, $1905.13374$].

5. Drivers, Causality, and Interpretational Advances

Recent work challenges older interpretations that gravitational retention is the dominant cause of the MZR. Statistical and machine-learning analyses using large IFU surveys (e.g., MaNGA) show that, once stellar mass is selected, the residual dependence of metallicity on dynamical mass or gravitational potential is minimal or even reversed [$2303.08145$].

• Direct Driver: The MZR is fundamentally a relation between the cumulative metal production (which scales with the integral of the star formation history, i.e., $M_*$) and the mass of stars formed. The basic proportionality is $Z \propto M_*/\int \mathrm{SFR}\,dt$, or in integral form:

$$M_* = \int \mathrm{SFR}\,dt \propto \int \frac{dZ}{dt}\,dt \propto Z$$

indicating that $M_*$ is effectively the time-integral of metal production and, by construction, should scale with metallicity, independent of details of galaxy potential well depth or M_dyn.

• Scatter and Stochasticity: The tightness of the MZR is not trivial; numerical models demonstrate that excessively bursty (stochastic) star formation—with large SFR RMS variations—destroys the observed relation, producing “chemical chaos” where $\sigma_Z$ can approach 1.4 dex, in stark disagreement with observed MZR scatter ($\sim0.2$–0.3 dex). Robust MZR and FMR relations at high-$z$ require moderate SFR stochasticity (RMS $\lesssim0.2$ dex) and efficient self-regulation through outflows with relatively low mass-loading factors [$2408.00061$].

6. Methodological Advances and Local “Time Machine” Probes

The evolution of the MZR at the extreme low-mass, high-$z$ regime is out of reach for direct emission-line spectroscopy. Instead, the metallicity distribution function (MDF) of ancient RR Lyrae stars in Local Group dwarfs can be used to reconstruct the “early-epoch” MZR (ee-MZR, Editor's term): for galaxies with at least a few RR Lyrae, the mean $[\mathrm{Fe/H}]$ of this old population reflects the chemical state at $z\gtrsim 3$ [$2503.04925$].

• This strategy allows the MZR to be probed down to $M_* \sim 10^4\,M_\odot$ at early epochs.
• The slope of the ee-MZR is found to be $\sim0.1$, steepening to $\sim0.3$ at the present day, with the difference reflecting continued chemical enrichment in more massive dwarfs after quenching or in the field, and effectively no metallicity evolution in “ultra–faint” dwarfs.
• This result is in strong agreement with high-$z$ emission line studies, validating the universality and temporal persistence of the mass–metallicity scaling.

7. Open Issues, Local–Global Connection, and Future Prospects

Despite the increasingly detailed observational and theoretical work, substantial open questions remain. The shape and normalization of the MZR depend on the details of feedback physics, the relative role of inflows vs. outflows, and the timescale and stochasticity of star formation regulation.

Spatially resolved observations (e.g., from CALIFA and MUSE) reveal variations in metallicity and its dependence on both local (surface mass density, star formation rate surface density) and global (integrated $M_*$) parameters, emphasizing that the galaxy–wide MZR emerges from both small- and large–scale processes [$1407.1315$, $2309.07919$]. Evolutionary changes in ISM structure, disk settling, and the development of metallicity gradients may all modify the observed relation.

High-redshift studies with JWST will continue to refine the universality and possible breaks in the MZR, FMR, and their scatter. The synergy between deep extragalactic spectroscopy, resolved galactic archaeology, and cosmological simulations is enabling new constraints on the metallicity–mass scaling as a fundamental benchmark for galaxy formation models.

---

Summary Table: Evolution and Drivers of the Mass–Metallicity Relation

Regime	Slope $\alpha$ (typ.)	Physics Dominant	Notable Effects
$M_* \lesssim 10^{9.5}$	$0.1$–$0.2$	Metal loss via efficient feedback	Outflows dominate, inefficient enrichment
$M_* \gtrsim 10^{9.5}$	$0.3$–$0.5$	Metal retention	ISM stabilizes, winds less effective
High-$z$ ($z\gtrsim2$)	Similar, but downshifted	Strong inflow, dilution	Lower Z at fixed $M_*$, higher SFR, burstier SFR
Low-$z$ ($z\sim0$)	Flattened at high mass	Retention, quenching	Metallicity saturates, SF suppressed

---

This synthesis reflects the state-of-the-art understanding of the MZR as a fundamental constraint and probe of galaxy chemical evolution, encompassing both the “mean” and the “scatter,” and integrating diverse empirical and theoretical perspectives across cosmic history.