---
title: Extended Main-Sequence Turn-Offs in Clusters
url: https://www.emergentmind.com/topics/extended-main-sequence-turn-offs-emstos
type: topic
---

# Extended Main-Sequence Turn-Offs in Clusters

Extended main-sequence turn-offs (eMSTOs) are broadened, elongated, or otherwise structured main-sequence turn-off regions in the color–magnitude diagrams of star clusters, in contrast to the narrow MSTO expected for a single-age, single-metallicity simple stellar population. In young clusters, eMSTOs are frequently associated with broadened or split upper main sequences; in some intermediate-age systems they are accompanied by extended red clumps. Across the literature summarized here, eMSTOs have been identified in Magellanic Cloud clusters, Milky Way open clusters, and very young Galactic systems, and they are now treated as a major challenge to the canonical simple-stellar-population picture. Their interpretation remains contested among age spreads, stellar rotation, internal mixing, binaries, variability, and circumstellar dust, although stellar rotation is described in the recent review literature as the most plausible contemporary mechanism on a global basis [2401.08062].

## 1. Morphological definition and quantitative characterization

In a classical simple stellar population, the MSTO is a narrow locus because stars share one age and one metallicity. An eMSTO is instead a broad or fan-shaped MSTO region, often extending in both color and magnitude, and in some clusters it is accompanied by a broadened or split upper main sequence. In Magellanic Cloud clusters, nearly all clusters younger than $\sim 2.5~\mathrm{Gyr}$ show eMSTOs, while all clusters younger than $\sim 700~\mathrm{Myr}$ show broadened or split main sequences; the same age-dependent morphology is reported for Milky Way open clusters younger than $\sim 1.5~\mathrm{Gyr}$ [1811.01192].

A widely used operational procedure is to interpret the extended MSTO with non-rotating isochrones and derive a pseudo-age distribution. In the Gaia-based analysis of Galactic open clusters, the intrinsic age spread is defined as
$$
\sigma_{\rm AGE}=\sqrt{\sigma_{\rm AGE,obs}^2-\sigma_{\rm AGE,sim}^2},
$$
with
$$
\mathrm{FWHM}=2.355\,\sigma_{\rm AGE}.
$$
Under that interpretation, the twelve Galactic open clusters studied by Cordoni et al. span ages from $\sim 180~\mathrm{Myr}$ to $\sim 1.13~\mathrm{Gyr}$ and apparent FWHM age spreads from $\sim 75~\mathrm{Myr}$ to $\sim 270~\mathrm{Myr}$ [1811.01192]. In the Magellanic Clouds, Goudfrooij et al. reported MSTO FWHM widths corresponding to age spreads of $200$–$550~\mathrm{Myr}$ in a mass-limited sample of eighteen intermediate-age clusters [1410.3840].

The morphology is not uniform across stellar mass. In Galactic open clusters older than $\sim 700~\mathrm{Myr}$, the main sequence becomes narrow below the “MS kink” at $T_{\rm eff}\sim 7000\,\mathrm{K}$, while in the Large Magellanic Cloud cluster NGC 1831 the kink occurs at an initial stellar mass of $1.45 \pm 0.02\,M_\odot$ and marks the point below which the effects of rotation on the energy output are argued to become negligible at the metallicity of these clusters [1811.01192] [1807.04737].

## 2. Discovery space, environments, and demographic patterns

The phenomenon first became prominent in intermediate-age Magellanic Cloud clusters. NGC 411 provided the second Small Magellanic Cloud case after NGC 419, and the two clusters show strikingly similar eMSTOs despite present masses differing by a factor of $4$; the age range required to bracket the NGC 411 MSTO is $1.51$ to $2.19~\mathrm{Gyr}$ [1303.1361]. In the Large Magellanic Cloud, low-mass clusters also proved important: Piatti and Bastian reported eMSTOs in clusters with masses $<5000\,M_\odot$, more than five times less massive than the eMSTO clusters previously studied, and concluded that mass is not the controlling factor and that a large core radius is not a requisite for a cluster to exhibit an eMSTO [1603.06891].

The demographic picture was broadened by larger cluster samples. In a study of $37$ Large Magellanic Cloud eMSTO clusters, Piatti and Bastian found that core radii, masses, and age-relaxation time ratios are not related to the genesis of eMSTOs, whereas the FWHM at the MSTO correlates most strongly with cluster age [1608.04554]. This age dependence was later echoed in Galactic work: Gaia DR2 photometry and proper motions showed that twelve Milky Way open clusters display eMSTOs and/or broadened main sequences, and that the Milky Way sample follows the same apparent age-spread versus age relation as Magellanic Cloud clusters [1811.01192].

The age range of confirmed systems also expanded downward. NGC 1856 yielded the first evidence for an eMSTO in a young cluster, at $\sim 300~\mathrm{Myr}$, with a broadened MSTO consistent with prolonged star formation of about $150~\mathrm{Myr}$ and a split main sequence whose red and blue components host $33\pm5\%$ and $67\pm5\%$ of the main-sequence stars, respectively [1504.03252]. Even younger still, the double cluster $h$ and $\chi$ Persei, at $\sim 14~\mathrm{Myr}$, shows eMSTO regions in Gaia DR2 CMDs: stars with masses below $\sim 1.3\,M_\odot$ populate narrow main sequences, while more massive stars define the eMSTO [1904.02005].

## 3. Observational diagnostics and measurement strategies

The empirical case for eMSTOs depends on aggressive control of membership, reddening, field contamination, binaries, and photometric scatter. In Galactic open clusters, Gaia DR2 proper motions and parallaxes are used iteratively to define membership and reject field stars, after which differential reddening is corrected using color residuals along the main sequence and local-neighborhood methods [1811.01192]. In Magellanic Cloud cluster work, the standard approach is deep HST photometry, completeness characterization with artificial-star tests, and field-star subtraction using equal-area reference fields and CMD-space distance metrics [1410.3840].

A central diagnostic is the pseudo-age distribution. The usual procedure is to define a parallelogram around the MSTO, with one axis approximately parallel to isochrones and the other perpendicular to them, project the stars onto the age-sensitive axis, and convert that coordinate into age using a grid of isochrones. Goudfrooij et al. expressed the intrinsic broadening through quantities such as
$$
{\rm FWHM}_{\rm MSTO}=\sqrt{{\rm FWHM}_{\rm obs}^2-{\rm FWHM}_{\rm SSP}^2},
$$
where the SSP term is derived from Monte Carlo simulations including unresolved binaries and photometric errors [1410.3840]. The same MSTO-derived age distributions were then compared with subgiant-branch and red-clump morphologies, and in five intermediate-age Large Magellanic Cloud clusters the cross-SGB profiles were found to be consistent with the cross-MSTO profiles when the latter were interpreted as age distributions [1503.07862].

Spectroscopy adds a direct kinematic dimension. In NGC 5822, a SALT/RSS spectroscopic survey of MSTO stars at $R\sim 4000$ measured projected rotational velocities and found a clear correlation between CMD position and $v\sin i$: fast rotators lie on the red side of the eMSTO and slow rotators on the blue side [1904.03547]. More recent methodology extends beyond classical isochrone fitting. Johnston et al. introduced “isochrone-clouds,” defined as the union of many isochrones at the same age and metallicity but with different convective-core boundary mixing and radiative envelope mixing profiles, and fitted cluster stars by minimizing the Mahalanobis Distance to points in the cloud [1910.00591]. In a further development, synthetic “Base Stellar Populations” including stellar rotation were combined with a genetic algorithm to fit both eMSTO and extended-red-clump histograms in NGC 419 and NGC 1817 [2509.13601].

## 4. Principal interpretations and the central controversy

One major interpretation is that eMSTOs trace genuine internal age spreads. In the Magellanic Cloud sample analyzed by Goudfrooij et al., all eighteen intermediate-age clusters in the mass-limited sample feature eMSTOs wider than predicted by a simple stellar population, the fraction of red-clump stars in secondary red clumps scales with the fraction of MSTO stars having pseudo-ages $\leq 1.35$ Gyr, and the width of the pseudo-age distributions correlates with central escape velocity, both currently and at an age of $10~\mathrm{Myr}$ [1410.3840]. In related work, subgiant-branch and red-clump morphologies in NGC 1651, NGC 1783, NGC 1806, NGC 1846, and IC 2146 were argued to be consistent with the age-spread scenario and inconsistent with simulated simple stellar populations [1503.07862].

The competing interpretation is that stellar rotation is the main driver. Rotation alters stellar structure and observable properties through gravity darkening, inclination effects, and rotational mixing; in many observational studies the red side of the eMSTO or the red branch of a split main sequence is associated with fast rotators, while the blue side is associated with slow rotators. NGC 5822 provides a direct Milky Way example: the red side of the eMSTO is populated predominantly by fast-rotating stars and the blue side by slow or moderate rotators, and a synthetic coeval cluster with a realistic rotation distribution reproduces both the eMSTO morphology and the observed rotation–color relation [1904.03547]. Cordoni et al. further argued that the apparent age spread in Galactic open clusters increases with cluster age following the same relation as in Magellanic Cloud clusters, which they describe as demonstrating that rotation is the responsible mechanism [1811.01192]. The 2024 review summarizes the broader literature by stating that, among contemporary models and hypotheses, stellar rotation has been demonstrated as the most plausible mechanism to explain the occurrence of eMSs and eMSTOs [2401.08062].

A persistent difficulty is that rotation-only models do not always recover the full morphology. For NGC 1987 and NGC 2249, Monte Carlo simulations with Geneva SYCLIST isochrones confirm that a distribution of stellar rotation velocities yields an MSTO extent proportional to cluster age, but under the assumption of random viewing angles stellar rotation accounts for only $\sim 60\%$ and $\sim 40\%$ of the observed FWHM widths, respectively; a combination of stellar rotation velocities and stellar ages fits the observed eMSTO morphologies very well [1707.03847]. This suggests that rotation is often necessary, but not always sufficient, in intermediate-age clusters.

## 5. Additional mechanisms, age dependence, and cluster-to-cluster diversity

Very young clusters sharpen the limits of the simplest rotation picture. In $h$ and $\chi$ Persei, the observed color spread of the upper main sequence and eMSTO is far larger than photometric errors, yet non-emission-line B-type stars show no statistically significant correlation between color and $v\sin i$, in contrast with older clusters. Be stars, however, lie preferentially on the red side of the eMSTO, and the authors conclude that variable stars, binary interactions, and stellar rotation affect the eMSTO morphology of these very young clusters [1904.02005].

Internal mixing independent of surface rotation has been proposed as another route. Yang and Tian showed that varying the convective-core overshooting parameter $\delta_{\rm ov}$ among coeval stars can reproduce observed eMSTOs, and that the equivalent age spreads caused by OVCC are related to cluster age in good agreement with many observations; NGC 1856 is specifically reproduced by coeval populations with different $\delta_{\rm ov}$. At the same time, they stress that OVCC cannot result in a significant split of the main sequence of young star clusters at $m_U\lesssim 21$, whereas rapid rotation can make the split more significant [1701.05963]. Johnston et al. generalized the mixing perspective with isochrone-clouds calibrated by field-star asteroseismology and found that enhanced and varied internal mixing can explain a good fraction of the observed eMSTOs of NGC 1850 and NGC 884 in terms of one coeval population with a spread in core masses [1910.00591].

Circumstellar matter and mass loss have recently entered the discussion. In NGC 1783, UV-dim stars on the red side of the ultraviolet eMSTO were modeled as stars obscured by dust in rings formed at the periphery of excretion discs expelled during the Be stage; the authors argue that the entire eMSTO can be populated by dusty stars, with the reddest UV-dim stars representing the tail of maximum obscuration and line-of-sight alignment [2303.16049]. A later study of NGC 419 and NGC 1817 reached a related but distinct conclusion: synthetic clusters with weak rotational mixing provide the best fit to both the eMSTO and extended-red-clump features, whereas strong rotational mixing makes post-main-sequence stars too bright; a simple test further suggests that self-extinction by decretion discs in equator-on fast rotators could influence inferred rotation distributions and help reconcile projected-velocity discrepancies across the eMSTO [2509.13601].

## 6. Astrophysical implications and unresolved problems

The existence of eMSTOs has transformed the interpretation of cluster CMDs. In the Galactic open-cluster context, the term “multiple populations” often denotes populations with different rotation rates, and perhaps modest age differences or apparent age differences due to rotation, rather than the strong chemical anomalies characteristic of classical globular clusters [1811.01192]. This broadens the concept of internal complexity in star clusters and suggests that photometric multiplicity need not imply the same underlying physics in all age and mass regimes.

Cluster-scale structural thresholds remain controversial. On one side, low-mass Large Magellanic Cloud clusters show that mass is not the controlling factor and that a large core radius is not required for an eMSTO [1603.06891]. On the other, Goudfrooij et al. argued that eMSTOs in intermediate-age Magellanic Cloud clusters correlate with early escape velocity and that a threshold of $12$–$15~\mathrm{km\,s^{-1}}$ is relevant for gas retention [1410.3840]. A further empirical constraint comes from the main-sequence kink: in NGC 1831 the kink at $1.45 \pm 0.02\,M_\odot$ was interpreted as a lower limit to the mass below which the effects of rotation on the energy output become negligible at the metallicity of these clusters, linking the age dependence of eMSTO morphology to the onset of magnetic braking and changing envelope structure [1807.04737]. This suggests that both cluster-scale dynamics and stellar-scale physics must be considered simultaneously.

Several problems remain open. The recent review literature emphasizes stellar rotation, but the detailed literature also contains strong age-spread arguments based on red-clump morphology, subgiant-branch widths, and escape-velocity correlations, as well as alternative explanations based on overshooting, asteroseismic mixing constraints, variability, binaries, mass loss, and circumstellar dust [2401.08062]. A plausible implication is that eMSTOs are not produced by a single mechanism in every cluster. Future progress is explicitly tied in the literature to cluster asteroseismology, improved rotational-evolution calculations, direct measurements of $v\sin i$ and surface abundances, and multiwavelength constraints on decretion discs and dust, all aimed at mapping CMD morphology back to internal stellar physics rather than treating the eMSTO purely as an apparent star-formation history [1910.00591].

Source: https://www.emergentmind.com/topics/extended-main-sequence-turn-offs-emstos