Main-Sequence Turnoff

The main-sequence turnoff (often abbreviated MSTO) is the sharp knee or bluest inflection point on the Hertzsprung-Russell diagram or observational color–magnitude diagram (CMD) of a coeval stellar population—such as a globular cluster, an open cluster, or a dwarf spheroidal galaxy—where stars are currently exhausting hydrogen fuel (X_c → 0) at their centers and evolving off the core-hydrogen-burning main sequence onto the subgiant branch. Because a star's main-sequence nuclear lifetime scales inversely with a steep power of its initial mass (t_MS ∝ M / L ∝ M^−2.5 to M^−3.5), the luminosity and effective temperature of the main-sequence turnoff decrease monotonically as a stellar population ages, making the turnoff the most direct and model-reliable fundamental clock in observational astronomy.

First applied quantitatively by Allan Sandage, Martin Schwarzschild, and Fred Hoyle in 1952–1955 to date Galactic globular clusters such as Messier 13 and M3, theoretical isochrone fitting to the main-sequence turnoff established that the oldest halo globular clusters formed 12.0 to 13.5 billion years ago (z > 4), providing an independent astrophysical lower bound on the age of the universe consistent with cosmic microwave background measurements.

Nuclear Timescale and the Mass–Luminosity Relation

During its residence on the main sequence, a star maintains hydrostatic and thermal equilibrium by fusing four protons into a helium-4 nucleus (4 ¹H → ⁴He + 2 e⁺ + 2 ν_e + 26.73 MeV), releasing Q ≈ 6.3 × 10¹⁴ J per kilogram of hydrogen consumed. A star of total mass M can fuse only the hydrogen fraction f_core ≈ 0.10 to 0.15 residing inside its hot central core (T_c ≥ 10⁷ K) before reaching the Schönberg-Chandrasekhar limit or becoming stabilized by electron degeneracy. Consequently, the main-sequence nuclear lifetime t_MS is governed by the ratio of available core nuclear energy to bolometric luminosity L:

t_MS ≈ ( 0.007 f_core X M c² ) / L ≈ 10¹⁰ yr × ( M / M_⊙ ) × ( L / L_⊙ )⁻¹

Combining this relation with the empirical mass–luminosity scaling along the lower and intermediate main sequence (L / L_⊙ ≈ (M / M_⊙)^3.5 for 0.8 M_⊙ ≤ M ≤ 10 M_⊙) yields the canonical turnoff age scaling:

t_age = t_MS(M_TO) ≈ 10¹⁰ yr × ( M_TO / M_⊙ )^−2.5 ≈ 10¹⁰ yr × ( L_TO / L_⊙ )^−0.71

At any snapshot in time t_age after an instantaneous burst of star formation:

• Stars more massive than the turnoff mass (M > M_TO) have already completed core hydrogen burning, ascended the red giant and asymptotic giant branches, or terminated as compact remnants (white dwarfs, neutron stars, or black holes).
• Stars right at the turnoff mass (M ≈ M_TO) are exhausting central hydrogen (X_c < 0.01), igniting a thick hydrogen-burning shell around an inert helium core, and cooling toward the red subgiant branch.
• Stars less massive than the turnoff (M < M_TO) remain on the unevolved main sequence. (Apparent exceptions lying blueward and brighter than the MSTO are blue stragglers—stars rejuvenated by binary mass transfer or direct stellar collisions in dense cluster cores.)

The geometric shape of the main-sequence turnoff on an observational color–magnitude diagram changes systematically around a critical turnoff mass of M_TO ≈ 1.2 to 1.3 M_⊙ (corresponding to cluster ages of ~3–5 Gyr), reflecting the transition between two nuclear energy generation regimes:

Low-Mass Radiative Cores (M_TO < 1.2 M_⊙: Smooth Rounded Turnoff)

In old globular clusters (t_age ≈ 11–13.5 Gyr, M_TO ≈ 0.8 M_⊙), core hydrogen fusion proceeds via the proton–proton (p–p) chain, whose nuclear energy generation rate has a gentle temperature dependence (ε_pp ∝ ρ X² T⁴). Because the central radiative flux is modest, the core remains stable against convection. Hydrogen is depleted gradually from the very center outward, creating a smooth compositional gradient of mean molecular weight μ(r). As X_c approaches zero, hydrogen fusion transitions smoothly from the center to an adjacent thick shell without any sudden core collapse, producing a continuous, smoothly rounded turnoff arc into the subgiant branch.

Intermediate- and High-Mass Convective Cores (M_TO > 1.3 M_⊙: The Turnoff "Hook")

In younger open clusters (t_age < 3 Gyr, M_TO > 1.3 M_⊙), central temperatures exceed 1.7 × 10⁷ K and hydrogen burning is dominated by the catalytic CNO cycle, whose energy generation rate depends steeply on temperature (ε_CNO ∝ ρ X Z_CNO T^16–18). The intense concentration of luminosity at the very center drives a turbulent convective core that keeps the inner 10%–25% of the star chemically homogeneous.

When hydrogen is exhausted at the center of a convective core, it is exhausted simultaneously throughout the entire convective core volume. Deprived briefly of nuclear energy generation, the entire star undergoes a rapid overall Kelvin–Helmholtz gravitational contraction lasting ~1%–3% of t_MS until the shell immediately outside the exhausted core heats up enough to ignite hydrogen shell burning. During this brief contraction phase, both luminosity L and effective temperature T_eff rise sharply upward and blueward before reversing to the red, carving a characteristic zigzag "blueward hook" at the main-sequence turnoff of clusters like M67.

Absolute vs. Relative Cluster Chronometry

In theoretical stellar evolution codes (such as MESA, Dartmouth, PARSEC, and BaSTI), an isochrone is the locus in the (M_V, B − V) or (L, T_eff) plane occupied at a single age t by stars of identical initial chemical composition (X, Y, Z) spanning a continuum of initial masses. The most sensitive age diagnostic along any isochrone is the bolometric luminosity (or absolute magnitude M_V,TO) of the main-sequence turnoff: near t = 12 Gyr, a shift of ΔM_V,TO = +0.10 mag corresponds to a ~9%–10% (+1.1 Gyr) increase in derived cluster age.

Because measuring M_V,TO directly requires knowing the cluster distance modulus (m − M)_V and interstellar extinction E(B − V) (often constrained via Strömgren photometry or Gaia trigonometric parallaxes), astronomers employ two complementary relative age methods:

1. The Vertical Method (ΔV_TO^HB): Measures the magnitude difference between the main-sequence turnoff and the luminosity-stable horizontal branch (M_V,HB ≈ +0.5 to +0.8 mag, set by the constant ~0.47 M_⊙ core mass at the helium flash), eliminating distance and reddening uncertainties.
2. The Horizontal Method (Δ(B − V)_TO–RGB): Measures the color separation between the bluest point of the turnoff and the base of the Red Giant Branch along the Hayashi track, which shrinks monotonically with increasing age.

Sensitivity to Helium (Y), α-Elements, Atomic Diffusion, and Core Overshooting

Precision turnoff chronometry at the ≤5% level requires accounting for four microphysical processes in stellar interiors:

• Heavy-Element Metallicity ([Fe/H]) and α-Enhancement ([α/Fe]): Higher abundances of iron-peak elements and CNO/α-elements (O, Ne, Mg, Si) increase bound-free radiative opacity κ and CNO catalytic efficiency, shifting the turnoff to cooler temperatures and fainter magnitudes at fixed age.
• Helium Mass Fraction (Y): In globular clusters exhibiting multiple stellar populations (enriched in helium up to Y ≈ 0.30–0.38 by prior generations of asymptotic giant branch or fast-rotating massive stars), higher helium raises the mean molecular weight μ, increasing luminosity (L ∝ μ^7.5) and shifting the turnoff to fainter magnitudes at fixed age.
• Gravitational Settling (Atomic Diffusion): Over gigayear timescales in low-mass radiative cores, helium and heavy elements slowly sink toward the center relative to hydrogen under gravity, raising central μ and accelerating hydrogen exhaustion, which reduces derived globular cluster ages by ~0.8–1.2 Gyr (~8%) compared to non-diffusive models.
• Convective Core Overshooting: In stars with M_TO > 1.2 M_⊙, convective plumes possess finite momentum at the Schwarzschild convective boundary and penetrate a distance d_ov ≈ 0.1–0.2 H_P into the overlying radiative zone, mixing fresh hydrogen into the core and extending t_MS by 15%–25%.

1. Sandage, A. R., & Schwarzschild, M. (1952). "Inhomogeneous Stellar Models. II. Models with Exhausted Cores in Gravitational Contraction." The Astrophysical Journal, 116, 463–476. https://doi.org/10.1086/145632
2. VandenBerg, D. A., Bolte, M., & Stetson, P. B. (1996). "The Age of the Galactic Globular Cluster System." Annual Review of Astronomy and Astrophysics, 34(1), 461–510. https://doi.org/10.1146/annurev.astro.34.1.461
3. Chaboyer, B., Demarque, P., Kernan, P. J., & Krauss, L. M. (1998). "The Age of Globular Clusters in Light of Hipparcos: Resolving the Age Problem?" The Astrophysical Journal, 494(1), 96–110. https://doi.org/10.1086/305201
4. Dotter, A., Chaboyer, B., Jevremović, D., Kostov, V., Baron, E., & Ferguson, J. W. (2008). "The Dartmouth Stellar Evolution Database." The Astrophysical Journal Supplement Series, 178(1), 89–101. https://doi.org/10.1086/589654