Stars, From Birth to Remnant Part 2: Mechanical Equilibrium and the Equation of State

Stars: From Birth to Remnant

Chapter 2: Mechanical Equilibrium and the Equation of State

Introduction

In Chapter 1, we looked at the star from the outside. Flux, luminosity, magnitude, parallax—all of them concerned either the light reaching us from the star or its position in the sky. The star's interior was still a black box. In this chapter, we open that box.

A star is a sphere of gas that does not collapse under its own gravity. What prevents the collapse is a pressure gradient that increases inward; the central idea really is that simple. But once we write it as an equation and begin asking, “what produces this pressure?”, the matter becomes more serious. Pressure in stellar matter comes from three distinct sources: the thermal collisions of particles (ideal-gas pressure), the push exerted by the sea of photons (radiation pressure), and the quantum-mechanical resistance created by the Pauli principle, which prevents electrons from being squeezed together into the same states (degeneracy pressure). Each dominates under different conditions, each rests on a different law of physics, and each takes center stage at some point in a star's life.

We will proceed as follows. First, we will derive the two fundamental differential equations of stellar structure: mass conservation and hydrostatic equilibrium. We will then calculate the three sources of pressure one by one, deriving the degenerate-electron pressure step by step from Fermi–Dirac statistics in particular, because it will be indispensable later for white dwarfs and neutron stars. We will combine all three to write the equation of state (EOS). Next, from statistical mechanics, we will derive the Saha equation, which answers the question “which atom ionizes, and when?” The chapter will conclude with two powerful general tools: the virial theorem, which tells us why a star has “negative heat capacity,” and the characteristic timescales that determine which physics dominates at different stages of a star's life.

The goal is this: by the end of the chapter, we will have a complete mathematical framework describing how stellar matter behaves. Once we add energy transport in Chapter 3, we will have assembled all the equations needed to solve for a star's internal structure.

1 The Mass-Conservation Equation

The first equation is the easiest: it follows directly from the star's geometry and contains no physics.

Derivation (Mass conservation)

Let us assume that the star is spherically symmetric, an excellent approximation for an isolated, non-rotating star whose magnetic field is not dominant. The density $\rho$ is a function only of the radius $r$.

Since the volume of the thin spherical shell between $r$ and $r + dr$ is $4\pi r^2\,dr$, its mass is
$dm = 4\pi r^2 \rho\, dr.$

Here $m(r)$ denotes the total mass from the center out to radius $r$—the “mass inside $r$.” Thus $m(0) = 0$ and $m(R) = M$ (the star's total mass). The differential ratio above takes the simple form
$\boxed{\frac{dm}{dr} = 4\pi r^2 \rho}$

Physical Interpretation

This equation is not physics but bookkeeping. The mass of a spherical shell equals its surface area times its thickness times its density; the equation is the one-line expression of that fact. The word “conservation” is somewhat grandiose: what it really says is that, if mass is neither created nor destroyed, $m(r)$ must accumulate solely as the integral of the density. Even during nuclear burning in stellar evolution, the mass defect changes the total mass by less than one part in a million; for practical purposes, “mass is conserved.”

1.1 Lagrangian coordinates: $m$ as the independent variable

It is often more robust to write the structure equations with $m$, rather than $r$, as the independent variable. The reason is this: over its lifetime, a star can change its radius dramatically (its radius increases by a factor of one hundred as it becomes a red giant). But the mass coordinate measured from the center identifies which shell we are talking about and remains unchanged throughout the evolution. A Lagrangian view (following fixed $m$), rather than an Eulerian view (tracking a fixed $r$), is far more natural in evolutionary calculations.

Derivation (Lagrangian form)

Let us invert the equation $dm/dr = 4\pi r^2 \rho$:
$\boxed{\frac{dr}{dm} = \frac{1}{4\pi r^2 \rho}}$

This time $r$ is the dependent variable (it tells us where the shell sits), while the mass coordinate $m$ is the independent variable. As it runs from $m = 0$ to $m = M$, the radius extends from the center to the surface.

Side Note

Choosing between Eulerian and Lagrangian coordinates is not, by itself, a physical decision; nevertheless, numerical stellar-solution codes almost invariably prefer the Lagrangian description. That is because the composition of a shell—for example, its “hydrogen mass fraction $X$”—is tied closely to the mass in that shell, not to $r$. As hydrogen burns, $X$ decreases; but tracking this by asking “what is $X$ at $r = 10^8$ m?” is a headache, because the shell may have swollen and expanded. Asking “what is $X$ at $m = 0{,}3M$?” is much cleaner.

2 Hydrostatic Equilibrium

The second equation is the first true equation of stellar physics: at every point, the inward pull of gravity and the outward push of pressure cancel each other. We promised it in Chapter 1; now we deliver.

Derivation (Hydrostatic equilibrium)

We again work under the assumption of spherical symmetry. Consider a thin spherical shell between $r$ and $r+dr$. Two forces act on each unit area of this shell:

Gravity. All the matter between the shell's inner and outer surfaces is pulled toward the center by the enclosed mass $m(r)$. The gravitational force per unit area on the shell is
$f_{\text{grav}} = -g(r) \times (\text{birim alan başına kütle}) = -\frac{Gm(r)}{r^2}\rho\, dr.$
The minus sign tells us that the force points in the direction of decreasing $r$—inward.

Pressure. The shell's lower surface is under pressure $P(r)$, while its upper surface is under pressure $P(r+dr)$. Taking upward as positive, the net pressure force per unit area is
$f_{\text{bas}} = P(r) - P(r + dr) = -dP = -\frac{dP}{dr}\, dr.$
If the pressure is greater inside and smaller outside, then $dP/dr < 0$, so $-dP/dr > 0$ and the net force points outward.

Equilibrium. If the shell is neither collapsing nor expanding, the total force is zero:
$-\frac{dP}{dr}\, dr - \frac{Gm\rho}{r^2}\, dr = 0.$
Dividing the equality by $dr$ and rearranging gives
$\boxed{\frac{dP}{dr} = -\frac{Gm(r)\,\rho(r)}{r^2}}$
This is the equation of hydrostatic equilibrium.

Physical Interpretation

What this equation says is extremely simple, but its consequences are profound. Every layer in a star must support the pile of material above it. “Supporting the pile” means that the pressure in that layer must be high enough; pressure must therefore increase inward. The minus sign says exactly this: as $r$ increases—as we move outward—$P$ decreases.

If, at some point, the pressure gradient cannot overcome gravity, the shell there accelerates inward and collapse begins. Conversely, if the gradient exceeds gravity, the shell expands outward. In an ideal star, exact balance holds everywhere, hence the term “hydrostatic” (static = stationary). Except during very brief events such as supernova collapse and the helium flash, hydrostatic equilibrium holds throughout a star's life.

2.1 Lagrangian form

From the chain rule:

Derivation (Lagrangian form)

$\frac{dP}{dm} = \frac{dP}{dr}\cdot\frac{dr}{dm} = \left(-\frac{Gm\rho}{r^2}\right)\cdot\left(\frac{1}{4\pi r^2\rho}\right).$
The factors of $\rho$ cancel:
$\boxed{\frac{dP}{dm} = -\frac{Gm}{4\pi r^4}}$
Notice that $\rho$ does not appear explicitly in this form; the dependence on density is contained implicitly through $r(m)$. To see the density, we must return to the first form. But the equation itself looks clean.

2.2 A rough estimate of the central pressure

To solve exactly for a star's central pressure, all the structure equations and the EOS must be solved together. Hydrostatic equilibrium alone, however, gives a rough estimate.

Derivation (Rough central pressure)

Let us integrate the hydrostatic equation from $r = 0$ to $r = R$, taking $P(R) \approx 0$ (small surface pressure) as the boundary condition:
$P_c = P(0) - P(R) = \int_0^R \frac{Gm(r)\,\rho(r)}{r^2}\, dr.$

To estimate the right-hand side, assume the star has a uniform density $\bar\rho = 3M/(4\pi R^3)$. Then $m(r) = (4\pi/3)\, r^3\bar\rho$, and the integrand is
$\frac{G\cdot(4\pi/3)r^3\bar\rho\cdot\bar\rho}{r^2} = \frac{4\pi G\bar\rho^2}{3}\, r.$
Integrating from $0$ to $R$:
$P_c \approx \frac{4\pi G\bar\rho^2}{3}\cdot\frac{R^2}{2} = \frac{2\pi G\bar\rho^2 R^2}{3}.$
Writing $\bar\rho$ in terms of $M$ and $R$ and simplifying:
$P_c \approx \frac{2\pi G}{3}\left(\frac{3M}{4\pi R^3}\right)^2 R^2 = \frac{3GM^2}{8\pi R^4} \approx 0{,}12\cdot\frac{GM^2}{R^4}.$

Example

For the Sun, using $M_\odot = 2{,}0 \times 10^{30}$ kg and $R_\odot = 7{,}0 \times 10^8$ m,
$P_c^{(\text{kaba})} \approx 0{,}12\cdot\frac{(6{,}67\times10^{-11})(2{,}0\times10^{30})^2}{(7{,}0\times10^8)^4} \approx 1{,}3\times10^{14}\ \text{Pa}.$
The actual value given by the standard solar model is approximately $2{,}5\times10^{16}$ Pa. Our estimate is therefore two orders of magnitude too low. The reason is clear: the actual density rises rapidly toward the center; the uniform-density assumption badly underestimates the true weight of the matter in the central region. Even so, obtaining the correct order of magnitude matters here: the relation $P_c \propto GM^2/R^4$ provides a general scaling law among stars, while the correct prefactor depends on the star's internal density profile.

Side Note

The reason the “rough” estimate is so inaccurate is a useful pedagogical warning. At the center of a star, the density is roughly 100 times the mean density. Because gravity is sensitive to $\rho^2$, this difference enters the central pressure as a factor of ten thousand. In other words, our estimate really says, “this would be the central pressure if the density were distributed uniformly”; a real star is far more centrally concentrated.

3 The Three Sources of Pressure

The hydrostatic-equilibrium equation contains $dP/dr$; to solve it, we must ask, “what is $P$?” Pressure in stellar matter has three distinct physical origins. In this section, we derive each one in turn.

3.1 Ideal-gas pressure

In most stars, the interior matter behaves like a classical gas: free particles moving at Boltzmann-distributed speeds, colliding and scattering. In this regime, the pressure is given by the ideal-gas law.

Derivation (Ideal-gas equation)

From the kinetic theory of gases, the pressure for an isotropic particle distribution is
$P = \frac{1}{3} n \langle p \cdot v\rangle$
where $n$ is the particle number density, $p$ the momentum, and $v$ the velocity. For classical (non-relativistic) particles, $p = mv$ and $p\cdot v = mv^2 = 2\cdot\tfrac{1}{2}mv^2$, so the momentum–velocity product is twice the kinetic energy. Therefore:
$P = \frac{2}{3} n \langle E_{\text{kin}}\rangle.$
For a Maxwell–Boltzmann distribution, $\langle E_{\text{kin}}\rangle = \tfrac{3}{2}kT$, so
$P = \frac{2}{3} n \cdot \frac{3}{2} kT = nkT.$
This is the familiar equation of state of an ideal gas.

In a stellar context, we want to convert particle number density into mass density. If the mean particle mass is $\bar m = \mu m_u$ ($\mu$ is the mean molecular weight and $m_u$ the atomic mass unit), then $n = \rho/(\mu m_u)$, and
$\boxed{P_{\text{gas}} = \frac{\rho}{\mu m_u}\, kT}$

Derivation (Mean molecular weight $\mu$)

$\mu$ depends on the composition and ionization state of the stellar matter. In a star's interior, the matter is fully ionized; every atom contributes both its nucleus and all its electrons as free particles.

Let the composition be specified by mass fractions: $X$ (hydrogen), $Y$ (helium), and $Z$ (all remaining “metals”), with $X + Y + Z = 1$. In the fully ionized state:

• Hydrogen ($A = 1$): each atom contributes 1 proton + 1 electron = 2 particles, with a mass of 1 unit.
• Helium ($A = 4$): 1 nucleus + 2 electrons = 3 particles, with a mass of 4 units.
• Typical metal ($A \approx 2Z,\ Z \gg 1$): 1 nucleus + $Z$ electrons $\approx Z + 1$ particles, with a mass of $A \approx 2Z$ units; the particle-to-unit ratio is $\approx 1/2$.

The total number of particles per unit mass is
$\frac{1}{\mu} = \underbrace{X\cdot\frac{2}{1}}_{H} + \underbrace{Y\cdot\frac{3}{4}}_{He} + \underbrace{Z\cdot\frac{1}{2}}_{\text{metaller}} = 2X + \frac{3Y}{4} + \frac{Z}{2}.$
For the solar composition $(X, Y, Z) \approx (0{,}71,\ 0{,}27,\ 0{,}02)$:
$\frac{1}{\mu_\odot} \approx 2(0{,}71) + 0{,}75(0{,}27) + 0{,}5(0{,}02) \approx 1{,}63 \implies \mu_\odot \approx 0{,}61.$
For pure, fully ionized hydrogen, $\mu = 1/2$: the proton and electron together give an average of half a unit per particle.

Physical Interpretation

The term “mean molecular weight” is misleading: this number does not measure the mass of molecules, but the mean mass per free particle in the plasma. Electrons count toward the particle number in an ionized gas, whereas a molecular mass does not count them separately. In stellar interiors, $\mu$ is typically between $0{,}5$ and $0{,}7$; roughly speaking, this means “two particles per unit mass.” Take care in calculations: treating the same material as a neutral atomic gas gives $\mu \approx 1{,}3$, a very different number.

3.2 Radiation pressure

The temperature in a stellar interior is so high that the photons themselves also exert pressure. In most main-sequence stars this pressure is a small correction, but in very massive stars it becomes dominant and sets the limit to the star's existence—the Eddington limit. Its derivation begins by recalling that photons carry momentum.

Derivation (Pressure and energy density for a photon gas)

We wrote the pressure of an isotropic particle distribution as $P = \tfrac{1}{3} n\langle p \cdot v\rangle$ (see 3.1). For a photon, $p = E/c$ and $v = c$, so $p \cdot v = E$. Thus
$P_{\text{ışın}} = \frac{1}{3} n \langle E\rangle.$
But $n\langle E\rangle = u$, the total energy density per unit volume. Therefore
$P_{\text{ışın}} = \frac{u}{3}$
Under the assumption of isotropy—which is exact locally to a good approximation in a stellar interior, where the photons are close to thermal equilibrium—this result is exact. All that remains is to relate $u$ to temperature.

Derivation (Blackbody energy density)

For a blackbody—an ideal absorber containing photons in thermal equilibrium at every frequency—the integral of the energy per unit volume obtained from the Planck function is
$u(T) = \int_0^\infty \frac{8\pi h\nu^3/c^3}{e^{h\nu/kT} - 1}\, d\nu = \frac{8\pi^5 k^4}{15\, c^3 h^3}\, T^4 \equiv aT^4.$
Here
$a \equiv \frac{8\pi^5 k^4}{15\, c^3 h^3} \approx 7{,}57 \times 10^{-16}\ \text{J m}^{-3}\,\text{K}^{-4}$
is the radiation constant. Its relation to the Stefan–Boltzmann constant $\sigma$ is $a = 4\sigma/c$ (derived in the second part of Chapter 1).

Derivation (Radiation pressure)

Combining the two results gives
$\boxed{P_{\text{ışın}} = \frac{1}{3} aT^4}$

Physical Interpretation
Radiation pressure differs from ideal-gas pressure in two fundamental ways. First, it depends entirely on temperature; density does not appear at all. This is because the number of photons in thermal equilibrium depends only on temperature (the amount of gas in the star is irrelevant). Second, it grows very rapidly, as $T^4$. Ideal-gas pressure is $\propto \rho T$, so their ratio scales as
$\frac{P_{\text{ışın}}}{P_{\text{gas}}} \propto \frac{T^3}{\rho}.$
At the center of the Sun, this ratio is $\sim 10^{-3}$, making radiation pressure a small correction. But at the center of a $50M_\odot$ star, the temperature exceeds $10^8$ K, the density falls, and the ratio approaches 1. At that point, the star's existence hangs by a thread: the outward push of its radiation begins to overcome its own gravity. The “Eddington limit” is the quantitative expression of this threshold; it will be derived in Chapter 6.

3.3 Degenerate-electron-gas pressure

There are two places in a star where matter does not behave like a classical ideal gas: during advanced stages of evolution—in a white-dwarf core, a red-giant core, or a neutron star—the density grows so high that the classical picture of electrons “colliding and scattering” breaks down and the constraint imposed by the Pauli principle takes over. Two electrons cannot occupy the same quantum state, so at high density the electrons fill the low-energy states and are forced into states of progressively higher momentum. This quantum-mechanical momentum produces a pressure that is independent of temperature.

This pressure has one especially important property: cooling does not extinguish it. In a classical gas, pressure falls when the temperature falls; in a degenerate electron gas, however, the electrons must remain in momentum states even if the temperature falls all the way to absolute zero, because the Pauli principle does not allow them to be compressed into lower states. This is why a white dwarf does not collapse even after cooling for billions of years without producing nuclear energy. Let us derive this step by step.

Derivation (Density of states in phase space)

The smallest phase-space volume allowed by the Heisenberg principle for a single particle is $h^3$: $\Delta x\,\Delta p \gtrsim h$ in all three dimensions. This means that only one momentum state can fit into a phase-space volume of size $h^3$. Electrons, being spin-1/2 particles, can occupy each state in pairs (spin up + spin down), so the density of states in momentum space is
$g(p)\, dp = \frac{2 \cdot 4\pi p^2}{h^3}\, dp = \frac{8\pi p^2}{h^3}\, dp.$
This is the number of available states per unit volume with momentum in the interval $[p, p + dp]$. To obtain the physical electron density, it must be multiplied by an occupation function specifying how fully these states are occupied.

The occupation from Fermi–Dirac statistics is
$f(p) = \frac{1}{\exp\left(\dfrac{E(p) - \mu}{kT}\right) + 1}.$
Here $\mu$ is the chemical potential. In the classical limit it becomes strongly negative, making $f$ very small (the Maxwell–Boltzmann limit); in the fully degenerate limit it is positive and equals the Fermi energy $E_F$. The latter is the limit of interest here.

Derivation (Fully degenerate limit: the Fermi sphere)
“Fully degenerate” means that the temperature is sufficiently low—or the density sufficiently high—that $kT \ll E_F$. Then $f(p)$ becomes a step function:
$f(p) = \begin{cases} 1, & p \le p_F \\ 0, & p > p_F \end{cases}$
Thus every momentum state below the Fermi momentum $p_F$ is occupied, while every state above it is empty. In momentum space, the electrons fill a sphere centered at the origin and having radius $p_F$: the Fermi sphere.

The number of electrons per unit volume is the number of states filled by this sphere:
$n_e = \int_0^{p_F} \frac{8\pi p^2}{h^3}\, dp = \frac{8\pi}{3h^3}\, p_F^3.$
Conversely, expressing $p_F$ in terms of $n_e$:
$p_F = h\left(\frac{3n_e}{8\pi}\right)^{1/3} = \hbar\left(3\pi^2 n_e\right)^{1/3}.$
The second expression follows by using $\hbar = h/(2\pi)$ and is the standard form.

Derivation (Pressure of a fully degenerate electron gas)

Let us write the classical kinetic-theory formula $P = \tfrac{1}{3} n\langle p \cdot v\rangle$ in continuous form, using distribution integrals:
$P_e = \frac{1}{3}\int_0^{p_F} \frac{8\pi p^2}{h^3}(p \cdot v)\, dp = \frac{8\pi}{3h^3}\int_0^{p_F} p^3\, v(p)\, dp.$
Here $v(p)$ is the velocity of an electron with momentum $p$; its form depends on whether the electron is relativistic. We consider two limits.

Non-relativistic (NR) limit. If the electron speed is much less than the speed of light ($p \ll m_e c$), the classical relation $v = p/m_e$ applies.

Derivation (NR degeneracy pressure)

$P_e^{\text{NR}} = \frac{8\pi}{3h^3 m_e}\int_0^{p_F} p^4\, dp = \frac{8\pi}{3h^3 m_e}\cdot\frac{p_F^5}{5} = \frac{8\pi p_F^5}{15 h^3 m_e}.$
Now substitute $p_F = \hbar(3\pi^2 n_e)^{1/3}$: $p_F^5 = \hbar^5(3\pi^2)^{5/3} n_e^{5/3}$ and $h^3 = (2\pi\hbar)^3 = 8\pi^3\hbar^3$.
$P_e^{\text{NR}} = \frac{8\pi}{15 \cdot 8\pi^3 \hbar^3 m_e}\cdot \hbar^5 (3\pi^2)^{5/3} n_e^{5/3} = \frac{\hbar^2}{15\pi^2 m_e}(3\pi^2)^{5/3} n_e^{5/3}.$
Using $(3\pi^2)^{5/3}/\pi^2 = 3^{5/3}\pi^{4/3}$ and $15 = 5 \cdot 3$, we can simplify the prefactor; the most commonly encountered closed form is
$\boxed{P_e^{\text{NR}} = \frac{(3\pi^2)^{2/3}}{5}\frac{\hbar^2}{m_e} n_e^{5/3}}$

Ultra-relativistic (UR) limit. If the density rises until the Fermi momentum satisfies $p_F \gg m_e c$, the electrons move near the speed of light ($v \approx c$) and $E = pc$.

Derivation (UR degeneracy pressure)

$P_e^{\text{UR}} = \frac{8\pi c}{3h^3}\int_0^{p_F} p^3\, dp = \frac{8\pi c}{3h^3}\cdot\frac{p_F^4}{4} = \frac{2\pi c\, p_F^4}{3h^3}.$
Making the same substitution, $p_F^4 = \hbar^4(3\pi^2)^{4/3} n_e^{4/3}$ and $h^3 = 8\pi^3\hbar^3$:
$P_e^{\text{UR}} = \frac{2\pi c\hbar^4}{3 \cdot 8\pi^3 \hbar^3}(3\pi^2)^{4/3} n_e^{4/3} = \frac{c\hbar}{12\pi^2}(3\pi^2)^{4/3} n_e^{4/3}.$
Since $(3\pi^2)^{4/3}/\pi^2 = 3^{4/3}\pi^{2/3}$ and $3^{4/3}/12 = 3^{1/3}/4$, the simplest form is
$\boxed{P_e^{\text{UR}} = \frac{(3\pi^2)^{1/3}}{4}\hbar c\, n_e^{4/3}}$

Physical Interpretation

The two limits have different density dependences, and this difference is the fundamental reason white dwarfs have a Chandrasekhar mass limit. In the NR limit, $P \propto n_e^{5/3}$; as density increases, pressure rises even faster, so resistance against gravity is always available. In the UR limit, $P \propto n_e^{4/3}$; pressure now grows more slowly with density—or, more precisely, with the same exponent as the version of gravity that also scales as $\rho^{4/3}$ under polytropic conditions. Once the two are matched, the pressure gradient can no longer balance gravity at every density: the star is either stable or unstable, with nothing in between. This delicate balance produces the Chandrasekhar mass $\approx 1{,}4M_\odot$ (fully derived in Chapter 13).

Derivation (Conversion to mass density)

In a stellar model, we want the pressure in terms of $\rho$ rather than $n_e$. In a fully ionized plasma, the electron density is related to the mass density by
$n_e = \frac{\rho}{\mu_e m_u},$
where $\mu_e$ is the mean molecular weight per electron. For typical metals with $A \approx 2Z$, $\mu_e \approx 2$; for pure hydrogen, $\mu_e = 1$. Substituting and writing the pressure in terms of $\rho$:
$P_e^{\text{NR}} = K_{\text{NR}}\left(\frac{\rho}{\mu_e}\right)^{5/3}, \qquad P_e^{\text{UR}} = K_{\text{UR}}\left(\frac{\rho}{\mu_e}\right)^{4/3},$
$K_{\text{NR}} = \frac{(3\pi^2)^{2/3}}{5}\frac{\hbar^2}{m_e\, m_u^{5/3}}, \qquad K_{\text{UR}} = \frac{(3\pi^2)^{1/3}}{4}\frac{\hbar c}{m_u^{4/3}}.$

Side Note

There is an objective surprise here: degenerate-electron pressure is completely independent of temperature. This is expected, because the Pauli principle forces electrons into high-momentum states without requiring any temperature. But the result comes at a price: degenerate matter is insensitive to temperature. If heat increases in a white dwarf, the pressure does not respond (to a good approximation); in a classical gas, by contrast, increased heat means increased pressure and expansion. Degenerate matter cannot expand in that way. The result is that, if nuclear burning begins in a degenerate region, the heat rises but the matter does not expand and cool. The temperature rises further, the burning accelerates, and the heat rises still more. This is a thermal-runaway response, the mechanism at the heart of the helium flash and Type Ia supernovae. We will lay out the full picture in Chapters 8 and 12.

4 Equation of State (EOS): Putting Everything Together

In a typical stellar interior, two pressure sources coexist: ideal-gas pressure and radiation pressure. Degeneracy pressure becomes important only in special stages and regions (white dwarfs and the advanced cores of giant stars). To write a fully general EOS, we must add them together.

The general form is

$$P_{\text{toplam}} = P_{\text{gas}} + P_{\text{ışın}} + P_{\text{dejenere}}.$$

In main-sequence stars, degeneracy can be neglected, and the EOS takes the practical form

$$\boxed{P = \frac{\rho kT}{\mu m_u} + \frac{1}{3} aT^4.}$$

A single dimensionless parameter is enough to track the relative contributions of the two pressure sources in stellar modeling.

Definition (The $\beta$ parameter)

The fraction of the total pressure contributed by gas pressure is
$\beta \equiv \frac{P_{\text{gas}}}{P_{\text{toplam}}}, \qquad 1 - \beta = \frac{P_{\text{ışın}}}{P_{\text{toplam}}}.$
$\beta = 1$ is pure gas pressure, $\beta = 0$ pure radiation pressure; $0 < \beta < 1$ is a mixture.

Physical Interpretation

$\beta$ carries an extraordinary amount of information about a star's character. In low-mass main-sequence stars (the Sun and below), $\beta \approx 1$ and radiation pressure is negligible. Around $20M_\odot$, $\beta \approx 0{,}8$; radiation supplies 20 percent of the pressure. In stars of $50M_\odot$ and above, $\beta \lesssim 0{,}5$. The limit $\beta \to 0$ is precisely the Eddington limit: the mass threshold beyond which a star cannot remain bound against the outward thrust of its own radiation. The Eddington limit is the upper bound on stellar mass, and the $\beta$ parameter is its numerical harbinger.

5 Ionization Equilibrium: The Saha Equation

So far, we have assumed stellar matter to be either “fully ionized” (in the interior) or “neutral” (in the atmosphere). The actual transition region—below the stellar photosphere, at the base of the atmosphere—is partially ionized; there we need an equation that tells us what fraction of an atom is ionized. This is the equation that Meghnad Saha derived in 1920, revolutionizing the interpretation of stellar spectra.

Our aim is to relate the ionized fraction of hydrogen atoms in a region—for simplicity, we consider only one element—to the temperature, density, and ionization energy of hydrogen.

5.1 Preparation: the individual components

The system is in thermal and chemical equilibrium. There are three “chemical species”: the neutral hydrogen atom ($H$), the proton ($p$), and the electron ($e$). The reaction between them is

$$H \rightleftharpoons p + e^-$$

and at equilibrium it proceeds at the same rate in both directions. In thermodynamic language, the condition for this equilibrium is that the chemical potentials satisfy

$$\mu_H = \mu_p + \mu_e.$$

In the ideal-gas limit, the chemical potential of each species has a familiar form. For one type of particle (of mass $m$, including its internal energy levels), classical statistical mechanics gives

$$\mu = -kT\ln\left[\frac{z}{n}\right], \qquad z = \left(\frac{2\pi mkT}{h^2}\right)^{3/2} Z_{\text{iç}},$$

where $n$ is the particle number density, $z$ is the single-particle quantum concentration, and $Z_{\text{iç}}$ is the partition function over the particle's internal states (electronic, rotational, and so on).

Derivation (Choosing the zero of energy)

We must choose the zero point of energy for the species carefully; if we neglect this, the ionization energy enters the equation with the wrong sign.

Zero = a free $p$ and a free $e$, both at rest and infinitely far apart. Under this choice:

• Free proton at rest: $E = 0$.
• Free electron at rest: $E = 0$.
• Ground state of hydrogen (1s): the electron is bound around the proton, with binding energy $\chi_H = 13{,}6$ eV. Thus the atom's energy is $E_H = -\chi_H$.

The internal partition function of hydrogen then receives a contribution only from the ground state (we neglect excited levels):
$Z_{\text{iç}}^H = g_H \exp\left(-\frac{E_H}{kT}\right) = g_H \exp\left(\frac{\chi_H}{kT}\right).$
The binding energy appears with a positive exponent because the bound state lies below the reference level. For the proton and electron, the internal states come only from spin:
$Z_{\text{iç}}^p = g_p = 2, \qquad Z_{\text{iç}}^e = g_e = 2.$
For the hydrogen ground state, the electron spin states (2) multiply the proton spin states (2):
$g_H = 4.$

5.2 Derivation of the Saha equation

Derivation (The Saha equation)

Write the three chemical potentials in the form $\mu = -kT\ln(z/n)$ and insert them into the equilibrium condition:
$-kT\ln\frac{z_H}{n_H} = -kT\ln\frac{z_p}{n_p} - kT\ln\frac{z_e}{n_e}.$
Canceling $kT$ and combining the logarithms:
$\ln\frac{n_p n_e}{n_H} = \ln\frac{z_p z_e}{z_H} \implies \frac{n_p n_e}{n_H} = \frac{z_p z_e}{z_H}.$
Now substitute the expressions for $z$. Since $m_p \approx m_H$ in the translational part, their ratio is extremely close to 1; in the end, only the electron's translational factor remains:
$\frac{z_p z_e}{z_H} = \left(\frac{2\pi m_e kT}{h^2}\right)^{3/2}\cdot\frac{g_p g_e}{g_H \exp(\chi_H/kT)}.$
The ratio of internal statistical weights is $g_p g_e/g_H = 2\cdot2/4 = 1$. Therefore
$\boxed{\frac{n_p\, n_e}{n_H} = \left(\frac{2\pi m_e kT}{h^2}\right)^{3/2}\exp\left(-\frac{\chi_H}{kT}\right)}$
This is the simplest form of the Saha equation for hydrogen.

Physical Interpretation

The structure of the equation can be summarized in words: the left-hand side is the ratio of the “ionized” population to the “neutral” population; the right-hand side contains two multiplicative factors. The exponential part is the expected Boltzmann factor: if $\chi_H/kT$ is large, ionization is energetically suppressed. But the equation also contains a prefactor, $(2\pi m_e kT/h^2)^{3/2}$, which is the “quantum-mechanically available phase-space volume for a free electron.” As the temperature increases, not only does the Boltzmann factor surge; the phase space open to the electron also grows. This is why hydrogen becomes ionized not at the $\approx 160{,}000$ K suggested by the naive Boltzmann estimate $kT \approx \chi_H$, but at much lower temperatures, $\approx 6000$–$10{,}000$ K. Ionization begins before the phase-space cost has been paid in full.

Example

To develop a numerical feel, let us evaluate the right-hand side of the Saha equation for the solar photosphere ($T \approx 5800$ K):
$\left(\frac{2\pi m_e kT}{h^2}\right)^{3/2} \approx 2{,}4\times10^{21}\cdot T^{3/2}\ \text{m}^{-3} \approx 2{,}4\times10^{21}\cdot(5800)^{1{,}5}\ \text{m}^{-3} \approx 1{,}1\times10^{27}\ \text{m}^{-3}.$
$\chi_H/kT = 13{,}6/(8{,}6\times10^{-5}\cdot5800) \approx 27{,}2$, so $\exp(-27{,}2) \approx 1{,}5\times10^{-12}$. The net value of the right-hand side is $\approx 1{,}7\times10^{15}$ m$^{-3}$. For the ionization fraction $x \equiv n_p/(n_p + n_H)$, let us rearrange the left-hand side:
$\frac{x^2}{1 - x}\cdot(n_p + n_H) = 1{,}7\times10^{15}\ \text{m}^{-3}.$
Taking a typical total hydrogen density in the photosphere of $\approx 10^{22}$ m$^{-3}$, we obtain $x^2/(1-x) \approx 1{,}7\times10^{-7}$ and hence $x \approx 4\times10^{-4}$. Only four parts in ten thousand of the hydrogen in the solar photosphere are ionized; the rest is neutral. The “help from phase space” is important, but 5800 K is still not enough for complete ionization.

5.3 General form: an arbitrary ion

We can generalize the Saha equation to any ionization stage. For equilibrium between the $i$-times-ionized and $i+1$-times-ionized states:

$$\frac{n_{i+1}\, n_e}{n_i} = \frac{2U_{i+1}}{U_i}\left(\frac{2\pi m_e kT}{h^2}\right)^{3/2}\exp\left(-\frac{\chi_i}{kT}\right).$$

Here $U_i$ is the internal partition function of the $i$-times-ionized state (including excited electronic levels), and $\chi_i$ is the ionization energy required to pass from that state to the next. The leading factor of “2” comes from the spin states of the free electron. For hydrogen, taking $U_p/U_H = 1/2$ recovers the special form above.

Side Note

The Saha equation revolutionized the interpretation of stellar spectra. Without it, one cannot solve the puzzle of the OBAFGKM sequence—why hydrogen lines are weak in hot stars, still weak in cool stars, and very strong in between (in class A). The answer is that hydrogen lines depend not on how much neutral hydrogen there is, but on whether excited neutral hydrogen is present. That, in turn, is the product of both the degree of ionization and the Boltzmann excitation fraction; the competition between the two peaks in class A ($\approx 10{,}000$ K). We will return to these details in the second part of Chapter 1.

6 The Virial Theorem

The closest thing in stellar physics to a single formula that “explains everything” is the virial theorem. For any self-gravitating system in hydrostatic equilibrium, there is a fixed relation—independent of the system's details—between its total thermal energy and its total gravitational potential energy. The consequences are surprising; “negative heat capacity,” in particular, is one of the strangest yet deepest results in stellar physics.

Derivation (The virial theorem from hydrostatic equilibrium)

Multiply both sides of the hydrostatic-equilibrium equation by $4\pi r^3\, dr$ and integrate from $0$ to $R$:
$\int_0^R 4\pi r^3\frac{dP}{dr}\, dr = -\int_0^R 4\pi r^3\cdot\frac{Gm\rho}{r^2}\, dr.$

Right-hand side. Simplifying,
$-\int_0^R Gm\cdot 4\pi r\rho\, dr = -\int_0^R \frac{Gm}{r}\cdot 4\pi r^2\rho\, dr = -\int_0^M \frac{Gm}{r(m)}\, dm \equiv \Omega.$
This integral is the definition of the gravitational potential energy: the sum of the gravitational binding energy of each shell due to the mass $m$ inside it (with a minus sign because the system is bound):
$\Omega = -\int_0^M \frac{Gm}{r}\, dm < 0.$

Left-hand side. Integration by parts:
$\int_0^R 4\pi r^3\frac{dP}{dr}\, dr = \left[4\pi r^3 P\right]_0^R - \int_0^R 12\pi r^2 P\, dr.$
Boundary terms: at $r = 0$, $r^3 = 0$; at $r = R$, the surface pressure $P(R) \approx 0$. The boundary term vanishes. What remains is
$\int_0^R 4\pi r^3\frac{dP}{dr}\, dr = -3\int_0^R P\cdot 4\pi r^2\, dr = -3\int_V P\, dV.$
Combining both sides gives
$-3\int_V P\, dV = \Omega \iff 3\int_V P\, dV + \Omega = 0.$
This is the general virial expression for a star in hydrostatic equilibrium.

Derivation (For an ideal gas: $2K + \Omega = 0$)

For an ideal gas, there is a simple relation between pressure and thermal kinetic energy per unit volume. For a classical (non-relativistic) gas:
$P = nkT = \frac{2}{3}\cdot n\cdot\frac{3}{2}kT = \frac{2}{3}u_{\text{kin}},$
because each particle carries $\tfrac{3}{2}kT$ of thermal kinetic energy. Thus
$\int_V P\, dV = \frac{2}{3}\int_V u_{\text{kin}}\, dV = \frac{2}{3}K,$
where $K \equiv \int u_{\text{kin}}\, dV$ is the star's total thermal kinetic energy. Substituting:
$3\cdot\frac{2}{3}K + \Omega = 0 \iff \boxed{2K + \Omega = 0}$
This is the celebrated form of the virial theorem.

Derivation (Total energy and $E = -K = \Omega/2$)

The star's total energy is thermal kinetic plus gravitational potential energy:
$E = K + \Omega.$
From the virial theorem, $\Omega = -2K$, hence
$E = K + (-2K) = -K, \qquad \text{eşdeğer biçimde}\quad E = \frac{\Omega}{2}.$

6.1 Negative heat capacity

We now arrive at the most counterintuitive consequence the virial theorem gives to stellar physics.

Physical Interpretation

A star loses energy through radiation; its total energy $E$ decreases over time ($\Delta E < 0$). Since $E = -K$,
$\Delta K = -\Delta E > 0.$
The star heats up as it loses energy. Its thermal kinetic energy increases; its temperature rises. This is the exact opposite of ordinary heat capacity: an object generally cools when it loses heat. A star heats up when it loses heat. This behavior is called negative heat capacity. The reason is neither nuclear nor magical; it is pure gravitational mechanics. As the star loses energy, it contracts; contraction releases some gravitational potential energy; the virial theorem requires half of that energy to be converted into heat and the other half to leave as radiation.

The practical consequences of this result dominate the chapters ahead:

• Protostar phase (Kelvin–Helmholtz). As a gas cloud contracts while radiating heat, its temperature rises. This rise continues until the temperature required to initiate nuclear burning is reached. The very radiation trying to cool the star is what ignites it; that is the irony.
• Main-sequence stability. If a star increases its nuclear-energy production slightly, some heat escapes outward, the core expands and cools, and the nuclear rate falls again. Negative heat capacity plays a thermostatic role here, making the star appear steady.
• Degenerate region. Where pressure is insensitive to temperature (a white-dwarf core or a red-giant core), the conditions of the virial theorem are violated. If nuclear burning begins in a degenerate region, the heat rises, but because the matter cannot expand, the temperature runs out of control: this is the mechanical origin of the helium flash and Type Ia supernovae (Chapters 8 and 12).

Side Note

Negative heat capacity may seem to contradict thermodynamic theorems that apply to an independent object. There is no contradiction: the usual definition of heat capacity is for closed systems independent of gravity. A star is not closed (it radiates), nor is it independent of gravity. Once gravity enters, “heat” and “mechanical work” acquire a different relationship, reversing the sign of the ordinary rule. The same phenomenon reappears in black-hole thermodynamics (Chapter 15).

7 Characteristic Timescales

A star is governed by three different physical processes: mechanical equilibrium, thermal adjustment, and nuclear burning. All three occur on different timescales, and the enormous separation among them lies at the very heart of what makes stellar physics mathematically tractable.

7.1 Dynamical timescale $\tau_{\text{dyn}}$

If a star lost its mechanical equilibrium—if, for example, its pressure vanished instantaneously—how long would it take to collapse under its own gravity?

Derivation (Free-fall time)

Neglect pressure completely and model the star as a sphere of gas with radius $R$, mass $M$, and uniform density $\bar\rho = 3M/(4\pi R^3)$. A particle at the surface is attracted by all the enclosed mass, so its motion is governed by $\ddot r = -GM/r^2$. The energy integral gives the exact time for the radius to fall to zero:
$\tau_{ff} = \frac{\pi}{2}\sqrt{\frac{R^3}{2GM}} = \sqrt{\frac{3\pi}{32\, G\bar\rho}}$
Discarding the numerical prefactor and retaining only the order of magnitude:
$\boxed{\tau_{\text{dyn}} \sim \sqrt{\frac{R^3}{GM}} \sim \frac{1}{\sqrt{G\bar\rho}}}$

Example

For the Sun, $R^3/(GM) \approx 2{,}6\times10^6$ s$^2$, so $\tau_{\text{dyn}}^\odot \approx 1600$ s $\approx 27$ minutes. If the star could not re-establish mechanical equilibrium within an hour, visibly dramatic events would result. When a star explodes or a neutron star forms, core collapse does indeed occur on this order of magnitude (seconds).

7.2 Thermal (Kelvin–Helmholtz) timescale $\tau_{\text{KH}}$

If the star lost all its nuclear sources, how long would it take to radiate its existing gravitational binding energy at its present luminosity?

Derivation (Kelvin–Helmholtz time)

From the virial theorem, the star's total energy is $E = \Omega/2$. In order of magnitude, $|\Omega| \sim GM^2/R$, and therefore $|E| \sim GM^2/(2R)$. If the star radiates this energy at luminosity $L$, the characteristic time is
$\boxed{\tau_{\text{KH}} \sim \frac{GM^2}{RL}}$

Example

For the Sun, $GM^2/R \approx 4\times10^{41}$ J and $L_\odot \approx 3{,}8\times10^{26}$ W, so $\tau_{\text{KH}}^\odot \approx 10^{15}$ s $\approx 3\times10^7$ years. Nineteenth-century physicists Kelvin and Helmholtz argued that the Sun could be no older than this, because gravitational contraction was the only energy source then known. Geology pointed to more than ten times that age; the inconsistency was resolved only with the discovery of nuclear energy in the 1930s.

7.3 Nuclear timescale $\tau_{\text{nuc}}$

When a star converts hydrogen into helium, it transforms approximately 0.7% of the mass—the mass defect—into energy. For how long can the star draw on this source while it has usable hydrogen available?

Derivation (Nuclear timescale)

Let the star have total mass $M$, let the fraction present as hydrogen in the core be $f_H \sim 0{,}1$ (only the core burns; the number depends on the stellar model), and let the energy efficiency of converting hydrogen to helium be $\eta = 0{,}007$. The total nuclear reserve is
$E_{\text{nuc}} = \eta\, f_H\, M c^2.$
If this energy is released at the star's present luminosity,
$\boxed{\tau_{\text{nuc}} \sim \frac{\eta\, f_H\, M c^2}{L}}$

Example

For the Sun, $\eta f_H M_\odot c^2 = 0{,}007\cdot0{,}1\cdot(2\times10^{30})(3\times10^8)^2 \approx 1{,}3\times10^{44}$ J. With $L_\odot = 3{,}8\times10^{26}$ W, $\tau_{\text{nuc}}^\odot \approx 3{,}4\times10^{17}$ s $\approx 10^{10}$ years. This is the Sun's estimated main-sequence lifetime; the Sun is now approximately $4{,}6\times10^9$ years old and is midway through its main-sequence life.

7.4 The enormous separation among the three timescales

Numerically:

$$\tau_{\text{dyn}}^\odot \sim 10^3\ \text{s}, \qquad \tau_{\text{KH}}^\odot \sim 10^{15}\ \text{s}, \qquad \tau_{\text{nuc}}^\odot \sim 10^{17}\ \text{s}.$$

Thus $\tau_{\text{dyn}} \ll \tau_{\text{KH}} \ll \tau_{\text{nuc}}$. In order of magnitude, $\tau_{\text{KH}}/\tau_{\text{dyn}} \sim 10^{12}$ and $\tau_{\text{nuc}}/\tau_{\text{KH}} \sim 10^2$.

Physical Interpretation

This enormous separation is the basis of why stellar modeling is possible. A star establishes mechanical equilibrium within $\tau_{\text{dyn}} \sim$ hours; no slower process can disrupt that balance quickly enough. It adjusts thermally over $\tau_{\text{KH}} \sim$ millions of years; nuclear evolution is very slow by comparison. Nuclear evolution unfolds over $\tau_{\text{nuc}} \sim$ billions of years.

Conclusion:

When calculating stellar evolution, we may assume that the star is in mechanical equilibrium at every instant; the resulting error is roughly of order $(\tau_{\text{dyn}}/\tau_{\text{KH}})^2 \sim 10^{-24}$. Likewise, we may average thermal equilibrium over nuclear timescales. This justifies numerical stellar-evolution codes solving an equilibrium model at each timestep under the “quasi-static” assumption and then advancing to the next. Most classical results in stellar physics rest on this assumption, and their accuracy comes from the great separation among the three timescales.

There are exceptions. During supernova core collapse, hydrostatic equilibrium fails abruptly and $\tau_{\text{dyn}}$ dominates: everything is over within seconds. In the helium flash, degenerate matter cannot make its mechanical response ($\tau_{\text{dyn}}$) keep pace with its thermal evolution; the temperature runs away on a timescale shorter than $\tau_{\text{dyn}}$. These exceptions are the dramatic events we will encounter later.

Next Chapter

So far, we have derived the two fundamental differential equations describing the mechanical equilibrium (hydrostatic equilibrium) and equation of state (the three sources of pressure + EOS) of stellar matter. With the Saha equation, we established ionization equilibrium; with the virial theorem, the thermal–mechanical relation. One more equation is needed to complete the picture of stellar structure: how is energy transported? How does heat produced in a star's core reach the surface—by radiation, by convection, or, in special cases, by conduction? This is the subject of Chapter 3. We will derive the equation for each of the three modes of transport, lay out the Schwarzschild and Ledoux criteria that determine which mode dominates under which conditions, and finally bring together the four fundamental structure equations (mass, hydrostatic equilibrium, energy generation, and energy transport) to close the mathematical problem of stellar internal structure.