Fermi-Dirac distribution

The Fermi-Dirac distribution follows from Fermi-Dirac statistics and is $n(E)=[e^{\beta (E-\mu )}+1{]}^{-1}$.

The mean occupation of a fermionic one-particle state of energy $E$ isAt zero temperature this becomes the step function $1_{E<E_F}$.

The Fermi window measures sensitivity of occupation to energy near the chemical potential. It is positive, even about $\mu$, has width of order $k_{B}T$, and integrates to one over the full energy line. Its centred variance is ${\pi }^2(k_{B}T{)}^2/3$. A band restricted to nonnegative energy contains weight $n_F(0,\mu )$ instead of exactly one; extending to the full line is exponentially accurate when $\mu \gg k_{B}T$. At zero temperature the window tends to a Dirac delta function at the chemical potential.

A density of states $g(ϵ)=g_0[1+\alpha \cos (2\pi ϵ/\Delta )]$ has zero-temperature oscillatory grand potential $g_0\alpha {\Delta }^2\cos (2\pi \mu /\Delta )/(4{\pi }^2)$. Its Fermi-window convolution multiplies that oscillation by $R_T$ when the lower band edge is far below the chemical potential. The Fourier transform of the even window removes its sine component and gives the title factor. At high $X$, $R_T\sim 2X{e}^{-X}$; at small $X$, $R_T=1-X^2/6+O(X^4)$. The corresponding thermal change relative to zero temperature has amplitude proportional to $1-R_T$, rather than $R_T$.

For the sinusoidal density-of-states model, divide the oscillatory amplitude of $\Phi (T,\mu )-\Phi (0,\mu )$ by its smooth magnitude $g_0{\pi }^2(k_{B}T{)}^2/6$. This gives the title ratio. Its limit as $T\to 0$ is $|\alpha |$, although both changes vanish at exactly zero temperature. The expansion is $|\alpha |[1-7X^2/60+O(X^4)]$, and its large-$X$ envelope is $6|\alpha |/X^2$. A fixed chemical potential adds the actual cosine phase, which can make the oscillatory contribution zero even when the envelope is nonzero. Exponential suppression of the full oscillatory potential therefore must not be confused with this algebraic thermal-change ratio.

When $E-\mu \gg kT$, the unit term in the denominator of the Fermi-Dirac distribution is negligible. The resulting exponentially small occupation is the Boltzmann approximation. It justifies the usual nondegenerate semiconductor carrier formulas; it fails near strongly occupied band states.

The Sommerfeld expansion approximates integrals weighted by the Fermi-Dirac distribution when $k_{B}T\ll \mu$ and $f$ is smooth near $\mu$. Applying it to the number integral at fixed particle number gives $\mu -E_F=-({\pi }^2/6)(k_{B}T{)}^2{g}^{\prime }(E_F)/g(E_F)+\cdots$. The sign depends on the slope of the density of states. For an unbounded parabolic band, $g(E)\propto E^{d/2-1}$, yielding opposite quadratic shifts in one and three dimensions. In two dimensions the leading algebraic shift vanishes, but the finite band-bottom cutoff still gives an exponentially small negative shift.

For a three-dimensional parabolic band and fixed positive $\mu -E_c$, the Sommerfeld expansion gives $n(T)=n(0)[1+{\pi }^2(k_{B}T/(\mu -E_c){)}^2/8+O(T^4)]$. This increase differs from the usual fixed-density calculation, in which the chemical potential decreases with temperature. The exact density is proportional to $T^{3/2}{F}_{1/2}[(\mu -E_c)/(k_{B}T)]$.

The temperature derivative of the Fermi occupation has opposite signs below and above $\mu$. Pairing energies $\mu \pm x$ leaves a positive contribution because the density of states increases with energy. The unpaired higher-energy tail is also positive. Hence the density increases at fixed $\mu$ for an unbounded three-dimensional parabolic band.

Thermal electron-positron pairs in chemical equilibrium with photons have opposite chemical potentials when the one-particle energy includes rest energy. For negligible degeneracy and $k_{B}T\ll m_e{c}^2$, the zero-chemical-potential density per charge species is $n_0=2[m_e{k}_{B}T/(2\pi {\hslash }^2){]}^{3/2}{e}^{-m_e{c}^2/(k_{B}T)}$. In that nondegenerate limit, charge neutrality and the classical equilibrium product $n_{-}{n}_{+}=n_0^2$ give $n_{+}=(\sqrt{n_b^2+4n_0^2}-n_b)/2$, with net electron density $n_b$. Pairs are important when $n_0\gtrsim n_b$. At arbitrary temperature the zero-chemical-potential marker uses the integral $n_0=({\pi }^2{\hslash }^3{)}^{-1}{\int }_0^{\infty }{p}^2[\exp (\sqrt{m_e^2{c}^4+p^2{c}^2}/k_{B}T)+1{]}^{-1}dp$. Its ultrarelativistic limit is $3\zeta (3)(k_{B}T/\hslash c{)}^3/(2{\pi }^2)$. A nonzero net chemical potential shifts the exact abundance boundary; the zero-potential curve is an approximate crossover marker.

For $s>0$,This follows from $(e^x+1{)}^{-1}=(e^x-1{)}^{-1}-2(e^{2x}-1{)}^{-1}$. In three spatial dimensions it gives the fermion-to-boson ratios $3/4$ for ultrarelativistic particle number and $7/8$ for energy density.

A Fermi gas consists of noninteracting fermions whose one-particle states are occupied according to the Fermi-Dirac distribution.

For a three-dimensional ideal gas with quadratic dispersion and fixed particle number, the total density of states at the Fermi energy is $g(E_F)=3N/(2E_F)$. The Sommerfeld coefficient then gives $C_V=({\pi }^2/2)N{k}_B(T/T_F)$ to leading order, where $T_F=E_F/k_B$. The fixed-number energy correction is $\Delta U=({\pi }^2/4)N(k_{B}T{)}^2/E_F$. Only states near the Fermi surface change occupation, explaining the small active fraction.

A model with separate filled proton and neutron momentum seas in a nuclear mean field. Each spin-half species has two spin states, so counting momentum-space cells gives $p_{F,i}=\hslash (3{\pi }^2{n}_i{)}^{1/3}$. It describes typical nucleon motion and a kinetic contribution to nuclear symmetry energy, but binding requires additional interactions.

The motion of particles occupying the momentum sea of a Fermi gas, especially nucleons in an atomic nucleus. It persists at zero temperature and broadens scattering kinematics relative to a stationary-nucleon approximation.

An electron gas confined in one spatial direction, with motion available in the other two. Its electron number per unit area is $n_s=N/A$. The ideal Fermi gas approximation neglects interactions between electrons; an actual two-dimensional electron gas can have important interactions. Confinement alone does not require a magnetic field or Landau levels.

State counting with total degeneracy $g$ gives a constant two-dimensional density of states $gm/(2\pi {\hslash }^2)$ above a parabolic conduction band edge. Integrating the Boltzmann occupation yields the displayed areal carrier density. For one valley and two spin states the coefficient is $m{k}_{B}T/(\pi {\hslash }^2)$. Applying the same count to holes gives $np=N_c{N}_v{e}^{-E_g/(k_{B}T)}$ and the intrinsic density $n_i=\sqrt{N_c{N}_v}{e}^{-E_g/(2k_{B}T)}\propto T{e}^{-E_g/(2k_{B}T)}$, distinct from the three-dimensional $T^{3/2}$ prefactor.

A gate separated by dielectric thickness $d$ has geometric capacitance per area $C/A=ϵ/d$ in the parallel-plate approximation. An electron-density change is $\Delta n_s=(ϵ/ed)\Delta V_g$ with the stated polarity/reference convention. Fringing, trapped charge, depletion and finite quantum capacitance can modify the relation. An absolute gate voltage additionally requires the zero-density offset or another calibration.

A magnetic field curves ballistic carrier trajectories between an injector and a collector. For openings separated along a specular edge by $L$, peaks occur near $L=2j{r}_c$ with $r_c=\hslash k_F/(e|B|)$, giving $|B_j|=2j\hslash k_F/(eL)$. Field polarity and carrier sign select the connection direction. Electron phase coherence can produce fine structure but is not required for the basic semiclassical focusing geometry.

In a parabolic band, transport mobility and mean free path obey $\ell =\mu \hslash k_F/e$. The rough condition $\ell \gtrsim L$ gives $\mu \gtrsim eL/(\hslash k_F)$, whereas surviving a first semicircle uses path length $\pi L/2$. Clear peaks usually need a larger margin and a specular boundary, so this is an observability scale rather than a universal sharp threshold.

A short electrostatically defined constriction connecting two electron reservoirs. Transverse confinement produces transport modes, and sufficiently narrow contacts can collimate carriers for electron-focusing experiments. Its quantum conductance depends on mode transmission; an opening described only as narrower than a micrometre is not automatically in a single-mode limit.

A parallel field lifts spin degeneracy and separates threshold energies. The step size may change from $2e^2/h$ to $e^2/h$ when the two spin branches are resolved. Finite layer thickness can additionally alter orbital confinement.

At zero temperature, a rigid spin-degenerate transmission spectrum with symmetric source/drain bias has ${\mu }_{S,D}=E_F\pm eV/2$ and the displayed differential conductance. Half-integer plateaus in units $2e^2/h$ occur when the two bias-window edges sample different channel counts. The bias convention and electrostatic response matter.

In a rigid spectrum, source and drain threshold lines in a bias-versus-gate map satisfy $E_n(V_g)={\mu }_S$ or ${\mu }_D$. Where adjacent thresholds align with opposite edges of the bias window, their spacing equals the window width $e|V|$. For symmetric bias and spin degeneracy, neighboring channel counts at those edges give a differential electrical conductance halfway between integer $2e^2/h$ plateaus. The dc ratio $I/V$ is a different observable; barrier motion and heating can modify the interpretation.

A channel supporting ballistic electron transport carries incident populations set by different source and drain chemical potentials, rather than an everywhere-equilibrated local distribution with an Ohmic voltage gradient. Its electrostatic barrier responds self-consistently to contacts, gates and electrostatic screening. A symmetric rigid-barrier model places reservoir energies at $E_F\pm eV/2$ relative to the barrier. Asymmetric partition or bias-dependent transmission changes the weights in the differential electrical conductance, and must be included when extracting an energy scale.

For rigid spin-degenerate channel transmissions and ${\mu }_{S,D}=E_F\pm eV/2$, differentiating $I=(2e/h){\sum }_n{\int }_{{\mu }_D}^{{\mu }_S}{T}_n(E)dE$ gives $G_d=(e^2/h){\sum }_n[T_n({\mu }_S)+T_n({\mu }_D)]$. If the two edges sample neighboring integer channel counts, $G_d$ is a half-integer multiple of $2e^2/h$. Explicit bias dependence of the potential adds terms; the apparent fraction does not imply fractional carriers.

A locally separable potential with an unstable longitudinal maximum and stable transverse confinement. Mode $n$ sees barrier top $E_n=U_0+\hslash {\omega }_y(n+1/2)$ and transmission $T_n(E)=[1+\exp (-2\pi (E-E_n)/(\hslash {\omega }_x)){]}^{-1}$. It is a single-particle smooth-barrier model; interactions and bias-induced reshaping require additions.

Slow widening of a constriction approximately conserves the transverse mode action, reducing transverse momentum as the width $W$ increases. A lowering band bottom also increases longitudinal kinetic energy. Together they can narrow the angular distribution of an emitted electron beam. An abrupt narrow aperture alone can instead diffract strongly.

An injector and collector aligned across an open two-dimensional electron gas. A direction-sensitive signal between them probes ballistic electron transport. Contact resistance and diffuse current backgrounds must be separated from the direct beam contribution.

Two surface electrodes whose depletion regions define a narrow conducting opening in an underlying two-dimensional electron gas. Gate voltage changes the transverse confinement and channel transmission without physically etching the conducting channel to its electrical width.

An electron Fermi gas whose occupation is governed mainly by the Pauli exclusion principle, with thermal energy small compared with its kinetic Fermi energy. At zero temperature its momentum states are filled up to $p_F$. In the ultrarelativistic limit $E/V=3nc{p}_F/4\propto n^{4/3}$, with $n=4\pi g_s{p}_F^3/(3h^3)$. This supplies the pressure supporting a white dwarf until the relativistic mass limit is approached.

Neglecting interactions and thermal pairs, the Fermi-Dirac distribution $f(p)=[\exp ((\epsilon (p)-{\mu }_{\mathrm{kin}})/(k_{B}T))+1{]}^{-1}$ with kinetic energy $\epsilon (p)=\sqrt{m_e^2{c}^4+p^2{c}^2}-m_e{c}^2$ determines the electron contribution to a stellar equation of state. The kinetic chemical potential is fixed by the electron number density, not an independently chosen pressure fraction:The electron kinetic internal energy per volume is $({\pi }^2{\hslash }^3{)}^{-1}{\int }_0^{\infty }{p}^2\epsilon (p)f(p)dp$. In the nondegenerate limit $P_e=n_e{k}_{B}T$; at zero temperature the integrals become the equation of state of a cold electron gas. These are limits of one component and must not be added together as independent electron pressures. When electron-positron thermal pair abundance matters, both charge species must be included, with charge neutrality fixing their net density.

For density of states $AV{ϵ}^a$, $a>-1$, the grand canonical ensemble has $\log \mathcal{Z}=V{T}^{a+1}f(\mu /T)$. Strict monotonicity of Fermi gas entropy per particle makes $\mu /T$ constant at fixed entropy and particle number. Therefore $V{T}^{a+1}$ and $P{V}^{(a+2)/(a+1)}$ are constant along a reversible adiabatic process.

For density of states proportional to ${ϵ}^a$, $a>-1$, and positive temperature, set $q=\mu /(kT)$, $n(x)=(1+e^{x-q}{)}^{-1}$, $b=a+1$ and $M_j={\int }_0^{\infty }{x}^{a+j}n(1-n)dx$. Differentiation and integration by parts giveThe inequality is strict Cauchy-Schwarz inequality, since $x$ is not constant under this positive weight. Thus fixed entropy per particle forces fixed $\mu /T$. This is the injectivity step behind the adiabatic scaling of an ideal Fermi gas.

The Fermi energy is the highest occupied one-particle energy in the ground state of a noninteracting Fermi gas at zero temperature, measured here relative to the lowest one-particle energy. The Pauli exclusion principle allows only one fermion in each complete one-particle quantum state, including its spin, so adding more particles fills progressively higher energy eigenstates. A cold gas can therefore have substantial kinetic energy even when its temperature is zero. In the thermodynamic limit, with the same energy reference, $E_F=\mu (T=0)$ where $\mu$ is the chemical potential.^[ref]

For a uniform, three-dimensional gas of noninteracting electrons, with spin one-half, equal populations of the two spin states, no applied magnetic field, and nonrelativistic particles of mass $m=m_e$,Here $n$ is the total number density, $V$ the volume, $p_F$ the Fermi momentum, $\hslash =h/(2\pi )$ the reduced Planck constant, and $k_B$ the Boltzmann constant. The electron Fermi temperature $T_F$ expresses the same energy scale in kelvin; it is not the actual gas temperature. Strong degeneracy of a degenerate electron gas means $T\ll T_F$. The momentum formula counts states and survives changing the energy–momentum relation, whereas $E_F=p_F^2/(2m)$ requires $p_F\ll mc$.^[ref]

The reference matters: the displayed $E_F$ excludes rest energy and sets the zero at $p=0$. Including rest energy shifts both the one-particle energy and the zero-temperature chemical potential by $m{c}^2$, leaving occupations unchanged. More generally, adding a constant $C$ to every one-particle energy also adds $C$ to $\mu$; the Fermi-Dirac distribution depends only on their difference. In a finite quantum system the chemical potential at zero temperature can lie between the last occupied and first empty energy eigenstates; its equality with the filling edge is a continuum statement.

At positive temperature, $E_F$ remains the zero-temperature scale associated with the specified number density. The Fermi level $\mu (T)$ must instead satisfy the particle-number conditionwhere $g(E)$ is the extensive density of states, including both spin states. Smearing the Fermi-Dirac distribution while arbitrarily holding $\mu =E_F$ generally changes $N$. This distinction matters when interpreting electron heat capacity and any experiment that infers temperature from an occupation profile.^[ref]

The Fermi temperature expresses a Fermi energy as a temperature scale, $T_F=E_F/k_B$, where $k_B$ is the Boltzmann constant. In the convention $E_F>0$ measured from the relevant band minimum, $T\ll T_F$ is the regime of a strongly degenerate Fermi gas. For an interacting Fermi liquid, an effective Fermi scale can be introduced through its quasiparticle dispersion; its convention should be specified.

Choose $n=10^{28}{\mathrm{m}}^{-3}$ for a uniform noninteracting electron model with two equally populated spin states. This is an original worked example, not a measured number density assigned to an actual material. Using $m_e\simeq 9.10938\times 10^{-31}\mathrm{kg}$, $\hslash \simeq 1.05457\times 10^{-34}\mathrm{J}\mathrm{s}$, and $k_B=1.380649\times 10^{-23}\mathrm{J}{\mathrm{K}}^{-1}$ givesThe conversion uses $1\mathrm{eV}=1.602176634\times 10^{-19}\mathrm{J}$: an electronvolt is an energy unit, while a joule is its SI counterpart. The dimensions provide a quick check: $n^{1/3}$ is inverse length, $\hslash n^{1/3}$ has units of momentum, $p_F^2/m$ has units of energy, and $E_F/k_B$ has units of temperature. The ratio $p_F/(m_{e}c)\simeq 0.002574$ verifies the nonrelativistic limit for this model, where $c$ is the speed of light.

Increasing this chosen number density eightfold doubles the Fermi momentum and quadruples the Fermi energy, giving approximately $6.7701\mathrm{eV}$. The exponent $2/3$ has a geometric origin: the number of occupied states grows as the cube of a momentum radius, but kinetic energy of nonrelativistic particles grows as its square. This makes Fermi energy a useful bridge between a count of particles and a mechanical energy scale.

The kinetic electron Fermi temperature is $T_F=[\sqrt{m_e^2{c}^4+p_F^2{c}^2}-m_e{c}^2]/k_B$, where the Fermi momentum $p_F=\hslash (3{\pi }^2{n}_e{)}^{1/3}$ counts the net electrons. Strong degeneracy requires $T\ll T_F$. With mean molecular weight per electron ${\mu }_e$ and $n_e=\rho /({\mu }_e{m}_u)$, $T_F=(m_e{c}^2/k_B)[\sqrt{1+(\rho /{\rho }_{\ast }{)}^{2/3}}-1]$. Its nonrelativistic and ultrarelativistic limits scale as ${\rho }^{2/3}$ and ${\rho }^{1/3}$, respectively.

Fermi momentum is the magnitude of the momentum at the boundary of filled one-particle states in an isotropic zero-temperature Fermi gas. For the uniform three-dimensional two-component electron model, a short state count givesTo see the factors, impose periodic boundary conditions on a cubic box of side $L$, so $V=L^3$ and each momentum component is $p_i=2\pi \hslash j_i/L$ for an integer $j_i$. Each allowed momentum point occupies a cell of volume $(2\pi \hslash /L{)}^3$ in momentum space. Divide the sphere volume $4\pi p_F^3/3$ by this cell volume and multiply by two for the two spin states. Replacing the discrete count with a volume is the large-box approximation; surface and shell effects are neglected.^[ref]

The factor two is not automatic for every Fermi gas. If only one internal state is populated, its number density $n_{\sigma }$ instead satisfies $p_{F,\sigma }=\hslash (6{\pi }^2{n}_{\sigma }{)}^{1/3}$. For two equally populated states, $n_{\sigma }=n/2$ recovers the displayed result. A Fermi surface in a crystal can have a more complicated shape, so a single spherical Fermi momentum is a property of this isotropic model.

The group velocity of an excitation evaluated on the Fermi surface. For an isotropic dispersion relation $ϵ(p)$, $v_F=(dϵ/dp{)}_{p_F}$, where $p_F$ is the Fermi momentum.

The Fermi wavevector is the radius of the occupied sphere in wavevector coordinates for an isotropic Fermi gas. In the uniform two-component electron model, $k_F=(3{\pi }^{2}n{)}^{1/3}$ and $E_F={\hslash }^2{k}_F^2/(2m_e)$ for nonrelativistic particles. Its units are inverse length, whereas the Fermi momentum has units of momentum.

The wavelength associated with a specified Fermi wavevector. A local-density response normally requires variation on a scale large compared with $k_F^{-1}$; numerical factors such as $2\pi$ must be kept explicit when comparing a variation scale with the full wavelength.

The Fermi sea is the zero-temperature many-fermion quantum state with every one-particle state below the Fermi energy occupied and every state above it empty. For a uniform isotropic Fermi gas, the occupied region in momentum space is a solid sphere; its boundary is the Fermi surface. These are momenta, not positions: filling the sphere does not put the particles in a ball in ordinary space. Opposite momentum contributions cancel, so the gas at rest has zero total momentum even though its total kinetic energy is positive.

The filled states are forced upward in energy by the Pauli exclusion principle. This is different from the zero-point energy of one particle confined by a boundary. At fixed number density, the Fermi energy remains positive as the box grows to the thermodynamic limit, while the lowest box confinement energy tends to zero. Low-energy excitations move particles from occupied states near the Fermi surface into nearby empty states; deep states have no nearby available destination. This explains why only a small fraction of electrons contribute to the low-temperature electron heat capacity.^[ref]

For the uniform noninteracting electron model with two spin states and kinetic energy of nonrelativistic particles $E=p^2/(2m_e)$, count all available one-particle states up to an energy $E\ge 0$:This extensive density of states has units of inverse energy; $g(E)/V$ is the density of states per unit volume. Availability and occupation are different: $g(E)$ is positive above $E_F$, but at zero temperature those states are empty. The occupied energy distribution is $g(E)f(E)$, with the Fermi-Dirac distribution reduced to $f(E)=1$ below $E_F$ and $f(E)=0$ above it. The convention at the exact boundary does not affect a continuum integral.

With the dimensionless quantity $x=E/E_F$, the occupied count per $dx$, divided by the total particle number, is $3\sqrt{x}/2$ for $0<x<1$. Its area is one. Weighting this by $E$ gives the internal energy and its mean per particle:Thus the highest occupied energy and the mean energy differ. For the Fermi energy of a model electron gas, $U/N\simeq 1.0155\mathrm{eV}$. At fixed $N$, $E_F\propto V^{-2/3}$, so the derivative $P=-(\partial U/\partial V{)}_N$ gives the electron degeneracy pressure $P=2n{E}_F/5$. The same state-counting geometry therefore predicts a nonzero pressure at zero temperature.

In 2001, Andrew G. Truscott, Randall G. Hulet, and collaborators observed degeneracy pressure in trapped lithium. A Zeeman slower fed a dual magneto-optical trap; optical pumping prepared atoms for a cloverleaf apparatus for magnetic trapping of neutral atoms. Microwave evaporation removed energetic lithium-7 bosons; elastic collisions cooled lithium-6 fermions by sympathetic cooling.^[ref]

Sequential absorption imaging recorded both trapped clouds on a camera. Fits to axial profiles used the boson cloud as a thermometer. At common temperature, the nearly equal magnetic trapping potentials provided a comparison: the clouds had similar sizes near $T/T_F=1$, but the fermion cloud remained broader as cooling reached $T/T_F=0.25$. Figure 3 compared its radius with ideal Fermi gas and classical predictions.^[ref]

Varying the wait before imaging tested thermal equilibrium. Finite imaging resolution restricted the analysis to the axial direction; fits neglected interaction-induced density shifts. Cooling eventually suffered from the dwindling boson heat capacity. The result supports quantum occupation effects, without directly resolving individual occupied states or reaching zero temperature.^[ref]

These were neutral alkali atoms in a spatially varying trap. The Fermi energy of a model electron gas instead uses uniform electrons, two equally populated spin states, and the electron mass. The measured trapped-atom radii are not measurements of that electron model's $p_F$ or $E_F$; comparing their numerical Fermi energies requires the appropriate particle mass, internal-state populations, and confining potential.

For spin-one-half particles in three dimensions with dispersion $\epsilon =pc$, the density of states isAt zero temperature,

Degeneracy pressure is the pressure of a fermionic system that remains at zero temperature because the Pauli exclusion principle forces particles to occupy states with nonzero momentum.

Electron degeneracy pressure supports dense objects when electrons fill momentum states up to a Fermi surface. In the nonrelativistic limit, a fixed-mass self-gravitating object has the characteristic degenerate scaling $R\propto M^{-1/3}$.

For a nonrelativistic three-dimensional ideal electron gas at zero temperature, the pressure is $p=2n{E}_F/5$ and the electronic bulk modulus is $B=2n{E}_F/3$. This is the electronic model contribution, not automatically the total bulk modulus of a real crystal.

For a nonrelativistic zero-temperature free electron model with fixed particle number, electron degeneracy pressure obeys $P\propto (N/V{)}^{5/3}$. Consequently its bulk modulus is $K=-V(\partial P/\partial V{)}_N=5P/3=2n{E}_F/3$. At temperatures small compared with the electron Fermi temperature, the leading isothermal compressibility is $1/K$. This electronic contribution alone omits ionic binding and interaction corrections in a real metal.

For completely degenerate, nonrelativistic electrons in fully ionized hydrogen and helium with hydrogen mass fraction $X$, $n_e=(1+X)\rho /(2m_p)$. Filling the spin-degenerate momentum states to $p_F$ gives $n_e=8\pi p_F^3/(3h^3)$ and $P_e=(1/3){\int }_0^{p_F}(p^2/m_e)(8\pi p^2/h^3)dp=8\pi p_F^5/(15m_e{h}^3)$. Eliminating $p_F$ gives $P_e=K{\rho }^{5/3}$, whereIt is a polytropic index $3/2$ relation. Complete degeneracy suppresses thermal dependence, while the nonrelativistic assumption requires $p_F\ll m_{e}c$.

The condition $p_F=m_{e}c$ gives ${\rho }_{\ast }=[{\mu }_e{m}_u/(3{\pi }^2)](m_{e}c/\hslash {)}^3\simeq 9.74\times 10^5{\mu }_e\mathrm{g}\mathrm{c}{\mathrm{m}}^{-3}$. Degenerate electrons are nonrelativistic for $\rho \ll {\rho }_{\ast }$ and ultrarelativistic for $\rho \gg {\rho }_{\ast }$. This density threshold differs from thermal relativity in a nondegenerate gas.

For noninteracting electrons at zero temperature, define $x=p_F/(m_{e}c)$ with Fermi momentum $p_F=\hslash (3{\pi }^2{n}_e{)}^{1/3}$. The isotropic momentum flux gives electron degeneracy pressureIts nonrelativistic and ultrarelativistic limits arerespectively. In a fully ionized stellar gas, $n_e=\rho /({\mu }_e{m}_u)$, using mean molecular weight per electron ${\mu }_e$. This neglects finite-temperature effects, interactions, and changes in composition.

Pauli paramagnetism is the weak positive magnetic response produced when a magnetic field shifts the two spin Fermi surfaces of a degenerate electron gas. At zero temperature, if $g(E_F)$ includes both spin states, its total magnetic moment and susceptibility are

For a smooth single-spin density of states $g$ near a positive Fermi energy and fixed total particle number, the Sommerfeld expansion gives $\mu ={ϵ}_F-({\pi }^2/6)(g_F^{\prime }/g_F)(k_{B}T{)}^2+O(T^4)$. The spin contribution to the total-moment magnetic susceptibility, defined by differentiating with respect to $B$, is $2{\mu }_B^2\int g(E)[-n^{\prime }(E)]dE$. Expanding about $\mu$ and including the shift of chemical potential gives the displayed formula. The $g_F^{\prime \prime }$ term is thermal broadening; the $-g_F^{\prime 2}/g_F$ term enforces fixed particle number. A band edge or singular density of states within $k_{B}T$ of ${ϵ}_F$ requires a separate analysis. Orbital magnetic response is excluded.

For noninteracting spin-half particles with dispersion $E=cp$, the two spin populations at common chemical potential $\epsilon$ are $V(\epsilon \pm \mu B{)}^3/(6{\pi }^2{\hslash }^3{c}^3)$. Fixed total number changes $\epsilon$ only at quadratic order in $B$. Their population difference is therefore $3N\mu B/(c{p}_F)$ to first order, giving the displayed magnetic moment. Only the specified spin-energy splitting is included, not orbital magnetic effects.

For nonrelativistic electrons in area $A$, with spin degeneracy $g_s=2$ and energy $E={\hslash }^2{k}^2/(2m)$, the density of one-particle states is independent of energy:

For a constant total density of states $D$ above a zero-energy band bottom, integrating the Fermi-Dirac occupation gives $N=D{k}_{B}T\ln (1+e^{\mu /(k_{B}T)})$. With ${ϵ}_F=N/D$, inversion gives the displayed chemical potential. It approaches ${ϵ}_F$ from below with an exponentially small correction at low temperature, rather than the usual three-dimensional quadratic correction. This formula assumes an ideal parabolic band extending to arbitrarily high energy.

Each allowed wavevector occupies area $(2\pi {)}^2/A$ under periodic boundary conditions, for each independent spin state. For parabolic dispersion, the number below energy $ϵ$ is $g_{s}Amϵ/(2\pi {\hslash }^2)$. Thus the often-quoted coefficient $Am/(2\pi {\hslash }^2)$ is per spin. Unpolarized electrons have $g_s=2$; the same convention must be used in particle counting and the Fermi energy. A linear Dirac dispersion has a different energy dependence.

A quadratic-dispersion Fermi gas in two dimensions has constant density of states $g$. At zero temperature, $N=g{E}_F$ and $U=g{E}_F^2/2$. Hence the mean energy is half the Fermi energy, independent of a common spin-degeneracy factor.

When the density of states is constant and the chemical potential is fixed at $E_F$, the thermal gain above $E_F$ cancels the thermal depletion below it after the energy range is extended to the whole real line. The finite lower edge leaves only an exponentially small correction of order $e^{-\beta E_F}$.

For constant density of states, weighting the difference between the Fermi-Dirac distribution and the zero-temperature step by $E-E_F$ produces a convergent integral that scales as ${\beta }^{-2}$. Hence the leading low-temperature energy correction is proportional to $T^2$.

For a two-dimensional Fermi gas of $N$ spin-one-half particles with Fermi energy $E_F$, the constant density of states makes the low-temperature internal energy correction and heat capacity at constant areaThus its low-temperature heat capacity obeys a power law with exponent one.