Chiral Order Parameters, Current Algebra, and Pions

Spontaneous $SU(N_f)_L\times SU(N_f)_R\to SU(N_f)_V$ breaking produces $N_f^2-1$ pseudoscalar Goldstone modes; for two light flavors they are the pions. The axial-current pole, the partially conserved axial current relation, the quark condensate, and the near-zero Dirac spectrum are mutually connected diagnostics, but they are not interchangeable observables and their order-of-limits and renormalization conventions must be kept explicit.

Required background. Chiral Symmetry in QCD supplies the currents and breaking pattern; Goldstone’s Theorem: Hypotheses and the Pole Argument supplies the spectral reasoning behind the massless pole.

Helpful background. Explicit Breaking and Pseudo-Goldstone Modes supplies the perturbation away from the exact chiral limit.

The order parameter and its limits

Work first with two degenerate light quarks, $m_u=m_d=\widehat m$, and define the renormalized condensate per flavor by

$$\Sigma(\mu)=-\langle\bar u u\rangle_\mu =-\langle\bar d d\rangle_\mu>0$$

in the convention where the aligned QCD vacuum has a negative scalar condensate. More operationally,

$$\Sigma=\lim_{\widehat m\to0^+}\lim_{V\to\infty} \frac{1}{2V}\frac{\partial}{\partial\widehat m}\ln Z.$$

The thermodynamic limit must come first. At finite $V$ and zero source, symmetry averaging makes a noninvariant one-point function vanish; reversing the limits removes spontaneous symmetry breaking rather than measuring it.

The scalar density renormalizes, so $\Sigma(\mu)$ is not by itself scheme independent. In a mass-independent continuum scheme, $Z_mZ_S=1$, and therefore

$$m_q(\mu)\Sigma_q(\mu)$$

is renormalization-group invariant. Cutoff regulators that break chiral symmetry may also require additive subtractions. Consequently, quoting a condensate requires its scheme, scale, flavor normalization, and chiral extrapolation. Ward identities and physical amplitudes are the invariant outputs.

The matrix order parameter $\langle\bar q_R^{,j}q_L^{,i}\rangle\propto\delta^{ij}$ is invariant under $L=R$ and breaks $N_f^2-1$ generators. Goldstone’s theorem therefore supplies that many massless pseudoscalars when the quark masses vanish. The anomalous singlet axial generator is excluded, so it does not add a ninth mode for $N_f=3$.

The axial-current pole

For $SU(2)$ generators $T^a=\tau^a/2$, define

$$A_a^\mu=\bar q\gamma^\mu\gamma_5T^a q, \qquad P_a=\bar q,i\gamma_5T^a q.$$

An infinitesimal axial rotation changes $P_b$ by a scalar density. Its vacuum commutator with the axial charge is therefore proportional to $\delta_{ab}\Sigma$. If that expectation value is nonzero, the Fourier transform of $\langle T A_a^\mu(x)P_b(0)\rangle$ must contain a state whose pole reaches $p^2=0$ in the chiral limit. Isolating the lightest pole defines the decay-constant convention

$$\langle0|A_a^\mu(0)|\pi^b(p)\rangle =iF_\pi p^\mu\delta_a^{\ b}.$$

Some literature rescales both the axial generator and decay constant by $\sqrt2$. Formulae translate consistently only if that normalization is changed everywhere. Here $T^a=\tau^a/2$ fixes the convention.

With degenerate nonzero masses, the non-singlet Ward identity becomes

$$\partial_\mu A_a^\mu=2\widehat m,P_a.$$

Writing

$$\langle0|P_a(0)|\pi^b(p)\rangle=G_\pi\delta_a^{\ b}$$

and taking the one-pion matrix element gives the partially conserved axial current relation

$$F_\pi m_\pi^2=2\widehat m,G_\pi.$$

“Partially conserved” means that the divergence is controlled by the explicit mass term. It does not mean that the physical pion is exactly massless.

The Gell-Mann–Oakes–Renner relation

The integrated axial Ward identity relates the pseudoscalar susceptibility to the scalar condensate. With the normalizations above, its leading singular part is

$$2\widehat m\int d^4x\, \langle P_a(x)P_b(0)\rangle =\delta_{ab}\Sigma+O(\widehat m).$$

Near the chiral limit, the pion pole contributes $G_\pi^2/m_\pi^2$. Substituting that contribution and then using $F_\pi m_\pi^2=2\widehat mG_\pi$ yields

$$F_\pi^2m_\pi^2 =2\widehat m\Sigma+O(\widehat m^2) =(m_u+m_d)\Sigma+O(m_q^2).$$

This is the two-flavor Gell-Mann–Oakes–Renner relation. It is a leading chiral expansion, not an exact identity at physical quark masses. Its scale dependence cancels between $m_q$ and $\Sigma$, while $F_\pi$ and $m_\pi$ are physical. The original current-algebra argument and its assumptions are given in Gell-Mann, Oakes, and Renner 1968, pp. 2195–2199.

The same result appears at tree level in the chiral Lagrangian as $m_\pi^2=B(m_u+m_d)$ with $\Sigma=F^2B$ in the chiral limit. Loop and higher-order low-energy constants then supply the controlled corrections; see Chiral Lagrangians and Low-Energy QCD.

The Banks–Casher spectral diagnostic

In Euclidean space, let $\lambda_n$ be the real eigenvalues of the Hermitian massless Dirac operator $i\gamma^\mu D_\mu$, paired as $\pm\lambda$ away from exact zero modes. Define the spectral density per four-volume by

$$\rho(\lambda)=\lim_{V\to\infty}\frac1V \left\langle\sum_n\delta(\lambda-\lambda_n)\right\rangle.$$

The condensate at positive mass has the spectral representation

$$\Sigma(m)=\int_{-\infty}^{\infty}d\lambda\, \rho(\lambda)\frac{m}{\lambda^2+m^2}.$$

Since $m/(\lambda^2+m^2)\to\pi\delta(\lambda)$ as $m\to0^+$,

$$\Sigma=\pi\rho(0).$$

This is the Banks–Casher relation Banks and Casher 1980, pp. 103–109. It explains how a macroscopic accumulation of near-zero eigenvalues survives even though exact zero modes have vanishing density at fixed topology in the thermodynamic limit.

The limit order is again decisive: form $\rho(\lambda)$ after $V\to\infty$, then take $\lambda\to0$. At finite volume the spectrum is discrete and $\rho(0)$ is not a spontaneous order parameter. The density and scalar density must also be renormalized consistently; Banks–Casher does not turn a scheme-dependent condensate into a direct observable.

Diagnostic, relation, and observable

Object	Status	Essential qualification
$\langle\bar q q\rangle_\mu$	Vacuum order parameter	Scheme, scale, flavor normalization, source, and limit order
$\rho(0)$	Equivalent spectral diagnostic under Banks–Casher hypotheses	Infinite-volume and consistent Dirac/spectral renormalization
Axial-current pion pole	Physical current matrix element	Generator and $F_\pi$ normalization
PCAC	Renormalized Ward identity and its matrix elements	Quark-mass and pseudoscalar-density conventions
GMOR	Leading chiral relation	Corrections begin beyond leading order; $m_q\Sigma$ is invariant
Pion mass and scattering amplitudes	Direct observables	Electromagnetic and isospin-breaking corrections when comparing to data

The table gives a translation path: a scheme-dependent order parameter enters invariant Ward identities, which determine low-energy constants and observable amplitudes. A numerical spectrum can support the spectral diagnostic only after volume, mass, regulator, and continuum effects are controlled.

Common pitfalls

Taking the chiral limit before the volume limit. This yields a symmetric finite-volume state and erases the order parameter. Introduce a small mass or source, take $V\to\infty$, and only then remove it.

Quoting $\Sigma$ without a scheme. The scalar density runs. Quote $\Sigma(\mu)$ in a named scheme or use an invariant combination such as $m_q(\mu)\Sigma(\mu)$.

Calling GMOR exact. It is the leading term of a controlled low-energy expansion. State the chiral order of omitted terms and use the same decay-constant normalization on both sides.

Exercise

Assume the pseudoscalar susceptibility is pion-pole dominated and use $2\widehat m\,\chi_P=\Sigma+O(\widehat m)$ together with $F_\pi m_\pi^2=2\widehat mG_\pi$. Derive the leading quark-mass scaling of $m_\pi^2$.

Solution

Pion-pole dominance gives $\chi_P=G_\pi^2/m_\pi^2+O(1)$. Therefore

$$2\widehat m\frac{G_\pi^2}{m_\pi^2}=\Sigma+O(\widehat m).$$

Substitute $G_\pi=F_\pi m_\pi^2/(2\widehat m)$:

$$F_\pi^2m_\pi^2=2\widehat m\Sigma+O(\widehat m^2).$$

Thus $m_\pi^2\propto\widehat m$ while $m_\pi\propto\sqrt{\widehat m}$ at leading order.

References

• Banks, Tom, and A. Casher. “Chiral Symmetry Breaking in Confining Theories.” Nuclear Physics B 169 (1980): 103–125. DOI.
• Gell-Mann, Murray, Robert J. Oakes, and B. Renner. “Behavior of Current Divergences under $SU(3)\times SU(3)$.” Physical Review 175 (1968): 2195–2199. DOI.
• Weinberg, Steven. The Quantum Theory of Fields. Volume II: Modern Applications. Cambridge: Cambridge University Press, 1996, §19.4. DOI.