Pion–Nucleon Sigma Term

• Pion–nucleon sigma term is a key QCD observable that measures the scalar light-quark content of nucleons and the impact of chiral symmetry breaking on nucleon mass.
• It is extracted via methods including Roy–Steiner equations, pionic-atom analyses, and effective field theories that correlate πN scattering data with nucleon structure.
• Current determinations converge near 58–60 MeV, influencing predictions in dark matter detection, in-medium chiral dynamics, and lattice QCD simulations.

The pion–nucleon sigma term, $\sigma_{\pi N}$, is a low-energy QCD observable defined by the scalar light-quark matrix element in the nucleon and, equivalently, by the Feynman–Hellmann derivative of the nucleon mass with respect to the average light-quark mass. It quantifies the contribution of explicit chiral-symmetry breaking to the nucleon mass, coincides with the scalar form factor at $t=0$, and sits at the intersection of $\pi N$ scattering, pionic atoms, chiral effective theory, lattice QCD, and scalar nucleon couplings relevant for dark-matter phenomenology (Alarcón et al., 2013, Hoferichter et al., 2016).

1. Operator definition and physical interpretation

In the isospin-symmetric limit, the sigma term is conventionally written as

$$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$

or, in equivalent notation,

$$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$

By the Feynman–Hellmann theorem,

$$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$

so $\sigma_{\pi N}$ measures the response of the nucleon mass to a variation of the light-quark masses (Alarcón et al., 2013, Ren et al., 2017).

This definition admits two closely related interpretations. First, $\sigma_{\pi N}$ is the nucleon’s isoscalar scalar form factor at zero momentum transfer, $\sigma(t=0)$. Second, it is the part of the nucleon mass generated by explicit, rather than spontaneous, chiral-symmetry breaking. The quantity therefore fixes how strongly Higgs-generated light-quark masses feed into the nucleon mass, while the remainder is dominated by gluonic dynamics and dynamical chiral symmetry breaking (Hoferichter et al., 2016, Liang et al., 15 Aug 2025).

The sigma term is also tied to the scalar flavor structure of the nucleon. In phenomenological discussions it is often combined with the strange sigma term, $\sigma_s = m_s\langle N|\bar s s|N\rangle$, or with the strangeness ratio

$t=0$0

because the size of $t=0$1 constrains how much strange scalar density can be accommodated in the nucleon (Alarcón et al., 2013, Huang et al., 2019).

2. Relation to $t=0$2 scattering and the Cheng–Dashen theorem

Phenomenological determinations of $t=0$3 proceed through the isoscalar $t=0$4 amplitude. The key low-energy theorem is the Cheng–Dashen relation, which connects the Born-subtracted isoscalar amplitude at the unphysical Cheng–Dashen point, $t=0$5, to the scalar form factor. In the dispersive formulation used in modern analyses,

$t=0$6

where $t=0$7 and $t=0$8 are subthreshold coefficients of the Born-subtracted isoscalar amplitude (Hoferichter et al., 2015).

The practical difficulty is that the Cheng–Dashen point lies outside the physical $t=0$9 region. Roy–Steiner equations solve this problem by combining analyticity, unitarity, and crossing symmetry into a controlled extrapolation from physical data to the subthreshold region. In that framework, $\pi N$0 becomes tightly correlated with the $\pi N$1-wave $\pi N$2 scattering lengths. One representative Roy–Steiner relation is

$\pi N$3

with

$\pi N$4

and reference scattering lengths

$\pi N$5

Using pionic-atom input, this program yields

$\pi N$6

(Hoferichter et al., 2016, Hoferichter et al., 2015).

The same dispersive machinery can be anchored directly to low-energy $\pi N$7 scattering rather than pionic atoms. A Roy–Steiner-based fit to low-energy $\pi N$8 cross sections extracted scattering lengths consistent with the pionic-atom determination and obtained

$\pi N$9

thereby reinforcing the “large-$$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$0” phenomenology from an independent data set (Elvira et al., 2017).

3. Pionic atoms, in-medium amplitudes, and nuclear observables

Pionic atoms provide a distinct route to $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$1 because the strong interaction shifts and broadens Coulombic $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$2 atomic levels in a way that depends on the $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$3 threshold amplitudes. In the Ericson–Ericson optical potential, the crucial quantity is the in-medium isovector $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$4-wave amplitude $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$5. Using the Tomozawa–Weinberg relation and a finite-density Gell-Mann–Oakes–Renner relation, Friedman and Gal wrote

$$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$6

so that a decrease of $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$7 and of the quark condensate with density enhances $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$8 (Friedman et al., 2019).

A global fit to 98 pionic-atom level shifts and widths across the periodic table, including deeply bound states in Sn isotopes and $$\sigma_{\pi N} = \hat m\,\langle N|\bar u u + \bar d d|N\rangle, \qquad \hat m = \frac{m_u+m_d}{2},$$9Pb, gave

$$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$0

That extraction was reported to be robust against variations of neutron-density parameters, $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$1-wave modeling, and other optical-potential terms. Higher-order corrections to the leading-density relation were found to involve partial cancellations, implying only a few percent overall systematic uncertainty (Friedman et al., 2019).

The pionic-atom method probes in-medium chiral dynamics rather than only free $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$2 scattering. In that setting, $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$3 governs the leading density dependence of

$$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$4

and therefore the partial restoration of chiral symmetry in nuclei. Because pionic atoms mainly sample densities around $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$5–$$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$6, the linear-density approximation is comparatively stable in practice (Friedman et al., 2019).

Deeply bound pionic atoms sharpen the same idea at the level of selected observables. A later analysis of Sn isotopes found that the $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$7–$$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$8 binding-energy gap and the $$\sigma_{\pi N} = m_q\,\langle N|\bar u u + \bar d d|N\rangle.$$9 width are particularly sensitive to $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$0. For realistic present-day experimental errors, the gap $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$1 was estimated to constrain $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$2 at the level of roughly $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$3 MeV, provided neutron densities and optical-potential parameters are sufficiently controlled. The same study also emphasized a strong correlation between $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$4, the real part of $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$5, and neutron-density profiles, so the method is intrinsically coupled to nuclear-structure systematics (Ikeno et al., 2022).

4. Effective-field-theory extractions and lattice-QCD strategies

Covariant baryon chiral perturbation theory (BChPT) in the EOMS scheme provides a direct link between $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$6 amplitudes and the sigma term through the low-energy constant $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$7. At $$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$8, the nucleon mass behaves as

$$\sigma_{\pi N} = \hat m\,\frac{\partial m_N}{\partial \hat m} = M_\pi^2\,\frac{\partial m_N}{\partial M_\pi^2},$$9

so that

$\sigma_{\pi N}$0

at leading order, with a calculable loop correction at the same chiral order. Fits of the covariant EOMS $\sigma_{\pi N}$1 amplitude with explicit $\sigma_{\pi N}$2 to modern partial-wave analyses gave

$\sigma_{\pi N}$3

while the older Karlsruhe analysis yielded a smaller value near $\sigma_{\pi N}$4 MeV. The higher value was associated with modern meson-factory and pionic-atom input (Alarcón et al., 2013, Camalich et al., 2011).

Lattice QCD accesses $\sigma_{\pi N}$5 in two main ways. The direct method computes the scalar three-point function and extracts

$\sigma_{\pi N}$6

whereas the Feynman–Hellmann method fits the nucleon mass as a function of $\sigma_{\pi N}$7 and differentiates with respect to $\sigma_{\pi N}$8 (Gupta et al., 2022).

These strategies have not always agreed numerically. A covariant SU(2) EOMS analysis of ETMC nucleon masses obtained

$\sigma_{\pi N}$9

and found that including a virtual $\sigma_{\pi N}$0 does not change the result qualitatively (Ren et al., 2017). By contrast, a heavy-baryon SU(3) $\sigma_{\pi N}$1 analysis that fit $\sigma_{\pi N}$2 phase shifts together with octet-baryon masses reported

$\sigma_{\pi N}$3

together with a very small nucleon strangeness content, $\sigma_{\pi N}$4 (Huang et al., 2019).

A more recent two-loop EOMS treatment of the nucleon mass, applied to $\sigma_{\pi N}$5 lattice data, reported

$\sigma_{\pi N}$6

That analysis identified intermediate $\sigma_{\pi N}$7 rescattering effects, which enter only at two-loop order, as the mechanism by which the long-standing lattice–dispersive tension can be naturally resolved (Liang et al., 15 Aug 2025).

5. Tension, systematics, and disputed determinations

For much of the last decade, the central controversy concerned the gap between phenomenological extractions near $\sigma_{\pi N}$8 MeV and several direct lattice calculations near the physical point. A Roy–Steiner review collected representative lattice results such as $\sigma_{\pi N}$9 MeV from BMW, $\sigma(t=0)$0 MeV from $\sigma(t=0)$1QCD, $\sigma(t=0)$2 MeV from ETMC, and $\sigma(t=0)$3 MeV from RQCD, and characterized the mismatch with $\sigma(t=0)$4 MeV as a tension of roughly $\sigma(t=0)$5 or more (Hoferichter et al., 2016).

A major proposed explanation is excited-state contamination in lattice nucleon correlators. A ChPT-guided reanalysis argued that direct lattice calculations of $\sigma(t=0)$6 are contaminated by multihadron $\sigma(t=0)$7 and $\sigma(t=0)$8 states, whose contributions are large and negative at realistic source–sink separations. In that picture, standard fits return values near $\sigma(t=0)$9 MeV, but fits that incorporate the multihadron spectrum move the result toward $\sigma_s = m_s\langle N|\bar s s|N\rangle$0 MeV (Gupta et al., 2022). A subsequent study summarized the same conclusion more explicitly: $\sigma_s = m_s\langle N|\bar s s|N\rangle$1 and $\sigma_s = m_s\langle N|\bar s s|N\rangle$2 states each contribute about $\sigma_s = m_s\langle N|\bar s s|N\rangle$3 MeV to $\sigma_s = m_s\langle N|\bar s s|N\rangle$4, and including the $\sigma_s = m_s\langle N|\bar s s|N\rangle$5 as an explicit degree of freedom does not alter that conclusion (Gupta et al., 2023).

A second systematic issue is the convention used for the isospin limit. A recent analysis pointed out that phenomenology conventionally defines the isospin limit with the charged pion mass, while lattice QCD usually adopts the neutral pion mass. In ChPT this mismatch induces

$\sigma_s = m_s\langle N|\bar s s|N\rangle$6

which should be included when comparing lattice and phenomenological results. The same work updated the Roy–Steiner-plus-pionic-atom benchmark, using the latest pionic-hydrogen width, to

$\sigma_s = m_s\langle N|\bar s s|N\rangle$7

(Hoferichter et al., 2023).

A separate controversy concerns the low-energy CHAOS data set. One analysis of CHAOS $\sigma_s = m_s\langle N|\bar s s|N\rangle$8 data extracted

$\sigma_s = m_s\langle N|\bar s s|N\rangle$9

placing the sigma term near the lower edge of the historical range (Stahov et al., 2012). A later comment argued that the angular distribution of the CHAOS $t=0$00 differential cross sections is incompatible in shape with the rest of the modern low-energy $t=0$01 data and that this problem must be resolved before any extrapolation into the unphysical Cheng–Dashen region can be trusted (Matsinos et al., 2013).

The range of representative determinations illustrates both the methodological diversity and the historical spread:

Approach	Result	Source
Roy–Steiner + pionic atoms	$t=0$02 MeV	(Hoferichter et al., 2015)
Low-energy $t=0$03 scattering + Roy–Steiner	$t=0$04 MeV	(Elvira et al., 2017)
Covariant EOMS BChPT fit to $t=0$05 PWAs	$t=0$06 MeV	(Alarcón et al., 2013)
Global pionic-atom fit	$t=0$07 MeV	(Friedman et al., 2019)
Covariant SU(2) EOMS fit to lattice $t=0$08	$t=0$09 MeV	(Ren et al., 2017)
HB SU(3) fit to $t=0$10 phase shifts and octet masses	$t=0$11 MeV	(Huang et al., 2019)
Two-loop EOMS extrapolation of $t=0$12 lattice data	$t=0$13 MeV	(Liang et al., 15 Aug 2025)

6. Phenomenological consequences and current outlook

Because $t=0$14 controls the scalar light-quark content of the nucleon, it propagates directly into scalar nucleon couplings. In the Roy–Steiner framework, a sigma term near $t=0$15 MeV leads to

$t=0$16

$t=0$17

with immediate implications for Higgs-mediated and other scalar interactions in dark-matter direct detection (Hoferichter et al., 2015). More generally, both phenomenological and lattice-oriented discussions identify $t=0$18 as a dominant hadronic input to spin-independent WIMP–nucleon scattering and to scalar contributions in $t=0$19 conversion and EDM analyses (Hoferichter et al., 2016, Gupta et al., 2022).

In nuclear matter, $t=0$20 fixes the leading density dependence of the quark condensate and thus the strength of partial chiral restoration. Pionic-atom analyses near $t=0$21 MeV imply a noticeable decrease of $t=0$22 and $t=0$23 already at densities around $t=0$24, which feeds directly into the in-medium renormalization of the isovector $t=0$25 amplitude $t=0$26 (Friedman et al., 2019). Continuum DSE studies that infer $t=0$27 from the density dependence of the chiral condensate likewise find values near $t=0$28 MeV, reinforcing the link between the sigma term and in-medium chiral dynamics (Huang et al., 2019).

The present status is more convergent than the older “small-$t=0$29” versus “large-$t=0$30” dichotomy suggests. Phenomenological analyses based on Roy–Steiner equations, low-energy $t=0$31 scattering, pionic atoms, and covariant BChPT continue to cluster near $t=0$32–$t=0$33 MeV (Elvira et al., 2017, Hoferichter et al., 2023). Several recent developments on the lattice side—explicit treatment of multihadron excited states, correction of isospin-limit conventions, and two-loop chiral extrapolations with $t=0$34 rescattering—move the preferred range upward and reduce the discrepancy (Gupta et al., 2022, Hoferichter et al., 2023, Liang et al., 15 Aug 2025).

The sigma term therefore remains a benchmark quantity for the mutual consistency of hadronic phenomenology, effective field theory, and lattice QCD. Its importance lies not only in the number itself, but in the fact that $t=0$35 encodes, in a single scalar matrix element, how chiral symmetry breaking propagates from the QCD Lagrangian into nucleon mass, $t=0$36 amplitudes, nuclear observables, and scalar couplings beyond the Standard Model.