---
title: Deeply Bound Pionic-Atom States
url: https://www.emergentmind.com/topics/deeply-bound-pionic-atom-states
type: topic
---

# Deeply Bound Pionic-Atom States

Deeply bound pionic-atom states are atomic states of a $\pi^-$ bound to a nucleus in which the pion occupies low-lying orbitals such as $1s$, $2p$, or, in some theoretical studies, $2s$, with binding energies of several MeV and widths of order a few hundred keV. They are described by a Klein–Gordon equation in the Coulomb field of the nucleus plus a complex pion–nucleus optical potential, and they are formed not through the ordinary atomic X-ray cascade but through recoil-less nuclear reactions such as $(d,{}^3{\rm He})$. Their quantitative description has made them a precision laboratory for the in-medium $\pi N$ interaction, neutron-density distributions, neutron skins, the renormalization of the isovector amplitude $b_1$, and partial restoration of chiral symmetry in nuclear matter [2508.02267].

## 1. Bound-state definition and optical-potential framework

A pionic atom is described by the Klein–Gordon equation for a $\pi^-$ of reduced mass $\mu$ moving in the Coulomb field $V_c(r)$ of the nucleus plus a complex optical strong-interaction potential $V_{\rm opt}(r)$:
$$
(\nabla^2 + [E - V_c(r)]^2 - \mu^2 - \Pi(E,r))\,\phi(r) = 0,
$$
with $\Pi(E,r)\simeq 2\mu V_{\rm opt}(r)$. At leading order in the nuclear density, the $s$-wave part of $V_{\rm opt}(r)$ is parametrized as
$$
V_{\rm opt}(r) = -\,4\pi\,(1 + m_\pi/M)\,[\,b_0\,\rho(r) + b_1\,\delta\rho(r)\,]
-\,4\pi\,(1 + m_\pi/2M)\,B_0\,\rho^2(r) + (\text{$p$-wave terms}),
$$
where $\rho(r)=\rho_p(r)+\rho_n(r)$, $\delta\rho(r)=\rho_n(r)-\rho_p(r)$, $M$ is the nucleon mass, and $B_0$ is a phenomenological absorption parameter $(\sim 1\,{\rm fm}^4)$ [2508.02267].

The negative isovector scattering length $b_1$ is repulsive. Because of this repulsion, low-lying states such as $1s$ are pushed partly out of the nucleus, which keeps the absorptive imaginary part of $V_{\rm opt}(r)$ small. The result is that deeply bound states can remain narrow even in very heavy nuclei. The same framework is often written in the Ericson–Ericson form, with explicit $p$-wave terms and additional parameters $c_0$, $c_1$, $C_0$, and $\lambda$, in analyses that solve the bound-state problem and then embed it in formation-reaction calculations [2204.09211].

Free-space $\pi N$ scattering lengths used in this context are quoted as
$$
b_0^{\rm free}\simeq -\,0.020\,m_\pi^{-1},\qquad
b_1^{\rm free}\simeq -\,0.0885\,m_\pi^{-1},
$$
in one summary of global pionic-atom fits, while a sigma-term analysis uses
$$
b_1^{\rm free}=-0.0861\,m_\pi^{-1},\qquad f_\pi=92.4~{\rm MeV},
$$
within its chosen implementation [2508.02267; 2204.09211]. The medium dependence of the isovector term is commonly expressed through Weise’s formula,
$$
b_1(\rho)=b_1^{\rm free}/[1-\sigma_{\pi N}\rho/(m_\pi^2f_\pi^2)],
$$
which directly connects the optical potential to finite-density chiral dynamics [2508.02267].

## 2. Formation mechanism and spectroscopic method

Deeply bound $1s$ and $2p$ states cannot be reached by atomic X-ray cascade because higher levels such as $3d$ and $4f$ are so strongly absorbed that radiative yields vanish. Their observation therefore relies on recoil-less “stripping” reactions at forward angles, which implant a $\pi^-$ almost at rest in the residual nucleus. The canonical example is
$$
{}^{208}{\rm Pb}(d,{}^3{\rm He}){}^{207}{\rm Pb}\otimes\pi^-,
\qquad E_d\simeq 600~{\rm MeV},
$$
with the kinematics tuned to a momentum transfer $q\simeq 20~{\rm MeV}/c$ [2508.02267].

The standard experimental configuration uses deuterons of $\sim 600~{\rm MeV}$ incident on a thin lead or tin target of a few ${\rm mg/cm^2}$. The outgoing ${}^3{\rm He}$ is momentum-analyzed in a high-resolution magnetic spectrometer at $0^\circ$, with typical $\Delta p/p \simeq 5\times 10^{-4}$, corresponding to $\Delta E \simeq 200$–$400~{\rm keV}$ (FWHM). The missing-mass spectrum,
$$
M_x = [M_{\rm target}+E_d-E_{{}^3{\rm He}}],
$$
shows peaks at the pion binding energies. Variants of this method have been applied at GSI and, more recently, at RIKEN for Sn isotopes at $E_d\simeq 503~{\rm MeV}$, with comparable or improved resolution [2508.02267].

The reaction is commonly treated in DWIA or effective-number formalisms. In finite-angle studies, different combinations of pion-bound and neutron-hole states dominate at different scattering angles because of the matching condition. Near the recoil-free limit, $1s$ formation is strongest; at finite angles, $2p$ and other configurations become more prominent. Representative calculations for ${}^{122}{\rm Sn}(d,{}^3{\rm He})$ at $T_d=500~{\rm MeV}$ show that the dominant structures shift from $(1s_\pi\otimes 3s_{1/2}^{-1})$ at $0^\circ$ to $2p$-dominated configurations at $2^\circ$ and $5^\circ$ [1112.2450].

High-resolution theoretical studies further indicate that non-yrast states such as $2s$ are expected to be visible in $(d,{}^3{\rm He})$ spectra. Their simultaneous observation with the ground $1s$ state is helpful both for reducing uncertainties associated with neutron wave functions and for reducing experimental uncertainties associated with calibration of the absolute excitation energy [1107.5918].

## 3. Measured binding energies, widths, and spectral systematics

Representative experimental values for deeply bound states in heavy nuclei are summarized below. These quantities are extracted by fitting missing-mass peaks to Lorentzian line shapes folded with the experimental resolution function [2508.02267].

| System | $B_{\rm exp}$ (MeV) | $\Gamma_{\rm exp}$ (MeV) |
|---|---:|---:|
| ${}^{207}{\rm Pb}$ $1s$ | $6.26\pm0.10$ | $0.39\pm0.04$ |
| ${}^{206}{\rm Pb}$ $2p$ | $4.90\pm0.10$ | $0.30\pm0.05$ |
| ${}^{121}{\rm Sn}$ $1s$ | $7.11\pm0.09$ | $0.74\pm0.20$ |
| ${}^{121}{\rm Sn}$ $2p$ | $5.02\pm0.04$ | $0.28\pm0.06$ |

Within the Klein–Gordon plus optical-potential framework, theoretical predictions reproduce both $B$ and $\Gamma$ within $\simeq 100~{\rm keV}$ [2508.02267]. This level of agreement is central to the interpretation of deeply bound states as controlled probes of the $s$-wave $\pi N$ interaction in nuclei.

A separate high-statistics study of the ${}^{122}{\rm Sn}(d,{}^3{\rm He}){}^{121}{\rm Sn}\otimes\pi^-$ spectrum observed the atomic $1s$ and $2p$ states as distinct peak structures in the missing-mass spectrum, and reported that the $2p$ state in a Sn nucleus was observed for the first time. The same experiment measured the spectrum at finite reaction angles for the first time and determined the formation cross sections between $0^\circ$ and $2^\circ$ [1708.07621].

Systematics across nuclei reinforce several recurring features. Heavy nuclei support deeply bound $1s$ levels with MeV-scale binding energies, yet the widths remain modest because the repulsive $b_1$ term reduces the pion’s overlap with the absorptive nuclear interior. This is the technical basis for the phrase “deeply bound” in this field: the states are deeply bound in energy, but not strongly broadened by absorption [2508.02267].

## 4. In-medium $\pi N$ dynamics, $\sigma_{\pi N}$, and chiral symmetry restoration

The most consequential interpretation of deeply bound pionic-atom spectroscopy is its sensitivity to the density dependence of the isovector $s$-wave amplitude $b_1$. In the low-density chiral description, the medium dependence is written as
$$
b_1(\rho)=b_1^{\rm free}\biggl[1-\frac{\sigma_{\pi N}}{m_\pi^2f_\pi^2}\rho\biggr]^{-1},
$$
and the isoscalar channel is supplemented by a density-dependent double-scattering correction,
$$
b_0(\rho)=b_0^{\rm free}
-\varepsilon_1\,\frac{3}{2\pi}\bigl[b_0^{{\rm free}\,2}+2\,b_1(\rho)^2\bigr]
\Bigl(\frac{3\pi^2}{2}\rho\Bigr)^{1/3},
$$
in one explicit implementation [2204.09211].

Global fits of “normal” and “deeply bound” pionic-atom X-ray level shifts and widths across the periodic table $(A=12$–$238)$ extract
$$
\sigma_{\pi N}=(57\pm7)\ {\rm MeV},\qquad f_\pi=92.2\ {\rm MeV},
$$
while the isoscalar amplitude is less well constrained but is consistent with
$$
b_0\simeq -0.010\pm0.005\,m_\pi^{-1}.
$$
These fits are presented as clear experimental evidence for reduction of the pion decay constant and partial restoration of the QCD chiral condensate $\langle\bar q q\rangle$ in the nuclear medium [2508.02267].

The sensitivity of specific observables to $\sigma_{\pi N}$ has been quantified explicitly. For tin isotopes, numerical solutions show that the $1s$ binding energy $B_\pi(1s)$ and width $\Gamma_\pi(1s)$ vary almost linearly with $\sigma_{\pi N}$ over $25$–$60~{\rm MeV}$. For ${}^{123}{\rm Sn}$, the sensitivities are
$$
\Delta B_\pi(1s)\approx 6.2~{\rm keV},\quad
\Delta\Gamma_\pi(1s)\approx 5.9~{\rm keV},\quad
\Delta[B_\pi(1s)-B_\pi(2p)]\approx 4.5~{\rm keV}
$$
per $1~{\rm MeV}$ change in $\sigma_{\pi N}$. For the lighter and more sensitive ${}^{111}{\rm Sn}$, they are
$$
\Delta B_\pi(1s)\approx 7.5~{\rm keV},\quad
\Delta\Gamma_\pi(1s)\approx 12.9~{\rm keV},\quad
\Delta[B_\pi(1s)-B_\pi(2p)]\approx 5.8~{\rm keV}
$$
per $1~{\rm MeV}$ change in $\sigma_{\pi N}$ [2204.09211].

Using experimental precisions of roughly $\Delta B_\pi(1s)\sim 80~{\rm keV}$, $\Delta\Gamma_\pi(1s)\sim 40~{\rm keV}$, and $\Delta[B_\pi(1s)-B_\pi(2p)]\sim 10$–$15~{\rm keV}$, the corresponding $\sigma_{\pi N}$ uncertainties are estimated as $\sim 11~{\rm MeV}$ from $B_\pi(1s)$ alone in ${}^{111}{\rm Sn}$, $\sim 3.1~{\rm MeV}$ from $\Gamma_\pi(1s)$ alone, and $\approx 1.7$–$3.3~{\rm MeV}$ from the $1s$–$2p$ energy gap in ${}^{111,123}{\rm Sn}$. This identifies the gap and the $1s$ width in lighter Sn isotopes as the best observables for a few-MeV determination of $\sigma_{\pi N}$, provided the neutron density is known accurately [2204.09211].

Large-scale analyses that fit 100 pionic-atom data points, including deeply bound states, reaffirm the density-dependent renormalization of $b_1$. In one such fit, keeping $b_1$ density-independent gives
$$
b_1\approx -0.095\pm0.002\,m_\pi^{-1},
$$
which is $5$–$6\sigma$ more repulsive than the free value $b_1=-0.0885\,m_\pi^{-1}$. Imposing the chiral ansatz for $b_1(\rho)$ restores full agreement with the free $\pi N$ value and reduces $\chi^2$ by $4$–$5$ units. The same analysis infers
$$
f_\pi(\rho_0)/f_\pi\approx 0.78\pm0.05,
$$
corresponding to a $\sim 20$–$25\%$ drop of $f_\pi^2$ in nuclear matter [1405.7133].

## 5. Neutron densities, neutron skins, and effective density

Because the isovector term $b_1\,\delta\rho(r)$ shifts levels in proportion to the local neutron–proton density difference, deeply bound pionic levels in isotope chains serve as probes of neutron density distributions and neutron skins. In Sn, Pb, and Zr isotopes, precise measurements of $1s$ and $2p$ level shifts have been presented as sensitive to neutron skins of
$$
\Delta r_{np}\simeq 0.1\text{–}0.2~{\rm fm}
$$
[2508.02267].

In a 100-point global analysis, neutron densities were parameterized by
$$
r_n-r_p=\gamma\frac{N-Z}{A}+\delta,
$$
with best fits yielding $\gamma=1.00\pm0.15~{\rm fm}$ and implying for ${}^{208}{\rm Pb}$
$$
r_n-r_p\approx 0.18\pm0.03~{\rm fm}.
$$
This result was reported to be in line with parity-violation and proton-scattering results [1405.7133].

A complementary perspective is provided by the effective density sampled by the atomic pion. Defining the overlapping density
$$
S(r)=\rho(r)\,|R_{nl}(r)|^2\,r^2,
$$
and the peak position $r_e$ by $dS/dr=0$, one obtains $\rho_{\rm eff}=\rho(r_e)$. Numerical calculations across $A=20$–$240$ and for states $1s$, $2p$, $3d$, and others give
$$
\rho_{\rm eff}\simeq 0.08\text{–}0.11~{\rm fm}^{-3}\simeq 0.5\text{–}0.7\,\rho_0.
$$
This near-constancy means that atomic pions probe nearly the same $\sim \tfrac12\rho_0$ over a wide range of nuclei and orbitals [1107.5918].

That property has a dual significance. It stabilizes the interpretation of extracted in-medium amplitudes, since different states are not sampling widely different densities. At the same time, it limits direct differential access to the density dependence away from $\rho_{\rm eff}$. A related study concluded that extracting $b_1$ at densities different from $\rho_{\rm eff}$ requires $\Delta B\lesssim 10~{\rm keV}$ accuracy together with simultaneous $1s$–$2s$ data in order to break Seki–Masutani correlations [1107.5918].

## 6. Reaction-theory discrepancies and refined experimental strategies

Although the level energies and widths are well described, the formation reaction still contains unresolved issues. In the angular-distribution measurement of ${}^{122}{\rm Sn}(d,{}^3{\rm He}){}^{121}{\rm Sn}\otimes\pi^-$, the observed reaction-angle dependence of each state was reproduced in shape by theoretical calculations, but the absolute magnitude of the pionic $1s$ formation cross section showed a significant discrepancy. At $\langle\theta\rangle\approx 0.25^\circ$,
$$
I_{1s}^{\rm exp}\simeq (0.85\pm0.08\pm0.30)\,\mu{\rm b/sr},\qquad
I_{1s}^{\rm th}\simeq 5.0\,\mu{\rm b/sr},
$$
corresponding to
$$
C_{1s}\approx 0.17\pm0.03.
$$
For the $2p$ state at the same angle,
$$
I_{2p}^{\rm exp}\simeq (0.48\pm0.05\pm0.14)\,\mu{\rm b/sr},\qquad
I_{2p}^{\rm th}\simeq 0.61\,\mu{\rm b/sr},
$$
giving
$$
C_{2p}\approx 0.79\pm0.10.
$$
Possible origins listed for the $1s$ suppression include higher-order reaction mechanisms, inadequacies in the DWIA distorted waves or spectroscopic factors, sensitivity to the short-range $s$-wave part of the optical potential, and many-body in-medium modifications of the $\pi N$ vertex beyond the standard Ericson–Ericson term [1708.07621].

Parameter correlations also limit what can be extracted from spectroscopy. The same observables that constrain $\sigma_{\pi N}$ correlate strongly with the two-nucleon absorption parameter ${\rm Re}\,B_0$. For ${\rm Re}\,B_0<0$, contours of constant $B_\pi(1s)-B_\pi(2p)$ are nearly parallel in the $(\sigma_{\pi N},{\rm Re}\,B_0)$ plane, making separation difficult; neutron-density uncertainties can shift the best-fit $\sigma_{\pi N}$ by several MeV; and the angular ratio $1s/2p$ in calculated $(d,{}^3{\rm He})$ formation spectra is almost flat in $\sigma_{\pi N}$ and cannot be used alone to determine it [2204.09211].

One refined strategy is to use odd-neutron targets. In ${}^{117}{\rm Sn}(d,{}^3{\rm He}){}^{116}{\rm Sn}(0^+)$, the picked-up neutron is the extra $s_{1/2}$ neutron, so the final pionic level is a pure $(n\ell)\otimes 0^+$ configuration and does not suffer additional shifts from residual neutron-hole–pion couplings. Numerical predictions for this case give, for ${}^{116}{\rm Sn}$,
$$
1s:\ B_{1s}\approx 6.2~{\rm MeV},\ \Gamma_{1s}\approx 0.30~{\rm MeV},\ \text{cross section}\simeq 50~{\rm nb/(sr\,MeV)},
$$
$$
2p:\ B_{2p}\approx 3.1~{\rm MeV},\ \Gamma_{2p}\approx 0.15~{\rm MeV},\ \text{cross section}\simeq 20~{\rm nb/(sr\,MeV)},
$$
before experimental resolution. With a $300~{\rm keV}$ FWHM instrumental resolution, the peaks remain well separated, and for a beam intensity of $10^{12}~d/{\rm s}$ and a $1~{\rm mg/cm^2}$ Sn target, the expected yield is $\sim 10^3$ events/day in the $1s$ region within $300~{\rm keV}$, sufficient to determine $B_{1s}$ to $\lesssim 50~{\rm keV}$ and $\Gamma_{1s}$ to $\lesssim 100~{\rm keV}$ [1304.0598].

Taken together, these developments define the present status of deeply bound pionic-atom states. The basic bound-state phenomenology is robust; the spectroscopy has reached the level at which $1s$, $2p$, and potentially $2s$ states can be separated with sub-MeV resolution; the in-medium renormalization of $b_1$ and the associated partial restoration of chiral symmetry are supported by global fits and isotope studies; and the remaining challenges lie primarily in reaction theory, correlated optical-potential parameters, and precision control of neutron-density inputs [2508.02267].

Source: https://www.emergentmind.com/topics/deeply-bound-pionic-atom-states