Resonance photoproduction of pionic atoms at the Gamma Factory
Extracted source text · arXiv 2010.06912
% ****** Start of file apssamp.tex ******
%
% This file is part of the APS files in the REVTeX 4.2 distribution.
% Version 4.2a of REVTeX, December 2014
%
% Copyright (c) 2014 The American Physical Society.
%
% See the REVTeX 4 README file for restrictions and more information.
%
% TeX'ing this file requires that you have AMS-LaTeX 2.0 installed
% as well as the rest of the prerequisites for REVTeX 4.2
%
% See the REVTeX 4 README file
% It also requires running BibTeX. The commands are as follows:
%
% 1) latex apssamp.tex
% 2) bibtex apssamp
% 3) latex apssamp.tex
% 4) latex apssamp.tex
%
\documentclass[%
reprint,
%superscriptaddress,
%groupedaddress,
%unsortedaddress,
%runinaddress,
%frontmatterverbose,
%preprint,
%preprintnumbers,
%nofootinbib,
%nobibnotes,
%bibnotes,
amsmath,amssymb,
aps,
%pra,
%prb,
%rmp,
%prstab,
%prstper,
%floatfix,
]{revtex4-2}
\usepackage{graphicx}% Include figure files
\usepackage{dcolumn}% Align table columns on decimal point
\usepackage{bm}% bold math
\usepackage{color}
\usepackage{amsmath}%[fleqn]
\usepackage{wasysym}
\usepackage{float}
\usepackage{array,url,kantlipsum}
\usepackage{verbatim}
\usepackage{xcolor}
\usepackage{CJK}
\usepackage{textcomp}
\newcolumntype{P}[1]{>{\centering\arraybackslash}p{#1}}
\usepackage{array}
\usepackage{hyperref}
\hypersetup{
colorlinks,
citecolor=blue,
filecolor=blue,
linkcolor=blue,
urlcolor=blue
}
\newcolumntype{x}[1]{>{\centering\arraybackslash\hspace{0pt}}p{#1}}
\newcommand{\DB}[1]{\textcolor{blue}{#1}}
\newcommand{\VF}[1]{\textcolor{red}{#1}}
\newcommand{\JJ}[1]{\textcolor{orange}{#1}}
\newcommand{\OT}[1]{\textcolor{green}{#1}}
\newcommand{\brown}[1]{\textcolor{rgb:red,4;green,2;yellow,-1}{#1}}
\begin{document}
%\preprint{APS/123-QED}
%\title{Manuscript Title:\\with Forced Linebreak}% Force line breaks with \\
%\thanks{A footnote to the article title}%
\title{Resonance photoproduction of pionic atoms at the Gamma Factory}
\author{Victor V.\,Flambaum}
\affiliation{School of Physics, University of New South Wales, Sydney 2052, Australia}
\affiliation{Johannes Gutenberg-Universit{\"a}t Mainz, 55128 Mainz, Germany}
\affiliation{Helmholtz-Institut, GSI Helmholtzzentrum f{\"u}r Schwerionenforschung, 55128 Mainz, Germany}
%\affiliation{The New Zealand Institute for Advanced Study, Massey University Auckland, 0632 Auckland, New Zealand}
\author{Junlan Jin}
\affiliation{Department of Modern Physics, University of Science and Technology of China, Hefei 230026, China}
\author{Dmitry Budker}
\affiliation{Johannes Gutenberg-Universit{\"a}t Mainz, 55128 Mainz, Germany}
\affiliation{Helmholtz-Institut, GSI Helmholtzzentrum f{\"u}r Schwerionenforschung, 55128 Mainz, Germany}
\affiliation{Department of Physics, University of California, Berkeley, California 94720, USA}
\date{\today}% It is always \today, today,
% but any date may be explicitly specified
\begin{abstract}
We present a possibility of direct resonance production of pionic atoms (Coulomb bound states of a negative pion and a nucleus) with a rate of up to
$\sim 10^{10}$
per second using the gamma-ray beams from the Gamma Factory.
\end{abstract}
\maketitle
\section{Introduction}
The pionic atom \cite{Backenstoss1970_pionic_atom} consists of a negative pion trapped in the Coulomb potential of an atomic nucleus.
Such systems provide great opportunities to study strong interaction and derive information on nuclear structure. Theoretical study of energy levels in pionic atoms started long ago \cite{Deser1954_En,Brueckner1955_En} and initiated extensive theoretical and experimental studies - see, for example,
recent experiment on laser spectroscopy of pionic $^4$He of Ref.\,\cite{Hori2020_spectra_helium} and references therein.
% Kaonic atoms have also been studied \cite{Friedman2007_widths,Curceanu2019_kaonic_width}, including
% their energy-level widths and kaon radius
% \cite{Barnea2020_kaon_deuterium_width,Michel2020_kaon_radius}.
A conventional production mechanism of pionic atoms involves creation of free negative pions which are then captured by nuclei. Here, we explore a possibility of direct production of pion-nucleus bound states by a monochromatic gamma-ray beam with the energy tuned to that of the bound state. This monochromatic gamma-ray beam is expected at the proposed Gamma Factory (GF) \cite{Krasny2015,Budker2020_AdP_GF} currently studied within the CERN Physics Beyond Colliders program.
It is noted that photoproduction of pionic atoms was once put forward by C. Tzara \cite{TZARA1970_photoproduction_pion}. Here, we introduce different approaches to evaluating photoproduction cross-sections, and
making use of experimental data for free-pion (both charged and neutral) photoproduction, we extend the analysis to a range of nuclei.
Estimates presented below show that the pionic-atom production rate $p$ may, in principle, reach $\sim 10^{10}$ atoms per second, i.e. exceed the production rate at existing facilities
($\sim 10^{5}$ pionic atoms per second - see, for example, Ref.\,\cite{Hori2020_spectra_helium}) by many orders of magnitude. Specific experimental arrangements, the discussion of which is beyond the scope of the present paper, may
reduce this gain in the production rate but the problem certainly deserves a preliminary investigation which is the aim of the present note.
\section{Estimate of the production rate}
We assume an arrangement, where the gamma-rays impinge onto a fixed target. Photoproduction of pionic atoms is realized through the reaction $\gamma+ n \rightarrow p + \pi^-$ within a nucleus, i.e.,
\begin{equation}
\gamma + {}^A_{Z_i}\textrm{X} \rightarrow (^A_Z\textrm{X}' +\pi^-)_{nl}\,,
\end{equation}
where ${}^A_{Z_i}$X and $^A_Z\textrm{X}'$ are the initial and final nucleus, respectively (both in their nuclear ground state \footnote{In principle, pionic atoms with final nuclei in excited states can also be resonantly produced with higher photon energies.}),
$Z_i$ and $Z= Z_i+1$ are the corresponding atomic numbers, $A$ is the number of nucleons which is the same for the initial nucleus and the final one, $n$ is the pionic atom principal quantum number, and $l$ is the angular momentum quantum number. We will focus mainly on $ns$ states since $\pi^-$ in $ns$ states have larger probability densities inside the nucleus and thus larger cross-sections for their production.
An example of the reaction is
\begin{equation}
\label{HHe}
\gamma + {}^3\textrm{H} \rightarrow (^3\textrm{He} +\pi^-)_{ns}\,.
\end{equation}
Similar reactions occur with heavier nuclei.
% We assume that $\pi^-$ is in the $ns$ bound state of the pionic atom since an $s$-wave pion has a non-zero probability to be produced inside the nucleus.
Assuming that the initial atom is at rest, we obtain an equation for the resonant photon energy
$E_{\gamma}= m c^2 + E_{CB} + \Delta \epsilon_{ns} + E_k + \Delta Mc^2$, where $m c^2=$139.57\,MeV is the rest energy of the negative pion, $c$ is the speed of light,
\begin{equation}
E_{CB}\approx -\frac{\mu c^2}{2}\left(\frac{Z\alpha}{n}\right)^2
\end{equation}
is the energy of $\pi^-$ in the nuclear Coulomb field, $\mu=mM/(m+M)$ is the reduced mass, $M$ is the mass of the final nucleus, $\alpha=1/137$ is the fine structure constant, $\Delta \epsilon_{ns}$ is the energy shift due to the strong interaction (see, for example, Ref.\,\cite{Friedman2007_widths}),
%,\JJ{except a different sign is used here}
$E_k \approx p^2/[2(m+M)]$ is the kinetic energy of the pionic atom with momentum $p=E_{\gamma}/c$, and $\Delta M c^2$ is the change of the mass energy of the final nucleus relative to the initial one. In heavy atoms the relativistic formula for $E_{CB}$
should be used \footnote{Note that the relativistic formula for the energy of a spin-zero particle in the case of the point-like nucleus breaks down for $Z\geq 137/2$, so the finite nucleus potential should be used when $Z$ is close to or above 67}. Production of $1s$, $2p$ and several other low-lying $ns$ states in heavy elements, so called deeply bound pionic states, may be especially interesting, since they can provide information on the pion behavior inside the nucleus and cannot be populated via an X-ray cascade in the capture process of a free pion due to the strong pion absorption in higher states (see, for example, Ref.\,\cite{Friedman2007_widths}). Note that the population of the higher $ns$ states in the cascade is also suppressed since the cascade tends to populate the highest orbital angular momentum (circular) states, $l=n-1$.
Deeply bound pionic atoms were experimentally studied in a series of measurements such as those in Refs.\,\cite{Yamazaki_1996_deeply_bound_states,Gilg2000_deeply_bound_pionic_207Pb,Geissel2002_deeply_bound_205Pb} via the $^{206,208}$Pb(d,$^3$He) transfer reaction. Other proposals for producing deeply bound pionic atoms
by methods other than resonance photoproduction are discussed in Refs.\,\cite{Nieves1993_radiative_trapping,Raywood1997_208Pb_radiative_capture,Tain2004_gamma-ray_ALBA}.
Pionic atoms in low-lying atomic states are usually short lived with typical
lifetimes of $\tau=\hbar/\Gamma_{tot}<10^{-16}$\,s ($\hbar$ is the reduced Planck constant, $\Gamma_{tot}$ is the total width of the pionic state). The negative pion is absorbed by the nucleus, which is followed by
%nuclear fission with various channels
emission of neutrons, protons, gamma rays, etc.
The produced pionic atoms can
be detected via measuring the fission products following the pion
absorption, where the challenge is to minimize the backgrounds directly induced by the high-energy photons.
We note that detection of gamma rays as a diagnostic for radiative capture of free pions by nuclei was used in Ref.\,\cite{Strandberg2020_threshold_pion_from_deuteron} and, in principle, is also applicable for bound pions in pionic atoms.
%With these facts, one of the possible ways of detecting produced pionic atoms could be to detect the fission signal due to pion absorption and suppress the background induced by high-energy photons meanwhile.}
\subsection{Photoproduction cross-section for monochromatic photons}
\label{Subsec:monochromatic_cross-section}
The on-resonance cross-section for absorption of monochromatic photons is \cite{Landau1958}
\begin{equation}
\label{Eq:res_cross_section}
\sigma_0=2 \pi\frac{2 J+1}{2 I+1} \left (\frac{\hbar c}{E_\gamma}\right )^2\frac{\Gamma_{\gamma}}{\Gamma_{tot}},
\end{equation}
where $I$ and $J$ are the spin quantum numbers of the initial and final nucleus, respectively (we assume $I\approx J$ in the following), $E_{\gamma}\approx 140$\,MeV is the photon energy,
$\Gamma_{\gamma}$ is the partial width of the produced pionic state $(^A_Z\textrm{X}' +\pi^-)_{nl}$ corresponding to its
decay via radiative pion capture with the final nucleus
%radiative decay to
${}^A_{Z_i}$X in the ground state and $\Gamma_{tot}$ is the total width of the pionic state.
%which is dominated by the pion absorption (states with very large principal quantum number $n$ may be an exception).
Unfortunately, we are not aware of any calculation of the ratio $\Gamma_{\gamma}/\Gamma_{tot}$. However, it is instructive to start from a few simple arguments indicating what value we may expect for this ratio.
The width $\Gamma_{\gamma}$ appears in the Breit-Wigner formula from the entrance channel where the photon is absorbed and the pion is created. The same partial width $\Gamma_{\gamma}$ appears in the decay channel where the pion is absorbed and the photon is emitted. The total width of the pionic atom $\Gamma_{tot}$ for $ns$ states is dominated by pion absorption (states with large principal quantum number $n$ may be an exception). The partial width $\Gamma_{\gamma}$ corresponds to an additional photon
radiated in the process of pion absorption.
The photon emission brings us
an extra factor $\alpha$ in the rate of the process. Therefore, $\Gamma_{\gamma}/\Gamma_{tot} \propto \alpha = 0.007$ \footnote{Surprisingly, such a naive estimate gives a correct result for the decay $\Delta \to N \pi \gamma$, where $\Gamma_{\gamma}/\Gamma_{tot}\approx 0.006$ \cite{Zyla2020_Particle_Data} }.
The assumption $\Gamma_{\gamma}/\Gamma_{tot} \sim \alpha$ would give us $\sigma_0 \sim 1.\times10^{-27}$\,cm$^2$.
As we see in Table\,\ref{tab:sigmap_light_nuclei}, this rough estimate is actually valid for the resonance cross-section in light nuclei. For heavy nuclei $\sigma_0$ is smaller.
Indeed, according to Refs.\,\cite{Petrukhin1965_radiative_capture,Delorme1966_radiative_capture}, radiative capture of a pion by the nuclues is typically 2$\%$ of the total capture rate. Here 2$\%$ can only be used as an upper estimate of the ratio $\Gamma_{\gamma}/\Gamma_{tot}$ in Eq.\,\eqref{Eq:res_cross_section}, since after the radiative capture the final nucleus may be not in the ground state (i.e., it could be in an excited or unbound state). For heavier nuclei containing a larger number of nucleons, there are more channels in the radiative capture process so that this ratio may be suppressed significantly.
Though we cannot directly calculate the resonant cross-section for monochromatic photons due to the lack of knowledge of $\Gamma_{\gamma}/\Gamma_{tot}$, we will calculate the integrated cross-section which we actually need for getting production rates of pionic atoms at the GF as discussed below.
% Therefore, we roughly estimate that $\Gamma_{\gamma}/ \Gamma_{tot}\approx 2\%/A \approx 0.01/Z$ for light nuclei with $Z\leq 10$, where $A\approx 2Z$ is the atomic mass number.
% (For heavy nuclei, this ratio should be suppressed significantly, but we expect it to gradually reach a saturation instead of showing $1/A$-dependence.)
% (Heavier nuclei will be considered below.)
%(see pion absorption i, n 3He).
% For light nuclei with $Z \leq 7$, the branching ratio can be estimated as $\Gamma_{\gamma}/\Gamma_{tot}\approx 1\%$. This gives $\sigma_0 \sim 1.\times10^{-27}$\,cm$^2$ with assuming $J\approx I$. For heavier nuclei, $\Gamma_{\gamma}/\Gamma_{tot}\approx 1\%$ may only be used as an upper estimate.
\subsection{Integrated cross-section for producing bound pions is related to free-pion production}
For photoproduction of pionic atoms in an $ns$ state with the total width $\Gamma_{tot}(Z,n)$,
the dependence of the absorption cross-section on photon energy is
\begin{equation}
\label{Eq:crosec_energy_dependence_NT}
\sigma(E)=\sigma_0 \frac{(\Gamma_{tot}/2)^2}{(\Gamma_{tot}/2)^2+(E-E_{\gamma})^2}.
\end{equation}
This gives the integrated cross-section
\begin{equation}
\int \sigma(E) dE=\frac{\pi}{2} \sigma_0 \Gamma_{tot}(Z,n).
\end{equation}
The cross-section for the production of bound pions may be related to the cross-section $\sigma_p$ for producing free pions with a small pion momentum $p$ and with the final nucleus in its ground state as \cite{TZARA1970_photoproduction_pion}
\begin{equation}
\label{Eq:crosec_bound_pion_v1_NT}
\int \sigma(E) dE=\sigma_p \frac{(Z\alpha)^2}{n^3} K mc^2,
\end{equation}
where $K= 1-\exp{(-2\pi Z\alpha mc/p)}$. For the zero pion momentum ($p=0$), the factor becomes $K=1$.
If we use $\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)/n^3$ (see Sec.\,\ref{Subsec:pion_absorption_width}), we have
\begin{equation}
\label{Eq:crosec_bound_pion_NT}
\frac{\pi}{2}\sigma_0 \Gamma_{tot}(Z,1)=\sigma_p (Z\alpha)^2 K mc^2.
\end{equation}
We see that the resonance cross-section $\sigma_0$ for the $ns$ level production does not depend on $n$, i.e., as expected, $\Gamma_{\gamma}$ and $\Gamma_{tot}$ have the same $n$ dependence - see Eq.\,\eqref{Eq:res_cross_section}.
% \JJ{
% It is noted that for producing free pions from heavy nuclei near threshold, $K\approx 1$ so that $\sigma_p$ is almost a constant independent of $p$ according to Eq.\,\eqref{Eq:crosec_bound_pion_NT}. For example,
% %for $p<100\,$MeV$/c$ and $Z>50$, we have $0.96<K<1$ which means
% $\sigma_p$ is almost a constant for $p$ from 0 to 100\,MeV$/c$ for a given $Z$ ($Z>50$).
% }
%(A potential problem on the asymptotic behavior of $\sigma_p$ [Eq.\,\eqref{Eq:crosec_bound_pion_NT}] for heavy nuclei: $K= 1-\exp{(-2\pi Z\alpha mc/p)}= 1-\exp{(-2\pi Z\alpha \sqrt{m c^2/(2T_\pi)})}$ For heavy nuclei
%with $Z> 40$, $K\approx 1$ when $T_\pi < 70$\,MeV so that
%$\sigma_p$ should be constant near threshold instead of showing the typical steep increase above threshold such as in %\cite{Sakamoto1990_133Ba_from_133Cs,Sakamoto1989_51Cr_from_51V}.)
Note that Eqs.\,\eqref{Eq:crosec_bound_pion_v1_NT} and \eqref{Eq:crosec_bound_pion_NT} are based on the relation between the wavefunctions of a slow free pion and a bound pion at the nucleus. Indeed, in the vicinity of the nucleus the total energy of both the bound and free pions may be neglected in comparison with the large interaction energy.
Therefore, Schr{\"o}dinger equations for the bound and free pions are the same and their wavefunctions are proportional to each other. The proportionality coefficient is determined from the comparison of the bound and free Coulomb wavefunctions outside the nucleus where there is negligible strong interaction.
%are determined by the normalization of the bound and free wavefunctions which is determined by the large distance where the potential is the Coulomb one.
This is why the relations of Eqs.\,\eqref{Eq:crosec_bound_pion_v1_NT} and \eqref{Eq:crosec_bound_pion_NT} are determined by the Coulomb wavefunctions only and are insensitive to the strong interaction. Strictly speaking, this statement is accurate for the wavefunctions with large principal quantum numbers $n$ and small free-pion momenta $p$, which are proportional to the zero-energy solution up to large distances defined by the energy $|E|\ll Ze^2/r$; here $E=E_{CB}$ or $E=p^2/2m$ for bound and free pions, respectively.
However, in atoms with $Z<37$ the results may be extended down to $n=1$ since $1s$ orbital is located mainly outside the nucleus there and the Coulomb-like $1/n^3$ relation between $ns$ and $1s$ pion densities at the nucleus is reasonably accurate. These relations are inaccurate in heavy atoms with $Z>37$ where $1s$ orbital sits mainly inside the nucleus. Therefore, for the small principal quantum number $n$ and $Z>37$, Eqs.\,\eqref{Eq:crosec_bound_pion_v1_NT} and \eqref{Eq:crosec_bound_pion_NT} may provide only an order-of-magnitude estimate.
\subsection{Cross-section for producing free pions near threshold}
\label{Subsec:free_pion_cross-section}
The free pion photoproduction $(\gamma,\pi^-)$ cross-sections have been measured for many nuclei.
What we need to calculate bound pion production cross-sections is $\sigma_p$ for producing free pions near threshold with the final nucleus directly in its ground state, for which related experimental studies are limited.
% Data on $\sigma_{p=0}$ for light nuclei can be found in Refs.\,\cite{Singham1981_threshold_crosec, Min1976_threshold_11C,Epstein1978_threshold_12N,DeCarlo1980_threshold_14O,Singham1979_threshold_7Li_12C}:
% 8\,$\mu$b for $^{7}\textrm{Li}(\gamma, \pi^-) ^{7}\textrm{Be}_{g.s.}$,
% 4\,$\mu$b for $^{11}\textrm{B}(\gamma, \pi^-) ^{11}\textrm{C}_{g.s.}$,
% 4\,$\mu$b for $^{12}\textrm{C}(\gamma, \pi^-) ^{12}\textrm{N}_{g.s.}$,
% 0.2\,$\mu$b for $^{14}\textrm{N}(\gamma, \pi^-) ^{14}\textrm{O}_{g.s.}$, where $g.s.$ refers to the ground state.
Data on $\sigma_{p=0}$ for light nuclei from
%experiments or calculations in
Refs.\,\cite{Bernstein1976_pion_threshold_12C_12N,Bosted1979_near_threshold_cross-section,Singham1981_threshold_crosec, Min1976_threshold_11C,Epstein1978_threshold_12N,DeCarlo1980_threshold_14O,Singham1979_threshold_7Li_12C} are listed in Table\,\ref{tab:sigmap_light_nuclei}.
\begin{table}[t]
\centering
\begin{tabular}{c c c c c c}
\hline
\hline
${}_{Z_i}^A$X &${}_Z^A\textrm{X}'_{g.s.}$ &$\sigma_{p=0}$\,($\mu$b) &$\sigma_0$\,($\mu$b) &$[10^3\times\Gamma_\gamma/ \Gamma_{tot}]$ &rate (s$^{-1})$\\
\hline \\[-0.2cm]
$_3^{7}\textrm{Li}$ &$_4^{7}\textrm{Be}_{g.s.}$ &8 &1200 &9.05 &6.0$\times10^9$\\
$_5^{11}\textrm{B}$ &$_6^{11}\textrm{C}_{g.s.}$ &4 &260 &1.90 &2.7$\times10^9$\\
$_6^{12}\textrm{C}$ &$_7^{12}\textrm{N}_{g.s.}$ &4 &200 &1.50 &2.6$\times10^9$\\
$_7^{14}\textrm{N}$ &$_8^{14}\textrm{O}_{g.s.}$ &0.2 &8 &0.057 &1.3$\times10^8$\\
\hline
\hline
\end{tabular}
\caption{Parameters of the ${}^A_{Z_i}$X$(\gamma, \pi^-){}^A_Z\textrm{X}'$ reaction for light nuclei. Data for the free $\pi^-$ production at threshold ($\sigma_{p=0}$) in the third column are from \cite{Bernstein1976_pion_threshold_12C_12N,Bosted1979_near_threshold_cross-section,Singham1981_threshold_crosec, Min1976_threshold_11C,Epstein1978_threshold_12N,DeCarlo1980_threshold_14O,Singham1979_threshold_7Li_12C}. Here $^A_Z\textrm{X}'_{g.s.}$ means the final nucleus is in the ground state. $\sigma_0$ is the resonant cross-section for bound $\pi^-$ production. The last column gives the production rate of pionic atom $^A_Z\textrm{X}'$ in $1s$ state expected at the GF by use of Eq.\,\eqref{Eq:max_production_rate_from_free_pion} (see Sec.\,\ref{Subsec:GF_pion_NT}).}
\label{tab:sigmap_light_nuclei}
\end{table}
%JJ{More exactly, $\sigma_{p=0}$ data on Z=4,7,8 are read from figures in the review paper Ref.\,\cite{Singham1981_threshold_crosec}. $\sigma_{p=0}$ on Z=6 is from Ref.\,\cite{Min1976_threshold_11C}.
%(I checked $\sigma_{p=0}$ data. For the examples in the table,
%they are exactly what we need though the experimental results may have large uncertainty. As for $\sigma_{p}$ data for producing neutral pions, they are %also right. Actually, they are indeed what we want by definition of the coherent production process.)
%}
Photoproduction of free pions from heavy nuclei was studied by use of a bremsstrahlung-photon beam and radiochemical measurements, for example, in Refs.\,\cite{Sakamoto1989_51Cr_from_51V,Sakamoto1990_133Ba_from_133Cs,Blomqvist1978_197Hg_from_197Au,Sakamoto1999_complex_nuclei}.
According to Ref.\,\cite{Sakamoto1999_complex_nuclei}, yields for $(\gamma, \pi^-)$ from nuclei with $A > 40$ are almost $A$-independent at bremsstrahlung end-point energies $E_0$=250\,MeV, 400\,MeV and 800\,MeV. In Ref.\,\cite{Sakamoto1999_complex_nuclei}, the authors did not use quasi-monoenergetic gamma-rays with near-threshold energy and the final nuclei could be in excited states instead of only the ground state, so we can only take $\sigma_{p=0}$ to be $Z$-independent for nuclei with $Z\geq 20$ as a rough estimate.
%\JJ{
Calculations of $\sigma_p$ for the $^{197}$Au($\gamma,\pi^-$)$^{197}$Hg$_{g.s.}$ reaction in Ref.\,\cite{Blomqvist1978_197Hg_from_197Au} produce $\sigma_p\approx 1\,\mu$b at $p\approx92\,$MeV$/c$ while experimental data presented in Ref.\,\cite{Blomqvist1978_197Hg_from_197Au} show a significantly larger result, up to $\sigma_p\sim 100\,\mu$b. However, radiochemical measurements are slow, therefore all excited nuclei have enough time to decay to the ground state or to metastable isomeric state, so only these two outcomes, accumulated ground state or accumulated isomeric state, have been experimentally separated. %Ref.\,\cite{Blomqvist1978_197Hg_from_197Au} authors included contributions from several excited states which then decay fast to the ground state. So, the cross-section for producing final nuclei directly in the ground state should be smaller. But it is noted that in Ref.\,\cite{Blomqvist1978_197Hg_from_197Au} the experimental data are significantly larger than those expected from the theoretical calculation.
Therefore, we suggest $\sigma_{p=0}\approx 1\,\mu$b as an order-of-magnitude estimate for producing free pions from $^{197}$Au near threshold with produced $^{197}$Hg directly in its ground state. Together with the $Z$-independence noted above, we apply the estimate $\sigma_{p=0}(Z)\approx 1\,\mu$b %to $20\leq Z\leq 92$ cases, and further extrapolate this
to $9\leq Z\leq 92$ cases.
Cross-sections for $(\gamma, \pi^-)$ reaction can also be estimated using experimental data on $(\gamma, \pi^0)$ reaction.
% \JJ{An advantage of $(\gamma, \pi^0)$ reaction is that in photoproduction of $\pi^0$ with the final nucleus in the ground state same as the initial nucleus there is the coherent enhancement, leading to $A^2$-dependence of the production cross-sections.
% % neglecting the strong interaction between $pi^0$ and the nucleus \cite{Arends1986_pion0_delta_region,Boffi1991_coherent_pion0}.
% Cross-sections due to other channels with excited nuclei in the final states, which are not enhanced, are smaller and may be negligible.}
An advantage of $(\gamma, \pi^0)$ reaction is that in photoproduction of $\pi^0$ with the final nucleus in the ground state same as the initial nucleus, there is coherent enhancement $\sim A$ in the production cross-sections. This means that in the near-threshold area we do not need to worry about other channels with excited nuclei in the final states, which are not enhanced. Coherent enhancement is strongly suppressed in photoproduction of $\pi^-$ with a neutron in the initial nucleus converted into a proton. This gives a factor $(A-Z)/A^2\approx 1/(4Z)$ in the ratio of $\sigma_p(\pi^-)/\sigma_p(\pi^0)$ \footnote{There are indications that, due to pion repulsion from the nucleus only surface nucleons contribute to the coherent cross-section, see, for example, Ref.\,\cite{Boffi1991_coherent_pion0}. If this is the case, we may estimate the suppression factor as $1/(4Z)^{2/3}$.}.
%(If we don't use near-threshold data on $\pi^0$, this method may be meaningless since there could already be data for producing $\pi^-$ ``far from" the threshold.)
Coulomb interaction between the nucleus and $\pi^-$ adds the Sommerfeld factor \cite{TZARA1970_photoproduction_pion} to the cross-section near the threshold. Therefore, for producing $\pi^-$ and $\pi^0$ at the same momentum we get
\begin{equation}
\label{Eq:neutral_pion}
\sigma_p(\pi^-)\approx \frac{1}{4Z} \frac{2\pi f}{1-\exp{(-2\pi f)}}\sigma_p(\pi^0),
\end{equation}
%%%\end{equation}
%According to Eq.\,\eqref{Eq:crosec_bound_pion_NT}, $\sigma_{p=0}=\sigma_p K$, $K= 1-\exp{(-2\pi f)}$ where
%$f= Z\alpha mc/p =Z\alpha \sqrt{140\,\textrm{MeV}/(2E_\pi)}$.
%\JJ{
where $f= Z\alpha mc/p$ and $m$ is the mass of $\pi^-$.
Note that in this relation momenta of $\pi^-$ and $\pi^0$ are the same but photon energies for producing them are different since masses of $\pi^-$ and $\pi^0$ are different.
With cross-section data for coherent photoproduction of $\pi^0$ read from figures in Ref.\,\cite{Boffi1991_coherent_pion0} (at $E_\gamma\approx 146$\,MeV), we get for producing $\pi^-$ at threshold [$\sigma_{p=0}(\pi^-)= \sigma_p(\pi^-) K$, see Eq.\,(\ref{Eq:crosec_bound_pion_v1_NT})] $\sigma_{p=0}(Z=6)\approx 1.0$\,$\mu$b,
%(at the same order of magnitude of the experimental $4\,\mu$b),
$\sigma_{p=0}(Z=20)\approx 3.2$\,$\mu$b, and $\sigma_{p=0}(Z=82)\approx 3.5$\,$\mu$b.
%}
Within expected order-of-magnitude accuracy these estimates are comparable to the estimate presented above, $\sigma_{p=0}(Z)\approx 1\,\mu$b for $9\leq Z\leq 92$.
% where $f= Z\alpha mc/p =Z\alpha \sqrt{mc^2/(2T_\pi)}$ and $T_\pi$ is the kinetic energy of the free pion. \VF{Note that in this relation momenta of $\pi^-$ and $\pi^0$ are the same but the photon energies are different since masses of $\pi^-$ and $\pi^0$ are different.}
%When relating the cross-section of producing $\pi^-$ to that of $\pi^0$ using (exp
%\begin{equation}
% \sigma_p(\pi^-)= \frac{2\pi f}{1-\exp{(-2\pi f)}}\sigma_p(\pi^0),
%\end{equation}
%we have
%\begin{equation}
% \sigma_{p=0}= 2\pi f \sigma_p(\pi^0)=2\pi Z\alpha \sqrt{\frac{mc^2}{2 T_\pi}} \sigma_p(\pi^0).
%\end{equation}
%With data on $\sigma_p(\pi^0)$ of producing $\pi^0$ using $E_\gamma= 200$\,MeV photons (roughly, $T_\pi= 65$\,MeV) from Ref.\,\cite{Krusche2002_pion0}, we would get for $\pi^-$ $\sigma_{p=0}(Z=6)\approx 86$\,$\mu$b (much larger than the experimental datum $4\mu$b), $\sigma_{p=0}(Z=20)\approx 697$\,$\mu$b, $\sigma_{p=0}(Z=41)\approx 2600$\,$\mu$b, $\sigma_{p=0}(Z=82)\approx 7400$\,$\mu$b. The $\pi^-$ cross-section
%$\sigma_{p=0}$ derived in this way is larger than what we expect from the upper limit (2$\%$) of $\Gamma_\gamma/ \Gamma_{tot}$, especially for heavy nuclei.
%It is noted that for photoproduction of $\pi^0$ with the final nucleus in the ground state same as the initial nucleus there is the coherent enhancement with approximately $A^2$-dependence of the production cross-sections. This enhancement may be strongly suppressed in photoproduction of $\pi^-$ with a neutron in the initial nucleus converted into a proton. Therefore, we modify the original estimate by introducing an extra factor $(A-Z)/A^2\approx 1/(4Z)$ as
% With cross-section data for producing $\pi^0$ near the threshold using $E_\gamma\approx 146$\,MeV photons ($T_{\pi^0}\approx 11$\,MeV)
% read from figures in Ref.\,\cite{Boffi1991_coherent_pion0}, we get for $\pi^-$ $\sigma_{p=0}(Z=6)\approx 2.1$\,$\mu$b (close to the experimental $4\,\mu$b), $\sigma_{p=0}(Z=20)\approx 3.5$\,$\mu$b, and $\sigma_{p=0}(Z=82)\approx 3.5$\,$\mu$b.
Using $\sigma_{p=0}\approx 1\,\mu$b (for $9\leq Z\leq 92$) and $\Gamma_{tot}(Z,1)$ (Sec.\,\ref{Subsec:pion_absorption_width}), we can get the resonant cross-section $\sigma_0$, the width ratio $\Gamma_\gamma/ \Gamma_{tot}$ and the partial width $\Gamma_\gamma$ as shown in Figs.\,\ref{fig:sigma0}, \ref{fig:branching_ratio} and \ref{fig:partial_width}, respectively.
Figs.\,\ref{fig:sigma0} and \ref{fig:branching_ratio}
look similar since $\sigma_0$ is proportional to $\Gamma_\gamma/ \Gamma_{tot}$ according to Eq.\,\eqref{Eq:res_cross_section}. For light nuclei with $Z\leq 8$, $\sigma_0$ and $\Gamma_\gamma/ \Gamma_{tot}$ are obtained with corresponding $\sigma_{p=0}$ data - see Table\,\ref{tab:sigmap_light_nuclei}.
%\JJ{$\Gamma_\gamma(Z,n)\approx 0.037 Z^2/n^3 \sigma_{p=0}$/($1\mu$b)\,eV}
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{sigma0.pdf}
\caption{Resonant cross-section $\sigma_0$ for producing $ns$ pionic states with final nuclei with $9\leq Z\leq 92$.}
\label{fig:sigma0}
\end{figure}
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{1000branching_ratio.pdf}
\caption{Width ratio $\Gamma_\gamma/ \Gamma_{tot}$ (multiplied by $10^3$) of $ns$ pionic states with final nuclei with $9\leq Z\leq 92$.}
\label{fig:branching_ratio}
\end{figure}
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{Gamma_gamma.pdf}
\caption{Partial width $\Gamma_\gamma$ of the $1s$ pionic state with final nuclei with $9\leq Z\leq 92$.}
\label{fig:partial_width}
\end{figure}
With known $\sigma_{p=0}$, integrated cross-sections for producing pionic atoms in $ns$ states can be derived directly using Eq.\,\eqref{Eq:crosec_bound_pion_v1_NT}. Then we can discuss production rates expected at the GF \cite{Krasny2015,Budker2020_AdP_GF}.
%(see, Refs.\,\cite{Krasny2015,Budker2020_AdP_GF}).
\subsection{Production rate with GF photons}
\label{Subsec:GF_pion_NT}
At the GF, laser photons will be sent
onto an incoming beam of partially stripped ions (PSI) with a high relativistic factor $\gamma$. Here $\gamma= 1/\sqrt{1-\beta^2}$ can be set between $\gamma\approx 200$ and $\gamma\approx 3000$ (in the case of the LHC ring), where $\beta=v/c\approx 1$ and $v$ is the speed of the ions. In the ion frame, the energy of incident photons is boosted by a factor of $\approx2\gamma$ and atomic-level transitions in the PSI can be excited. The secondary photons emitted by the excited ions, are seen in the lab frame to be predominantly emitted in the direction of propagation of the PSI. Their energy is boosted by another factor of up to $\approx 2\gamma$, and can be tuned by changing $\gamma$ and the energy of the laser photons.
Neglecting the energy spread of the ions which is $\approx 10^{-4}$ and can be reduced further to $\approx 10^{-6}$ as projected with the application of laser cooling of the PSI, secondary photons at the GF are almost uniformly distributed up to the maximal energy. It will be tuned to $E_{\gamma}\approx 140$\,MeV for producing pionic atoms. Because the energy of
those photons is correlated with their propagation angle, after proper collimation, the effective photon flux can be estimated as
\begin{equation}
j_{eff}\approx j \frac{\Gamma_{ph}}{E_{\gamma}},
\end{equation}
where $j= 10^{17}$\,s$^{-1}$ is the expected total photon flux at the GF and $\Gamma_{ph}$ is the photon-distribution width of GF photons after the collimation. We can see that a reduction in $\Gamma_{ph}$ comes at a cost of the effective photon flux.
After the collimation, GF photons are nearly uniformly distributed between $E_{\gamma}-\Gamma_{ph}/2$ and $E_{\gamma}+\Gamma_{ph}/2$.
Considering the photon-distribution width, the effective photon-absorption cross-section of the initial nucleus for creating pionic atoms in the $ns$ state with the atomic number $Z$ is
%[see Eq.\,(\ref{Eq:crosec_energy_dependence_NT})]
\begin{equation}
\label{cutoff_NT}
\sigma_{eff}(Z_i,n)=
%\int_{E_{\gamma}-\Gamma_{ph}/2}^{E_{\gamma}+\Gamma_{ph}/2}
%\int \sigma(E)g(E) dE=
\sigma_0 \frac{\Gamma_{tot}(Z,n)}{\Gamma_{ph}}\arctan{\frac{\Gamma_{ph}}{\Gamma_{tot}(Z,n)}}.
\end{equation}
To make better use of photon fluxes at the GF, i.e., get larger $j_{eff}\sigma_{eff}$, we should choose $\Gamma_{ph}\geq \Gamma_{tot}(Z,n)$, which leads to $\sigma_{eff}(Z_i,n)\approx \sigma_0\Gamma_{tot}(Z,n)/\Gamma_{ph}$.
When producing pionic atoms with 140\,MeV photons, there are other competing processes. Various contributions to the photon attenuation cross-section are shown in Ref.\,\cite{Zyla2020_Particle_Data}. In our case, pair production in the nuclear field and Compton scattering off electrons are dominant.
According to Ref.\,\cite{Berestetskii_QED},
the cross-section due to scattering of 140\,MeV photons by electrons is
\begin{equation}
\sigma_{scat} \approx 2\pi Z_i r_e^2 \frac{m_e c^2}{2E_{\gamma}}\left(\ln{\frac{2E_{\gamma}}{m_e c^2}}+\frac{1}{2}\right),
\end{equation}
and the cross-section due to pair production in the nuclear field is
\begin{equation}
\label{Eq:nuclear_pair_production_NT}
\sigma_{pp} \approx \frac{28}{9} Z_i^2 \alpha r_e^2 \left(\ln{\frac{2E_{\gamma}}{m_e c^2}}-\frac{109}{42}\right),
\end{equation}
where
$r_e = 2.8$\,fm is the classical electron radius, and $m_ec^2=0.511$\,MeV is the rest energy of an electron.
Thus, we estimate the total cross-section as
\begin{equation}
%\label{Eq:total_attenuation}
\sigma_{tot}(Z_i)\approx \sigma_{scat}+\sigma_{pp}\approx 0.6\times10^{-26}(Z_i^2+Z_i)\,\textrm{cm}^2.
\end{equation}
When all GF photons (after collimation) are absorbed in the target, the maximal production rate is
\begin{equation}
\label{Eq:max_production_rate_NT}
%\begin{split}
p_{max}(Z,n) \approx j_{eff} \frac{\sigma_{eff}(Z_i,n)}{\sigma_{tot}(Z_i)}= j \frac{\Gamma_{tot}(Z,n)}{E_{\gamma}}\frac{\sigma_0}{\sigma_{tot}(Z_i)}.
\end{equation}
Substituting $\sigma_0$ from Eq.\,\eqref{Eq:res_cross_section} or the integrated cross-section from Eq.\,\eqref{Eq:crosec_bound_pion_NT}, we get %respectively
\begin{equation}
\label{Eq:max_production_rate_from_partial_width}
p_{max}(Z,n) \approx
2 \pi\frac{2 J+1}{2 I+1}\frac{\Gamma_{\gamma}(Z,n)}{E_{\gamma}}
\frac{(\hbar c/E_{\gamma})^2}{\sigma_{tot}(Z_i)}j
\end{equation}
and
\begin{equation}
\label{Eq:max_production_rate_from_free_pion}
p_{max}(Z,n)\approx \frac{2}{\pi} \frac{\sigma_{p=0}(Z)}{\sigma_{tot}(Z_i)} \frac{Z^2 \alpha^2}{n^3}j\approx 5.6\times10^8 \frac{\sigma_{p=0}(Z)/1\mu\textrm{b}}{n^3}\,\textrm{s}^{-1}
\end{equation}
%with $\sigma_{p=0}$ in the unit of $\mu$b
\footnote{There is an extra coefficient $Z/(Z-1)$ in Eq.\,\eqref{Eq:max_production_rate_from_free_pion} that should be taken into account for light nuclei.}.
The production rate obtained with Eq.\,\eqref{Eq:max_production_rate_from_free_pion} is proportional to $\sigma_{p=0}$.
With the approximation
$\sigma_{p=0}(Z)\approx 1\,\mu$b for $9\leq Z\leq 92$, we get
$p_{max}(Z,1)\approx 5.6\times10^8$\,s$^{-1}$. Production rates for light nuclei in $1s$ states are presented in Table\,\ref{tab:sigmap_light_nuclei}.
%Victor
%We can also use Eq.\,\eqref{Eq:max_production_rate_NT} and the rough estimate $\Gamma_\gamma/ \Gamma_{tot}\leq \alpha$ as noted in Sec.\,\ref{Subsec:monochromatic_cross-section} to estimate the production rates. In this case, we need the total width of the pionic state. Widths for $ns$ pionic states are discussed in Sec.\,\ref{Subsec:pion_absorption_width}.
%Figure\,\ref{fig:ratio_cross_section_new} shows an estimate of the maximal production rate $p_{max}(Z,n)$ of pionic atoms in some low $ns$ states with $9\leq Z\leq 92$.
The production rate could reach up to $\sim 10^{10}$ per second.
Such high production rate of pionic atoms would be a significant improvement compared with other production methods (see, for example, Ref.\,\cite{Hori2020_spectra_helium} where $\sim 10^5$ pionic helium atoms per second are produced at a 590\,MeV proton facility).
%\begin{figure}[!htpb]\centering
% \includegraphics[width=1.0\linewidth]{pion_production_rate_Z_from_9.pdf}
% \caption{Estimate of the maximal production rate of pionic atoms in $1s$, $2s$, $3s$ and $6s$ states utilizing photons from the GF, see %Eq.\,\eqref{Eq:max_production_rate_NT}.
% %The maximal production rate is $p_{max}\approx j\sigma_{eff}(Z_i,n)/ \sigma_{tot}(Z_i)$, where the expected photon flux at the GF is $j = %10^{17}$\,s$^{-1}$.
% }
% \label{fig:ratio_cross_section_new}
%\end{figure}
To suppress the background induced by high-energy photons, $\Gamma_{ph}$ should be reduced by collimation but remain larger than $\Gamma_{tot}$ to obtain high production rates (the total widths for $ns$ pionic states $\Gamma_{tot}$ are discussed in Sec.\,\ref{Subsec:pion_absorption_width}). Therefore, $\Gamma_{ph}=\Gamma_{tot}$ is a good choice of $\Gamma_{ph}$ except when $\Gamma_{tot}$ is smaller than the lower limit of $\Gamma_{ph}$ which could be $\approx$140\,eV at the GF. It is also possible to produce pionic atoms in various bound states at the same time utilizing more photons and achieving even higher total production rates.
To realize this we can tune
%\usepackage{}$\Gamma_{ph}\approx \Gamma_{tot}(Z,1)/2 + E_{1s}(Z)\approx E_{1s}(Z)$
$\Gamma_{ph}\approx E_{1s}(Z)$, where $E_{1s}$ is the $1s$ binding energy of the pionic atom (see, for example, Refs.\,\cite{Toki1989_pion_width_1s,Friedman2007_widths}).
%\subsection{Photoproduction of kaonic atoms}
%Kaonic atoms have been less studied than pionic atoms, so this problem may be especially interesting.
%The rest energy of negative kaons is $m_{K^-}c^2=494$\,MeV, so we need much higher photon energies than those for pionic atoms.
%This may, in principle, be achieved with higher-energy laser photons. At the GF, the energy of secondary photons is $\approx 4\gamma^2$ times larger than that of primary photons with $\gamma$ up to 3000. When $\gamma=3000$, primary photons should have energies of 13.7\,eV to get 494\,MeV secondary photons.
%The basic theory for photoproduction of kaonic atoms is similar to that for pionic atoms mentioned above though the numbers are different.
%see, for example, \cite{Friedman2007_widths}).
%\JJ{
%The maximal production rate $p_{max}(Z,n)$ at the GF can be estimated using the first part of Eq.\,\eqref{Eq:max_production_rate_from_free_pion}. Now, when applying a beam of GF photons with energies of $E_{\gamma}\approx 500$\,MeV,
%}
%the total cross-section for photon attenuation is dominated by pair production in the nuclear field
%\cite{Zyla2020_Particle_Data}. With Eq.\,\eqref{Eq:nuclear_pair_production_NT}, we estimate the total cross-section as
%\begin{equation}
%\label{Eq:total_attenuation_kaon_NT}
% \sigma_{tot}(Z_i)\approx \sigma_{pp}\approx 0.90\times10^{-26}Z_i^2\,\textrm{cm}^2.
%\end{equation}
%\JJ{
%So we get\\
%$p_{max}(Z,n)\approx 3.7\times10^8 \sigma_{p=0}(Z)(\mu\textrm{b})/n^3$\,s$^{-1}$ \footnote{There is an extra coefficient $[Z/(Z-1)]^2$ that should be taken into account for light nuclei.}.
%with $\sigma_{p=0}(Z)$ expressed in the unit of $\mu$b.
%The maximal production rates can be obtained directly if data for producing free kaons near threshold ($\sigma_{p\approx0}$) were available.
%}
%\JJ{
%If we use Eq.\,\eqref{Eq:max_production_rate_NT} together with the rough estimate $\Gamma_\gamma/ \Gamma_{tot}\approx \alpha$ (at least suitable for light atoms) noted in Sec.\,\ref{Subsec:monochromatic_cross-section} and the total width of the kaonic state discussed in Sec.\,\ref{Subsec:kaon_absorption_width} (Or we can move this to Appendix), we can get rough estimate of the production rates.
%Figure\,\ref{fig:Kaon_ratio_cross_section} shows an estimate of the maximal production rate $p_{max}(Z,n)$ of kaonic atoms in some low $ns$ states, which is up to over $10^{9}$ per second.
%}
%As for the choice of the energy-distribution width %$\Gamma_{ph}$ of GF photons, considerations are similar with the case of pionic atoms (see Sec.\,\ref{Subsec:GF_pion_NT}).
\subsection{Estimate of the pion absorption width}
\label{Subsec:pion_absorption_width}
% The energy distribution of photons from the GF has a finite width $\Gamma_{ph}$.\JJ{(We may not mention $\Gamma_{ph}$ of GF photons here.) } We should compare this width with
The width of the atomic energy level $\Gamma_{tot}$ for $ns$ states is dominated by pion absorption.
As an approximation we may assume that $\Gamma_{tot}$ is proportional to $Z^4/n^3$, where $Z^3/n^3$ is from the probability density of the $ns$-wavefunction at the origin , $\psi_{ns}^2(0)$, and the extra factor $Z$ appears if each proton may absorb the pion.
According to Ref.\,\cite{Backenstoss1974_X-ray_helium}, $\Gamma_{tot}$ of $(^4\textrm{He} +\pi^-)_{1s}$ is 45\,eV. This gives us an estimate of the absorption width in the vicinity of $Z= 2$:
\begin{equation}
\label{Eq:Width_approx}
\Gamma_{tot}(Z,n)\approx 45\frac{Z^4}{16n^3}\,\textrm{eV}.
\end{equation}
However, this formula strongly overestimates pion absorption width for $Z \gg 1$. Due to the repulsive strong interaction between the pion in $1s$ state and the nucleus,
%there is saturation of the width $\Gamma_{tot}$, i.e., the increase of $\Gamma_{tot}$ can be small
%or even negative
the increase of $\Gamma_{tot}$ could be small as $Z$ increases \cite{Friedman_1985_width_heavy_atoms}. Thus the formula above needs to be adjusted.
By fitting data for the $1s$ level width of pionic atoms with $Z$ up to 82 from Refs.\,\cite{Toki1989_pion_width_1s,Konijn1990_widths_nl,Friedman2007_widths} (see Fig.\,\ref{fig:width_fitting}), we approximate the pion absorption width of the $1s$ level as
\begin{equation}
\label{Eq:1s_width}
\Gamma_{tot}(Z,n=1)\approx \frac{16Z^4}{Z^3+70Z^2-950Z+11000}\,\textrm{keV}.
\end{equation}
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{Pion_width_fitting.pdf}
\caption{Fitting of $1s$ width data from Refs.\,\cite{Konijn1990_widths_nl,Friedman2007_widths} shown as red dots on the graph. When width data for different isotopes with the same atomic number $Z$ are given, the average value of the widths is used for the fitting. Data for $^{40}$Ca and $^{90}$Zr are from Ref.\,\cite{Toki1989_pion_width_1s}. $1s$ width for $^4$He is from Ref.\,\cite{Backenstoss1974_X-ray_helium}. The yellow curve and blue curve represent Eq.\,\eqref{Eq:Width_approx} and Eq.\,\eqref{Eq:1s_width}, respectively.}
\label{fig:width_fitting}
\end{figure}
For $ns$ states with $n\geq2$, which have smaller widths and longer lifetimes,
%facilitating experimental detection of produced pionic atoms,
the pion absorption width can be estimated as $\Gamma_{ns}=\Gamma_{1s}/n^3$ %\JJ{$\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)/n^3$, we may improve this paragraph literally if we want}.
The $1/n^3$ dependence for the width is violated in heavy atoms since the pion in the $1s$ state of heavy atoms has orbital radius close to the nuclear radius and the nucleus cannot be approximated as a point-like center.
For pionic atoms in $ns$ states with $Z/n^{3/2} \leq 37$ the radius of the corresponding Bohr orbit [$\approx 200 n^2/Z$ fm] is larger than the nucleus [$\approx1.2(2.5Z)^{1/3}$\,fm]. For $Z\leq 37$
we estimate the width $\Gamma_{tot}(Z,n)$ as
% For pionic atoms in $ns$ states with $Z/n^{3/2} \leq 37$ the radius of the corresponding Bohr orbit (200 $n^2/Z$ fm) is larger than the nucleus [$\approx1.2(2.5Z)^{1/3}$\,fm]. For $Z<37$
\begin{equation}
\label{Eq:ns_width}
\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)\frac{1}{n^3}.
\end{equation}
For pionic atoms with $Z >37$ this estimate of the total width $\Gamma_{ns}$ is not applicable since the $1/n^3$ dependence follows from the Coulomb wavefunctions but $1s$ function in this case is not the Coulomb one. Therefore, for $Z >37$ we may use Eq.\,\eqref{Eq:ns_width} only as an order-of-magnitude estimate for $\Gamma_{tot}(Z,n)$.
% We see that the width of low $ns$ states significantly exceeds the expected ultimate photon distribution width $\Gamma_{ph}\approx 140$\,eV at the GF (see Sec.\,\ref{Subsec:GF_photon}). \JJ{( We do not need to compare the widths if we plan to tune $\Gamma_{ph}$ according to $\Gamma_{tot}(Z,n)$.)} For example, width data for the $ns$ states with $n = 1,2,3$ of $^{208}$Pb are presented in Ref.\,\cite{Konijn1990_widths_nl}. They are all much larger than $140$\,eV.
Besides $ns$ states, pionic atoms in other $nl$ states can also be produced.
When the radii of corresponding Bohr orbits exceed the nuclear size,
the widths of these states can be estimated as $\Gamma_{nl}/\Gamma_{ns}\approx R_{nl}(r_0)^2/ R_{ns}(r_0)^2$. Here $R_{nl}$ is the radial wavefunction of hydrogen-like atoms, and $R_{nl}(r_0)^2/ R_{ns}(r_0)^2$ gives the ratio of probabilities of a pion in $nl$ and $ns$ state to occur on the nuclear surface. For an estimate, $\Gamma_{nl}/\Gamma_{ns}\sim (Z r_0/a_0)^{2l}$, where $r_0\approx 1.2 (2.5Z)^{1/3}$\,fm is the nuclear radius, and $a_0\approx 200$\,fm is the $1s$ Bohr radius of the pionic hydrogen atom. There are some $nl$-level width data for pionic atoms in Ref.\,\cite{Friedman2007_widths}, and we see that the simple estimate $\Gamma_{nl}/\Gamma_{ns}\sim (Z r_0/a_0)^{2l}$ is consistent with the data for $2p$-level widths within one order of magnitude.
The widths for $3d$ and $4f$ states should be estimated using more accurate expression $\Gamma_{nl}/\Gamma_{ns}\approx R_{nl}(r_0)^2/ R_{ns}(r_0)^2$.
% When the radii of corresponding Bohr orbits exceed the nuclear size,
% %\JJ{(For nl($l>0$) states in atoms with $Z<100$ it is true)}
% the widths of these states can be estimated as $\Gamma_{nl}/\Gamma_{ns}\approx P_{nl}/P_{ns}$.
% Here $P_{nl}$
% %$P_{nl}=\int_{r\leq r_0} R_{nl}(r) r^2 dr$
% is the probability for a pion in the $nl$ state to be inside the nucleus.
% %$r_0\approx 1.2 (2.5Z)^{1/3}$\,fm is the nuclear radius and $R_{nl}(r)$ is the radial wavefunction of the pionic hydrogen atom.
% %\JJ{(For higher $nl$ states we need be careful whether the pion absorption width is still dominant or not.)}
% %For axample, for $2p$ states, $\Gamma_{2p}/\Gamma_{2s}\approx (1/20)(Z r_0/a_0)^2$ %$\Gamma_{2p}/\Gamma_{2s}\approx (1/20)(Z r_0/a_0)^2\approx 5\times10^{-6}Z^2$
% For an estimate, $\Gamma_{nl}/\Gamma_{ns}\sim (Z r_0/a_0)^{2l}$, where $r_0\approx 1.2 (2.5Z)^{1/3}$\,fm is the nuclear radius, and $a_0\approx 200$\,fm is the $1s$ Bohr radius of the pionic hydrogen atom.
% \VF{There are some $nl$-level width data for pionic atoms in Ref.\,\cite{Friedman2007_widths}, and we see that simple estimate $\Gamma_{nl}/\Gamma_{ns}\sim (Z r_0/a_0)^{2l}$ is consistent with the data for $2p$-level widths within an order-of-magnitude.
% The widths for $3d$ and $4f$ states should be estimated using more accurate expression $\Gamma_{nl}/\Gamma_{ns}\approx P_{nl}/P_{ns}$.
% %are significantly overestimated using our approximation, which could be due to the width data being for heavy nuclei while our estimate, in principle, only works for light nuclei.
% }
%\JJ{(Where is $\Gamma_{nl}/\Gamma_{ns}\sim (Z r_0/a_0)^{2l}$ from?)}
% \JJ{We may also talk about deeply bound $2p$ states. For $Z<92$ we are intereted in, $2p$ states may be approximated as Coulomb bound states due to their Bohr orbital radius larger than the nucleus.}
% \DB{Note that when the pionic-atom Coulomb radius is comparable to the nuclear size, pionic atoms can be produced not only in $ns$ states, but also $np$ and higher orbitals. When the Coulomb radius exceeds the nuclear size, the relative suppression factor for production of an $nl$ state relative to $ns$ can be estimated as}
% \begin{equation}
Atomic levels also have a radiative width. Following the calculation for hydrogen \cite{BetheSalpeter} we obtain the following estimate of the radiative width for the pionic $ns$-levels with $n>2$: $\Gamma_{rad} \approx 10^{-4} Z^4/n^3$\,eV. This width is much smaller than the pion-absorption width. Note that for large $n$ the width due to emission of Auger electrons may exceed the radiative width. However, for $n <\sqrt{m_\pi/ m_e}\approx 16$ the pionic atom de-excites primarily by an X-ray cascade, which has electric dipole character to a good approximation - see, for example, Ref.\,\cite{Ericson1988}.
%\subsection{\JJ{Estimate of the kaon absorption width}}
%\label{Subsec:kaon_absorption_width}
%By fitting data for the $1s$ level width of four kaonic atoms: deuterium, carbon, nickel and lead from Refs.\,\cite{Friedman1999_kaon_antiproton,Friedman1999_deep_bound_kaon,Curceanu2019_kaonic_width} (see Fig.\,\ref{fig:kaon_width_fitting_NT}),
%\JJ{ do we need to present the width numbers here?}
%we approximate the kaon absorption width of the $1s$ level as
%\begin{equation}
%\label{Eq:kaon_1s_width_NT}
% \Gamma_{tot}(Z,n=1)\approx \frac{20Z^4}{Z^3-9Z^2+100Z-60}\,\textrm{keV}.
%\end{equation}
%
%\begin{figure}[!htpb]\centering
% \includegraphics[width=1.0\linewidth]{Kaon_width_fitting.pdf}
% \caption{Fitting of $1s$ width data (red dots on the graph) for kaonic atoms. The width data for carbon, nickel and lead are from Refs.\,\cite{Friedman1999_deep_bound_kaon,Friedman1999_kaon_antiproton}; the $1s$ width of kaonic deuterium is from Ref.\,\cite{Curceanu2019_kaonic_width}. }
% \label{fig:kaon_width_fitting_NT}
%\end{figure}
%
%The Coulomb radius of the kaonic state is 3.5 times smaller than that of the corresponding pionic state, therefore, only $ns$ states with $Z/n^{3/2} \leq 14$ may be approximately treated as the Coulomb bound states.
%For $Z\leq 14$ the kaon absorption width for higher $ns$ states can be estimated
%as $\Gamma_{tot}(Z,n) \approx \Gamma_{tot}(Z,1)/n^3$.
%For kaonic atoms with $Z > 14$ this estimate of the total width $\Gamma_{tot}(Z,n)$ is not applicable since the $1/n^3$ dependence in Eq.\,\eqref{Eq:ns_width} follows from the Coulomb wavefunctions but the $1s$ function in this case is not the Coulomb one. Therefore, we may use Eq.\,\eqref{Eq:ns_width} for $Z > 14$ as an order-of-magnitude estimate of $\Gamma_{tot}(Z,n)$ only.
%
%Similar to the case of pionic atoms (Sec.\,\ref{Subsec:pion_absorption_width}), besides $ns$ states, kaonic atoms in other $nl$ states can also be produced. The estimate of widths of $nl$ states is similar with the case of pionic atoms except now for kaonic atoms $a_0\approx 55$\,fm \footnote{Note that for light kaonic atoms, $a_0$ should be obtained considering the reduced mass of the kaonic atom.}.
\section{Conclusion}
Estimates presented above indicate that the direct resonance photoproduction of pionic atoms in $ns$ states at the Gamma Factory would provide several orders of magnitude larger number of these atoms than any existing facility. The $ns$ states are especially sensitive and important for the study of the interaction between the pion and nucleons, nuclear structure and nuclear forces forming the structure, including, for example, the neutron skin problem related to the prediction of the neutron-star equation of state and maximal neutron-star mass.
This study of the strong interaction effects is, in fact, the main aim of
the pion-atom production. Production of low-lying states in heavy elements, so called deeply bound pionic states, may be especially interesting, since they can provide information on the pion behavior inside the nucleus and cannot be populated via an X-ray cascade in the capture process of a free pion due to the strong pion absorption in higher states. Note that the population of the higher $ns$ states via an X-ray cascade is also suppressed since the cascade tends to
populate the highest orbital angular momentum (circular) states, $l=n-1$ (see, for example, Ref.\,\cite{Friedman2007_widths}).
% According to the estimate above, the extremely narrow energy spread, $\approx 10^{-6}$, may be unnecessary for photoproduction of pionic and kaonic atoms with broad width, i.e., $\Gamma_{tot} > \Gamma_{ph}$.
% Without laser cooling at the GF, the energy spread of GF photons will be $\approx 10^{-4}$. It gives the photon-distribution width $\Gamma_{ph}\approx 14$\,KeV for creating pionic atoms and $\Gamma_{ph}\approx 50$\,KeV for creating kaonic atoms.
% %According to estimates presented above,
% For producing $ns$ states in pionic atoms, $\Gamma_{tot}(Z,n) > \Gamma_{ph}$ is satisfied when $\sqrt[3]{0.4Z^4/(Z^3-20Z^2+400Z+800)} > n$.
% For kaonic atoms, it is $\sqrt[3]{0.4Z^4/(Z^3-9Z^2+100Z-60)} > n$.
% In these cases, the maximal production rate $p_{max} \approx j\sigma_0/\sigma_{tot}(Z_i)$, which is the same as that obtained with laser cooling.
% In more general cases, $p_{max}\approx j\sigma_{eff}(Z_i,n)/\sigma_{tot}(Z_i)$.
% %, where $\sigma_{eff}(Z_i,n)$ is given in Eq.\,\eqref{cutoff}.
% With broader photon-distribution width, the effective cross-section %$\sigma_{eff}(Z_i,n)$
% as well as the production rate would be smaller. In the ``worst" cases where $\Gamma_{ph} > 100\Gamma_{tot}(Z,n)$, the maximal production rate will be two orders of magnitude smaller than that obtained with laser cooling.
\section*{Acknowledgments}
The authors are grateful to Carlos Bertulani, Catalina Curceanu, Eli Friedman, Avraham Gal, Mikhail Gorshteyn, Masaki Hori, M. Witold Krasny, and Vladimir G.\,Zelevinsky for helpful information and discussions. The work of VF is supported by the Australian Research Council grants DP190100974 and DP20010015 and the Gutenberg Fellowship. The work of DB was supported in part by the DFG Project ID 390831469: EXC 2118 (PRISMA+ Cluster of Excellence).
\bibliography{Pionic_Atoms}
\end{document}
%------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\subsection{Production rate with GF photons}
\label{Subsec:GF_pion}
In the GF (see, Ref.\,\cite{Budker2020_AdP_GF}), light from a laser beam will be sent
onto an incoming beam of partially stripped ions (PSI) with a high relativistic factor $\gamma$. Here $\gamma= 1/\sqrt{1-\beta^2}$ can be up to $\approx 3000$ (in the case of the LHC ring), where $\beta=v/c\approx 1$ and $v$ is the speed of the ions. In the ion frame, the energy of incident photons is boosted by a factor of $\approx2\gamma$ and atomic-level transitions in the PSI can be excited. The secondary photons emitted by the excited ions, are seen in the lab frame to be predominantly emitted in the direction of propagation of the PSI. Their energy is boosted by another factor of up to $\approx 2\gamma$, and can be tuned by changing $\gamma$ and the energy of the laser photons.
Neglecting the energy spread of the ions which is $\approx 10^{-4}$ and can be reduced further to $\approx 10^{-6}$ as projected with the application of laser cooling of the PSI, the energy of secondary GF photons is almost uniformly distributed up to the maximal energy, i.e., $E_{\gamma}\approx 140$\,MeV for producing pionic atoms. Because the energy of the secondary GF photons is correlated with their propagation angle, after proper collimation, the effective photon flux can be estimated as
\begin{equation}
j_{eff}\approx j \frac{\Gamma_{ph}}{E_{\gamma}},
\end{equation}
where $j= 10^{17}$s$^{-1}$ is the expected total photon flux at the GF and $\Gamma_{ph}$ is the energy-distribution width of GF photons after the collimation. We can see that a reduction in $\Gamma_{ph}$ comes at a cost of the effective photon flux.
% When applying a beam of photons on a fixed target, the absorption cross-section for production of pionic atoms is
% \begin{equation}
% \sigma(E)=\sigma_0 \frac{(\Gamma_{tot}/2)^2}{(\Gamma_{tot}/2)^2+(E-E_{\gamma})^2},
% \end{equation}
% where $\sigma_0$ is the on-resonance cross-section defined in Eq.\,\eqref{Eq:res_cross_section}, $E$ is the photon energy.
%and $\Delta E\approx140$\,MeV is the energy resonant for creating pionic atoms.
After the collimation, the energy of GF photons can be assumed to be uniformly distributed between $E_{\gamma}-\Gamma_{ph}/2$ and $E_{\gamma}+\Gamma_{ph}/2$.
The dependence of the absorption cross-section on the energy of photons is
\begin{equation}
\label{Eq:crosec_energy_dependence}
\sigma(E)=\sigma_0 \frac{(\Gamma_{tot}/2)^2}{(\Gamma_{tot}/2)^2+(E-E_{\gamma})^2},
\end{equation}
where $\sigma_0$ is the on-resonance cross-section defined in Eq.\,\eqref{Eq:res_cross_section}, $E$ is the photon energy.
Considering the photon-distribution width, the effective photon-absorption cross-section of the initial nucleus for creating pionic atoms in the $ns$ state with the atomic number $Z$ is
\begin{equation}
\label{cutoff}
\sigma_{eff}(Z_i,n)=
%\int_{E_{\gamma}-\Gamma_{ph}/2}^{E_{\gamma}+\Gamma_{ph}/2}
%\int \sigma(E)g(E) dE=
\sigma_0 \frac{\Gamma_{tot}(Z,n)}{\Gamma_{ph}}\arctan{\frac{\Gamma_{ph}}{\Gamma_{tot}(Z,n)}}.
\end{equation}
To make better use of photon fluxes at the GF, i.e., get larger $j_{eff}\sigma_{eff}$, we should choose $\Gamma_{ph}\geq \Gamma_{tot}(Z,n)$, which leads to $\sigma_{eff}(Z_i,n)\approx \sigma_0\Gamma_{tot}(Z,n)/\Gamma_{ph}$.
The cross-section for the pion atom production may be related to the cross-section $\sigma_p$ for the free pion photoproduction for a small pion momentum $p$ \cite{TZARA1970_photoproduction_pion}:
\begin{equation}
\label{Eq:crosec_bound_pion}
\int \sigma_n dE=\frac{\pi}{2}\sigma_0 \Gamma_{tot}(Z,n)=\sigma_p \frac{(Z\alpha c)^2}{n^3} K m,
\end{equation}
where $K= 1-\exp{(-2\pi Z\alpha mc/p)}$. For the zero pion momentum $p$ the factor $K=1$. The result looks even simpler if we use $\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)/n^3$:
\begin{equation}
\label{Eq:crosec_bound_pion1}
\frac{\pi}{2}\sigma_0 \Gamma_{tot}(Z,1)=\sigma_p (Z\alpha c)^2 K m.
\end{equation}
We see that the resonance cross-section $\sigma_0$ for the $ns$ level production does not depend on $n$, i.e., as expected, $\Gamma_{\gamma}$ and $\Gamma_{tot}$ have the same $n$ dependence - see Eq. (\ref{Eq:res_cross_section}).
Note that Eqs. (\ref{Eq:crosec_bound_pion}) and (\ref{Eq:crosec_bound_pion1}) have been derived under assumption that the strong interaction affects wavefunctions of the bound pion and low energy free pion at the nucleus in approximately the same way, so this effect of the strong interaction cancels out in the ratio of the cross-sections. The accurate relation may actually be established for $ns$ levels with $n \gg 1 $ and small pion momentum cross-sections since both pion wavefunctions in this case are proportional to the zero energy pion wavefunction at small distance and differ only in the normalization coefficients which are defined by large distances where the interaction is the pure Coulomb one. As we noted above, the $1/n^3$ extrapolation to $1s$ level is reasonably accurate for $Z<37$. For $Z>37$ and small $n$ levels this is only an order of magnitude estimate.
The free pion photoproduction cross-section $\sigma_p$ has been measured for many nuclei.
\JJ{According to Refs.\,\cite{Singham1981_threshold_crosec, Min1976_threshold_11C,Epstein1978_threshold_12N,DeCarlo1980_threshold_14O,Singham1979_threshold_7Li_12C}, $\sigma_{p=0}$ is approximately
(4.6\,$\mu$b) for $^{7}\textrm{Li}(\gamma, \pi^-) ^{7}\textrm{Be}_{g.s.}$,
4\,$\mu$b for $^{11}\textrm{B}(\gamma, \pi^-) ^{11}\textrm{C}_{g.s.}$,
4\,$\mu$b for $^{12}\textrm{C}(\gamma, \pi^-) ^{12}\textrm{N}_{g.s.}$,
0.2\,$\mu$b for $^{14}\textrm{N}(\gamma, \pi^-) ^{14}\textrm{O}_{g.s.}$} (data for heavy nuclei?).
-----------------------------------------------------------
% \JJ{
% (Rough estimate of $\sigma_{p=0}$ for photoproduction of $\pi^-$ using data on $\pi^0$)
% According to Eq.\,\eqref{Eq:crosec_bound_pion1}, $\sigma_{p=0}=\sigma_p K$, $K= 1-\exp{(-2\pi f)}$ where $f= Z\alpha mc/p =Z\alpha \sqrt{140\,\textrm{MeV}/(2E_\pi)}$.
% When relating the cross-section of producing $\pi^-$ to that of $\pi^0$ using
% \begin{equation}
% \sigma_p(\pi^-)= \frac{2\pi f}{1-\exp{(-2\pi f)}}\sigma_p(\pi^0),
% \end{equation}
% we have
% \begin{equation}
% \sigma_{p=0}= 2\pi f \sigma_p(\pi^0)=2\pi Z\alpha \sqrt{\frac{140}{2 E_\pi}} \sigma_p(\pi^0).
% \end{equation}
% With data on $\sigma_p(\pi^0)$ of producing $\pi^0$ using $E_\gamma= 200$\,MeV photos (roughly, $E_\pi= 65$\,MeV), I get $\sigma_{p=0}(Z=6) =86$\,$\mu$b (much larger than the corresponding experimental datum $4\mu$b), $\sigma_{p=0}(Z=20) =697$\,$\mu$b, $\sigma_{p=0}(Z=41) =2600$\,$\mu$b, $\sigma_{p=0}(Z=82) =7400$\,$\mu$b. ($\sigma_{p=0}(Z=82) =7400$\,$\mu$b gives $\sigma_0(Z=82)=310$\,mb, much larger than 1\,mb in Eq.\,\eqref{Eq:res_cross_section})
% }
----------------------------------------------------------------------
When producing pionic atoms with 140\,MeV photons, there are other competing processes. Various contributions to the photon attenuation cross-section are shown in Ref.\,\cite{Zyla2020_Particle_Data}. Among these contributions, pair production in the nuclear field and Compton scattering off electrons are dominant for 140\,MeV photons interacting with atoms in the target. According to Ref.\,\cite{Berestetskii_QED},
the cross-section due to scattering of 140\,MeV photons by electrons is
\begin{equation}
\sigma_{scat} \approx 2\pi Z_i r_e^2 \frac{m_e c^2}{2E_{\gamma}}\left(\ln{\frac{2E_{\gamma}}{m_e c^2}}+\frac{1}{2}\right),
\end{equation}
and the cross-section due to pair production in the nuclear field is
\begin{equation}
\label{Eq:nuclear_pair_production}
\sigma_{pp} \approx \frac{28}{9} Z_i^2 \alpha r_e^2 \left(\ln{\frac{2E_{\gamma}}{m_e c^2}}-\frac{109}{42}\right),
\end{equation}
where $Z_i$ is the number of electrons in the initial atom, $r_e = 2.8$\,fm is the classical electron radius, and $m_ec^2=0.511$\,MeV is the rest energy of the electron.
%and $E_{\gamma}\approx 140$\,MeV.
% we gelt the cross-section due to Compton scattering as $0.61\times 10^{-26}Z$\,cm$^2$ and the cross-section due to pair production in the nuclear field as $0.64 \times10^{-26}Z^2$\,cm$^2$.
Thus, we estimate the total cross-section as
\begin{equation}
%\label{Eq:total_attenuation}
\sigma_{tot}(Z_i)\approx \sigma_{scat}+\sigma_{pp}\approx 0.6\times10^{-26}(Z_i^2+Z_i)\,\textrm{cm}^2.
\end{equation}
The maximal production rate of pionic atoms with atomic number $Z$ ($Z= Z_i+1$) is
\begin{equation}
\label{Eq:max_production_rate_new_with_sigma0}
\begin{split}
p_{max} \approx j_{eff} \frac{\sigma_{eff}}{\sigma_{tot}(Z_i)}= j \frac{\Gamma_{tot}(Z,n)}{E_{\gamma}}\frac{\sigma_0}{\sigma_{tot}(Z_i)}\\
\JJ{
%=\frac{2}{\pi} j \frac{\sigma_p(Z)}{\sigma_{tot}(Z_i)} K_p(Z) \frac{Z^2 \alpha^2}{n^3}
=\frac{2}{\pi} j \frac{\sigma_{p=0}(Z)}{\sigma_{tot}(Z_i)} \frac{Z^2 \alpha^2}{n^3} }.
%j\frac{\sigma_{eff}(Z_i,n)}{\sigma_{tot}(Z_i)}
\end{split}
\end{equation}
\JJ{(for getting the production rates, we only need $\sigma_{p=0}(Z)$. Maybe fit the experimental data)}
Substituting $\sigma_0$ from Eq.\,\eqref{Eq:res_cross_section}, we get
\begin{equation}
\label{Eq:max_production_rate_new}
p_{max} \approx
%j \frac{\Gamma_{ph}}{E_{\gamma}}\frac{\Gamma_{tot}}{\Gamma_{ph}}\frac{\sigma_0}{\sigma_{tot}(Z_i)}=
2 \pi\frac{2 J+1}{2 I+1}
%\left (\frac{\hbar c}{E_\gamma}\right )^2
\frac{\Gamma_{\gamma}(Z,n)}{E_{\gamma}}\frac{(\hbar c/E_{\gamma})^2}{\sigma_{tot}(Z_i)}j
%j\frac{\sigma_{eff}(Z_i,n)}{\sigma_{tot}(Z_i)}
\end{equation}
when all GF photons impinging on the target are absorbed.
Figure\,\ref{fig:ratio_cross_section_new} shows an estimate of the maximal production rate of pionic atoms in the $1s$ state and other low $ns$ states with $Z$ up to 92. The production rate $p$ could reach up to over $10^{10}$ per second. Such high production rate of pionic atoms would be a significant improvement compared with other production methods (see, for example, Ref.\,\cite{Hori2020_spectra_helium} where $\sim 10^5$ pionic helium atoms per second are produced at a 590\,MeV proton facility).
To suppress the background induced by high-energy photons, $\Gamma_{ph}$ should be reduced by collimation but remain larger than $\Gamma_{tot}$ to obtain higher production rate. Therefore, $\Gamma_{ph}=\Gamma_{tot}$ is a good choice of $\Gamma_{ph}$ except when $\Gamma_{tot}$ is smaller than the lower limit of $\Gamma_{ph}$ which could be 140\,eV at the GF. It is also possible to produce pionic atoms in various bound states at the same time utilizing more photons and achieving even higher total production rate.
%\JJ{(We may not expect improvement in the order of magnitude from this.)}.
To realize this we can tune
%$\Gamma_{ph}\approx \Gamma_{tot}(Z,1)/2 + E_{1s}(Z)\approx E_{1s}(Z)$
$\Gamma_{ph}\approx E_{1s}(Z)$, where $E_{1s}$ is the $1s$ binding energy of the pionic atom (see, for example, Refs.\,\cite{Toki1989_pion_width_1s,Friedman2007_widths}).
%with atomic number $Z$ in the $1s$ state.
% \JJ{I will add texts about binding energy soon, with one or two values as an example.}
% \JJ{All useless photons, i.e., photons with energies off-resonance should be cut off by collimation to reduce the background. }
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{pion_production_rate_Z_from_9.pdf}
\caption{Estimate of the maximal production rate of pionic atoms in $1s$, $2s$, $3s$ and $6s$ states utilizing photons from the GF, see Eq.\,\eqref{Eq:max_production_rate_NT}.
%The maximal production rate is $p_{max}\approx j\sigma_{eff}(Z_i,n)/ \sigma_{tot}(Z_i)$, where the expected photon flux at the GF is $j = 10^{17}$\,s$^{-1}$.
}
\label{fig:ratio_cross_section_new}
\end{figure}
% \subsection{Influence of the photon energy distribution}
% \label{Subsec:GF_photon}
% It is projected that, applying laser cooling on partially stripped ions, the energy spread of
% %secondary
% photons from the GF will be reduced from $\approx 10^{-4}$ to $\approx 10^{-6}$, leading to a photon distribution width $\Gamma_{ph}\approx 140$\,eV when $E_{\gamma}\approx 140$\,MeV. When $\Gamma_{ph}=\Gamma_{tot}(Z,n)$, we have $n=\sqrt[3]{40Z^4/(Z^3-20Z^2+400Z+800)}$.
% %(for $Z\leq 30$)
% % \footnote{In the case of the GF without laser cooling, when the photon energy spread is $\approx 10^{-4}$, the boundary value of $n$ will be five times smaller: $n=\sqrt[3]{0.4Z^4/(Z^3-20Z^2+400Z+800)}$}.
% Solving equation $\Gamma_{tot}(Z,n=1)=\Gamma_{ph}\approx 140$\,eV we obtained $Z\approx2.5$. For producing pionic atoms in the $1s$ state with $Z\geq4$, $\Gamma_{tot}\geq5\Gamma_{ph}$, thus nearly all GF photons can be considered as on-resonance.
% For pionic atoms in the $1s$ state with small $Z$ ($Z=2,3$) or in states with large $n$ ($n > 1$), $\Gamma_{ph}$ can be close to or larger than $\Gamma_{tot}$.
% If we apply a beam of photons on a fixed target, the absorption cross-section for production of pionic atoms is
% \begin{equation}
% \sigma(E)=\sigma_0 \frac{(\Gamma_{tot}/2)^2}{(\Gamma_{tot}/2)^2+(E-\Delta E)^2},
% \end{equation}
% where $\sigma_0$ is the on-resonance cross-section defined in Eq.\,\eqref{Eq:res_cross_section}, $E$ is the photon energy and $\Delta E\approx140$\,MeV is the energy resonant for creating pionic atoms.
% The exact energy distribution of GF photons is unknown. For estimate we assume the Lorentzian distribution for the photon energy as
% \begin{equation}
% g(E)=\frac{1}{\pi}\frac{\Gamma_{ph}/2}{(\Gamma_{ph}/2)^2+(E-\Delta E)^2},
% \end{equation}
% where $\Gamma_{ph}$ is the photon distribution width stated above. The effective photon absorption cross-section of the initial nucleus for creating pionic atoms in the $ns$ state with the atomic number $Z$ can be derived as
% \begin{equation}
% \label{cutoff}
% \sigma_{eff}(Z_i,n)= \int_{-\infty}^{+\infty} \sigma(E)g(E) dE=\frac{\sigma_0}{1+\Gamma_{ph}/\Gamma_{tot}(Z,n)}.
% \end{equation}
% Therefore, considering the energy distribution of GF photons, the effective cross-section as well as the production rate of pionic atoms will be divided by a factor $1+(\Gamma_{ph}/\Gamma_{tot})$ and become smaller compared with the case of producing pionic atoms with monochromatic photons on resonance.
% \subsection{Production rate of pionic atoms}
% When producing pionic atoms with 140\,MeV photons, there are other competing processes. Various contributions to the photon attenuation cross-section are shown in Ref.\,\cite{Zyla2020_Particle_Data}. Among these contributions, pair production in the nuclear field and Compton scattering off electrons are dominant for 140\,MeV photons interacting with atoms in the target. According to Ref.\,\cite{Berestetskii_QED},
% the cross-section due to scattering of 140\,MeV photons by electrons is
% \begin{equation}
% \sigma_{scat} \approx 2\pi Z_i r_e^2 \frac{m_e c^2}{2E_{\gamma}}\left(\ln{\frac{2E_{\gamma}}{m_e c^2}}+\frac{1}{2}\right),
% \end{equation}
% and the cross-section due to pair production in the nuclear field is
% \begin{equation}
% \label{Eq:nuclear_pair_production}
% \sigma_{pp} \approx \frac{28}{9} Z_i^2 \alpha r_e^2 \left(\ln{\frac{2E_{\gamma}}{m_e c^2}}-\frac{109}{42}\right),
% \end{equation}
% where $Z_i$ is the number of electrons in the initial atom, $r_e = 2.8$\,fm is the classical electron radius, $m_ec^2=0.511$\,MeV is the rest energy of the electron, and $E_{\gamma}\approx 140$\,MeV.
% % we gelt the cross-section due to Compton scattering as $0.61\times 10^{-26}Z$\,cm$^2$ and the cross-section due to pair production in the nuclear field as $0.64 \times10^{-26}Z^2$\,cm$^2$.
% Thus, we estimate the total cross-section as
% \begin{equation}
% \label{Eq:total_attenuation}
% \sigma_{tot}(Z_i)\approx \sigma_{scat}+\sigma_{pp}\approx 0.6\times10^{-26}(Z_i^2+Z_i)\,\textrm{cm}^2.
% \end{equation}
% The maximal production rate of pionic atoms with atomic number $Z$ ($Z= Z_i+1$) is
% \begin{equation}
% \label{Eq:max_production_rate}
% p_{max} \approx j\frac{\sigma_{eff}(Z_i,n)}{\sigma_{tot}(Z_i)}
% \end{equation}
% when all GF photons are absorbed in the target.
% %Here $Z_i = Z-1$ is the atomic number of initial atoms in the target before interacting with photons,
% %in the target with length long enough
% Here $j = 10^{17}$\,s$^{-1}$ is the expected photon flux at the GF. Figure\,\ref{fig:ratio_cross_section} shows an estimate of the maximal production rate of pionic atoms in the $1s$ state and other low $ns$ states \footnote {The production of high $ns$ states is suppressed by the increase of $\Gamma_{ph}/\Gamma_{tot}$ in the denominator of Eq.\,\eqref{cutoff}.} with $Z$ up to 92 \footnote{Note that although Eq.\,\eqref{Eq:1s_width} may not be a good estimate of $\Gamma_{tot}$ for pionic atoms with $Z > 50$, Eq.\,\eqref{Eq:max_production_rate} is still applicable to low $ns$ states of these pionic atoms because $\Gamma_{ph}/\Gamma_{tot} \ll 1$ is negligible and $\sigma_{eff}\approx\sigma_0$ in these cases.}. The production rate $p$ could reach up to over $10^{15}$ per second. Such high production rate of pionic atoms would be a significant improvement compared with other production methods (see, for example, Ref.\,\cite{Hori2020_spectra_helium} where $p \approx 10^5$ pionic helium atoms per second are produced at a 590\,MeV proton facility).
% \begin{figure}[!htpb]\centering
% \includegraphics[width=1.0\linewidth]{10.13Pion_cross_section_ratio.pdf}
% \caption{Estimate of the maximal production rate of pionic atoms in $1s$, $2s$, $3s$ and $6s$ states utilizing photons from the GF, see Eq.\,\eqref{Eq:max_production_rate}. The maximal production rate is $p_{max}\approx j\sigma_{eff}(Z_i,n)/ \sigma_{tot}(Z_i)$, where the expected photon flux at the GF is $j = 10^{17}$\,s$^{-1}$. }
% \label{fig:ratio_cross_section}
% \end{figure}
%\subsection{Photoproduction of kaonic atoms}
%
%Kaonic atoms have been less studied than pionic atoms, so this problem may be especially interesting.
%The rest energy of negative kaons is $m_{K^-}c^2=494$\,MeV, so we need much higher photon energies than those for pionic atoms.
%This may, in principle, be achieved with higher-energy laser photons. At the GF, the energy of secondary photons is $\approx 4\gamma^2$ times larger than that of primary photons with $\gamma$ up to 3000. When $\gamma=3000$, primary photons should have energies of 13.7\,eV to get 494\,MeV secondary photons.
%The basic theory for photoproduction of kaonic atoms is similar to that for pionic atoms mentioned above though the numbers are different (see, for example, \cite{Friedman2007_widths}).
Assuming that $\Gamma_{\gamma}/\Gamma_{tot} \sim \alpha\approx0.007$ and using Eq.\,\eqref{Eq:res_cross_section}, we obtain the resonant cross-section for creating kaonic atoms as $\sigma_0 \sim 1. \times 10^{-28}$\,cm$^2$,
an order of magnitude smaller than that for pionic atoms.
By fitting data for the $1s$ level width of four kaonic atoms: deuterium, carbon, nickel and lead from Refs.\,\cite{Friedman1999_kaon_antiproton,Friedman1999_deep_bound_kaon,Curceanu2019_kaonic_width} (see Fig.\,\ref{fig:kaon_width_fitting}),
%\JJ{ do we need to present the width numbers here?},
we approximate the kaon absorption width of the $1s$ level as
\begin{equation}
\label{Eq:kaon_1s_width}
\Gamma_{tot}(Z,n=1)\approx \frac{20Z^4}{Z^3-9Z^2+100Z-60}\,\textrm{keV}.
\end{equation}
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{Kaon_width_fitting.pdf}
\caption{Fitting of $1s$ width data (red dots on the graph) for kaonic atoms. The width data for carbon, nickel and lead are from Refs.\,\cite{Friedman1999_deep_bound_kaon,Friedman1999_kaon_antiproton}; the $1s$ width of kaonic deuterium is from Ref.\,\cite{Curceanu2019_kaonic_width}. }
\label{fig:kaon_width_fitting}
\end{figure}
The Coulomb radius of the kaonic state is 3.5 times smaller than that of the corresponding pionic state, therefore, only $ns$ states with $Z/n^{3/2} \leq 14$ may be approximately treated as the Coulomb bound states.
For $Z\leq 14$ the kaon absorption width for higher $ns$ states can be estimated
as $\Gamma_{tot}(Z,n) \approx \Gamma_{tot}(Z,1)/n^3$.
%according to Eq.\,\eqref{Eq:ns_width}.
%\begin{equation}
%\label{Eq:kaon_ns_width}
% \Gamma_{tot}(Z,n) \approx \Gamma_{tot}(Z,1)\frac{1}{n^3}.
%\end{equation}
% =\frac{20Z^4}{Z^3-9Z^2+100Z-60}\frac{1}{n^3}\,\textrm{KeV}.
For kaonic atoms with $Z > 14$ this estimate of the total width $\Gamma_{tot}(Z,n)$ is not applicable since the $1/n^3$ dependence in Eq.\,\eqref{Eq:ns_width} follows from the Coulomb wavefunctions but the $1s$ function in this case is not the Coulomb one. Therefore, we may use Eq.\,\eqref{Eq:ns_width} for $Z > 14$ as an order-of-magnitude estimate of $\Gamma_{tot}(Z,n)$ only.
Similar to the case of pionic atoms (Sec.\,\ref{Subsec:pion_absorption_width}), besides $ns$ states, kaonic atoms in other $nl$ states can also be produced. The estimate of widths of $nl$ states is similar with the case of pionic atoms except now for kaonic atoms $a_0\approx 55$\,fm \footnote{Note that for light kaonic atoms, $a_0$ should be obtained considering the reduced mass of the kaonic atom.}.
% Besides $ns$ states, kaonic atoms in other $nl$ states can also be produced. When the radii of corresponding Bohr orbits exceed the nuclear size,
% %\JJ{(For nl($l>0$) states in atoms with $Z<100$ it is true)}
% widths of these states can be estimated as $\Gamma_{nl}/\Gamma_{ns}\approx P_{nl}/P_{ns}$.
% Here $P_{nl}$
% %$P_{nl}=\int_{r\leq r_0} R_{nl}(r) r^2 dr$
% is the probability for a kaon in the $nl$ state to be inside the nucleus.
% %\JJ{(For higher $nl$ states we need be careful whether the pion absorption width is still dominant or not.)}
% For example, $\Gamma_{np}/\Gamma_{ns}\approx (Z r_0/a_0)^2$, where $r_0\approx 1.2 (2.5Z)^{1/3}$\,fm is the nuclear radius, and $a_0= 55$\,fm is the $1s$ Bohr radius \JJ{Actaully 84\,fm due to the reduced mass, put this into footnote, for lighter kaonci atoms...} of the kaonic hydrogen atom.
%We see that the width of low $ns$ states is expected to exceed the photon distribution width $\Gamma_{ph}\approx 0.5$\,KeV at the GF by few orders of magnitude.
The radiative width of atomic levels in kaonic atoms is expected to be 3.5 times larger than that in pionic atoms (see Sec.\,\ref{Subsec:pion_absorption_width}). It is much smaller than the kaon absorption width.
% Consider applying a beam of GF photons with photon energies $E_{\gamma}\approx 500$\,MeV and energy distribution width $\Gamma_{ph}\approx 0.5$\,KeV. For photoproduction of the $1s$ state in kaonic atoms with $Z > 1$, $\Gamma_{tot}\geq 6\Gamma_{ph}$. Therefore, almost all GF photons are on resonance. For producing higher $ns$ states, $\Gamma_{ph}$ can be close to or larger than $\Gamma_{tot}$. When $\Gamma_{tot}(Z,n)= \Gamma_{ph}\approx 0.5$\,KeV, we have $n= \sqrt[3]{40Z^4/(Z^3-9Z^2+100Z-60)}$. The influence of the photon-distribution width is shown in Eq.\,\eqref{cutoff} where now $\sigma_0 \approx 1. \times 10^{-28}$\,cm$^2$, $\Gamma_{ph}\approx 0.5$\,KeV, and $\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)/n^3$ with $\Gamma_{tot}(Z,1)$ given in Eq.\,\eqref{Eq:kaon_1s_width} for kaonic atoms.
Consider applying a beam of GF photons with energies of $E_{\gamma}\approx 500$\,MeV to produce kaonic atoms.
For the interaction between 500\,MeV photons and atoms in a fixed target, the total cross-section for photon attenuation is dominated by pair production in the nuclear field
\cite{Zyla2020_Particle_Data}. With Eq.\,\eqref{Eq:nuclear_pair_production}, we estimate the total cross-section as
\begin{equation}
\label{Eq:total_attenuation_kaon}
\sigma_{tot}(Z_i)\approx \sigma_{pp}\approx 0.90\times10^{-26}Z_i^2\,\textrm{cm}^2.
\end{equation}
% With $\sigma_{eff}(Z_i,n)$ and $\sigma_{tot}(Z_i)$ for kaonic atoms derived above, the maximal production rate of kaonic atoms can be estimated by using Eq.\,\eqref{Eq:max_production_rate}. Figure\,\ref{fig:Kaon_ratio_cross_section} shows an estimate of the maximal production rate of kaonic atoms in
% %the $1s$, $2s$, $3s$ and $6s$ states.
% some low $ns$ states.
% The production rate $p$ is also up to $\approx 10^{15}$ per second.
The maximal production rate of kaonic atoms can be estimated by using %Eq.\,\eqref{Eq:max_production_rate_new} or
Eq.\,\eqref{Eq:max_production_rate_new_with_sigma0}, where now for kaonic atoms we have $\sigma_0 \sim 1. \times 10^{-28}$\,cm$^2$, $E_{\gamma}\approx 500$\,MeV, and $\Gamma_{tot}(Z,n)\approx \Gamma_{tot}(Z,1)/n^3$ with $\Gamma_{tot}(Z,1)$ given in Eq.\,\eqref{Eq:kaon_1s_width}. Figure\,\ref{fig:Kaon_ratio_cross_section} shows an estimate of the maximal production rate of kaonic atoms in
some low $ns$ states.
The production rate $p$ is up to over $10^{9}$ per second.
As for the choice of the energy-distribution width $\Gamma_{ph}$ of GF photons, considerations are similar with the case of pionic atoms (see Sec.\,\ref{Subsec:GF_pion}).
\begin{figure}[!htpb]\centering
\includegraphics[width=1.0\linewidth]{kaon_production_rate_Z_from_9.pdf}
\caption{Estimate of the maximal production rate of kaonic atoms in the $1s$, $2s$, $3s$ and $6s$ states utilizing photons from the GF.
%The maximal production rate is $p_{max}\approx j\sigma_{eff}(Z_i,n)/ \sigma_{tot}(Z_i)$, where the expected photon flux at the GF is $j = 10^{17}$\,s$^{-1}$.
}
\label{fig:Kaon_ratio_cross_section}
\end{figure}
% \begin{figure}[!htpb]\centering
% \includegraphics[width=1.0\linewidth]{10.13Kaon_cross_section_ratio.pdf}
% \caption{Estimate of the maximal production rate of kaonic atoms in the $1s$, $2s$, $3s$ and $6s$ states utilizing photons from the GF. The maximal production rate is $p_{max}\approx j\sigma_{eff}(Z_i,n)/ \sigma_{tot}(Z_i)$, where the expected photon flux at the GF is $j = 10^{17}$\,s$^{-1}$.}
% \label{fig:Kaon_ratio_cross_section}
% \end{figure}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\JJ{
%(Unfinished...)
The integrated cross-section obtained with \eqref{Eq:crosec_energy_dependence} is
\begin{equation}
\label{Eq:integrated cross-section our}
\Sigma(Z,n)=\int \sigma(E) dE= \frac{\pi}{2}\sigma_0 \Gamma_{tot}(Z,n).
\end{equation}
}
\JJ{
According to \cite{TZARA1970_photoproduction_pion}, the production amplitude of a pion in $ns$ states can be approximated as
\begin{equation}
M_{if}\approx \psi_{ns}(0)\int f(\Vec{r}) d\Vec{r}=\psi_{ns}(0)\mu_{if},
\end{equation}
where $f(\Vec{r})$ includes the nuclear and incoming particle wavefunctions and the transition operator, and $\int f(\Vec{r}) d\Vec{r}$ is defined as $\mu_{if}$ for brevity. We study production rates of pionic atoms with final nuclei only in the ground state.
%%%%
(I am not clear about the $f(r)$ or $\mu_{if}$ here. It seems to depend on many things including the final state of the pion?)
%%%%%%
The integrated cross-section for photoproduction of pions in $ns$ states is
\begin{equation}
\label{Eq:crosec_bound_pion}
\Sigma =\int \sigma_n dE=\frac{2\pi}{\hbar c} |\psi_{ns}(0)|^2 |\mu_{if}|^2= \frac{2}{\hbar c} \frac{1}{n^3 a^3} |\mu_{if}|^2,
\end{equation}
where %$a= \hbar/(\alpha Z m c)$
$a= a_0/ Z$
is the $1s$ orbital radius of the pion.
%\Sigma |\mu_{if}| here, what does it contain? can it be cancelled out with the formula below?
The cross-section for producing free pions is
\begin{equation}
\label{Eq:crosec_free_pion}
% \sigma_k=\frasection_freec{8\pi^2}{\h_pionbar c} \Sigma |\mu_{if}|^2 |\psi_k(0)|^2\rho_f .
\sigma_k=\frac{8\pi^2}{\hbar c} \rho_f |\psi_k(0)|^2 |\mu_{if}|^2,
\end{equation}
where for free pions at low momentum,
%$\rho_f=p\bar{m}/(2\pi\hbar)^3$
$\rho_f=pm/(2\pi\hbar)^3$ is (???), and $|\psi_k(0)|^2= 2\pi\gamma/(1-e^{-2\pi\gamma})$ is the Sommerfeld factor (?)
%with $\gamma= Z\alpha m c/ p= 1/(a_0 k)$.
with $\gamma= 1/(a k)$.
Here, $k$ is the magnitude of the wave vector of the free pion.}
\JJ{
So the ratio of the two cross-sections above is
\begin{equation}
\frac{\Sigma}{\sigma_k}=
\frac{1}{4\pi}\frac{1}{\rho_f}\frac{|\psi_{ns}(0)|^2}{|\psi_k(0)|^2}=
%\frac{1}{4\pi^2}\frac{1}{n^3a^3}\frac{1}{|\psi_k(0)|^2\rho_f}
%\frac{\hbar^2}{n^3 a^2 m} (1-e^{-2\pi\gamma}).
\frac{(\alpha Z c)^2}{n^3} m (1-e^{-2\pi\gamma}).
\end{equation}
( again $\mu_{if}$, is it the same for two cross-sections above? Otherwise, this factor cannot be cancelled out.)
Next: data for cross-sections of producing free pions with photon energy ``near" the threshold.
}
-------------------------------------------------
\JJ{The maximal production rate of kaonic atoms can be estimated by using Eq.\eqref{Eq:max_production_rate}. When $\Gamma_{tot}\gg \Gamma_{ph}\approx 0.5$\,KeV (e.g., for photoproduction of kaonic atoms with $Z > 1$ in the $1s$ state), the maximal production rate of these kaonic atoms is approximately
\begin{equation}
p_{max} = j \frac{\sigma_0}{\sigma_{tot}(Z_i)}.
\end{equation}
}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% \JJ{For these $ns$ states, total widths are expected to be
% $\Gamma_{tot}(Z,n)= \Gamma_{tot}(Z,1)/n^3$. When $\Gamma_{tot}(Z,n)= \Gamma_{ph}\approx 0.5$\,keV, we have $n=\sqrt[3]{40Z^4/(Z^3-9Z^2+100Z-60)}$.
%Their maximal production rate can be estimated by Eq.\,\eqref{Eq:max_production_rate} with using parameters derived above for kaonic atoms. $\sigma_{tot}$ for production of kaonic atoms is the same as that for pionic atoms.
% }
% \JJ{With $\Gamma_{tot}(Z,n)$ and $\Gamma_{ph}$ for kaonic atoms derived above, the influence of photon distribution width is shown in Eq.\,\eqref{cutoff}. The total cross-section for photon attenuation, $\sigma_{tot}(Z_i)$ given in Eq.\,\eqref{Eq:total_attenuation} is the same for creating kaonic atoms and pionic atoms. The maximal production rate of kaonic atoms can be gotten by using Eq.\eqref{Eq:max_production_rate}. When $\Gamma_{tot}\gg \Gamma_{ph}\approx 0.5$\,keV (e.g., for photoproduction of kaonic atoms with $Z > 1$ in the $1s$ state), the maximal production rate of these kaonic atoms is approximately
% \begin{equation}
% p_{max} = j \frac{\sigma_0}{\sigma_{tot}(Z_i)}.
% \end{equation}
% }
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% For example, consider photoproduction of kaonic atoms in the $1s$ state.
% %According to Ref.\,\cite{Curceanu2019_kaonic_width}, $\Gamma_{tot}\approx 750$\,eV for a kaonic deuterium in the $1s$ state. As explained in Sec.\,\ref{Subsec:pion_absorption_width}, $\Gamma_{tot}$ is expected to increase with $Z$, therefore,
% According to Eq.\,\eqref{Eq:kaon_1s_width}, $\Gamma_{tot}\gg \Gamma_{ph}\approx 0.5$\,keV when $Z > 1$. So, almost all GF photons are on resonance. The maximal production rate of these kaonic atoms is
% \begin{equation}
% p_{max} = j \frac{\sigma_0}{\sigma_{tot}(Z_i)},
% \end{equation}
% where $\sigma_{tot}(Z_i)$ is the same as that in Eq.\eqref{Eq:max_production_rate} and $\sigma_0 \approx 1. \times 10^{-28}$\,cm$^2$.
\JJ{With width data about kaonic atoms in $nl(l=n-1)$ states from $2p$ to $8j$ in Ref.\,\cite{Friedman1994_kaonic_nl_widths}, we can expect that for kaonic atoms in the $1s$ state with $Z\geq 6$, $\Gamma_{tot}(Z,n=1)\gg \Gamma_{ph} = 494$\,eV. So, the maximal production rate of these kaonic atoms is
\begin{equation}
p_{max} = j \frac{\sigma_0}{\sigma_{tot}(Z_i)},
\end{equation}
where $\sigma_{tot}(Z_i)$ is the same as that in Eq.\eqref{Eq:max_production_rate}. Since $\sigma_0(kaon)\approx 0.1\sigma_0(pion)$, the production rate of these kaonic atoms will be one order of magnitude smaller than that of corresponding pionic atoms which we already knew from the estimate above.}
Additional items:
\begin{itemize}
% \item Try K-atoms \JJ{\cite{Friedman2007_widths}} if it is more interesting. Basic theory is the same but numbers are different.
\item
%Consider photoproduction of deeply bound pionic states?
\cite{Yamazaki_1996_deeply_bound_states,Friedman_1985_width_heavy_atoms,Kienle2004_deep_bound_state} Prof. Avraham Gal claimed, `` Production of $1s$ states in heavy elements, the so called deeply bound pionic states would be very interesting".
%( the key to success was the creation of a pion at rest )
\end{itemize}
%High energy leads to other ``competitive" reactions. If they do not absorb many photons, we will still have high production rate but with background noise. Broad widths may make the atoms hard to be detected because of their short lifetime.
%We lack the \Gamma_{\gamma} info and only have \Gamma_{tot}.
% We found the ratio of the widths $\Gamma_{\gamma}/\Gamma_{tot}\approx\alpha\approx0.007$ in pion production by photons on nuclei (\JJ{ How to get this, from Particle Data \cite{Groom2000_Particle_Data}? More explanations may be in need.}).
% \section{Conclusions and outlook}
% \label{Section:Conclusions}
% \section*{Acknowledgments}
% The authors are grateful to Carlos Bertulani, Eli Friedman, Avraham Gal and Masaki Hori for helpful information and discussions.
\section*{Appendix}
\appendix
\section{Experimental considerations (for internal use)}
\subsection{background noise}
Some widths are awfully large of order $\leq10^{-19}$ seconds. The atom is thus destroyed very rapidly, perhaps before it can traverse any significant amount of space and make it to the detector. Indeed, it is better to consider light atoms to have a longer lifetime. \JJ{(I am not sure. People are able to detect deeply bound pionic states which are likely to have shorter lifetime.)}
% The atom is (using a very rough approximation) ideally created in the kinetic energy + momentum matching condition, so initially the atom is not stationary but in-flight within a dense target containing many other atoms and nuclei. If the gamma ray energy were tuned near threshold the free pion production is suppressed; nevertheless the free pion and atomic bound state production cross-sections must be of similar order so that spectroscopic information is needed to distinguish the two types of signals.
As the atom traverses the material, its remaining outer electrons are rapidly ionized and so its effective Z charge changes. If the atom’s kinetic energy exceeds the binding energy of the pion, also the pion is ionized by interactions with the surrounding atoms/nuclei. This implies that pionic atoms of sufficiently high-Z will need to be used; but this is unfavourable from the point of view that the 1s state width will increase roughly as a function of the overlap integral between the 1s wavefunction and the nucleus.
If the target is thick, the incoming gamma rays cause various types of nuclear excitation, fission, etc. which produces a background of secondary gamma rays and nuclear fragments. This background (which gets worse for higher Z materials) makes signal identification difficult. So a thin target is preferred, but then the production rate is reduced.
In terms of a zero-degree nuclear reaction, when the incoming photon behaves like a 1s plane wave, and the recoiling nucleus scatters into the theta=0 laboratory angle. The target must be so thin that Compton scattering and energy loss etc. is negligible. My intuition would seem to say that to obtain a clean signal it is insufficient to place a detector near theta=0, I would need to tag and reject other background sources by placing particle identification capability around the target.
We tune the photon energy to the resonance in the moving final system, including both the energy and momentum conservation, i.e. the energy conservation includes kinetic energy of the bound system.
For sufficiently large $Z$ it is not enough energy to produce free $\pi^-$ since bound $\pi^-$ energy is smaller.
The binding energy of $1s$ states is 0.004$Z^2$\,MeV. It exceeds kinetic energy of the final system 10\,MeV/A \JJ{(??)} for $Z\geq13$.
%\JJ{(Do we need to confine atoms with $Z\geq13$ ?)}
Even if it does not exceed, this is the threshold process, free $\pi^-$ production is very strongly suppressed. Delta resonance width and fermi motion are completely irrelevant to the possibility of free pion production since the total energy is fixed. If photon energy is below threshold, free pion can not be produced.
\JJ{Pionic atoms are usually produced by stopping a beam of free $\pi-$ in a target. The production rate will be limited by the intensity of the $\pi-$ beam. For example, according to Ref.\,\cite{Hori2020_spectra_helium}, the highest average intensity of free $\pi-$ beam available worldwide is $\approx 10^7$\,s$^{-1}$ provided at the Paul Scherrer Institute. It can produce $\approx 10^5$ pionic helium atoms per second.}
% Now, we consider producing pionic atoms in $1s$ states with
% %secondary
% photons from GF. By applying laser cooling on partially stripped ions, the energy spread of secondary photons from GF will hopefully be reduced from $\approx $ 10$^{-4}$ to $\approx $ 10$^{-6}$, leading to photon distribution width $\Gamma_{ph} = 140$\,eV when $E_{\gamma}=140$\,MeV. Solving equation $\Gamma_{tot}=\Gamma_{ph}= 140$\,eV we obtain $Z\approx2.5$. For producing pionic atoms in $1s$ state with $Z\geq4$, $\Gamma_{tot}\geq5\Gamma_{ph}$ thus nearly all GF photons can be considered as on-resonance.
% At GF, photon flux $j=10^{17}$\,ph/s. Utilizing a
% %1\,cm$\times$1\,cm$\times$100\,cm
% target containing $\approx 10^{25}$ nuclei ($Z\geq4$) interacting with photons, the photoproduction rate of pionic atoms is $\approx 10^{15}$ per second.
% For example, pionic atoms in Table A.2 of \cite{Konijn1990_widths_nl} all have widths $\geq1$\,keV. So, according to the estimate above their production rate can reach $\approx 10^{15}$ per second. Such high production rate of pionic atoms may be a significant improvement compared with that of other ways (see, for example Ref. \cite{Hori2020_spectra_helium} where $\approx 10^{5}$ pionic atoms per second are produced at the 590\,MeV proton facility).
{Texts here may be deleted later}
\JJ{Stated by Prof. Avraham Gal, `` Pion absorption in pionic atoms occurs dominantly on two nucleons in $s$-wave situations, not $p$-wave, and has little to do with the $\Delta$-resonance."} Pion absorption/production is dominated by $\Delta$-resonance. $\Delta$ may decay to 99.4\% nucleon and pion (then pion may generate nuclear processes) or 0.6\% nucleon, pion and gamma.
We found the ratio of the widths $\Gamma_{\gamma}/\Gamma_{tot}=0.006$ in Particle Data \cite{Groom2000_Particle_Data}. The same widths appear in pion production by photons on nuclei. The photon is absorbed by a neutron with excitation of $\Delta$-resonance which then decays to proton and pion- in the bound state, $\gamma + n \rightarrow \Delta \rightarrow ( p+\pi^-)$. Total width $\Gamma_{tot} = \Gamma_{a}$ (pion absorption width). $(\pi^- + p) \rightarrow \Delta \rightarrow n + \pi^0$. In free space this bound state decay has a very small phase volume since released energy is $\approx 3$\,MeV only, very strong suppression of the total width. In nuclei it is the virtual pion which is absorbed with nuclear decay, no phase volume suppression. When this process happens inside the nucleus, we assume that the ratio of the widths is approximately the same as in the $\Delta$ peak. This is an order of magnitude estimate. The accuracy is the same as in estimate of the ratio as $\approx \alpha=1/137=0.007$ since photon emission involves electric charge $e^2$ in the probability. Released kinetic energy in the $\Delta$ peak = $\Delta$ mass 1232\,MeV - 940\,MeV (neutron mass) - 135\,MeV ($\pi^0$ mass) = 157\,MeV ($\geq\pi$ mass), i.e. no phase volume suppression in the $\Delta$ width = 117\,MeV.
%Producing macroscopic pionic atoms with various host nuclei efficiently will be greatly appreciated in related studies. In the following, we present a possibility of a very promising production rate, up to $\approx 10^{15}$ pionic atoms per second, with an upcoming powerful tool, the Gamma Factory (GF) \cite{Budker2020_AdP_GF} currently studied within the CERN Physics Beyond Colliders program.
is produced in for produced pionic atoms with the pion in the bound state, $Q_m\approx E_{CB}+\JJ{\epsilon_{1s}}-E_{\gamma}$, where $E_{CB}$ is the Coulomb binding energy and $\epsilon_{1s}$ is the energy shift due to the strong interaction. \JJ{(The strong interaction shift is smaller than $\left|E_{CB}\right|$ and can be neglected in our estimate)}
\begin{equation}
E_{CB}\approx -\frac{\mu c^2}{2}\left(\frac{Z\alpha}{n}\right)^2,
\end{equation}
where $\mu=mM/(m+M)$ is the reduced mass with $m$ the mass of $\pi^-$ and $M$ the mass of the nucleus, $c$ the speed of light, $Z$ the atomic number, $\alpha=1/137$ the fine structure constant and $n$ the principal quantum number. With $\left|E_{CB}\right|\ll \left|{Q_m}\right|$, we have $Q_m\approx-E_{\gamma}$, thus $E_{\gamma}\approx140$\,MeV. \JJ{ ( $\left|E_{CB}\right|\ll \left|{Q_m}\right|$ is true for not too heavy pionic atoms ($Z\leq70$) or atoms with large $n$. We may care about atoms with widths smaller than 1\,MeV, which seems to be satisfied only when $Z\leq100$. We can consider heavy atoms separately if we want since as Prof. Avraham Gal claimed, `` Production of $1s$ states in heavy elements so called ``deeply bound pionic states" would be very interesting. )}
\JJ{ Other ways of producing pionic atoms and production rates. Need a ref here.} According to Prof. Masaki Hori, they get $\geq3\times10^5$\,s$^{-1}$ at the 590\,MeV proton facility.
Specific examples:
With width data from the Table 1 in \cite{Friedman2007_widths}, the production rates of ( $^{A=20,22}$Ne + $\pi^-$ )$_{1s}$, ( Na + $\pi^-$ )$_{1s}$, ( Mg + $\pi^-$ )$_{1s}$, ( Si + $\pi^-$ )$_{1s}$, ( Sn + $\pi^-$ )$_{1s}$, ( Pb + $\pi^-$ )$_{1s}$ are all around 10$^{15}$ s$^{-1}$.
For $1s$ width data of 16O, 40Ca, 90Zr, 208Pb (from Prof. Masaki Hori without refs where the data are from), they are all $\geq1$\,keV. So these pionic atoms will have the high production rate mentioned above.
\JJ{The $1/n^3$ dependence seems not suitable to $^{208}$Pb. How about $ns$ states of other pionic atoms? Do we need a table of $ns$ widths? Maybe we can also give the production rate in the table, though for states with $\Gamma_{ns}\gg\Gamma_{ph}$ the production rates are almost the same.}
\begin{table*}[t]
\centering
\begin{tabular}{l c c c c c c c}
\hline
\hline
Z &$^{A}$X &$1s$ widths (keV) &$1s$ binding energy (keV)\\
\hline \\[-0.2cm]
6 &$^{16}$O &7 &218 \\
20 &$^{40}$Ca &105 &1108 \\
40 &$^{90}$Zr &296 &2970 \\
82 &$^{208}$Pb &632 &6959 \\
\hline
\hline
\end{tabular}
\caption{Widths and binding energies of pionic atoms in $1s$ states \cite{Friedman2007_widths}. ( Data in first four lines are from Prof. Masaki Hori. But he did not provide the refs where he got the data...)}
\label{tab: Widths}
\end{table*}
%apsrev4-2.bst 2019-01-14 (MD) hand-edited version of apsrev4-1.bst
%Control: key (0)
%Control: author (8) initials jnrlst
%Control: editor formatted (1) identically to author
%Control: production of article title (0) allowed
%Control: page (0) single
%Control: year (1) truncated
%Control: production of eprint (0) enabled
\begin{thebibliography}{43}%
\makeatletter
\providecommand \@ifxundefined [1]{%
\@ifx{#1\undefined}
}%
\providecommand \@ifnum [1]{%
\ifnum #1\expandafter \@firstoftwo
\else \expandafter \@secondoftwo
\fi
}%
\providecommand \@ifx [1]{%
\ifx #1\expandafter \@firstoftwo
\else \expandafter \@secondoftwo
\fi
}%
\providecommand \natexlab [1]{#1}%
\providecommand \enquote [1]{``#1''}%
\providecommand \bibnamefont [1]{#1}%
\providecommand \bibfnamefont [1]{#1}%
\providecommand \citenamefont [1]{#1}%
\providecommand \href@noop [0]{\@secondoftwo}%
\providecommand \href [0]{\begingroup \@sanitize@url \@href}%
\providecommand \@href[1]{\@@startlink{#1}\@@href}%
\providecommand \@@href[1]{\endgroup#1\@@endlink}%
\providecommand \@sanitize@url [0]{\catcode `\\12\catcode `\$12\catcode
`\&12\catcode `\#12\catcode `\^12\catcode `\_12\catcode `\%12\relax}%
\providecommand \@@startlink[1]{}%
\providecommand \@@endlink[0]{}%
\providecommand \url [0]{\begingroup\@sanitize@url \@url }%
\providecommand \@url [1]{\endgroup\@href {#1}{\urlprefix }}%
\providecommand \urlprefix [0]{URL }%
\providecommand \Eprint [0]{\href }%
\providecommand \doibase [0]{https://doi.org/}%
\providecommand \selectlanguage [0]{\@gobble}%
\providecommand \bibinfo [0]{\@secondoftwo}%
\providecommand \bibfield [0]{\@secondoftwo}%
\providecommand \translation [1]{[#1]}%
\providecommand \BibitemOpen [0]{}%
\providecommand \bibitemStop [0]{}%
\providecommand \bibitemNoStop [0]{.\EOS\space}%
\providecommand \EOS [0]{\spacefactor3000\relax}%
\providecommand \BibitemShut [1]{\csname bibitem#1\endcsname}%
\let\auto@bib@innerbib\@empty
%</preamble>
\bibitem [{\citenamefont {Backenstoss}(1970)}]{Backenstoss1970_pionic_atom}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {G.}~\bibnamefont
{Backenstoss}},\ }\bibfield {title} {\bibinfo {title} {Pionic atoms},\
}\href
{https://www.annualreviews.org/doi/pdf/10.1146/annurev.ns.20.120170.002343}
{\bibfield {journal} {\bibinfo {journal} {Ann. Rev. Nucl. Sci.}\ }\textbf
{\bibinfo {volume} {20}},\ \bibinfo {pages} {467} (\bibinfo {year}
{1970})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Deser}\ \emph {et~al.}(1954)\citenamefont {Deser},
\citenamefont {Goldberger}, \citenamefont {Baumann},\ and\ \citenamefont
{Thirring}}]{Deser1954_En}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {S.}~\bibnamefont
{Deser}}, \bibinfo {author} {\bibfnamefont {M.~L.}\ \bibnamefont
{Goldberger}}, \bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Baumann}},\ and\ \bibinfo {author} {\bibfnamefont {W.}~\bibnamefont
{Thirring}},\ }\bibfield {title} {\bibinfo {title} {Energy level
displacements in pi-mesonic atoms},\ }\href
{https://doi.org/10.1103/PhysRev.96.774} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev.}\ }\textbf {\bibinfo {volume} {96}},\ \bibinfo {pages}
{774} (\bibinfo {year} {1954})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Brueckner}(1955)}]{Brueckner1955_En}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.~A.}\ \bibnamefont
{Brueckner}},\ }\bibfield {title} {\bibinfo {title} {Energy levels of
$\ensuremath{\pi}$-mesonic atoms},\ }\href
{https://doi.org/10.1103/PhysRev.98.769} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev.}\ }\textbf {\bibinfo {volume} {98}},\ \bibinfo {pages}
{769} (\bibinfo {year} {1955})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Hori}\ \emph {et~al.}(2020)\citenamefont {Hori},
\citenamefont {Aghai-Khozani}, \citenamefont {S{\'o}t{\'e}r}, \citenamefont
{Dax},\ and\ \citenamefont {Barna}}]{Hori2020_spectra_helium}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {M.}~\bibnamefont
{Hori}}, \bibinfo {author} {\bibfnamefont {H.}~\bibnamefont {Aghai-Khozani}},
\bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {S{\'o}t{\'e}r}}, \bibinfo
{author} {\bibfnamefont {A.}~\bibnamefont {Dax}},\ and\ \bibinfo {author}
{\bibfnamefont {D.}~\bibnamefont {Barna}},\ }\bibfield {title} {\bibinfo
{title} {Laser spectroscopy of pionic helium atoms},\ }\href
{https://doi.org/10.1038/s41586-020-2240-x} {\bibfield {journal} {\bibinfo
{journal} {Nature}\ }\textbf {\bibinfo {volume} {581}},\ \bibinfo {pages} {37
} (\bibinfo {year} {2020})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Krasny}(2015)}]{Krasny2015}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {M.~W.}\ \bibnamefont
{Krasny}},\ }\href@noop {} {\emph {\bibinfo {title} {{The Gamma Factory
proposal for CERN}}}},\ \bibinfo {type} {Tech. Rep.}\ (\bibinfo
{institution} {CERN},\ \bibinfo {year} {2015})\ \Eprint
{https://arxiv.org/abs/1511.07794} {arXiv:1511.07794 [hep-ex]} \BibitemShut
{NoStop}%
%%CITATION = ARXIV:1511.07794;%%
\bibitem [{\citenamefont {Budker}\ \emph {et~al.}(2020)\citenamefont {Budker},
\citenamefont {Crespo L{\'o}pez-Urrutia}, \citenamefont {Derevianko},
\citenamefont {Flambaum}, \citenamefont {Krasny}, \citenamefont {Petrenko},
\citenamefont {Pustelny}, \citenamefont {Surzhykov}, \citenamefont
{Yerokhin},\ and\ \citenamefont {Zolotorev}}]{Budker2020_AdP_GF}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {D.}~\bibnamefont
{Budker}}, \bibinfo {author} {\bibfnamefont {J.~R.}\ \bibnamefont {Crespo
L{\'o}pez-Urrutia}}, \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont
{Derevianko}}, \bibinfo {author} {\bibfnamefont {V.~V.}\ \bibnamefont
{Flambaum}}, \bibinfo {author} {\bibfnamefont {M.~W.}\ \bibnamefont
{Krasny}}, \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Petrenko}},
\bibinfo {author} {\bibfnamefont {S.}~\bibnamefont {Pustelny}}, \bibinfo
{author} {\bibfnamefont {A.}~\bibnamefont {Surzhykov}}, \bibinfo {author}
{\bibfnamefont {V.~A.}\ \bibnamefont {Yerokhin}},\ and\ \bibinfo {author}
{\bibfnamefont {M.}~\bibnamefont {Zolotorev}},\ }\bibfield {title} {\bibinfo
{title} {Atomic physics studies at the gamma factory at cern},\ }\href
{https://doi.org/10.1002/andp.202000204} {\bibfield {journal} {\bibinfo
{journal} {Annalen der Physik}\ ,\ \bibinfo {pages} {2000204}} (\bibinfo
{year} {2020})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Tzara}(1970)}]{TZARA1970_photoproduction_pion}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {C.}~\bibnamefont
{Tzara}},\ }\bibfield {title} {\bibinfo {title} {Effect of the coulomb
potential on the ${\pi}^{\ensuremath{-}}$ photoproduction},\ }\href
{https://doi.org/https://doi.org/10.1016/0550-3213(70)90290-7} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics B}\ }\textbf {\bibinfo
{volume} {18}},\ \bibinfo {pages} {246 } (\bibinfo {year}
{1970})}\BibitemShut {NoStop}%
\bibitem [{Note1()}]{Note1}%
\BibitemOpen
\bibinfo {note} {In principle, pionic atoms with final nuclei in excited
states can also be resonantly produced with higher photon
energies.}\BibitemShut {Stop}%
\bibitem [{\citenamefont {Friedman}\ and\ \citenamefont
{Gal}(2007)}]{Friedman2007_widths}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {E.}~\bibnamefont
{Friedman}}\ and\ \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Gal}},\
}\bibfield {title} {\bibinfo {title} {In-medium nuclear interactions of
low-energy hadrons},\ }\href
{https://doi.org/https://doi.org/10.1016/j.physrep.2007.08.002} {\bibfield
{journal} {\bibinfo {journal} {Physics Reports}\ }\textbf {\bibinfo {volume}
{452}},\ \bibinfo {pages} {89 } (\bibinfo {year} {2007})}\BibitemShut
{NoStop}%
\bibitem [{Note2()}]{Note2}%
\BibitemOpen
\bibinfo {note} {Note that the relativistic formula for the energy of a
spin-zero particle in the case of the point-like nucleus breaks down for
$Z\geq 137/2$, so the finite nucleus potential should be used when $Z$ is
close to or above 67}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Yamazaki}\ \emph {et~al.}(1996)\citenamefont
{Yamazaki}, \citenamefont {Hayano}, \citenamefont {Itahashi}, \citenamefont
{Oyama}, \citenamefont {Gillitzer}, \citenamefont {Gilg}, \citenamefont
{Kn{\"u}lle}, \citenamefont {M{\"u}nch}, \citenamefont {Kienle},
\citenamefont {Schott}, \citenamefont {Geissel}, \citenamefont {Iwasa},\ and\
\citenamefont {M{\"u}nzenberg}}]{Yamazaki_1996_deeply_bound_states}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {T.}~\bibnamefont
{Yamazaki}}, \bibinfo {author} {\bibfnamefont {R.~S.}\ \bibnamefont
{Hayano}}, \bibinfo {author} {\bibfnamefont {K.}~\bibnamefont {Itahashi}},
\bibinfo {author} {\bibfnamefont {K.}~\bibnamefont {Oyama}}, \bibinfo
{author} {\bibfnamefont {A.}~\bibnamefont {Gillitzer}}, \bibinfo {author}
{\bibfnamefont {H.}~\bibnamefont {Gilg}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Kn{\"u}lle}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {M{\"u}nch}}, \bibinfo {author} {\bibfnamefont
{P.}~\bibnamefont {Kienle}}, \bibinfo {author} {\bibfnamefont
{W.}~\bibnamefont {Schott}}, \bibinfo {author} {\bibfnamefont
{H.}~\bibnamefont {Geissel}}, \bibinfo {author} {\bibfnamefont
{N.}~\bibnamefont {Iwasa}},\ and\ \bibinfo {author} {\bibfnamefont
{G.}~\bibnamefont {M{\"u}nzenberg}},\ }\bibfield {title} {\bibinfo {title}
{Discovery of deeply bound ${\pi}^{-}$ states in the
${}^{208}\mathrm{Pb}{(d,}^{3}\mathrm{He})$ reaction},\ }\href
{https://doi.org/10.1007/s002180050101} {\bibfield {journal} {\bibinfo
{journal} {Zeitschrift f{\"u}r Physik A Hadrons and Nuclei}\ }\textbf
{\bibinfo {volume} {355}},\ \bibinfo {pages} {219} (\bibinfo {year}
{1996})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Gilg}\ \emph {et~al.}(2000)\citenamefont {Gilg},
\citenamefont {Gillitzer}, \citenamefont {Kn\"ulle}, \citenamefont {M\"unch},
\citenamefont {Schott}, \citenamefont {Kienle}, \citenamefont {Itahashi},
\citenamefont {Oyama}, \citenamefont {Hayano}, \citenamefont {Geissel},
\citenamefont {Iwasa}, \citenamefont {M\"unzenberg},\ and\ \citenamefont
{Yamazaki}}]{Gilg2000_deeply_bound_pionic_207Pb}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {H.}~\bibnamefont
{Gilg}}, \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Gillitzer}},
\bibinfo {author} {\bibfnamefont {M.}~\bibnamefont {Kn\"ulle}}, \bibinfo
{author} {\bibfnamefont {M.}~\bibnamefont {M\"unch}}, \bibinfo {author}
{\bibfnamefont {W.}~\bibnamefont {Schott}}, \bibinfo {author} {\bibfnamefont
{P.}~\bibnamefont {Kienle}}, \bibinfo {author} {\bibfnamefont
{K.}~\bibnamefont {Itahashi}}, \bibinfo {author} {\bibfnamefont
{K.}~\bibnamefont {Oyama}}, \bibinfo {author} {\bibfnamefont {R.~S.}\
\bibnamefont {Hayano}}, \bibinfo {author} {\bibfnamefont {H.}~\bibnamefont
{Geissel}}, \bibinfo {author} {\bibfnamefont {N.}~\bibnamefont {Iwasa}},
\bibinfo {author} {\bibfnamefont {G.}~\bibnamefont {M\"unzenberg}},\ and\
\bibinfo {author} {\bibfnamefont {T.}~\bibnamefont {Yamazaki}},\ }\bibfield
{title} {\bibinfo {title} {Deeply bound ${\ensuremath{\pi}}^{\ensuremath{-}}$
states in ${}^{207}\mathrm{Pb}$ formed in the
${}^{208}\mathrm{Pb}{(d,}^{3}\mathrm{He})$ reaction. i. experimental method
and results},\ }\href {https://doi.org/10.1103/PhysRevC.62.025201} {\bibfield
{journal} {\bibinfo {journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume}
{62}},\ \bibinfo {pages} {025201} (\bibinfo {year} {2000})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {Geissel}\ \emph {et~al.}(2002)\citenamefont
{Geissel}, \citenamefont {Gilg}, \citenamefont {Gillitzer}, \citenamefont
{Hayano}, \citenamefont {Hirenzaki}, \citenamefont {Itahashi}, \citenamefont
{Iwasaki}, \citenamefont {Kienle}, \citenamefont {M\"unch}, \citenamefont
{M\"unzenberg}, \citenamefont {Schott}, \citenamefont {Suzuki}, \citenamefont
{Tomono}, \citenamefont {Weick}, \citenamefont {Yamazaki},\ and\
\citenamefont {Yoneyama}}]{Geissel2002_deeply_bound_205Pb}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {H.}~\bibnamefont
{Geissel}}, \bibinfo {author} {\bibfnamefont {H.}~\bibnamefont {Gilg}},
\bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Gillitzer}}, \bibinfo
{author} {\bibfnamefont {R.~S.}\ \bibnamefont {Hayano}}, \bibinfo {author}
{\bibfnamefont {S.}~\bibnamefont {Hirenzaki}}, \bibinfo {author}
{\bibfnamefont {K.}~\bibnamefont {Itahashi}}, \bibinfo {author}
{\bibfnamefont {M.}~\bibnamefont {Iwasaki}}, \bibinfo {author} {\bibfnamefont
{P.}~\bibnamefont {Kienle}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {M\"unch}}, \bibinfo {author} {\bibfnamefont
{G.}~\bibnamefont {M\"unzenberg}}, \bibinfo {author} {\bibfnamefont
{W.}~\bibnamefont {Schott}}, \bibinfo {author} {\bibfnamefont
{K.}~\bibnamefont {Suzuki}}, \bibinfo {author} {\bibfnamefont
{D.}~\bibnamefont {Tomono}}, \bibinfo {author} {\bibfnamefont
{H.}~\bibnamefont {Weick}}, \bibinfo {author} {\bibfnamefont
{T.}~\bibnamefont {Yamazaki}},\ and\ \bibinfo {author} {\bibfnamefont
{T.}~\bibnamefont {Yoneyama}},\ }\bibfield {title} {\bibinfo {title} {Deeply
bound $1\mathit{s}$ and $2\mathit{p}$ pionic states in $^{205}${Pb} and
determination of the $\mathit{s}$-wave part of the pion-nucleus
interaction},\ }\href {https://doi.org/10.1103/PhysRevLett.88.122301}
{\bibfield {journal} {\bibinfo {journal} {Phys. Rev. Lett.}\ }\textbf
{\bibinfo {volume} {88}},\ \bibinfo {pages} {122301} (\bibinfo {year}
{2002})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Nieves}\ and\ \citenamefont
{Oset}(1993)}]{Nieves1993_radiative_trapping}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {J.}~\bibnamefont
{Nieves}}\ and\ \bibinfo {author} {\bibfnamefont {E.}~\bibnamefont {Oset}},\
}\bibfield {title} {\bibinfo {title} {Direct production of pionic atoms from
radiative trapping of pions in flight},\ }\href
{https://doi.org/https://doi.org/10.1016/0375-9474(93)90665-K} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics A}\ }\textbf {\bibinfo
{volume} {553}},\ \bibinfo {pages} {595 } (\bibinfo {year}
{1993})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Raywood}\ \emph {et~al.}(1997)\citenamefont
{Raywood}, \citenamefont {Lange}, \citenamefont {Jones}, \citenamefont
{Pavan}, \citenamefont {Sevior}, \citenamefont {Hutcheon}, \citenamefont
{Olin}, \citenamefont {Ottewell}, \citenamefont {Yen}, \citenamefont {Lee},
\citenamefont {Sim}, \citenamefont {Altman}, \citenamefont {Friedman},\ and\
\citenamefont {Trudel}}]{Raywood1997_208Pb_radiative_capture}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.~J.}\ \bibnamefont
{Raywood}}, \bibinfo {author} {\bibfnamefont {J.~B.}\ \bibnamefont {Lange}},
\bibinfo {author} {\bibfnamefont {G.}~\bibnamefont {Jones}}, \bibinfo
{author} {\bibfnamefont {M.}~\bibnamefont {Pavan}}, \bibinfo {author}
{\bibfnamefont {M.~E.}\ \bibnamefont {Sevior}}, \bibinfo {author}
{\bibfnamefont {D.~A.}\ \bibnamefont {Hutcheon}}, \bibinfo {author}
{\bibfnamefont {A.}~\bibnamefont {Olin}}, \bibinfo {author} {\bibfnamefont
{D.}~\bibnamefont {Ottewell}}, \bibinfo {author} {\bibfnamefont
{S.}~\bibnamefont {Yen}}, \bibinfo {author} {\bibfnamefont {S.~J.}\
\bibnamefont {Lee}}, \bibinfo {author} {\bibfnamefont {K.~S.}\ \bibnamefont
{Sim}}, \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Altman}},
\bibinfo {author} {\bibfnamefont {E.}~\bibnamefont {Friedman}},\ and\
\bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Trudel}},\ }\bibfield
{title} {\bibinfo {title} {Search for deeply bound pionic states in
${}^{208}${Pb} via radiative atomic capture of negative pions},\ }\href
{https://doi.org/10.1103/PhysRevC.55.2492} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume} {55}},\ \bibinfo
{pages} {2492} (\bibinfo {year} {1997})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Tain}\ \emph {et~al.}(2004)\citenamefont {Tain} \emph
{et~al.}}]{Tain2004_gamma-ray_ALBA}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {J.~L.}\ \bibnamefont
{Tain}} \emph {et~al.},\ }\href
{https://intranet.cells.es/static/Gamma_rays.pdf} {\bibinfo {title} {Proposal
for the {C}onstruction of a {Gamma-Ray Beam Line} at the {S}panish
{Synchrotron} {ALBA}}} (\bibinfo {year} {2004})\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Strandberg}\ \emph {et~al.}(2020)\citenamefont
{Strandberg}, \citenamefont {Fissum}, \citenamefont {Annand}, \citenamefont
{Briscoe}, \citenamefont {Brudvik}, \citenamefont {Cividini}, \citenamefont
{Clark}, \citenamefont {Downie}, \citenamefont {England}, \citenamefont
{Feldman}, \citenamefont {Glazier}, \citenamefont {Hamilton}, \citenamefont
{Hansen}, \citenamefont {Isaksson}, \citenamefont {Al~Jebali}, \citenamefont
{Kovash}, \citenamefont {Kudryavtsev}, \citenamefont {Lensky}, \citenamefont
{Lipschutz}, \citenamefont {Lundin}, \citenamefont {Meshkian}, \citenamefont
{Middleton}, \citenamefont {Myers}, \citenamefont {O'Donnell}, \citenamefont
{O'Rielly}, \citenamefont {Oussena}, \citenamefont {Preston}, \citenamefont
{Schr\"oder}, \citenamefont {Seitz}, \citenamefont {Strakovsky},
\citenamefont {Taragin},\ and\ \citenamefont
{Tarasov}}]{Strandberg2020_threshold_pion_from_deuteron}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {B.}~\bibnamefont
{Strandberg}}, \bibinfo {author} {\bibfnamefont {K.~G.}\ \bibnamefont
{Fissum}}, \bibinfo {author} {\bibfnamefont {J.~R.~M.}\ \bibnamefont
{Annand}}, \bibinfo {author} {\bibfnamefont {W.~J.}\ \bibnamefont {Briscoe}},
\bibinfo {author} {\bibfnamefont {J.}~\bibnamefont {Brudvik}}, \bibinfo
{author} {\bibfnamefont {F.}~\bibnamefont {Cividini}}, \bibinfo {author}
{\bibfnamefont {L.}~\bibnamefont {Clark}}, \bibinfo {author} {\bibfnamefont
{E.~J.}\ \bibnamefont {Downie}}, \bibinfo {author} {\bibfnamefont
{K.}~\bibnamefont {England}}, \bibinfo {author} {\bibfnamefont
{G.}~\bibnamefont {Feldman}}, \bibinfo {author} {\bibfnamefont {D.~I.}\
\bibnamefont {Glazier}}, \bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Hamilton}}, \bibinfo {author} {\bibfnamefont {K.}~\bibnamefont {Hansen}},
\bibinfo {author} {\bibfnamefont {L.}~\bibnamefont {Isaksson}}, \bibinfo
{author} {\bibfnamefont {R.}~\bibnamefont {Al~Jebali}}, \bibinfo {author}
{\bibfnamefont {M.~A.}\ \bibnamefont {Kovash}}, \bibinfo {author}
{\bibfnamefont {A.~E.}\ \bibnamefont {Kudryavtsev}}, \bibinfo {author}
{\bibfnamefont {V.}~\bibnamefont {Lensky}}, \bibinfo {author} {\bibfnamefont
{S.}~\bibnamefont {Lipschutz}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Lundin}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Meshkian}}, \bibinfo {author} {\bibfnamefont {D.~G.}\
\bibnamefont {Middleton}}, \bibinfo {author} {\bibfnamefont {L.~S.}\
\bibnamefont {Myers}}, \bibinfo {author} {\bibfnamefont {D.}~\bibnamefont
{O'Donnell}}, \bibinfo {author} {\bibfnamefont {G.~V.}\ \bibnamefont
{O'Rielly}}, \bibinfo {author} {\bibfnamefont {B.}~\bibnamefont {Oussena}},
\bibinfo {author} {\bibfnamefont {M.~F.}\ \bibnamefont {Preston}}, \bibinfo
{author} {\bibfnamefont {B.}~\bibnamefont {Schr\"oder}}, \bibinfo {author}
{\bibfnamefont {B.}~\bibnamefont {Seitz}}, \bibinfo {author} {\bibfnamefont
{I.~I.}\ \bibnamefont {Strakovsky}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Taragin}},\ and\ \bibinfo {author} {\bibfnamefont {V.~E.}\
\bibnamefont {Tarasov}} (\bibinfo {collaboration} {PIONS@MAX-lab
Collaboration}),\ }\bibfield {title} {\bibinfo {title} {Near-threshold
${\ensuremath{\pi}}^{\ensuremath{-}}$ photoproduction on the deuteron},\
}\href {https://doi.org/10.1103/PhysRevC.101.035207} {\bibfield {journal}
{\bibinfo {journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume} {101}},\
\bibinfo {pages} {035207} (\bibinfo {year} {2020})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Landau}\ and\ \citenamefont
{Lifshitz}(1958)}]{Landau1958}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {L.~D.}\ \bibnamefont
{Landau}}\ and\ \bibinfo {author} {\bibfnamefont {E.~M.}\ \bibnamefont
{Lifshitz}},\ }\href@noop {} {\emph {\bibinfo {title} {Quantum Mechanics}}}\
(\bibinfo {publisher} {Pergamon Press Ltd},\ \bibinfo {year}
{1958})\BibitemShut {NoStop}%
\bibitem [{Note3()}]{Note3}%
\BibitemOpen
\bibinfo {note} {Surprisingly, such a naive estimate gives a correct result
for the decay $\Delta \to N \pi \gamma $, where $\Gamma _{\gamma }/\Gamma
_{tot}\approx 0.006$ \cite {Zyla2020_Particle_Data}}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Petrukhin}\ and\ \citenamefont
{Prokoshkin}(1965)}]{Petrukhin1965_radiative_capture}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {V.~I.}\ \bibnamefont
{Petrukhin}}\ and\ \bibinfo {author} {\bibfnamefont {Y.~D.}\ \bibnamefont
{Prokoshkin}},\ }\bibfield {title} {\bibinfo {title} {Radiative capture of
stopped negative pions by nuclei},\ }\href
{https://doi.org/https://doi.org/10.1016/0029-5582(65)90796-0} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics}\ }\textbf {\bibinfo {volume}
{66}},\ \bibinfo {pages} {669 } (\bibinfo {year} {1965})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {Delorme}\ and\ \citenamefont
{Ericson}(1966)}]{Delorme1966_radiative_capture}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {J.}~\bibnamefont
{Delorme}}\ and\ \bibinfo {author} {\bibfnamefont {T.}~\bibnamefont
{Ericson}},\ }\bibfield {title} {\bibinfo {title} {Radiative
${\pi}^{\ensuremath{-}}$ capture in nuclei},\ }\href
{https://doi.org/https://doi.org/10.1016/0031-9163(66)91357-6} {\bibfield
{journal} {\bibinfo {journal} {Physics Letters}\ }\textbf {\bibinfo {volume}
{21}},\ \bibinfo {pages} {98 } (\bibinfo {year} {1966})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {Bernstein}\ \emph {et~al.}(1976)\citenamefont
{Bernstein}, \citenamefont {Paras}, \citenamefont {Turchinetz}, \citenamefont
{Chasan},\ and\ \citenamefont
{Booth}}]{Bernstein1976_pion_threshold_12C_12N}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {A.~M.}\ \bibnamefont
{Bernstein}}, \bibinfo {author} {\bibfnamefont {N.}~\bibnamefont {Paras}},
\bibinfo {author} {\bibfnamefont {W.}~\bibnamefont {Turchinetz}}, \bibinfo
{author} {\bibfnamefont {B.}~\bibnamefont {Chasan}},\ and\ \bibinfo {author}
{\bibfnamefont {E.~C.}\ \bibnamefont {Booth}},\ }\bibfield {title} {\bibinfo
{title} {Reaction
$^{12}\mathrm{C}(\ensuremath{\gamma},{\ensuremath{\pi}}^{\ensuremath{-}})^{12}\mathrm{N}$
near the threshold},\ }\href {https://doi.org/10.1103/PhysRevLett.37.819}
{\bibfield {journal} {\bibinfo {journal} {Phys. Rev. Lett.}\ }\textbf
{\bibinfo {volume} {37}},\ \bibinfo {pages} {819} (\bibinfo {year}
{1976})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Bosted}\ \emph {et~al.}(1979)\citenamefont {Bosted},
\citenamefont {Blomqvist},\ and\ \citenamefont
{Bernstein}}]{Bosted1979_near_threshold_cross-section}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {P.~E.}\ \bibnamefont
{Bosted}}, \bibinfo {author} {\bibfnamefont {K.~I.}\ \bibnamefont
{Blomqvist}},\ and\ \bibinfo {author} {\bibfnamefont {A.~M.}\ \bibnamefont
{Bernstein}},\ }\bibfield {title} {\bibinfo {title} {Total cross section for
($\ensuremath{\gamma}$, ${\ensuremath{\pi}}^{\ensuremath{-}}$) reactions in
the $\ensuremath{\Delta}$ region},\ }\href
{https://doi.org/10.1103/PhysRevLett.43.1473} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev. Lett.}\ }\textbf {\bibinfo {volume} {43}},\ \bibinfo
{pages} {1473} (\bibinfo {year} {1979})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Singham}\ and\ \citenamefont
{Tabakin}(1981)}]{Singham1981_threshold_crosec}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {M.}~\bibnamefont
{Singham}}\ and\ \bibinfo {author} {\bibfnamefont {F.}~\bibnamefont
{Tabakin}},\ }\bibfield {title} {\bibinfo {title} {Charged pion
photoproduction from light nuclei},\ }\href
{https://doi.org/https://doi.org/10.1016/0003-4916(81)90144-5} {\bibfield
{journal} {\bibinfo {journal} {Annals of Physics}\ }\textbf {\bibinfo
{volume} {135}},\ \bibinfo {pages} {71 } (\bibinfo {year}
{1981})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Min}\ \emph {et~al.}(1976)\citenamefont {Min},
\citenamefont {Stoler}, \citenamefont {Trentalange}, \citenamefont {Winhold},
\citenamefont {Yergin}, \citenamefont {Bernstein}, \citenamefont
{Turchinetz},\ and\ \citenamefont {Sastry}}]{Min1976_threshold_11C}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Min}}, \bibinfo {author} {\bibfnamefont {P.}~\bibnamefont {Stoler}},
\bibinfo {author} {\bibfnamefont {S.}~\bibnamefont {Trentalange}}, \bibinfo
{author} {\bibfnamefont {E.~J.}\ \bibnamefont {Winhold}}, \bibinfo {author}
{\bibfnamefont {P.~F.}\ \bibnamefont {Yergin}}, \bibinfo {author}
{\bibfnamefont {A.~M.}\ \bibnamefont {Bernstein}}, \bibinfo {author}
{\bibfnamefont {W.}~\bibnamefont {Turchinetz}},\ and\ \bibinfo {author}
{\bibfnamefont {K.~S.~R.}\ \bibnamefont {Sastry}},\ }\bibfield {title}
{\bibinfo {title}
{$^{11}\mathrm{B}(\ensuremath{\gamma},{\ensuremath{\pi}}^{\ensuremath{-}})^{11}\mathrm{C}$
cross section near threshold},\ }\href
{https://doi.org/10.1103/PhysRevC.14.807} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume} {14}},\ \bibinfo
{pages} {807} (\bibinfo {year} {1976})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Epstein}\ \emph {et~al.}(1978)\citenamefont
{Epstein}, \citenamefont {Singham},\ and\ \citenamefont
{Tabakin}}]{Epstein1978_threshold_12N}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {G.~N.}\ \bibnamefont
{Epstein}}, \bibinfo {author} {\bibfnamefont {M.~K.}\ \bibnamefont
{Singham}},\ and\ \bibinfo {author} {\bibfnamefont {F.}~\bibnamefont
{Tabakin}},\ }\bibfield {title} {\bibinfo {title}
{$^{12}\mathrm{C}(\ensuremath{\gamma},{\ensuremath{\pi}}^{\ensuremath{-}})^{12}\mathrm{N}_{\mathrm{g}.\mathrm{s}.}$
near threshold},\ }\href {https://doi.org/10.1103/PhysRevC.17.702} {\bibfield
{journal} {\bibinfo {journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume}
{17}},\ \bibinfo {pages} {702} (\bibinfo {year} {1978})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {DeCarlo}\ \emph {et~al.}(1980)\citenamefont
{DeCarlo}, \citenamefont {Freed}, \citenamefont {Rhodes}, \citenamefont
{B\"ulow}, \citenamefont {Jonsson}, \citenamefont {Lindgren},\ and\
\citenamefont {Pettersson}}]{DeCarlo1980_threshold_14O}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {V.}~\bibnamefont
{DeCarlo}}, \bibinfo {author} {\bibfnamefont {N.}~\bibnamefont {Freed}},
\bibinfo {author} {\bibfnamefont {W.}~\bibnamefont {Rhodes}}, \bibinfo
{author} {\bibfnamefont {B.}~\bibnamefont {B\"ulow}}, \bibinfo {author}
{\bibfnamefont {G.~G.}\ \bibnamefont {Jonsson}}, \bibinfo {author}
{\bibfnamefont {K.}~\bibnamefont {Lindgren}},\ and\ \bibinfo {author}
{\bibfnamefont {R.}~\bibnamefont {Pettersson}},\ }\bibfield {title}
{\bibinfo {title} {Photoproduction of ${\ensuremath{\pi}}^{\ensuremath{-}}$
on $^{14}\mathrm{N}$},\ }\href {https://doi.org/10.1103/PhysRevC.21.1460}
{\bibfield {journal} {\bibinfo {journal} {Phys. Rev. C}\ }\textbf {\bibinfo
{volume} {21}},\ \bibinfo {pages} {1460} (\bibinfo {year}
{1980})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Singham}\ \emph {et~al.}(1979)\citenamefont
{Singham}, \citenamefont {Epstein},\ and\ \citenamefont
{Tabakin}}]{Singham1979_threshold_7Li_12C}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {M.~K.}\ \bibnamefont
{Singham}}, \bibinfo {author} {\bibfnamefont {G.~N.}\ \bibnamefont
{Epstein}},\ and\ \bibinfo {author} {\bibfnamefont {F.}~\bibnamefont
{Tabakin}},\ }\bibfield {title} {\bibinfo {title} {Photoproduction of
{0-200-MeV Pions} from $^{12}\mathrm{C}$ and $^{7}\mathrm{Li}$},\ }\href
{https://doi.org/10.1103/PhysRevLett.43.1476} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev. Lett.}\ }\textbf {\bibinfo {volume} {43}},\ \bibinfo
{pages} {1476} (\bibinfo {year} {1979})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Sakamoto}\ \emph {et~al.}(1989)\citenamefont
{Sakamoto}, \citenamefont {Yoshida}, \citenamefont {Kubota}, \citenamefont
{Fukasawa}, \citenamefont {Kunugise}, \citenamefont {Hamajima}, \citenamefont
{Shibata},\ and\ \citenamefont {Fujiwara}}]{Sakamoto1989_51Cr_from_51V}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Sakamoto}}, \bibinfo {author} {\bibfnamefont {M.}~\bibnamefont {Yoshida}},
\bibinfo {author} {\bibfnamefont {Y.}~\bibnamefont {Kubota}}, \bibinfo
{author} {\bibfnamefont {T.}~\bibnamefont {Fukasawa}}, \bibinfo {author}
{\bibfnamefont {A.}~\bibnamefont {Kunugise}}, \bibinfo {author}
{\bibfnamefont {Y.}~\bibnamefont {Hamajima}}, \bibinfo {author}
{\bibfnamefont {S.}~\bibnamefont {Shibata}},\ and\ \bibinfo {author}
{\bibfnamefont {I.}~\bibnamefont {Fujiwara}},\ }\bibfield {title} {\bibinfo
{title} {Chromium and titanium isotopes produced in photonuclear reactions of
vanadium, revisited},\ }\href
{https://doi.org/https://doi.org/10.1016/0375-9474(89)90156-5} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics A}\ }\textbf {\bibinfo
{volume} {501}},\ \bibinfo {pages} {693 } (\bibinfo {year}
{1989})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Sakamoto}\ \emph {et~al.}(1990)\citenamefont
{Sakamoto}, \citenamefont {Hamajima}, \citenamefont {Soto}, \citenamefont
{Kubota}, \citenamefont {Yoshida}, \citenamefont {Kunugise}, \citenamefont
{Masatani}, \citenamefont {Shibata}, \citenamefont {Imamura}, \citenamefont
{Furukawa},\ and\ \citenamefont {Fujiwara}}]{Sakamoto1990_133Ba_from_133Cs}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Sakamoto}}, \bibinfo {author} {\bibfnamefont {Y.}~\bibnamefont {Hamajima}},
\bibinfo {author} {\bibfnamefont {M.}~\bibnamefont {Soto}}, \bibinfo {author}
{\bibfnamefont {Y.}~\bibnamefont {Kubota}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Yoshida}}, \bibinfo {author} {\bibfnamefont
{A.}~\bibnamefont {Kunugise}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Masatani}}, \bibinfo {author} {\bibfnamefont
{S.}~\bibnamefont {Shibata}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Imamura}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Furukawa}},\ and\ \bibinfo {author} {\bibfnamefont
{I.}~\bibnamefont {Fujiwara}},\ }\bibfield {title} {\bibinfo {title} {Barium
and xenon isotope yields in photopion reactions of $^{133}\mathrm{Cs}$},\
}\href {https://doi.org/10.1103/PhysRevC.42.1545} {\bibfield {journal}
{\bibinfo {journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume} {42}},\
\bibinfo {pages} {1545} (\bibinfo {year} {1990})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Blomqvist}\ \emph {et~al.}(1978)\citenamefont
{Blomqvist}, \citenamefont {B{\"u}low}, \citenamefont {Fredrikson},
\citenamefont {Johnsson}, \citenamefont {Jonsson}, \citenamefont {Lindgren},
\citenamefont {Nilsson}, \citenamefont {Petersson}, \citenamefont {Glomset},
\citenamefont {Freed},\ and\ \citenamefont
{Rhodes}}]{Blomqvist1978_197Hg_from_197Au}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {I.}~\bibnamefont
{Blomqvist}}, \bibinfo {author} {\bibfnamefont {B.}~\bibnamefont
{B{\"u}low}}, \bibinfo {author} {\bibfnamefont {A.}~\bibnamefont
{Fredrikson}}, \bibinfo {author} {\bibfnamefont {B.}~\bibnamefont
{Johnsson}}, \bibinfo {author} {\bibfnamefont {G.~G.}\ \bibnamefont
{Jonsson}}, \bibinfo {author} {\bibfnamefont {K.}~\bibnamefont {Lindgren}},
\bibinfo {author} {\bibfnamefont {M.}~\bibnamefont {Nilsson}}, \bibinfo
{author} {\bibfnamefont {R.}~\bibnamefont {Petersson}}, \bibinfo {author}
{\bibfnamefont {O.}~\bibnamefont {Glomset}}, \bibinfo {author} {\bibfnamefont
{N.}~\bibnamefont {Freed}},\ and\ \bibinfo {author} {\bibfnamefont
{W.}~\bibnamefont {Rhodes}},\ }\bibfield {title} {\bibinfo {title} {Pion
photoproduction on $^{197}${Au} to ground and isomeric states in
$^{197}${Hg}},\ }\href {https://doi.org/10.1007/BF01415199} {\bibfield
{journal} {\bibinfo {journal} {Z Physik A}\ }\textbf {\bibinfo {volume}
{288}},\ \bibinfo {pages} {313} (\bibinfo {year} {1978})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {Sakamoto}\ \emph {et~al.}(1999)\citenamefont
{Sakamoto}, \citenamefont {Sarkar}, \citenamefont {Oura}, \citenamefont
{Haba}, \citenamefont {Matsumura}, \citenamefont {Miyamoto}, \citenamefont
{Shibata}, \citenamefont {Furukawa},\ and\ \citenamefont
{Fujiwara}}]{Sakamoto1999_complex_nuclei}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {K.}~\bibnamefont
{Sakamoto}}, \bibinfo {author} {\bibfnamefont {S.~R.}\ \bibnamefont
{Sarkar}}, \bibinfo {author} {\bibfnamefont {Y.}~\bibnamefont {Oura}},
\bibinfo {author} {\bibfnamefont {H.}~\bibnamefont {Haba}}, \bibinfo {author}
{\bibfnamefont {H.}~\bibnamefont {Matsumura}}, \bibinfo {author}
{\bibfnamefont {Y.}~\bibnamefont {Miyamoto}}, \bibinfo {author}
{\bibfnamefont {S.}~\bibnamefont {Shibata}}, \bibinfo {author} {\bibfnamefont
{M.}~\bibnamefont {Furukawa}},\ and\ \bibinfo {author} {\bibfnamefont
{I.}~\bibnamefont {Fujiwara}},\ }\bibfield {title} {\bibinfo {title} {Mass
yield features of $(\ensuremath{\gamma},{\ensuremath{\pi}}^{+})$ and
$(\ensuremath{\gamma},{\ensuremath{\pi}}^{\ensuremath{-}}xn)$ reactions for
$x=0-9$ on complex nuclei in the \ensuremath{\Delta} region},\ }\href
{https://doi.org/10.1103/PhysRevC.59.1497} {\bibfield {journal} {\bibinfo
{journal} {Phys. Rev. C}\ }\textbf {\bibinfo {volume} {59}},\ \bibinfo
{pages} {1497} (\bibinfo {year} {1999})}\BibitemShut {NoStop}%
\bibitem [{Note4()}]{Note4}%
\BibitemOpen
\bibinfo {note} {There are indications that, due to pion repulsion from the
nucleus only surface nucleons contribute to the coherent cross-section, see,
for example, Ref.\protect \tmspace +\thinmuskip {.1667em}\cite
{Boffi1991_coherent_pion0}. If this is the case, we may estimate the
suppression factor as $1/(4Z)^{2/3}$.}\BibitemShut {Stop}%
\bibitem [{\citenamefont {Boffi}\ \emph {et~al.}(1991)\citenamefont {Boffi},
\citenamefont {Bracci},\ and\ \citenamefont
{Christillin}}]{Boffi1991_coherent_pion0}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {S.}~\bibnamefont
{Boffi}}, \bibinfo {author} {\bibfnamefont {L.}~\bibnamefont {Bracci}},\ and\
\bibinfo {author} {\bibfnamefont {P.}~\bibnamefont {Christillin}},\
}\bibfield {title} {\bibinfo {title} {Coherent ${(\gamma, \pi^0)}$ on
nuclei},\ }\href {https://doi.org/10.1007/BF02820558} {\bibfield {journal}
{\bibinfo {journal} {Nuov Cim A}\ }\textbf {\bibinfo {volume} {104}},\
\bibinfo {pages} {843} (\bibinfo {year} {1991})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {{Particle Data Group}}\ \emph
{et~al.}(2020)\citenamefont {{Particle Data Group}}, \citenamefont {Zyla}
\emph {et~al.}}]{Zyla2020_Particle_Data}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibnamefont {{Particle Data
Group}}}, \bibinfo {author} {\bibfnamefont {P.~A.}\ \bibnamefont {Zyla}},
\emph {et~al.},\ }\bibfield {title} {\bibinfo {title} {Review of particle
physics},\ }\href {https://doi.org/10.1093/ptep/ptaa104} {\bibfield
{journal} {\bibinfo {journal} {Progress of Theoretical and Experimental
Physics}\ }\textbf {\bibinfo {volume} {2020}},\ \bibinfo {pages} {083C01}
(\bibinfo {year} {2020})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Berestetskii}\ \emph {et~al.}(1982)\citenamefont
{Berestetskii}, \citenamefont {Pitaevskii},\ and\ \citenamefont
{Lifshitz}}]{Berestetskii_QED}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {V.~B.}\ \bibnamefont
{Berestetskii}}, \bibinfo {author} {\bibfnamefont {L.~P.}\ \bibnamefont
{Pitaevskii}},\ and\ \bibinfo {author} {\bibfnamefont {E.~M.}\ \bibnamefont
{Lifshitz}},\ }\href@noop {} {\emph {\bibinfo {title} {Quantum
Electrodynamics}}}\ (\bibinfo {publisher} {Pergamon Press Ltd},\ \bibinfo
{year} {1982})\BibitemShut {NoStop}%
\bibitem [{Note5()}]{Note5}%
\BibitemOpen
\bibinfo {note} {There is an extra coefficient $Z/(Z-1)$ in Eq.\protect
\tmspace +\thinmuskip {.1667em}\protect \textup {\hbox {\mathsurround \z@
\protect \normalfont (\ignorespaces \ref
{Eq:max_production_rate_from_free_pion}\unskip \@@italiccorr )}} that should
be taken into account for light nuclei.}\BibitemShut {Stop}%
\bibitem [{\citenamefont {Toki}\ \emph {et~al.}(1989)\citenamefont {Toki},
\citenamefont {Hirenzaki}, \citenamefont {Yamazaki},\ and\ \citenamefont
{Hayano}}]{Toki1989_pion_width_1s}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {H.}~\bibnamefont
{Toki}}, \bibinfo {author} {\bibfnamefont {S.}~\bibnamefont {Hirenzaki}},
\bibinfo {author} {\bibfnamefont {T.}~\bibnamefont {Yamazaki}},\ and\
\bibinfo {author} {\bibfnamefont {R.}~\bibnamefont {Hayano}},\ }\bibfield
{title} {\bibinfo {title} {Structure and formation of deeply-bound pionic
atoms},\ }\href
{https://doi.org/https://doi.org/10.1016/0375-9474(89)90154-1} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics A}\ }\textbf {\bibinfo
{volume} {501}},\ \bibinfo {pages} {653 } (\bibinfo {year}
{1989})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Backenstoss}\ \emph {et~al.}(1974)\citenamefont
{Backenstoss}, \citenamefont {Egger}, \citenamefont {{von Egidy}},
\citenamefont {Hagelberg}, \citenamefont {Herrlander}, \citenamefont {Koch},
\citenamefont {Povel}, \citenamefont {Schwitter},\ and\ \citenamefont
{Tauscher}}]{Backenstoss1974_X-ray_helium}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {G.}~\bibnamefont
{Backenstoss}}, \bibinfo {author} {\bibfnamefont {J.}~\bibnamefont {Egger}},
\bibinfo {author} {\bibfnamefont {T.}~\bibnamefont {{von Egidy}}}, \bibinfo
{author} {\bibfnamefont {R.}~\bibnamefont {Hagelberg}}, \bibinfo {author}
{\bibfnamefont {C.}~\bibnamefont {Herrlander}}, \bibinfo {author}
{\bibfnamefont {H.}~\bibnamefont {Koch}}, \bibinfo {author} {\bibfnamefont
{H.}~\bibnamefont {Povel}}, \bibinfo {author} {\bibfnamefont
{A.}~\bibnamefont {Schwitter}},\ and\ \bibinfo {author} {\bibfnamefont
{L.}~\bibnamefont {Tauscher}},\ }\bibfield {title} {\bibinfo {title} {Pionic
and muonic {X}-ray transitions in liquid helium},\ }\href
{https://doi.org/https://doi.org/10.1016/0375-9474(74)90637-X} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics A}\ }\textbf {\bibinfo
{volume} {232}},\ \bibinfo {pages} {519 } (\bibinfo {year}
{1974})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Friedman}\ and\ \citenamefont
{Soff}(1985)}]{Friedman_1985_width_heavy_atoms}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {E.}~\bibnamefont
{Friedman}}\ and\ \bibinfo {author} {\bibfnamefont {G.}~\bibnamefont
{Soff}},\ }\bibfield {title} {\bibinfo {title} {Calculations of deeply bound
pionic states in heavy and superheavy atoms},\ }\href
{https://doi.org/10.1088/0305-4616/11/3/003} {\bibfield {journal} {\bibinfo
{journal} {Journal of Physics G: Nuclear Physics}\ }\textbf {\bibinfo
{volume} {11}},\ \bibinfo {pages} {L37} (\bibinfo {year} {1985})}\BibitemShut
{NoStop}%
\bibitem [{\citenamefont {Konijn}\ \emph {et~al.}(1990)\citenamefont {Konijn},
\citenamefont {{De Laat}}, \citenamefont {Taal},\ and\ \citenamefont
{Koch}}]{Konijn1990_widths_nl}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {J.}~\bibnamefont
{Konijn}}, \bibinfo {author} {\bibfnamefont {C.}~\bibnamefont {{De Laat}}},
\bibinfo {author} {\bibfnamefont {A.}~\bibnamefont {Taal}},\ and\ \bibinfo
{author} {\bibfnamefont {J.}~\bibnamefont {Koch}},\ }\bibfield {title}
{\bibinfo {title} {An improved parametrization of the optical potential for
pionic atoms},\ }\href
{https://doi.org/https://doi.org/10.1016/0375-9474(90)90435-O} {\bibfield
{journal} {\bibinfo {journal} {Nuclear Physics A}\ }\textbf {\bibinfo
{volume} {519}},\ \bibinfo {pages} {773 } (\bibinfo {year}
{1990})}\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Bethe}\ and\ \citenamefont
{Salpeter}(1957)}]{BetheSalpeter}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {H.~A.}\ \bibnamefont
{Bethe}}\ and\ \bibinfo {author} {\bibfnamefont {E.~E.}\ \bibnamefont
{Salpeter}},\ }\href@noop {} {\emph {\bibinfo {title} {Quantum Mechanics of
One- and Two-Electron Atoms}}}\ (\bibinfo {publisher} {Springer-Verlag
Berlin Heidelberg},\ \bibinfo {year} {1957})\BibitemShut {NoStop}%
\bibitem [{\citenamefont {Ericson}\ and\ \citenamefont
{Weise}(1988)}]{Ericson1988}%
\BibitemOpen
\bibfield {author} {\bibinfo {author} {\bibfnamefont {T.}~\bibnamefont
{Ericson}}\ and\ \bibinfo {author} {\bibfnamefont {W.}~\bibnamefont
{Weise}},\ }\href@noop {} {\emph {\bibinfo {title} {Pions and Nuclei}}}\
(\bibinfo {publisher} {Clarendon Press},\ \bibinfo {year}
{1988})\BibitemShut {NoStop}%
\end{thebibliography}%