1: %\documentstyle[epsfig,preprint,prc,aps,epsfig]{revtex}
2:
3:
4: \documentstyle[preprint,tighten,prc,aps,epsfig]{revtex}
5:
6: %\documentclass[twocolumn,showpacs]{revtex4}
7: %\documentclass[draft,showpacs]{revtex4}
8: %\usepackage{psfig}
9:
10: \newcommand{\be}{\begin{equation}}
11: \newcommand{\ee}{\end{equation}}
12: \newcommand{\bfeps}{\mbox {\boldmath $\epsilon$}}
13: \newcommand{\bfp}{\mbox {\boldmath $p$}}
14: \newcommand{\bfk}{\mbox {\boldmath $k$}}
15: \newcommand{\bfs}{\mbox {\boldmath $S$}}
16: \newcommand{\bfg}{\mbox {\boldmath $G$}}
17: \newcommand{\bfzeta}{\mbox {\boldmath $\zeta$}}
18: \newcommand{\bfsigma}{\mbox {\boldmath $\sigma$}}
19: \newcommand{\bftau}{\mbox {\boldmath $\tau$}}
20: \newcommand{\bfphi}{\mbox {\boldmath $\varphi$}}
21: \newcommand{\dsdo}{\frac {d\sigma}{d\Omega}}
22: \newcommand{\dso}{d\sigma /d\Omega}
23: \newcommand{\aofo}{$a^0_0$$\,\leftrightarrow\,$$f_0$}
24: \newcommand{\mix}{$a^0_0\,$$-$$f_0$\, }
25: \newcommand{\pieta}{$\pi^0$$-$$\eta$}
26: \newcommand{\aofomix}{\mbox {\boldmath $a_0\,$-$f_0$}}
27: \newcommand{\pndao}{\mbox {\boldmath $pn\to d a_0$}}
28: \newcommand{\vpn}{\mbox {\boldmath $\vec p n\to d a_0$}}
29: \newcommand{\dxl}{\overleftarrow{\partial_{\nu}}}
30: \newcommand{\dxr}{\overrightarrow{\partial_{\nu}}}
31: \newcommand{\dxb}{\stackrel{\leftrightarrow}{\partial}}
32:
33:
34: \begin{document}
35: \preprint{FZJ-IKP-TH-2002-05}
36: \title{Aspects of \aofomix\ mixing in the reaction \vpn\ }
37:
38: \author{A.E. Kudryavtsev$^1$, V.E. Tarasov$^1$,
39: J. Haidenbauer$^2$, C. Hanhart$^2$, and J. Speth$^2$}
40: \address{$^1$Institute of Theoretical and Experimental Physics, \\
41: B. Cheremushkinskaya 25, 117259 Moscow, Russia \\
42: $^2$Institut f\"ur Kernphysik, Forschungszentrum J\"ulich,
43: D-52425 J\"ulich }
44:
45: \date{\today}
46: \maketitle
47:
48: \begin{abstract}
49: Some aspects of \mix mixing effects in the reaction
50: ${\vec p} n\to da^0_0$ with perpendicular polarized proton beam are
51: discussed. An angular--asymmetry parameter $A$ is defined to study those
52: effect. It is shown that, for energies close to the production threshold,
53: the angular--asymmetry parameter $A(\theta, \varphi)$ is proportional to
54: the \mix mixing amplitude for arbitrary polar and azimuthal angles
55: $\theta$ and $\varphi$ of the outgoing $a_0$ meson.
56: This statement is also valid for arbitrary
57: energies, but then only at polar angles $\theta=0^0$ and $\theta=90^0$.
58: The mass dependence of the differential cross section $d\sigma/dm_{\pi^0\eta}$
59: in the reaction $pn\to d\pi^0\eta$ in the presence of \mix mixing
60: is also discussed.
61: \end{abstract}
62:
63: \pacs{13.60.Le; 13.75.-n; 14.40.Cs}
64:
65: % 1
66: \section{Introduction}
67: The nature of the lightest, virtually mass-degenerate, scalar mesons
68: $a_0$ (980) ($I^G J^{PC}=1^- 0^{++}$) and $f_0$ (980) ($0^+ 0^{++}$)
69: is an important and still unsolved problem of hadron physics.
70: The quark structure of
71: these mesons is not well established at present time.
72: This issue is also closely related to the very interesting problem
73: of $a_0\,$-$f_0$ mixing. A dynamical mechanism for this mixing
74: close to $K\bar K$ threshold was suggested around 20 years ago in
75: Ref.~\cite{Acha}.
76: Since that time a number of papers have been published in which different
77: aspects of $a_0\,$-$f_0$ mixing and the possibilities to measure
78: this effect have been discussed,
79: see, for example, Refs.~\cite{Barnes,Acha1,Krepl,Kerb,Kud,Grish}.
80: Recently, in Ref.~\cite{Close}, some new arguments were presented
81: in support of a fairly large \mix-mixing intensity.
82: Analyzing the experimental
83: data~\cite{Barb,Kirk} on the exclusive production of the $a_0$(980),
84: $a_2$(1320) and $f_0$(980) resonances in $pp$ collisions at
85: $\sqrt{s}=29.1$~GeV, the authors of Ref.~\cite{Close} came to the
86: conclusion that $80\pm 25$~\% of all $a^0_0$ mesons come from the
87: $f_0$(980).
88:
89: The results of Ref.~\cite{Close} provide certainly a
90: fairly strong indication for a large \mix\ mixing. However, for more
91: solid conclusions, in particular about the quantitative
92: value of the mixing, a thorough analysis is needed.
93: In this context it is important to keep in mind that the mixing
94: can manifest itself in many different physical processes and observables.
95: Thus, in this paper we want to discuss another
96: consequence of the \mix mixing effect, namely the angular
97: asymmetry of the reaction
98: \be
99: p\,n \to d\, a^0_0/f_0 \, .
100: \label{1}
101: \ee
102: with unpolarized and polarized proton beam.
103: Furthermore we study the influence of the finite width of the
104: scalar mesons on the unpolarized as well as the polarized differential
105: cross section. All these investigations are performed with special emphasis
106: on how to extract informations on the $a_0/f_0$ mixing.
107:
108: In the present paper we will address the $a_0$ and $f_0$ mesons
109: as resonances. However,
110: it should be stressed that all the results derived in this paper
111: do not need any assumption about the nature of those scalar mesons.
112: Even if these scalar mesons are dynamically generated
113: \cite{Jans,Oller97}, in the proximity
114: of the pole the scattering matrix can still be well approximated
115: by the propagation of quasiparticles corresponding to elementary
116: fields. The true nature of the
117: propagating object is then hidden in the effective parameters
118: of those elementary fields. This was shown rigorously by Weinberg
119: for the case where inelastic channels are absent \cite{weinberg}.
120:
121: The paper is organized as follows. In Section 2 we review some common
122: properties of the production of unstable particles. We start out
123: from the case of a single particle and then, in Sect. 3, generalize
124: the formalism to the two channel case relevant for the present study.
125: Section 4 contains a discussion of the general structure of the primary
126: production amplitude for small excess energies.
127: Section 5 is devoted to the study of different observables.
128: The main emphasis is put on the $\pi \eta$ final state. Two differential
129: observables are analyzed in detail, namely the double
130: differential cross section $d^2\sigma/d\Omega/dm^2$ as well as the
131: analyzing power. The former observable was studied recently in a series of
132: papers, however, without investigating the $m^2$ dependence more thoroughly.
133: The latter, on the other hand, was not discussed in detail before at all.
134: We close with a short summary and conclusions.
135:
136: % 2
137: \section{The production of unstable particles}
138:
139: Let us start with the general expression for the cross section of the
140: reaction $NN \to dX \ ,$ where $X$ denotes the decay products of an
141: unstable meson.
142: We start the discussion by assuming the presence of one mesonic resonance
143: only. The generalization to more than one resonances will be done
144: in the next section. The main purpose of this section is to
145: introduce our notation and to remind the reader on the impact of
146: a finite decay width on observables.
147:
148: The reaction cross section in the center-of-mass (CM) system
149: be can expressed as
150: \be
151: d\sigma_X = (2\pi)^4 \, \frac{1}{4|\vec p \,|\sqrt{s}} \,
152: |{\cal A}_X|^2 d\Phi_{\mu+1}(P;p_d,k_1,...,k_\mu) \ .
153: \label{dsig}
154: \ee
155: Here $\vec p $ denotes the CM momentum of the initial protons and
156: $P$ the total initial four-momentum. The reaction amplitude ${\cal A}_X$
157: is given by
158: \be
159: {\cal A}_X = W_X(k_1,...,k_\mu)G(m^2){\cal M} \ .
160: \label{adef}
161: \ee
162: where the primary production amplitude is denoted by ${\cal M}$.
163: We assume the unstable meson, whose propagation is described
164: by $G(m^2)$, to decay into the $\mu$ particles of the final state $X$
165: through the vertex function $W_X$. $m^2$ is the total invariant
166: mass of the final $\mu$-particles system and is given by $m^2=(\sum k_i)^2$,
167: with $k_i$, $i=1,...,\mu$ being the four-momenta of the $\mu$ decay particles.
168: The phase space of the final $\mu$ particles and the final deuteron (with
169: the four-momentum $p_d$) is defined as
170: \begin{eqnarray}
171: \nonumber
172: d\Phi_{\mu+1}(P;p_d,k_1,...,k_\mu) &=& \delta^4(P-p_d-k)\,
173: \frac{d^3p_d}{(2\pi)^3 2E_d}
174: \prod_i \frac{d^3k_i}{(2\pi)^3 2\omega_i} \\
175: &=&(2\pi)^3dm^2d\Phi_{2}(P;p_d,k)\, d\Phi_{\mu}(k;k_1,...,k_\mu)
176: \label{dphi}
177: \end{eqnarray}
178: where $E_d$ and $\omega_i$ are the energies of the final deuteron and
179: the $i$-th decay particle, respectively, and $k = \sum k_i$.
180: The latter recursive formula
181: allows to study the propagation and the decay of the unstable particles
182: independently of the production mechanism itself.
183:
184: It is convenient to introduce what we would like to call
185: partial spectral functions $\rho_X$, which is given by
186: \begin{eqnarray}
187: \rho_X(m^2) :=
188: (2\pi )^3\int d\Phi_\mu(k;k_1,...,k_\mu)\,\, |G(m^2)\, W_X|^2 \ .
189: \label{rho}
190: \end{eqnarray}
191: Note that unitarity demands that
192: \be
193: {\sum_{X}} \, \rho_X (m^2)
194: = -\frac{1}{\pi}{\rm Im} G(m^2) \ , \ \ \int dm^2 {\sum_{X}} \, \rho_X (m^2)
195: = 1 \ ,
196: \label{rhoeig}
197: \ee
198: where the sum runs over all decay channels of the unstable particle
199: that are open at a certain given value of $m^2$.
200:
201: Using Eqs. (\ref{dsig}) -- (\ref{rho}) we can write
202: \be
203: \frac{d^2\sigma_X}{dm^2d\Omega_k}
204: =\frac{1}{64\pi^2}\left(\frac{|\vec k \, |}{|\vec p \, |}\right)
205: \frac{1}{s}|{\cal M}|^2\rho_X(m^2) \ ,
206: \label{csform}
207: \ee
208: where $\vec k$ is the relative momentum of the meson
209: system $X$ with respect to the deuteron. Naturally, $|\vec k \, |$ is
210: a function of $m^2$, namely
211: \be
212: %!!!m^2 and M_d^2 !!!
213: |\vec k \, |=\frac{\lambda^{1/2}(s,m^2,M_d^2)}{2\sqrt{s}},
214: \label{momenta}
215: \ee
216: where $M_d$ denotes the mass of the deuteron and
217: the function $\lambda$ is given by
218: \be
219: \lambda(x,y,z)=(x-y-z)^2-4yz.
220: \ee
221:
222: In this paper we neglect a possible final state interaction (FSI) of the
223: produced mesons with the deuteron (this issue was discussed for the reaction
224: $pp\to d a^+_0$ in Ref.~\cite{Oset}).
225: %
226: Then, close to threshold, the energy dependence of the matrix element $\cal M$
227: is determined by the lowest possible orbital angular momentum $l$ of the meson
228: system $X$ with respect to the deuteron. Specifically, it will
229: be given by $k^l$ with $k=|\vec k \, |$ being the corresponding relative
230: momentum.
231:
232: Let us now consider the case that interests us in the present study, namely
233: the decay of a scalar resonance or quasiparticle into two pseudoscalar
234: mesons. For any given interaction of the
235: scalar mesons with the pseudoscalars, the physical two point functions for the
236: scalar mesons can then be constructed by solving the Dyson equation,
237: where the self-energy $\Sigma$ is given by the leading order two pseudoscalar
238: loop diagrams. The equation is illustrated in Fig.~\ref{dyson}.
239: The strongest energy dependence of the loops is introduced by the
240: unitarity cut and its analytical continuation below the threshold.
241: The remaining piece can be assumed constant and can be absorbed in the
242: physical mass $M_R$. Under these assumptions the physical propagators of
243: the scalar mesons can be described by a Flatt\'e form \cite{flatte},
244: namely by
245: \begin{equation}
246: G(m^2) = \frac{1}{m^2-M_R^2 + iM_R {\sum\limits_{X}} \Gamma_X} \ ,
247: \label{prop}
248: \end{equation}
249: where
250: $$
251: \Gamma_X=\frac{|W_X|^2 | \vec k_X \, |}{8\pi M_R^2} \, ,
252: $$
253: with $\vec k_X$ being the CM momentum of the meson system $X$.
254: Below the threshold of the production of the final state $X$ the
255: analytic continuation of $\Gamma_X$ is to be used.
256:
257: % 3
258: \section{The mixing amplitude}
259:
260: We now turn to the case that is relevant for the present paper,
261: namely when there are two propagating unstable particles with
262: different isospins. Then the self-energy $\Sigma$ exhibits a matrix
263: structure in isospin space.
264: If we now allow the propagating particles to mix then the self-energy
265: matrix develops non-diagonal elements. Note, since the mixing particles
266: $a_0$ and $f_0$ are essentially mass degenerate
267: all the mixing should be completely dominated by its effect on the
268: meson propagation itself.
269: Along the lines of this reasoning we
270: completely neglect any isospin violation in the primary production amplitude
271: $\cal M$ and in the coupling of the pseudoscalar mesons to the resonances.
272: In short, we assume the resonances to be produced in pure isospin
273: states. The physical states, however, are those that propagate.
274: %For that purpose the physical propagator is to be diagonalized.
275: The physical states of $a_0$ and $f_0$ are consequently no longer pure
276: isospin states.
277:
278: We start from the inverse propagator in the isospin basis,
279: \begin{eqnarray}
280: {\mathbf{G^{-1}}} (m^2) = \left(\begin{array}{cc}
281: (m^2-m_1^2-\Sigma_{11}) & -\Sigma_{10} \\ -\Sigma_{01} & (m^2-m_0^2-\Sigma_{00} )
282: \end{array} \right)
283: =: \left( \begin{array}{cc}
284: g_{11} & g_{12} \\ g_{12} & g_{22}
285: \end{array} \right) \ ,
286: \nonumber
287: \end{eqnarray}
288: where $m_1$ and $m_0$ denote the bare masses in the isospin 1 and 0
289: channel, respectively.
290: The diagonal elements of the self-energy are given by
291: \begin{eqnarray}
292: \Sigma_{11} &=& \bar \Pi_{\bar K K}^{1}+\Pi_{\pi\eta} \\
293: \Sigma_{00} &=& \bar \Pi_{\bar K K}^{0}+\Pi_{\pi\pi}
294: \end{eqnarray}
295: and the isospin breaking is generated by
296: \be
297: \Sigma_{10} = \Sigma_{01} = \alpha +
298: \Pi_{K^+ K^-}-\Pi_{\bar K^0 K^0} \ .
299: \label{smix}
300: \ee
301: In these formulas $\Pi_{xy}$ denote loops integrals with particles
302: $x$ and $y$ propagating. The loops are assumed to be renormalized --
303: the renormalization constants are either absorbed into the physical
304: masses of the scalar mesons or into the complex quantity $\alpha$
305: (see also the discussion of this question in Ref.~\cite{Acha2}).
306: The loops denoted by $\bar \Pi^I_{\bar K K}$ are to be read as
307: containing contributions from charged kaons as well as neutral kaons,
308: coupled to the isospin $I$ as indicated in the superscript.
309: It is also implied that the physical masses are used in those
310: loop integrals. The mass difference between the charged and the
311: neutral kaons is the source for the isospin breaking generated by
312: the loops of Eq. (\ref{smix}). Without this difference the two kaon
313: loops would cancel each other.
314: The complex quantity $\alpha$ contains the effect induced by
315: $\pi$-$\eta$ mixing \cite{Tipp}, as for instance discussed in Ref.~\cite{Kud},
316: as well as a possibly existing direct $a_0$--$f_0$ transition.
317: Naturally, the imaginary part of $\alpha$, induced by the \pieta\
318: transition, can be easily estimated as was done in Ref.~\cite{Kud}.
319:
320: The propagation of the physical states is described by the eigenvalues
321: of $\mathbf{G}$, i. e.
322: \be
323: G_{a_0}=2/(g_{11}+g_{22}+r) \ \mbox{and} \ G_{f_0}=2/(g_{11}+g_{22}-r) \ ,
324: \label{physprop}
325: \ee
326: where $r = \sqrt{(g_{11}-g_{22})^2+4g_{12}^2}$.
327: The physical masses of the scalar mesons $a_0$ and $f_0$ are given
328: by the real part of the zeros of Eq.~(\ref{physprop}).
329:
330: As usual the full propagator $\mathbf{G}$ can be expressed in terms of
331: the physical propagators $G_{a_0}$ and $G_{f_0}$ plus a parameter for
332: the mixing, which can be identified as
333: $
334: \delta = 2g_{12}/(g_{11}-g_{22}) \ .
335: $
336: In the case of $\rho$--$\omega$ mixing or $\pi$--$\eta$ mixing
337: $\delta$ is guaranteed to be small due to a large denominator $g_{11}-g_{22}$.
338: In the former case there is a large difference in the widths whereas
339: in the latter the masses of the mixing particles are rather different.
340: However, the experimental evidence that is available for the
341: masses and widths of the scalar mesons $a_0$ and $f_0$ suggests that
342: here both these quantities might be very similar \cite{PDG}.
343: Thus, even $g_{11}=g_{22}$ is not excluded.
344: Also, we want to emphazise that
345: $\delta$ can not be interpreted as a mixing angle because
346: it is necessarily complex valued.
347:
348: For convenience we write the primary production amplitude
349: $\cal M$ as a vector in isospin space,
350: \begin{eqnarray}
351: {\mathbf{M}} =
352: \left( \begin{array}{c}
353: {\cal M}_1 \\ {\cal M}_0
354: \end{array} \right) \ ,
355: \nonumber
356: \end{eqnarray}
357: and the final production vertex $W_X$ as a matrix
358: \begin{eqnarray}
359: {\mathbf{W}} =
360: \left( \begin{array}{cc}
361: W_1^{(\pi \eta)} & W_1^{(\pi \pi)} \\ W_0^{(\pi \eta)} & W_0^{(\pi \pi)}
362: \end{array} \right) \ .
363: \nonumber
364: \end{eqnarray}
365: Note that here $W_1^{(\pi \eta)}$ is the vertex
366: for the coupling of the $\pi \eta$ system to an isospin-one particle as
367: given by an elementary Lagrangian, whereas
368: $W_0^{(\pi \eta)}$ is the vertex where the $\pi \eta$ system mixes
369: first into the $\pi \pi$ system which then couples to an isospin-zero
370: particle. We thus find
371: for the complete production amplitude of the final states $X$
372: = $(\eta \pi)$ or $(\pi\pi)$
373: (c.f. Eq. (\ref{adef}) for the one channel situation)
374:
375: \begin{eqnarray}
376: {\mathbf{A}} = {\mathbf{W}}^\dagger{\mathbf{G}}{\mathbf{M}}
377: = {\mathbf{W}}^\dagger \frac{1}{g_{11}g_{22}-g_{12}^2}
378: \left( \begin{array}{cc}
379: g_{22} & -g_{12} \\ -g_{12} & g_{11}
380: \end{array} \right) {\mathbf{M}} =
381: {\mathbf{W}}^\dagger G_{a_0} G_{f_0} \left( \begin{array}{cc}
382: g_{22} & -g_{12} \\ -g_{12} & g_{11}
383: \end{array} \right) {\mathbf{M}} \ .
384: \nonumber
385: \end{eqnarray}
386:
387:
388: Basically all investigations on the \mix so far don't use
389: the propagators of the physical states, i.e. they don't
390: use the eigenvalues of $\mathbf{G}$. Rather they assume that
391: the $a_0$ and $f_0$ mesons still propagate as proper isospin
392: states.
393: This leads to expressions that are very easy to interpret and
394: which are illustrated in Fig. 2 for the case of $X=(\pi \eta)$.
395: Within the formalism developed above this assumption
396: corresponds to the approximation
397: \begin{equation}
398: G_{f_0}G_{a_0}=\frac{1}{g_{11}g_{22}-g_{12}^2} \simeq
399: \frac{1}{g_{11}g_{22}} = \tilde G_{f_0} \tilde G_{a_0},
400: \label{gapprox}
401: \end{equation}
402: where $\tilde G_{f_0}$ and $\tilde G_{a_0}$ are now propagators of these
403: mesons in the isospin $0$ and $1$ states, respectively.
404: Indeed, the results presented in
405: Ref.~\cite{Kud} suggest that
406: $$
407: \frac{g_{12}^2}{g_{11}g_{22}}<10^{-2} \ ,
408: $$
409: i.e. that the approximation (\ref{gapprox}) is not unreasonable.
410:
411: Since the main focus of the paper is the $\pi \eta$ channel it is convenient
412: to introduce a mixing parameter $\xi$ by
413: \be
414: \xi(m) = \Sigma_{01}(m)\, \tilde G_{f_0}(m^2),
415: \label{52}
416: \ee
417: %=\frac{\Sigma_{01}(m)}{m^2-M^2_{f_0} +i M_{f_0} \Gamma_{f_0}(m^2)}\, ,
418: where $\Sigma_{01}$ was defined in Eq.~(\ref{smix}) and is related to the
419: \mix transition amplitude. With this definition we get for the
420: transition amplitude for the reaction $pn \to d\pi \eta$
421: \begin{equation}
422: {\cal A}_{(\pi \eta)} = W^{(\pi \eta)}_1\tilde G_{a_0}({\cal M}_1
423: +\xi (m){\cal M}_0)+ W^{(\pi \eta)}_0\tilde G_{f_0}{\cal M}_0
424: + W^{(\pi \eta)}_0\xi (m) \tilde G_{a_0} {\cal M}_1 \ .
425: \label{mmm}
426: \end{equation}
427: The last term involves mixing in $W_0^{(\pi\eta)}$ as well as in $\xi (m)$
428: and is therefore a second order correction.
429: Close inspection reveals that in the region $m\simeq 2 m_K$ the
430: third term in Eq.~(\ref{mmm}) should be suppressed by a factor
431: $$
432: \frac{ g_{f_0\pi\pi} |\lambda_{\pi^0\eta}| m_{a_0} \Gamma_{a_0}}
433: { g_{a_0\pi\eta} \Sigma_{01} (m^2_{\eta}-m^2_{\pi})} \approx 0.12
434: $$
435: compared to the second term. In this estimation
436: we used the values $\Sigma_{01}\approx 5000$~MeV$^2$ determined from
437: $K\bar K$ decay loops (see Fig.~2 in Ref.~\cite{Kud}) at
438: $m\simeq 2m_K$ and the $\pi\eta$ mixing amplitude
439: $\lambda_{\pi^0\eta}\simeq -5000$~MeV$^2$
440: (see Ref.~\cite{Kud} and references therein). We also use the
441: masses and widths $m_{a_0}=m_{f_0}=980$~MeV$/c^2$ and
442: $\Gamma_{a_0}=\Gamma_{f_0}=50$~MeV$/c^2$ and coupling constants $g_{a_0\pi\eta}$
443: and $g_{f_0\pi\pi}$ that are determined from the equations
444: $\Gamma_{a_0}= g^2_{a_0\pi\eta} q_{\pi\eta}/(8\pi m^2_{a_0})$ and
445: $\Gamma_{f_0}= 3 g^2_{f_0\pi\pi} q_{\pi\pi}/(16\pi m^2_{f_0})$,
446: where $q_{\pi\eta}$ and $q_{\pi\pi}$ are corresponding relative momenta.
447: Consequently, both terms -- the third and the last in Eq.~(\ref{mmm})
448: -- will be neglected in what follows.
449:
450: % 4
451: \section{Structure of the primary production amplitude}
452:
453: As we argued in the former section one can safely assume
454: the primary production amplitudes $\cal M_I$ as isospin conserving.
455: In this section we demonstrate how to construct the effective
456: interaction relevant for the transitions $pn\to d+(scalar \ meson)$ in
457: the close-to-threshold regime.
458:
459: As pointed out in Refs.~\cite{Gri,Kud}, if the scalar meson has isospin 1
460: it can only be produced in a $p$ wave. Since the initial $pn$ system
461: has to be in an isotriplet state and has to have odd parity, the Pauli
462: principle requires that it is also in a spin-triplet state.
463: Thus, the effective transition operator for the isovector final state
464: has to be linear in $\bfk$ and has to have an odd power of $\bfp$
465: (hereafter $\bfk$ and $\bfp$ denote the final and initial relative
466: three-momenta in the considered reaction, respectively).
467: It also has to be linear in both the spin
468: $\bfs :=\phi^T_1\sigma_2\bfsigma\phi_2$ of the initial nucleons pair
469: and the polarization vector $\bfeps$ of the outgoing deuteron. These
470: constraints lead to the reaction amplitude of the following type
471: \be
472: {\cal M}_1 = a\, (\bfp\cdot\bfs)\, (\bfk\cdot\bfeps^*)
473: + b\, (\bfp\cdot\bfk)\, (\bfs\cdot\bfeps^*)
474: + c\, (\bfk\cdot\bfs)\, (\bfp\cdot\bfeps^*)
475: + d\, (\bfp\cdot\bfs)\, (\bfp\cdot\bfeps^*)\, (\bfk\cdot\bfp)\, .
476: %\phantom{xx}
477: \label{8}
478: \ee
479: The coefficients $a$, $b$, $c$ and $d$ are independent scalar amplitudes.
480: They may depend on the total CM energy and are necessarily complex since
481: they contain the initial state interaction. For the isospin 1 amplitude
482: relevant here the phase induced by the two nucleon unitarity cuts
483: can be related to the $NN$ phase shifts, as pointed out in
484: Ref.~\cite{kanzoandi}. However, for the isoscalar initial state
485: (required for the production of the $f_0$ meson, cf. below) there is no
486: phase-shift analysis available at the relevant energies due to a lack of
487: high energy $pn$ scattering experiments \cite{Arndt}. Therefore, in this
488: paper we will not consider the effects induced by the initial state
489: interaction. Let us mention, however, that their effects do not influence
490: the qualitative aspects discussed here. Note, that
491: the scalar amplitudes $a$, $b$, $c$ and $d$ are expected to have quite
492: smooth energy dependence and, accordingly, can be considered as constant
493: in the near threshold region.
494:
495: In Ref. \cite{Oset} the transition amplitude was given in a partial wave
496: decomposed form. Obviously the two descriptions are equivalent, however,
497: we regard Eq. (\ref{8}) as more convenient for the construction of polarization
498: observables.
499:
500: The amplitude relevant for the near-threshold production of an
501: isoscalar scalar meson can be constructed along the same lines and is given by
502: \be
503: {\cal M}_0 = f\, (\bfs\cdot\bfeps^*)
504: + g\, (\bfp\cdot\bfs)\, (\bfp\cdot\bfeps^*)\, .
505: \label{81}\ee
506: Similar expressions for the amplitudes~(\ref{8}) and~(\ref{81})
507: can be found in Ref.~\cite{Grish,Kondr}. Note, however, that the terms
508: proportional to $g$ and $d$ were omitted in that work.
509:
510: It is obvious from Eqs.~(\ref{8}) and (\ref{81}) that
511: ${\cal M}_1 \propto k$ and ${\cal M}_0 \propto const.$
512: near the threshold. For the isospin conserving case we therefore get
513: $$
514: \frac{d\sigma}{dm^2} (pn\to da_0) \propto Q^{3/2} \ \mbox{and}
515: \ \frac{d\sigma}{dm^2} (pn\to df_0) \propto Q^{1/2} \ ,
516: $$
517: where $Q(m) = \sqrt{s}-M_d-m$ is the excess energy. Thus, if we study the
518: production of an isospin 1 final state, namely $\pi^0\eta$, the mixing
519: with the isoscalar ($f_0$) state will be kinematically enhanced \cite{Kud}.
520:
521: % 5
522: \section{Study of observables}
523:
524: \subsection{Invariant mass spectrum}
525: Naturally the simplest observable is just the invariant mass
526: spectrum of the reaction $pn\to dX$. As long as we look at
527: energies/values of $m^2$ such that $Q(m^2)$ is small, the spectrum
528: should show a resonant peak $\sim {\sum\limits_{X}}\rho_X$
529: reflecting the $m^2$ dependence of the $f_0$ propagator,
530: cf. Eq. (\ref{rhoeig}).
531: It is possible to deduce the branching ratios of a particular resonance
532: by a fit of the Flatt\'e distribution to the mass spectrum. This was
533: demonstrated, e.g., for the case of the $a_0^+$ in Ref.~\cite{BNL}.
534: Since so far little is known about the $f_0$ this very easy to measure
535: observable is very interesting.
536: In addition, the number of all events in this peak is proportional to
537: %$k |{\cal M}_0|^2\, |G_{f_0}(m^2)|^2 \,m\Gamma_{f_0}(m^2)$
538: %(see Eqs.~(\ref{csform}), (\ref{prop}) and (\ref{81})), where
539: $$
540: |{\cal M}_0|^2= 3 |f|^2 + 2\,{\rm Re}(fg^*) p^2 + |g|^2 p^4\, ,
541: $$
542: c.f. Eqs. (\ref{rhoeig}), (\ref{csform}) and (\ref{81}).
543: %
544: Note that the knowledge of the amplitudes $f$ and $g$ is important,
545: if we want to deduce quantitative informations about the $f_0/a_0$ mixing
546: amplitude from the reaction $pn\to dX$.
547:
548: Definitely the most interesting observables are those concerning
549: the channel $pn\to d\pi \eta$,
550: since here the isospin conserving amplitude enters
551: in a $p$ wave whereas the isoscalar amplitude, that can be mixed in via
552: isospin violation, enters in an $s$ wave. This is the kinematical
553: enhancement mentioned at the end of the last section.
554:
555: Let us discuss the mass distribution $d\sigma/dm_{\pi^0\eta}$ for the
556: $\pi^0\eta$ system in the reaction $pn\to d\pi^0\eta$. Note that near
557: the threshold of the reaction $pn\to d a^0_0$ this mass distribution
558: should be sensitive to the magnitude of \mix mixing as well as to
559: the mixing mechanism.
560:
561: Consider first the limiting case of isospin conservation, so that the
562: \mix mixing is absent. In this case the mass spectrum
563: of the $\pi^0\eta$
564: system should coincide with the spectrum of the $\pi^+\eta$ system from
565: the reaction $pp\to d a^+_0\,$ -- apart from an overall (isospin)
566: factor of 0.5 \cite{Grish}. The amplitude of the reaction
567: $pn\to d\pi^0\eta$ is then given by the diagram of Fig.~2a.
568: In this case the invariant
569: $\pi^0\eta$-mass distribution is determined by the phase space
570: and $\rho_{(\pi \eta)}$, which was defined in Eq. (\ref{rho}), and is
571: dominated by the $a_0$ propagator (c.f.
572: Eq.~(\ref{prop})). The $\pi^0\eta$-mass distribution obtained under these
573: asumptions, is shown in Fig.~3 by the dotted line. The calculations
574: are done at $T_p=2645$~MeV$/c$ where experimental results can be
575: expected from the ANKE collaboration soon \cite{Kondr}.
576: The parameters used here are: $M_{a_0}=980$ MeV$/c^2$,
577: $\Gamma_{a_0}=50$ MeV$/c^2$. The mass-dependent width is described by a
578: Flatt\'e-like form $\Gamma_{a_0}(m)=\Gamma_{a_0} +\Gamma_{K\bar K}(m)$,
579: which takes into account the $K\bar K$ decay channel
580: (see Eq.~(8) in Ref.\cite{Kud}).
581: As we discussed in the previous
582: Sections, the isospin conserving production of the $a^0_0$ meson
583: can take place only in a $p$ wave and hence is suppressed near threshold.
584:
585: If isospin is not conserved and \mix mixing is present, then the
586: $\pi \eta$ system can be
587: produced also in an $s$ wave through the $f_0$. The
588: corresponding diagram is shown in Fig.~2b.
589: For this mechanism the $\pi^0\eta$ mass
590: spectrum is determined not only by the $a^0_0$ propagator,
591: the partial wave of production and the phase space but also
592: by the $f_0$ propagator and by the mass dependence of the nondiagonal
593: self-energy matrix element $\Sigma_{01} (m)$
594: (cf. Eqs.~(\ref{smix}) and (\ref{52})):
595: \begin{eqnarray}
596: \nonumber
597: \xi(m) = \Sigma_{01}(m)\, \tilde G_{f_0}(m^2) \approx
598: \frac{\Sigma_{01}(m)}{m^2-M^2_{f_0} +i M_{f_0} \Gamma_{f_0}(m^2)}\, .
599: \end{eqnarray}
600: For illustrative purposes we will use
601: $M_{f_0}=980$ MeV$/c^2$ and the same Flatt\'e-like width
602: that was used for the $a_0$ meson above, i.e.
603: $\Gamma_{f_0}(m^2)=\Gamma_{a_0}(m^2)$.
604:
605: Of course, now the invariant mass plot becomes sensitive to the specific
606: mixing mechanism. For example, if we assume that
607: the leading contribution to $\Sigma_{01} (m)$ comes from the direct
608: \mix transition mechanism, i. e. the term $\alpha$ in Eq.~(\ref{smix}),
609: then there should be no drastic dependence on the invariant mass
610: $m$ of the $\pi^0\eta$ system.
611: The $\pi^0\eta$-mass spectrum, corresponding to the mechanism of Fig.~2b
612: and with $\Sigma_{01} (m)=const$, is shown in Fig.~3 by the dashed line.
613: (All the distributions are normalized to 1 at the maximal values!)
614: On the other hand,
615: if $\Sigma_{01} (m)$ is determined mainly by the $K\bar K$ loop (the
616: term $\Pi_{K^+ K^-}$$\,-\,$$\Pi_{\bar K^0 K^0}$ in Eq.~(\ref{smix})),
617: it should be strongly enhanced near the $K\bar K$ threshold (see, e.g.,
618: Refs.~\cite{Acha,Krepl,Kud}).
619: The corresponding $\pi^0\eta$-mass distribution
620: is shown in Fig.~3 by the solid line.
621:
622: Thus, we conclude that the $\pi^0\eta$ mass spectrum is rather sensitive
623: to the \mix mixing mechanism. That is why the study of
624: $d\sigma/dm_{\pi^0\eta}$ in the reaction $pn\to d\pi^0\eta$ should
625: allow to shed light on the nature of the lightest scalar mesons
626: $a_0(980)$ and $f_0(980)$.
627:
628: % 6
629: \subsection{Analyzing power for the process $pn\to d\pi\eta$}
630: \subsubsection{Near threshold}
631:
632: In our approximation the total amplitude $\cal M$ of the resonant
633: $\pi \eta$--production
634: with \mix mixing effects taken into account may be written as
635: (c.f. Eq. (\ref{mmm}))
636: \be
637: {\cal M} = {\cal M}_1 + \xi {\cal M}_0\, ,
638: \label{51}
639: \ee
640: where the amplitudes ${\cal M}_1$ and ${\cal M}_0$ are given by the
641: Eqs.~(\ref{8}) and (\ref{81}) and the mixing parameter $\xi$ was
642: defined in Eq. (\ref{52}).
643:
644: Let us introduce the polarization of one of the nucleons $\bfzeta$
645: through $\bfzeta=\phi^+_1\bfsigma\phi_1\,$ or
646: $\,\phi_1\phi^+_1 =(1+\bfzeta\cdot\bfsigma)/2$.
647: The matrix element~(\ref{51}) squared and averaged (summed)
648: over the polarizations of the initial neutron (final deuteron) is then
649: given by
650: $$
651: \overline {|{\cal M}|^2}=\frac{1}{2}
652: \left[\left( |a|^2 +|c|^2\right)p^2 k^2 + 3 |B|^2 + |D|^2 p^4 \right]
653: +(\bfp\cdot\bfk)\, {\rm Re}H
654: $$
655: \be
656: + p^2\, {\rm Re}\left[a^* D\, (\bfp\cdot\bfk) + B^* D\right]
657: + (\bfzeta\cdot[\bfk\times\bfp\,])\, {\rm Im}H \, ,
658: \label{54}
659: \ee
660: where
661: $$
662: B=b\, (\bfp\cdot\bfk) +f\,\xi\, ,
663: \phantom{xx}
664: D=d\, (\bfp\cdot\bfk) +g\,\xi\, ,
665: \phantom{xx}
666: H= a^* B + a^* c\, (\bfp\cdot\bfk) + B^* c +D^* c\, p^2\, .
667: $$
668: Let the polarization vector $\bfzeta$ be directed along the $x$-axis
669: ($\bfzeta\perp\bfp$). Then, in terms of the angles $\theta$ and $\varphi$,
670: we have $\bfp\cdot\bfk=pk\cos\theta\,$ and
671: $\bfzeta\cdot[\bfk\times\bfp\,]=\zeta\, p k \sin\theta\,\sin\varphi\,$
672: ($\zeta=|\bfzeta|$). Using those expressions Eq.~(\ref{54}) may be
673: written as
674: %
675: \vspace{1mm}
676: \be
677: \overline {|{\cal M}(\theta,\varphi)|^2}=
678: C_0 + C_1 \cos\theta +C_2 \cos^2\theta
679: + \,\zeta\, \sin\theta\,\sin\varphi\, (D_0 + D_1\cos\theta)\, ,
680: \label{55}\ee
681: where
682: $$
683: C_0 = \frac{1}{2} \left( |a|^2 +|c|^2\right) p^2 k^2
684: + \left( |f|^2 + \frac{1}{2} |f+p^2 g|^2 \right)\, |\xi|^2\, ,
685: $$ $$
686: C_1 = pk\, {\rm Re} \left(\left[ (a+b+c+p^2 d)^* (f+p^2 g)
687: + 2b^* f \right]\,\xi \right)\, ,
688: $$
689: \be
690: C_2 = p^2 k^2\, \left[ |b|^2 +\frac{1}{2} |b+p^2 d|^2
691: + {\rm Re} \left( a^* c + (a+c)^* (b+p^2 d) \right) \right]\, ,
692: \label{56}\ee
693: $$
694: D_0 =p k\, {\rm Im}\left( \left[ a^* f -c^* (f+p^2 g)\right]
695: \xi \right)\, ,
696: $$ $$
697: D_1 =p^2 k^2\, {\rm Im} (a^* b + a^* c + b^* c + p^2 d^{\,*} c)\, .
698: $$
699: \vspace{1mm}
700: We can now generalize the definition of the asymmetry $A$
701: given in Ref.~\cite{Kud} to the polarized situation.
702: Using the short hand notation
703: \be
704: \sigma (m;\theta, \varphi) \equiv \frac{d^2\sigma}
705: {dm^2d\Omega } (\theta, \varphi)\, .
706: %\phantom{xx}
707: \label{57}\ee
708: we now define the angular--asymmetry parameter $A$ through
709: \begin{eqnarray}
710: \nonumber
711: A(m;\theta,\varphi)&=&
712: \frac{\sigma (m;\theta, \varphi) -\sigma (m;\pi-\theta, \varphi+\pi)}
713: {\sigma (m;\theta, \varphi) +\sigma (m;\pi-\theta, \varphi+\pi)}\, \\
714: &=&\frac{C_1 \cos\theta\,
715: +\,\zeta\, D_0 \sin\theta\,\sin\varphi }
716: {C_0 + C_2 \cos^2\theta
717: + \,\zeta\, D_1 \cos\theta\,\sin\theta\,\sin\varphi}\, .
718: \label{58}
719: \end{eqnarray}
720: It follows from Eqs.~(\ref{56}) and~(\ref{58}) that $C_1=D_0=0$ for $\xi=0$
721: and thus
722: \be
723: A(m;\theta, \varphi)_{\xi=0}=0\,
724: \label{591}
725: \ee
726: when isospin is conserved.
727:
728: Let us discuss some specific features of the expression~(\ref{58}) in
729: the following. First note that there are two different terms.
730: The first term does not depend on the
731: polarization $\zeta$ of the proton beam and on the angle $\varphi$.
732: Therefore, Eq.~(\ref{58}) implies that
733: \be
734: A(m;\theta=0^0, \varphi)=\frac{C_1}{C_0+C_2} \ .
735: \label{59}
736: \ee
737: This particular result for the
738: asymmetry $A$ was derived earlier in Ref.~\cite{Kud}
739: but not in terms of the amplitudes $a$, $b$, $c$, $d$, $f$ and $g$.
740: The discussion in Refs.~\cite{Grish,Kondr}, on the other hand,
741: takes into account only the amplitudes $a$, $b$, $c$, and $f$,
742: but not the amplitudes $d$ and $g$.
743:
744: At $\theta=90^0$ and for $\zeta\ne 0$ we get
745: \be
746: A(m;\theta=90^0, \varphi)=\zeta\,\frac{D_0}{C_0}\,\sin\varphi\, =
747: \frac{2\zeta\,p k\, {\rm Im}\left( \left[ a^* f -c^* (f+p^2 g)\right]\,
748: \xi \right) \sin\varphi} {\left( |a|^2 +|c|^2\right) p^2 k^2
749: + \left( 2 |f|^2 + |f+p^2 g|^2 \right)\, |\xi|^2}\, .
750: \label{510}
751: \ee
752:
753: Obviously
754: the isospin-breaking effect for different angles $\theta$ and
755: $\varphi$ depends on different combinations of the basic amplitudes.
756: Thus, experimental information on the asymmetry obtained with polarized
757: beam could allow to deduce additional constraints on the amplitudes $a$,
758: $b$, $c$, $d$, $f$, and $g$ of the $a_0$- and $f_0$ production reactions.
759: With regard to that let us mention that
760: the model of $a_0$-meson production discussed in
761: Ref.~\cite{Kud}, which was based on the impulse approximation, leads to
762: $A(m;\theta=90^0, \varphi)\equiv 0$ -- because in that model $a = c = 0$.
763:
764: Throughout this paper we make the assumption that isospin breaking
765: takes place only in the propagation of the scalar mesons. This implies
766: that the mixing strength is the same for all partial waves.
767: On the other hand, if a significant mixing takes place also in the
768: initial $NN$ interaction or in the primary production amplitude
769: ${\cal M}_1$, there is no reason to expect this mixing as being
770: partial-wave independent. Thus the study of polarization observables allows
771: to examine this basic model assumption in a clean way.
772:
773: In addition it will be very interesting
774: to study the $m$ dependence of $A(m;\theta,\varphi)$ in detail.
775: By construction $A$ projects on the isospin breaking pieces of
776: the amplitude. Therefore, the $m$ dependence of $A$ gives direct access
777: to the $m$ dependence of $\Sigma_{01}(m)$ and thus to the mixing
778: mechanism.
779:
780: % 7
781: \subsubsection{Higher energies}
782:
783: As mentioned above the asymmetry $A(m;\theta, \varphi)$ (Eq.~(\ref{58}))
784: will become zero if \mix mixing is not included, i.e. for $\xi=0$. This
785: result is in line with the observation made in Ref.~\cite{Miller}
786: that the asymmetry $A(\theta)$ is identical to zero in the reaction
787: $pn\to d\pi^0$ for the unpolarized case.
788: Let us emphasize, however, that the vanishing of the asymmetry
789: $A(m;\theta, \varphi)$ in the limit $\xi=0$ for the polarized case is true
790: only near threshold, where the $a_0$ meson is produced in a $p$ wave and
791: the $f_0$ meson in an $s$ wave, see Eqs.~(\ref{8}) and (\ref{81}).
792: The expressions~(\ref{8}) and (\ref{81}) do not take into account
793: contributions from higher partial waves,
794: specifically they do not include, e.g., the $d$ wave production of the
795: $a_0$ meson and the $p$ wave production of the $f_0$ meson.
796: In the following we want to consider this question in more detail,
797: assuming that isospin is conserved.
798:
799: The amplitudes for $a_0$- and $f_0$ production in the
800: reaction~(\ref{1}) can be written in the most general form as
801: \be
802: {\cal M}_I =\phi^T_1\sigma_2\, (F_I +\bfg_I\cdot\bfsigma)\,\phi_2 \, ,
803: \label{511}
804: \ee
805: where $\phi_{1,2}$ are the spinors of the nucleons. The two terms
806: $F_I$ and $\bfg_I$ correspond to $S_{NN}=0$ and $S_{NN}=1\,$,
807: respectively, where $S_{NN}$ is the total spin of the initial $NN$
808: system. The subscript $I$ denotes the isospin of the produced scalar
809: meson. The scalar and vector
810: functions $F_I$ and $\bfg_I$ depend on the vectors $\bfp$, $\bfk$ and
811: $\bfeps$. Let $L_{NN}$ and $L_{Md}$ be the angular momenta in the
812: initial $NN$ and in the final meson$+d$ systems, respectively. It follows
813: from the conservation of quantum numbers together with the required
814: antisymmetry of the system with respect to the initial nucleons that
815: the $F_1$ ($\bfg_1$) term in Eq.~(\ref{511}) should contain only the
816: contributions from even (odd) values of $L_{NN}$ and $L_{Md}$.
817: One can also see that the $F_0$ ($\bfg_0$) term should only contain the
818: contributions from odd (even) values of $L_{NN}$ and $L_{Md}$.
819: In the isovector ($I=1$) case the $s$ wave ($L_{NN}=L_{Md}=0$) is
820: forbidden, i.e. the contributions to the $F_1$ term start with a
821: $d$ wave.
822: For the case where one of the nucleons is polarized, the expression
823: for the squared matrix element~(\ref{511}) is given by the form
824: \be
825: \overline {|{\cal M}_I|^2}=\frac{1}{2} \left[ |F_I|^2
826: + \left( [ F_I \bfg^*_I + F^*_I \bfg_I ]\cdot\bfzeta\right)
827: + (\bfg^*_I \cdot \bfg_I)
828: + i \left(\bfzeta\cdot [\bfg_I \times \bfg^*_I ]\right)\right]\, ,
829: \label{512}
830: \ee
831: where $\bfzeta$ is the proton polarization vector.
832: The second term in the
833: expression~(\ref{512}) corresponds to the interference between the
834: $F_I$ and $\bfg_I$ amplitudes. Thus, this term is an
835: odd function of the final momentum $\bfk$.
836: All other terms in Eq.~(\ref{512}) are even functions of
837: $\bfk$. We see that the second term is responsible for the angular
838: asymmetry in the reaction~(\ref{1}) even if isospin is conserved.
839: It vanishes in the case of
840: unpolarized protons, i.e. $A(m;\theta, \varphi)\equiv 0$ for $\bfzeta=0$.
841:
842: Note that the amplitudes~(\ref{8}) and (\ref{81}) are special cases of
843: the general form Eq.~(\ref{511}) with $F_I\equiv 0$.
844: Also, setting $F_I\equiv 0$ in
845: Eq.~(\ref{512}) we always get $A(m;\theta,\varphi)\equiv 0$ if isospin is
846: conserved. The amplitudes~(\ref{8}) and (\ref{81}) correspond
847: to the $\bfg_I$ terms in Eq.~(\ref{512}) in the lowest-order
848: approximation with respect to the final momentum $\bfk$, i.e. keeping
849: only contributions that are at most linear in $\bfk$.
850: The next-order term with respect to $\bfk$ in the amplitude for
851: $a_0$-meson production is the $d$-wave contribution. This term (of the
852: order $\sim k^2$) contributes to $F_1$ in Eq.~(\ref{511}) and
853: modifies the amplitude ${\cal M}_1$ given in Eq. (\ref{8}), i.e.
854: \be
855: {\cal M}_1 \to {\cal M}_1 + F_1 \,\, \phi^T_1\sigma_2\,\phi_2 \, ,
856: \phantom{xxx}
857: F_1 =e\, (\bfeps^*\cdot [\bfp\times\bfk]) (\bfp\cdot\bfk)
858: \label{513}
859: \ee
860: where $e$ is a scalar amplitude. It is clear that the asymmetry
861: $A(m;\theta, \varphi)$~(\ref{58}), calculated with the
862: amplitude~(\ref{513}), will get non-zero contributions $\sim k^3$ for
863: small $k$ due to the interference between $p$ and $d$ waves.
864: It follows from Eq.~(\ref{513}) that $F_1=0$ at $\sin\theta=0\,$ or
865: $\cos\theta=0\,$.
866: In general, the $F_1$ term contains all contributions with even values
867: $L_{NN}, L_{Md} =2,4,6,...\,$. Thus, each contribution should include
868: the factor $\bfeps^*\cdot [\bfp\times\bfk]$
869: as well as the factor $\bfp\cdot\bfk$. Therefore, even in the
870: general case $F_1=0$ if $\sin\theta=0\,$ or $\cos\theta=0\,$.
871: Accordingly, for the case of conserved isospin we get
872: \be
873: A(m;\theta=0,\varphi)\equiv 0
874: \label{5131}
875: \ee
876: at any $k$ -- and not only near threshold (compare this result with
877: that from Eq.~(\ref{591})). This result (\ref{5131}) confirms a general
878: statement made in Ref.~\cite{Miller} on the absence of a forward--backward
879: asymmetry for the reaction $pn\to d X^0$ in the limit of isospin
880: conservation.
881:
882: It is remarkable that in addition to the condition~(\ref{591}) for the
883: case of polarized proton beam we get also the nontrivial result that
884: \be
885: A(m;\theta=90^0,\varphi)\equiv 0
886: \label{514}
887: \ee
888: if isospin is conserved.
889: Therefore, any deviation of $A(m;\theta=90^0,\varphi)$ from zero
890: is a direct indication for \mix mixing. Note that
891: this statement does not depend on the number of partial waves
892: taken into account, i.e. it is valid for any excess energy and not
893: only near threshold. Consequently, a
894: measurement of the asymmetry at $\theta=90^0$
895: should provide us with evidence on the \mix mixing amplitude.
896: We want to remark that this information is to be considered as
897: complementary to the one that can be obtained with unpolarized beam.
898: %
899: It should be mentioned that the vanishing of the asymmetry at
900: $\theta=90^0$ for the reaction $pn\to d a^0_0\,$ does not follow from
901: the theorem formulated in Ref.~\cite{Miller} and is new.
902:
903: Finally, let us discuss the situation at $\theta\ne 0^0$ and
904: $\theta\ne 90^0$.
905: If the amplitude for $a^0_0$-meson production is taken
906: of the form~(\ref{51}), i.e. isospin-breaking effects are included,
907: then the leading-order contribution to the asymmetry
908: $A(m;\theta,\varphi)$ are of the order $\sim \xi k$
909: (see Eqs.~(\ref{56}) and (\ref{58})). The
910: next-to-leading terms in the amplitude ${\cal M}_1$
911: (see Eq.~(\ref{513})) give rise to contributions of the order $\sim k^3$
912: to the asymmetry from isospin-conserving terms. Thus, when studying
913: isospin-breaking effects over a wide region of $\theta$, one should
914: consider $a^0_0$-meson production in a rather narrow region of
915: relative momenta $k$, i.e. at small excess energies,
916: in order to suppress contributions to $A(m;\theta,\varphi)$ from
917: isospin-conserving amplitudes with higher angular momenta.
918:
919: % 9
920: \section{Summary}
921:
922: Let us summarize the main results of this paper.
923: The most sensitive observables for examining the \mix mixing amplitude are:
924:
925: \begin{itemize}
926: \item[{\it i})]
927: the angular asymmetry $A(m;\theta, \varphi)$, as defined by Eq.~(\ref{58})
928: of this paper
929: \item[{\it ii})]
930: the distribution of the effective mass of the $\pi^0\eta$ system near the
931: $a^0_0$ threshold
932: \end{itemize}
933: %
934: With regard to the
935: angular asymmetry $A(m;\theta, \varphi)$ for the channel
936: $pn\to d\pi^0\eta$, it is expected to be rather large, i.e. in the order of
937: $\sim 10^{-1}$. For the case of an unpolarized beam the forward-backward
938: asymmetry was estimated earlier in Ref.~\cite{Kud}. The study of the
939: asymmetry $A(m;\theta, \varphi)$ with perpendicular polarized proton
940: beam should shed additional light on the mechanism of $a_0$ production
941: in the reaction $pn\to d a^0_0$. As we have shown the various basic
942: amplitudes for this reaction give different contributions to
943: $A(m;\theta, \varphi)$ at different angles.
944:
945: Since by constuction $A$ projects on the isospin breaking pieces of
946: the amplitude, the $m$ dependence of $A$ gives direct access
947: to the $m$ dependence of $\Sigma_{01}(m)$ and thus to the mixing
948: mechanism.
949:
950:
951: Note, in experiments with polarized beams
952: isospin-breaking effects give contributions $\sim \xi k$ to
953: the asymmetry $A(m;\theta, \varphi)$ in leading-order
954: with respect to the final momentum $\bfk$. Isospin-conserving amplitudes
955: give contributions $\sim k^3$ due to $d$-wave terms. Thus when
956: studying isospin-breaking effects over a wider range of $\theta$ and
957: $\varphi$, one should consider $a^0_0$-meson production
958: at small momenta $k$, or small excess energies. Accordingly,
959: when studying the reaction $pn\to d\pi^0\eta$ near the
960: threshold of $a_0$ production, one should limit the invariant mass $m$
961: of the $\pi^0\eta$ system in a narrow region near threshold.
962: Note that the contribution of the $d$-wave to the reaction $pn\to d\,a_0$
963: may be monitored by a simultaneous measurement of the differential
964: cross section.
965:
966: If isospin is conserved, then asymmetry effects should be absent
967: at $\theta=0^0$ or $\theta=90^0$, even in the polarized case.
968: The angle $\theta=0^0$ is not conclusive
969: for the reaction with polarized beam, since all polarization
970: effects are proportional to $\zeta\sin\theta\cos\varphi$ and vanish at
971: $\sin\theta=0$. Thus, the asymmetry for $\theta=0^0$ coincides with
972: the one for unpolarized nucleons.
973: The case $\theta=90^0$ is much more interesting
974: when studying the isospin-breaking effects in the reaction with
975: polarized protons, since polarization effects are maximal at
976: this angle. Anyway, in either case ($\theta=0$ and $\theta=90^0$) a
977: non-vanishing
978: asymmetry $A(m;\theta, \varphi)$ can come only from isospin-breaking
979: effects.
980:
981: Very important informations on \mix mixing can be also expected from
982: a study of the mass distribution $d\sigma/dm_{\pi^0\eta}$ for the
983: $\pi^0\eta$ system in the reaction $pn\to d\pi^0\eta$. If the
984: isospin is conserved then
985: the mass spectrum of the $\pi^0\eta$ system from this reaction
986: should coincide with the one for the $\pi^+\eta$ system from the reaction
987: $pp\to d a^+_0\,$. Differences in the mass distributions of the
988: $\pi^0\eta$ and the $\pi^+\eta$ systems
989: are a clear signal that the isospin is not conserved and that
990: \mix mixing is present. The mass distribution $d\sigma/dm_{\pi^0\eta}$
991: for the $\pi^0\eta$ system is also extremely sensitive to the mechanism
992: of \mix mixing. For example, in the case where \mix mixing is
993: dominated by the $K\bar K$ loop, the $\pi^0\eta$ mass
994: spectrum will be strongly enhanced near the $K\bar K$ threshold.
995:
996: Thus, we conclude that the study of the reaction $pn\to d a^0_0$ is
997: extremely useful to understand the mechanism and the magnitude of the
998: \mix mixing amplitude. The knowledge of this mixing amplitude should
999: allow us to
1000: shed light on the nature and the quark content of the lightest
1001: scalar mesons $a_0(980)$ and $f_0(980)$.
1002:
1003: Several interesting observables where not discussed in this
1004: paper and are to be studied in a subsequent work.
1005: One example is the differential cross section
1006: for $pn\to dK^+ K^-$.
1007: Since this final state does not have definite isospin the differential
1008: cross section develops a forward--backward asymmetry at larger
1009: excess energies $Q$.
1010: Models predict that the kaon loops are an important source for
1011: the mixing \cite{Krepl}. Therefore, the measurement of
1012: a forward--backward asymmetry for the process $pn\to dK\bar K$ (the sum
1013: of $pn\to dK^+ K^-$ and $pn\to dK^0 \bar K^0$) should help to pin down the
1014: \mix mixing amplitude. Corresponding measurements of these
1015: decay channels should be feasible at the ANKE and TOF facilities,
1016: respectively, at the COSY accelerator in J\"ulich.
1017:
1018: Finally let us consider the reaction
1019: $dd\to {^4}{\rm He}\,\pi^0\eta$. Evidently, here isospin is not
1020: conserved. Near the $a_0$ threshold this reaction should
1021: proceed only through the chain
1022: $dd\to {^4}{\rm He}\,f_0\to {^4}{\rm He}\,a^0_0\to {^4}{\rm He}\,\pi^0\eta$.
1023: Thus, the $\pi^0\eta$ mass spectrum of this reaction should
1024: be even more sensitive to the \mix mixing mechanism, than the one
1025: for the reaction $pn\to d a^0_0$ \cite{Grish}.
1026:
1027: \vfill
1028: \section*{Acknowledgments}
1029: The authors are thankful to L.A.~Kondratyuk and M.~B\"uscher
1030: for stimulating discussions.
1031: This work was partly supported by the DFG-RFBI grant No. 02-02-04001
1032: (436 RUS 113/652/1-1) and by the RFBR grant No. 00-15-96562.
1033:
1034:
1035: %\newpage
1036:
1037: \begin{thebibliography}{99}
1038: %
1039: \bibitem{Acha} N.N.~Achasov, S.A.~Devyanin, and G.N.~Shestakov,
1040: Phys. Lett. {\bf B88}, 367 (1979); Yad.Fiz. {\bf 33},
1041: 1337 (1981) [Sov. J. Nucl. Phys. {\bf 33}, 715 (1981)].
1042: %
1043: \bibitem{Barnes} T. Barnes, Phys. Lett. {\bf B165}, 434 (1985).
1044:
1045: \bibitem{Acha1} N.N.~Achasov and G.N.~Shestakov,
1046: Phys. Rev. D {\bf 56}, 212 (1997).
1047:
1048: %
1049: \bibitem{Krepl} O. Krehl, R. Rapp, and J. Speth,
1050: Phys. Lett. {\bf B390}, 23 (1997).
1051: %
1052: \bibitem{Kerb} B.~Kerbikov and F.~Tabakin,
1053: Phys. Rev. C {\bf 62}, 064601 (2000).
1054: %
1055: \bibitem{Kud} A.E.~Kudryavtsev and V.E.~Tarasov,
1056: JETP Lett. {\bf 72}, 401 (2000); {\tt nucl-th/0102053}.
1057: %
1058: \bibitem{Grish} V.Yu.~Grishina, L.A.~Kondratyuk, M.~B\"uscher, W.~Cassing,
1059: and H. Str\"oher, Phys. Lett. {\bf B 521},
1060: 217 (2001).
1061: %
1062: \bibitem{Close} F.E.~Close and A.~Kirk,
1063: Phys. Lett. {\bf B489}, 24 (2000).
1064: %
1065: \bibitem{Barb} D.~Barberis, et al.,
1066: Phys. Lett. {\bf B488}, 225 (2000).
1067: %
1068: \bibitem{Kirk} A.~Kirk, Phys. Lett. {\bf B489}, 29 (2000).
1069: %
1070: \bibitem{Jans} G. Janssen, B. Pierce, K. Holinde, and J. Speth,
1071: Phys. Rev. D {\bf 52}, 2690 (1995).
1072: %
1073: \bibitem{Oller97} J.A. Oller and E. Oset,
1074: Nucl. Phys. {\bf A620}, 438 (1997);
1075: Nucl. Phys. {\bf A652}, 407 (1999).
1076: %
1077: \bibitem{weinberg} S.~Weinberg,
1078: Phys. Rev. {\bf 130}, 776 (1963);
1079: Phys. Rev. {\bf 131}, 400 (1963).
1080: %
1081: \bibitem{Oset} E.~Oset, J.A.~Oller, and Ulf-G.~Mei{\ss}ner,
1082: Eur. Phys. J. {\bf A12}, 435 (2001).
1083: %
1084: \bibitem{flatte} S.M.~Flatt\'e, Phys. Lett. {\bf 63}, 224 (1976).
1085: %
1086: \bibitem{Acha2} N.N.~Achasov, S.A.~Devyanin, and G.N.~Shestakov,
1087: Yad. Fiz. {\bf 32}, 1098 (1980)
1088: [Sov. J. Nucl. Phys. {\bf 32}, 566 (1980)].
1089: %
1090: \bibitem{Tipp} W.B. Tippens et al.,
1091: Phys. Rev. C {\bf 63}, 052001 (2001).
1092: %
1093: \bibitem{PDG} Particle Data Group: D.E. Groom et al.,
1094: Eur. Phys. J. {\bf C15}, 1 (2000).
1095:
1096: \bibitem{Gri} V.Yu.~Grishina, L.A.~Kondratyuk, E.L.~Bratkovskaya,
1097: M.~B\"uscher, and W.~Cassing, Eur. Phys. J. {\bf A9},
1098: 277 (2000).
1099: %
1100: \bibitem{kanzoandi} C.~Hanhart and K.~Nakayama,
1101: Phys. Lett. {\bf B454}, 176 (1999).
1102: %
1103: \bibitem{Arndt} R.A. Arndt, I.I. Strakovsky, and R.L. Workman,
1104: Phys. Rev. C {\bf 62}, 034005 (2000).
1105: %
1106: \bibitem{Kondr} M.~B\"uscher et al., {\it Investigation of neutral
1107: scalar mesons $a_0/f_0$ with ANKE, COSY proposal \# 97} (2001),
1108: http://ikpd15.ikp.kfa-juelich.de:8085/doc/Proposals.html
1109: %
1110: \bibitem{BNL} S.~Teige, et al.,
1111: Phys. Rev. D {\bf 59}, 012001 (1999).
1112: %
1113: \bibitem{Miller} G.A.~Miller, B.M.K.~Nefkens, and I.~\v Slaus,
1114: Phys. Rep. {\bf 194}, 1 (1990).
1115: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1116: \end{thebibliography}
1117:
1118: \newpage
1119:
1120: \begin{figure}
1121: \caption{Graphical representation of the Dyson equation.
1122: The thin solid line denotes the bare propagator whereas the
1123: thick solid line stands for the dressed/physical propagator. The sum
1124: appearing in the expression for the self energy is assumed to run over
1125: all decay channels $X$.}
1126: \label{dyson}
1127: \end{figure}
1128:
1129: \begin{figure}
1130: \caption{Different contributions to $\pi \eta$ production in leading
1131: order in the mixing. Here $f_0-a_0$ mixing is denoted by a circle and
1132: $\pi-\eta$ mixing is denoted by a cross.
1133: Note that $f_0-a_0$ mixing can occur via $K\bar K$ loops as well
1134: as via direct mixing, as explained in Sect. 3. The numbers in the
1135: primary production vertices (denoted by a big circle) indicate the
1136: isospin of the relevant amplitude.}
1137: \label{pieta}
1138: \end{figure}
1139:
1140: \begin{figure}
1141: \caption{Spectra of the $\pi^0\eta$ invariant mass for the reaction
1142: $pn\to d\pi^0\eta$ at $T_p=2645$~MeV$/c$. The dotted line corresponds
1143: to a reaction mechanism where
1144: an $a_0$ meson is produced in a $p$-wave and then decays into the
1145: $\pi^0\eta$ system.
1146: The dashed line corresponds to the production of an $f_0$ meson
1147: in an $s$-wave that mixes into an $a_0$ with $\Sigma_{10} =const$, cf.
1148: Eq. (\ref{smix}).
1149: The solid line shows results where $\Sigma_{10}$ is evaluated
1150: from the $K\bar K$ mixing mechanism.
1151: Note that all the distributions are normalized
1152: to 1 at their maximal value.}
1153: \label{fig3}
1154: \end{figure}
1155:
1156: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1157:
1158: \newpage
1159: \vglue 1cm
1160: \begin{figure}[tb]
1161: \vspace{3cm}
1162: \special{psfile=dyson.eps
1163: hoffset= 20 voffset =10 hscale=50 vscale=50}
1164: \vspace{4mm}
1165: \center{FIG. 1}
1166: \end{figure}
1167:
1168: \vskip 5cm
1169:
1170: \begin{figure}[tb]
1171: \vspace{4cm}
1172: \special{psfile=pieta.eps
1173: hoffset= 30 voffset
1174: =10 hscale=80 vscale=80}
1175: \vspace{4mm}
1176: \center{(a) \hskip 4cm (b) \hskip 4cm (c)}
1177: \vspace{8mm}
1178: \center{FIG. 2}
1179: \end{figure}
1180:
1181: \newpage
1182:
1183: \vglue 1cm
1184: \begin{figure}[t]
1185: \vspace{13cm}
1186: \special{psfile=spectm.eps
1187: hoffset= -30 voffset
1188: =-120 hscale=80 vscale=80}
1189: \vspace{4mm}
1190: \center{FIG. 3}
1191: \end{figure}
1192: %
1193: \end{document}
1194: