1: \documentclass[11pt]{article}
2: \usepackage{epsfig}
3: \usepackage{palatino}
4: %\usepackage{axodraw}
5: \usepackage{graphicx}
6: \oddsidemargin -0.5truecm
7: \textwidth 16.5cm
8: \begin{document}
9:
10: \title{$e^+e^-\to e^+e^-\pi^0\pi^0$ at DA$\Phi$NE}
11: \author{F.~Nguyen$^a$, F.~Piccinini$^b$ and A.D. Polosa$^c$\\
12: $^a$Dip. Fisica, Universit\`a ``Roma Tre'', and Sez. INFN, via della Vasca Navale 84, Roma, Italy\\
13: $^b$INFN, Sezione di Pavia and\\ Dip. Fisica
14: Nucleare e Teorica , via Bassi 6, Pavia, Italy\\
15: $^c$INFN, Sezione di Roma, p. A. Moro 2, Roma, Italy}
16: %\date{\today}
17: \maketitle
18: \begin{abstract}
19: The production of the $\sigma(500)$
20: meson in $\gamma\gamma\to\pi^0\pi^0$ is studied.
21: In particular, the KLOE data collected during the DA$\Phi$NE run at
22: $\sqrt{s}=1$~GeV are appropriate to this purpose
23: because of the strong reduction of Kaon backgrounds.\newline
24: %The $e^+e^-\to e^+e^-\pi^0\pi^0$
25: %channel is studied assuming the $\sigma$ to be a 4-quark meson.\newline
26: {\bf Preprint No.} FNT/T/2006-02\newline
27: {\bf Keywords} Hadron Spectroscopy, Phenomenological Quark Models\newline
28: {\bf PACS} 12.39.-x
29: \end{abstract}
30: \newpage
31:
32: \section{Introduction}
33: One of the first attempts to describe nucleon-pion interactions within a
34: spontaneously broken $SU(2)_L\otimes SU(2)_R$ theory was the linear sigma
35: model of Gell-Mann and Levy~\cite{gell}.
36: The idea is that of writing a Lagrangian of the $\psi=(p~n)$ fields
37: introducing chirally invariant terms of
38: the form $\bar{\psi}_L\Sigma\psi_R$ with $\Sigma$ transforming
39: linearly under chiral transformations: $\Sigma\to L\Sigma R^\dag$.
40: Since the defining $SU(2)$ representation is pseudoreal, one can write
41: $\Sigma$ in terms of only four real parameters. The
42: linear sigma model parameterization is indeed
43: $\Sigma=\sigma{\bf 1}+i{\mathbf \tau}^{a}\pi^a$, where $\pi^a$ are the three
44: isospin components of the pion and $\sigma$ is a scalar field possibly
45: associated to some particle in the spectrum. Upon spontaneous symmetry breaking
46: the nucleon gets a mass $m_N=F_\pi g_{\pi NN}$ in agreement with
47: the Goldberger-Treiman relation with $g_A$ fixed to be 1
48: (whereas the physical value is 1.26).
49: $F_\pi$ is the constant, with dimension of a mass, appearing
50: in the potential of the
51: model, the well known Mexican hat potential $V=\lambda/4[(\sigma^2+
52: \mathbf{\pi}^2)-F_\pi^2]^2$. One can show that $F_\pi$
53: so defined coincides with the pion decay constant.
54:
55: The drawback is that the artificial $\sigma$ field appears
56: to be coupled to pions and nucleons suggesting the existence of
57: a new particle to be looked for.
58: The natural process where a $\sigma$ contribution is expected to be important
59: is the $\pi\pi\to\pi\pi$ elastic channel.
60: Unfortunately experimental studies have never provided
61: over the years a clear signal for it and the assessment of $\sigma$ has
62: become more and more controversial.
63: The indication coming from $\pi\pi$ collision studies is that,
64: if the $\sigma$ can be considered as a resonance at all, it ought to be
65: an extremely broad (short lived) state.
66: Very recently~\cite{leut}
67: it has been shown that the $\pi\pi$ scattering amplitude
68: contains a pole with the quantum numbers of vacuum, the $\sigma$, with a mass
69: of $M_\sigma=441^{+16}_{-8}$~MeV and a
70: width $\Gamma_\sigma=544^{+25}_{-18}$~MeV.
71: The $\sigma$ has been looked for also in $D$ decays by the E791 Collaboration
72: at Fermilab~\cite{e791}.
73: From the $D\to 3\pi$ Dalitz plot analysis, E791 finds that almost the
74: $46\%$ of the width is due to $D\to\sigma\pi$ with a
75: $M_\sigma=478\pm 23 \pm 17$~MeV and
76: $\Gamma_\sigma=324\pm 40\pm 21$~MeV.
77: BES~\cite{bes} has looked for $\sigma$ in $J/\psi\to\omega\pi^+\pi^-$
78: giving a mass value of $M_\sigma=541\pm 39$~MeV and a width of
79: $\Gamma_\sigma=252\pm 42$~MeV. For a summary of experimental
80: data see~\cite{pdg}.
81:
82: Trying to extend the linear sigma model from $SU(2)_L\otimes SU(2)_R$
83: to $SU(3)_L\otimes SU(3)_R$ in order to include the strange sector
84: ($\psi=(u~d~s)$), one encounters the problem that the
85: fundamental representation of $SU(3)$ is complex and
86: a parameterization for the $\Sigma$ field having the same form
87: of the one given above requires 18 real parameters.
88: One possibility is to construct
89: a generalized sigma model with 9 scalar and 9 pseudoscalar fields~\cite{schec}.
90: Diverse solutions of this kind have been investigated in the literature but
91: none of them has proved to be effective in explaining data.
92:
93: The most successful
94: approach to build a theory of pions at low energies is that of
95: Callan-Coleman-Wess-Zumino (CCWZ)
96: to define $\Sigma=\exp(2iT^a\pi^a/F_\pi)$ in terms
97: of the 8 Goldstones arising from $SU(3)_L\otimes SU(3)_R\to SU(3)^{\rm diag}$.
98: With this definition of $\Sigma$,
99: pions do not transform linearly under $\Sigma\to L\Sigma R^\dag$.
100: The non-linear realization (one of the infinite realizations, all equivalent
101: at the level of physical results --the CCWZ theorem)
102: of chiral symmetry has proven to be extremely successful in the building
103: of Chiral Perturbation Theory (ChPT)~\cite{ecker},
104: the standard effective approach
105: to describe pion interactions at low energies.
106:
107: The non-linear sigma model excludes the
108: $\sigma$ field by construction, therefore the role of a $\sigma$ particle in
109: $\pi\pi$ interactions, as well as in heavy-light meson decays, calls
110: for understanding. Above all, is this a real particle associated
111: to some field in an effective Lagrangian (an exotic states Lagrangian
112: for example) or is it just a pion
113: rescattering effect?
114:
115: The problem of assessing the existence and nature of this state is not
116: confined to low energy phenomenology. Just to mention a possible relevant
117: physical scenario in which $\sigma$ could play a role,
118: consider the contamination of
119: $B\to\sigma\pi$ in $B\to\rho\pi$ decays (possible because of the large
120: $\sigma$ width). This could sensibly affect the isospin analysis for
121: the CKM-$\alpha$ angle extraction~\cite{bsig}. Similarly recent studies of the
122: $\gamma$ angle through a Dalitz analysis of neutral $D$ decays
123: need the presence of a $\sigma$ resonance in the fit~\cite{babsig}.
124:
125: In this paper we want to highlight the possibility that $\sigma$ could be
126: found in $e^+e^-$ collisions at DA$\Phi$NE now running at a center of
127: mass energy of 1~GeV, a region where the $\phi$
128: backgrounds are considerably diminished. In particular we examine
129: the $e^+e^-\to e^+ e^-\pi^0\pi^0$, $\gamma$-fusion channel.
130: This could represent the cleanest technique available for an independent measure
131: of the $\sigma$.
132:
133: \section{The $\gamma\gamma\to\pi^0\pi^0$ channel.}
134:
135: Our aim is to suggest to extract a $\sigma$ signal in $e^+e^-\to
136: e^+e^-\pi^0\pi^0$
137: where also ChPT predictions are very solid:
138: the $\gamma\gamma\to\pi^0\pi^0$ channel has been determined to
139: two loop accuracy in the region of photon-photon c.o.m. energy from
140: about $2 m_\pi$ up to $700$~MeV~\cite{sainio}. In the same energy region the
141: $\gamma\gamma\to\pi^+\pi^-$ is affected by a large background $\gamma\gamma\to\mu^+\mu^-$.
142:
143: To perform our calculation of this process
144: we will assume that the $\sigma$ is not a standard meson, but a
145: crypto-exotic state~\cite{scalmai}. In particular we assume that $\sigma$ is
146: $[qq][\bar q \bar q]$ where the parentheses indicate a diquark bound
147: in the ${\bf \bar 3}_c$ attractive color channel~\cite{wil}.
148: Assuming a common spatial
149: configuration and spin~0 for the diquark we have a ${\bf\bar 3}_f$ configuration
150: for the diquark and ${\bf\bar 3}_f$ for the antidiquark because of Fermi
151: statistics. We can therefore expect to have a full nonet of states
152: ${\bf 3}_f\otimes {\bf \bar 3}_f={\bf 1}_f\oplus {\bf 8}_f$ whose exotic
153: content is `crypted' (we would equally have such a nonet with a
154: standard $q\bar q$ assignation). Such hypothesis about the $\sigma$
155: structure is particularly appealing for two reasons. {\bf 1.} It allows to
156: obtain a mass spectrum for the scalar mesons
157: ($\sigma, f_0(980), a_0(980), \kappa(900)$) which shows an inverted pattern
158: with respect to the $q\bar q$ one. The experimental
159: spectrum for these states, provided the $\kappa$ is confirmed, resembles the
160: inverted pattern. {\bf 2.}
161: Considering the $\sigma$ to be a meson with a different body plan,
162: qualitatively different from pions, allows to formulate an independent
163: effective theory for scalar meson dynamics~\cite{scalmai} escaping
164: the problem of the emergence of $\sigma$ in ChPT.
165:
166: Our calculation will be performed using the full matrix element derived
167: by the Feynman amplitudes in Fig.~\ref{f:diag}(a). To further test
168: our results, we also adopt the double equivalent photon approximation,
169: see Fig.~\ref{f:diag}(b), largely used in the past for studying this
170: kind of processes~\cite{brodsky,panca}.
171:
172: \begin{figure}[htb]
173: \begin{center}
174: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
175: height=3truecm, width=7truecm,
176: figure=fig1.eps}%\vskip1cm
177: \hspace{1in}%
178: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
179: height=3truecm, width=4truecm,
180: figure=fig2.eps}%\vskip1cm
181: \caption{\footnotesize
182: (a) The Feynman diagrams for $e^+e^-\to e^+e^-\pi^0\pi^0$.
183: In the $t$-channel one photon is exchanged between electron and positron
184: and a $\sigma$ is coupled to it. The black dot indicates the
185: decay of $\sigma\to\pi^0\pi^0$ where the two pions are allowed to decay
186: $\pi^0\to\gamma\gamma$. The $\gamma$'s are eventually measured
187: in the detector. The final-particle phase space in our calculation
188: is the 4-body $e^+e^-\pi^0\pi^0$. In the $s$-channel a virtual photon emits a
189: $\sigma$ and another virtual photon, eventually decaying into $e^+e^-$.
190: (b) The $\sigma(\gamma\gamma\to\pi^0\pi^0)$ cross section is factorized
191: with the distribution functions of the photon in the electron. Such
192: approximation is strictly valid for collinear photons, being less and less
193: applicable once one allows for a $p_\perp\neq 0$ of the electron.
194: }
195: \label{f:diag}
196: \end{center}
197: \end{figure}
198:
199:
200: \begin{figure}[htb]
201: \begin{center}
202: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
203: height=7truecm, width=10truecm,
204: figure=fig3.eps}%\vskip1cm
205: \caption{\footnotesize
206: The vertex $\sigma\gamma\gamma$ is obtained assuming
207: Vector-Meson-Dominance (VMD): the $\sigma$ decays
208: to $\rho\rho$ and each $\rho$ converts to $\gamma$ according to its $f_\rho$
209: decay constant. Our microscopic picture of the decay process of
210: a $\sigma$ to $\rho\rho$ is a kind of
211: tunneling of a quark which escapes its diquark shell to meet an antiquark
212: from the antidiquark forming a standard color singlet $q\bar q$ meson.
213: The higher the barrier to be crossed the stronger is the diquark energy
214: binding.
215: The interaction strength $a$ is expressed as the coupling of a
216: gauge-invariant term $a/M_\sigma^2 \phi F^{\mu\nu}F_{\mu\nu}$, $\phi$
217: being the field associated to the $\sigma$~\cite{scalmai}.
218: }
219: \label{f:vert}
220: \end{center}
221: \end{figure}
222:
223: \begin{figure}[htb]
224: \begin{center}
225: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
226: height=7truecm, width=10truecm,
227: figure=miss.eps}%\vskip1cm
228: \caption{\footnotesize
229: The solid line represents the so
230: called double equivalent photon approximation; see Fig.~1(b)~\cite{panca}.
231: Dashed and dot-dashed represent the $t$ and $s+t$ channel respectively
232: in the Feynman diagram computation of Fig.~1(a).
233: By $m^*$ we denote the invariant mass of the $e^+e^-$ final state pair
234: (missing mass). With a moderate statistics simulation, the four-body phase
235: space of the matrix element calculation converges with more difficulty
236: with respect to the two body phase space of the double equivalent
237: photon approximation.
238: }
239: \label{f:mmass}
240: \end{center}
241: \end{figure}
242:
243: The double equivalent photon approximation is strictly valid when the
244: photons emitted by the electrons (see Fig.~\ref{f:diag}(b)) are collinear.
245: On the other hand the results of our, straightforward,
246: calculation are formally valid also
247: for final state electrons selected with a certain $p_\perp$. In such a way
248: one should be able, at least in principle, to perform
249: the experimental analysis also in absence of forward
250: electron detectors (electron tagging).
251:
252: The model dependent part of the calculation is in the $\gamma\gamma\sigma$
253: vertex. In Fig.~\ref{f:vert} we show how we do parameterize such interaction.
254: Because of gauge invariance, the vertex has the form $a/M_\sigma^2 \phi
255: F^{\mu\nu}F_{\mu\nu}$, $\phi$ representing the $\sigma$ field. In a previous
256: paper~\cite{scalmai} the method for the computation of the coupling of
257: a cryptoexotic $\sigma$ to $\rho$ mesons is traced. The coupling $a$ used
258: here needs not to be equivalent to the ${\cal A}=2.6$~GeV coupling fitted
259: in~\cite{scalmai}. ${\cal A}$ defines the decay vertex $\sigma\to\pi\pi$.
260: Since we cannot fit $a$ from data we treat it as a free
261: parameter. This doesn't affect our analysis since we will only discuss
262: normalized distributions. In other words the 4-quark
263: hypothesis does not alter quantitatively the nature of our conclusions
264: though it is intimately connected to the qualitative picture of the
265: microscopic dynamics of the $\sigma \to \rho\rho$ decay.
266: The $\rho$ mesons are coupled to photons using a VMD Ansatz,
267: see Fig.~\ref{f:vert}.
268:
269: The $\sigma$ propagation is described by a simple Breit-Wigner function
270: with mass and width taken from E791 data.
271: The cross section can be plotted as a function of several variables.
272: We choose to show the distributions in the missing mass $m^{*}$,
273: i.e., the invariant
274: mass of the final state $e^+e^-$ pair and in terms of
275: the invariant mass of the incoming $\gamma\gamma$, $W_{\gamma\gamma}$.
276:
277: \section{Results}
278: The reason for plotting the cross section distribution as a function
279: of the missing mass $m^*$ of the final $e^+e^-$ pair, Fig.~\ref{f:mmass} is that using
280: this variable we can get rid of the most relevant background we have,
281: namely $e^+e^-\to\omega\pi^0\to \pi^0\pi^0\gamma$. This is peaked around
282: zero in the $m^*$ variable and the distribution smearing could still
283: allow the
284: signal from background extraction.
285:
286: $\gamma\gamma\to\pi^0\pi^0$ has been computed in ChPT
287: to two-loop level accuracy~\cite{sainio}. We can as well plot the normalized distribution
288: $1/\sigma*d\sigma/dW_{\gamma\gamma}$, $W_{\gamma\gamma}$ being
289: the photon-photon center of mass energy, see Fig.~\ref{f:chpt}.
290: The two-loop ChPT cross section $\sigma(\gamma\gamma\to
291: \pi^0\pi^0)$ is convoluted with photon distribution functions as prescribed by
292: the double equivalent photon approximation.
293:
294: \begin{figure}[htb]
295: \begin{center}
296: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
297: height=7truecm, width=10truecm,
298: figure=comp-chpt.eps}%\vskip1cm
299: \caption{\footnotesize
300: $W_{\gamma\gamma}$ is the photon-photon center of mass energy. The solid line
301: is the $e^+e^-\to e^+e^-\pi^0\pi^0$ cross section distribution calculated
302: convoluting the two-loop $\sigma (\gamma\gamma\to\pi^0\pi^0)$ cross
303: section with the photon distribution functions according to the
304: double equivalent photon approximation. Dashed, dot-dashed and dotted lines
305: represent the same quantity
306: where the two pions come from the decay of a $\sigma$ produced in photon-photon
307: fusion. In particular the dashed curve is computed using the $\sigma$ parameters
308: found by E791 ($M_\sigma=478$~MeV and $\Gamma_\sigma=324$~MeV)~\cite{e791},
309: the dot-dashed corresponds to an $M_\sigma=441$~MeV and
310: $\Gamma_\sigma=544$~MeV (from the $\pi\pi$ scattering amplitude
311: analysis~\cite{leut}) while the dotted curve is computed
312: according to BES results ($M_\sigma=541$~MeV and $\Gamma_\sigma=252$~MeV)~\cite{bes}.
313: }
314: \label{f:chpt}
315: \end{center}
316: \end{figure}
317:
318: It is not surprising that the effect of a Breit-Wigner resonance
319: makes this distribution different from what expected in
320: ChPT where the $\sigma$ can only be simulated by an effect of strong
321: interaction in the $\pi\pi$ channel.
322: Assuming that $\sigma$ is a propagating particle, with either 4-quark or
323: 2-quark structure, qualitatively changes the $d\sigma/dW_{\gamma\gamma}$
324: with respect to ChPT predictions.
325: Anyway the $W_{\gamma\gamma}$ region where our calculation differs more
326: sensibly from ChPT could be affected by interference effects with $\gamma\gamma\to f_0(980)\to \pi^0\pi^0$. However to address this issue will require precise data at much higher $\sqrt{s}$ values.
327: %@@@@@
328:
329: In Fig.~\ref{f:combi} we compare our results
330: to the only existing data set for the $\sigma(\gamma\gamma\to\pi^0\pi^0)$, namely the 1993
331: Crystal Ball data points. The solid line represent the two-loop ChPT result . The dashed line
332: is $\sigma(\gamma\gamma\to\sigma\to\pi^0\pi^0)_{\rm B.W.}$ where the $a$ parameter has been fitted
333: to the cross section value of the lowest $W_{\gamma\gamma}$ experimental point.
334: The dotted line is the same as the dashed re-weighting the amplitude by an Adler zero
335: factor~\cite{adler} $(s-s_A)$ where $s_A=0.5m_\pi^2$.
336: The resonant contribution should be properly combined with the chiral loop amplitudes to show
337: a more realistic result: the two contributions do not exclude each other.
338:
339: Our results are certainly dependent on the Breit-Wigner (BW) parameterization of the $\sigma$
340: contribution (we have also analyzed the possibility of implementing the BW Ansatz with a
341: $\sigma$ comoving width).
342: However, given the poor resolution expected for low energy photons and the actual absence of an electron tagging device, we believe that our simple approach is appropriate enough to motivate
343: an experimental analysis in this channel: Fig.~\ref{f:combi} certainly enforces the case to require more precise data. The use of a Breit-Wigner parameterization of $\sigma$ at E791 and BES is probably more challenged by the fact that charged pions momenta are more precisely measured in these experiments, allowing a quite good resolution on invariant masses.
344: Moreover the recent experience at BaBar~\cite{cavoto} where the $D(D_s)\to 3\pi$ Dalitz plots are currently under study, has shown preliminarly that the use of more sophisticated approaches to the parameterization of the isoscalars (e.g. K-Matrix) are not providing more stringent and convincing results in the extraction of the $D\to \sigma\pi $ channel with respect to what obtainable via simple BW Ansatz.
345:
346: \begin{figure}[htb]
347: \begin{center}
348: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
349: height=6.5truecm, width=9truecm,
350: figure=4qpuro.eps}%\vskip1cm
351: %\vspace{-0.5truecm}
352: \caption{\footnotesize
353: Data points are cross section values for $\gamma\gamma\to\pi^0\pi^0$ obtained by Crystal Ball
354: (1993). The solid curve represent the two-loop ChPT result . The dashed line
355: is $\sigma(\gamma\gamma\to\sigma\to\pi^0\pi^0)_{\rm B.W.}$.
356: The dotted line is the same as the dashed re-weighting the amplitude by an Adler zero
357: factor~\cite{adler} $(s-s_A)$ where $s_A=0.5m_\pi^2$.
358: Data uncertainties
359: certainly call for a more precise measurement of this channel.
360: }
361: \label{f:combi}
362: \end{center}
363: \end{figure}
364:
365:
366: \section{Backgrounds}
367:
368: \begin{figure}[htb]
369: \begin{center}
370: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
371: height=8truecm, width=8truecm,
372: figure=smearom.eps}%\vskip1cm
373: \hspace{0.001in}%
374: \epsfig{%bbllx=0.5cm,bblly=16cm,bburx=20cm,bbury=23cm,
375: height=8truecm, width=8truecm,
376: figure=smearom-pi0pi0.eps}%\vskip1cm
377: \vspace{-1.5truecm}
378: \caption{\footnotesize
379: In the left panel we compare the background $e^+e^-\to \omega\pi^0$
380: distribution (the one centered around zero) to the
381: signal (the broader distribution) as functions of
382: the missing mass $m_{\rm miss}\equiv m^*$. In the background we have five photons
383: in the final state. We pick up four of them reconstructing
384: $\pi^0\pi^0\to 4\gamma$ and perform the smearing, as explained in the
385: text.
386: Even after smearing the signal and the $\omega\pi^0$
387: background appear to have different shapes in $m^*$.
388: In the right panel we perform the same comparison using another variable,
389: namely, the invariant mass of the $\pi^0\pi^0$ system. The peak is in
390: correspondence of two back-to-back $\pi^0$'s; the shape at $\sim 280$~MeV
391: is the $2 m_\pi$ invariant mass.
392: }
393: \label{f:smear}
394: \end{center}
395: \end{figure}
396:
397:
398: Here we proceed to a list of the relevant backgrounds which have to be
399: taken under control in the data analysis. In fact, even if KLOE
400: is collecting data at $\sqrt{s}=1$~GeV, some of the $\phi$ decays
401: with at least 4 photons in the final state have a rate comparable
402: to the signal one (we require at least
403: $4\gamma$'s in the final state).
404:
405: \begin{itemize}
406: \item $\phi\to\eta\gamma\to 3\pi^0\gamma.$ At a center of mass energy of
407: $1$~GeV the $\phi$ is strongly reduced. The cross section
408: for $e^+e^-\to\phi\to\eta\gamma$ drops by about a factor of
409: 30~\cite{Akhmetshin:2004gw}
410: from $~20$~nb at $\sqrt{s}=M_\phi$ down to $0.67$~nb
411: at $\sqrt{s}\simeq1$~GeV;
412: since the branching ratio $BR(\eta\to 3\pi^0)\simeq 32.5\%$,
413: we have $\sigma(e^+e^-\to\phi\to\eta\gamma\to 3\pi^0\gamma)\sim 0.2$~nb,
414: at $\sqrt{s}\simeq1$~GeV.
415: \item $\phi\to K_S K_L.$ The cross section in this case drops from
416: $1350$~nb to $11$~nb according to the indications
417: of CMD-2~\cite{Akhmetshin:1999ym}.
418: This background is particularly dangerous when $K_S$ decays into $2\pi^0$'s
419: ($BR(K_S\to 2\pi^0)\simeq31\%$) and $K_L$ escapes detection.
420: Considering that in KLOE at $\sqrt{s}=1$~GeV, about $75\%$ of $K_L$'s decay inside the detector,
421: and that about $70\%$ of the surviving ones interact in the Electromagnetic Calorimeter
422: we get a further reduction of the $\sigma(e^+e^-\to\phi\to
423: K_S K_L\to\pi^0\pi^0 K_L)$ (with $K_L$ undetected)
424: which amounts to $~0.2$~nb.
425: \item $\phi\to\pi^0\pi^0\gamma.$ Starting from the
426: KLOE result~\cite{Aloisio:2002bt}
427: $BR(\phi\to f_0(980)\gamma)\simeq 10^{-4}$ at $\sqrt{s}=M_\phi$,
428: and assuming that $\sigma(e^+e^-\to\phi\to f_0(980)\gamma)$
429: gets reduced by a factor 30 at $\sqrt{s}=1$~GeV,
430: we estimate a cross section of about
431: $\sim 10$~pb for this channel.
432: \item $\phi\to \pi^0\eta\gamma.$ Given the KLOE result~\cite{Aloisio:2002bs}
433: $BR(\phi\to a_0(980)\gamma)\simeq 7\times10^{-5}$ at $\sqrt{s}=M_\phi$,
434: and taking into account $BR(\eta\to\gamma\gamma)\simeq 39.4\%$ and the
435: reduction by a factor of 30, we estimate a cross section of about
436: $\sim 4$~pb for this channel.
437: \item $e^+e^-\to\eta e^+e^-\to \pi^0\pi^0\pi^0 e^+e^-.$
438: We estimate a cross section for this process amounting to $\sigma\simeq 13$~pb.
439: Such a background should be removable imposing that the 4
440: reconstructed photons belong to 2 pions going
441: back-to-back in the transverse plane.
442: \end{itemize}
443:
444: \subsection{The $e^+e^-\to \omega\pi^0$ channel}
445: Contrary to $\phi$ decays, the cross section for
446: this process tends to increase from $\sigma=(0.51\pm0.07)$~nb
447: at $\sqrt{s}=1.02$~GeV to $\sigma=(0.56\pm 0.14)$~nb
448: at $\sqrt{s}=1.005$~GeV, as measured by the
449: SND~\cite{Achasov:1999wr} experiment. Moreover the peak at $~0.6$~GeV in the $\pi\pi$
450: invariant mass, as shown in the right panel of Fig.~6, corresponds to events with 2 pions
451: emerging back-to-back in the transverse plane.
452:
453: Since the fraction of background events in which the fifth photon is lost has the same experimental
454: signature of the signal, namely $2\pi^0$'s collinear in the transverse plane plus missing energy,
455: we performed a dedicated study of this channel.
456: We consider $e^+e^-\to \omega\pi^0\to\pi^0\pi^0\gamma$
457: to proceed via VMD $e^+e^-\to\gamma^*\to\rho\to\omega\pi^0$.
458: This channel has a missing mass peaked at zero.
459: However, the energy resolution of the photons coming from the
460: $2\pi^0$'s modifies this simple pattern.
461:
462: Every photon energy distribution is convoluted
463: with a Gaussian function, where the standard deviation
464: is the energy resolution function of the KLOE
465: Electromagnetic Calorimeter~\cite{Adinolfi:2002zx}:
466: $$
467: \frac{\sigma_E}{E} ~ ~=~ ~ \frac{5.7\%}{\sqrt{E~[\mathrm{GeV}]}}.
468: $$
469: The same procedure is applied to the 4 photons
470: generated by the signal $e^+e^-\to\pi^0\pi^0 e^+e^-$.
471:
472: Upon such a smearing procedure, no significant
473: change is introduced in the signal distribution,
474: while the $\delta$-function in
475: the missing mass gets broadened but still does not
476: superimpose to the
477: signal region, see Fig.~\ref{f:mmass} and Fig.~\ref{f:smear}
478: where the cross section value
479: $\sigma(e^+e^-\to\omega\pi^0\to\pi^0\pi^0\gamma)=0.6$~nb is used.
480:
481: It is also clear that the experimental analysis is going to
482: deal not only with the photon energy resolution, but
483: also with the photon efficiency as a function
484: of the energy and of the polar angle
485: given the correct pairing
486: of the final photons to the parent $\pi^0$'s. A full
487: simulation of the detector would be in order for a detailed
488: discussion of these issues.
489: Nevertheless, it is reasonable to guess that the same
490: sources of inefficiency are shared by the
491: signal and the background processes.
492:
493: \section{Conclusions}
494: We believe that KLOE in the low energy DA$\Phi$NE
495: run has a concrete opportunity to find (or disprove) the
496: $\sigma$ in a clean experimental channel, $\gamma\gamma\to\pi^0\pi^0$.
497: The scope of our feasibility study is to underscore this possibility and to
498: suggest some data analysis strategies. The crucial interest should be
499: that of investigating the nature of the $\sigma$: is it a particle or
500: an effect of chiral dynamics?
501: The comparison with ChPT predictions for
502: $\gamma\gamma\to\pi^0\pi^0$ illustrates that if a broad resonance
503: is indeed produced in $\gamma\gamma$ it could be detectable in the
504: low energy $\gamma\gamma$ region, see Fig.~\ref{f:chpt},
505: from the distribution slopes.
506: As soon as smearing is introduced the two curves (resonance-type versus ChPT) will tend to
507: superimpose. This certainly calls for a more selective analysis of data
508: making use of forward detectors to tag electrons. Such devices, which at the moment
509: are not part of the experimental apparatus, could be integrated in view of possible future luminosity upgrades.
510:
511: \section*{Acknowledgements}
512: We wish to thank F.~Ambrosino, D.~Babusci, C.~Bloise, F.~Ceradini, G.~Capon,
513: G. D'Ambrosio,
514: S.~Giovannella, M.~Sainio and G.~Pancheri
515: for useful discussions. In particular we wish to
516: thank L.~Maiani for his comments on the manuscript.
517:
518: \begin{thebibliography}{99}
519: \bibitem{gell} M. Gell-Mann, L\'{e}vy, Nuov. Cim. {\bf 16}, 705 (1960)
520: \bibitem{leut} I. Caprini, G. Colangelo and H. Leutwyler,
521: [arXiv:hep-ph/0512364]
522: \bibitem{pdg} S.~Eidelman {\it et al.} PDG, Phys. Lett. {\bf B592},1 (2004).
523: \bibitem{e791} E.M.~Aitala {\it et al.} [E791 Collab.], Phys. Rev. Lett.
524: {\bf 86},770 (2001).
525: \bibitem{bes} M.~Ablikim {\it et al.} [BES Collab.], Phys. Lett.
526: {\bf B598}, 149 (2004).
527: \bibitem{schec} A.H.~Fariborz, R.~Jora and J.~Schechter,
528: Int. J. Mod. Phys. {\bf A20}, 6178 (2005).
529: \bibitem{ecker} G. Ecker, {\it Chiral Perturbation Theory},
530: [arXiv:hep-ph/9501357]
531: \bibitem{bsig} A.~Deandrea and A.D.~Polosa, Phys. Rev. Lett.
532: {\bf 86}, 216 (2001); S.~Gardner and U.~Meissner, Phys. Rev. {\bf D65},
533: 094004 (2002); I.~Bigi [arXiv:hep-ph/0601167].
534: \bibitem{babsig} B.~Aubert {\it et al.}, [BaBar Collab.], Phys. Rev. Lett.
535: {\bf 95}, 121802 (2005).
536: \bibitem{sainio} J.~Gasser, M.A.~Ivanov and M.~Sainio, Nucl. Phys.
537: {\bf B728}, 31 (2005); see also
538: S.~Bellucci, J.~Gasser and M.~Sainio Nucl. Phys. {\bf B423}, 80 (1994);
539: erratum-ibid. {\bf B431}, 413-414 (1994).
540: \bibitem{scalmai} L.~Maiani, F.~Piccinini, A.D.~Polosa, V.~Riquer,
541: Phys. Rev. Lett. {\bf 93}, 212002 (2004), [arXiv:hep-ph/0407017]
542: \bibitem{wil}
543: R.L.~Jaffe and F.~Wilczek, Phys. Rev. Lett. {\bf 91} (2003) 232003.
544: \bibitem{brodsky} S.~Brodsky, T.~Kinoshita and H.Terazawa, Phys. Rev.
545: {\bf D4}, 1532 (1971).
546: \bibitem{panca} G.~Alexander {\it et al.}, Nuovo Cimento {\bf 107}, 837 (1994).
547: \bibitem{adler} S.~Weinberg, Phys. Rev. Lett. {\bf 17}, 616 (1966); See also the recent review by
548: D.~Bugg, [arXiv:hep-ex/0510021] and G. Pancheri and Y.N. Srivastava, Lett. Nuovo Cim. {\bf 13},
549: 221 (1975).
550: \bibitem{cavoto} G.~Cavoto, private communications.
551: \bibitem{Akhmetshin:2004gw}
552: R.~R.~Akhmetshin {\it et al.} [CMD-2 Collab.],
553: %``Study of the processes e+ e- $\to$ eta gamma, pi0 gamma $\to$ 3gamma in the
554: %cm. energy range 600-MeV - 1380-MeV at CMD-2,''
555: Phys.\ Lett.\ B {\bf 605}, 26 (2005).
556: \bibitem{Akhmetshin:1999ym}
557: R.~R.~Akhmetshin {\it et al.} [CMD-2 Collab.],
558: %``Measurement of phi meson parameters in K0(L) K0(S) decay mode with CMD-2,''
559: Phys.\ Lett.\ B {\bf 466}, 385 (1999)
560: [Erratum-ibid.\ B {\bf 508}, 217 (2001)].
561: \bibitem{Aloisio:2002bt}
562: A.~Aloisio {\it et al.} [KLOE Collab.],
563: %``Study of the decay Phi $\to$ pi0 pi0 gamma with the KLOE detector,''
564: Phys.\ Lett.\ B {\bf 537}, 21 (2002).
565: \bibitem{Aloisio:2002bs}
566: A.~Aloisio {\it et al.} [KLOE Collab.],
567: %``Study of the decay Phi $\to$ eta pi0 gamma with the KLOE detector,''
568: Phys.\ Lett.\ B {\bf 536}, 209 (2002).
569: \bibitem{Achasov:1999wr}
570: M.~N.~Achasov {\it et al.}, [SND Collab.]
571: %``Investigation of the e+ e- $\to$ omega pi0 $\to$ pi0 pi0 gamma reaction in
572: %the energy domain near the Phi meson,''
573: Nucl.\ Phys.\ B {\bf 569}, 158 (2000).
574: \bibitem{Adinolfi:2002zx}
575: M.~Adinolfi {\it et al.}, [KLOE Collab.]
576: %``The KLOE electromagnetic calorimeter,''
577: Nucl.\ Instrum.\ Meth.\ A {\bf 482}, 364 (2002).
578: %\bibitem{cmd2} cmd2
579: %\bibitem{lucini} C.~Alexandrou, P.~DeForcrand, B. Lucini, hep-lat/0509113.
580: \end{thebibliography}
581: \end{document}
582:
583:
584:
585: