1: \documentclass[prd,aps,preprintnumbers,showpacs]{revtex4}

2: \usepackage{epsfig}

3: \bibliographystyle{apsrev}

4: %\oddsidemargin =2 truecm

5: %\usepackage[active]{srcltx}

6: \usepackage{graphicx}

7: \usepackage{bm}

8: \newcommand{\spur}[1]{\not\! #1 \,}

9: \newcommand{\be}{\begin{equation}}

10: \newcommand{\dd}{\displaystyle}

11: \newcommand{\ee}{\end{equation}}

12: \newcommand{\bea}{\begin{eqnarray}}

13: \newcommand{\eea}{\end{eqnarray}}

14: \newcommand{\abs}[1]{\left\vert#1\right\vert}

15: \def\qq{Q\!\!\!\!Q}

16: \def\starteq{%\vspace{.1in}

17: \begin{eqnarray}}

18: \def\fineq{\end{eqnarray}}

19: \let\un=\b

20: \let\ced=\c

21: \let\du=\d

22: \let\um=\H

23: \let\sll=\l

24: \let\Sll=\L

25: \let\slo=\o

26: \let\Slo=\O

27: \let\tie=\t

28: \let\br=\u

29: \let\SEC=\S

30: \def\cA{{\cal A}}

31: \def\cB{{\cal B}}

32: \def\cC{{\cal C}}

33: \def\cD{{\cal D}}

34: \def\cE{{\cal E}}

35: \def\cF{{\cal F}}

36: \def\cG{{\cal G}}

37: \def\cH{{\cal H}}

38: \def\cI{{\cal I}}

39: \def\cJ{{\cal J}}

40: \def\cK{{\cal K}}

41: \def\cL{{\cal L}}

42: \def\cM{{\cal M}}

43: \def\cN{{\cal N}}

44: \def\cO{{\cal O}}

45: \def\cP{{\cal P}}

46: \def\cQ{{\cal Q}}

47: \def\cR{{\cal R}}

48: \def\cS{{\cal S}}

49: \def\cT{{\cal T}}

50: \def\cU{{\cal U}}

51: \def\cV{{\cal V}}

52: \def\cW{{\cal W}}

53: \def\cX{{\cal X}}

54: \def\cY{{\cal Y}}

55: \def\cZ{{\cal Z}}

56: % Greek alphabet

57: \def\a{\alpha}

58: \def\b{\beta}

59: % \def\c{\chi}

60: \def\d{\delta}      \def\D{\Delta}

61: \def\e{\epsilon}

62: \def\eps{\varepsilon}

63: \def\f{\phi}        \def\F{\Phi}

64: \def\vf{\varphi}

65: \def\g{\gamma}      \def\G{\Gamma}

66: \def\h{\eta}

67: \def\i{\iota}

68: \def\j{\psi}        \def\J{\Psi}

69: \def\k{\kappa}

70: \def\l{\lambda}   \def\L{\Lambda}

71: \def\m{\mu}

72: \def\n{\nu}

73: \def\o{\omega}   \def\O{\Omega}

74: \def\p{\pi}      \def\P{\Pi}

75: \def\q{\theta}   \def\Q{\Theta}

76: \def\r{\rho}

77: \def\s{\sigma}   \def\S{\Sigma}

78: \def\t{\tau}

79: \def\u{\upsilon}  \def\U{\Upsilon}

80: \def\x{\xi}      \def\X{\Xi}

81: \def\z{\zeta}

82: \def\dag{^{\dagger}{}}

83: \def\rarr{\rightarrow}

84: \def\larr{\leftarrow}

85: \def\iff{\leftrightarrow}

86: \def\conj{^{\raisebox{-.4cm}{\normalsize *}}}

87: \newcommand{\nn}{\nonumber}

88: \newcommand{\de}{\partial}

89: \def\Black{\special{color cmyk 0 0 0 1.}}

90:  \def\AliasBlue{\special{color cmyk 1. 1  0.2 0.3}}

91:  \def\Blue{\special{color cmyk 1. 0.8 0 0.3}}

92:  \def\Brown{\special{color cmyk 0 0.4 0.6 0.4}}

93: 

94: \begin{document}

95: %\begin{titlepage}

96: \preprint{BARI-TH 496/04}

97: \preprint{DSF-2004/25 (Napoli)}

98: 

99: \title{Final state interactions for $B\to VV$ charmless decays}

100: 

101: 

102: \author{\textbf{Massimo Ladisa, Vincenzo Laporta, Giuseppe Nardulli}}

103: \affiliation{Dipartimento di Fisica dell'Universit{\`a} di Bari, Italy\\

104: Istituto Nazionale di Fisica Nucleare, Sezione di Bari, Italy}

105: %

106: \author{\textbf{Pietro Santorelli}}

107: \affiliation{Dipartimento di Scienze Fisiche, Universit{\`a} di

108: Napoli "Federico II", Italy\\

109: Istituto Nazionale di Fisica Nucleare, Sezione di Napoli, Italy}

110: 

111: 

112: \begin{abstract}

113: We estimate final state interactions in the

114: $B$-meson decays into two light vector mesons by the Regge model.

115: We consider Pomeron exchange and charmed Regge trajectories that

116: can relate intermediate charmed particles to the final state. The

117: Regge poles have various helicity-flip residues, which allows a

118: change from the longitudinal to transverse  polarization. In this

119: way a significant reduction of the longitudinal polarization

120: fraction can be produced. In the factorization approximation we

121: find agreement with recent data from the BaBar and Belle

122: collaborations in the $B\to K^*\phi$ decay channel, as a result of

123: an appropriate choice of semileptonic form factors and Regge

124: exchanges. On the other hand, data for the $K^*\r$ decay channels

125: appear more elusive. The soft effects discussed in the present

126: paper are based on a model of Regge trajectories that is shown to

127: reproduce correctly in the non-charmed case the Regge

128: phenomenology of light mesons.

129: \end{abstract}

130: \pacs{13.25.Hw}

131: 

132: \maketitle

133: 

134: %\setcounter{page}{1}

135: \section{Introduction \label{sec:-1}}

136: Recent data from the Babar and Belle collaborations

137: \cite{Aubert:2003mm,Aubert:2003xc,belle:2003jf,Zhang:2003up} on

138: the $B$ decays into two light vector mesons $B\to V\,V$ have

139: produced considerable theoretical interest, see e.g. the recent

140: papers

141: \cite{Kagan:2004uw,Grossman:2003qi,Colangelo:2004rd,Kagan:2004ia}.

142: Data for the Branching Ratios (BR) and  the polarization fractions

143: for a few decay channels are reported in Table~\ref{table0}.

144: \begin{table}[h]

145: \begin{center}

146: \begin{tabular}{|c|c|c|c|c|}

147: \hline Decay mode &BR & $\Gamma_L/\Gamma$&$\Gamma_\perp/\Gamma$

148: &Ref.\\

149: \hline

150: $B^+\to\rho^0 K^{*+}$&$(10.6^{+3.0}_{-2.6}\pm2.4)\times10^{-6}$\,

151: &$0.96^{+0.04}_{-0.15}\pm0.04$&

152: &\cite{Aubert:2003mm},\cite{Aubert:2003xc}\\

153: \hline$B^+\to\phi K^{*+}$&$(9.5\pm1.7)\times10^{-6}$&$0.46\pm0.12\pm0.03$&

154: &\cite{Aubert:2003mm},\cite{belle:2003jf}\\

155: \hline $B^0\to\phi

156: K^{*0}$&$(10.7\pm1.2)\times10^{-6}$&$0.58\pm0.06$&$0.41\pm0.10\pm0.02$

157: &\cite{Aubert:2003mm},\cite{belle:2003jf}\\

158: \hline $B^+\to\rho^0

159: \rho^{+}$&$(26.2\pm6.2)\times10^{-6}$&$0.96\pm0.07$&

160: &\cite{Aubert:2003mm},\cite{Aubert:2003xc},\cite{Zhang:2003up}\\

161: \hline $B^0\to\rho^+

162: \rho^-$&$(25^{+7+5}_{-6-6})\times10^{-6}$&$0.98^{+0.02}_{-0.08}\pm0.03$&

163: &\cite{Aubert:2003mm},\cite{Aubert:2003xc}\\

164: \hline

165: \end{tabular}

166: \end{center}\caption{

167: Survey of experimental results for $B$ decays into two light

168: vector mesons. Data  of Refs.\cite{Aubert:2003mm} and

169: \cite{Aubert:2003xc} are  from the BaBar

170: collaboration. Data from Refs. \cite{belle:2003jf} and \cite{Zhang:2003up}

171: are from the Belle Collaboration. There is also an

172: upper bound ${\cal B}(B^0 \to \rho^0

173: \rho^0)\le 2.1 \times 10^{-6}$ from BaBar

174: \cite{Aubert:2003mm}.

175: \label{table0}}

176: \end{table}

177: 

178:  In these decay modes the two vector mesons have the same helicity; therefore three

179:  different polarization states are possible,

180:  one longitudinal ($L$) and two transverse, corresponding to helicities $\l=0$ and

181:  $\l=\pm 1$. We define the corresponding amplitudes as $A_{0,\pm}$. It is easy to

182:  show \cite{Kagan:2004uw} that in the large $m_b$ limit longitudinal

183:  polarization must be enhanced by a factor of $m_b$ in the amplitude.

184:  This can be seen at the quark level, the reason being that transverse

185:  polarizations are generated by helicity flip and this implies an extra

186:  factor $m_V/m_b$. Alternatively, at the hadronic level,

187:  by making use of naive factorization, one traces back the enhancement of

188:  the longitudinal polarization to the presence of an extra factor

189:  $m_B/m_V+{\cal O}((m_V/m_B)^2)$ in the longitudinal polarization vectors.

190:  The enhancement in the longitudinal amplitude

191:  is of one power of $m_b$. In fact the term with

192: two powers is multiplied by the difference

193: $A_1-A_2$ where  $A_1$ and $A_2$ are the usual axial form factors \cite{Bauer:1986bm}

194: for the transition

195:  $B\to V$ computed at $q^2=m_{V}^2$. This difference vanishes in the

196:  high $m_b$ limit, see e.g. \cite{Charles:1998dr}.

197: 

198: Decay modes $B^0 \to \rho^+ \rho^-$

199: and  $B^+ \to \rho^0 \rho^+, \rho^0 K^{*+}$ confirm this prediction

200: as the final states are

201: dominated by the longitudinal polarization. On the other hand,

202: for the observed $B \to \phi K^*$ transitions, the longitudinal

203: amplitude width gives about $~50\%$ of the rate. This is a major puzzling

204:  feature because, as argued above, in the naive factorization,

205: one expects also for the $\phi K^*$ channels

206: \be

207: \frac{\Gamma_L}{\Gamma}=\frac{|A_0|^2}{|A_0|^2+|A_+|^2+|A_-|^2}=

208: 1-{\cal O}\left(\frac{1}{m^2_b}\right)\ .

209: \ee

210: Another surprising aspect concerns the transverse polarization defined by

211: \be

212: \Gamma_\perp\propto |A_\perp|^2\, ,

213: \ee

214: where $A_{\perp,\parallel}=(A_+\mp A_-)/\sqrt 2$. Moreover, one expects \cite{Kagan:2004uw}

215: \be

216:  \frac{\Gamma_\perp}{\Gamma_\parallel}=

217:  1+{\cal O}\left(\frac{1}{m_b}\right)

218:  \label{naive}\, .

219:  \ee

220:  This

221:  expectation is based on naive factorization and large energy  relations

222:  \cite{Charles:1998dr}. The former assumption implies that

223:  in the large $m_b$ limit  $A_-$ is proportional to $(A_1-V)/(A_1+V)+{\cal O}(m_V/m_B)$ where

224:  $A_1$ and $V$ are the usual form factors \cite{Bauer:1986bm} for the transition

225:  $B\to K^*$ computed at $q^2=m_{\phi}^2$. The latter implies that

226: $(A_1-V)/(A_1+V)={\cal O}(1/m_b)$, a prediction confirmed

227: by different numerical computations

228: based on lattice QCD or QCD sum rules. As a consequence $A_-/A_+={\cal O}(1/m_b)$

229: and (\ref{naive}) should hold. A result from Belle is reported in

230: Table \ref{table0}, which is at odds with (\ref{naive}); a smaller

231: result from BaBar \cite{Gritsan:2004new}

232: \be

233: \frac{\Gamma_\perp}{\Gamma}=0.27\pm 0.07\pm 0.02

234: \label{babarr}

235: \ee

236: has been also presented, which might be compatible with (\ref{naive}).

237: 

238: All previous considerations do not take into account final state

239: interactions (FSI) that, although power suppressed, might nevertheless

240: produce sizeable effects. In \cite{Colangelo:2004rd}

241: a parametrization of these effects is considered. It assumes a major role of

242: the so-called charming penguin diagrams. They are

243: Feynman diagrams with  intermediate $D\bar D$ states

244: \cite{Colangelo:1989gi,Ciuchini:1997hb,Isola:2001ar,Isola:2001bn,Isola:2003fh}.

245: The suppression due to their non-factorizable status might be

246: compensated by the Cabibbo-Kobayashi-Maskawa (CKM) enhancement.

247: The computational scheme used in \cite{Colangelo:2004rd}

248: and in previous computations for $B$ decays into two light particles

249: \cite{Isola:2001ar,Isola:2001bn,Isola:2003fh} is chiral perturbation

250: theory for light and heavy mesons (for a review see \cite{Casalbuoni:1997pg}).

251: Since  momenta of the final particles are hard, the application of the method

252: involves large extrapolations and the method is questionable

253: (see e.g. \cite{Kagan:2004ia}). A possible answer consists of  introducing

254: some correction. In \cite{Isola:2001ar,Isola:2001bn,Isola:2003fh} such correction

255: was modelled by a form factor evaluated by a constituent quark model. In

256: \cite{Colangelo:2004rd} a phenomenological parameter $r$ is introduced to weight

257: the charming penguin contribution. Both approaches result in a suppression

258: of charming penguin contributions, but uncertainties are  huge.

259: 

260: The reason for the suppression can be traced back to the Regge theory.

261: In fact, given the rather large energy involved (${\sqrt s}=m_B$),

262: the Regge approach should be

263: a good approximation for the  final state interactions and provide

264:  a computational scheme of the rescattering effects,

265:  be they elastic or inelastic. The former are described by Pomeron exchange

266:   \cite{Donoghue:1996hz}, \cite{Nardulli:1997ht}.

267:  Among the latter, one may still assume dominance

268:  of the charming penguin amplitudes, due to the CKM enhancement, but

269:  they should be evaluated {\it via} a reggeized amplitude.

270: In this context the suppression of the charming penguin

271: arises because the Regge charmed

272: trajectories have a negative intercept $\alpha(0)$ and therefore a suppression factor

273: $(s/s_0)^{\alpha(0)}$ ($s_0\simeq 1$ GeV$^2$, a threshold). The advantage of the

274: Regge approach is to evaluate the intermediate charmed states contributions

275: not by Feynman diagrams, but by unitarity diagrams and the Watson's theorem \cite{Watson:1952ji}.

276: In this way the remarks connected to the extrapolation of chiral theory to

277: hard momenta are by-passed.

278: 

279: The aim of this paper is to give an evaluation of the final

280:  state interactions for $B\to VV$ using

281: the Regge model. Clearly also the numerical values of the bare (no

282: FSI) amplitudes are important. We compute them in the

283: factorization approximation and therefore we need a set of

284: semileptonic form factors as an input. We use a determination

285: based on Light Cone QCD sum rules \cite{Ball:2003rd} that

286: encompasses both the $B\to K^*$ and the  $B\to\r$ transitions. For

287: $B\to K^*$ we also use results obtained by QCD sum rules

288: \cite{Colangelo:1996jv}. These results are summarized in section

289: \ref{sec:0}.

290:  In section \ref{sec:1} we discuss

291: rescattering effects as parametrized by the Regge model.  We

292: include in the parametrization the Pomeron and charmed Regge

293: trajectories and we give estimates of the Regge residues. In

294: section \ref{sec:2} we present our numerical results and discuss

295: them. The basic result we find is that, due to the presence of

296: helicity-flip Regge residues, longitudinal polarization fractions

297: are in general reduced, except for  the $B\to\r\r$ decay modes.

298: Differences between the channels $B\to K^*\phi$ and $B\to K^*\r$

299: might depend on the choice of the form factors, therefore we

300: discuss differences induced by this choice as well. We argue that

301: for the $B\to K^*\phi$ decay mode an interplay between form

302: factors and Regge parameters can produce theoretical results

303: compatible with the data with no need to introduce new physics

304: effects. For the $B\to K^*\r$ the results are more elusive and

305: more work is needed especially for the determination of the

306: semileptonic form factors. Finally in the Appendix we give an

307: outline of the model used to compute the parameters of the charmed

308: Regge trajectories.

309: \section{Bare amplitudes \label{sec:0}}

310: In this section we introduce bare amplitudes, i.e.

311: matrix elements of the weak hamiltonian with no final state interaction effects.

312: We consider two sets of weak bare amplitudes. The first set includes

313: the bare amplitudes for the $B$ decays into two light vector mesons.

314:  We consider the channels $B\to \r\r$, $B\to K^*\r$, $B\to K^{*}

315:  \phi$. Since this calculation is straightforward

316:  we limit our presentation to the numerical results.

317: 

318: For the relevant Wilson coefficients we use

319:   $a_1=1.05$, $a_2=0.053$,  and

320:  $(a_3,a_4,a_5,a_7,a_9,a_{10})$ $=(48,-439-77i,-45,+0.5-1.3i,-94-1.3i,-14-0.4i)$

321:   $\times 10^{-4}$ \cite{Ali:1998eb}. We use  two different choices of form factors

322:  for the weak transition $B\to V$. The first choice is based on Light Cone QCD sum rules

323:  \cite{Ball:2003rd}. The corresponding results for the bare amplitudes for $B\to K^*\r$

324:  and $B\to \r\r$

325:  are reported in  Table \ref{table1bis}. We also consider the QCD Sum Rules results of Ref.

326:  \cite{Colangelo:1996jv}. Since in this paper only the $b\to s$ transitions

327:  are considered, these results

328:  can be employed only for $B\to K^*\phi$ decays. The numerical results obtained

329:  by the two sets of form factors for the  $B\to K^*\phi$ decay channel are reported in Table

330:  \ref{table1ter}.

331: \begin{table}[ht]

332: \begin{center}

333: \begin{tabular}{||c|c|c||c|c|c||}

334: \hline

335: Process & $\lambda$& $A_{b,\lambda}$

336: &Process& $\lambda$&$A_{b,\lambda}$\\

337: \hline

338: $B^+\to K^{*0}\rho^+$&0&$-0.62+3.14\,i$ &

339: $B^+\to \r^{0}\rho^+$&0&$+4.21-2.60\,i$

340: \\

341: \hline

342: $B^+\to K^{*0}\rho^+$&+&$+0.17-0.86\,i$&

343: $B^+\to \r^{0}\rho^+$&+&$-1.01+0.62\,i$\\

344: \hline

345: $B^+\to K^{*0}\rho^+$&-&$-0.02\,i$ &

346: $B^+\to \r^{0}\rho^+$&-&$-0.01+0.01\,i$\\

347: \hline

348: $B^+\to K^{*+}\rho^0$&0&$+0.54+2.64\,i$&

349: $B^0\to \r^{+}\rho^-$&0&$+5.60-3.93\,i$\\

350: \hline

351: $B^+\to K^{*+}\rho^0$ &$+$&$-0.15-0.70\,i$

352: & $B^0\to \r^{+}\rho^-$&+&$-1.34+0.94\,i$\\

353: \hline

354: $B^+\to K^{*+}\rho^0$&-&$-0.01\,i$&

355: $B^0\to \r^{+}\rho^-$&-&$-0.02+0.01\,i$\\

356: \hline

357: \end{tabular}

358: \end{center}\caption{

359: Bare helicity amplitudes for

360: $B\to K^*\r$ and $B\to \r\r$ with form factors from Ref. \cite{Ball:2003rd}.

361: Results in $10^{-8}$ GeV.

362: \label{table1bis}}

363: \end{table}

364: 

365: 

366: 

367: \begin{table}[ht]

368: \begin{center}

369: \begin{tabular}{||c|c|c||c||}

370: \hline Process & $\lambda$ &

371: $A_{b,\lambda}$(Ref. \cite{Ball:2003rd}) &

372: $A_{b,\lambda}$(Ref. \cite{Colangelo:1996jv})\\

373: \hline

374: $B\to K^*\phi$  & $0$ & $-0.83+3.85\,i$ &  $-0.48+2.24\,i$\\

375: \hline

376: $B\to K^*\phi$  & $+$ & $+0.27-1.25\,i$ &  $+0.28-1.30\,i$\\

377: \hline

378: $B\to K^*\phi$  & $-$ & $-0.03\,i$      &  $+0.01-0.06\,i$\\

379: \hline

380: \end{tabular}

381: \end{center}\caption{

382: Bare helicity amplitudes for $B\to K^*\phi$ with two different choices

383: of form factors \cite{Ball:2003rd} and \cite{Colangelo:1996jv}.

384: Results in $10^{-8}$ GeV.

385: \label{table1ter}}

386: \end{table}

387: We note the numerical relevance of

388: $A_0$ and  the suppression of $A_-$ in comparison with $A_+$. We

389:  also note numerical differences

390: in Table \ref{table1ter} arising from the different choice of form factors.

391: 

392: The second set of amplitudes includes the bare decays of $B^+$ into two charmed mesons,

393: $D_s^{(*)+}\, {\bar D}^{(*)0}$. For the $B\to\r\r$ decay channel we consider

394: $B$ decay into two charmed non strange mesons. The matrix elements

395: of this second set are computed in the factorization

396: approximation, using the Heavy Quark Effective Theory at leading

397: order, with the following  Isgur-Wise function

398: $\xi(\omega)=4/(1+\omega)^2$. For the CKM matrix elements we use

399: PDG data \cite{Eidelman:2004wy}.

400: 

401: Numerical results for the  bare helicity amplitudes

402: $A_{b,\lambda}$ for the channels $B^+\to D_s^{(*)+} {\bar

403: D}^{(*)0}$

404:  are reported in Table  \ref{table1} .

405: 

406: 

407: 

408: \begin{table}[!!!!h]

409: \begin{center}

410: \begin{tabular}{||c|c|c||}

411: \hline

412: Process&$\lambda$&$A^{(k)}_{b,\lambda}$\

413: \\

414: \hline

415: $B^+\to D^+_s\bar D^{0}$&0&$+1.32\,i$

416: \\

417: \hline

418: $B^+\to D^+_s\bar D^{*0}$&0 & $-1.15\,i$\\

419: \hline

420: $B^+\to D_s^{*+}\bar D^{0}$&0 &$-1.15\,i$\\

421: \hline

422: $B^+\to D_s^{*+}\bar D^{*0}$&0 &$+1.32\,i$\\

423: \hline $B^+\to D_s^{*+}\bar D^{*0}$&$+$&$-1.02\,i$\\

424: \hline $B^+\to D_s^{*+}\bar D^{*0}$ &$-$

425: & $-0.42\,i$ \\

426: \hline

427: \end{tabular}

428: \end{center}\caption{

429: Bare helicity amplitudes for $B$ decays into charmed mesons. The four channels

430: $B^+\to D_s^{+}\bar D^{0}$, $B^+\to D_s^{*+}\bar D^{0}$, $B^+\to D_s^{*+}\bar D^{0}$, and

431:   $B^+\to D_s^{*+}\bar D^{*0}$ are

432: identified respectively by the index $k=1,2,3,4$. Units are $10^{-6}$ GeV.

433: \label{table1}}

434: \end{table}

435: 

436: One can immediately notice the CKM enhancement (typically by a

437: factor $10^2$) of the amplitudes in Table \ref{table1} in

438: comparison with those of  Table \ref{table1bis}.  For $B$ decay

439: into two non-strange charmed mesons the enhancement is smaller.

440: 

441: \section{Final state interactions and Regge behavior \label{sec:1}}

442:  Corrections to the bare amplitudes due to final state interactions are taken into

443: account by means of the Watson's theorem \cite{Watson:1952ji}:

444: \be

445: A = \sqrt{S}A_{b}

446: \label{amp}

447: \ee

448: where $S$ is the

449: $S$-matrix, $A_{b}$ and $A$ are the bare and the full amplitudes.

450: 

451: The two-body $S$-matrix elements are given by

452: \starteq

453: S^{(I)}_{ij} = \d_{ij} + 2i\sqrt{\r_{i}\r_{j}}\

454: T^{(I)}_{ij}(s)~~,

455: \label{s-matrix}

456: \fineq

457: \par\noindent

458: where $i,j$ run over all the channels involved in the final state interactions.

459: 

460: The $J=0$, isospin $I$ amplitude $

461: T^{(I)}_{ij}(s)$ is obtained by projecting the $J=0$ angular

462: momentum out of the amplitude $T^{(I)}_{ij}(s,t)$:

463: \starteq

464: T^{(I)}_{ij}(s) =  {1 \over 16\p}{s \over \sqrt{\ell_{i}\ell_{j}}}

465: \int^{t_{-}}_{t_{+}}dt \,T^{(I)}_{ij}(s,t) ~~.

466: \label{S-wave}

467: \fineq

468: The momenta $\r_j\,,\ell_j$ are defined below.

469: For the channel $B\to K^*\phi$ we only have the $I=1/2$ transition amplitudes, for

470: $B\to K^*\r$ both $I=1/2$ and $I=3/2$ are involved. For $B\to \r\r$ we can have $I=0$

471: and $I=2$.

472: 

473: As discussed in the Introduction, given the rather large value of

474: $s=m^2_B$, a Regge approximation for the transition amplitude

475: seems adequate. We will therefore include first of all the Pomeron

476: term, which contributes to the elastic channels. For the inelastic

477: channels we include only channels whose bare amplitudes are

478: prominent. For $B\to K^*\phi$ and $B\to K^*\r$ they should be  the

479: amplitudes for the transition $B\to D^{(*)} D_s^{(*)}$ since these

480: channels are Cabibbo enhanced, see table \ref{table1}. For the

481: $I=0$ amplitude of the decay $B\to\r^+\r^-$ there is no Cabibbo

482: enhancement; most probably other trajectories should be

483: considered, but we consider here again charmed trajectories mainly

484:  for the sake of comparison.   The

485: transition from a state with two charmed mesons to a state with

486: two light particles can only occur {\it via} charmed Regge poles.

487: Therefore in the present approximation we will include, besides

488: the Pomeron, only the charmed Regge trajectories.

489: 

490: \subsection{Pomeron contribution}

491: To begin with we  consider the Pomeron contribution. We write

492: \be

493: S={\cal P}

494: \label{P0}

495: \ee

496: neglecting for the time being

497: non leading Regge trajectories and inelastic terms. We have

498: \be

499: {\cal P}=1+2iT^{{\cal

500: P}}(s)\,,\hskip1cm T^{{\cal P}}(s)=\frac{1}{16\p s

501: }\int_{-s+4m^2_V}^{0}T^{{\cal P}}(s,t) dt\, .

502: \label{p1}

503: \ee

504: The following parametrization can be used

505: \cite{Nardulli:1997ht}, \cite{Donnachie:1992ny}:

506: \starteq

507: T^{{\cal P}}(s,t)=~-~ \b^{P}g(t) \left({s \over s_{0}}

508: \right)^{\a_{P(t)}} e^{ -i{\p \over 2}\a_{P}(t)}~~,

509: \label{P}

510: \fineq

511: with $s_{0} = 1\,\rm GeV^{2}$ and

512: \starteq

513: \a_{P}(t) = 1.08

514: + 0.25t   \qquad \qquad (t {\rm \ in \ GeV^{2}})~~,

515: \label{alphaP}

516: \fineq

517: as given by fits to hadron-hadron scattering total cross

518: sections. The product $\b^{P}\cdot

519: g(t) = \b^{\cal P}(t)$ represents the Pomeron residue; for the

520: $t$-dependence we assume \cite{Nardulli:1997ht}, \cite{Donnachie:1992ny}

521: \starteq

522: g(t) = {1 \over (1-

523: t/m_{\r}^{2})^{2}} \simeq e^{2.8t}~~.

524: \label{gt}

525: \fineq

526: To get $\b^{\cal P}$ we use factorization and

527: the additive

528: quark counting rule.

529: For the $B\to K^*\rho$ channel residue factorization gives

530:  \be

531:  \b^{\cal P}_{K^*\rho}=\b^{\cal P}_\r\b^{\cal

532: P}_{K^*}\ .

533: \label{fa}

534: \ee

535: The two residues appearing in (\ref{fa}) can be computed by the additive

536: quark counting rule. This gives \cite{Nardulli:1997ht}

537: \bea

538: \beta^{\cal P}_{\r}&\sim& \beta^{\cal P}_{\p}=2\b^{\cal P}_{u}\, ,\nonumber\\

539: \beta^{\cal P}_{K^*}&\sim& \beta^{\cal P}_{K}=\b^{\cal P}_{u}+

540: \b^{\cal P}_{s}\, .

541: \eea

542: The basis of the additive quark counting rule is given

543:  by pion-proton and proton-proton high energy scattering data.

544: From these data  \cite{Nardulli:1997ht},\cite{Zheng:1995mg}:

545: $\b^{\cal P}_{\p}\sim 2/3 \b^{\cal P}_{p}\sim 5.1$.

546: From $Kp$ high energy scattering data, one finds

547: $\b^{\cal P}_{s}\sim 2/3\b^{\cal P}_{u}$. Therefore

548: \be

549: \b^{\cal P}_{K^*\rho}\approx 22\ .

550: \ee

551: For the $B\to K^*\phi$ channel we have, instead of

552:  Eq. (\ref{fa}), $ \b^{\cal P}_{K^*\phi}=

553: \b^{\cal P}_\phi\b^{\cal

554: P}_{K^*}\ , \label{fa1}$ and, numerically,

555: $ \b^{\cal P}_{K^*\phi}\approx 14\ .

556:  \label{fa10}$

557: Finally for the $B\to \rho\rho$ channel we get

558: $ \b^{\cal P}_{\r\rho}\approx 26\ .\label{fa2}$

559: 

560: Using (\ref{p1}) in the approximation (\ref{P0}) violates unitarity.

561: We can observe from this that inelasticity effects are important in the determination

562: of the FSI phases \cite{Donoghue:1996hz}. As matter of fact, let us

563:  parametrize them as in Ref. \cite{Donoghue:1996hz} by one effective state,

564:  with no extra phases. Then

565: the $S$-matrix should be written as follows (neglecting a

566: small phase $\varphi=\,-0.01$ in $\sqrt{\cal P}$):

567: \be (B\to K^*\r)\hskip2cm S\approx\left(%

568: \begin{array}{cc}

569:   0.64 & 0.77 i \\

570:  0.77 i & 0.64 \\

571: \end{array}%

572: \right) \ ,\hskip1cm

573:  \sqrt{S}\approx\left(%

574: \begin{array}{cc}

575:   0.80& 0.62(1+ i) \\

576:  0.62(1+ i) & 0.80 \\

577: \end{array}%

578: \right) \ ;\ee

579: 

580: \be (B\to K^*\phi)\hskip2cm S\approx\left(%

581: \begin{array}{cc}

582:   0.77& 0.64\,i \\

583:  0.64\, i & 0.77 \\

584: \end{array}%

585: \right)\ , \hskip1.0cm \sqrt S\approx\left(%

586: \begin{array}{cc}

587:   0.88& 0.57(1+i) \\

588:   0.57(1+i) & 0.88 \\

589: \end{array}%

590: \right) \ ;\ee

591: \be (B\to \r\r)\hskip2cm  S\approx\left(%

592: \begin{array}{cc}

593:   0.58& 0.82\,i \\

594:  0.82\, i & 0.58 \\

595: \end{array}%

596: \right)\ , \hskip1.0cm \sqrt S\approx\left(%

597: \begin{array}{cc}

598:   0.76& 0.64(1+i) \\

599:   0.64(1+i) & 0.76 \\

600: \end{array}%

601: \right) \ .\ee

602: This shows that even neglecting the effect of

603: the non leading Regge trajectories, final state interactions

604:  due to inelastic effects parameterized by the Pomeron

605:  exchange can produce sizeable strong phases.

606:  This result agrees with the analogous findings of Refs.

607:  \cite{Nardulli:1997ht}  and \cite{Donoghue:1996hz}.

608: 

609: \subsection{Regge trajectories\label{reg}}

610: Let us now consider the contribution of the charmed Regge

611: trajectories. They are present in the $B\to K^*\phi$ amplitude, in

612: the

613:  $I=1/2$  amplitude  for the decay $B\to K^*\r$ and in $I=0$ amplitude

614:   for $B\to\r^+\r^-$.

615:  They do not contribute to the $B^+\to \r^+\r^0$ decay channel

616: because the final state has $I=2$ and the two charmed mesons can

617: only have either $I=0$ or $I=1$.

618: 

619: 

620: Including charmed Regge trajectories the  $S$ matrix can be

621: written for the generic  $B\to VV$ case as follows

622: \be

623: S={\cal

624: P}+{\cal D}+{\cal D}^*\, .

625: \label{regg}

626: \ee

627: Here ${\cal P}$ is the

628: Pomeron contribution discussed above; we note that ${\cal

629: P}_{ij}={\cal P}_{i}\delta_{ij}$. $ {\cal D}$ and ${\cal D}^*$ are

630: reggeized amplitudes corresponding to the $ {D}$ and ${D}^*$ Regge

631: trajectories. For $B\to K^*\phi$, $K^*\r$ decays they connect the

632: state $| K^* V\rangle$ to the other states $| D_s D\rangle$, $|

633: D_s D^*\rangle$, $| D^*_s D\rangle$, $| D^*_s D^*\rangle$. For

634: $B^0\to \r^+\r^-$ they connect the state $|\r^+\r^-\rangle$ to the

635: states $| D^{+} D^{-}\rangle$, $| D^{*+} D^{-}\rangle$, $| D^{+}

636: D^{*-}\rangle$, $| D^{*+} D^{*-}\rangle$.  Since all the Regge

637: contributions are exponentially suppressed one  can make the

638: approximation

639: \be

640: \sqrt{S}\approx \sqrt{\cal P}+\frac 1 2 {\cal

641: P}^{-1/2}\left({\cal D}+{\cal D}^*\right)\, .

642: \ee

643: 

644: We discuss explicitly the case $B\to K^* \phi$, the treatment of

645: the other cases  being similar. Neglecting other inelastic effects

646: we consider a $5\times 5$ $S_{ij}$ matrix, where $i=1,2,3,4,5$

647: represent respectively the states $| D_s D\rangle$, $| D_s

648: D^*\rangle$, $| D^*_s D\rangle$, $| D^*_s D^*\rangle$, $| K^* \phi

649: \rangle$. The kinematic factors are as follows:

650: $\ell_j=s(s-4m^2_D)\ ,j=1,\cdots,4$, $\ell_5=s^2$,

651: $\r_j=\sqrt{(s-4m^2_D)/s}\ ,j=1,\cdots,4$, $\r_5=1$. We work in

652: the approximation $m^2_D=m^2_{D_s}=m^2_{D^*}=m^2_{D^*_s}$ and

653: $m_{K^*}\,=\,m_\r=m_\phi\,=\,m_V\sim 0.9$ GeV with $s=m^2_B$. The integration limits

654: are for the elastic contribution: $t_-=0$ and $t_+=-s+4m_V^2$;  for

655: the inelastic contributions: $t_\mp=-s/2+m^2_V+m^2_D\pm s/2\sqrt{1-4m_V^2/s}

656: \sqrt{1-4m_D^2/s}$. Therefore

657: \be

658: A(B \rarr K^{*}\phi)_{\lambda}=

659: \sqrt{{\cal P}}A^{(5)}_{b\,,\lambda}\ +\

660: \sum_{k=1}^4\,\sum_{\lambda_i}\,\frac 1 2 {\cal

661: P}^{-1/2}\left({\cal D}\ +\ {\cal

662: D}^*\right)^{(k)}_{\lambda_i\,,\lambda} A_{b\,,\lambda_i}^{(k)}\, .

663: \ee

664: Let us consider the Regge amplitudes $R={\cal D}\ , {\cal D}^*$:

665: \be

666: R^{(k)}(s)=\frac{2i}{16\p}\sqrt[4]

667: {\frac{s-4m^2_D}{s}}\frac{1}{\sqrt{s(s-4m^2_D)}} \int_{t_+}^{t_-}

668: dt R^{(k)}(s,t)\, ,

669: \ee

670: with $ t_{\pm}$ given above. We assume the general parametrization

671: \starteq

672: R^{(k)}(s,t)= -\b^{R} {1 + (-)^{s_{R}}

673: e^{ -i\p\a_{R}(t)} \over 2}\G(l_{R} - \a_{R}(t))(\a ')^{1 - l_{R}}

674: (\a 's)^{\a _{R}(t)}

675: \label{R}

676: \fineq

677: as suggested in \cite{Irving:1977ea}. We notice the Regge poles

678: at $l_{R} - \a_{R}(t)=0,-1,-2,\cdots$; for

679: $R={\cal D}$ we have $s_R=\ell_R=0$ and,

680:  for $R= {\cal D}^*$,  $s_R=\ell_R=1$.

681: Near $t = m_{R}^{2}$, (\ref{R}) reduces to \starteq R \approx

682: \b^{R}{s^{s_{R}} \over ( t-m_{R}^{2} )} \label{Rapprox} \fineq

683: which allows to identify $\b^{R} $ as the product of two on-shell

684: coupling constants. Let us write \be \beta^R=\beta_{D_s^{(*)}

685: K^*}\beta_{D^{(*)} V}\ .\ee Using the effective lagrangian

686: approach \cite{Casalbuoni:1996pg} we can compute the residues by

687: identifying them with the coupling constants for $t\sim m^2_R$.

688: The numerical results are in

689:  Table \ref{table2}; they are obtained

690:  by the following values of the constants defined in \cite{Casalbuoni:1996pg}:

691:  $g_V=5.8$, $\l=0.56$

692: GeV$^{-1}$, $\b=0.9$, which represents an updated fit, see \cite{Isola:2003fh}

693: for details.

694: 

695: \vskip0.5cm

696: \begin{table}[ht]

697: \begin{center}

698: \begin {minipage}{6.5in}

699: \begin{tabular}{||c|c|c||c|c||}

700:   % after \\: \hline or \cline{col1-col2} \cline{col3-col4} ...

701: \hline

702: Residue &  ${\cal D}$ & \rm{ Num. values} & ${\cal D}^*$&  \rm{ Num. values}\\

703: \hline

704: $\beta_{DV}^{0}$ & $ \beta\, g_V\,m_V/\sqrt 2$&  +2.84& 0 &0\\

705: \hline

706: $\beta_{DV}^{\pm}$\ &\ $\beta\, g_V\,m_D$ &$+9.76$&

707: $\sqrt 2 \lambda g_V\,m_D $ &$+8.59$\\

708: \hline

709: $\beta_{D^*V}^{0,0}$  &   $0$& $0

710: $ &$\displaystyle\frac{m_V\,g_V}{2\,m_D}(\beta-4\lambda m_D)

711: $&$-3.93$ \\

712: \hline

713: $\beta_{D^*V}^{\pm,\pm}$ &

714: $\sqrt{2}\,\lambda \,g_V\, m^2_D$ &  +16.1& $2\,\lambda\,g_V\, m_D

715: $ &$+12.1$\\

716: \hline

717: $\displaystyle\beta_{D^*V}^{0,\pm}$  &

718: $2\,\lambda\, g_V\,m^2_D  $ & +22.7 &

719: $\displaystyle \frac{g_V}{\sqrt 2}

720: \left(\beta-2\lambda m_D\right)

721: $ &$-4.90$ \\

722: \hline

723: $\beta_{D^*V}^{\pm,0}$  &

724: $ 2\,\lambda\, g_V\,m_V\,m_D  $& +9.35 &$0

725: $& 0\\

726:  \hline

727: $\beta_{D^*V}^{\pm,\mp}$  &  $\sqrt{2}\,\lambda \,g_V\, m^2_D$

728: &+16.1&$0  $&0 \\

729:   \hline

730: \end{tabular}

731: \end {minipage}

732: \end{center}

733: \caption{

734: Regge residues $\b^{\l_i,\l}$ of the poles  ${\cal D}^*$ and

735: ${\cal D}$; $\l_i$ is the polarization of the charmed meson, $\l$

736: that of the light vector meson. The numerical results are obtained

737: with $g_V=5.8$, $\l=0.56$ GeV$^{-1}$, $\b=0.9$, see

738: \cite{Isola:2003fh}. $SU(3)_f$ nonet symmetry has been assumed for

739: the evaluation of the residues.\label{table2}}

740: \end{table}

741: 

742: It is important to stress that, differently form the Pomeron,

743:  which is mainly helicity conserving,

744:  Regge charmed trajectories have several helicity-flip residues.

745:   In particular they can change longitudinal

746:  into transverse polarizations, see e.g. $\displaystyle\beta_{D^*V}^{0,\pm}$. In the next section

747:  we will discuss these effects in the context of $B\to VV $ decays.

748: 

749: Next step is the calculation of the trajectories $\alpha_{D}(t)$

750: and $\alpha_{D^*}(t)$. To do this we use the model described in

751: the Appendix. For charmed mesons such as $D$ and $D^*$ (we do not

752: distinguish between $D$ and $D_s$)  the Regge trajectories are not

753: linear in the whole $t$ range. Due to the exponential suppression

754: of the high $|t|$ range, only  the small $t$ region is of interest

755: and here we can approximately write\be

756: \alpha_{D}=s_D+\alpha_0+\alpha^\prime t\,,\hskip1cm

757: \alpha_{D^*}=s_{D^*}+\alpha_0+\alpha^\prime t\ ,\ee with $s_D=0$,

758: $s_{D^*}=1$. From the results of the Appendix we have

759: $\alpha_0=-1.8$. As to the value of the slope $ \alpha^\prime$,

760: linearizing the trajectory as in the dashed-dotted line of Fig.

761: \ref{fig:2} (right side)

762:  one gets $\alpha^\prime=0.33$ GeV$^{-2}$.

763: This formula overestimates the mass of the $D,\,D^*$ system by 15\%.

764: Fitting these masses would give the value $\alpha^\prime=0.45$ GeV$^{-2}$;

765: from these results one can estimate a range 0.33-0.45 GeV$^{-2}$ for  $\alpha^\prime$.

766: In a conservative vein we double this range

767: to take into account other  theoretical uncertainties (arising from the model

768: adopted in the Appendix)

769: and use therefore

770: \be\alpha_0=-1.8\,,\hskip1cm

771: \alpha^\prime\,=\,(0.39\,\pm \,0.12) \ {\rm GeV} ^{-2}\ . \label{eq:27}\ee

772: 

773: \section{Numerical results and discussion\label{sec:2}}

774: In the present approach one can identify two kinds of theoretical

775: uncertainties, one related to the choice of the semileptonic $B\to

776: V$ form factors, the other to the parameters of the Regge

777: amplitudes. To start with we fix the form factors, by choosing the

778: numerical results  obtained by the Light Cone sum rules approach

779: \cite{Ball:2003rd}. The bare amplitudes can be found in Tables

780: \ref{table1bis} and \ref{table1ter}; for the Regge parameters we

781: use eq. (\ref{eq:27}), limiting the analysis to the central value

782: and the two extremes of the $\a^\prime$ range. The numerical

783: results are presented in Table \ref{Ball1}.

784: 

785: \begin{table}[ht]

786: \begin{center}

787: \begin{tabular}{|c|c|c|c|c|}

788: \hline

789: Process   & $\Gamma_L/\Gamma$&  $\Gamma_\perp/\Gamma$ &

790: $\Gamma_\parallel/\Gamma$  & B.R.\, $\times 10^6$  \\

791: \hline

792: $B\to K^*\phi $    &

793: $\begin{array}{c} 0.31\\0.73\\0.87\end{array}$  &

794: $\begin{array}{c} 0.1\\0.06\\0.05\end{array}$  &

795: $\begin{array}{c} 0.59\\0.20\\0.08\end{array}$  &

796: $\begin{array}{c} 35\\15\\12\end{array}$\\

797: \hline

798: $B\to K^{*+}\r^0$  &

799: $\begin{array}{c} 0.26\\0.72\\0.89\end{array}$  &

800: $\begin{array}{c} 0.12\\0.08\\0.06\end{array}$  &

801: $\begin{array}{c} 0.61\\0.20\\0.05\end{array}$  &

802: $\begin{array}{c} 19\\6.2\\4.9\end{array}$\\

803: \hline

804: $B\to K^{*0}\r^+$  &

805: $\begin{array}{c} 0.18\\0.61\\0.84\end{array}$  &

806: $\begin{array}{c} 0.11\\0.07\\0.05\end{array}$  &

807: $\begin{array}{c} 0.71\\0.32\\0.11\end{array}$  &

808: $\begin{array}{c} 35\\9.9\\7.2\end{array}$\\

809: \hline

810: $B^+\to\r^+\r^0 $ &

811: $0.95$  &

812: $0.03$  &

813: $0.03$  &

814: $14$\\

815: \hline

816: $B^0\to\r^+\r^- $ &

817: $\begin{array}{c} 0.86\\0.93\\0.94\end{array}$  &

818: $\begin{array}{c} 0.01\\0.02\\0.02\end{array}$  &

819: $\begin{array}{c} 0.13\\0.05\\0.04\end{array}$  &

820: $\begin{array}{c} 28\\26\\26\end{array}$\\

821:  \hline

822: \end{tabular}

823: \end{center}\caption{

824: Results for the various $B$ decay channels obtained with the form factors as in

825: \cite{Ball:2003rd}. For each decay channel the Regge slope of the charmed trajectories

826: is, from top to bottom, $\alpha^\prime$ = (0.27,0.39,0.51) GeV$^{-2}$.

827: There is no Regge contribution to $B^+\to\r^+\r^0 $ so that the only FSI effect here

828: is the elastic one. All the Branching Ratios are evaluated using $\tau_B\ =\ 1.67$ ps.\label{Ball1}}

829: \end{table}

830: 

831: We note that for $B\to K^*V$ decays the smallest value of

832: $\a^\prime$ corresponds to the largest contribution from the Regge

833: trajectories. The Table \ref{Ball1} shows that the

834:  main effect of the Regge poles for the channels with a $K^*$ in the final state

835:  is the reduction

836: of the longitudinal polarization fraction and the increase of the

837: other fractions. This is due to the effect mentioned in subsection

838: \ref{reg}, i.e. the existence of helicity flip residues changing

839: longitudinal into transverse polarizations. On the other hand the

840: transverse polarization fraction $\G_\perp/\G$ cannot increase

841: beyond 10\%-12\%.

842: 

843: A more detailed numerical analysis for the channel $B\to

844: K^{*+}\r^0$ can be found in Fig. \ref{figkstar} that shows the

845: variation of the branching ratio and the polarization fractions

846: versus the slope $\a^\prime$. A detailed analysis of the $K^*\phi$

847: channel will be presented  below. Finally we observe that, as

848: expected, the $\r\r$ channels get negligible or vanishing

849: contribution from the Regge poles. For the $\r^+\r^-$ channel this

850: is due to the absence of the Cabibbo enhancement; for the

851: $\r^+\r^0$ decay mode there is no Regge contribution because the

852: final state  is a pure $I=2$ state. If the explanation presented

853: here is valid, the large fraction $\G_L/\G$ for $B\to\r\r$ (see

854: Table \ref{table0}) is a consequence of the negligible role of the

855: charmed Regge poles.

856: \begin{figure}[ht]

857: \centerline{

858: \begin{tabular}{cc}

859: \epsfig{file=BrKpiustarro0.eps,width=8truecm } &

860: \epsfig{file=GLKpiustarro0Ball.eps,width=8truecm  }

861: \end{tabular}}

862: \caption{Comparison between theoretical predictions based on the

863: form factors of Ref. \cite{Ball:2003rd} and the experimental data

864: for the channel $B^+\to K^{+*}\rho^0$.  On the left: The

865: $B.R.\times 10^6$ as a function of the parameter $\alpha^\prime$

866: giving the slope of the charmed Regge trajectories. Continuous

867: straight lines give the experimental interval (see Table

868: \ref{table0}). On the right: Continuous lines give the fraction

869: $\G_L/\G$, the dashed line is the fraction $\G_\perp/\G$ and the

870: dotted line the fraction $\G_\parallel/\G$. Units of

871: $\alpha^\prime$ are GeV$^{-2}$. We use $\tau_B\ =\ 1.67$ ps. \label{figkstar}}

872: \end{figure}

873: 

874: 

875: Let us comment on the discrepancies between the theoretical

876: results of Table \ref{Ball1}, Fig. \ref{figkstar} and experimental

877: data. They might point to new physics, but a simpler possibility

878: is an interplay between final state interactions and form factors.

879: An example is offered by the channel $K^*\phi$. Here to get a

880: small longitudinal polarization fraction one would prefer a small

881: value of $\a^\prime$. Small values of $\a^\prime$, however,

882: produce  values for the branching ratio much higher than the data.

883: As discussed in the Introduction, the dominance of $\G_L/\G$ is a

884: consequence of the chiral structure of the currents in the

885: standard model and the $m_b\to\infty$ limit. The only way to

886: reduce it without invoking new physics is to show the existence of

887: soft effects in the standard model. Our analysis shows that these

888: effects can be generated by rescattering effects parametrized  by

889: the Regge model.  On the other hand the high value of the

890: branching ratio might be due to the choice  of the form factors of

891: Ref. \cite{Ball:2003rd}. To deal with this problem we consider a

892: different set of form factors. Since in $B\to VV$ decays only

893: small $q^2$ are involved, the Lattice QCD determinations are

894: unreliable because of the large extrapolations they would need.

895: The best alternative to Light Cone sum rules are the traditional

896: QCD sum rules. In \cite{Colangelo:1996jv} such a calculation is

897: performed; we have reported these predictions in Table

898: \ref{table1bis}. The authors in \cite{Colangelo:1996jv} only

899: consider the $b\to s$ transition, and therefore their results can

900: be only applied to the channel $B\to \phi K^*$ (in \cite{Ball:1997rj}

901: also the transition $b\to u$ was considered, but with

902: a set of parameters different from \cite{Colangelo:1996jv}, which

903: renders the comparison difficult). In figures \ref{col1} and \ref{col2}

904: we report the branching ratio and  the fraction

905: $\G_L/\G$ as functions of the slope $\a^\prime$ of the charmed Regge

906: trajectories for the parametrizations \cite{Colangelo:1996jv} and \cite{Ball:2003rd}.

907: As to the other  polarization fractions , $\G_\parallel/\G$ and

908: $\G_\perp/\G$ for the form factors of Ref. \cite{Ball:2003rd} can

909: be found in Table \ref{Ball1}. The analogous figures for the form

910: factors of  \cite{Colangelo:1996jv} are

911: $\G_\perp/\G=(0.12,\,0.12,\,0.12)$ and

912: $\G_\parallel/\G=(0.74,\,0.40,\,0.20)$ for

913: $\a^\prime=(0.27,\,0.39,\,0.51)$ GeV$^{-2}$.

914: 

915: \begin{figure}[ht]

916: \centerline{ \epsfig{file=BrKstarphiConfronto.eps,width=8truecm }

917: } \caption{ $BR(B\to K^*\phi)\times 10^6$ as a function of the

918: parameter $\alpha^\prime$ giving the slope of the charmed Regge

919: trajectories. Continuous straight lines give the experimental

920: interval (see Table \ref{table0}). Dashed lines are theoretical

921: predictions based on the form factors of ref.\cite{Ball:2003rd};

922: dotted lines are based on the form factors of ref.

923: \cite{Colangelo:1996jv}. Units of $\alpha^\prime$ are GeV$^{-2}$.

924: \label{col1}}

925: \end{figure}

926: 

927: \begin{figure}[ht]

928: \centerline{ \epsfig{file=GLKstarphiConfronto.eps,width=8truecm }

929: } \caption{ $\G_L/\G (B\to K^*\phi)$ as a function of the

930: parameter $\alpha^\prime$ giving the slope of the charmed Regge

931: trajectories. Continuous straight lines give the experimental

932: interval (see Table \ref{table0}). Dashed lines are theoretical

933: predictions based on the form factors of ref.\cite{Ball:2003rd};

934: dotted lines are based on the form factors of ref.

935: \cite{Colangelo:1996jv}. Units of $\alpha^\prime$ are

936: GeV$^{-2}$. We use $\tau_B\ =\ 1.67$ ps. \label{col2}}

937: \end{figure}

938: 

939: Fig. \ref{col1} shows that much smaller values of the branching

940: ratio are  obtained by the QCD sum rules form factors  of Ref.

941: \cite{Colangelo:1996jv} (dotted lines) for $\a^\prime$ around its

942: central value; Fig \ref{col2} shows that for the same values of

943: $\a^\prime$ satisfactory results for  $\G_L/\G$ are obtained.

944: Finally $\G_\perp/\G$ gets a value smaller, but still compatible

945: within 2$\s$, with the BaBar result (\ref{babarr}).

946: 

947: These results show that data on the polarization fractions

948:  for the $B\to K^*\phi$ decay

949: channel might be explained as the effect of two converging

950: factors: an appropriate set of form factors

951: \cite{Colangelo:1996jv} and the presence of final state

952: interactions computed by the Regge approach.  It remains to be

953: seen however how QCD sum rules and Regge phenomenology would work

954: for the other channels $B\to K^*\r$ and $B\to\r\r$, an issue we

955: plan to deal with in the future.

956: 

957: 

958: \section{Appendix: a model for Regge trajectories \label{app:0}}

959: 

960:  For non charmed Regge poles,

961: $\a_{R}(t)$ is phenomenologically

962: given by

963: \starteq

964:  \a_{R}(t) = s_R + \a '(t -m^2_R)

965: \label{alphaR}

966: \fineq with the

967:  universal slope: $\a '= 0.93 \, \rm GeV^{-2}$.

968:  This behavior can be

969:   reproduced by a potential model, which  gives the Regge formula

970:   the meaning of an exchange of infinitely

971:   many particles. To compute their  spectrum one solves a Schroedinger equation

972:   and one shows that the slope is related to the string tension

973:   responsible of the confining part of the potential.

974: 

975: To be more specific we consider a bound state comprising a quark and an antiquark.

976: For the bound state problem we consider the Salpeter

977: equation (Schr\"odinger equation with relativistic kinematics):

978: \be

979: \left(\sqrt{{\bf p_1}^2+m_1^2}+\sqrt{{\bf p_2}^2+m_2^2}+V(r)\right)\psi({\bf r })

980: = E\psi({\bf r})

981: \ee

982: where $m_j$ are quark masses, $V(r)$ is the potential

983: energy  and $r=|{\bf r_1}-{\bf r_2}|$.

984: In the meson rest frame $E$ is the meson mass $M$ and

985: ${\bf p_1}^2={\bf p_2}^2=-\hbar^2\nabla^2$.

986: The bound state equation can be solved for arbitrarily large

987: quantum numbers $n$ and $\ell$ (radial and orbital quantum numbers)

988: by the WKB method \cite{Cea:1982rg}. We assume the potential

989: \be

990: V(r)=V_0+\mu^2 r

991: \label{vo}

992: \ee

993: which is of course a simplification, but however able to show

994: the linearity of Regge trajectories; $\mu$ is the string tension and

995: $V_o$ is a constant negative term mimicking a repulsive core.

996: 

997: The WKB formula for this potential gives the result

998: \cite{Cea:1982rg} (in units of string tension)

999: \be \displaystyle

1000: \int_0^\sigma dr\ \sqrt{\frac{(M-V(r))^2}{4}-\frac{m_1^2+m_2^2}{2}

1001: +\frac{(m_1^2-m_2^2)^2}{4(M-V(r))^2}}\ =\ \pi\left( n+\frac \ell 2

1002: + \frac 3 4\right)\hskip.2cm\

1003: \ee

1004: where $\sigma=M-V_0-m_1-m_2$. For

1005: light quarks we take for the constituent quark masses the values

1006: $m_1=m_2=300$ MeV. In units of string tension we get, for

1007: $V_0=-1.923$ and $n=0$ the Chew-Frautschi plot of Fig.

1008: \ref{fig:2} (left side) showing the almost linear Regge

1009: trajectory. It is given by Eq. (\ref{alphaR}) with $s_R=0$ since

1010: no spin term has been added to the potential in (\ref{vo}). For

1011: $\alpha=\ell$ we find successive squared meson masses. The Regge

1012: trajectory has slope $\alpha^\prime=0.25/\mu^2$. For $\mu=0.52$

1013: GeV we get a phenomenologically acceptable value of

1014: $\alpha^\prime\sim 0.9$ GeV$^{-2}$. The mass of low-lying meson

1015: with $\ell=0$ made up by up/down quarks turns out to be $\sim 800$

1016: MeV, an acceptable value as well.

1017: \begin{figure}[ht]

1018: \centerline{

1019: \epsfxsize=6cm\epsfbox{regge3bis.eps}\hskip1cm\epsfxsize=6.5cm\epsfbox{regge5.eps} }

1020: \caption {{ On the left: Chew-Frautschi plot for a $q\bar q$ meson of mass $M$

1021:  made up by light quarks with

1022: masses of 0.577 in units of string tension ($\sim 300$ MeV). On the horizontal

1023: axes $t=M^2$ in units of string tension, on the

1024: the vertical axis the Regge trajectory

1025: $\alpha(t)$ approximately given by

1026: $\alpha(t)=s_R+\alpha^\prime (t-m^2_R)$ with $s_R=0$, see text. The meson masses

1027: correspond to successive values of $\alpha=\ell=0\,,1\,,\cdots$.

1028: On the right: Chew-Frautschi plot for a $q\bar Q$ meson of mass $M$

1029: made up by a charm quark and a light quark ($m_c=1.7 $ GeV, $m_q=0.3 $ GeV).

1030: On the horizontal axes $t=M^2$ in units of string tension, on the

1031: the vertical axis the Regge trajectory. Solid line gives the

1032: WKB solution of the bound state equation; dashed-dotted line a

1033: linear approximation in the small $t$ region.

1034: \label{fig:2} }}

1035: \end{figure}

1036: 

1037: We can now repeat the exercise for the $D, D^*$ mesons. We use again  $m_1=300 $ MeV and

1038: put for the charm quark $m_2=m_c=1.7 $ GeV. The result is reported on the right in

1039: Fig. \ref{fig:2}. One can observe some deviations from linearity.

1040: The minimum value is given by kinematics as

1041: $t=4$ in units of the string tension.  Solid line gives the

1042: WKB solution of the bound state equation; for small values we can still approximate

1043: $\alpha(t)$ by a straight line (dashed-dotted line in the figure),

1044: with $\alpha(t)=\alpha_0+\alpha^\prime t$

1045: and $\alpha_0=-1.8$, $\alpha^\prime= 0.09/\mu^2=0.33 $ GeV$^{-2}$.

1046: 

1047: 

1048: 

1049: %\bibliographystyle{apsrev}

1050: %\bibliography{charming}

1051: \begin{thebibliography}{30}

1052: \expandafter\ifx\csname

1053: natexlab\endcsname\relax\def\natexlab#1{#1}\fi

1054: \expandafter\ifx\csname bibnamefont\endcsname\relax

1055:   \def\bibnamefont#1{#1}\fi

1056: \expandafter\ifx\csname bibfnamefont\endcsname\relax

1057:   \def\bibfnamefont#1{#1}\fi

1058: \expandafter\ifx\csname citenamefont\endcsname\relax

1059:   \def\citenamefont#1{#1}\fi

1060: \expandafter\ifx\csname url\endcsname\relax

1061:   \def\url#1{\texttt{#1}}\fi

1062: \expandafter\ifx\csname

1063: urlprefix\endcsname\relax\def\urlprefix{URL }\fi

1064: \providecommand{\bibinfo}[2]{#2}

1065: \providecommand{\eprint}[2][]{\url{#2}}

1066: 

1067: \bibitem[{\citenamefont{Aubert et~al.}(2003)}]{Aubert:2003mm}

1068: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Aubert}} \bibnamefont{et~al.}

1069:   (\bibinfo{collaboration}{BABAR}), \bibinfo{journal}{Phys. Rev. Lett.}

1070:   \textbf{\bibinfo{volume}{91}}, \bibinfo{pages}{171802}

1071:   (\bibinfo{year}{2003}), \eprint{hep-ex/0307026}.

1072: 

1073: \bibitem[{\citenamefont{Aubert et~al.}(2004)}]{Aubert:2003xc}

1074: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Aubert}} \bibnamefont{et~al.}

1075:   (\bibinfo{collaboration}{BABAR}), \bibinfo{journal}{Phys. Rev.}

1076:   \textbf{\bibinfo{volume}{D69}}, \bibinfo{pages}{031102}

1077:   (\bibinfo{year}{2004}), \eprint{hep-ex/0311017}.

1078: 

1079: \bibitem[{\citenamefont{Chen et~al.}(2003)}]{belle:2003jf}

1080: \bibinfo{author}{\bibfnamefont{K.~F.} \bibnamefont{Chen}} \bibnamefont{et~al.}

1081:   (\bibinfo{collaboration}{Belle}), \bibinfo{journal}{Phys. Rev. Lett.}

1082:   \textbf{\bibinfo{volume}{91}}, \bibinfo{pages}{201801}

1083:   (\bibinfo{year}{2003}), \eprint{hep-ex/0307014}.

1084: 

1085: \bibitem[{\citenamefont{Zhang et~al.}(2003)}]{Zhang:2003up}

1086: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Zhang}} \bibnamefont{et~al.}

1087:   (\bibinfo{collaboration}{Belle}), \bibinfo{journal}{Phys. Rev. Lett.}

1088:   \textbf{\bibinfo{volume}{91}}, \bibinfo{pages}{221801}

1089:   (\bibinfo{year}{2003}), \eprint{hep-ex/0306007}.

1090: 

1091: \bibitem[{\citenamefont{Kagan}(2004{\natexlab{a}})}]{Kagan:2004uw}

1092: \bibinfo{author}{\bibfnamefont{A.~L.} \bibnamefont{Kagan}}

1093:   (\bibinfo{year}{2004}{\natexlab{a}}), \eprint{hep-ph/0405134}.

1094: 

1095: \bibitem[{\citenamefont{Grossman}(2004)}]{Grossman:2003qi}

1096: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Grossman}},

1097:   \bibinfo{journal}{Int. J. Mod. Phys.} \textbf{\bibinfo{volume}{A19}},

1098:   \bibinfo{pages}{907} (\bibinfo{year}{2004}), \eprint{hep-ph/0310229}.

1099: 

1100: \bibitem[{\citenamefont{Colangelo et~al.}(2004)\citenamefont{Colangelo,

1101:   De~Fazio, and Pham}}]{Colangelo:2004rd}

1102: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Colangelo}},

1103:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{De~Fazio}}, \bibnamefont{and}

1104:   \bibinfo{author}{\bibfnamefont{T.~N.} \bibnamefont{Pham}},

1105:   \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B597}},

1106:   \bibinfo{pages}{291} (\bibinfo{year}{2004}), \eprint{hep-ph/0406162}.

1107: 

1108: \bibitem[{\citenamefont{Kagan}(2004{\natexlab{b}})}]{Kagan:2004ia}

1109: \bibinfo{author}{\bibfnamefont{A.~L.} \bibnamefont{Kagan}}

1110:   (\bibinfo{year}{2004}{\natexlab{b}}), \eprint{hep-ph/0407076}.

1111: 

1112: \bibitem[{\citenamefont{Bauer et~al.}(1987)\citenamefont{Bauer, Stech, and

1113:   Wirbel}}]{Bauer:1986bm}

1114: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Bauer}},

1115:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Stech}}, \bibnamefont{and}

1116:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Wirbel}},

1117:   \bibinfo{journal}{Z. Phys.} \textbf{\bibinfo{volume}{C34}},

1118:   \bibinfo{pages}{103} (\bibinfo{year}{1987}).

1119: 

1120: \bibitem[{\citenamefont{Charles et~al.}(1999)\citenamefont{Charles, Le~Yaouanc,

1121:   Oliver, Pene, and Raynal}}]{Charles:1998dr}

1122: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Charles}},

1123:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Le~Yaouanc}},

1124:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Oliver}},

1125:   \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Pene}}, \bibnamefont{and}

1126:   \bibinfo{author}{\bibfnamefont{J.~C.} \bibnamefont{Raynal}},

1127:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D60}},

1128:   \bibinfo{pages}{014001} (\bibinfo{year}{1999}), \eprint{hep-ph/9812358}.

1129: 

1130: \bibitem[{\citenamefont{Gritsan}(2004)}]{Gritsan:2004new}

1131: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Gritsan}}

1132:   (\bibinfo{collaboration}{BABAR}) (\bibinfo{year}{2004}),

1133:   \eprint{http://costard.lbl.gov~gritsan/RPM/BABAR-COLL-0028.pdf}.

1134: 

1135: \bibitem[{\citenamefont{Colangelo et~al.}(1990)\citenamefont{Colangelo,

1136:   Nardulli, Paver, and Riazuddin}}]{Colangelo:1989gi}

1137: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Colangelo}},

1138:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}},

1139:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Paver}}, \bibnamefont{and}

1140:   \bibinfo{author}{\bibnamefont{Riazuddin}}, \bibinfo{journal}{Z. Phys.}

1141:   \textbf{\bibinfo{volume}{C45}}, \bibinfo{pages}{575} (\bibinfo{year}{1990}).

1142: 

1143: \bibitem[{\citenamefont{Ciuchini et~al.}(1997)\citenamefont{Ciuchini, Franco,

1144:   Martinelli, and Silvestrini}}]{Ciuchini:1997hb}

1145: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ciuchini}},

1146:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Franco}},

1147:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Martinelli}},

1148:   \bibnamefont{and}

1149:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Silvestrini}},

1150:   \bibinfo{journal}{Nucl. Phys.} \textbf{\bibinfo{volume}{B501}},

1151:   \bibinfo{pages}{271} (\bibinfo{year}{1997}), \eprint{hep-ph/9703353}.

1152: 

1153: \bibitem[{\citenamefont{Isola et~al.}(2001)\citenamefont{Isola, Ladisa,

1154:   Nardulli, Pham, and Santorelli}}]{Isola:2001ar}

1155: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Isola}},

1156:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ladisa}},

1157:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}},

1158:   \bibinfo{author}{\bibfnamefont{T.~N.} \bibnamefont{Pham}}, \bibnamefont{and}

1159:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Santorelli}},

1160:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D64}},

1161:   \bibinfo{pages}{014029} (\bibinfo{year}{2001}), \eprint{hep-ph/0101118}.

1162: 

1163: \bibitem[{\citenamefont{Isola et~al.}(2002)\citenamefont{Isola, Ladisa,

1164:   Nardulli, Pham, and Santorelli}}]{Isola:2001bn}

1165: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Isola}},

1166:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ladisa}},

1167:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}},

1168:   \bibinfo{author}{\bibfnamefont{T.~N.} \bibnamefont{Pham}}, \bibnamefont{and}

1169:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Santorelli}},

1170:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D65}},

1171:   \bibinfo{pages}{094005} (\bibinfo{year}{2002}), \eprint{hep-ph/0110411}.

1172: 

1173: \bibitem[{\citenamefont{Isola et~al.}(2003)\citenamefont{Isola, Ladisa,

1174:   Nardulli, and Santorelli}}]{Isola:2003fh}

1175: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Isola}},

1176:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ladisa}},

1177:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}}, \bibnamefont{and}

1178:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Santorelli}},

1179:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D68}},

1180:   \bibinfo{pages}{114001} (\bibinfo{year}{2003}), \eprint{hep-ph/0307367}.

1181: 

1182: \bibitem[{\citenamefont{Casalbuoni

1183:   et~al.}(1997{\natexlab{a}})}]{Casalbuoni:1997pg}

1184: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Casalbuoni}}

1185:   \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rept.}

1186:   \textbf{\bibinfo{volume}{281}}, \bibinfo{pages}{145}

1187:   (\bibinfo{year}{1997}{\natexlab{a}}), \eprint{hep-ph/9605342}.

1188: 

1189: \bibitem[{\citenamefont{Donoghue et~al.}(1996)\citenamefont{Donoghue, Golowich,

1190:   Petrov, and Soares}}]{Donoghue:1996hz}

1191: \bibinfo{author}{\bibfnamefont{J.~F.} \bibnamefont{Donoghue}},

1192:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Golowich}},

1193:   \bibinfo{author}{\bibfnamefont{A.~A.} \bibnamefont{Petrov}},

1194:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~M.}

1195:   \bibnamefont{Soares}}, \bibinfo{journal}{Phys. Rev. Lett.}

1196:   \textbf{\bibinfo{volume}{77}}, \bibinfo{pages}{2178} (\bibinfo{year}{1996}),

1197:   \eprint{hep-ph/9604283}.

1198: 

1199: \bibitem[{\citenamefont{Nardulli and Pham}(1997)}]{Nardulli:1997ht}

1200: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}} \bibnamefont{and}

1201:   \bibinfo{author}{\bibfnamefont{T.~N.} \bibnamefont{Pham}},

1202:   \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B391}},

1203:   \bibinfo{pages}{165} (\bibinfo{year}{1997}), \eprint{hep-ph/9610525}.

1204: 

1205: \bibitem[{\citenamefont{Watson}(1952)}]{Watson:1952ji}

1206: \bibinfo{author}{\bibfnamefont{K.~M.} \bibnamefont{Watson}},

1207:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{88}},

1208:   \bibinfo{pages}{1163} (\bibinfo{year}{1952}).

1209: 

1210: \bibitem[{\citenamefont{Ball}(2003)}]{Ball:2003rd}

1211: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Ball}},

1212:   \bibinfo{journal}{ECONF} \textbf{\bibinfo{volume}{C0304052}},

1213:   \bibinfo{pages}{WG101} (\bibinfo{year}{2003}), \eprint{hep-ph/0306251}.

1214: 

1215: \bibitem[{\citenamefont{Colangelo et~al.}(1996)\citenamefont{Colangelo,

1216:   De~Fazio, Santorelli, and Scrimieri}}]{Colangelo:1996jv}

1217: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Colangelo}},

1218:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{De~Fazio}},

1219:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Santorelli}},

1220:   \bibnamefont{and}

1221:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Scrimieri}},

1222:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D53}},

1223:   \bibinfo{pages}{3672} (\bibinfo{year}{1996}), \eprint{hep-ph/9510403}.

1224: 

1225: \bibitem[{\citenamefont{Ali et~al.}(1998)\citenamefont{Ali, Kramer, and

1226:   Lu}}]{Ali:1998eb}

1227: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ali}},

1228:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Kramer}}, \bibnamefont{and}

1229:   \bibinfo{author}{\bibfnamefont{C.-D.} \bibnamefont{Lu}},

1230:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D58}},

1231:   \bibinfo{pages}{094009} (\bibinfo{year}{1998}), \eprint{hep-ph/9804363}.

1232: 

1233: \bibitem[{\citenamefont{Eidelman et~al.}(2004)}]{Eidelman:2004wy}

1234: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Eidelman}} \bibnamefont{et~al.}

1235:   (\bibinfo{collaboration}{Particle Data Group}), \bibinfo{journal}{Phys.

1236:   Lett.} \textbf{\bibinfo{volume}{B592}}, \bibinfo{pages}{1}

1237:   (\bibinfo{year}{2004}).

1238: 

1239: \bibitem[{\citenamefont{Donnachie and Landshoff}(1992)}]{Donnachie:1992ny}

1240: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Donnachie}} \bibnamefont{and}

1241:   \bibinfo{author}{\bibfnamefont{P.~V.} \bibnamefont{Landshoff}},

1242:   \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B296}},

1243:   \bibinfo{pages}{227} (\bibinfo{year}{1992}), \eprint{hep-ph/9209205}.

1244: 

1245: \bibitem[{\citenamefont{Zheng}(1995)}]{Zheng:1995mg}

1246: \bibinfo{author}{\bibfnamefont{H.-q.} \bibnamefont{Zheng}},

1247:   \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B356}},

1248:   \bibinfo{pages}{107} (\bibinfo{year}{1995}), \eprint{hep-ph/9504360}.

1249: 

1250: \bibitem[{\citenamefont{Irving and Worden}(1977)}]{Irving:1977ea}

1251: \bibinfo{author}{\bibfnamefont{A.~C.} \bibnamefont{Irving}} \bibnamefont{and}

1252:   \bibinfo{author}{\bibfnamefont{R.~P.} \bibnamefont{Worden}},

1253:   \bibinfo{journal}{Phys. Rept.} \textbf{\bibinfo{volume}{34}},

1254:   \bibinfo{pages}{117} (\bibinfo{year}{1977}).

1255: 

1256: \bibitem[{\citenamefont{Casalbuoni

1257:   et~al.}(1997{\natexlab{b}})}]{Casalbuoni:1996pg}

1258: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Casalbuoni}}

1259:   \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rept.}

1260:   \textbf{\bibinfo{volume}{281}}, \bibinfo{pages}{145}

1261:   (\bibinfo{year}{1997}{\natexlab{b}}), \eprint{hep-ph/9605342}.

1262: 

1263: \bibitem[{\citenamefont{Ball and Braun}(1997)}]{Ball:1997rj}

1264: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Ball}} \bibnamefont{and}

1265:   \bibinfo{author}{\bibfnamefont{V.~M.} \bibnamefont{Braun}},

1266:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D55}},

1267:   \bibinfo{pages}{5561} (\bibinfo{year}{1997}), \eprint{hep-ph/9701238}.

1268: 

1269: \bibitem[{\citenamefont{Cea et~al.}(1982)\citenamefont{Cea, Colangelo,

1270:   Nardulli, Paiano, and Preparata}}]{Cea:1982rg}

1271: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Cea}},

1272:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Colangelo}},

1273:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Nardulli}},

1274:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Paiano}}, \bibnamefont{and}

1275:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Preparata}},

1276:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D26}},

1277:   \bibinfo{pages}{1157} (\bibinfo{year}{1982}).

1278: 

1279: \end{thebibliography}

1280: 

1281: \end{document}

1282: