hep-ph0404004/vap.tex
1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %%%%%%                                               %%%%%%
3: %%    The VAP Green Function in the Resonance Region     %%
4: %%%%%%                                               %%%%%%
5: %%%%%%     Version: 31/5/04                           %%%%%%
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7: 
8: \documentclass[12pt,a4paper]{article}
9: \usepackage{graphicx,cite}
10: \oddsidemargin -0.29cm
11: \textwidth 16.6cm
12: \textheight 21.5cm
13: \topmargin -0.3cm
14: \bibliographystyle{phlb}
15: 
16: \unitlength1cm
17: \pagestyle{plain}
18: 
19: \newcommand{\figwidth}{8cm}
20: \newcommand{\figheight}{6.5cm}
21: 
22: \newcommand{\Frac}[2]{\frac{\displaystyle #1}{\displaystyle #2}}
23: \newcommand{\ep}{\epsilon}
24: \newcommand{\nn}{\nonumber}
25: \newcommand{\IM}{\mbox{\rm Im}}
26: \newcommand{\eqn}[1]{(\ref{#1})}
27: \newcommand{\mev}{\mbox{\rm MeV}}
28: \newcommand{\gev}{\mbox{\rm GeV}}
29: \newcommand{\Li}{\mbox{\rm Li}_2}
30: \newcommand{\MSb}{{\overline{\rm MS}}}
31: \newcommand\lsim{\mathrel{\rlap{\lower4pt\hbox{\hskip1pt$\sim$}}
32:     \raise1pt\hbox{$<$}}}
33: \newcommand\gsim{\mathrel{\rlap{\lower4pt\hbox{\hskip1pt$\sim$}}
34:     \raise1pt\hbox{$>$}}}
35: 
36: \newcommand{\VAP}{\langle V \! A P\rangle}
37: \newcommand{\order}{{\cal O}}
38: 
39: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
40: %  Some definitions 
41: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
42: \newcommand{\chpt}{CHPT}
43: \newcommand{\CQg}{\left( {\langle\bar q q \rangle^{(2)}(\mu)\over 
44: f_\pi^3}\right)}
45: \newcommand{\CQ}{\left( {\langle\bar q q \rangle^{(2)}(\mu)\over f^2}\right)}
46: \newcommand{\ds}{\displaystyle}
47: \newcommand{\bea}{\begin{eqnarray}}
48: \newcommand{\eea}{\end{eqnarray}}
49: \newcommand{\beq}{\begin{equation}}
50: \newcommand{\eeq}{\end{equation}}
51: \newcommand{\nl}{\nonumber\\}
52: \newcommand{\dg}{\dagger}
53: \newcommand{\hl}{\hline}
54: \newcommand{\vsl}{v\hskip-.2cm/}
55: \newcommand{\real}{{\rm Re}}
56: \newcommand{\imag}{{\rm Im}}
57: \newcommand{\hepph}[1]{{\tt hep-ph/#1}} 
58: \newcommand{\hepex}[1]{{\tt hep-ex/#1}}
59: \newcommand{\heplat}[1]{{\tt hep-lat/#1}}  
60: %
61: \newcommand{\cO}{{\cal O}}
62: \newcommand{\cL}{{\cal L}}
63: \newcommand{\cQ}{{\cal Q}}
64: \newcommand{\cA}{{\cal A}}
65: \newcommand{\mbf}{\mathbf} 
66: \newcommand{\mrm}{\mathrm} 
67: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
68: \begin{document}
69: 
70: \begin{titlepage}
71: 
72: \begin{flushright}
73: {\small\sf  MAP-295 \\ IFIC/04-12\\FTUV/04-0401\\UWThPh-2004-6}
74: \end{flushright}
75: 
76: \vspace*{1.5cm} 
77: \begin{center}
78: {\Large\bf $\VAP$ Green Function in the Resonance Region$^*$}
79: \\[20mm]
80: 
81: {\normalsize\bf \sc V. Cirigliano$^{1}$, G. Ecker$^2$, M.~Eidem\"uller$^{3}$, 
82: A. Pich$^{3}$ and J. Portol\'es$^{3}$  }\\
83:  
84: \vspace{1cm} 
85: ${}^{1}$ Department of Physics, California Institute of Technology\\ 
86: Pasadena, California 91125, USA\\[10pt]
87: ${}^{2)}$ Institut f\"ur Theoretische Physik, Universit\"at 
88: Wien\\ Boltzmanngasse 5, A-1090 Vienna, Austria \\[10pt]
89: ${}^{3)}$ Departament de F\'{\i}sica Te\`orica, IFIC, CSIC --- 
90: Universitat de Val\`encia \\ 
91: Edifici d'Instituts de Paterna, Apt. Correus 22085, E-46071 
92: Val\`encia, Spain \\
93: \end{center}
94: 
95: \vfill
96: 
97: \begin{abstract}
98: \noindent
99: We analyse the $\VAP$ three-point function of vector, axial-vector and
100: pseudoscalar currents. In the spirit of large $N_C$, a resonance
101: dominated Green function is confronted with the leading high-energy
102: behaviour from the operator product expansion. The matching is shown
103: to be fully compatible with a chiral resonance Lagrangian and it
104: allows to determine some of the chiral low-energy constants of
105: $\cO(p^6)$. 
106: \end{abstract}
107: 
108: \vfill
109: 
110: %\noindent
111: %{\it Keywords}: resonance chiral QCD\\
112: %{\it PACS}: ?
113: 
114: \noindent 
115: *~Work supported in part by HPRN-CT2002-00311 (EURIDICE) and by
116: Acciones Integradas, HU2002-0044 (MCYT, Spain), Project No. 19/2003 
117: (Austria).
118: \end{titlepage}
119: 
120: 
121: 
122: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
123: % Beginning of the paper
124: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
125: 
126: %%%%%%%%%%%%%%%%%%%  Introduction %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
127: 
128: \addtocounter{page}{1} 
129: 
130: %%%%%%%%%%%%%%%%%%%%% Changes %%%%%%%%%%%%%%%%%%%%
131: 
132: 
133: 
134: \paragraph{1.}
135: %\section{Introduction}
136: 
137: Following the work of Moussallam \cite{Moussallam:1997xx} and 
138: of Knecht and Nyffeler \cite{Knecht:2001xc}, we
139: reanalyse the three-point function of vector, axial-vector and
140: pseudoscalar currents. The procedure of matching between low and high
141: energies is especially transparent for Green functions like the
142: $\VAP$ correlator that are order parameters of chiral symmetry
143: breaking. 
144: Working in the chiral limit and at leading order in the $1/N_C$ expansion, 
145: the matching is 
146: performed by saturating the operator product expansion (OPE) with a 
147: certain number of resonance
148: multiplets in accordance with the Minimal Hadronic Ansatz \cite{MHA}.
149: An alternative way consists in using directly a chiral
150: resonance Lagrangian to perform the matching, getting thereby also
151: information on some of the resonance couplings of such a Lagrangian.
152: 
153: In Ref.~\cite{Knecht:2001xc} it was claimed that, 
154: unlike the situation at $\cO(p^4)$ \cite{Ecker:1989te,Ecker:1989yg}, a minimal
155: chiral resonance Lagrangian using the Proca formalism for spin-1 fields 
156: \cite{Prades:1994ys} is unable to recover the asymptotic behaviour of 
157: the Green functions considered, in particular the $\VAP$ correlator.
158: This claim should be contrasted with the results of 
159: Ref.~\cite{Ruiz-Femenia:2003hm} for the $\langle VVP \rangle$
160: correlator where a chiral resonance Lagrangian, using the
161: antisymmetric tensor formalism of spin-1 fields, was found to be 
162: consistent with the leading asymptotic behaviour of QCD.
163: 
164: An important bonus of the matching procedure is 
165: the prediction of resonance contributions to couplings of 
166: the Chiral Perturbation Theory (CHPT) Lagrangian at ${\cal O}(p^6)$ 
167: \cite{Fearing:1996ga,Bijnens:1999sh,Bijnens:2001bb,Ebertshauser:2001nj}. 
168: With 90 such low-energy constants (LECs) in the even-intrinsic parity 
169: Lagrangian for three flavours, such predictions are indispensable at
170: present to do
171: phenomenology to next-to-next-to-leading order in the low-energy
172: expansion. Compatibility between the constraints from QCD at high
173: energies and a suitable chiral resonance Lagrangian, as already 
174: established at $\cO(p^4)$ \cite{Ecker:1989te,Ecker:1989yg}, would 
175: allow for a consistent Lagrangian treatment of resonance
176:   contributions at least up to $O(p^6)$.
177: With this motivation in mind, we first show that a
178: chiral resonance Lagrangian with an appropriate set of
179: resonance fields is fully consistent with QCD constraints for the 
180: $\VAP$ Green function to be specified below. We then go on to 
181: determine six of the LECs of
182: $\cO(p^6)$  from $V, A$ and $P$ resonance contributions.   
183:  
184: The approach employed in this letter should be seen as an
185:   approximation to large-$N_C$ QCD. The approximation consists in the
186:   choice of a hadronic ansatz and in a set of QCD short-distance
187:   constraints to be satisfied. We adopt the following guidelines:
188: \begin{itemize} 
189: \item {\it Hadronic ansatz}: We include all lowest-lying resonance
190:   multiplets that can contribute to the given Green function. This
191:   choice is motivated by the well-founded assumption that the
192:   low-lying hadronic spectrum determines the chiral LECs that govern
193:   the low-energy behaviour of the Green function.  
194: 
195: \item {\it Short-distance constraints}: Working with a finite number
196:   of resonance multiplets, conflicts may arise between different types
197:   of asymptotic constraints \cite{Bijnens:2003rc}. We assign the
198:   highest priority to the constraints dictated by the OPE to leading
199:   order in inverse powers of large momenta. We then consider the
200:   constraints implied by the high-momentum behaviour of all hadronic
201:   form factors (as predicted by the quark counting rules
202:   \cite{Lepage:1979zb,Lepage:1980fj,Brodsky:1981rp}) with on-shell
203:   Goldstone modes and photons. We do not consider form factors with
204:   external resonance states.
205: 
206: \end{itemize} 
207:   
208: 
209: 
210: %%%%%%%%%%%%%%%%%%% Definitions %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
211: \paragraph{2.}  
212: %\section{The $\langle V A P \rangle$ Green function and its chiral
213: %expansion} 
214: 
215: The $\VAP$ three-point function in momentum space is defined as
216: \begin{eqnarray}
217: \label{VAPdef}
218: (\Pi_{V\!AP})_{\mu\nu}^{abc}(p,q) = \int d^4x \int d^4y \ e^{i(p \cdot x + 
219: q \cdot y)} \langle 0 \vert T \{ V_\mu^a(x) A_\nu^b(y) P^c(0) \} \vert 0
220: \rangle \,,
221: \end{eqnarray}
222: with $SU(3)$ octet vector, axial-vector and pseudoscalar currents 
223: \begin{eqnarray}
224: \label{currents}
225: V_\mu^a = \bar \psi \gamma_\mu \displaystyle\frac{\lambda^a}{2} \psi \,, 
226: & \quad
227: A_\mu^a = \bar \psi \gamma_\mu \gamma_5
228: \displaystyle\frac{\lambda^a}{2}  \psi \,, 
229: & \quad
230: P^a = \bar \psi i \gamma_5 \displaystyle\frac{\lambda^a}{2} \psi \,. 
231: \end{eqnarray}
232: It satisfies the chiral Ward identities \cite{Moussallam:1997xx,Gasser:1984yg}
233: \begin{eqnarray}
234: \label{WardIdentities}
235: p^{\mu}(\Pi_{V\!AP})_{\mu\nu}^{abc}(p,q) &=& 
236: \langle{\overline\psi}\psi\rangle_0 f^{abc}\,\Bigg[\frac{q_{\nu}}{q^2}\,-\,
237: \frac{(p+q)_{\nu}}{(p+q)^2}\Bigg]\, ,\nn\\
238: q^{\nu}(\Pi_{V\!AP})_{\mu\nu}^{abc}(p,q) &=&
239: \langle{\overline\psi}\psi\rangle_0 f^{abc}\,\frac{(p+q)_{\mu}}{(p+q)^2}\, ,
240: \end{eqnarray}
241: where $\langle{\overline\psi}\psi\rangle_0$ denotes the
242: quark condensate in the chiral limit.  The general solution
243: of these Ward identities, taking into account the QCD symmetries
244: $SU(3)_V$, parity and time reversal, is 
245: \cite{Moussallam:1997xx,Knecht:2001xc}
246: \begin{eqnarray}
247: \label{GeneralSolution}
248: %\! \! \! \! \! \! \! \! \! \! \! \! \!
249: (\Pi_{V\!AP})_{\mu\nu}^{abc}(p,q) & = & 
250:   f^{abc}\,\Bigg\{ \langle{\overline\psi}\psi\rangle_0 \left[
251: \frac{(p+2q)_\mu q_\nu}{q^2 (p+q)^2} - \frac{g_{\mu\nu}}{(p+q)^2}\right] 
252:  \nn\\
253: %&& \; \; \; \; \;\; \; \; \; \; \; \; \;\; \; \;\; \;  
254: && + P_{\mu\nu}(p,q) {\cal F}(p^2,q^2,(p+q)^2) 
255: %\nn \\
256: %&& \; \; \; \; \; \; \; \; \; \;\; \; \;\; \; \; \; \;
257: + Q_{\mu\nu}(p,q) {\cal G}(p^2,q^2,(p+q)^2) \Bigg\} \, .
258: \end{eqnarray}
259: The transverse tensors $P_{\mu\nu}$ and 
260: $Q_{\mu\nu}$ are defined as
261: \begin{eqnarray}
262: \label{PQdef}
263: P_{\mu\nu}(p,q) &=& q_\mu p_\nu - (p \cdot q) g_{\mu\nu} \, , \nn\\ 
264: Q_{\mu\nu}(p,q) &=& p^2 q_\mu q_\nu + q^2 p_\mu p_\nu - (p \cdot q)
265: p_\mu q_\nu - p^2 q^2 g_{\mu\nu} \, .  
266: \end{eqnarray}
267: 
268: %\subsection{Chiral expansion}
269: The behaviour of the invariant functions ${\cal F}$ and ${\cal G}$ 
270: at small momentum transfers is governed by the contributions from 
271: Goldstone boson intermediate states. As one-particle
272: exchange dominates in the limit $N_C \to \infty$, we only need to
273: keep the corresponding Goldstone boson poles and the polynomial terms
274: involving the LECs. In the basis of Ref. \cite{Bijnens:1999sh} 
275: for the LECs of ${\cal O}(p^6)$ one finds \cite{Knecht:2001xc}
276: ($F$ is the pion decay constant in the chiral limit)
277: \begin{eqnarray}
278: \label{ChPTSolution}
279: \! 
280: {\cal F}^{\mrm{CHPT}}(p^2,q^2,(p+q)^2) & = &  \frac{4 
281: \langle{\overline\psi}\psi\rangle_0}{F^2 (p+q)^2} 
282: \Bigg[ \, L_9 + L_{10}   + \left( C_{78} - \frac{5}{2}C_{88} - C_{89} 
283: + 3C_{90} \right) p^2 \nl
284: &&  \hspace*{-2cm} \left. + \left( C_{78} - 2C_{87} + 
285: \frac{1}{2}C_{88}\right) q^2 
286:  + \left(C_{78} + 4C_{82} - \frac{1}{2}C_{88}\right)
287: (p+q)^2 \right] \,+ \cO(p^8) , \nn \\ 
288: {\cal G}^{\mrm{CHPT}}(p^2,q^2,(p+q)^2) & = & \frac{4 
289: \langle{\overline\psi}\psi\rangle_0}{F^2 q^2 (p+q)^2} 
290: \left[ L_9 + 2( -C_{88} +  C_{90}) p^2  \right. \nn\\ 
291: &&  \left. + (2C_{78} - C_{89} + C_{90}) q^2  - 2  C_{90} (p+q)^2
292: \right] \,+ \cO(p^8) \,.
293: \end{eqnarray}
294: 
295: 
296: %%%%%%%%%%%%%%%%%%%% Short-distance analysis %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
297: \paragraph{3.}  
298: %\section{Operator product expansion and $V, A, P$ exchange}
299: 
300: We next recall the properties of the $\VAP$ correlator at short
301: distances. Following Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}, 
302: the analysis will be restricted to the leading term in the OPE.
303: Being an order parameter of chiral symmetry
304: breaking, the behaviour of the $\VAP$ Green function at short
305: distances is smoother than expected from naive power counting
306: arguments. The leading contributions at large momenta are proportional
307: to the quark condensate, which has the same anomalous dimension
308: as the pseudoscalar current. Hence the corresponding Wilson coefficient
309: in the OPE of the $\VAP$ correlator does not have an anomalous
310: dimension and, therefore, QCD corrections should modify the following 
311: results only mildly.
312: \par
313: As shown by Knecht and Nyffeler \cite{Knecht:2001xc}, two
314: short-distance limits are of interest here. In the first case, the two
315: momenta $p$ and $q$ in the correlator (\ref{VAPdef}) become 
316: simultaneously large. In position space this amounts to the
317: situation where the space-time arguments of the three operators
318: tend towards the same point at the same rate ($x \sim y \sim 0$). 
319: Restricting the discussion as always to the leading term in the OPE, 
320: one obtains \cite{Moussallam:1997xx,Knecht:2001xc}
321: \begin{eqnarray}
322: \label{ShortDistanceGeneral}
323: \lim_{\lambda\to\infty}(\Pi_{V\!AP})_{\mu\nu}^{abc}(\lambda p,\lambda q)
324: & = & \frac{\langle{\overline\psi}\psi\rangle_0}{\lambda^2}\,f^{abc}\,
325: \frac{1}{p^2q^2(p+q)^2}\,\bigg\{\,
326: p^2(p+2q)_{\mu}q_{\nu} \nn\\* 
327: && -g_{\mu\nu}p^2q^2 +\frac{1}{2}(p^2-q^2-(p+q)^2)P_{\mu\nu}-Q_{\mu\nu}\,
328: \bigg\} +\,{\cal O}\left(\frac{1}{\lambda^4}\right)\,
329: \end{eqnarray}
330: and therefore
331: \begin{eqnarray}
332: \label{ShortDistanceFG}
333: \lim_{\lambda \to \infty} {\cal F}((\lambda p)^2, (\lambda q)^2, (\lambda p
334: +\lambda q)^2) &=& \frac{\langle{\overline\psi}\psi\rangle_0}
335: {2\lambda^4} \, 
336: \frac{p^2 - q^2 - (p+q)^2}{p^2 q^2 (p+q)^2} + 
337: \order\left(\frac{1}{\lambda^6}\right) \,, \nn\\
338: \lim_{\lambda \to \infty} {\cal G}((\lambda p)^2, (\lambda q)^2, (\lambda p
339: + \lambda q)^2) &=& - \frac{\langle{\overline\psi}\psi\rangle_0}
340: {\lambda^6} \,\frac{1}{p^2 q^2 (p+q)^2} + 
341:  \order\left(\frac{1}{\lambda^8}\right) \,.
342: \end{eqnarray}
343: \par
344: The second situation of interest corresponds to the case where the
345: relative distance between only two of the three operators involved
346: becomes small. We refer to Ref.~\cite{Knecht:2001xc} for a complete
347: discussion of the various cases where different two-point functions 
348: arise. It turns out that many of the resulting conditions are not
349: independent when taken together with the constraint 
350: (\ref{ShortDistanceGeneral}). Therefore, we only reproduce the
351: following short-distance condition from Ref.~\cite{Knecht:2001xc},
352: which, together with (\ref{ShortDistanceGeneral}), leads to a complete
353: set of leading-order high-energy constraints: 
354: \begin{eqnarray}
355: \label{ShortDistance1}
356: \lim_{\lambda \to \infty}
357: (\Pi_{V\!AP})_{\mu\nu}^{abc}(\lambda p,q-\lambda p) &=& 
358: -\,\frac{1}{\lambda}\,f^{abc}\,\langle{\overline\psi}\psi\rangle_0
359: \,\frac{p_{\mu}q_{\nu}+p_{\nu}q_{\mu}-(p\cdot q)g_{\mu\nu}}{p^2 q^2}
360: + \, {\cal O}\left(\frac{1}{\lambda^2}\right)\,.
361: \end{eqnarray}
362: In terms of the invariant functions ${\cal F}$ and ${\cal G}$, this 
363: asymptotic behaviour implies 
364: \begin{eqnarray}
365: \label{FGShortDistance1}
366: \lim_{\lambda \to \infty}
367: {\cal F}((\lambda p)^2,(q-\lambda p)^2,q^2) &=&
368: \frac{\langle{\overline\psi}\psi\rangle_0}{\lambda^2 p^2}\, 
369: \left[{\cal F}^{(0)}(q^2)\,+\, \frac{1}{\lambda}\,\frac{p\cdot q}{p^2}
370: \,{\cal F}^{(1)}(q^2)
371: \,+\,\order\left({1\over \lambda^2}\right)\right] \,, \nn\\
372: \lim_{\lambda \to \infty}
373: {\cal G}((\lambda p)^2,(q-\lambda p)^2,q^2) &=&
374: \frac{\langle{\overline\psi}\psi\rangle_0}{(\lambda^2 p^2)^2}\, 
375: \left[{\cal G}^{(0)}(q^2)\,+\, \frac{1}{\lambda}\,\frac{p\cdot q}{p^2}
376: \,{\cal G}^{(1)}(q^2)
377: \,+\,\order\left({1\over \lambda^2}\right)\right] \,, 
378: \end{eqnarray}
379: with
380: \begin{eqnarray}
381: \label{FGShortDistance2}
382: {\cal F}^{(0)}(q^2)-{\cal G}^{(0)}(q^2) = 
383: \displaystyle\frac{1}{q^2} \, , \quad & \quad
384: {\cal F}^{(1)}(q^2)-{\cal G}^{(1)}(q^2)+{\cal G}^{(0)}(q^2)= 
385: \displaystyle\frac{2}{q^2}\,.
386: \end{eqnarray}
387: \par
388: The $\VAP$ correlator is also related to the $\Gamma_{\rm VA}$ and
389: $\Gamma_{\rm VP}$ vertex functions \cite{Knecht:2001xc,Moussallam:1997xx}.
390: The short-distance behaviour of these
391: vertex functions gives additional constraints on the parameters of 
392: ${\cal F}$ and ${\cal G}$. However, using the expansion of 
393: $\Gamma_{\rm VA}$ and $\Gamma_{\rm VP}$ consistently up to 
394: ${\cal O}(1/\lambda)$,
395: those constraints are equivalent to the limit where one momentum of $\VAP$ 
396: becomes large and therefore they do not provide new information at leading
397: order.
398: \par
399: In order to solve the short-distance conditions
400: (\ref{ShortDistanceGeneral}) and (\ref{ShortDistance1}), we propose
401: the following ansatz inspired by large $N_C$ that is a generalization
402: of the one used in Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}:
403: \begin{eqnarray}
404:  \label{FGansatz}
405: {\cal F}(p^2,q^2,(p+q)^2) &=&
406: \frac{\langle{\overline\psi}\psi\rangle_0}{(p^2-M_V^2)(q^2-M_A^2)}\nn\\*
407: &&\times \left[a_0+\frac{b_1+b_2 p^2 + b_3 q^2}{(p+q)^2}
408: +\frac{c_1 + c_2 p^2 + c_3 q^2}{(p+q)^2-M_P^2}\right] ~,\nn\\*
409: {\cal G}(p^2,q^2,(p+q)^2) &=&
410: \frac{\langle{\overline\psi}\psi\rangle_0}{(p^2-M_V^2)q^2}
411: \Bigg[\frac{d_1 + d_2 q^2}{(p+q)^2(q^2-M_A^2)}
412: + \frac{f}{(p+q)^2-M_P^2}\Bigg] ~.
413: \end{eqnarray}
414: This ansatz differs from the one in
415: Refs.~\cite{Moussallam:1997xx,Knecht:2001xc} by
416: the inclusion of a nonet of pseudoscalar resonances with mass $M_P$
417: (remember that we always work in the chiral limit). The ansatz of 
418: Refs.~\cite{Moussallam:1997xx,Knecht:2001xc} is recovered in the 
419: limit $M_P \to \infty$, i.e. by dropping the parameters 
420: $c_1,c_2,c_3,f$ that specify the
421: contributions from pseudoscalar resonance exchange.
422:  While the ansatz of Refs.~\cite{Moussallam:1997xx,Knecht:2001xc} 
423: was designed to match the leading short-distance constraints with the
424: minimal resonance content, we include all lowest-lying resonance
425: multiplets that can contribute to the  LECs of ${\cal O} (p^6)$.
426: Our approach appears more natural when
427: attempting to construct an explicit Lagrangian realization for the 
428: resonance interactions (see discussion below).
429: \par
430: The parameters in (\ref{FGansatz}) fall into two classes:
431: \begin{itemize} 
432: \item[1)] The dimensionless parameters $a_0,b_2,b_3,c_2,c_3,d_2,f$ are
433: constrained by the OPE conditions (\ref{ShortDistanceGeneral}) 
434: and (\ref{ShortDistance1}).
435: \item[2)] The parameters $b_1,c_1,d_1$ with squared mass dimension are not
436:   affected by the (leading-order) OPE conditions. As discussed in the 
437:   next paragraph,
438:   they can be constrained  by asymptotic
439:   conditions on various form factors when one or two pions are put
440:   on-shell \cite{Moussallam:1997xx,Knecht:2001xc}.
441: \end{itemize} 
442: 
443: The short-distance conditions (\ref{ShortDistanceGeneral}) and 
444: (\ref{ShortDistance1}) yield the following set of six linear equations
445: for the seven parameters in class 1:
446: \begin{equation} 
447:  \label{OPEeq}
448: \begin{tabular}{lll} 
449: \quad $a_0 = -\displaystyle\frac{1}{2}$ , \qquad & \quad $b_2+c_2 =
450: \displaystyle\frac{1}{2}$ , \qquad & \quad 
451: $b_3+c_3 = -\displaystyle\frac{1}{2}$ , \nn \\[.2cm] 
452: \quad $d_2+f = -1$ , \qquad & \quad $b_2+b_3-d_2 = 1$ ,
453: \qquad & \quad $2 b_2-d_2 = 2$ .
454: \end{tabular} 
455: \end{equation}  
456: We use these equations to express six of the parameters in
457: terms of $b_3$:
458: \begin{equation}
459: \label{OPEsol} 
460: \begin{tabular}{lll} 
461: $a_0 = -\displaystyle\frac{1}{2}$ ,\hspace*{1cm}  & $b_2 = 1 + b_3$ , 
462: \hspace*{1cm} & $d_2 = 2 b_3$ , \\[.2cm]  
463: $c_2 = -\displaystyle\frac{1}{2} - b_3$ , \hspace*{1cm}  & 
464: $c_3 = c_2$ ,
465: \hspace*{1cm} & $f = 2 c_2$ .
466: \end{tabular} 
467: \end{equation} 
468: Setting the pseudoscalar exchange parameters $c_2,c_3$ and $f$ to zero
469: (or, equivalently, letting $M_P \to \infty$ in the ansatz
470: (\ref{FGansatz})), we 
471: recover the solution of Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}:
472: \begin{equation} 
473: \label{KNsol}
474: \begin{tabular}{llll} 
475: $a_0 = -\displaystyle\frac{1}{2}$ ,\hspace*{1cm}  & 
476: $b_2 = \displaystyle\frac{1}{2}$ , \hspace*{1cm} &  
477: $b_3 = -\displaystyle\frac{1}{2}$ , \hspace*{1cm} &
478: $d_2 = - 1$ .
479: \end{tabular} 
480: \end{equation} 
481: 
482: 
483: \paragraph{4.}  
484: %\section{Form factors and Weinberg sum rules}
485: Additional information on the parameters in  Eq.~(\ref{FGansatz})
486: can be obtained by putting one or two momenta in the $\VAP$ Green
487: function (\ref{VAPdef}) on the pion mass shell
488: \cite{Moussallam:1997xx,Knecht:2001xc}. The form factors appearing in 
489: the resulting vertex functions are not directly constrained by QCD but 
490: there are strong theoretical arguments \cite{Lepage:1979zb,Lepage:1980fj,
491: Brodsky:1981rp} for the form factors in question to fall off at least
492: like $1/q^2$ for large momentum transfers.
493: 
494: The two dimensional parameters $b_1,d_1$ were determined in
495: this way by Moussallam \cite{Moussallam:1997xx}, making also use of
496: the two Weinberg sum rules \cite{Weinberg:1967kj}. The inclusion of 
497: pseudoscalar resonances does not affect those results:
498: \begin{eqnarray}
499: \label{b1d1} 
500: b_1 = M_A^2 - M_V^2 ~, \qquad & \qquad d_1 = 2 M_A^2 ~.
501: \end{eqnarray}
502: 
503: Although we cannot determine $c_1$ from consideration of a pionic form
504: factor we can fix the remaining dimensionless parameter $b_3$
505: in this way. For this purpose, we consider the axial form factor
506: $G_A(t)$ governing the matrix element 
507: $\langle \gamma | A_{\mu} | \pi \rangle$ \cite{Gasser:1984yg}.
508: Extracting $G_A(t)$ from the
509: Green function (\ref{VAPdef}) by setting $p^2=0$ and $(p+q)^2=0$
510: (massless pion), one finds in terms of the parameters defined in 
511: (\ref{FGansatz}) 
512: \begin{equation}
513: \label{ganew} 
514: G_A(t) = \Frac{F^2}{M_V^2} \; \Frac{b_1+b_3 t}{M_A^2-t} ~.
515: \end{equation} 
516: Demanding that the form factor $G_A(t)$ vanishes for large $t$ 
517: \cite{Lepage:1979zb,Lepage:1980fj,Brodsky:1981rp,Ecker:1989yg}, we
518: obtain
519: \begin{equation}
520: \label{b3} 
521: b_3=0~.
522: \end{equation}
523: Therefore, the solution (\ref{KNsol})
524: \cite{Moussallam:1997xx,Knecht:2001xc} 
525: is not compatible with the asymptotic
526: vanishing of $G_A(t)$. The value $b_3= -1/2$ in (\ref{KNsol})
527: is also at the origin
528: of the very small partial width obtained 
529: \cite{Moussallam:1997xx,Knecht:2001xc} for the decay $a_1 \to
530: \pi \gamma$. The decay matrix element is governed 
531: by the combination $b_1 + b_3 M_A^2$. With the solution (\ref{KNsol}),
532: this matrix element is proportional to 
533: $(M_A^2 - 2 M_V^2)/2$ and therefore suppressed 
534: compared to our solution with $b_3=0$ where the same matrix element is 
535: given by $b_1= M_A^2 - M_V^2$. The numerical value of the decay width
536: $\Gamma(a_1 \to \pi \gamma)$ will be discussed later.
537: \par
538:  The $\VAP$ Green function also contributes to the decay
539: $\tau \rightarrow 3 \pi \nu_{\tau}$. One can study the
540: axial-vector form factor contributing to this process and require that
541: it vanishes like $1/q^2$ for large momentum transfer. This procedure
542: provides the conditions
543: \begin{equation}
544: b_2 = 1 \; , \; \; \; \; \;  \; \; \; b_3=0 \; , \;\; \; \; \; 
545:  \; \; \; d_2 = 0 
546: \end{equation}
547: that are consistent with the results in Eqs.~(\ref{OPEsol}) and
548: (\ref{b3}). 
549: \par
550:  The discussion above is related to a general point raised 
551: in Ref.~\cite{Bijnens:2003rc}. 
552: There it was claimed that for a given three-point function 
553: that is an order parameter of chiral symmetry breaking,
554: any large-$N_C$ inspired ansatz with a {\em finite number} of resonance
555: multiplets will fail to reproduce simultaneously (i) the leading OPE
556: constraints and (ii) the $1/q^2$ asymptotic behaviour of  {\em all}
557: hadronic form factors (appearing as residues of two-particle poles in
558: the Green function).  
559: Our explicit construction shows that with a reasonable number of
560: resonance multiplets one can still fulfill both the leading OPE 
561: constraints and the correct asymptotic behaviour of form factors 
562: involving Goldstone modes and on--shell photons 
563: ($F_V^{\pi}(t)$ and $G_{A}(t)$ in our case).
564: 
565: \paragraph{5.}  
566: %\section{Chiral resonance Lagrangian}
567: 
568: We now turn to an explicit realization of our solutions
569: (\ref{OPEsol}), (\ref{b1d1}) and (\ref{b3}) in terms of a chiral 
570: resonance Lagrangian. Such a Lagrangian was introduced in 
571: Refs.~\cite{Ecker:1989te,Ecker:1989yg} to investigate the LECs of
572: $\cO(p^4)$. That Lagrangian has to be extended when going up to
573: $\cO(p^6)$ where bilinear resonance couplings also contribute.
574: 
575: In the notation of Refs.~\cite{Ecker:1989te,Ecker:1989yg}, the
576: kinetic terms of the Lagrangian restricted to vector, axial-vector and
577: pseudoscalar resonance fields ($V(1^{--})$, $A(1^{++})$ and
578: $P(0^{-+})$) are given by
579: \begin{eqnarray}
580: {\cal L}_{ \rm kin}^R &=& -\frac{1}{2} \langle
581: \nabla^\lambda R_{\lambda\mu} \nabla_\nu R^{\nu\mu} - \frac{M_R^2}{2}
582: R_{\mu\nu} R^{\mu\nu} \rangle \; \; \; , \; \; \;  \; \; R \, = \, V,A 
583: \; , \nn \\
584: {\cal L}_{ \rm kin}^P &=& \frac{1}{2} \langle \nabla^{\mu} P \nabla_{\mu} P
585: - M_P^2 P^2 \rangle \; .
586: \end{eqnarray}
587: $V_{\mu\nu}$ and $A_{\mu\nu}$ are antisymmetric tensor fields
588: describing nonets of spin-1 mesons and $P$ is a pseudoscalar
589: (nonet) field. The brackets $\langle \dots \rangle$ denote a 
590: three-dimensional trace in flavour space. In the large-$N_C$ limit
591: where multiple trace terms are suppressed, the interaction terms
592: linear in the resonance fields  and with the minimal number of
593: derivatives and mass insertions are given by \cite{Ecker:1989te}
594: \begin{eqnarray}
595: \label{lag1}
596: {\cal L}_2^{V,A,P} & = &  \frac{F_V}{2\sqrt{2}} \langle V_{\mu\nu}
597: f_+^{\mu\nu}\rangle + i\,\frac{G_V}{\sqrt{2}} \langle V_{\mu\nu} u^\mu
598: u^\nu\rangle +  \frac{F_A}{2\sqrt{2}} \langle A_{\mu\nu}
599: f_-^{\mu\nu}\rangle + i \, d_{m} \, \langle P \chi_{-} \rangle ~.
600: \end{eqnarray}
601: The chiral fields $u^\mu,f_+^{\mu\nu},f_-^{\mu\nu},\chi_{-}$ are
602: defined as usual \cite{Ecker:1989te,Ecker:1989yg} in terms of
603: Goldstone fields and external fields. The coupling
604: constants $F_V$, $G_V$, $F_A$ and $d_m$ are real.
605: 
606: To account for LECs of $\cO(p^6)$, we must also include couplings of 
607: Goldstone bosons with two resonance fields. As for the linear
608: couplings (\ref{lag1}), we only include terms with the minimal number
609: of derivatives and mass insertions. 
610: Although there is a priori no guarantee that the various
611: asymptotic constraints discussed previously can be satisfied with such 
612: a minimal Lagrangian this approach proved to be successful at 
613: $\cO(p^4)$ \cite{Ecker:1989te,Ecker:1989yg}.
614: 
615: The vector--axial-vector bilinear terms were already
616: introduced in Ref.~\cite{Dumm:2003ku}:\\
617: \begin{equation} 
618: \label{OVA}
619: \begin{tabular}{lll} 
620: ${\cal L}_2^{VA}  =  \sum_{i=1}^{5} \, \lambda_i^{VA} \, 
621: {\cal O}_i^{VA}$  & \hspace*{1cm} &  \\[.2cm] 
622: ${\cal O}_1^{VA} =  \langle \,  [ \, V^{\mu\nu} \, , \, 
623: A_{\mu\nu} \, ] \,  \chi_- \, \rangle$ , & &
624: ${\cal O}_2^{VA}  =  i\,\langle \, [ \, V^{\mu\nu} \, , \, 
625: A_{\nu\alpha} \, ] \, h_\mu^{\;\alpha} \, \rangle$ , \\[.2cm] 
626: ${\cal O}_3^{VA} =   i \,\langle \, [ \, \nabla^\mu V_{\mu\nu} \, , \, 
627: A^{\nu\alpha}\, ] \, u_\alpha \, \rangle$ , & &
628: ${\cal O}_4^{VA}  =  i\,\langle \, [ \, \nabla_\alpha V_{\mu\nu} \, , \, 
629: A^{\alpha\nu} \, ] \,  u^\mu \, \rangle$ , \\[.2cm] 
630: ${\cal O}_5^{VA} =   i \,\langle \, [ \, \nabla_\alpha 
631: V_{\mu\nu} \, , \, 
632: A^{\mu\nu} \, ] \, u^\alpha \, \rangle$ .& & \\[.1cm] 
633: \end{tabular} 
634: \end{equation}
635: The pseudoscalar--vector bilinear couplings are given by \\[.1cm] 
636: \begin{equation} 
637: \label{OPV}
638: \begin{tabular}{lll} 
639: ${\cal L}_2^{PV} =  \sum_{i=1}^{2} \, \lambda_i^{PV} \, 
640: {\cal O}_i^{PV}$  & \hspace*{1cm} & \\[.2cm] 
641: ${\cal O}_1^{PV} =  i \, \langle \,  [ \, \nabla^\mu P \, , \, 
642: V_{\mu\nu} \, ] \,  u^\nu \, \rangle$ , & &
643: ${\cal O}_2^{PV}   =  i \, \langle \,  [ \, P \, , \, 
644: V_{\mu\nu} \, ] \,  f^{\mu\nu}_- \, \rangle$ .\\[.1cm] 
645: \end{tabular} 
646: \end{equation} 
647: Finally, only one bilinear term with pseudoscalar and axial-vector 
648: fields contributes to our Green function:
649: \begin{eqnarray}
650: \label{OPA}
651: {\cal L}_2^{PA} & = &  \lambda_1^{PA} \, 
652: {\cal O}_1^{PA}  \nl
653: {\cal O}_1^{PA} &  = & i \, \langle \,  [ \, P \, , \, 
654: A_{\mu\nu} \, ] \,  f^{\mu\nu}_+ \, \rangle \,.
655: \end{eqnarray}
656: Adding the lowest-order chiral Lagrangian \cite{Gasser:1985gg}, 
657: we obtain the following chiral resonance Lagrangian to be used for the 
658: calculation of the $\VAP$ Green function:
659: \begin{equation}
660: \label{TotalLagrangian}
661: {\cal L}_{\mrm{CHRL}} \, = \, \frac{F^2}{4}\langle u_{\mu}
662: u^{\mu} + \chi _+ \rangle  + 
663: {\cal L}_{ \mrm{kin}}^{V,A,P}  +  {\cal L}_2^{V,A,P}  +  
664: {\cal L}_2^{VA}  +  {\cal L}_2^{PV}  + 
665: {\cal L}_2^{PA} \; .
666: \end{equation}
667: 
668: \paragraph{6.}  
669: %\section{The $\VAP$ Green function in chiral resonance theory}
670: \begin{figure}
671: \begin{center}
672: \includegraphics[height=100pt,width=100pt,angle=0]{fig1a.eps}
673: \hspace*{0.5cm}
674: \includegraphics[height=100pt,width=100pt,angle=0]{fig1b.eps}
675: \hspace*{0.5cm}
676: \includegraphics[height=100pt,width=100pt,angle=0]{fig1c.eps}
677: \caption{\label{fig:OneResonance}
678: Single-resonance exchange: $\pi$ denotes a Goldstone boson, $V$ and 
679: $A$ stand for vector and axial-vector resonances, respectively.}
680: \end{center}
681: \end{figure}
682: Lowest-order Goldstone boson exchange provides
683: the first two terms in the $\VAP$ correlator (\ref{GeneralSolution}) 
684: that drive the chiral Ward identities (\ref{WardIdentities}).
685: Next we compute the diagrams shown in 
686: Fig.~\ref{fig:OneResonance} where a single resonance is exchanged. 
687: There is no contribution from pseudoscalar resonance exchange in this
688: case \cite{Ecker:1989te}. 
689: The diagrams in Fig.~\ref{fig:OneResonance} give rise to 
690: \begin{eqnarray}
691:  \label{L1FGResults}
692: {\cal F}^{V,A}(p^2,q^2,(p+q)^2) &=& 
693: \frac{\langle{\overline\psi}\psi\rangle_0}{(p+q)^2 (p^2-M_V^2)}
694: \left(\frac{F_V^2-2 F_V G_V}{F^2}
695: -  \frac{p^2-M_V^2}{q^2-M_A^2}\frac{F_A^2}{F^2}\right)~, \nn\\
696: {\cal G}^{V,A}(p^2,q^2,(p+q)^2) &=& 
697: \frac{\langle{\overline\psi}\psi\rangle_0}{q^2 (p+q)^2 (p^2-M_V^2)} 
698: \frac{- 2 F_V G_V}{F^2} \,.
699: \end{eqnarray}
700: 
701: The double-resonance contributions to the $\VAP$ Green function are
702: described by the diagrams in Fig.~\ref{fig:TwoResonances}.
703: Summing up single- and double-resonance exchange contributions, the
704: final result can be given in terms of the parameters defined in the
705: general ansatz (\ref{FGansatz}):
706: \begin{equation} 
707:  \label{FGParameters}
708: \begin{tabular}{ll} 
709: $a_0 =  -2 \sqrt{2}\displaystyle\frac{F_V F_A}{F^2}\lambda_0$ ,  &
710: $b_1 = M_V^2\displaystyle\frac{F_A^2}{F^2}-M_A^2 
711: \displaystyle\frac{F_V^2-2F_V G_V}{F^2}$ , \nn\\
712: $b_2 =  -\displaystyle\frac{F_A^2}{F^2}+
713: 2\sqrt{2}\displaystyle\frac{F_V F_A}{F^2}\lambda'$ , &
714: $b_3 =  \displaystyle\frac{F_V^2-2F_V G_V}{F^2} +
715: 2\sqrt{2}\displaystyle\frac{F_V F_A}{F^2}\lambda''$ , \nn\\
716: $c_1 = -M_V^2 c_2 - M_A^2 c_3$ , &
717: $c_2 = - 8\sqrt{2}\,\displaystyle\frac{F_A d_m}{F^2}\, \lambda_1^{PA}$ , \nn\\
718: $c_3 = 8\sqrt{2}\,\displaystyle\frac{F_V d_m}{F^2}\, 
719: \left(\displaystyle\frac{\lambda_1^{PV}}{2}+\lambda_2^{PV}\right)$ , &
720: $d_1 = \displaystyle\frac{2 F_V G_V}{F^2} M_A^2$ , \nl
721: $d_2 = - \displaystyle\frac{2 F_V G_V}{F^2} + 
722: 2\sqrt{2}\, \displaystyle\frac{F_V F_A}{F^2}(\lambda'+\lambda'')$ ,
723: \qquad  &
724: $f = 4\sqrt{2}\,\displaystyle\frac{F_V d_m}{F^2}\, \lambda_1^{PV}$~,
725: \end{tabular} 
726: \end{equation} 
727: where we have used the definitions \cite{Dumm:2003ku}
728: \begin{eqnarray}
729:  \label{LambdaDef}
730: \sqrt{2} \lambda_0 &=& -4\lambda_1^{VA}-\lambda_2^{VA}
731: -\frac{\lambda_4^{VA}}{2}-\lambda_5^{VA}~, \nn\\
732: \sqrt{2} \lambda' &=& \lambda_2^{VA}-\lambda_3^{VA}
733: +\frac{\lambda_4^{VA}}{2}+\lambda_5^{VA}~, \nn\\
734: \sqrt{2} \lambda'' &=& \lambda_2^{VA}
735: -\frac{\lambda_4^{VA}}{2}-\lambda_5^{VA}~.
736: \end{eqnarray}
737: \begin{figure}
738: \begin{center}
739: \includegraphics[height=100pt,width=100pt,angle=0]{fig2a.eps}
740: \hspace*{0.5cm}
741: \includegraphics[height=100pt,width=100pt,angle=0]{fig2b.eps}
742: \hspace*{0.5cm}
743: \includegraphics[height=100pt,width=100pt,angle=0]{fig2c.eps}\\[4mm]
744: \includegraphics[height=100pt,width=100pt,angle=0]{fig2d.eps}
745: \hspace*{0.5cm}
746: \includegraphics[height=100pt,width=100pt,angle=0]{fig2e.eps}
747: \caption{\label{fig:TwoResonances}
748: Double-resonance exchange contribution to the $\VAP$ Green function; 
749: $P$ denotes a pseudoscalar resonance.}
750: \end{center}
751: \end{figure}
752: %%%%%%%%%%%%%%%%%%% Short-distance constraints %%%%%%%%%%%%%%%%%%%
753: 
754: As pointed out in  Ref.~\cite{Ecker:1989yg}, the short-distance
755: structure of QCD can be used to constrain the couplings of the 
756: chiral resonance Lagrangian by matching the asymptotic 
757: behaviour of two-point functions, form factors and scattering 
758: amplitudes with the results from resonance exchange (at leading 
759: order in the $1/N_C$ expansion and assuming a single nonet of $V$ 
760: and $A$ resonances each). In this way one finds
761: the relations \cite{Ecker:1989yg,Weinberg:1967kj} 
762: \begin{eqnarray} 
763: \label{FVGVFA}
764: F_V G_V = F^2, \qquad & F_V^2 - F_A^2 = F^2 , \qquad &
765: F_V^2 M_V^2 = F_A^2 M_A^2~,
766: \end{eqnarray} 
767: allowing to express the couplings $F_V$, $G_V$ and $F_A$ in terms of
768: $F$, $M_V$ and $M_A$. From a similar joint analysis of the scalar form factor 
769: \cite{Jamin:2000wn,Jamin:2001zq} and the SS-PP sum rules 
770: \cite{Golterman:1999au} one gets, assuming again only one nonet of 
771: $S$ and $P$ resonances each \cite{Pich:2002xy},
772: \begin{equation}
773: \label{dm}
774: d_m = \Frac{F}{2 \sqrt{2}} \; \; .
775: \end{equation}
776: 
777: The ten relations in (\ref{FGParameters}) can now be compared with the
778: previous results (\ref{OPEsol}), (\ref{b1d1}) and (\ref{b3}). From the
779: equations for $a_0,b_2,b_3$ we extract the combinations of 
780: coupling constants $\lambda_0,\lambda',\lambda''$ that satisfy the 
781: relation $4 \lambda_0 = \lambda' + \lambda''$.
782: %\begin{equation} 
783: %4\lambda_1^{VA} + 3\lambda_2^{VA} - \lambda_3^{VA} +
784: %\displaystyle\frac{\lambda_4^{VA}}{2} + \lambda_5^{VA} = 0~. 
785: %\end{equation}
786: The equation for $d_2$ is then automatically satisfied.
787: The relations for $c_2,c_3,f$ fix the coupling
788: constants $\lambda_1^{PV},\lambda_2^{PV},\lambda_1^{PA}$. The
789: equations for $b_1,d_1$ are consistent with (\ref{b1d1}) and there is
790: a new relation for the dimensional parameter $c_1$:
791: \begin{equation}
792: \label{c1} 
793: c_1 = \displaystyle\frac{1}{2} (M_V^2 + M_A^2)~. 
794: \end{equation}
795:  
796: The predictions of the chiral resonance Lagrangian 
797: (\ref{TotalLagrangian}) are fully consistent with the OPE and form
798: factor constraints (\ref{OPEsol}), (\ref{b1d1}) and (\ref{b3}). In
799: other words, at leading order in $1/N_C$ and considering for each
800: current in the Green function only one multiplet of resonances with
801: the same quantum numbers, the chiral resonance
802: Lagrangian provides a $\VAP$ Green function with the correct
803: asymptotic behaviour dictated by QCD.
804: 
805: Using (\ref{FVGVFA}) and (\ref{dm}), we get the following final
806: results for the coupling constants of the chiral Lagrangians 
807: (\ref{OVA}), (\ref{OPV}) and (\ref{OPA}), depending only on the masses
808: $M_V$ and $M_A$:
809: \begin{equation}
810: \label{lambda}
811: \begin{tabular}{lclcl}
812: $\lambda' = \displaystyle\frac{M_A}{2 \sqrt{2} M_V}$ , & \hspace*{1cm} &    
813: $\lambda'' = \displaystyle\frac{M_A^2 - 2 M_V^2}{2 \sqrt{2} M_V M_A}$ ,
814:   & \hspace*{1cm} & $4 \lambda_0 = \lambda' + \lambda''$ , \nn \\[.4cm] 
815: $\lambda_1^{PV}= - 4 \lambda_2^{PV}$ , & \hspace*{1cm} &
816: $\lambda_2^{PV} = \displaystyle\frac{\sqrt{M_A^2 - M_V^2}}{8 M_A}$ , &
817: \hspace*{1cm} & 
818: $\lambda_1^{PA} = \displaystyle\frac{\sqrt{M_A^2 - M_V^2}}{8 M_V}$~.
819: \end{tabular}
820: \end{equation}
821: 
822: We now come back to the axial form factor $G_A(t)$ in the matrix 
823: element $\langle \gamma | A_{\mu} | \pi \rangle$. With
824: single-resonance exchange only, this form factor is given by 
825: \cite{Ecker:1989yg}
826: \begin{equation}
827: \label{gaold}
828: G_A(t) = \Frac{2F_V G_V - F_V^2}{M_V^2} + \Frac{F_A^2}{M_A^2-t} \; \;.
829: \end{equation}
830: Requiring $G_A(t)$ to vanish for $t \to \infty$ implies the relation
831: $F_V = 2 G_V$, one version of the so-called KSFR relation 
832: \cite{Kawarabayashi:1966kd,Riazuddin:1966sw}. The inclusion of 
833: bilinear resonance couplings modifies the form factor as given
834: in Eq.~(\ref{ganew}) with $b_1=M_A^2 - M_V^2, ~b_3=0$, and it induces
835: a correction to the KSFR relation:
836: \begin{eqnarray} 
837: \displaystyle\frac{2 F_V G_V - F_V^2}{2 F^2} = 1 -
838: \displaystyle\frac{F_V^2}{2 F^2} = \displaystyle\frac{M_A^2 - 2 M_V^2}
839: {2(M_A^2 - M_V^2)}  \; \; .
840: \end{eqnarray}
841: With $M_A=1.23$ GeV, $M_V=0.771$ GeV \cite{Hagiwara:2002fs}, the 
842: right-hand side takes the value $\simeq 0.18$. The
843: partial decay width  $\Gamma(a_1 \to \pi \gamma)$ is now 
844: \begin{equation}
845: \label{width}
846:  \Gamma(a_1 \to \pi \gamma) = \displaystyle\frac{\alpha M_A}{24}
847:   \left( \displaystyle\frac{M_A^2}{M_V^2} - 1 \right)^3
848:  \left(1 - \displaystyle\frac{M_\pi^2}{M_A^2} \right)^3~. 
849: \end{equation}
850: With the same values for $M_V,M_A$ and with the physical pion mass,
851: we obtain $\Gamma(a_1 \to \pi \gamma) = 1.33$ MeV, in reasonable
852: agreement with the experimental value $640 \pm 246$ keV 
853: \cite{Zielinski:1984au}. 
854: It should be noted that the width is very sensitive to $M_A$. 
855: A value of $M_A=\sqrt{2}M_V$, as required by the KSFR relation, 
856: would give $\Gamma(a_1 \to \pi \gamma) = 316$ keV
857: and for $M_A=1.2$ GeV, a value extracted from  $\tau\to 3 \pi\nu_\tau$
858: data \cite{Dumm:2003ku}, one obtains $\Gamma(a_1 \to \pi \gamma) = 1.01$ MeV.
859: In comparison, the decay
860: width in the scenario of Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}
861: is strongly suppressed with respect to (\ref{width}) by a factor
862: \begin{equation} 
863: \displaystyle\frac{(M_A^2 - 2 M_V^2)^2}{4(M_A^2 - M_V^2)^2} \simeq
864: 0.03 ~.
865: \end{equation} 
866: Although the corresponding width is more than an order of magnitude
867: smaller than the listed value \cite{Hagiwara:2002fs} the experimental 
868: situation (discussed in Ref.~\cite{Moussallam:1997xx}) remains to be 
869: settled. However, we are confident that future experiments will be
870: able to decide between two predictions that differ by a factor 30.
871: \par
872: As commented above, the relations (\ref{lambda}) also have a bearing on 
873: the decays
874: $\tau \to  3 \pi \nu_\tau$. As shown in Ref.~\cite{Dumm:2003ku}, the
875: requirement that the $J=1$ axial spectral function vanishes for large
876: momentum transfer implies certain values for $\lambda'$,
877: $\lambda''$. Those values coincide\footnote{In Ref.~\cite{Dumm:2003ku} 
878: the KSFR relation $F_V=2 G_V$ was adopted implying $M_A^2 = 2 M_V^2$.} 
879: with the corresponding results in Eq.~(\ref{lambda}). The 
880: coupling $\lambda_0$ was extracted in that reference from a fit to the 
881: spectrum and branching ratio of the decay. However, the fitted value 
882: turns out to be too large and, as discussed in that reference, carries 
883: a big uncertainty due to the fact that in the $\tau \rightarrow 3 \pi 
884: \nu_{\tau}$ amplitude the  coupling $\lambda_0$ always appears
885: multiplied by a factor $M_{\pi}^2$.
886: 
887:  
888: 
889: %%%%%%%%%%%% Results for low-energy constants %%%%%%%%%%%%%%%%%%%
890: \paragraph{7.}  
891: %\section{Low-energy constants from matching}
892: 
893: Having established the compatibility between the QCD short-distance
894: constraints and the chiral resonance Lagrangian, we can now use 
895: the results (\ref{OPEsol}), (\ref{b1d1}) and (\ref{b3}) (with the 
896: additional relation (\ref{c1}) for
897: $c_1$) to compare with the low-energy expansion 
898: (\ref{ChPTSolution}) of the $\VAP$ Green function. It turns out that
899: all LECs appearing in (\ref{ChPTSolution}) can be determined
900: separately in this way:
901: \begin{equation} 
902:  \label{LowEnergySolution}
903: \begin{tabular}{lll} 
904: $L_9$ = $\displaystyle\frac{F^2}{2 M_V^2}$  , & \hspace*{.1cm} &
905: $L_{10}$ =  $- \displaystyle\frac{F^2 (M_A^2 + M_V^2)}{4 M_V^2 M_A^2}$ 
906:  ,  \\[.3cm] 
907: $C_{78} = \displaystyle\frac{F^2(3 M_A^2 + 4 M_V^2)}{8 M_V^4 M_A^2} 
908: - \displaystyle\frac{F^2}{16 M_V^2 M_P^2}$  , & \hspace*{.1cm} &
909: $C_{82} = - \displaystyle\frac{F^2(4 M_A^2 + 5 M_V^2)}{32 M_V^4
910:   M_A^2} 
911: - \displaystyle\frac{F^2}{32 M_A^2 M_P^2}$  ,  \\[.3cm] 
912: $C_{87} = \displaystyle\frac{F^2(M_A^4 + M_V^4 + M_A^2 M_V^2)}
913: {8 M_V^4 M_A^4}$  ,  & \hspace*{.1cm} &
914: $C_{88} = - \displaystyle\frac{F^2}{4 M_V^4} + 
915: \displaystyle\frac{F^2}{8 M_V^2 M_P^2}$  ,  \\[.3cm] 
916: $C_{89} = \displaystyle\frac{F^2(3 M_A^2 + 2 M_V^2)}{4 M_V^4 M_A^2}$ 
917:  ,  & \hspace*{.1cm} & 
918: $C_{90} = \displaystyle\frac{F^2}{8 M_V^2 M_P^2}$  .
919: \end{tabular} 
920: \end{equation} 
921: 
922: The results for $L_9$ and $L_{10}$, the LECs of $\cO(p^4)$, coincide 
923: with those in Ref.~\cite{Ecker:1989yg}. The LECs of $\cO(p^6)$
924: differ from the ones in Ref.~\cite{Knecht:2001xc}, first of all by terms
925: involving the mass $M_P$ of the pseudoscalar resonance nonet. However,
926: even in the limit $M_P \to \infty$ a  small difference remains 
927: for the LECs
928: $C_{78}$ and $C_{82}$. The reason for this difference is
929: that the short-distance limit and the limit $M_P \to \infty$ do not
930: commute, as it is evident from the analysis of Eq.~(\ref{OPEeq}). 
931: Since $M_P \simeq 1.3 \, \mbox{GeV}$ is rather heavy, the pseudoscalar
932: contributions to the LECs are not large, ranging from $0.02 \times 10^{-4}$
933: to $0.09 \times 10^{-4}$ for $F^2 \, C_i$. All contributions from 
934: pseudoscalar resonances originate from the $f$ parameter in 
935: Eq.~(\ref{FGansatz}), except for $C_{82}$ which also gets a contribution
936: from the $c_1$ term.
937: \par
938: Our large-$N_C$ determination of these LECs cannot reproduce their
939: scale dependence (a next-to-leading-order effect in the $1/N_C$ counting). 
940: We have checked that in all cases the variation of the renormalized
941: LECs between $\mu = M_K$ and $\mu = 1$ GeV remains within a 
942: range of $30 \,\%$. Although there is no reason a priori that this
943: result will be valid for all LECs our findings quantify the
944: reliability of the estimates (\ref{LowEnergySolution}) for 
945: phenomenological applications.
946: 
947: 
948: \paragraph{8.}  
949: %\section{Conclusions}
950: 
951: We have extended the ansatz of Moussallam, Knecht and Nyffeler
952: \cite{Moussallam:1997xx,Knecht:2001xc} for the $\VAP$ Green function
953: in the intermediate energy region by including the lowest-lying nonet
954: of pseudoscalar resonances. Since the model has more parameters
955: it trivially satisfies all short-distance constraints discussed 
956: in Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}. \\[.2cm] 
957: The distinctive features of our solution are the following:
958: \begin{itemize} 
959: \item The axial form factor in the matrix element for the 
960:   decay $\pi \to e \nu_e \gamma$ vanishes for $t \to \infty$.
961: \item We obtain a partial decay width $\Gamma(a_1 \to \pi \gamma)$ 
962:   in reasonable agreement with the experimental value 
963:   \cite{Zielinski:1984au} but more than a factor 30 bigger
964:   than the prediction of Refs.~\cite{Moussallam:1997xx,Knecht:2001xc}.
965: \item The asymptotic vanishing of the axial spectral function relevant 
966:   for the decay $\tau \to  3 \pi \nu_\tau$ \cite{Dumm:2003ku} is
967:   compatible with the model. 
968: \item The LECs of $\cO(p^6)$ determined by the matching procedure
969:   differ in general from the ones derived in Ref.~\cite{Knecht:2001xc}
970:   even in the limit $M_P \to \infty$ for the mass of the pseudoscalar
971:   resonance nonet. 
972: \item The solution of the short-distance constraints is consistent
973:   with a minimal chiral resonance Lagrangian with vector, axial-vector
974:   and pseudoscalar resonances where the spin-1 mesons are described
975:   by antisymmetric tensor fields.
976: \end{itemize}
977: 
978: \paragraph{Acknowledgements}
979: \noindent 
980: We wish to thank Roland Kaiser and Pedro D. Ruiz-Femenia for useful 
981: discussions on the topic of this paper. We also thank Eduardo de Rafael,
982: Marc Knecht, Bachir Moussallam, Andreas Nyffeler and Joaquim Prades for
983: comments on the manuscript.
984: V.C. is supported by a Sherman Fairchild Fellowship from Caltech. 
985: M.E. thanks the European Union for financial support under contract 
986: no. HPMF-CT-2001-01128. This work has been supported in part by 
987: MCYT (Spain) under grant FPA2001-3031 and by ERDF funds from the
988: European Commission.
989: 
990: \vspace*{1cm} 
991: %\newpage
992: %\bibliography{ref_vap}
993: \begin{thebibliography}{99}
994: 
995: \bibitem{Moussallam:1997xx} B.~Moussallam, {\em Nucl. Phys.} {\bf B504}
996: (1997) 381.
997: 
998: \bibitem{Knecht:2001xc} M.~Knecht and A.~Nyffeler, {\em Eur. Phys. J.} 
999: {\bf C21} (2001) 659.
1000: 
1001: \bibitem{MHA} S.~Peris, M.~Perrottet and E.~de Rafael, {\em JHEP} {\bf 05}
1002: (1998) 011; \\
1003: S.~Peris, B.~Phily and E.~de Rafael, {\em Phys. Rev. Lett.} {\bf 86} (2001)
1004: 14; \\
1005: E.~de Rafael, {\em Nucl. Phys. Proc. Suppl.} {\bf 119} (2003) 71. 
1006: 
1007: \bibitem{Ecker:1989te} G.~Ecker, J.~Gasser, A.~Pich and E.~de~Rafael, 
1008: {\em Nucl. Phys.} {\bf B321} (1989) 311.
1009: 
1010: \bibitem{Ecker:1989yg} G.~Ecker {\em et al.}, {\em Phys. Lett.} {\bf B223}
1011: (1989) 425.
1012: 
1013: \bibitem{Prades:1994ys} J.~Prades, {\em Z. Phys.} {\bf C63} (1994) 491.
1014: 
1015: \bibitem{Ruiz-Femenia:2003hm} P.D.~Ruiz-Femen\'{\i}a, A.~Pich and J.~Portol\'es,
1016: {\em JHEP} {\bf 07} (2003) 003.
1017: 
1018: \bibitem{Fearing:1996ga} H.W.~Fearing and S.~Scherer, {\em Phys. Rev.} 
1019: {\bf D53} (1996) 315.
1020: 
1021: \bibitem{Bijnens:1999sh} J.~Bijnens, G.~Colangelo and G.~Ecker, {\em JHEP}
1022: {\bf 02} (1999) 020.
1023: 
1024: \bibitem{Bijnens:2001bb} J.~Bijnens, L.~Girlanda and P.~Talavera, 
1025: {\em Eur. Phys. J.} {\bf C23} (2002) 539.
1026: 
1027: \bibitem{Ebertshauser:2001nj} T.~Ebertshauser, H.W.~Fearing and S.~Scherer,
1028: {\em Phys. Rev.} {\bf D65} (2002) 054033.
1029: 
1030: \bibitem{Bijnens:2003rc} J.~Bijnens, E.~G\'amiz, E.~Lipartia and J.~Prades,
1031: {\em JHEP} {\bf 04} (2003) 055.
1032: 
1033: \bibitem{Lepage:1979zb} G.P.~Lepage and S.J.~Brodsky, {\em Phys. Lett.}
1034: {\bf B87} (1979) 359.
1035: 
1036: \bibitem{Lepage:1980fj} G.P.~Lepage and S.J.~Brodsky, {\em Phys. Rev.}
1037: {\bf D22} (1980) 2157.
1038: 
1039: \bibitem{Brodsky:1981rp} S.J.~Brodsky and G.P.~Lepage, {\em Phys. Rev.}
1040: {\bf D24} (1981) 1808.
1041: 
1042: \bibitem{Gasser:1984yg} J.~Gasser and H.~Leutwyler, {\em Ann. Phys.} {\bf 158}
1043: (1984) 142.
1044: 
1045: \bibitem{Weinberg:1967kj} S.~Weinberg, {\em Phys. Rev. Lett.} {\bf 18} (1967)
1046: 507.
1047: 
1048: \bibitem{Dumm:2003ku} D.~G\'omez Dumm, A.~Pich and J.~Portol\'es, 
1049: {\em Phys. Rev.} {\bf D69} (2004) 073002.
1050: %{\em arXiv:hep-ph/0312183}.
1051: 
1052: \bibitem{Gasser:1985gg} J.~Gasser and H.~Leutwyler, {\em Nucl. Phys.} 
1053: {\bf B250} (1985) 465.
1054: 
1055: \bibitem{Jamin:2000wn} M.~Jamin, J.A.~Oller and A.~Pich, {\em Nucl. Phys.}
1056: {\bf B587} (2000) 331.
1057: 
1058: \bibitem{Jamin:2001zq} M.~Jamin, J.A.~Oller and A.~Pich, {\em Nucl. Phys.}
1059: {\bf B622} (2002) 279.
1060: 
1061: \bibitem{Golterman:1999au} M.F.L.~Golterman and S.~Peris, {\em Phys. Rev.}
1062: {\bf D61} (2000) 034018.
1063: 
1064: \bibitem{Pich:2002xy} A.~Pich, in {\em  Phenomenology of 
1065: large $N_C$ QCD}, ed. R.F.~Lebed (World Scientific, 2002) p. 239,
1066: {\em arXiv:hep-ph/0205030}.
1067: 
1068: \bibitem{Kawarabayashi:1966kd} K.~Kawarabayashi and M.~Suzuki, 
1069: {\em Phys. Rev. Lett.} {\bf 16} (1966) 255.
1070: 
1071: \bibitem{Riazuddin:1966sw} Riazuddin and Fayyazuddin, {\em Phys. Rev.} 
1072: {\bf 147} (1966) 1071.
1073: 
1074: \bibitem{Hagiwara:2002fs} K.~Hagiwara {\em et al.}, {\em Phys. Rev.}
1075: {\bf D66} (2002) 010001.
1076: 
1077: \bibitem{Zielinski:1984au} M.~Zielinski {\em et al.}, {\em Phys. Rev. Lett.}
1078: {\bf 52} (1984) 1195.
1079: 
1080: 
1081: \end{thebibliography}
1082: 
1083: \end{document}
1084: