1: %\documentclass[aps,prd,showpacs,preprintnumbers,eqsecnum,amsmath,amssymb,nofootinbib]{revtex4}
2: %\renewcommand{\topfraction}{1}
3: %\renewcommand{\bottomfraction}{1}
4: %\renewcommand{\textfraction}{0}
5: %\renewcommand{\floatpagefraction}{0}
6: \documentclass[aps,preprintnumbers,eqsecnum,amsmath,amssymb,nofootinbib]{revtex4}
7: \usepackage{graphicx}% Include figure files
8: \usepackage{dcolumn}% Align table columns on decimal point
9: \usepackage{bm}% bold math
10: %\usepackage{simpleConference}
11: \usepackage{epsfig}
12: %\usepackage{showkeys}
13: \renewcommand{\thesection}{\arabic{section}}
14: \renewcommand{\thetable}{\arabic{table}}
15: \renewcommand{\thefigure}{\arabic{figure}}
16: \renewcommand{\tablename}{Table}
17: \renewcommand{\figurename}{Fig.}
18: \renewcommand{\theequation}{\thesection.\arabic{equation}}
19: \begin{document}
20: %\preprint{}
21: %\preprint{RM3-TH/04-7}
22: \title{Heavy-to-light form factors: sum rules on the light cone and beyond}
23: \author{Wolfgang Lucha$^{a}$, Dmitri Melikhov$^{a,b}$ and Silvano Simula$^{c}$}
24: \affiliation{
25: $^a$ Institute for High Energy Physics, Austrian Academy of Sciences, Nikolsdorfergasse 18, A-1050, Vienna, Austria\\
26: $^b$ Nuclear Physics Institute, Moscow State University, 119992, Moscow, Russia\\
27: $^c$ INFN, Sezione di Roma III, Via della Vasca Navale 84, I-00146, Roma, Italy}
28: \date{\today}
29: \begin{abstract}
30: We report the first systematic analysis of the off-light-cone effects in
31: sum rules for heavy-to-light form factors. These effects are investigated in a model
32: based on scalar constituents, which allows a technically rather simple
33: analysis but has the essential features of the analogous QCD calculation.
34: The correlator relevant for the extraction of the heavy-to-light form factor
35: is calculated in two different ways:
36: first, by adopting the full Bethe--Salpeter amplitude of the light meson
37: and, second, by performing the expansion of this amplitude near the light cone $x^2=0$.
38: We demonstrate that the contributions to the correlator from the light-cone term $x^2=0$ and the
39: off-light-cone terms $x^2\ne 0$ have the same order in the $1/m_Q$ expansion.
40: The light-cone correlator, corresponding to $x^2=0$, is shown to systematically
41: overestimate the full correlator, the difference being $\sim \Lambda_{\rm QCD}/\delta$,
42: with $\delta$ the continuum subtraction parameter of order 1 GeV.
43: Numerically, this difference is found to be
44: $10\div 20$\%.
45: \end{abstract}
46: %\pacs{11.55.Hx, 12.38.Lg, 11.10.St}
47: \maketitle
48: \vspace{-.5cm}
49: \section{Introduction}
50: QCD sum rules on the light cone \cite{lcsr} have been extensively applied to various exclusive
51: form factors, in particular, to weak heavy-to-light transition form factors
52: (see, e.g., \cite{ck}).
53: Within this method, the relevant correlators are obtained as power
54: expansions in $x^2$ near the
55: light cone (LC) $x^2=0$ in terms of the distribution amplitudes of increasing twist.
56: In practice, however, only few lowest-twist distribution amplitudes are known with a
57: reasonable accuracy.
58: Thus one has to rely on calculations which take into account only lowest powers of
59: $x^2$, at most up to terms linear in $x^2$, without a systematic study
60: of the effects related to higher powers of $x^2$ (\cite{ck,bb,bz} and references therein).
61: However, the off-LC $(x^2\ne 0)$ contributions to the correlator are not parametrically
62: suppressed compared to the contribution evaluated at $x^2=0$:
63: for instance, Braun and Halperin \cite{bh} studied the pion elastic form
64: factor with light-cone sum rules and found that contributions
65: to the correlator from terms
66: corresponding to higher powers of $x^2$ in the pion Bethe--Salpeter (BS) amplitude
67: $\langle 0|T\bar u(x)\gamma_\mu\gamma_5 d(0)|\pi(p)\rangle$
68: have in general the same $\sim 1/q^4$ behaviour.
69: As we show here, a similar situation occurs for the heavy-to-light form factor:
70: the contributions to the correlator coming from
71: the LC and the off-LC terms in the BS amplitude of the light meson
72: have the same dependence in $1/m_Q$, $m_Q$ being the heavy-quark mass.
73: Therefore, relying on calculations performed at $x^2=0$ does not seem to be safe
74: and the effects of all powers of $x^2$ should be summed and taken into account in the sum rule.
75:
76: In this paper, we study the correlator relevant for the extraction of the
77: heavy-to-light form factors with sum rules and propose a method which allows one to obtain and
78: take into account the full $x^2$-dependence of the BS amplitude of the light meson.
79: For the sake of argument, we present the analysis for the model with scalar constituents,
80: the heavy scalar $Q$ of the mass $m_Q$ and the light scalar $\varphi$ of the mass $m$
81: (which we name ``quarks'' throughout the paper). We study the weak transition of a heavy spinless
82: bound state $M_Q(Q\varphi)$ to the light spinless bound state $M(\varphi\varphi)$ (which we refer to as
83: mesons) induced by the weak heavy-to-light $Q\to\varphi$ quark transition.
84: The analysis of this model is technically simpler but at the same time
85: allows one to study some essential features of the corresponding QCD case.
86:
87: The paper is organized as follows:
88:
89: In Sec.~\ref{sect_bs}, we consider the BS amplitude of the light meson
90: $\Psi_{\rm BS}(x,p)=\langle 0|T\varphi(x)\varphi(0)|M(p)\rangle$ and its expansion in powers of $x^2$.
91: %
92: In Sec.~\ref{sect_ff}, we study the correlator
93: $i\int d^4x\exp(ipx) \langle 0|T \varphi(x)Q(x)Q(0)\varphi(0)|M(p')\rangle$,
94: relevant for the extraction of the
95: $M_Q\to M$ form factor within the method of sum rules. First, we show that the use
96: of the Nakanishi representation for
97: the BS amplitude of the light meson amounts to a
98: summation of all the off-LC effects. We refer to the correlator obtained by this procedure
99: as the full correlator. Second, we make use of the
100: expansion of the BS amplitude near the LC and obtain a different form
101: of the correlator corresponding to this LC expansion.
102: We show that for the quark--quark interaction dominated by
103: one-boson exchange at small distances (similar to QCD, where the
104: interaction is dominated by one-gluon exchange) any term corresponding to $(x^2)^n$
105: in the BS amplitude gives a contribution to the correlator which behaves as $1/m_Q^2$,
106: independent of $n$. Keeping the term $x^2=0$ only, gives the LC correlator.
107: We then discuss the procedure of making the Borelization of the correlator and obtaining the sum rule for the full and the LC
108: correlators.
109: %
110: Section~\ref{numerical} presents the numerical analysis of the full and the LC correlators for
111: charm and beauty decays. It is shown that at $q^2\simeq 0$ the LC correlator systematically overestimates the full
112: correlator, the difference being numerically $10\div 20$\%, independent of the mass of the heavy quark.
113: %
114: Section~\ref{conclusions} summarizes our results.
115:
116: \section{\label{sect_bs}The Bethe--Salpeter amplitude and its expansion near the light cone}
117: The BS amplitude is defined according to
118: \begin{eqnarray}
119: \label{BS}
120: \Psi_{\rm BS}(x,p')=\langle 0|T\varphi(x)\varphi(0)|M(p')\rangle=\Psi(x^2,xp',p'^2=M^2).
121: \end{eqnarray}
122: As a function of the variable $xp'$, the amplitude may be represented by the Fourier integral
123: \begin{eqnarray}
124: \label{nak1}
125: \Psi_{\rm BS}(x,p')=
126: \int\limits_0^1 d\xi \exp({-i\xi p'x})K(x^2,\xi),
127: \end{eqnarray}
128: where the $\xi$-integration runs from $0$ to $1$ as follows from the analytic properties of
129: Feynman diagrams. Nakanishi proposed to parametrize the kernel
130: $K(x^2,\xi)$ as \cite{nakanishi}
131: \begin{eqnarray}
132: \label{nak2}
133: K(x^2,\xi)=\frac{1}{(2\pi)^4 i}\int\limits_0^\infty dz\, G(z,\xi)
134: \int d^4k'
135: \frac{\exp({-ik'x})}{[\,z+m^2-\xi(1-\xi)M^2-k'^2-i0]^3},
136: \end{eqnarray}
137: where the function $G(z,\xi)$ has no singularities in the integration regions in $z$ and $\xi$.
138: The BS equation for the function $\Psi_{\rm BS}$ leads to an equation for
139: the function $G(z,\xi)$.
140:
141: The $k'$-integral in (\ref{nak2}) is the second derivative w.r.t.\ $\mu^2$ of the
142: Feynman propagator of a scalar particle of mass $\mu$
143: [with $\mu^2=z+m^2-\xi(1-\xi)M^2$] in coordinate space. Making use of the explicit expression for
144: this propagator \cite{bogolyubov}, we obtain the following expansion near the light cone $x^2=0$:
145: \begin{eqnarray}
146: \label{prop}
147: \frac{1}{(2\pi)^4 i}\int d^4k'\exp({-ik'x})
148: \frac{1}{(\mu^2-k'^2-i0)^3}&=&
149: \frac12\frac{\partial^2}{(\partial \mu^2)^2}
150: \left\{\frac{1}{(2\pi)^4 i}\int d^4k'\exp({-ik'x})
151: \frac{1}{\mu^2-k'^2-i0}\right\}
152: \nonumber\\
153: &=&\frac{1}{32\pi^2 \mu^2}+\frac{x^2}{128\pi^2}\left[1-2\gamma_E-\log\left({-\mu^2 x^2}/{4}\right)\right]+
154: O(x^4).
155: \end{eqnarray}
156: Interestingly, this is not a pure power series but it contains also terms involving $\log(-x^2)$.
157: Inserting (\ref{prop}) into (\ref{nak2}) leads to the light-cone expansion of $K(x^2,\xi)$:
158: \begin{eqnarray}
159: \label{logseries}
160: K(x^2,\xi)=g_0(\xi)&+&x^2 \left[\,g_1(\xi)+\log(-x^2 m^2)h_1(\xi)\right]
161: +x^4 \left[\,g_2(\xi)+\log(-x^2 m^2)h_2(\xi)\right] +\cdots.
162: \end{eqnarray}
163: The functions $g_n$ and $h_n$ here may be expressed in terms of $G(z,\xi)$.
164: For instance,
165: \begin{eqnarray}
166: \label{2.6}
167: g_0(\xi)&=&\frac{1}{32\pi^2}\int\limits_0^\infty dz\, G(z,\xi) \frac{1}{z+m^2-\xi(1-\xi)M^2},
168: \nonumber\\
169: g_1(\xi)&=&\frac{1}{128\pi^2}\int\limits_0^\infty dz\, G(z,\xi)
170: \left[\,1-2\gamma_E-\log\left(\frac{z+m^2-\xi(1-\xi)M^2}{4m^2}\right)\right],
171: \nonumber\\
172: h_1(\xi)&=&-\frac{1}{128\pi^2}\int\limits_0^\infty dz\, G(z,\xi).
173: \end{eqnarray}
174: The presence of the logarithmic terms leads to complications:
175: Let us go back to Eq.~(\ref{nak2}) and expand the exponential
176: $\exp({-ik'x})$. Only the first term corresponding to $x^2=0$ (the light-cone) is finite, whereas all
177: higher terms lead to divergent $k'$-integrals; this is a consequence of the presence of
178: the terms $\log(-x^2)$. After performing the Wick rotation in the $k'$-space, we may cut
179: the $k'_E$-integration at $k_E'^2=\Lambda^2$.
180: In this way, we obtain a regularized pure {\it power} series in $x^2$ for $K(x^2,\xi)$
181: and thus for $\Psi_{\rm BS}(x,p')$.
182: The introduction of the cutoff $\Lambda$ in the momentum-space integrals is equivalent to the
183: introduction of a regulator $\lambda$ in coordinate space:
184: \begin{eqnarray}
185: \log(-x^2 m^2)\to \log(\lambda-x^2 m^2).
186: \end{eqnarray}
187: The initial expansion (\ref{logseries}) is reproduced by setting $\lambda\to 0$.
188: For a nonzero $\lambda$, we can represent $\log(\lambda-x^2 m^2)$ as a power series in
189: $x^2$ and insert this expansion in (\ref{logseries}), obtaining
190: \begin{eqnarray}
191: K(x^2,\xi)=\phi_0(\xi)+x^2 \phi_1(\xi,\lambda)+x^4 \phi_2(\xi,\lambda)+\cdots,
192: \end{eqnarray}
193: with the first three distribution amplitudes $\phi_n$ given by
194: \begin{eqnarray}
195: \label{2.9}
196: \phi_0(\xi)&=& g_0(\xi), \nonumber\\
197: \phi_1(\xi,\lambda)&=& g_1(\xi)+\log(\lambda)h_1(\xi), \nonumber\\
198: \phi_2(\xi,\lambda)&=& g_2(\xi)+\log(\lambda)h_2(\xi)-\frac{1}{\lambda}m^2 h_1(\xi).
199: \end{eqnarray}
200: As a result, instead of the series expansion for the
201: BS wave function in terms of powers and logarithms of $x^2$ with finite coefficients, we obtained
202: a pure power series, in which only the first term, corresponding to $x^2=0$, is finite,
203: whereas all higher distribution amplitudes depend on the regulator parameter $\lambda$ and
204: become infinite in the limit $\lambda\to 0$ ($\Lambda\to\infty$).
205: The regularised LC expansion of the BS amplitude reads
206: \begin{eqnarray}
207: \label{lcexpansion}
208: \Psi_{\rm BS}(x,p')=\sum_{n=0}^{\infty}(x^2)^n\int\limits_0^1
209: d\xi \exp({-i p'x\xi})\phi_n(\xi,\lambda).
210: \end{eqnarray}
211: Notice that in our case the full BS amplitude does not depend on the
212: regulator parameter $\lambda$.
213: However, any truncation of the series at some $n$ leads to an explicit
214: dependence of $\Psi_{\rm BS}(x,p)$ on $\lambda$ in
215: such a way, that it diverges as $\lambda\to 0$.
216:
217: The BS amplitude in momentum space reads
218: \begin{eqnarray}
219: \label{bsmom}
220: \Psi_{\rm BS}(k,p')&=&\int d^4x \exp({ikx})\Psi_{\rm BS}(x,p')=%\nonumber\\&=&
221: \frac{1}{i}\int\limits_0^1 d\xi \int\limits_0^\infty dz\,
222: \frac{G(z,\xi)}{[\,z+m^2-\xi(1-\xi)M^2-(k-\xi p')^2-i0]^3}.
223: \end{eqnarray}
224: The kernel $G(z,\xi)$ determines now the bound-state properties. For instance,
225: the decay constant of a bound state $M$ is given by the relation
226: \begin{eqnarray}
227: f_M=\langle 0|\varphi(0)\varphi(0)|M(p')\rangle= \Psi_{\rm BS}(x=0,p').
228: \end{eqnarray}
229: Performing the $k'$-integration in Eq.~(\ref{nak2}) for $x=0$ leads to
230: \begin{eqnarray}
231: f_M=\frac1{32\pi^2}
232: \int\limits_0^1 d\xi \int\limits_0^\infty dz
233: \frac{G(z,\xi)}{z+m^2-\xi(1-\xi)M^2}.
234: \end{eqnarray}
235: The light-cone wave function is related to the kernel $G(z,\xi)$ as follows
236: \cite{karmanov}:
237: \begin{eqnarray}
238: \label{2.14}
239: \Psi_{\rm LC}(\xi,k_\perp)=\xi(1-\xi)\int\limits_0^\infty dz\frac{G(z,\xi)}
240: {\left[\,z+k_\perp^2+m^2-\xi(1-\xi)M^2\right]^2}.
241: \end{eqnarray}
242: Respectively, the distribution amplitudes $\phi_n$ may be calculated as $k_\perp^2$-moments of
243: $\Psi_{\rm LC}(\xi,k_\perp)$.
244: For the function (\ref{2.14}), only the zero $k_\perp^2$-moment leading to $\phi_0$ is
245: finite, all higher moments require a cut-off in $k_\perp^2$ precisely as in (\ref{2.9}).
246:
247:
248: For the case of an interaction between the constituents via exchange of a massless boson,
249: the solution to the BS equation in the ladder
250: approximation takes a simple form \cite{karmanov}:
251: \begin{eqnarray}
252: G(z,\xi)=\delta(z)G(\xi), \qquad G(\xi)=\xi(1-\xi)f(\xi).
253: \end{eqnarray}
254: The function $f(\xi)$ here is a smooth function which takes finite nonzero values at
255: the end-points. Only the end-point behavior is crucial
256: for the heavy-to-light correlator, therefore the precise form of the function $f(\xi)$ and its normalization
257: are not essential for our analysis; of importance for us is only the fact that
258: $f(\xi=0)\ne 0$. Hereafter we just set $G(\xi)=m^2\xi(1-\xi)$ and do not care about
259: the overall factor $\propto f(0)$. This factor turns out to be the same in the full and the light-cone
260: correlators which we calculate and compare in the next section.
261:
262: Closing this section, let us emphasize that the $\delta$-functional $z$-dependence of the kernel $G(z,\xi)$
263: leads to a nontrivial $k_\perp^2$-dependence of the light-cone wave function
264: \begin{eqnarray}
265: \label{2.16}
266: \Psi_{\rm LC}(\xi,k_\perp)=
267: \frac{\xi(1-\xi)G(\xi)}{\left[\,k_\perp^2+m^2-\xi(1-\xi)M^2\right]^2}.
268: \end{eqnarray}
269: The $1/k_\perp^4$ tail of the bound-state LC wave function is a consequence of the
270: massless-boson exchange (and is similar in this respect to the power-like behavior of the pion
271: light-cone wave function in QCD).
272:
273: %\newpage
274: \section{\label{sect_ff}
275: Heavy-to-light correlator and the decay form factor}
276: In this section, we consider the correlator
277: \begin{eqnarray}
278: \Gamma(p^2,q^2)=i \int d^4x \exp({ipx})\langle 0|T \varphi(x)Q(x) Q(0)\varphi(0)|M(p')\rangle
279: \end{eqnarray}
280: and its Borel transform in the variable $p^2$ relevant for the
281: extraction of the $M_Q\to M$ form factor.
282: Since $p'^2=M^2$ is fixed, this correlator depends on the two variables $p^2$ and $q^2$.
283:
284: \subsection{Borel sum rule for $\Gamma(p^2,q^2)$}
285: Making use of the hadronic degrees of freedom, by inserting the complete
286: system of hadron states into the correlator, we obtain the so-called phenomenological
287: representation for
288: $\Gamma(p^2,q^2)$, which we denote by $\Gamma_{\rm phen}(p^2,q^2)$:
289: \begin{eqnarray}
290: \label{phen}
291: \Gamma_{\rm phen}(p^2,q^2)&=&
292: \frac{1}{(2\pi)^4}\int d^4x \exp({ipx})\langle 0|\varphi(x)Q(x)|M_Q(\tilde p)\rangle
293: \frac{d^4\tilde p}{M_Q^2-\tilde p^2-i0}
294: \langle M_Q(\tilde p)|Q(0)\varphi(0)|M(p')\rangle+
295: {\rm hadr.\, cont.}
296: \nonumber\\
297: &=&F_{M_Q\to M}(q^2)\frac{1}{M_Q^2-p^2-i0}f_{M_Q}+\int\limits_{s_{\rm cont}}^\infty \frac{ds}{s-p^2-i0}
298: \Delta_{\rm phen}(s,q^2),
299: \end{eqnarray}
300: where $s_{\rm cont}$ is the threshold of the hadronic continuum, containing the
301: $M_Q$-meson plus light mesons,
302: i.e.,
303: $M_Q M$, $M_QMM$, etc.\ states. Respectively, $s_{\rm cont}=(M_Q+M)^2$.
304: The decay constant $f_{M_Q}$ of the $M_Q$-meson is defined as
305: \begin{eqnarray}
306: f_{M_Q}=\langle 0|T \varphi(0)Q(0)|M_Q(p)\rangle,
307: \end{eqnarray}
308: and the form factor is given by
309: \begin{eqnarray}
310: F_{M_Q\to M}(q^2)=\langle M_Q(p)|Q(0)\varphi(0)|M(p')\rangle, \qquad q\equiv p-p'.
311: \end{eqnarray}
312: Eq.~(\ref{phen}) gives the single dispersion
313: representation for $\Gamma(p^2,q^2)$ in
314: terms of the hadron degrees of freedom.
315:
316: One can calculate the correlator $\Gamma(p^2,q^2)$ also by making
317: use of the underlying field-theoretic degrees of freedom and obtain in this way a different ---
318: theoretical --- expression for $\Gamma(p^2,q^2)$, which we call $\Gamma_{\rm th}(p^2,q^2)$.
319: This expression may also be written as a single dispersion representation in $p^2$:
320: \begin{eqnarray}
321: \label{disp_th}
322: \Gamma_{\rm th}(p^2,q^2)=\int \frac{ds}{s-p^2-i0}\Delta_{\rm th}(s,q^2).
323: \end{eqnarray}
324: Quark--hadron duality \cite{svz,shifman} postulates that $\Gamma_{\rm th}$ and $\Gamma_{\rm phen}$ are equal to each other
325: after a proper smearing is applied to both of them.
326: The smearing may be implemented, e.g., by applying the Borel transform \cite{svz}.
327: Calculating the Borel image of $\Gamma_{\rm th}$ and $\Gamma_{\rm phen}$ with the parameter
328: $2\mu_B^2$, such that $1/(s-p^2)\to \exp(-s/2\mu_B^2)$, we obtain the sum rule for $\Gamma(p^2,q^2)$:
329: \begin{eqnarray}
330: \label{sr_1}
331: f_{M_Q}\,F_{M_Q\to M}(q^2) \exp\left({-M_Q^2/2\mu_B^2}\right)+
332: \int {ds}\,\theta(s-s_{\rm cont})\exp\left({-s/2\mu_B^2}\right)\Delta_{\rm phen}(s,q^2)=
333: \int {ds}\exp\left({-s/2\mu_B^2}\right)\Delta_{\rm th}(s,q^2).\nonumber\\
334: \end{eqnarray}
335: Let us ignore for a moment the dependence of $\Delta$ on $q^2$ and
336: introduce the continuum subtraction point $s_0$
337: (an effective continuum threshold, different from the physical continuum threshold $s_{\rm cont}$)
338: according to
339: \begin{eqnarray}
340: \label{s0}
341: \int ds\,\theta(s-s_0) \exp\left({-s/2\mu_B^2}\right)\Delta_{\rm th}(s)=
342: \int ds\,\theta(s-s_{\rm cont})\exp\left({-s/2\mu_B^2}\right) \Delta_{\rm phen}(s).
343: \end{eqnarray}
344: For sufficiently large $s$, one expects $\Delta_{\rm th}(s)=\Delta_{\rm phen}(s)$
345: (up to oscillating terms \cite{shifman}).
346: The region near the continuum threshold $s\simeq s_{\rm cont}$ is
347: clearly below the duality region: recall that $\Delta_{\rm phen}(s_{\rm cont})=0$, while
348: $\Delta_{\rm th}(\rm s_{cont})>0$. Thus, in the vicinity of $s_{\rm cont}$,
349: $\Delta_{\rm th}>\Delta_{\rm phen}$.
350: As a result, the continuum subtraction $s_0$ as defined by (\ref{s0}) depends on the Borel parameter
351: (and on $q^2$)
352: and lies above the physical threshold: $s_0(\mu_B^2)>s_{\rm cont}$.
353: Then the sum rule (\ref{sr_1}) takes the form
354: \begin{eqnarray}
355: \label{srff}
356: f_{M_Q}\,F_{M_Q\to M}(q^2)=
357: \exp\left({ M_Q^2/2\mu_B^2}\right)\hat \Gamma_{\rm th}(\mu_B^2,q^2,s_0(\mu_B^2,q^2)),
358: \end{eqnarray}
359: with the correlator to which the $s$-cut is applied (hereafter referred to as the cut correlator)
360: \begin{eqnarray}
361: \label{3.9}
362: \hat \Gamma_{\rm th}(\mu_B^2,q^2,s_0)=
363: \int ds\, \theta(s<s_0)\,\exp\left({-s/2\mu_B^2}\right)\Delta_{\rm th}(s,q^2).
364: \end{eqnarray}
365: If one had known the $\mu_B$- and $q^2$-dependent solution $s_0(\mu_B^2,q^2)$ to (\ref{s0}),
366: which encodes the full nonperturbative dynamics,
367: the expression (\ref{srff}) would have been exact. Taking $s_0$ constant makes it approximate.
368: One may then impose the stability criterium which is expected to guarantee the extraction of the
369: true form factor in several different ways (see, e.g., \cite{svz,jamin,tk}).
370:
371: We shall now concentrate on different possibilities to calculate $\Gamma_{\rm th}(p^2,q^2)$ and
372: compare the corresponding results.
373:
374:
375:
376: \begin{figure}[t]
377: \begin{center}
378: \includegraphics[width=12cm]{correlator.eps}
379: \end{center}
380: \caption{\label{Fig:1}Feynman diagrams for the correlator $\Gamma(p,q)$.
381: The normal and bold lines denote the heavy $Q$ and the light $\varphi$ particles,
382: respectively, the wavy line corresponds to the scalar particle which
383: mediates the $Q$--$\varphi$ interaction (a scalar analog of the gluon in QCD);
384: $|M(p')\rangle$ is the state-vector of the $\varphi\varphi$ bound-state with the momentum $p'$.
385: The contribution of the first diagram is expressable through the BS wave function of the light
386: meson and the full $Q$-propagator. In the region of not very large $q$, $q^2\ll m_Q^2$,
387: the second and the third terms are suppressed by higher powers of the heavy-quark mass ($1/m_Q$ and $1/m_Q^2$, respectively)
388: compared to the first term and will not be considered.}
389: \end{figure}
390:
391:
392:
393:
394:
395:
396:
397: \subsection{Theoretical calculation of $\Gamma(p^2,q^2)$}
398: We now calculate this correlator using the underlying field-theoretic degrees of freedom,
399: and denote the result as $\Gamma_{\rm th}$. The corresponding Feynman diagrams are shown in
400: Fig.~\ref{Fig:1}. In the region of not too large timelike momentum transfer $q$, $q^2\ll m_Q^2$, the first diagram gives the main contribution whereas
401: the contributions of the other diagrams are suppressed by higher powers of the heavy-particle
402: mass $m_Q$. Neglecting the contributions of the power-suppressed diagrams leads to
403: \begin{eqnarray}
404: \label{3.5a}
405: \Gamma_{\rm th}(p^2,q^2)=i\int dx \exp({ipx}){\cal D}_{Q}(x)
406: \langle 0|T \varphi(x)\varphi(0)|M(p')\rangle,
407: \end{eqnarray}
408: where ${\cal D}_Q$ is the full propagator of the heavy quark.
409: Approximating the full propagator with the free propagator gives
410: \begin{eqnarray}
411: \label{3.5}
412: \Gamma_{\rm th}(p^2,q^2)=\frac1{(2\pi)^4}\int d^4k d^4x \exp\left({ipx-ikx}\right)\frac{1}{m_Q^2-k^2-i0}
413: \langle 0|T \varphi(x)\varphi(0)|M(p')\rangle.
414: \end{eqnarray}
415: We can proceed further in two different ways:
416:
417:
418:
419: A. First, the correlator (\ref{3.5}) may be expressed in terms of the BS amplitude
420: in momentum space
421: \begin{eqnarray}
422: \Gamma_{\rm th}(p^2,q^2)=\frac1{(2\pi)^4}\int d^4k
423: \frac{\Psi_{\rm BS}(k,p')}{m_Q^2-(p-k)^2-i0}.
424: \end{eqnarray}
425: Making use of (\ref{bsmom}) gives
426: \begin{eqnarray}
427: \Gamma_{\rm th}(p^2,q^2)=\frac1{(2\pi)^4 i}
428: \int\limits_0^1 d\xi \int\limits_0^\infty dz ~G(z,\xi)
429: \int \frac{d^4k}{\left[{z+m^2-\xi(1-\xi)M^2-(k-\xi p')^2-i0}\right]^3
430: [m_Q^2-(p-k)^2-i0]}.
431: \end{eqnarray}
432: Introducing the Feynman parameter $\alpha$ and performing the $k$-integration, we find
433: \begin{eqnarray}
434: \label{3.6}
435: \Gamma_{\rm th}(p^2,q^2)&=&\frac1{32\pi^2}
436: \int\limits_0^1 d\xi \int\limits_0^\infty dz ~G(z,\xi)\int\limits_0^1d\alpha (1-\alpha)^2
437: \frac{1}{\left[\mu^2(1-\alpha)+m_Q^2 \alpha -P^2\alpha(1-\alpha)\right]^2},\nonumber\\
438: &&\mu^2\equiv z+m^2-\xi(1-\xi)M^2,\qquad P^2=p^2(1-\xi)-\xi(1-\xi)M^2 +q^2\xi.
439: \end{eqnarray}
440: From this expression, it is straightforward to obtain the single dispersion representation for
441: $\Gamma_{\rm th}$ in the form
442: \begin{eqnarray}
443: \label{disp_th_2}
444: \Gamma_{\rm th}(p^2,q^2)=\int\limits_{(m_Q+m)^2}^\infty \frac{ds}{(s-p^2-i0)^2}\sigma_{\rm th}(s,q^2),
445: \end{eqnarray}
446: with
447: \begin{eqnarray}
448: \sigma_{\rm th}(s)&=&
449: \frac1{32\pi^2}
450: \int\limits_0^1 \frac{d\xi}{(1-\xi)^2} \int\limits_0^\infty dz~ G(z,\xi)
451: \int\limits_0^1 \frac{d\alpha}{\alpha^2}
452: \delta(s-s_\alpha), \nonumber\\
453: &&s_\alpha=\left(\frac{z+m^2-M^2\xi(1-\xi)}{\alpha}+\frac{m_Q^2}{1-\alpha}
454: \right)\frac{1}{1-\xi}+M^2\xi-q^2\frac{\xi}{1-\xi}.
455: \end{eqnarray}
456: The spectral density $\sigma_{\rm th}(s,q^2)$ vanishes at the lower limit of the $s$-integration at
457: $s=(m_Q+m)^2$. Notice that the quantity $m$ in the lower limit of the $s$-integration
458: is the mass of the light spectator. For more details about the analytic properties of the
459: three-point functions, we refer to \cite{lms}.
460:
461: Performing an integration by parts in (\ref{disp_th_2}), we obtain the spectral representation in the
462: ``standard'' form (\ref{disp_th}) with
463: $\Delta_{\rm th}(s,q^2)=\frac{\partial}{\partial s}\sigma_{\rm th}(s,q^2).$
464: The Borel images of the two representations (\ref{disp_th_2}) and
465: (\ref{disp_th}) read, respectively,
466: \begin{eqnarray}
467: \label{3.17}
468: \hat \Gamma_{\rm th}(\mu_B^2,q^2)=\frac{1}{2\mu_B^2}
469: \int\limits_{(m_Q+m)^2}^\infty {ds}\exp(-s/2\mu_B^2)\sigma_{\rm th}(s,q^2)
470: \end{eqnarray}
471: and
472: \begin{eqnarray}
473: \label{3.18}
474: \hat \Gamma_{\rm th}(\mu_B^2,q^2)=
475: \int\limits_{(m_Q+m)^2}^\infty {ds}\exp(-s/2\mu_B^2)\Delta_{\rm th}(s,q^2).
476: \end{eqnarray}
477: When the $s$-integration extends to infinity, the equality of both formulas is
478: demonstrated by integration by parts.
479:
480: We now want to obtain the cut Borel image of Eq.~(\ref{3.9}).
481: Recall, that the cut at $s_0$ is applied to the Borel transform of the spectral representation
482: for the correlator in the form (\ref{3.5}).
483: Then, the relation
484: \begin{eqnarray}
485: \label{surface}
486: \int\limits_{(m_Q+m)^2}^{s_0}{ds}\exp(-s/2\mu_B^2)\Delta_{\rm th}(s,q^2)=
487: \frac{1}{2\mu_B^2}\int\limits_{(m_Q+m)^2}^{s_0}{ds}\exp(-s/2\mu_B^2)\sigma_{\rm th}(s,q^2)
488: +\exp(-s_0/2\mu_B^2)\sigma_{\rm th}(s_0,q^2)%\nonumber\\
489: \end{eqnarray}
490: leads to the appearance of an additional surface term in the cut spectral representation of the form Eq.~(\ref{3.17}):
491: \begin{eqnarray}
492: \label{gammaborelfullcut}
493: \label{3.20}
494: \hat \Gamma_{\rm th}(\mu_B^2,q^2,s_0)&=&
495: \frac{1}{2\mu_B^2}\int\limits_{(m_Q+m)^2}^{s_0}{ds}\exp(-s/2\mu_B^2)\sigma_{\rm th}(s,q^2)
496: +\exp(-s_0/2\mu_B^2)\sigma_{\rm th}(s_0,q^2).
497: \end{eqnarray}
498: Moreover, if one works with the Borel image in the form (\ref{3.20}), the surface term gives the main
499: contribution to the correlator for large values of $\mu_B$.
500:
501: The sum rule for the form factor can now be written in two equivalent ways:
502: \begin{eqnarray}
503: \label{sr_full}
504: f_{M_Q} F_{M_Q\to M}(q^2)&=&
505: \frac{1}{2\mu_B^2}\int\limits_{(m_Q+m)^2}^{s_0}{ds}\exp\left(-\frac{s-M_Q^2}{2\mu_B^2}\right)
506: \sigma_{\rm th}(s,q^2)
507: +\exp\left(-\frac{s_0-M_Q^2}{2\mu_B^2}\right)\sigma_{\rm th}(s_0,q^2)\\
508: &=&
509: \int\limits_{(m_Q+m)^2}^{s_0}{ds}\exp\left(-\frac{s-M_Q^2}{2\mu_B^2}\right)\Delta_{\rm th}(s,q^2) .
510: \end{eqnarray}
511:
512:
513:
514: %\newpage
515: B. Another possibility to calculate $\Gamma_{\rm th}(p^2,q^2)$ is to substitute
516: the
517: light-cone expansion of $\Psi_{\rm BS}(x,p')$, Eq.~(\ref{lcexpansion}), into Eq.~(\ref{3.5}):
518: \begin{eqnarray}
519: \label{3.22}
520: \Gamma_{\rm th}(p^2,q^2)=
521: \frac1{(2\pi)^4}\int d^4k\, d^4x \exp\left({ipx-ikx}\right)\frac{1}{m_Q^2-k^2-i0}\sum_{n=0}^{\infty}(x^2)^n\int\limits_0^1
522: d\xi \exp({-i p'x\xi})\phi_n(\xi,\lambda).
523: \end{eqnarray}
524: The functions $\phi_i(\xi)$ are related to $G(z,\xi)$ via Eqs.~(\ref{2.6}) and (\ref{2.9}).
525: Performing the $x$ and $k$ integrations gives
526: \begin{eqnarray}
527: \label{3.23}
528: \Gamma_{\rm th}(p^2,q^2)=
529: \int\limits_0^1
530: \frac{d\xi ~\phi_0(\xi,\lambda)}{m_Q^2-(p-\xi p')^2}
531: -8m_Q^2 \int\limits_0^1\frac{d\xi ~\phi_1(\xi,\lambda)}{\left[m_Q^2-(p-\xi p')^2\right]^3}+\cdots.
532: \end{eqnarray}
533: Here $(p-\xi p')^2=p^2(1-\xi)-M^2\xi(1-\xi)+q^2\xi$.
534: This series (\ref{3.23}) may be written as the series of spectral representations
535: \begin{eqnarray}
536: \label{3.23a}
537: \Gamma_{\rm th}(p^2,q^2)=%\sum_{n=0}^\infty\Gamma^{(n)}(p^2,q^2)\equiv
538: \int\limits_{m_Q^2}^{\infty}
539: \frac{ds}{s-p^2}\Delta^{(0)}(s,q^2)+
540: \int\limits_{m_Q^2}^{\infty}\frac{ds}{(s-p^2)^3}\Delta^{(1)}(s,q^2)+\cdots,
541: \end{eqnarray}
542: with $\Delta^{(n)}$ calculable via $\phi_n$. For instance, for $q^2=0$ one finds
543: \begin{eqnarray}
544: \label{3.23b}
545: \Delta^{(0)}(s,q^2=0)=\frac{1}{m_Q^2}(1-\xi^*)\phi_0(\xi^*,\lambda),\quad
546: \Delta^{(1)}(s,q^2=0)=-8\frac{\phi_1(\xi^*,\lambda)}{1-\xi^*}, \qquad \xi^*=1-m_Q^2/s.
547: \end{eqnarray}
548: The corresponding uncut Borel image reads
549: \begin{eqnarray}
550: \label{3.24}
551: \hat \Gamma_{\rm th}(\mu_B^2,q^2)=
552: \int\limits_{m_Q^2}^{\infty}
553: ds\exp(-s/2\mu_B^2)\Delta^{(0)}(s,q^2)
554: +
555: \frac{1}{2(2\mu_B)^2}\int\limits_{m_Q^2}^{\infty}
556: ds\exp(-s/2\mu_B^2)\Delta^{(1)}(s,q^2)+\cdots,
557: \end{eqnarray}
558: or, equivalently,
559: \begin{eqnarray}
560: \label{3.24a}
561: \hat \Gamma_{\rm th}(\mu_B^2,q^2)=\int\limits_0^1 \frac{d\xi}{1-\xi}
562: \left[
563: \phi_0(\xi)-\frac{m_Q^2}{\mu_B^4}\frac{\phi_1(\xi,\lambda)}{(1-\xi)^2}+\cdots \right]
564: \exp({-s_\xi/{2\mu_B^2}}),
565: \end{eqnarray}
566: with
567: \begin{eqnarray}
568: s_\xi=\frac{m_Q^2}{1-\xi}+M^2\xi -q^2 \frac{\xi}{1-\xi}.
569: \end{eqnarray}
570: In (\ref{3.24}) and (\ref{3.24a}) the dots stand for terms corresponding to higher $n$.
571: Clearly, their contributions are suppressed with powers of the Borel parameter $\mu_B$ and vanish in the limit
572: $\mu_B\to\infty$.
573:
574: Usually, the Borel parameter has the following dependence on the heavy quark mass: $\mu_B^2=m_Q\beta$ \cite{ck},
575: where $\beta$ stays finite in the limit $m_Q\to\infty$. Then, the terms corresponding to higher $n$ are not
576: suppressed at $m_Q\to\infty$.
577:
578: Now, let us consider the cut Borel image of Eq.~(\ref{3.23a}).
579: As explained above, the cut is applied to the dispersion representation
580: in the form (\ref{3.5}). Therefore, if we want to work with the spectral densities $\Delta^{(i)}$,
581: we should take into account the surface terms, similar to Eq.~(\ref{3.20}).
582: The cut Borel image takes the form
583: \begin{eqnarray}
584: \label{cutlc}
585: \hat \Gamma_{\rm th}(\mu_B^2,q^2,s_0)&=&
586: \int\limits_{m_Q^2}^{s_0}
587: {ds}\exp(-s/2\mu_B^2)\Delta^{(0)}(s,q^2)
588: \\ \nonumber
589: &&+
590: \int\limits_{m_Q^2}^{s_0}
591: \frac{ds\exp(-s/2\mu_B^2)\Delta^{(1)}(s,q^2)}{2(2\mu_B)^2}
592: +
593: \left[
594: \frac{\partial\Delta^{(1)}(s_0,q^2)}{\partial s_0}
595: +
596: \frac{\Delta^{(1)}(s_0,q^2)}{(2\mu_B)^2}\right]\frac{\exp(-s_0/2\mu_B^2)}{2}
597: \\
598: \nonumber
599: &&+\mbox{contributions of terms corresponding to higher $n$}.
600: \end{eqnarray}
601: %The cut value $s_0^{\rm LC}$ is related to $s_0$ but may not coincide with it since
602: %the integration variables in the spectral representations (\ref{disp_th}) and (\ref{3.23a})
603: %are different from each other. The explicit dependence $s_0^{\rm LC}(s_0)$ unfortunatley could not be obtained.
604: %
605: Making use of the distribution amplitudes $\phi_n$, we rewrite the correlator (\ref{cutlc}) as
606: \begin{eqnarray}
607: \label{cutlc1}
608: \hat \Gamma_{\rm th}(\mu_B^2,q^2,s_0)&=&\int\limits_0^1 \frac{d\xi}{1-\xi}
609: \left[
610: \phi_0(\xi)-\frac{m_Q^2}{\mu_B^4}\frac{\phi_1(\xi,\lambda)}{(1-\xi)^2}+\cdots \right]
611: \exp({-s_\xi/{2\mu_B^2}})\theta(s_\xi <s_0)\nonumber\\
612: &-&4\exp(-s_0/{2\mu_B^2})
613: \left[
614: \frac{\phi_1(\xi_0)}{m_Q^2}
615: +\frac{\phi'_1(\xi_0)}{4\mu_B^2(1-\xi_0)}+\frac{\phi'_1(\xi_0)}{s_0}
616: \right]+\cdots,
617: \end{eqnarray}
618: where $\xi_0=1-m_Q^2/s_0$ and $\cdots$ stand for contributions of the terms corresponding
619: to $n\ge 2$.
620:
621: The light-cone correlator corresponds to the first term in the
622: expansions (\ref{cutlc}) and (\ref{cutlc1}), calculated with the cut parameter
623: $s_0^{\rm LC}$ specific for the LC approximation (see the next section for details):
624: \begin{eqnarray}
625: \label{gammaborellc}
626: \hat \Gamma_{\rm LC}(\mu_B^2,q^2,s_0^{\rm LC})=
627: \int\limits_{m_Q^2}^{s_0^{\rm LC}}
628: {ds}\exp(-s/2\mu_B^2)\Delta^{(0)}(s,q^2)%\nonumber\\
629: =\int\limits_0^1 \frac{d\xi}{1-\xi}
630: \phi_0(\xi)\exp({-s_\xi/{2\mu_B^2}})\theta(s_\xi <s_0^{\rm LC}).
631: \end{eqnarray}
632: Keeping only this term leads to the following sum rule:
633: \begin{eqnarray}
634: \label{sr_lc}
635: f_{M_Q} F_{M_Q\to M}(q^2)= \int\limits_0^1 \frac{d \xi}{1-\xi}
636: \phi_0(\xi)\exp\left(-\frac{s_\xi-M_Q^2}{2\mu_B^2}\right)\theta(s_\xi <s_0^{\rm LC} ).
637: \end{eqnarray}
638: Let us study the conditions under which the contributions of higher $n$ are parametrically
639: suppressed compared to the contribution of the light-cone $n=0$.
640:
641: 1. The heavy-meson mass is related to the heavy quark mass by $M_Q=m_Q+\varepsilon_Q$,
642: $\varepsilon_Q\simeq \Lambda_{\rm QCD}$.
643: One can then check that the dominant contribution to the integral comes from the end-point
644: region $\xi\sim \Lambda_{\rm QCD}/m_Q$. Therefore, in order to have contributions of higher $n$ suppressed by powers of
645: $1/m_Q$ compared to the $n=0$ contribution, one needs, e.g., the following end-point behavior:
646: %\vspace{-.3cm}
647: \begin{eqnarray}
648: \label{nb}
649: \phi_n(\xi)\sim \xi^{n+1}.
650: \end{eqnarray}
651: Such a behavior can be obtained, e.g., for the light-cone wave function
652: \begin{eqnarray}
653: \Psi_{\rm LC}(\xi,k_\perp^2)=\exp\left(-{k_\perp^2}/{2\beta_M^2\xi(1-\xi)}\right),
654: \end{eqnarray}
655: $\beta_M$ being the size parameter of the light meson.
656: In this specific case the functions $\phi_n(\xi)$ have the necessary behavior (\ref{nb}) in the
657: end-point region.
658: However, for realistic wave functions obtained
659: as solutions to the BS equation, the distribution amplitudes $\phi_n(\xi)$ for different $n$ have a similar behavior
660: in the end-point region and therefore the LC expansion for the form factor has no small parameter.
661: In this case, the term $n=0$ is not parametrically enhanced compared to terms of higher $n$, and,
662: in order to
663: obtain the form factor, terms for all $n$ should be summed.
664: In other words, the transverse motion of the light
665: quark is essential, the longitudinal light-cone distribution amplitude $g_0(\xi)$ is not sufficient, and
666: the knowlegde of the full wave function $\Psi(\xi,k_\perp)$ is necessary to obtain the $B\to M$ form factor.
667:
668: 2. One might expect to have a power suppression of the off-LC terms corresponding to $n\ge 1$
669: in the limit $\mu_B\to\infty$, within the ``local-duality'' sum rule \cite{radyushkin},
670: taking into account that in standard SVZ sum rules the choice $\mu_B\to\infty$ suppresses
671: condensate contributions.
672: Within the context of light-cone sum rules, the limit $\mu_B\to\infty$ is, however, tricky:
673: In the uncut light-cone correlator the limit $\mu_B\to\infty$ indeed suppresses the off-LC
674: effects ($x^2\ne 0$ terms). However, after applying the cut, this property is lost: obviously,
675: the surface terms in the cut correlator (\ref{cutlc}) corresponding to $n\ge 1$ remain finite in
676: the limit $\mu_B\to\infty$ and can give a sizeable contribution.
677: There is also another subtlety: A specific feature of the local-duality sum rule is that the
678: observables are determined
679: to great extent by the value of the continuum subtraction point $s_0$.
680: Therefore one cannot apply the
681: standard sum-rule stability criteria to extract the physical value of an observable.
682: Nevertheless, {\it assuming} the duality interval to be process-independent and
683: fixing $s_0$ from one observable allows one to predict other
684: observables.\footnote{We have shown recently that the local-duality sum rule
685: should provide a good description of the pion elastic form factor for
686: all spacelike momentum transfers \cite{lm}.} However, as we shall see in the next section,
687: when the standard criterion to fix the continuum subtraction $s_0$
688: is applied, its value for the full and for the LC correlators turn out to be different from each
689: other.
690:
691:
692: %\footnote
693:
694: Finally, we conclude that for the realistic case of the interaction dominated by one-boson
695: exchange at short distances,
696: the LC contribution does not dominate the cut correlator parametrically, i.e., the off-LC terms
697: are not suppressed by any large parameter compared to the LC term.
698: To understand how well the LC contribution numerically compares with the full result,
699: we calculate the correlator $\hat \Gamma_{\rm th}$, Eq.~(\ref{gammaborelfullcut}), and the
700: LC correlator, Eq.~(\ref{gammaborellc}), making use of the BS wave function obtained as
701: solution to the BS equation with a realistic potential dominated by a
702: one-boson exchange at small separations.
703:
704:
705: %\newpage
706:
707:
708: %\newpage
709: \section{\label{numerical}Numerical results}
710: In this section we address the case $q^2=0$, and therefore do not write explicitly the argument $q^2$.
711: For the BS kernel $G(z,\xi)=m^2 \delta(z)\xi(1-\xi)$ it is straightforward to calculate
712: the spectral densities $\Delta_{\rm th}$ and $\Delta_{\rm LC}$.
713: We shall analyse the correlators for the light meson in the final state.
714: We may therefore set $M=0$ and find
715: \begin{eqnarray}
716: \Delta_{\rm th}(s)&=&\frac{1}{s^3}\left[
717: (m_Q^2+m^2)\lambda^{1/2}(s,m_Q^2,m^2)+4 m_Q^2 m^2
718: \log\left(\frac{s-m_Q^2-m^2+\lambda^{1/2}(s,m_Q^2,m^2)}{2m_Q m}\right)
719: \right]\theta(s-(m_Q+m)^2),\nonumber\\
720: \Delta_{\rm LC}(s)&=&\frac{m_Q^2(s-m_Q^2)}{s^3}\theta(s-m_Q^2).
721: \end{eqnarray}
722: Here $\lambda(s,m_Q^2,m^2)=(s-m_Q^2-m^2)^2-4 m_Q^2 m^2$ and $\Delta_{\rm LC}(s)$ is just
723: $\Delta^{(0)}(s,q^2=0)$ of Eq.~(\ref{3.23a}).
724: Clearly, in the limit $m\to 0$, both quantites coincide.
725: Notice the relation
726: \begin{eqnarray}
727: \label{relbor}
728: \int\limits_{(m_Q+m)^2}^{\infty} ds\,\Delta_{\rm th}(s)=
729: \int\limits_{m_Q^2}^{\infty} ds\,\Delta_{\rm LC}(s),
730: \end{eqnarray}
731: which can be checked explicitly.
732:
733: Let us now discuss the values of the parameters to be used in numerical estimates.
734:
735: For the heavy-quark mass the situation is obvious: to study the beauty-meson decay, we choose
736: the value $m_Q=4.8$ GeV used in light-cone sum-rules \cite{bb,bz}. The same value
737: is employed in quark-model calculations \cite{stech}. To describe decay of the charm meson,
738: we set $m_Q=1.4$ GeV according to sum rules \cite{bp}.
739:
740: The relevant choice of the light-quark mass parameter $m$ requires clarification.
741: %We could have taken the values from \cite{karmanov} and obtain the numerical estimates for the correlators.
742: Our interest is to understand with the help of the
743: simplified model under consideration the corresponding QCD calculation. We therefore
744: choose the numerical parameters relevant for QCD.
745: Since the light-quark mass parameter $m$ appears in the framework of the BS equation, it should be
746: understood as the effective quark mass which takes into account nonperturbative effects related to
747: spontaneous chiral symmetry breaking in the soft region, i.e., the constituent quark mass.
748: In \cite{ms,msl}, the constituent quark mass was calculated through the
749: quark condensate in QCD with the result $m\simeq 220$ MeV at the chiral-symmetry breaking scale
750: $1$ GeV. The scale-dependence of the constituent mass of the light quark above 1 GeV was also
751: reported in \cite{ms}. In sum-rule analyses \cite{bb,bz} it was found, that the relevant infrared
752: factorization scale for a heavy-meson decay is $\sqrt{M_Q^2-m_Q^2}$. We therefore make use of the
753: constituent quark mass evaluated at this scale. Employing the results from \cite{ms}, we find
754: the relevant value of the constituent mass of the light quark $m=150$ MeV for beauty-meson decay, and
755: $m=200$ MeV for charm-meson decay. These values of the quark masses will be used in numerical estimates
756: below.
757:
758: \begin{figure}[b]
759: \begin{center}
760: \begin{tabular}{lr}
761: \includegraphics[width=8cm]{Fig_1_a.eps}
762: &
763: \includegraphics[width=8cm]{Fig_3_a.eps}
764: \end{tabular}
765: \caption{\label{Fig:2}Spectral densities $m_Q^2\Delta_{\rm th}(s)$ (solid red line) and $m_Q^2\Delta_{\rm LC}(s)$ (dashed blue line)
766: for parameter values corresponding to beauty-meson decay,
767: $m_Q=4.8$ GeV, $m=150$ MeV (left), and to charm-meson decay, $m_Q=1.4$ GeV, $m=200$ MeV (right).}
768: \end{center}
769: \end{figure}
770:
771: Fig.~\ref{Fig:2} shows the quantities $m_Q^2\Delta_{\rm th}$ and $m_Q^2\Delta_{\rm LC}$
772: for two cases: (a) $m_Q=4.8$ GeV and $m=150$ MeV relevant for $B$-decay and (b)
773: $m_Q=1.4$ GeV and $m=200$ MeV relevant for $D$-decay.
774: Important for the following is that the thresholds in $\Delta_{\rm th}$ and $\Delta_{\rm LC}$
775: do not coincide: in the light-cone spectral density the threshold is $m_Q^2$ whereas in the
776: full spectral density it is $(m_Q+m)^2$. The region near the threshold
777: provides the main contribution to the cut Borel-transformed correlators, therefore
778: the mismatch of the thresholds is responsible for the nonvanishing of the off-light-cone
779: effects in sum rules.
780:
781: Let us briefly address the {\it uncut} Borel-transformed full and LC correlators
782: \begin{eqnarray}
783: \hat \Gamma_{\rm th}(\mu_B^2)=\int\limits_{(m_Q+m)^2}^{\infty} ds\,\exp(-s/2 \mu_B^2)\,\Delta_{\rm th}(s),\qquad
784: %\nonumber\\
785: \hat \Gamma_{\rm LC}(\mu_B^2)=\int\limits_{m_Q^2}^{\infty} ds\,\exp(-s/2 \mu_B^2)\Delta_{\rm LC}(s).
786: \end{eqnarray}
787: Eq.~(\ref{relbor}) shows that for large values of the Borel mass $\mu_B$ these two quantites coincide.
788: The same conclusion may be obtained directly from Eq.~(\ref{3.24}):
789: the terms proportional to the off-LC distribution amplitudes are suppressed by powers of the Borel
790: parameter $\mu_B^2$. Thus, in the case of the uncut Borel transform there is a clear limit --- large
791: values of the Borel parameter --- in which the off-LC effects vanish, independently of the
792: specific value of the light-quark mass. However, the uncut Borel-transformed correlators contain
793: contributions of all possible hadronic states containing the heavy quark, and therefore
794: cannot provide information on the properties of the single heavy meson.
795:
796: %\newpage
797: %B-meson
798: \begin{figure}[t]
799: \begin{center}
800: \begin{tabular}{ll}
801: \includegraphics[width=8.5cm]{Fig_1_c.eps}&\includegraphics[width=8.5cm]{Fig_2_c.eps}\\
802: \includegraphics[width=8.5cm]{Fig_1_d.eps}&\includegraphics[width=8.5cm]{Fig_2_d.eps}\\
803: \hspace{.3cm}\includegraphics[width=8.5cm]{Fig_1_e.eps}&\hspace{.3cm}\includegraphics[width=8.5cm]{Fig_2_e.eps}
804: \end{tabular}
805: \caption{\label{Fig:3}Plots for the parameters corresponding to beauty-meson decay
806: $m_Q=4.8$ GeV, $m=150$ MeV and $\delta$ fixed by tuning
807: $\sqrt{\langle s \rangle}$ to $M_Q=5.27$ GeV at two different values of $\beta$:
808: The plots in the left column correspond to $\delta_{\rm LC}=0.96$ GeV and $\delta_{\rm th}=0.79$
809: GeV, which
810: are obtained from the relation $\sqrt{\langle s \rangle_{\rm LC}}=\sqrt{\langle s \rangle_{\rm th}}=5.27$ GeV
811: for $q^2=0$ and $\beta=0.5$ GeV;
812: the plots in the right column correspond to
813: $\delta_{\rm LC}=0.755$ GeV and $\delta_{\rm th}=0.69$ GeV,
814: which are obtained from the relation $\sqrt{\langle s \rangle_{\rm LC}}=\sqrt{\langle s \rangle_{\rm th}}=5.27$ GeV
815: for $q^2=0$ and $\beta=4$ GeV.
816: {\bf First row}: $\sqrt{\langle s \rangle_{\rm th}}$ (solid red line)
817: and $\sqrt{\langle s \rangle_{\rm LC}}$ vs $\beta$ (dashed blue
818: line). The horizontal (green) line is $M_Q=5.27$ GeV.
819: {\bf Second row}: $\widetilde\Gamma(\beta, q^2,\delta)$ vs $\beta$ [(\ref{gammas})] at $q^2=0$:
820: $\widetilde\Gamma_{\rm th}(\beta, q^2,\delta_{\rm th})$ (solid red line)
821: and
822: $\widetilde\Gamma_{\rm LC}(\beta, q^2,\delta_{\rm th})$ (dashed blue line).
823: {\bf Third row:} The ratio
824: $\hat\Gamma_{\rm th}(\beta, q^2,\delta_{\rm th})/\hat\Gamma_{\rm LC}(\beta, q^2,\delta_{\rm LC})$
825: vs $\beta$ for $q^2=0$.}
826: \end{center}
827: \end{figure}
828: Now, let us study the {\it cut} Borel-transformed correlators which are relevant for the
829: extraction of the form factors within QCD sum rules. We shall see that for the cut Borel-transformed correlator the
830: situation is different: namely, there is no physical limit in which the
831: off-LC effects are negligible.\footnote{As we have seen, the full and the LC spectral densities
832: coincide in the limit $m\to 0$. However, the parameter $m$ should be identified
833: with the effective quark mass, which stays finite of order $\Lambda_{\rm QCD}$ in the chiral limit.
834: Therefore,
835: the limit $m\to 0$ does not correspond to a realistic situation.}
836: %D-MESON
837: \begin{figure}[t]
838: \begin{center}
839: \begin{tabular}{ll}
840: \includegraphics[width=8.4cm]{Fig_3_c.eps}&\includegraphics[width=8.4cm]{Fig_4_c.eps}\\
841: \includegraphics[width=8.4cm]{Fig_3_d.eps}&\includegraphics[width=8.4cm]{Fig_4_d.eps}\\
842: \hspace{.3cm}\includegraphics[width=8.4cm]{Fig_3_e.eps}&\hspace{.3cm}\includegraphics[width=8.4cm]{Fig_4_e.eps}
843: \end{tabular}
844: \caption{\label{Fig:4}Plots corresponding to charm-meson decay $m_Q=1.4$ GeV,
845: for $\delta$ fixed by tuning $\sqrt{\langle s\rangle}$ to $M_Q=1.87$ GeV at $\beta=2$ GeV,
846: and for two different values of the light-quark mass:
847: $m=200$ MeV (left column) and $m=100$ MeV (right column).
848: {\bf First row:} $\sqrt{\langle s \rangle_{\rm LC}}$ and $\sqrt{\langle s \rangle_{\rm th}}$ [Eq.~(\ref{delta})]
849: vs $\beta$. The horizontal (green) line is $M_Q=1.87$ GeV.
850: The parameters $\delta_{\rm LC}=0.93$ GeV and
851: $\delta_{\rm th}=0.72$ GeV for $m=200$ MeV (left column) and
852: $\delta_{\rm th}=0.85$ GeV for $m=100$ MeV (right column)
853: are obtained by requiring that at $q^2=0$ and $\beta=2$ GeV
854: $\sqrt{\langle s \rangle_{\rm LC}}=\sqrt{\langle s \rangle_{\rm th}}=1.87$ GeV.
855: {\bf Second row:}
856: $\widetilde\Gamma(\beta, q^2,\delta)$ [Eq.~(\ref{gammas})] at $q^2=0$:
857: $\widetilde\Gamma_{\rm th}(\beta, q^2,\delta_{\rm th})$ (solid red line)
858: and
859: $\widetilde\Gamma_{\rm LC}(\beta, q^2,\delta_{\rm LC})$ (dashed blue line).
860: {\bf Third row:} The ratio
861: $\hat\Gamma_{\rm th}(\beta, q^2,\delta_{\rm th})/\hat\Gamma_{\rm LC}(\beta, q^2,\delta_{\rm LC})$ at
862: $q^2=0$.}
863: \end{center}
864: \end{figure}
865: We introduce the parameters $\delta$ related to the continuum
866: subtraction $s_0$ by
867: \begin{eqnarray}
868: s_0=(m_Q+\delta)^2.
869: \end{eqnarray}
870: To fix $\delta$, we follow the standard procedure \cite{bz}: namely, we require that the quantity
871: \begin{eqnarray}
872: \label{delta}
873: \langle s(\beta,q^2,\delta) \rangle \equiv
874: \frac{\displaystyle\int\limits_{s_{\rm low}}^{(m_Q+\delta)^2}ds\exp\left(-\frac{s-M_Q^2}{2\mu_B^2}\right) \,s\, \Delta(s,q^2)}
875: {\displaystyle\int\limits_{s_{\rm low}}^{(m_Q+\delta)^2}ds \exp\left(-\frac{s-M_Q^2}{2\mu_B^2}\right)\,\Delta(s,q^2)}\,,
876: \qquad \mu_B^2=m_Q\beta,
877: \end{eqnarray}
878: reproduces the heavy-meson mass, i.e.,
879: \begin{eqnarray}
880: \label{delta2}
881: \langle s(\beta,q^2,\delta) \rangle=M_Q^2,
882: \end{eqnarray}
883: both for the LC and the full spectral densities,
884: where $s_{\rm low}=m_Q^2$ for the LC correlator and $s_{\rm low}=(m_Q+m)^2$
885: for the full correlator.
886: The quantity $\delta$ as defined by (\ref{delta2}) depends on $q^2$ and $\mu_B$:
887: Eq.~(\ref{delta2}) is just the definition of the implicit function
888: $\delta(q^2,\mu_B^2)$. Such a procedure of fixing $\delta$ for the light-cone correlator
889: was employed e.g. in \cite{bz}.
890: Recall, however, that there is no unique way to fix $\delta$: one may
891: require instead that $\langle s^n \rangle=(M_Q^2)^n$ for $n>1$.
892: Moreover, since the spectral densities and the thresholds in Eq.~(\ref{delta}) are different
893: for $\delta_{\rm th}$ and $\delta_{\rm LC}$, also the numerical values of
894: $\delta_{\rm th}$ and $\delta_{\rm LC}$ obtained from Eq.~(\ref{delta2}) are different.
895: We discuss here only the case $q^2=0$. Taking into account the lack of a unique way to
896: introduce $\delta$, we shall not consider the $\mu_B$-dependent $\delta_{\rm th}$ and
897: $\delta_{\rm LC}$. We rather determine the constant values $\delta_{\rm th}$ and $\delta_{\rm LC}$
898: such that the relation (\ref{delta2}) is satisfied only for one specific value of $\beta$.
899:
900: Fig.~\ref{Fig:3} presents the numerical results for beauty-meson decay: $M_Q=5.27$ GeV$,
901: m_Q=4.8$ GeV and $m=150$ MeV. We plot the Borel curves for the full and for the LC correlators for
902: two different values of $\delta$:
903: %
904: The left column shows the results for $\delta_{\rm th}=0.86$ GeV and $\delta_{\rm LC}=0.96$ GeV.
905: In this case, the relation (\ref{delta2}) is fulfilled at the relatively low value $\beta=0.5$ GeV:
906: namely,
907: $\sqrt{\langle s \rangle_{\rm LC}}=\sqrt{\langle s \rangle_{\rm th}}=5.27$ GeV
908: for $q^2=0$ and $\beta=0.5$ GeV.
909: %
910: The right column present the results for $\delta_{\rm th}=0.72$ GeV and $\delta_{\rm LC}=0.755$ GeV.
911: In this case,
912: $\sqrt{\langle s \rangle_{\rm LC}}=\sqrt{\langle s \rangle_{\rm th}}=5.27$ GeV
913: for $q^2=0$ and $\beta=4$ GeV.
914: %
915: The first row shows $\sqrt{\langle s(\beta,\delta)\rangle}$ calculated with the LC and the
916: full correlators vs $\beta$.
917: %
918: The second row presents the quantity
919: \begin{eqnarray}
920: \label{gammas}
921: \widetilde \Gamma(\beta,q^2,\delta)=m_Q^2\exp\left(\frac{M_Q^2}{2\mu_B^2}\right)
922: \hat \Gamma(\mu_B^2,q^2,s_0), \quad \mu_B^2=m_Q\beta, \quad s_0=(m_Q+\delta)^2,
923: \end{eqnarray}
924: for the LC and the full correlators.
925: %
926: Finally, the third row gives the ratio
927: of the full to the LC correlators.
928:
929:
930: Fig.~\ref{Fig:4} gives the results for charm-meson decay: $M_Q=1.87$ GeV, $m_Q=1.4$ GeV and
931: $m=200$ MeV (left column).
932: To illustrate the influence of the light-quark mass on the off-LC effects, we
933: present also the results for $m=100$ MeV (right column).
934: The continuum subtraction parameter $\delta$ is fixed from the relation
935: $\sqrt{\langle s(\beta,\delta)\rangle}=M_Q$ at $\beta=2$ GeV.
936:
937: To show the origin of the difference between the cut full and light-cone correlators,
938: we consider the limit $m_Q\to \infty$ and $\mu_B\to\infty$, with
939: $\mu_B\gg m_Q$.
940: In this case explicit expressions for the correlators may be obtained:
941: \begin{eqnarray}
942: \label{4.8}
943: m_Q^2\hat \Gamma_{\rm LC}(\mu^2_B\to\infty,q^2=0,\delta_{\rm LC})&=&
944: 2\delta^2_{\rm LC}+O(\delta^3_{\rm LC}/m_Q),\nonumber \\
945: m_Q^2\hat \Gamma_{\rm th}(\mu^2_B\to\infty,q^2=0,\delta_{\rm th})&=&
946: 2\delta^2_{\rm th}-m^2 \left[\log\left(\frac{4 \delta^2_{\rm th}}{m^2}\right)+1\right]+O(m^4/\delta^2_{\rm th})
947: +O(\delta^3_{\rm th}/m_Q).
948: \end{eqnarray}
949: For $\mu_B\gg m_Q$ ($\beta\sim m_Q$), the uncut and the cut correlators (both full and LC)
950: behave quite differently for large $m_Q$:
951: \begin{eqnarray}
952: \Gamma(\mu_B^2\to\infty,m_Q^2)=O(1), \qquad
953: \Gamma(\mu_B^2\to\infty,m_Q^2,\delta)=O(\delta^2/m_Q^2).
954: \end{eqnarray}
955: Thus, the cut correlator picks up only a small
956: fraction of the full correlator from the region not far from the threshold.
957: (For $\beta \ll m_Q$, both the
958: cut and the uncut correlators have a similar behavior $\sim 1/m_Q^2$.)
959:
960:
961: Now, fixing
962: $\delta_{\rm th}$ and $\delta_{\rm LC}$ according to the standard procedure (\ref{delta2}),
963: we express these quantities via the binding energy of the heavy meson $\varepsilon_Q$ defined
964: according to
965: $
966: M_Q=m_Q+\varepsilon_Q$:\footnote{
967: Notice that $\delta_{\rm LC}$ turns out to be different
968: from $\delta_{\rm th}$. This has the following origin:
969: If we include $N$ terms in the LC expansion of the cut correlator (\ref{cutlc}), cut them at
970: $\delta^{(N)}_{\rm LC}$, and determine the latter from Eqs.~(\ref{delta}) and (\ref{delta2}),
971: then
972: $\lim\limits_{N\to\infty}\delta^{(N)}_{\rm LC}=\delta_{\rm th}$.
973: Since we have included only one $(n=0)$ term, we obtain $\delta^{(0)}_{\rm LC}\ne \delta_{\rm th}$.}
974: \begin{eqnarray}
975: \label{deltas}
976: \delta_{\rm LC}&=&\frac32 \varepsilon_Q,\nonumber\\
977: \delta_{\rm th}&=&\frac32 \varepsilon_Q-\frac{2m^2}{3\varepsilon_Q}\left[
978: \log\left(\frac{3 \varepsilon_Q}{m}\right)-1
979: \right]+\cdots,
980: %O\left((m/\varepsilon_Q)^3\log(m/\varepsilon_Q)\right),
981: \end{eqnarray}
982: \vspace{-.5cm}
983: leading to
984: \begin{eqnarray}
985: m_Q^2\hat \Gamma_{\rm LC}(\mu^2_B\to\infty,q^2=0,\delta_{\rm LC})&=&
986: \frac92\varepsilon_Q^2,\nonumber \\
987: m_Q^2\hat \Gamma_{\rm th}(\mu^2_B\to\infty,q^2=0,\delta_{\rm th})&=&
988: \frac92\varepsilon_Q^2-6m^2\log\left(\frac{3\varepsilon_Q}{\sqrt{e}m}\right)+\cdots,
989: \end{eqnarray}
990: \vspace{-.4cm}
991: and thus
992: \begin{eqnarray}
993: \frac{\hat \Gamma_{\rm th}(\mu^2_B\to\infty,q^2=0,\delta_{\rm th})}
994: {\hat \Gamma_{\rm LC}(\mu^2_B\to\infty,q^2=0,\delta_{\rm LC})}=
995: 1-\frac{4m^2}{3\varepsilon_Q^2 }\log\left(\frac{3\varepsilon_Q}{\sqrt{e}m}\right)+\cdots.
996: \end{eqnarray}
997: In the expressions above the dots denote terms containing higher powers of $m/\varepsilon_Q$.
998: This example illustrates that the off-LC effects may play an essential role in the cut correlators, as their contribution is
999: not suppressed by any large parameter: the quantites $m$ and $\varepsilon_Q$ have the same order of magnitude.
1000:
1001: Let us emphasize that we compare the full and the light-cone correlators
1002: evaluated at different values of the cut parameters $\delta_{\rm LC}$ and $\delta_{\rm th}$.
1003: From our point of view this very comparison is relevant if one wants to understand
1004: the error due to taking into account only the light-cone $(x^2=0)$ contribution
1005: to the correlator and neglecting terms containing higher powers of $x^2$.
1006:
1007: One could also compare the correlators for the same value of $\delta$. The difference between
1008: the full and the LC correlators is only slightly reduced in this case, the ratio
1009: still remaining $O(m^2/\delta^2)$. This can be seen by
1010: setting $\delta_{\rm LC}=\delta_{\rm th}$ in (\ref{4.8}).
1011:
1012: %\newpage
1013: %\vspace{1.2cm}
1014: The following lessons may be drawn from the results presented in this section:
1015: \begin{itemize}
1016: \item[a.]
1017: The off-LC effects play an essential role in the cut correlator, as they are not suppressed by any large
1018: parameter. Numerically, the difference between the full and the LC correlators,
1019: evaluated at the same value of the Borel parameter, is 10$\div$20\%.
1020: This difference is due to the off-LC effects.
1021: \item[b.] The functions
1022: $\widetilde\Gamma_{\rm LC}(\beta)$ and $\widetilde\Gamma_{\rm th}(\beta)$
1023: have the same shape, but $\widetilde\Gamma_{\rm LC}(\beta)$ lies well above
1024: $\widetilde\Gamma_{\rm th}(\beta)$, if the standard
1025: procedure of fixing $\delta_{\rm LC}$ and $\delta_{\rm th}$ (\ref{delta2}) is used.
1026: The values of both (th and LC) correlators obtained with $\delta_{\rm th}$ and $\delta_{\rm LC}$
1027: tuned at $\beta=0.5$ GeV (left column in Fig.~\ref{Fig:3})
1028: are greater than the values of the respective correlators
1029: obtained with $\delta_{\rm th}$ and $\delta_{\rm LC}$ tuned at $\beta=4$ GeV (right column in Fig.~\ref{Fig:3}).
1030: The local stability is better when one fixes $\delta$ from (\ref{delta2}) at a larger value
1031: of $\beta$ (right column in Fig. \ref{Fig:3}).
1032: In this case, both Borel curves for
1033: $\widetilde\Gamma_{\rm LC}$ and $\widetilde\Gamma_{\rm th}$ show a
1034: good stability in $\beta$. Nevertheless, still
1035: $\widetilde\Gamma_{\rm LC}$ is much larger than $\widetilde\Gamma_{\rm th}$! This illustrates that the Borel
1036: stability {\it per se} does not guarantee the extraction of the correct physical value.
1037:
1038: \item[c.]
1039: The difference between $\widetilde\Gamma_{\rm LC}$ and $\widetilde\Gamma_{\rm th}$
1040: %corresponding to off-LC effects,
1041: increases with increasing mass of the light quark.
1042: Therefore, this difference is expected to be greater for the heavy mesons $B_s$ and $D_s$,
1043: containing the strange $s$-quark, than for $B$ and $D$.
1044: \end{itemize}
1045:
1046: %\newpage
1047:
1048: \section{\label{conclusions}Conclusions}
1049: In this paper, we studied the correlator
1050: $$
1051: i\int dx \exp(ipx)\langle 0|T \varphi(x)Q(x) Q(0)\varphi(0)|M(p')\rangle,
1052: $$
1053: which is one of the basic objects for extracting the heavy-to-light form factor within the
1054: method of sum rules.
1055: We have shown that to leading $1/m_Q$-accuracy this correlator may be calculated through
1056: the BS amplitude of the light meson
1057: \vspace{-.3cm}
1058: $$
1059: \langle 0|T \varphi(x)\varphi(0)|M(p')\rangle.
1060: $$
1061: Expanding the BS amplitude near the light cone $x^2=0$
1062: generates the light-cone expansion of the correlator.
1063:
1064: Making use of the Nakanishi representation for the BS amplitude,\footnote{The Nakanishi
1065: representation leads to technical simplifications, but conceptually
1066: any other form of the BS amplitude may be used.}
1067: we obtained
1068: dispersion representations for the full and the
1069: LC correlators in terms of the kernel $G(z,\xi)$ of the Nakanishi representation.
1070: We studied the full and the light-cone correlators and their Borel transforms
1071: depending on the properties of the kernel $G(z,\xi)$.
1072:
1073: We then made use of the known solution for $G(z,\xi)$ in a model with
1074: light scalar particles interacting by an exchange of a massless boson. This relatively simple
1075: model provides a good laboratory for studying QCD since the corresponding bound-state wave
1076: functions have properties similar to the properties of hadron wave functions in QCD.
1077: We calculated the full and the light-cone correlators and their
1078: Borel transforms in the variable $p^2$, and studied these correlators for various
1079: prescriptions to fix the heavy-hadron continuum subtraction and in various regions
1080: of the parameters relevant for extracting the heavy-to-light form factors.
1081: This work thus represents the
1082: first systematic study of the off-light-cone effects in QCD sum rules for heavy-to-light form
1083: factors.
1084:
1085: %\vspace{.5cm}
1086: Our main results may be summarized as follows:
1087: \begin{itemize}
1088: \item[1.]
1089: We have seen that --- after performing the Borel transform --- the light-cone
1090: correlator provides numerically the bulk of the full
1091: correlator, although parametrically the off-LC effects are not suppressed
1092: compared to the LC contribution. This observation holds for various prescriptions of fixing the heavy-hadron continuum subtraction point
1093: (i.e., a cut applied to the correlator for isolating the contribution of the heavy hadron of interest in the initial state) and a wide range of masses of particles involved in the decay process.
1094: \item[2.]
1095: We demonstrated that, nevertheless, the difference between the cut full and the cut
1096: light-cone correlators always remains nonvanishing. For example, fixing the continuum subtraction points
1097: the by standard criteria, we have found the following relation for the cut
1098: Borel transforms of the full and the LC correlators for $m_Q\to\infty$,
1099: $\mu_B\to\infty$, and $q^2=0$:
1100: \begin{eqnarray}
1101: \label{result}
1102: \frac{\hat \Gamma_{\rm th}(\mu^2_B\to\infty,q^2=0,\delta_{\rm th})}
1103: {\hat \Gamma_{\rm LC}(\mu^2_B\to\infty,q^2=0,\delta_{\rm LC})}=
1104: 1-\frac{4m^2}{3\varepsilon_Q^2 }\log\left(\frac{3\varepsilon_Q}{\sqrt{e}m}\right)+\cdots,
1105: \end{eqnarray}
1106: the correction being always negative. Here $m$ is the effective constituent mass of the light quark,
1107: which emerges from the BS equation, $m\simeq \Lambda_{\rm QCD}$, and $\varepsilon_Q$ is the binding energy
1108: of the heavy meson $M_Q=m_Q+\varepsilon_Q$.
1109: Taking into account that the constituent quark
1110: mass remains finite in the chiral limit, we come to the following important conclusion:
1111: {\it In heavy-to-light decays, there exists no rigorous theoretical limit in which
1112: the cut LC correlator coincides with the cut full correlator}.
1113:
1114: Thus, the off-light-cone
1115: effects in sum rules for heavy-to-light correlators are not negligible and should be
1116: taken into account.
1117: %In other words, the knowledge of only the longitudinal lowest-twist distribution amplitudes
1118: %of the light hadrons involved in the weak decay is not sufficient: the transverse-motion effects
1119: %always provide nonvanishing contributions and should be taken into account.
1120: \item[3.]
1121: We note that the Borel curves for the full and the LC correlators have similar shapes.
1122: However the light-cone correlator systematically {\it overestimates} the full correlator, the difference at small $q^2$
1123: being $10\div 20$\% in a wide range of the heavy-quark mass relevant for charm and
1124: beauty decays. We want to point out that the similarity of the Borel
1125: curves for the full and the LC correlators implies that the systematic difference
1126: between the correlators cannot be diminished by a relevant choice of the criterion
1127: for extracting the heavy-to-light form factor.
1128:
1129: The observed effect might suggest a systematic uncertainty in the results for
1130: form factors obtained within light-cone sum rules.
1131: As follows from the relation (\ref{result}), this uncertainty is expected to be larger
1132: for decays of heavy mesons containing the strange quark,
1133: $B_s$ and $D_s$, than for the $B$ and $D$ mesons. This issue deserves further investigation.
1134: \end{itemize}
1135: Finally, we point out the following: Although the model, discussed here, in many aspects
1136: differs from QCD, this model mimics correctly those
1137: features which are essential for the effects discussed.
1138: Therefore, many of the results obtained in this paper are valid also for QCD.
1139: In particular, the expression (\ref{result}) suggests the following relationship between the light-cone
1140: and the full correlators in QCD for large values of $m_Q$ and $\mu_B$:
1141: \begin{eqnarray}
1142: \label{result_QCD}
1143: \frac{\hat \Gamma_{\rm th}(\mu^2_B,q^2,\delta_{\rm th})}{
1144: \hat \Gamma_{\rm LC}(\mu^2_B,q^2,\delta_{\rm LC})}=
1145: 1-O\left(\frac{\Lambda_{\rm QCD}}{\delta}\right).
1146: \end{eqnarray}
1147: In numerical estimates, we used the parameters relevant for $B$ and $D$ decays.
1148: We therefore believe that also the numerical estimates for higher-twist effects
1149: obtained in this work provide a realistic estimate for higher-twist effects in QCD.
1150:
1151: \vspace{.3cm}
1152: \noindent
1153: {\it Acknowledgments.}
1154: We are grateful to Vladimir Braun, Pietro Colangelo, and Alexander Khodjamirian
1155: for valuable comments on the preliminary version of the paper.
1156: We thank Vittorio Lubicz and Matthias Neubert for interesting and stimulating
1157: discussions.
1158: D.~M. gratefully acknowledges financial support from INFN, University
1159: ``Roma Tre'',
1160: and the Austrian Science Fund (FWF) under project P17692.
1161:
1162: \begin{thebibliography}{30}
1163: \bibitem{lcsr}I.~I.~Balitsky, V.~M.~Braun, and A.~V.~Kolesnichenko, Nucl.~Phys.~{\bf B312}, 509
1164: (1989); V.~M.~Braun and I.~Filyanov, Z.~Phys.~{\bf C44}, 157 (1989);
1165: V.~I.~Chernyak and I.~R.~Zhitnitsky, Nucl.~Phys.~{\bf B345}, 137 (1990).
1166: \bibitem{ck}P.~Colangelo and A.~Khodjamirian,
1167: {\it QCD sum rules: a modern perspective}, in {\em At the Frontier of Particle Physics},
1168: edited by M. Shifman (World Scientific, Singapore, 2001), vol.~3, p.~1495 [hep-ph/0010175].
1169: \bibitem{bb} P.~Ball and V.~M.~Braun, Phys. Rev. {\bf D58}, 094016 (1998).
1170: \bibitem{bz}P.~Ball and R.~Zwicky, Phys. Rev. {\bf D71}, 014015 (2005).
1171: \bibitem{bh} V.~M.~Braun and I.~Halperin, Phys. Lett. {\bf B328}, 457 (1994).
1172: \bibitem{nakanishi} N.~Nakanishi, Phys. Rev. {\bf 130}, 1230 (1963).
1173: \bibitem{bogolyubov} N.~N.~Bogolyubov and D.~V.~Shirkov, {\it Quantum Fields}, Nauka, Moscow (1980).
1174: \bibitem{karmanov} V.~A.~Karmanov and J.~Carbonell, Eur.~Phys.~J.~{\bf A27},1 (2006).
1175: \bibitem{svz} M.~Shifman, A.~Vainshtein, and V.~Zakharov, Nucl. Phys. {\bf B147}, 1 (1979).
1176: \bibitem{shifman} M.~Shifman, Prog. Theor. Phys. Suppl. {\bf 131}, 1 (1998) [hep-ph/9802214].
1177: \bibitem{jamin} M.~Jamin and B.~Lange, Phys. Rev. {\bf D65}, 056005 (2002).
1178: \bibitem{tk} T.~Kleinschmidt, hep-ph/0409039 (unpublished).
1179: \bibitem{lms}
1180: D.~Melikhov, Phys. Rev. {\bf D53}, 2460 (1996); Phys. Rev. {\bf D56}, 7089 (1997);
1181: Eur. Phys. J. direct {\bf C4}, 2 (2002) [hep-ph/0110087];
1182: W.~Lucha, D.~Melikhov, and S.~Simula, Phys. Rev. {\bf D75}, 016001 (2007).
1183: \bibitem{radyushkin}V.~A.~Nesterenko and A. V. Radyushkin, Phys. Lett. {\bf 115B}, 410 (1982).
1184: \bibitem{lm} W.~Lucha and D.~Melikhov, Phys. Rev. {\bf D73}, 054009 (2006).
1185: \bibitem{stech}
1186: D.~Melikhov and B.~Stech, Phys. Rev. {\bf D62}, 014006 (2000).
1187: \bibitem{bp}P.~Ball, Phys. Lett. {\bf B641}, 50 (2006).
1188: \bibitem{ms}
1189: D.~Melikhov and S.~Simula, Eur. Phys. Journal {\bf C37}, 437 (2004).
1190: \bibitem{msl}
1191: W.~Lucha, D.~Melikhov, and S.~Simula, Phys. Rev. {\bf D74}, 054004 (2006).
1192: \end{thebibliography}
1193: \end{document}
1194:
1195: 11.55.Hx - Sum rules,
1196: 12.38.Lg - other nonperturbative methods,
1197: 11.10.St bound and unstabel states, Bethe-Salpeter equations
1198:
1199: 13.20.-v semileptoniv decays of mesons
1200: 13.20.He decays of bottom mesons
1201: 03.65.Ge - solutions of wave equations
1202: