1: \newpage
2: \section{Introduction}
3: \label{sec:introduction}
4: Weak decays of hadrons provide an important source of information on
5: the parameters of the standard model, the structure
6: of weak currents, and internal structure of hadrons.
7: Therefore for many years weak decays have been in the focus
8: of experimental and theoretical investigations.
9:
10: In the last decade, the main emphasis has been laid on weak $B$ decays which
11: allow to measure the unknown parametrs of the Cabibbo-Kobayashi-Maskawa (CKM) matrix
12: describing the mixing of heavy $c$, $b$ and $t$ quarks, and the CP-violation:
13: semileptonic and nonleptonic $B$ decays induced
14: by the charged-current quark transition $b\to u, c$ provide a direct
15: access to the CKM matrix elements $V_{ub}$ and $V_{cb}$ and the weak
16: CP-violating phase.
17:
18: Rare semileptonic $B$ decays induced by the
19: flavour-changing neutral current transitions $b\to s$ and $b\to d$ measure
20: $V_{ts}$ and represent an important test of the standard model and of the physics beyond it.
21: Rare decays are forbidden at the tree level in the standard model and occur through
22: loop diagrams. Thus they provide the possibility to probe at the
23: relatively low energies of $B$ decays
24: the structure of the electroweak sector at large mass scales from contributions of virtual particles
25: in the loop diagrams.
26: Interesting information about the structure of the theory is contained in the
27: Wilson coefficents in the effective Hamiltonian which describes the $b\to s,d$ transition
28: at low energies. These Wilson coefficients take different values in different theories with
29: testable consequences in rare $B$ decays.
30:
31: However, along with the interesting parameters of the standard model, the decay rates and
32: differential distributions measurable in the decay process contain also quantites
33: related to the presence of hadrons in the decay process.
34: Therefore the extraction of the interesting standard model parameters
35: from the decay experiments requires reliable information on hadron structure and hadronic
36: amplitudes of the weak quark currents.
37:
38: Theoretical description of hadronic amplitudes of the quark currents is one of the key problems of particle
39: physics as such amplitudes provide a bridge between QCD formulated in the language of quarks and gluons and
40: observable phenomena which deal with hadrons. The difficulty in such calculations lie in the fact that hadron
41: formation occurs at large distances where perturbative QCD methods are not applicable and therefore
42: a nonperturbative consideration is necessary.
43:
44: The presence of heavy quarks with the masses much bigger than the confinement scale of QCD provide
45: important constraints upon the long distance effects due to a new spin-flavour symmetry which emerges
46: in this limit \cite{shifmanvoloshin,iwhh}. This symmetry leads to a heavy quark effective
47: theory (HQET)
48: \cite{hqet} for QCD with heavy quarks. HQET restricts the structure of the expansion
49: of the hadronic transition amplitudes in inverse powers of the heavy quark mass.
50:
51: In inclusive $B$ decays the combination of the heavy-quark and the operator product expansions allow one
52: to connect the decay rate of the quark bound in a heavy hadron to the decay rate of the free quark.
53: An important consequence of the operator product expansion is the appearance of the
54: corrections to the free quark decay rate only at the second power of the
55: $1/m_Q$ \cite{cgg,bsuv93}. Therefore, these corrections are numerically small.
56: Providing quite reliable predictions for the integrated rates, the theoretical approach based on the OPE turns out to be less efficient for the
57: description of the differential distributions. In this case a more precise information on the
58: details of the $b$ quark motion inside the $B$ meson is necessary.
59:
60: For the exclusive $b\to c$ decays, HQET provides constraints on the structure of the expansion of
61: the meson transition form factors in the inverse powers of the $b$ and $c$ quark masses
62: and leads to the appearance of universal process-independent form factors at each order
63: of the $1/m_Q$ expansion \cite{luke,neubert}. Moreover, heavy quark symmetry provides the absolute value of the
64: leading order universal form factor - the Isgur-Wise function at zero recoil (maximum momentum transfer)
65: thus allowing for a relible description of the $b\to c$ decay in this kinematical region.
66: HQET however cannot calculate the universal form factors as functions of the momentum transfer.
67: Moreover, the $1/m_c$ corrections for the
68: $b\to c$ transiton form factors turn out to be not small because the $c$ quark
69: is not sufficiently heavy.
70:
71: For the exclusive $b\to u,s$ transitions, when the final quark is light,
72: relations between the form factors describing meson transition induced by
73: different currents emerge in the region near zero recoil \cite{iwhl}.
74: In the opposite region of large recoils, where the final quark is fast, one can construct
75: another effective theory, so called large energy effective theory (LEET) which allows
76: the double expansion of the transition
77: form factors in inverse powers of $m_b$ and the energy $E$ of the light quark produced
78: in the weak decay \cite{leet}. LEET predicts the appearance of several universal form factors at the leading order
79: of the $1/E$ and $1/m_b$ expansion, but does not calculate these form factors, and also does not constrain
80: the structure of higher order corrections.
81:
82: Thus, heavy-quark symmetry provides important constraints for the form factors but does not allow to
83: calculate them in full detail. This task requires a detailed treatment of the nonperturbative effects.
84:
85: Theoretical approaches for calculating transition form factors are
86: quark models \cite{wsb,koerner,isgw,isgw2,jaus,jauswyler,dubin,beyer,faustov,orsay,aley,sim95,simula,gns,m1,mplb1,mplb2,m2},
87: QCD sum rules \cite{sr1,bbd,bhr,ball,sr4,sr7,colangelo,lcsr,lcsr1,braun,aliev,likhoded},
88: and lattice QCD \cite{lat1,lat2,lat3,lat4,lat5,lattice,ape,ukqcd,latg}.
89: Approaches combining different methods are also extensively used
90: \cite{lr,burdmandonoghue,stech,lat,anal1,anal2,lellouch,soares}.
91:
92: In spite of the considerable progress made in the recent years
93: theoretical uncertainties are still uncomfortably large, around 10-15\%.
94: The main difficulty in obtaining the full picture of the form factors for various decays and for
95: all $q^2$ lies in the fact that methods more directly related to QCD,
96: such as lattice QCD and QCD sum rules
97: have only a limited range of applicability, whereas the results from quark models are
98: sensitive to the details of the model and the parameters used.
99:
100: QCD sum rules are suitable for describing the low $q^2$
101: region of the form factors. Sum rule calculations require various inputs such
102: as the condensates or the distribution amplitudes of the light mesons produced
103: in the heavy meson decay. The higher $q^2$ region is hard to get and
104: higher order calculations are not likely to give real progress because of the
105: appearance of many new parameters. The accuracy of the method
106: cannot be arbitrarily improved because of the necessity to isolate the
107: contribution of the states of interest from others.
108:
109: Lattice QCD gives good results for the high $q^2$ region. But because of the
110: many numerical extrapolations involved this method does not provide for
111: a full picture of the form factors and for the relations between the various
112: decay channels.
113:
114: Quark models do provide such relations connecting different
115: processes through the meson soft wave functions, and give
116: the form factors in the full $q^2$ range.
117: However, quark models are not closely related to the QCD
118: Lagrangian (or at least this relationship is not well understood yet) and
119: therefore have input parameters which are not directly measurable and may
120: not be of fundamental significance.
121:
122: Since quark models are not directly deduced from QCD it is important to match
123: the results for the transition form factors obtained within quark models
124: to rigorous QCD predictions for these form factors in the specific limit of
125: heavy quark masses.
126:
127: The application of various versions of the constituent quark picture
128: to decay processes has a long history.
129: First models were based on a relativistic \cite{wsb} or
130: nonrelativistic \cite{koerner,isgw} considerations. They could not fully
131: incorporate the quark dynamics and spin structure and therefore could not satisfy
132: rigorous relations for the transition
133: form factors based of the spin-flavour symmetry in the heavy quark limit of QCD.
134:
135: A self-consistent relativistic treatment of the quark spins can be performed
136: within the light-front quark model \cite{lcqm,lcqm1}. The model allows for a calculation
137: of the reference-frame dependent partonic contribution to the form factor. However, the
138: non-partonic contribution cannot in general be calculated. At spacelike momentum transfers
139: the nonpartonic part can be killed by an appropriate choice of the reference frame. Therefore,
140: the partonic contribution calculated in this specific reference frame gives the full form factor.
141: At timelike momentum transfers the nonpartonic contribution does not vanish for any accessible
142: choice of the reference frame, and the knowledge of the frame-dependent partonic contribution is
143: not sufficient to determine the form factor.
144:
145: A relativistic dispersion approach to the decay processes based on the constituent
146: quark picture overcomes the above difficulties. It has been formulated in
147: \cite{m1,mplb1,mplb2,m2} and applied to the study of the long-distance effects in
148: many pocesses involving the $B$ meson: such as
149: exclusive semileptonic \cite{m1,mplb1,mplb2,m2,mb1,mb2,ms} and rare
150: \cite{mnsplb1,mns} $B$ decays, the $B\to\gamma l\nu$ form factor and the weak annihilation
151: in the rare radiative $B\to \rho\gamma$ decay \cite{bmns,ms1}, nonfactorizable corrections
152: for the $B^0-\bar B^0$ mixing \cite{mn}, calculation of the differential distributions in
153: inclusive $B\to X_cl\nu$ decays \cite{msimula}.
154:
155: The approach is based on a consistent treatment of the two-particle singularities of the
156: Feynman diagrams describing meson interactions. Amplitudes of these processes are given by the
157: relativistic spectral representations over the mass variables in terms of the wave functions
158: of the participating mesons and spectral densities of the corresponding Feynman diagrams.
159: In particular, meson transition form factors both at spacelike and timelike momentum transfers
160: are given by the double spectral representations.
161:
162: We discuss in this paper applications of the dispersion approach to various processes
163: involving heavy mesons, laying the main emphasis on the calculation of the form factors
164: for heavy meson decays.
165:
166: Let us highlight the main features of our dispersion formulation of the constituent quark
167: model:
168:
169: %\newpage
170: \vskip0.5cm
171: \centerline{\it 1. The physical picture}
172: \vskip0.5cm
173: \noindent
174: The constituent quark picture is based on the following phenomena expected from QCD:
175: \begin{itemize}
176: \item
177: %\noindent i)
178: the chiral symmetry breaking in the low-energy region which provides for
179: the masses of the constituent quarks;
180: \item
181: %\noindent ii)
182: a strong peaking of the nonperturbative meson wave functions
183: in terms of the quark momenta with a width of the order of
184: the confinement scale;
185: \item
186: %\noindent iii)
187: a $q\bar q$ composition of mesons in terms of constituent quarks.
188: \end{itemize}
189: As is well known, the $q\bar q$ component of the meson wave function in terms of
190: current quarks dominates the exclusive form factors in the deep inelastic region,
191: i.e. for large spacelike momentum transfers. A successesfull description of the
192: mass spectrum of mesons as $q\bar q$ states in terms of the constituent quarks
193: obtained in \cite{gi} shows that the $q\bar q$ approximation works well also
194: in the soft region, where one however has to take into account the transition
195: of the current quarks to constituent quarks. The $q\bar q$ meson component
196: in terms of the constituent quarks leads to a good description of the elastic
197: form factors at small and intermediate momentum transfers \cite{amn1,amn2}.
198: Therefore, one can expect the two-particle approximation to be quite reliable for
199: the description of the range of momentum transfers relevant for
200: heavy meson decays.
201:
202: An important shortcoming of previous quark model predictions was a strong
203: dependence of the results on the special form of the quark model
204: and on the parameter values. We demonstrate that once
205: \begin{itemize}
206: \item[(a)]
207: a proper relativistic formalism is used for the description of the transition
208: form factors and
209: \item[(b)]
210: the numerical parameters of the model are chosen properly (we discuss
211: criteria for such a proper choice below),
212: \end{itemize}
213: the quark model yields results
214: in full agreement with the existing experimental data and in accord with the
215: predictions of more fundamental theoretical approaches.
216: In addition, our approach allows to predict many other form factors which
217: have not yet been measured.
218:
219: \vskip0.5cm
220: \centerline{\it 2. The formalism}
221: \vskip0.5cm
222:
223: \noindent For the description of the transition form factors in their full
224: $q^2$-range and for various initial and final mesons, a fully relativistic
225: treatment is necessary. The dispersion formulation of the quark
226: model provides for such a relativisitc treatment and guarantees the correct
227: spectral and analytical properties of the obtained form factors.
228:
229: The form factors are given by the double spectral representations
230: over the variables $s_1$ and $s_2$, the squares of the invariant masses of the initial
231: and final $Q\bar q$ pairs, respectively.
232: The integrations in $s_1$ and $s_2$ run along the two-particle cuts in the complex $s_1$ and
233: $s_2$ planes.
234: The spectral functions of these spectral representations involve the
235: wave functions of the participating mesons and the double discontinuities of the
236: corresponding triangle Feynman diagrams.
237:
238: We start our analysis of the transition form factors from the spacelike region $q^2<0$,
239: where the double discontinuities can be calculated by means of the Landau-Cutkosky rules.
240:
241: The form factors in the decay region $q^2>0$ are obtained by the analytical
242: continuation in $q^2$. A specific feature of the timelike decay region of $q^2$,
243: is the
244: appearance of the anomalous cuts in the complex $s_1$ and $s_2$ planes and, respectively, of
245: the
246: anomalous contributions to the form factors. Let us point out that the anomalous contribution
247: as well as the normal one is completely determined by the wave functions of the participating
248: mesons in the physical region. The anomalous contribution is small for small positive $q^2$
249: but becomes increasingly important as $q^2$ rises.
250:
251: The form factors obtained by this procedure obey all rigorous constraints from QCD
252: on the structure of the long-distance corrections in the heavy quark limit:
253: they develop the correct heavy-quark expansion at
254: leading and next-to-leading orders in $1/m_Q$ in accordance with QCD
255: in the case when both quarks participating in the weak transition are heavy, i.e.
256: have masses much larger than the confinement scale of QCD.
257:
258:
259: For the heavy-to-light transition, i. e. when only the initial quark is treated as heavy,
260: the transition form factors of the dispersion approach
261: satisfy the relations between the form factors
262: of vector, axial-vector, and tensor currents valid at small recoil.
263: In the limit of the heavy-to-light transitions at small $q^2$ the form
264: factors obey the lowest order $1/m_Q$ and $1/E$ relations of the Large
265: Energy Effective Theory.
266:
267: We want to emphasise once more that form factors for meson decays
268: in the physical decay region $q^2>0$ are directly calculated through the
269: meson wave functions.
270:
271: \vskip0.5cm
272: \centerline{\it 3. Parameters of the model}
273: \vskip0.5cm
274:
275: \noindent
276: In previous applications of quark models the transition form factors
277: turned out to be sensitive to the numerical parameters, such as the
278: quark masses and the slopes of the meson soft wave functions.
279:
280: A possible way to control quark masses and the meson soft wave functions
281: is to use the lattice results for the $B\to \rho$
282: form factors at large $q^2$ as 'experimental' inputs.
283: The $b$ and $u$ constituent quark masses
284: and slope parameters of the $B$, $\pi$, and $\rho$ wave functions assuming for
285: them a simple Gaussian form are obtained through this procedure \cite{mb1}.
286: The Gaussian wave functions of the charm and strange mesons and the effective masses
287: $m_c$ and $m_s$ are fixed by fitting the measured rates
288: for the decays $D\to (K, K^*)l\nu$ \cite{ms}.
289:
290: With these few inputs, numerous predictions for the form
291: factors for the $D_{(s)}$ and $B_{(s)}$ decays into light mesons
292: are obtained \cite{ms} which nicely agree with the experimental results
293: at places where data are available. The calculated transition form factors
294: are also found to be in good agreement with the results from lattice QCD
295: and from sum rules in their regions of validity.
296:
297: Thus, in spite of the rather different masses and properties of mesons
298: involved in weak transitions, all existing data on the form factors can be
299: understood in the quark picture, i.e. all
300: form factors can be described by the few degrees of
301: freedom of constituent quarks. Details of the soft wave functions are not
302: crucial; only the spatial extention of these wave functions of order of
303: the confinement scale is important. In other words, only the meson radii are essential
304: for the decay processes.
305:
306: \vspace{.5cm}
307: The paper is organised as follows:
308:
309: %_______________________
310: \vspace{.5cm}
311:
312: In Chapter II we present details of the composite system description using spectral
313: representations over mass variables. The start with the amplitude of the constituent
314: interaction in the low-energy region and its analytical properties.
315: We then the consider the interaction of the bound state
316: with the external electromagnetic field and construct the gauge and relativistic invariant
317: amplitude of the bound state interaction. The elastic electromagnetic form factor is
318: discussed and the relativistic wave function is introduced. The normalization condition
319: for the wave function is obtained which corresponds to the electric charge conservation.
320:
321: Properties of pseudoscalar mesons are studied. A single spectral representation
322: for the weak leptonic decay constant and double spectral representations for
323: the elastic electromagnetic and the weak transition form factors in terms of the wave
324: function are obtained at $q^2<0$. For comparison with the light-cone quark model results
325: we rewrite our explicitly invariant spectral representations in terms of the light-cone
326: variables and demonstarte the equivalence of the form factors obtained within our dispersion
327: approach and the light-cone quark model for $q^2<0$.
328:
329: The weak transition form factor at $q^2>0$ is obtained by the
330: analytical continuation which discussed in full detail.
331:
332: We then consider the case when one of the quarks in the pseudoscalar meson is heavy,
333: $m_Q\to\infty$ and analyse the size of the $1/m_Q$ effects in the
334: form factors for quark masses in the range of the $b$ and $c$ quarks.
335:
336: %_______________________
337: \vspace{.5cm}
338:
339: Chapter III contains a detailed discussion of the weak transitions of the pseudoscalar mesons
340: to pseudoscalar and vector mesons. The double spectral representation for both cases are
341: constructed starting from $q^2<0$ and going to $q^2>0$ by the analytical continuation.
342: A procedure of fixing the subtraction prescription in the double spectral representations
343: is discussed.
344:
345: We then perform the $1/m_Q$ expansion of the dispersion form factors for the
346: heavy-to-heavy transition to next-to-leading order accuracy in $1/m_Q$.
347: For the heavy-to-light transition the expansion within the leading order in $1/m_Q$
348: is developed.
349: Full consistency with the structure of the heavy-quark expansion in QCD
350: for both of these cases to the orders considered is verified.
351:
352: Spectral representation for the Isgur-Wise function and subleading universal form factors
353: is obtained in terms of the wave function of an infinitely heavy meson. Numerical estimates
354: for the universal form factors are given.
355:
356: %_______________________
357: \vspace{.5cm}
358:
359: In Chapter IV the choice of the numerical paremeters of the model is discussed and form
360: factors for many $D(D_s)$ and $B(B_s)$ to light mesons are calculated.
361: Convenient parametrizations for the calculated form factors are given. Strong coupling
362: constants of heavy mesons are estimated by analysing the residues of the form factors
363: in the poles located beyond the physical region of meson decay.
364: The strong coupling constant obtaibed by this procedure are found in agreement with the
365: results of the direct calculation within our dispersion approach.
366:
367: A detailed comparison with the experimental data and results from other approaches
368: is presented. In all cases a good agreement with the available experimental data on form
369: factors and strong coupling constants is observed.
370:
371: %_______________________
372: \vspace{.5cm}
373:
374: In Chapter V the $B\to\gamma l\nu$ form factors and the weak annihilation in rare
375: $B\to\rho\gamma$ decay is analysed within the factorization approximation.
376: A detailed analysis of the contact terms in the weak annihilation amplitude is presented.
377: A new contribution missed in the previous analyses of the weak annihilation is reported.
378: Parameter-free numerical estimates of the relevant form factors are given.
379:
380: %_______________________
381: \vspace{.5cm}
382:
383: Chapter VI contains the analysis of the nonfactorizable effects in the $B^0-\bar B^0$ mixing
384: due to soft gluon exchanges. Assuming the dominance of the local gluon condensate, the
385: correction to factorization can be expressed in terms of the specific $B$ meson transition
386: form factors at zero momentum transfers. It is shown that the correction is strictly
387: negative independent of the values of the form factors. The form factors are calculated within the
388: dispersion approach and numerical estimates for them are obtained.
389:
390: %_______________________
391: \vspace{.5cm}
392:
393: In Chapter VII the application of the dispersion approach to inclusive $B\to X_c l\nu$ decays
394: is presented. Spectral representation for the integrated semileptonic rate in terms of the
395: $B$ meson wave function is constructed. The subtraction prescription is discussed and the
396: absence of the $1/m_Q$ correction in the ratio of the bound to free quark decay rates is
397: verified. Differential distributions are calculated in terms of the $B$ meson wave function.
398:
399: %_______________________
400: \vspace{.5cm}
401:
402: Conclusion summarises the main results.
403:
404: