1: %\documentstyle[aps,prd,epsfig]{revtex}
2: %\begin{document}
3: %{\em Appendix: The scalar form factor}
4:
5: \vspace{.5cm}
6: \underline{\em The scalar form factor}
7: \vspace{.5cm}
8:
9: Let us consider the form factor of the transition between two pseudoscalar
10: mesons induced by the scalar current
11: \begin{eqnarray}
12: \label{scalar1}
13: \langle P(M_2,p_2)|\bar q_1 q_2|P(M_1,p_1)\rangle &=&f_s(q^2)
14: \end{eqnarray}
15: For the analysis of the heavy-quark transition we introduce the velocity-dependent
16: form factor as follows
17: \begin{eqnarray}
18: \langle P(M_2,p_2)|\bar q_1 q_2|P(M_1,p_1)\rangle &=&\sqrt{M_1
19: M_2}(1+\omega)h_{f_s}(\omega)
20: \end{eqnarray}
21: such that
22: \begin{eqnarray}
23: f_s(q^2)=\sqrt{M_1 M_2}(1+\omega)h_{f_s}(\omega).
24: \end{eqnarray}
25: From HQET we find the following expansion
26: \begin{eqnarray}
27: \label{scalar2}
28: h_{f_s}=\xi+\left(\frac{1}{m_1}+\frac{1}{m_2}
29: \right)
30: \left[
31: \frac{\omega-1}{\omega+1}
32: \left(\frac{\bar\Lambda}2\xi-\xi_3\right)+\rho_1\right].
33: \end{eqnarray}
34: An important consequence of this expansion is the relation
35: \begin{eqnarray}
36: h_{f_s}(1)=1+O(1/m_Q^2),
37: \end{eqnarray}
38: which shows that the deviation of $h_{f_s}(1)$ from unity emerges only
39: at the $1/m_Q^2$ order, as well as for the form factors
40: $h_{f}$, $h_{f_+}$, and $h_{g_+}$ (\ref{luke1}).
41:
42: Notice that in the leading and subleading orders in $1/m_Q$
43: the following quantities have the same expansion
44: \begin{eqnarray}
45: f_s&\sim& 2(p_1p_2-m_1m_2)s+m_2 f_1+m_1 f_2
46: \nonumber\\
47: &\sim& 2(p_1p_2-M_1M_2)s+M_1 f_1+M_2 f_2.
48: \end{eqnarray}
49:
50: The form factor can be obtained according to the relation
51: \begin{eqnarray}
52: \label{class1}
53: f_s(q^2)=\frac{1}{m_2-m_1}
54: q^\mu\langle P(M_2,p_2)|\bar q_1 \gamma_\mu q_2|P(M_1,p_1)\rangle,
55: \end{eqnarray}
56: which follows from the equations of motion for quark fields. We then find
57: \begin{eqnarray}
58: \label{scalar7}
59: f_s(q^2)=\frac{1}{m_2-m_1}\left[(M_1^2-M_2^2)f_+(q^2)+q^2\,f_-(q^2)\right]
60: \end{eqnarray}
61: and
62: \begin{eqnarray}
63: h_{f_s}=\frac{1}{m_2-m_1}\left[
64: h_{f_+}(M_1-M_2)(\omega+1)
65: -h_{f_-}(M_1+M_2)(\omega-1)\right].
66: \end{eqnarray}
67: In this case the structure of the expansion (\ref{scalar2}) is satisfied
68: automatically.
69:
70: In the dispersion approach the double discontinuity of the form factor $f_s$
71: is found from the Feynman graph
72: \begin{eqnarray}
73: \tilde f_s^D=
74: 2m_3\left((m_1-m_2)^2-q^2\right)
75: +2m_2\left(s_2-(m_1-m_3)^2\right)
76: +2m_1\left(s_1-(m_2-m_3)^2\right)
77: -4m_1m_2m_3.
78: \end{eqnarray}
79: The double discontinuities of the form factors $f_+$, $f_-$ and $f_s$ obey the classical equation of motion
80: \begin{eqnarray}
81: \label{class2}
82: \tilde f^D_s(s_1,s_2,q^2)=\frac{1}{m_2-m_1}\left[(s_1-s_2)\tilde f_+(s_1,s_2,q^2)+q^2 \tilde f_-(s_1,s_2,q^2)\right].
83: \end{eqnarray}
84: As becomes clear from comparing Eqs.(\ref{class1}) and (\ref{class2}),
85: in order to satisfy the stucture of the expansion (\ref{scalar2}) the form factor
86: requires subtraction. The spectral density which includes the proper subtraction
87: term reads
88: \begin{eqnarray}
89: \tilde f_s=\tilde f_s^D+\left[(M_1^2-s_1)-(M_2^2-s_2)\right]\frac{\tilde f_+}{m_2-m_1}.
90: \end{eqnarray}
91: Equivalently, the behavior of the form factor $f_s$ can be brought in accordance
92: with HQET in the NLO by defining the subtraction as follows
93: \begin{eqnarray}
94: \tilde f_s=\tilde f_s^D+\left[(M_1^2-s_1)+(M_2^2-s_2)\right]\tilde \rho_{f_s},
95: \end{eqnarray}
96: provided $\tilde \rho_{f_s}\sim \tilde \rho_{g_0} \sim \tilde \rho_{a}$, Eqs.
97: (\ref{rhoa}) and (\ref{rhog0}).
98: %\end{document}
99:
100:
101:
102: