1: \documentclass[prd,twocolumn,aps,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2: %\documentstyle{article}
3: \usepackage{graphicx,graphics,color}% Include figure files
4: \usepackage{dcolumn}% Align table columns on decimal point
5: \usepackage{bm}% bold math
6: \usepackage{psfrag}% psfrag
7:
8: \begin{document}
9: %\draft
10: \newcommand{\be}{\begin{equation}}\newcommand{\ee}{\end{equation}}
11: \newcommand{\bea}{\begin{eqnarray}}\newcommand{\eea}{\end{eqnarray}}
12: \newcommand{\bc}{\begin{center}}\newcommand{\ec}{\end{center}}
13: \def\no{\nonumber}
14: \def\eq#1{Eq. (\ref{#1})}\def\eqeq#1#2{Eqs. (\ref{#1}) and (\ref{#2})}
15: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
16: \def\lsim{\raise0.3ex\hbox{$\;<$\kern-0.75em\raise-1.1ex\hbox{$\sim\;$}}}
17: \def\gsim{\raise0.3ex\hbox{$\;>$\kern-0.75em\raise-1.1ex\hbox{$\sim\;$}}}
18: \def\slash#1{\ooalign{\hfil/\hfil\crcr$#1$}}
19: \def\eff{\mbox{\tiny{eff}}}
20: \def\order#1{{\mathcal{O}}(#1)}
21: \def\etp{\eta^{\prime}}\def\etetp{\eta^{(\prime)}}
22: \def\bketp{B\to K \etp}\def\bket{B\to K \eta}\def\bketetp{B\to K \eta^{(\prime)}}
23: \def\bkzetp{B^0\to K^0 \etp}\def\bkzet{B^0\to K^0 \eta}\def\bkzetetp{B^0\to K^0 \eta^{(\prime)}}
24: \def\bk#1#2{\langle 0|#1|#2 \rangle}\def\la{\langle}\def\ra{\rangle}
25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
26: \preprint{CP3-06-15}
27: \title{Anomalous enhancement of a penguin hadronic matrix element in $\bketp$}
28: \author{
29: J.-M. Gerard\footnote{email address: gerard@fyma.ucl.ac.be} and
30: E. Kou\footnote{email address: ekou@fyma.ucl.ac.be}}
31: \address{Institut de Physique Theorique and
32: Centre for Particle Physics and Phenomenology (CP3)\\
33: Universit\'{e} Catholique de Louvain,
34: Chemin du Cyclotron 2,
35: B-1348 Louvain-la-Neuve, Belgium
36: }
37: \date{\today}
38: \begin{abstract}
39: We estimate the density matrix element for the $\pi^0, \eta$ and $\etp$ production from the vacuum in the large-$N_c$ limit.
40: As a consequence, we find that the QCD axial anomaly leads to highly
41: non-trivial corrections to the usual flavour SU(3) relations
42: between
43: $B^0 \to K^0\pi^0$, $B^0 \to K^0 \eta$ and $B^0 \to K^0 \eta^{\prime}$ decay amplitudes.
44: These corrections may explain why
45: the $B\to K \etp$ branching ratio is about six times larger than the
46: $B\to K \pi$ one.
47: \end{abstract}
48: \pacs{13.20.He}
49: \maketitle
50: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
51: First observations of a large branching ratio for $\bketp$ triggered numerous theoretical investigations within and beyond the standard model. The latest average of the available experimental data~\cite{hfag}:
52: \begin{equation}
53: Br(\bkzetp)= (64.9\pm 3.5)\times 10^{-6}
54: \label{eq:1}
55: \end{equation}
56: definitely confirms a sizable excess of $\etp$ compared with $\pi^0$, $Br(B^0\to K^0\pi^0)=(10.0\pm 0.6)\times 10^{-6}$, and $\eta$, $Br(B^0\to K^0\eta)<1.9\times 10^{-6}$. The fact that $\etp$ is mostly a flavour $SU(3)$-singlet naturally suggests mechanisms such as the so-called singlet contribution ($\etp$ production by gluon-gluon fusion from $b\to sg$ or $b\to sgg$)~\cite{singlet} or intrinsic charm contribution ($\etp$ production by $c\bar{c}$ annihilation from $b\to c\bar{c}s$)~\cite{charm}. However, these do not seem to explain why the $\bkzetp$ branching ratio is about six times larger than the $B^0\to K^0\pi^0$ one. More recently, the hadronic parameters for $\etp$ production have been re-examined and a couple of different solutions were proposed~\cite{BN, zupan,Li}. In this letter, we analyse an overlooked correction from the axial $U(1)$ anomaly in the hadronic matrix elements associated with $b\to s\bar{d}d, s\bar{s}s$.
57:
58: The derivative of the flavour-singlet axial current is given by:
59: \be
60: \partial_{\mu}j^{\mu 5}_0 =2i\sum_{q=u,d,s}m_q\bar{q}\gamma_5q+\frac{3\alpha_s}{4\pi} G_{\alpha\beta}\tilde{G}^{\alpha\beta}\label{eq:deriv}
61: \ee
62: where $G_{\alpha \beta}$ is the gluonic field strength tensor and $\tilde{G}^{\alpha\beta}$, its dual.
63: The mass term in the r.h.s. of Eq. (\ref{eq:deriv}) implies explicit flavour $SU(3)$ violation. This breaking alone would lead to the {\it ideal} mass relations
64: \be (M_{\eta}^2)_{\tiny ideal}=2M_K^2-M_{\pi}^2,
65: \quad (M_{\etp}^2)_{\tiny ideal}=M_{\pi}^2 \label{eq:ideal}
66: \ee
67: which are quite successful for the $(\phi, \omega, \rho)$ vector mesons but totally unrealistic for the $(\eta, \etp, \pi)$ pseudoscalar mesons~\cite{Weinberg}.
68: However, the second term in the r.h.s. of Eq. (\ref{eq:deriv}) represents the QCD anomaly which breaks the axial $U(1)$ symmetry to provide the $\etp$ with a mass around 1 GeV~\cite{tHooft}.
69:
70: The hadronic matrix element $\bk{G_{\alpha\beta}\tilde{G}^{\alpha\beta}}{\etetp}$ associated with the anomalous term in Eq. (\ref{eq:deriv}) has been considered in relation to the Zweig-suppressed $(J/\psi \to \etetp \gamma)$ radiative decays. In this letter, we aim at an estimate of the corresponding hadronic matrix elements for the first term in the r.h.s. of Eq. ({\ref{eq:deriv}). This will then allow us to compute the contribution of the dominant penguin density-density operator to the $(B^0\to K^0 \pi^0, \eta, \etp)$ hadronic decays in a way fully consistent with the axial $U(1)$ and flavour $SU(3)$ symmetry-breaking requirements.
71:
72: In a naive quark picture, the hadronic matrix elements $\bk{\bar{q}\gamma_5q}{\etp, \eta, \pi^0}$ are simply related through Clebsch-Gordan coefficients (C.G.). These coefficients are fixed by the $q\bar{q}$ content in the pseudoscalar wave-functions. However, corrections due to flavour $SU(3)$ but also to axial $U(1)$ violations have to be taken into account. In particular, we expect the following
73: generic form:
74: \be
75: \left.\frac{\bk{\bar{q}\gamma_5q}{\etetp}}{\bk{\bar{q}\gamma_5q}{\pi^0}}\right|_{q=u,d} = \mbox{ C.G.} \big\{1+\frac{M_{\etetp}^2-M_{\pi}^2}{\Lambda^2}\big\}. \label{eq:guess}
76: \ee
77: The appearance of the physical $\etp$ mass may, at first sight, be surprising. But this is in fact required by the axial $U(1)$ symmetry.
78: Indeed, in the absence of the axial anomaly, the {\it ideal} mass relations
79: given in Eq. (\ref{eq:ideal}) would consistently imply that the $\etp/\pi^0$ ratio in Eq (\ref{eq:guess}) is equal to $+1$ if $q=u$ and $-1$ if $q=d$ since the $\etp$ and $\pi^0$ wave-functions have the same quark content as the iso-singlet $\omega$ and the iso-triplet
80: $\rho$ in this fictitious world. The axial anomaly calls thus for a sizable correction to the $\bk{\bar{q}\gamma_5q}{\etp}$ hadronic matrix elements if the cut-off scale is what we naturally expect from the real QCD dynamics, namely $\Lambda=\order{1}$ GeV. So, these matrix elements have to be consistently extracted from a low-energy effective theory of QCD.
81:
82: Again with reference to the observed pattern for the pseudoscalar mass spectrum, let us consider the large-$N_c$ limit at each order in the (squared) momentum $p^2$. The genuine $U(3)_L\times U(3)_R$ chiral invariant structure of the effective non-linear theory implies then the following hierarchy~\cite{GK}:
83: \be
84: \order{p^0,1/N_c} > \order{p^2,1/\infty} > \order{p^4,1/\infty} \cdots \label{eq:hierarchy}
85: \ee
86: such that the full pseudoscalar mass spectrum naturally arises in three steps. The leading ($p^0$) term ensures the breaking of the $U(1)_A$ symmetry and provides the flavour-singlet $\eta_0$ with a mass around 1 GeV. The next-to-leading ($p^2$) term implies the usual $SU(3)_V$ mass splitting among the ($\pi, K, \eta_8$) flavour-octet. Eventually, next-to-next-leading ($p^4$) terms are needed to precisely reproduce the observed $\eta-\etp$ mass splitting.
87:
88: Starting from this effective theory of QCD in the large-$N_c$ limit, we may consistently express the anomalous operator of Eq. (\ref{eq:deriv}) purely in terms of the flavour-singlet field $\eta_0$. This allows us to extract the $\eta-\etp$ mixing angle associated with the diagonalization of the $\eta_8-\eta_0$ squared mass matrix:
89: \bea
90: \eta& =& \eta_8 \cos\theta - \eta_0\sin\theta \nonumber \\
91: \etp&=& \eta_8\sin\theta+\eta_0\cos\theta \label{eq:2-2-3}
92: \eea
93: from the well-measured ($J/\psi \to \etetp\gamma$) radiative decays~\cite{GK}. The extracted value is
94: $\theta=-(22\pm 1)^{\circ}$. Let us emphasise once again that in the absence of the axial anomaly, $\theta$ would have been equal to the {\it ideal} octet-singlet mixing angle $+35.3^{\circ}$ of the $\phi-\omega$, i.e.
95: \be
96: \cos\theta_{\tiny ideal} =\sqrt{2/3}, \quad
97: \sin\theta_{\tiny ideal}=\sqrt{1/3}. \label{eq:idealm}
98: \ee
99: since the ideal mass relations given in Eq. (\ref{eq:ideal}) correspond to the
100: wave-functions $\eta=-s\bar{s}$ and $\etp=(u\bar{u}+d\bar{d})/\sqrt{2}$.
101: This illustrates how $\etetp$ masses
102: and mixing are strongly correlated through the $U(1)_A$ symmetry.
103:
104: Similarly, we may also express the density operator of Eq. (\ref{eq:deriv}) in terms of the flavour-nonet field $\pi$:
105: \bea
106: &&\bar{q}^a\gamma_5 q^{b} \supset i \frac{f\ r}{2\sqrt{2}} \\
107: && \ \ \ \ \ \ \times\left[
108: 1-\frac{\partial_{\mu}\partial^{\mu}}{\Lambda_0^2}+r(m_a+m_b)\left(\frac{1}{\Lambda_1^2}-\frac{1}{4\Lambda_2^2}\right)\right] \pi^{ba}\no
109: \eea
110: where $a, b=1,2,3$ are the flavour indices.
111: The parameters $r$ and $f$ are related to the physical masses and decay constants as
112: \bea
113: &M_{\pi}^2=r m_q[1+M_{\pi}^2(\frac{2}{\Lambda_1^2}-\frac{1}{\Lambda_2^2})] &\\
114: &M_K^2=\frac{r (m_q+m_s)}{2}[1+M_K^2(\frac{2}{\Lambda_1^2}-\frac{1}{\Lambda_2^2})]&
115: \eea
116: and
117: \bea
118: &f_{\pi}=f[1+M_{\pi}^2(\frac{1}{\Lambda_0^2}+\frac{1}{2\Lambda_2^2})] &\\
119: &f_{K}=f[1+M_{K}^2(\frac{1}{\Lambda_0^2}+\frac{1}{2\Lambda_2^2})] &
120: \eea
121: if isospin symmetry is assumed, $q= u \mbox{ or } d$.
122: As a result, we obtain the following set of density matrix elements:
123: \bea
124: &\bullet&\la0| \bar{q} \gamma_5q | \pi^0 \ra =\pm i\frac{f_\pi}{2\sqrt{2}}\frac{M_{\pi}^2}{m_q}\label{eq:r-pi} \\
125: &\bullet&\la0| \bar{q}\gamma_5 s | K \ra =i{f_K}\frac{M_{K}^2}{m_s+m_q} \label{eq:r-k} \\
126: &\bullet&\la0| \bar{q}\gamma_5 q | \eta \ra = \label{eq:r-1}\\
127: & &i\frac{f_\pi}{2\sqrt{6}}\frac{M_{\pi}^2}{m_q}\left(c\theta -\sqrt{2}s\theta\right) \left[1+\frac{M_{\eta}^2-M_\pi^2} {\Lambda_0^2}\right]\nonumber \\
128: &\bullet&\la0| \bar{q}\gamma_5 q | \etp \ra = \label{eq:r-2}\\
129: &&i\frac{f_\pi}{2\sqrt{6}}\frac{M_{\pi}^2}{m_q}\left(\sqrt{2}c\theta +s\theta\right)\left[1+\frac{M_{\eta^{\prime}}^2-M_\pi^2} {\Lambda_0^2}\right] \nonumber \\
130: &\bullet&\la0| \bar{s}\gamma_5 s | \eta \ra = - i\frac{f_K}{\sqrt{3}}\frac{M_{K}^2}{m_s+m_q}\left(\sqrt{2}c\theta +s\theta\right)\label{eq:r-3} \\
131: &&\ \ \ \times \left[1+\frac{M_{\eta}^2-M_K^2}{\Lambda_0^2}+2(M_K^2-M_\pi^2)\left(\frac{1}{\Lambda_1^2}-\frac{1}{4\Lambda_2^2}\right)\right] \no \\
132: &\bullet&\la0| \bar{s}\gamma_5 s | \etp \ra = i\frac{f_K}{\sqrt{3}}\frac{M_{K}^2}{m_s+m_q}\left(c\theta -\sqrt{2}s\theta\right)\label{eq:r-4} \\
133: &&\ \ \ \times \left[1+\frac{M_{\etp}^2-M_K^2}{\Lambda_0^2}+2(M_K^2-M_\pi^2)\left(\frac{1}{\Lambda_1^2}-\frac{1}{4\Lambda_2^2}\right)\right] \no
134: \eea
135: where we introduce the abbreviation $(c\theta, s\theta)$ for $(\cos\theta, \sin\theta)$. On the basis of Eq. (\ref{eq:hierarchy}), we neglect $\order{1/\Lambda_i^4}$ corrections.
136: This parametrisation of the density matrix elements for $\pi^0, \eta$ and $\etp$ follows from a $U(3)_L\times U(3)_R$ invariant theory and is thus fully consistent with the $U(1)_A$ and $SU(3)_V$ symmetry requirements on the pseudoscalar mixing {\it and} masses in the isospin limit.
137:
138: Eqs. (\ref{eq:r-pi}) and (\ref{eq:r-k}) are the well-known hadronic matrix elements already derived in the large-$N_c$ limit using chiral perturbation theory~\cite{chiE}. These hadronic matrix elements feel the usual effects of $SU(3)_V$ violation on the decay constants and masses in the pseudoscalar flavour-octet.
139:
140: Eqs. (\ref{eq:r-1}) and (\ref{eq:r-2}) display the highly non-trivial effect of the $U(1)_A$ breaking: in the absence of the axial anomaly, the {\it ideal} masses (see Eq. (\ref{eq:ideal})) and mixing (see Eq. (\ref{eq:idealm})) would consistently imply $\bk{\bar{q}\gamma_5q}{\eta}=0$ and $\bk{\bar{q}\gamma_5q}{\etp}=\pm\bk{\bar{q}\gamma_5q}{\pi^0}$. They nicely confirm our original guess expressed in Eq. (\ref{eq:guess}).
141: Moreover, a global fit of
142: the pseudoscalar masses, mixing and decays constants has already been undertaken in our previous work~\cite{GK}. The resulting values for the $\Lambda_{0,1,2}$ cut-offs:
143: \be
144: \Lambda_0 \simeq 1.2 \ \mbox{GeV}, \ \ \ \Lambda_1\simeq 1.2\ \mbox{GeV}, \ \ \ \Lambda_2\simeq 1.3\ \mbox{GeV} \label{eq:v6-13}
145: \ee
146: are indeed all around one GeV, as anticipated.
147:
148: Eqs. (\ref{eq:r-3}) and (\ref{eq:r-4}) consistently combine the effects of $U(1)_A$ and $SU(3)_V$ violations. Our results for the scale-independent density matrix elements $2m_s \la 0|\bar{s}\gamma_5s|\etetp \ra$ are compared with previous works in Table 1.
149: \begin{table*}[t!]
150: \begin{center}
151: \scalebox{1.2}{\begin{tabular}{|c|c|c|c|c|}
152: \hline
153: &\multicolumn{2}{|c|}{this work} & \multicolumn{2}{|c|}{previous works} \\
154: \hline
155: $\theta=-22^{\circ}$& $U(1)_A\times SU(3)_V$ & $SU(3)_V$ & AG~\cite{AG}& BN~\cite{BN} \\
156: \hline\hline
157: $2im_s \la 0|\bar{s}\gamma_5s|\eta \ra$ & $+0.053 \pm 0.008$ &$+0.058$& $+0.057$& $+0.055$\\
158: \hline
159: $2im_s \la 0|\bar{s}\gamma_5s|\etp \ra$ &$-0.109 \pm 0.016$&$-0.069$ &$-0.071$&$-0.068$ \\
160: \hline
161: %$\zeta$& 1.11 &1.19& 1.25 & 1.20 \\
162: %\hline
163: %$\zeta^{\prime}$ &1.53&1.08&1.10&1.06\\
164: %\hline
165: \end{tabular}}
166: \caption{Comparison of numerical results for the $\bar{s}\gamma_5s$ density matrix elements in GeV$^3$ units. The second column includes flavour $SU(3)$ breaking with a realistic octet-singlet mixing angle, $\theta=-22^{\circ}$, but with the {\it ideal} mass relations for $\etetp$ (see Eq. (\ref{eq:ideal})). Our result agrees with the previous works in this peculiar limit. The first column includes full $U(1)_A$ and $SU(3)_V$ breaking effects. We find that the magnitude of the $\etp$ hadronic matrix element increases by about 60\% while the one for $\eta$ only decreases by about 10\%. }
167: \end{center}
168: \end{table*}
169: The second column includes flavour $SU(3)$ breaking with a realistic octet-singlet mixing angle but with the {\it ideal} mass relations for $\etetp$ (see Eq. (\ref{eq:ideal})).
170: We can see an excellent numerical agreement in this limit which is
171: rather peculiar since any realistic $\eta-\etp$ mixing excludes ideal
172: $\eta-\etp$ masses from the viewpoint of the $U(1)_A$ symmetry.
173: The first column includes full $U(1)_A$ and $SU(3)_V$ breaking effects. The magnitude of the $\etp$ hadronic matrix element increases then by about 60\% while the one for $\eta$ only decreases by about 10\%.
174: %This large correction for $\bk{\bar{s}\gamma_5s}{\etp}$ should not come as a surprise if one realizes from Eq. (\ref{eq:idealm}) that this hadronic matrix element exactly vanishes in the absence of the axial
175: %anomaly!
176: Theoretical uncertainties associated with higher order
177: $SU(3)_V$ and $U(1)_A$ corrections are expected to be $M_K^4/\Lambda^4\simeq 5$\% and $M_K^2M_{\etetp}^2/\Lambda^4\simeq 15$\%, respectively.
178:
179:
180: Eqs. (\ref{eq:r-pi}) to (\ref{eq:r-4}) turn out to be crucial for an estimate of the $\bkzetetp$ decay amplitudes. Indeed, their typical $M^2/m$ chiral enhancement, at the basis of the $\Delta I=1/2$ rule in $K\to \pi\pi$ decays~\cite{BBG}, is such that the
181: $Q_6\equiv \sum_{q}(\bar{b}_Lq_R)(\bar{q}_Rs_L)$ weak operator provides the main contribution for the $b\to s\bar{q}q$ penguin-induced hadronic processes involving two pseudoscalars in the final state.
182: (Notice that such is not the case for the vector-pseudoscalar final state, $B\to K^* \etetp$.)
183: In the limit where, diagrammatically,
184: $\bkzetetp$ come from $b\to s\bar{d}d$ and $b\to s\bar{s}s$ penguins, while $B^0\to K^0\pi^0$ comes only from $b\to s\bar{d}d$, we may thus write
185: \bea
186: &&\left. \frac{A(\bkzetetp)}{A(B^0\to K^0\pi^0)}\right|_{Q_6} \label{eq:q6}\\
187: &=& \frac{ \la K^0 |\bar{b}_Ls_R|{B^0}\ra}{\la \pi^0 |\bar{b}_Ld_R|{B^0}\ra}
188: \left[\frac{\la \etetp |\bar{b}_Ld_R|{B^0}\ra}{ \la K^0 |\bar{b}_Ls_R|{B^0}\ra}+\frac{\la\etetp|\bar{s}_Rs_L|0\ra}{\la K^0|\bar{d}_Rs_L|0\ra}\right]. \no
189: \eea
190: We already notice that in the absence of the axial anomaly, the $\etp$
191: wave-function has no $s\bar{s}$ component and the $b\to s\bar{s}s$ penguin
192: contribution to $B^0\to K^0\etp$ vanishes. In that {\it ideal} limit, the
193: $\etp/\pi^0$ ratio in Eq.(\ref{eq:q6}) is equal to $-1$ and thus $Br(\bkzetp) =
194: Br(B^0\to K^0\pi^0)$!
195:
196:
197: \begin{figure*}[t]
198: \begin{center}
199: \psfrag{u1}[c][l][1]{}
200: %\begin{minipage}{4cm}$U(1)$ breaking correction to hadronic matrix element\end{minipage}}
201: \psfrag{arr}[c][c][.8]{\begin{picture}(2,4)\put(0,0){\vector(0,1){12}}\end{picture}}
202: \psfrag{arrp}[c][c][.8]{\begin{picture}(2,4)\put(0,0){\vector(0,1){22}}\end{picture}}
203: \psfrag{x}[c][c][1]{$\zeta^{\prime}$}
204: \psfrag{y}[c][c][1]{$Br(\bkzet)/Br(B^0\to K^0 \pi^0)$}
205: \psfrag{yp}[c][c][1]{$Br(\bkzetp)/Br(B^0\to K^0 \pi^0)$}
206: \psfrag{exp}[c][c][1]{}
207: \psfrag{no}[c][c][1.1]{$\bullet$}\psfrag{yes}[c][c][2]{$\star$}
208: \psfrag{a}[c][c][1.2]{\rotatebox{23}{\bf /}}\psfrag{b}[c][c][1.2]{\rotatebox{23}{\bf /}}
209: \psfrag{noc}[c][c][0.9]{$SU(3)_V$}\psfrag{yesc}[c][c][0.9]{$U(1)_A\times SU(3)_V$}
210: \psfrag{1}[c][c][1]{1}\psfrag{1.2}[c][c][1]{1.2}\psfrag{1.3}[c][c][1]{1.3}\psfrag{1.4}[c][c][1]{1.4}\psfrag{1.5}[c][c][1]{1.5}\psfrag{1.6}[c][c][1]{1.6}\psfrag{1.7}[c][c][1]{1.7}\psfrag{1.8}[c][c][1]{1.8}\psfrag{2.2}[c][c][1]{2.2}
211: \psfrag{1}[c][c][1]{1}\psfrag{2}[c][c][1]{2}\psfrag{3}[c][c][1]{3}\psfrag{4}[c][c][1]{4}\psfrag{5}[c][c][1]{5}\psfrag{6}[c][c][1]{6}\psfrag{7}[c][c][1]{7}\psfrag{8}[c][c][1]{8}\psfrag{10}[c][c][1]{10}
212: \psfrag{0.2}[c][c][1]{0.2}\psfrag{0.175}[c][c][1]{}\psfrag{0.15}[c][c][1]{0.15}\psfrag{0.125}[c][c][1]{}
213: \psfrag{0.1}[c][c][1]{0.1}\psfrag{0.075}[c][c][1]{}\psfrag{0.05}[c][c][1]{0.05}\psfrag{0.025}[c][c][1]{}
214: \includegraphics[width=8cm]{bretp.eps}\hspace*{1cm}
215: %\includegraphics[width=8cm]{breta.eps}
216: \caption{
217: The $\bkzetp$ to $B^0\to K^0\pi^0$ ratio of branching ratios as a
218: function of the $\etp$ mass-dependent parameter $\zeta^{\prime}$, for a fixed value
219: of the mixing angle $\theta = -22^{\circ}$.
220: The usual flavour $SU(3)$ relation, $Br(\bkzetp)/Br(B^0\to K^0\pi^0)\simeq 3$, is obtained at $\zeta^{\prime}=1$. The dot corresponds to $SU(3)_V$ but no $U(1)_A$ corrections to $\zeta^{\prime}$, i.e. $\zeta^{\prime}=1.09$.
221: Finally, the star represents full $SU(3)_V$ and $U(1)_A$ corrections to $\zeta^{\prime}$, i.e. $\zeta^{\prime}=1.72\pm 0.26$. The shaded area
222: displays the current experimental range.
223: }
224: \label{fig:1}
225: \end{center}
226: \end{figure*}
227:
228: As a first order approximation, the effect of the $U(1)_A$ and $SU(3)_V$
229: violations on the pseudoscalar masses can be safely neglected compared to the $B$-mass scale. So, we simply express the $B$-to-light meson hadronic matrix elements in terms of C.G. and focus on the axial $U(1)$ and flavour $SU(3)$ breaking corrections for the vacuum-to-light meson transitions. In this approximation, we easily obtain:
230: \bea
231: &&\frac{A(\bkzet)}{A(B^0\to K^0\pi^0)} \label{eq:eta}\\
232: &=&-\sqrt{2}\left[(\sqrt{\frac{1}{6}}c\theta-\sqrt{\frac{1}{3}}s\theta)+(-\frac{2}{\sqrt{6}}c\theta-\sqrt{\frac{1}{3}}s\theta)\zeta\right]\no \\
233: &&\frac{A(\bkzetp)}{A(B^0\to K^0\pi^0)} \label{eq:etp}\\
234: &=& -\sqrt{2}\left[(\sqrt{\frac{1}{6}}s\theta+\sqrt{\frac{1}{3}}c\theta)+(-\frac{2}{\sqrt{6}}s\theta+\sqrt{\frac{1}{3}}c\theta)\zeta^{\prime}\right]\no
235: \eea
236: where the $U(1)_A$ and $SU(3)_V$ breaking effects associated with the pseudoscalar masses are fully encoded in the parameters
237: \be
238: \zeta^{(\prime)}\equiv 1+\frac{M_{\etetp}^2-M_{K}^2}{\Lambda_0^2}+2(M_K^2-M_{\pi}^2)(\frac{1}{\Lambda_1^2}-\frac{1}{4\Lambda_2^2}).
239: \ee
240: For a sensible value of the mixing angle, $\theta=-19.5^{\circ}$ (i.e. $\cos\theta=2\sqrt{2}/3$ and $\sin\theta=-1/3$), we have $\eta=(u\bar{u}+d\bar{d}-s\bar{s})/\sqrt{3}$ and $\etp=(u\bar{u}+d\bar{d}+2s\bar{s})/\sqrt{6}$, such that Eqs. (\ref{eq:eta}) and (\ref{eq:etp}) simply reduce to
241: \bea
242: \left.\frac{A(\bkzet)}{A(B^0\to K^0\pi^0)}\right|_{\theta=-19.5^{\circ}}
243: &=&-\sqrt{\frac{2}{3}}\left[1-\zeta\right] \label{eq:lipkineta}\\
244: \left.\frac{A(\bkzetp)}{A(B^0\to K^0\pi^0)}\right|_{\theta=-19.5^{\circ}}
245: &=&-\sqrt{\frac{1}{3}}\left[1+2\zeta^{\prime}\right] .
246: \eea
247: In the absence of the axial anomaly, the ideal relations given in Eq. (\ref{eq:ideal}) for
248: $M_{\etetp}$ imply $\zeta^{(\prime)}=1.41 (1.09)$. If in addition we assume $M_K = M_{\pi}$, then $\zeta=\zeta^{\prime}=1$ and
249: we recover the
250: well-known flavour $SU(3)$ relations between the $B^0\to K^0 [0^{-+}]$ branching ratios ($\etp : \eta : \pi^0= 3:0:1$)~\cite{su3}.
251: But again, the $U(1)_A$ symmetry requires physical $\etetp$ masses for a
252: physical $\eta-\etp$ mixing angle. Hence we have
253: \be
254: \zeta=1.29\pm 0.19, \quad \zeta^{\prime}=1.72\pm 0.26. \label{eq:nzeta}
255: \ee
256: The impact of the $U(1)$ anomaly on the $B^0 \to K^0\etp$ branching ratio
257: is displayed in Fig.1 for a mixing angle $\theta = -22^{\circ}$. The result
258: with only $SU(3)_V$ breaking corrections (a dot in Fig.1) is obtained
259: for $\zeta^{\prime} = 1.09$. As we have shown in Table 1, this peculiar limit
260: reproduces the numerical values of the hadronic matrix elements used
261: in the previous works. Our result which consistently includes the effect
262: of the $U(1)_A$ violation on the masses and mixing (a star in Fig.1) is
263: based on Eq. (\ref{eq:nzeta}) for $\zeta^{\prime}$. We observe a strong increase for the $B^0 \to
264: K^0 \etp$ branching ratio, in agreement with the data.
265: On the other hand, the relatively small $\eta$ mass induces a small $U(1)_A$ correction to $\zeta$, which does deviate from one mainly through the $SU(3)_V$ violation. As a result, the $\bkzet$ branching ratio stays well below the experimental bound due to the efficient cancellation in Eq. (\ref{eq:lipkineta}) but is, at the same time,
266: extremely sensitive to
267: the uncertainties on $\zeta$ and $\theta$. For $\zeta = 1.29\pm 0.19$ and $\theta =
268: -(22 \pm 1)^{\circ}$, its value runs between $0.01 \times10^{-6}$ and $1.10\times 10^{-6}$. At this
269: level, contributions from the other penguin operators might be non-negligible.
270:
271:
272:
273: In conclusion, we reconsidered the puzzle of large $\bketp$ branching ratio and pointed out a missing $U(1)_A$ breaking correction to the penguin hadronic matrix element. We estimated this correction in the framework of a low-energy effective theory of QCD in the large-$N_c$ limit. We provided the expression for all the density matrix elements relevant to the hadronic $B$ decays, in terms of physical masses and decay constants.
274: We found a rather large increase (60\%) of
275: $| \la 0|\bar{s}\gamma_5s|\etp \ra|$ and a moderate decrease (10\%) of $| \la 0|\bar{s}\gamma_5s|\eta \ra|$ compared to the previous works.
276: The sizable corrections for the $\etp$
277: should not come as a surprise since its hadronic matrix element
278: vanishes in the absence of the axial anomaly. This correction
279: may explain why the $\bkzetp$ branching ratio is about six times larger than the $B^0\to K^0\pi^0$ one.
280:
281: %We have illustrated then that this correction plays an important role to enhance the $\bkzetp$ branching ratio since it enters as the dominant contribution. We are not expecting such an enhancement in $B\to K^*\etetp$.
282:
283: \bigskip
284:
285: \begin{acknowledgments}
286: This work was supported by the Belgian
287: Federal Office for Scientific, Technical and Cultural Affairs through the
288: Interuniversity Attraction Pole P5/27.
289: \end{acknowledgments}
290:
291: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
292: \begin{thebibliography}{99}
293: %\cite{:2006bi}
294: \bibitem{hfag}
295: [Heavy Flavor Averaging Group (HFAG)],
296: %``Averages of b-hadron properties at the end of 2005,''
297: arXiv:hep-ex/0603003.
298: %%CITATION = HEP-EX 0603003;%%
299: \bibitem{singlet}
300: %\cite{Atwood:1997bn}
301: %\bibitem{Atwood:1997bn}
302: D.~Atwood and A.~Soni,
303: %``B --> eta' + X and the QCD anomaly,''
304: Phys.\ Lett.\ B {\bf 405} (1997) 150
305: [arXiv:hep-ph/9704357]. \\
306: %%CITATION = HEP-PH 9704357;%%
307: %\cite{Ahmady:1997fa}
308: %\bibitem{Ahmady:1997fa}
309: M.~R.~Ahmady, E.~Kou and A.~Sugamoto,
310: % ``Non-spectator contribution: A mechanism for inclusive B --> X/s eta' and
311: %exclusive B --> K* eta' decays,''
312: Phys.\ Rev.\ D {\bf 58} (1998) 014015
313: [arXiv:hep-ph/9710509]. \\
314: %%CITATION = HEP-PH 9710509;%%
315: %\cite{Ali:1997ex}
316: %\bibitem{Ali:1997ex}
317: A.~Ali, J.~Chay, C.~Greub and P.~Ko,
318: % ``Contribution of b --> s g g through the QCD anomaly in exclusive decays
319: %B+- --> (eta',eta) (K+-,K*+-) and B0 --> (eta',eta) (K0,K*0),''
320: Phys.\ Lett.\ B {\bf 424} (1998) 161
321: [arXiv:hep-ph/9712372].\\
322: %%CITATION = HEP-PH 9712372;%%
323: %\cite{Dighe:1995gq}
324: %\bibitem{Dighe:1995gq}
325: A.~S.~Dighe, M.~Gronau and J.~L.~Rosner,
326: %``Amplitude relations for $B$ decays involving $\eta$ and $\eta'$,''
327: Phys.\ Lett.\ B {\bf 367} (1996) 357
328: [Erratum-ibid.\ B {\bf 377} (1996) 325]
329: [arXiv:hep-ph/9509428].
330: %%CITATION = HEP-PH 9509428;%%
331: \bibitem{charm}
332: %\cite{Halperin:1997as}
333: %\bibitem{Halperin:1997as}
334: I.~E.~Halperin and A.~Zhitnitsky,
335: %``B --> K eta' decay as unique probe of eta' meson,''
336: Phys.\ Rev.\ D {\bf 56} (1997) 7247
337: [arXiv:hep-ph/9704412]. \\
338: %%CITATION = HEP-PH 9704412;%
339: %\cite{Petrov:1997yf}
340: %\bibitem{Petrov:1997yf}
341: A.~A.~Petrov,
342: %``Intrinsic charm of light mesons and CP violation in heavy quark decay,''
343: Phys.\ Rev.\ D {\bf 58} (1998) 054004
344: [arXiv:hep-ph/9712497].
345: %%CITATION = HEP-PH 9712497;%%
346: %\cite{Ali:1997nh}
347: \bibitem{BN}
348: M.~Beneke and M.~Neubert,
349: %``Flavor-singlet B decay amplitudes in QCD factorization,''
350: Nucl.\ Phys.\ B {\bf 651} (2003) 225
351: [arXiv:hep-ph/0210085].
352: %%CITATION = HEP-PH 0210085;%%
353: %\cite{Williamson:2006hb}
354: \bibitem{zupan}
355: A.~R.~Williamson and J.~Zupan,
356: %``Two body B decays with isosinglet final states in SCET,''
357: Phys.\ Rev.\ D {\bf 74} (2006) 014003
358: [Erratum-ibid.\ D {\bf 74} (2006) 03901]
359: [arXiv:hep-ph/0601214].
360: %%CITATION = HEP-PH 0601214;%%
361: %\cite{Charng:2006zj}
362: \bibitem{Li}
363: Y.~Y.~Charng, T.~Kurimoto and H.~n.~Li,
364: %``Gluonic contribution to B --> eta(') form factors,''
365: arXiv:hep-ph/0609165.
366: %%CITATION = HEP-PH 0609165;%%
367: \bibitem{Weinberg}
368: S.~Weinberg,
369: %``The U(1) Problem,''
370: Phys.\ Rev.\ D {\bf 11} (1975) 3583.
371: %%CITATION = PHRVA,D11,3583;%%
372: \bibitem{tHooft}
373: G.~'t Hooft,
374: %``Symmetry Breaking Through Bell-Jackiw Anomalies,''
375: Phys.\ Rev.\ Lett.\ {\bf 37} (1976) 8. \\
376: %%CITATION = PRLTA,37,8;%%
377: R.~Jackiw and C.~Rebbi,
378: %``Vacuum Periodicity In A Yang-Mills Quantum Theory,''
379: Phys.\ Rev.\ Lett.\ {\bf 37} (1976) 172. \\
380: %%CITATION = PRLTA,37,172;%%
381: %\cite{Callan:1976je}
382: C.~G.~Callan, R.~F.~Dashen and D.~J.~Gross,
383: %``The Structure Of The Gauge Theory Vacuum,''
384: Phys.\ Lett.\ B {\bf 63} (1976) 334.
385: %%CITATION = PHLTA,B63,334;%%
386: \bibitem{GK}
387: %\cite{Gerard:2004gx}
388: %\bibitem{Gerard:2004gx}
389: J.~M.~Gerard and E.~Kou,
390: %``eta - eta' masses and mixing: A large N(c) reappraisal,''
391: Phys.\ Lett.\ B {\bf 616} (2005) 85
392: [arXiv:hep-ph/0411292].
393: %%CITATION = HEP-PH 0411292;%%
394: \bibitem{chiE}
395: R.~S.~Chivukula, J.~M.~Flynn and H.~Georgi,
396: %``Polychromatic Penguins Don't Fly,''
397: Phys.\ Lett.\ B {\bf 171} (1986) 453. \\
398: %%CITATION = PHLTA,B171,453;%%
399: W.~A.~Bardeen, A.~J.~Buras and J.-M.~Gerard,
400: %``The Delta I = 1/2 Rule In The Large N Limit,''
401: Phys.\ Lett.\ B {\bf 180} (1986) 133.
402: %%CITATION = PHLTA,B180,133;%%
403: \bibitem{BBG}
404: %\cite{Bardeen:1986uz}
405: %\bibitem{Bardeen:1986uz}
406: W.~A.~Bardeen, A.~J.~Buras and J.~M.~Gerard,
407: %``THE K $\to$ pi pi DECAYS IN THE LARGE N LIMIT: QUARK EVOLUTION,''
408: Nucl.\ Phys.\ B {\bf 293} (1987) 787. \\
409: %%CITATION = NUPHA,B293,787;%%% phys lett B192 138
410: %\cite{Bardeen:1986vz}
411: %\bibitem{Bardeen:1986vz}
412: W.~A.~Bardeen, A.~J.~Buras and J.~M.~Gerard,
413: %``A CONSISTENT ANALYSIS OF THE DELTA I = 1/2 RULE FOR K DECAYS,''
414: Phys.\ Lett.\ B {\bf 192} (1987) 138.
415: %%CITATION = PHLTA,B192,138;%%
416: \bibitem{AG}
417: A.~Ali and C.~Greub,
418: % ``An analysis of two-body non-leptonic B decays involving light mesons in
419: %the standard model,''
420: Phys.\ Rev.\ D {\bf 57} (1998) 2996
421: [arXiv:hep-ph/9707251].
422: %%CITATION = HEP-PH 9707251;%%
423: %\cite{Beneke:2002jn}
424: \bibitem{su3}
425: %\cite{Lipkin:1998ew}
426: %\bibitem{Lipkin:1998ew}
427: H.~J.~Lipkin,
428: %``FSI rescattering in B+- decays via states with eta, eta' omega and Phi,''
429: Phys.\ Lett.\ B {\bf 433} (1998) 117. \\
430: %%CITATION = PHLTA,B433,117;%%
431: %\cite{Lipkin:1997ad}
432: %\bibitem{Lipkin:1997ad}
433: H.~J.~Lipkin,
434: % ``Penguins, trees and final state interactions in B decays in broken
435: %SU(3),''
436: Phys.\ Lett.\ B {\bf 415} (1997) 186
437: [arXiv:hep-ph/9710342]. \\
438: %%CITATION = HEP-PH 9710342;%%
439: %\cite{Dighe:1997hm}
440: %\bibitem{Dighe:1997hm}
441: A.~S.~Dighe, M.~Gronau and J.~L.~Rosner,
442: %``B decays involving eta and eta' in light of the B --> K eta' process,''
443: Phys.\ Rev.\ Lett.\ {\bf 79} (1997) 4333
444: [arXiv:hep-ph/9707521]. \\
445: %\cite{Dighe:1997wj}
446: %\bibitem{Dighe:1997wj}
447: A.~S.~Dighe, M.~Gronau and J.~L.~Rosner,
448: %``B decays to charmless V P final states,''
449: Phys.\ Rev.\ D {\bf 57} (1998) 1783
450: [arXiv:hep-ph/9709223]. \\
451: %%CITATION = HEP-PH 9709223;%%
452: %\cite{Datta:2001hd}
453: %\bibitem{Datta:2001hd}
454: A.~Datta, H.~J.~Lipkin and P.~J.~O'Donnell,
455: %``Weak decays to final states with radial excitation admixtures,''
456: arXiv:hep-ph/0102070.
457: %%CITATION = HEP-PH 0102070;%%
458: \end{thebibliography}
459: \end{document}
460:
461: