hep-ex0510066/gp.tex
1: \documentclass[12pt,epsfig,prd,showpacs]{revtex4}
2: \setlength{\oddsidemargin}{0.5cm}
3: \setlength{\topmargin}{0cm}
4: \setlength{\headheight}{0cm}
5: \setlength{\textheight}{23cm}
6: \setlength{\textwidth}{15cm}
7: 
8: \usepackage{graphicx}
9: \usepackage{epsfig}
10: \usepackage{dcolumn}
11: \usepackage{bm}
12: 
13: \newcommand{\etac}{\eta_c}
14: \newcommand{\etap}{\eta^{\prime}}
15: \newcommand{\jpsi}{J/\psi}
16: \newcommand{\pp}{\pi^0\pi^0}
17: \newcommand{\pppm}{\pi^+\pi^-}
18: \newcommand{\ar}{\rightarrow}
19: 
20: \begin{document}
21: \vspace{-30mm}
22: \title{\bf \boldmath Measurement of the branching fractions for
23: $J/\psi\rightarrow\gamma\pi^0$, $\gamma\eta$ and $\gamma\eta^{\prime}$}
24: \author{
25: M.~Ablikim$^{1}$,              J.~Z.~Bai$^{1}$,               Y.~Ban$^{11}$,
26: J.~G.~Bian$^{1}$,              X.~Cai$^{1}$,                  H.~F.~Chen$^{16}$,
27: H.~S.~Chen$^{1}$,              H.~X.~Chen$^{1}$,              J.~C.~Chen$^{1}$,
28: Jin~Chen$^{1}$,                Y.~B.~Chen$^{1}$,              S.~P.~Chi$^{2}$,
29: Y.~P.~Chu$^{1}$,               X.~Z.~Cui$^{1}$,               Y.~S.~Dai$^{18}$,
30: Z.~Y.~Deng$^{1}$,              L.~Y.~Dong$^{1}$$^{a}$,        Q.~F.~Dong$^{14}$,
31: S.~X.~Du$^{1}$,                Z.~Z.~Du$^{1}$,                J.~Fang$^{1}$,
32: S.~S.~Fang$^{2}$,              C.~D.~Fu$^{1}$,                C.~S.~Gao$^{1}$,
33: Y.~N.~Gao$^{14}$,              S.~D.~Gu$^{1}$,                Y.~T.~Gu$^{4}$,
34: Y.~N.~Guo$^{1}$,               Y.~Q.~Guo$^{1}$,               Z.~J.~Guo$^{15}$,
35: F.~A.~Harris$^{15}$,           K.~L.~He$^{1}$,                M.~He$^{12}$,
36: Y.~K.~Heng$^{1}$,              H.~M.~Hu$^{1}$,                T.~Hu$^{1}$,
37: G.~S.~Huang$^{1}$$^{b}$,       X.~P.~Huang$^{1}$,             X.~T.~Huang$^{12}$,
38: X.~B.~Ji$^{1}$,                X.~S.~Jiang$^{1}$,             J.~B.~Jiao$^{12}$,
39: D.~P.~Jin$^{1}$,               S.~Jin$^{1}$,                  Yi~Jin$^{1}$,
40: Y.~F.~Lai$^{1}$,               G.~Li$^{2}$,                   H.~B.~Li$^{1}$,
41: H.~H.~Li$^{1}$,                J.~Li$^{1}$,                   R.~Y.~Li$^{1}$,
42: S.~M.~Li$^{1}$,                W.~D.~Li$^{1}$,                W.~G.~Li$^{1}$,
43: X.~L.~Li$^{8}$,                X.~Q.~Li$^{10}$,               Y.~L.~Li$^{4}$,
44: Y.~F.~Liang$^{13}$,            H.~B.~Liao$^{6}$,              C.~X.~Liu$^{1}$,
45: F.~Liu$^{6}$,                  Fang~Liu$^{16}$,               H.~H.~Liu$^{1}$,
46: H.~M.~Liu$^{1}$,               J.~Liu$^{11}$,                 J.~B.~Liu$^{1}$,
47: J.~P.~Liu$^{17}$,              R.~G.~Liu$^{1}$,               Z.~A.~Liu$^{1}$,
48: F.~Lu$^{1}$,                   G.~R.~Lu$^{5}$,                H.~J.~Lu$^{16}$,
49: J.~G.~Lu$^{1}$,                C.~L.~Luo$^{9}$,               F.~C.~Ma$^{8}$,
50: H.~L.~Ma$^{1}$,                L.~L.~Ma$^{1}$,                Q.~M.~Ma$^{1}$,
51: X.~B.~Ma$^{5}$,                Z.~P.~Mao$^{1}$,               X.~H.~Mo$^{1}$,
52: J.~Nie$^{1}$,                  S.~L.~Olsen$^{15}$,            H.~P.~Peng$^{16}$,
53: N.~D.~Qi$^{1}$,                H.~Qin$^{9}$,                  J.~F.~Qiu$^{1}$,
54: Z.~Y.~Ren$^{1}$,               G.~Rong$^{1}$,                 L.~Y.~Shan$^{1}$,
55: L.~Shang$^{1}$,                D.~L.~Shen$^{1}$,              X.~Y.~Shen$^{1}$,
56: H.~Y.~Sheng$^{1}$,             F.~Shi$^{1}$,                  X.~Shi$^{11}$$^{c}$,
57: H.~S.~Sun$^{1}$,               J.~F.~Sun$^{1}$,               S.~S.~Sun$^{1}$,
58: Y.~Z.~Sun$^{1}$,               Z.~J.~Sun$^{1}$,               Z.~Q.~Tan$^{4}$,
59: X.~Tang$^{1}$,                 Y.~R.~Tian$^{14}$,             G.~L.~Tong$^{1}$,
60: G.~S.~Varner$^{15}$,           D.~Y.~Wang$^{1}$,              L.~Wang$^{1}$,
61: L.~S.~Wang$^{1}$,              M.~Wang$^{1}$,                 P.~Wang$^{1}$,
62: P.~L.~Wang$^{1}$,              W.~F.~Wang$^{1}$$^{d}$,        Y.~F.~Wang$^{1}$,
63: Z.~Wang$^{1}$,                 Z.~Y.~Wang$^{1}$,              Zhe~Wang$^{1}$,
64: Zheng~Wang$^{2}$,              C.~L.~Wei$^{1}$,               D.~H.~Wei$^{1}$,
65: N.~Wu$^{1}$,                   X.~M.~Xia$^{1}$,               X.~X.~Xie$^{1}$,
66: B.~Xin$^{8}$$^{b}$,            G.~F.~Xu$^{1}$,                Y.~Xu$^{10}$,
67: M.~L.~Yan$^{16}$,              F.~Yang$^{10}$,                H.~X.~Yang$^{1}$,
68: J.~Yang$^{16}$,                Y.~X.~Yang$^{3}$,              M.~H.~Ye$^{2}$,
69: Y.~X.~Ye$^{16}$,               Z.~Y.~Yi$^{1}$,                G.~W.~Yu$^{1}$,
70: C.~Z.~Yuan$^{1}$,              J.~M.~Yuan$^{1}$,              Y.~Yuan$^{1}$,
71: S.~L.~Zang$^{1}$,              Y.~Zeng$^{7}$,                 Yu~Zeng$^{1}$,
72: B.~X.~Zhang$^{1}$,             B.~Y.~Zhang$^{1}$,             C.~C.~Zhang$^{1}$,
73: D.~H.~Zhang$^{1}$,             H.~Y.~Zhang$^{1}$,             J.~W.~Zhang$^{1}$,
74: J.~Y.~Zhang$^{1}$,             Q.~J.~Zhang$^{1}$,             X.~M.~Zhang$^{1}$,
75: X.~Y.~Zhang$^{12}$,            Yiyun~Zhang$^{13}$,            Z.~P.~Zhang$^{16}$,
76: Z.~Q.~Zhang$^{5}$,             D.~X.~Zhao$^{1}$,              J.~W.~Zhao$^{1}$,
77: M.~G.~Zhao$^{10}$,             P.~P.~Zhao$^{1}$,              W.~R.~Zhao$^{1}$,
78: Z.~G.~Zhao$^{1}$$^{e}$,        H.~Q.~Zheng$^{11}$,            J.~P.~Zheng$^{1}$,
79: Z.~P.~Zheng$^{1}$,             L.~Zhou$^{1}$,                 N.~F.~Zhou$^{1}$,
80: K.~J.~Zhu$^{1}$,               Q.~M.~Zhu$^{1}$,               Y.~C.~Zhu$^{1}$,
81: Y.~S.~Zhu$^{1}$,               Yingchun~Zhu$^{1}$$^{f}$,      Z.~A.~Zhu$^{1}$,
82: B.~A.~Zhuang$^{1}$,            X.~A.~Zhuang$^{1}$,            B.~S.~Zou$^{1}$
83: \\(BES Collaboration)\\
84: \vspace{0.2cm}
85: $^{1}$ {\it Institute of High Energy Physics, Beijing 100049, People's Republic of China}\\
86: $^{2}$ {\it China Center for Advanced Science and Technology(CCAST), Beijing 100080, People's Republic of China}\\
87: $^{3}$ {\it Guangxi Normal University, Guilin 541004, People's Republic of China}\\
88: $^{4}$ {\it Guangxi University, Nanning 530004, People's Republic of China}\\
89: $^{5}$ {\it Henan Normal University, Xinxiang 453002, People's Republic of China}\\
90: $^{6}$ {\it Huazhong Normal University, Wuhan 430079, People's Republic of China}\\
91: $^{7}$ {\it Hunan University, Changsha 410082, People's Republic of China}\\
92: $^{8}$ {\it Liaoning University, Shenyang 110036, People's Republic of China}\\
93: $^{9}$ {\it Nanjing Normal University, Nanjing 210097, People's Republic of China}\\
94: $^{10}$ {\it Nankai University, Tianjin 300071, People's Republic of China}\\
95: $^{11}$ {\it Peking University, Beijing 100871, People's Republic of China}\\
96: $^{12}$ {\it Shandong University, Jinan 250100, People's Republic of China}\\
97: $^{13}$ {\it Sichuan University, Chengdu 610064, People's Republic of China}\\
98: $^{14}$ {\it Tsinghua University, Beijing 100084, People's Republic of China}\\
99: $^{15}$ {\it University of Hawaii, Honolulu, HI 96822, USA}\\
100: $^{16}$ {\it University of Science and Technology of China, Hefei 230026, People's Republic of China}\\
101: $^{17}$ {\it Wuhan University, Wuhan 430072, People's Republic of China}\\
102: $^{18}$ {\it Zhejiang University, Hangzhou 310028, People's Republic of China}\\
103: \vspace{0.4cm}
104: $^{a}$ Current address: Iowa State University, Ames, IA 50011-3160, USA\\
105: $^{b}$ Current address: Purdue University, West Lafayette, IN 47907, USA\\
106: $^{c}$ Current address: Cornell University, Ithaca, NY 14853, USA\\
107: $^{d}$ Current address: Laboratoire de l'Acc{\'e}l{\'e}ratear Lin{\'e}aire, Orsay, F-91898, France\\
108: $^{e}$ Current address: University of Michigan, Ann Arbor, MI 48109, USA\\
109: $^{f}$ Current address: DESY, D-22607, Hamburg, Germany\\
110: }
111: %%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
112: \begin{abstract}
113:  The decay modes
114:  $\jpsi\ar\gamma\pi^0, \gamma\eta$ and $\gamma\etap$ are analyzed using 
115:  a data sample of 58 million $\jpsi$ decays 
116:  collected with the BESII detector at  BEPC. The branching fractions are
117:  determined to be: 
118:  $Br(\jpsi\ar\gamma\pi^0)=(3.13^{+0.65}_{-0.44})\times10^{-5}$, 
119:  $Br(\jpsi\ar\gamma\eta)=(11.23\pm0.89)\times10^{-4}$, and 
120:  $Br(\jpsi\ar\gamma\etap)=(5.55\pm0.44)\times10^{-3}$, where the errors are combined 
121:  statistical and systematic errors. The ratio of partial widths 
122:  $\Gamma(\jpsi\ar\gamma\etap)/\Gamma(\jpsi\ar\gamma\eta)$ is measured to be $4.94\pm0.40$, and the 
123:  singlet-octet pseudoscalar mixing angle of $\eta-\etap$ system is determined to be 
124:  $\theta_{P}=(-22.08\pm0.81)^{\circ}$.
125: 
126: \end{abstract}
127: \pacs{13.25.Gv, 12.38.Qk, 14.40.Cs }
128: \maketitle
129: 
130: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
131: 
132: \vspace{5mm}
133: \section{INTRODUCTION}
134: 
135:  In flavor-$SU(3)$, the $\pi^0$, $\eta$ and $\etap$ mesons 
136:  belong to the same pseudoscalar nonet. 
137:  The physical states $\eta$ and $\etap$ are related to the $SU_f(3)$-octet state $\eta_8$ and the 
138:  $SU_f(3)$-singlet state $\eta_1$, via the usual mixing formulae:
139: $$
140: \eta=\eta_8\cos\theta_P-\eta_1\sin\theta_P,
141: $$
142: $$
143: \etap=\eta_8\sin\theta_P+\eta_1\cos\theta_P,
144: $$
145: where $\theta_P$ is the pseudoscalar mixing angle~\cite{theo,mixing}. 
146: The conventional estimate of $\eta - \etap$ mixing uses the quadratic mass matrix
147: \[
148: M^2 = \left(\begin{array}{cc}  M^2_{88} & M^2_{18} \\ M^2_{18} & M^2_{11}\end{array}\right),
149: \]
150: where  $M^2_{88}=\frac{1}{3}(4m^2_K-m^2_{\pi})$ is given by the Gell-Mann-Okubo mass formula. 
151: Diagonalization of this matrix gives
152: $$
153: \tan^2 \theta_P = \frac{M^2_{88}-m^2_{\eta}}{m^2_{\etap}-M^2_{88}} \Longrightarrow \theta_P \approx -10^{\circ}.
154: $$
155: With a linear mass matrix and the linear 
156: Gell-Mann-Okubo mass formula  $M_{88}=\frac{1}{3}(4m_K-m_{\pi})$, 
157: $\theta_P$ is computed to be about $-24^{\circ}$~\cite{mixing}. 
158: 
159: The mixing angle has been measured experimentally in different ways, and the value is around 
160: $-20^{\circ}$~\cite{mixing}. One of these measurements is based on $\jpsi$ radiative decays. 
161: In the limit where the OZI rule and $SU_f(3)$ symmetry are exact, one gets~\cite{thetap1}
162: $$
163: R =
164:  \frac{\Gamma(\jpsi\ar\gamma\etap)}
165:  {\Gamma(\jpsi\ar\gamma\eta)}=(\frac{p_{\etap}}{p_{\eta}})^3\cdot{\cot^2\theta_P},
166: $$
167: where $p_{\eta}$ and $p_{\etap}$ are the momenta of $\eta$ and $\etap$
168: in the $\jpsi$ Center of Mass System (CMS). 
169: 
170: The first-order perturbation theory~\cite{pertu1,pertu2} 
171: expression for the partial width  $\Gamma(\jpsi\ar\gamma + 
172:  pseudoscalar)$ is 
173: $$
174: \Gamma(\jpsi\ar\gamma + P) = \frac{1}{6}(\frac{2}{3})^2\alpha^4_s \alpha Q^2_c\frac{1}{M^3_{\jpsi}}(\frac{4R_{\jpsi}(0)}{\sqrt{4\pi M_{\jpsi}}})^2 (\frac{4R_{P}(0)}{\sqrt{4\pi M_{P}}})^2 x |H^P(x)|^2.
175: $$
176: Here $R_{\jpsi}(0)$ and $R_{P}(0)$ are the wave functions at the origin 
177: of the $\jpsi$ and the pseudoscalar with mass $M_{P}$, and $Q_c$ is the charge of the charmed
178: quark. The pseudoscalar helicity amplitude $H^P(x)$ depends on $x=1-(\frac{M_P}{M_{\jpsi}})^2$; 
179: numerically $x|H^P(x)|\approx55$ for $M_P=m_{\etap}$. $R_{\jpsi}(0)$ and $R_{P}(0)$ can be 
180: determined from the $\jpsi\ar e^+ e^-$ and $P\ar\gamma\gamma$ partial decay widths, 
181: respectively. 
182: Using the lowest-order QCD formula for $\alpha_s$, the $\jpsi\ar\gamma\etap$ decay width is 
183: calculated to be 213~eV, which is in agreement with the experimentally measured value. The value of $\Gamma(\jpsi\ar\gamma\eta)$
184: determined from the same 
185: formula disagrees with measurements.
186: Some models that assign a small admixture of $\eta$ and $\etap$ to other states have 
187: been proposed to explain the large value 
188: of the ratio $R=\Gamma(\jpsi\ar\gamma\etap)/\Gamma(\jpsi\ar\gamma\eta)$. 
189: For example, Ref.~\cite{ratio1}, which  assigns small $c \bar{c}$ contribution from $\eta_c$ in the $\eta$ and $\etap$ wave functions, 
190: predicts $R=3.9$;  Ref.~\cite{ratio2} gives a value of $R=5.1$ by 
191: considering some admixture of the $\iota(1440)$ to the $\eta$ and $\etap$. A precision 
192: measurement of $R$ could distinguish between these mixing models, as well as provide
193: a determination of the mixing angle 
194: $\theta_P$.  Experimental measurements of $Br(\jpsi\ar\gamma\eta)$ and $Br(\jpsi\ar\gamma\etap)$ 
195: were reported by the DESY-Heidelberg group~\cite{DESY}, the Crystal Ball~\cite{CBAL}, 
196: MarkIII~\cite{MARK3} and DM2~\cite{DM2}. 
197: 
198: The decay $\jpsi\ar\gamma\pi^0$ is suppressed because the photon can only be radiated from the 
199: final state quarks.  This branching fraction was measured by DASP~\cite{DASP} and Crystal 
200: Ball~\cite{CBAL}; the average of the measurements, $(3.9\pm1.3) \times 10^{-5}$~\cite{PDG04}, 
201: is in  agreement with the VMD prediction $3.3\times 10^{-5}$~\cite{VMD}. In contrast, the QCD 
202: multipole 
203: expansion theory~\cite{mutipole} predicts a value of $1\times10^{-6}$. 
204: 
205: In this paper, $\jpsi\ar\gamma\pi^0$ is studied using $\pi^0\ar\gamma\gamma$ decay, 
206: $\jpsi\ar\gamma\eta$ is measured using $\eta\ar\gamma\gamma$ and $\eta\ar\pi^0\pi^+\pi^-$ with 
207: $\pi^0\ar\gamma\gamma$, and $\jpsi\ar\gamma\etap$ is studied using $\etap\ar\gamma\gamma$, 
208: $\etap\ar\gamma\pi^+\pi^-$ and $\etap\ar\eta\pi^+\pi^-$ with $\eta\ar\gamma\gamma$. The analyses 
209: use a data sample that contains $58\times10^6$ $\jpsi$ decays collected with the 
210: updated BEijing Spectrometer 
211: (BESII) operating at the Beijing Electron Positron Collider (BEPC). 
212: 
213: \section{BES DETECTOR AND MONTE CARLO SIMULATION}
214: 
215: BESII is a large solid-angle magnetic spectrometer that is described in detail in
216: Ref.~\cite{bes2}. The momentum of  charged particles is measured
217: in a 40-layer cylindrical main drift chamber (MDC) with a
218: momentum resolution of $\sigma_{p}$/p=$1.78\%\sqrt{1+p^2}$ ($p$ in
219: GeV/c).  Particle identification is accomplished using specific
220: ionization ($dE/dx$) measurements in the drift chamber and
221: time-of-flight (TOF) information from a barrel-like array of 48
222: scintillation counters. The $dE/dx$ resolution is
223: $\sigma_{dE/dx}\simeq8.0\%$; the TOF resolution for Bhabha events is
224: $\sigma_{TOF}= 180$ ps.  Radially outside of the time-of-flight
225: counters is a 12-radiation-length barrel shower counter (BSC)
226: comprised of gas tubes interleaved with lead sheets. The
227: BSC measures the energy and direction of photons with resolutions of
228: $\sigma_{E}/E\simeq21\%\sqrt{E}$ ($E$ in GeV), $\sigma_{\phi}=7.9$
229: mrad, and $\sigma_{z}=2.3$ cm. The iron flux-return of the magnet is
230: instrumented with three double layers of proportional counters
231: that are used to identify muons.
232: 
233: A GEANT3-based Monte Carlo simulation package~\cite{simbes}, which simulates the
234: detector response including interactions of secondary particles in
235: the detector material, is used to determine detection efficiencies
236: and mass resolutions,  optimize
237: selection criteria, and estimate backgrounds.  Reasonable agreement
238: between data and MC simulation is observed for various calibration
239: channels, including $e^+e^-\to(\gamma) e^+ e^-$,
240: $e^+e^-\to(\gamma)\mu^+\mu^-$, $J/\psi\to p\bar{p}$, $J/\psi\to\rho\pi$
241: and $\psi(2S)\to\pi^+\pi^- J/\psi$, $J/\psi\to l^+ l^-$.
242: 
243: %%%%%%%%%%%%%%%%%%%%%%%
244: \section{DATA ANALYSIS}
245: 
246: \subsection{\boldmath $\jpsi\ar\gamma\gamma\gamma$}
247: In the $\jpsi\ar\gamma\gamma\gamma$ decay mode, there are no charged tracks in the final states.
248: Each candidate event is required to have three and only three photon candidates; the MC 
249: indicates that the number of these decays that produce final states with
250: more than three photon candidates is negligible.
251:  A photon candidate is defined as
252: a cluster in the BSC with an energy deposit of more than 50 MeV, and with an angle 
253: between the development direction of the cluster and the direction 
254: from the interaction point to the first hit layer of the BSC that is 
255:  less than 20$^{\circ}$. If two clusters have an opening angle 
256: that is less than 10$^{\circ}$ or have an 
257:  invariant mass that is less than 50 MeV$/c^2$, the lower energy cluster is regarded as a 
258: remnant  from the other and not a separate
259: photon candidate. A kinematic fit that conserves energy and momentum  is 
260:  applied to the three photon candidates, and $\chi^2\leq20$ is required. We also
261: require $|\cos\theta_v|<0.8$ and   $\theta_{min}>6^{\circ}$, 
262: where $\theta_v$ is the polar angle of a decay photon in the pseudoscalar's CMS 
263:  (shown in Fig. 1a), and $\theta_{min}$ is the minimum angle between any two of the three photon 
264:  candidates (shown in Fig. 1b).  This rejects background from the
265: continuum  $e^+e^-\ar\gamma\gamma(\gamma)$ process.
266: 
267: 
268: \begin{figure}[htbp]
269: \centerline{\psfig{file=1a.epsi,width=6cm,height=5cm}
270:             \psfig{file=1b.epsi,width=6cm,height=5cm}}
271: \parbox[top]{8cm}{(a)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~(b)}
272: \caption{Distribution of (a) $\cos\theta_v$ and (b) $\theta_{min}$. The open histograms are 
273: $\jpsi$ data, the shaded histograms are background from $e^+e^-\ar\gamma\gamma(\gamma)$, and the 
274: dashed lines are simulated $\jpsi\ar\gamma\pi^0\ar\gamma\gamma\gamma$ events (not normalized).}
275: \end{figure}
276: 
277: \subsubsection{$\jpsi\ar\gamma\pi^0,\pi^0\ar\gamma\gamma$}
278:  
279: Figure 2 shows the invariant mass distribution in the $\pi^0$ mass region
280: of the two photon candidates that have the smallest opening 
281: angle.  A peak at the $\pi^0$  mass is evident.
282: 
283: \begin{figure}[htbp]
284: \centerline{\psfig{file=2.epsi,width=6cm,height=5cm}}
285: \caption{Invariant mass distribution of the $\gamma\gamma$ with smallest opening angle for 
286: $\jpsi\ar\gamma\gamma\gamma$ candidate events. The solid squares with error bars are data, the 
287: histogram is the best fit described in the text, and the dashed line is the background.}
288: \end{figure}
289: 
290: From  MC studies, background channels that produce a peak in the 
291: $\pi^0$ signal region come mainly from channels with $5\gamma$  
292: final states, such as $\jpsi\ar\gamma\pi^0\pi^0$, via the $f_2(1270)$, $f_0(2100)$ etc. 
293: ($\jpsi\ar 4 \gamma$s violates C-parity).  These background sources are
294: studied using events where the number of photon candidates in the event is four. 
295: Four photon events are  selected and subjected to a four-constraint kinematic fit to 
296: $\jpsi\ar\gamma\gamma\gamma$,  using any three of the four photons;  the three-photon 
297: combination with the smallest $\chi^2$ is selected for the background study. Figure~3 
298: shows the  invariant mass 
299: distribution for the two photons with the smallest opening angle from four-photon 
300: events.  A peak is observed in the $\pi^0$ mass region that agrees with 
301: expectations from MC simulations that include all known modes that produce 
302: $5\gamma$ final states. 
303: However, since the known background channels do not account for the level of the observed
304: background in the data sample, a scale factor is introduced to scale the MC 
305: background predictions for fits to the distribution in Fig.~2. The 
306: scale factor depends strongly on which intermediate states are considered 
307: for $\jpsi\ar 5 \gamma$ decays; the 
308: difference between the scale factors determined from different channels is treated as  
309: a systematic uncertainty of the background subtraction.
310: 
311: Figure 2 is fit with a MC-simulated $\jpsi\ar\gamma\pi^0$ histogram for the signal, a MC-simulated 
312: $\jpsi\ar 5 \gamma$ background shape, as well as a shape of MC simulated phase space for other 
313: sources of backgrounds. The number of $\gamma\pi^0$ events determined from the fit is $586\pm51$. 
314: The MC-determined detection efficiency for
315: $\jpsi\ar\gamma\pi^0,\pi^0\ar\gamma\gamma$ is 
316: $\varepsilon = (32.80\pm0.21)$\%, where the error comes from the limited 
317: statistics of the MC sample.
318: 
319: \begin{figure}[htbp]
320: \centerline{\psfig{file=3.epsi,width=6cm,height=5cm}}
321: \caption{The invariant mass distribution of $\gamma\gamma$ pairs with the smallest opening 
322: angle in  $\jpsi\ar\gamma\gamma\gamma$ events selected from the four photon event sample.}
323: \end{figure}
324: 
325: 
326: \subsubsection{$\jpsi\ar\gamma\eta,\eta\ar\gamma\gamma$}
327: Figure 4 shows the invariant mass distribution of the two photon candidates with
328: the  smallest opening  angle in the $\eta$ mass region, where an $\eta$ peak is evident.
329: 
330: \begin{figure}[htbp]
331: \centerline{\psfig{file=4.epsi,width=6cm,height=5cm}}
332: \caption{Invariant mass distribution of the $\gamma\gamma$ with the smallest opening 
333: angle of $\jpsi\ar\gamma\gamma\gamma$ candidates. Solid squares with error bars are 
334: data, the histogram is the fit result, and the dashed line is the background.}
335: \end{figure}
336: 
337: 
338: The $\gamma\gamma$ invariant mass distribution of Fig.~4 is fit with a 
339: histogram from MC-simulated
340: $\jpsi\ar\gamma\eta, \eta\ar\gamma\gamma$ events and a second order Legendre polynomial background
341: function.  The  fit yields a signal of 9096$\pm$133 $\eta$s.   The 
342: MC-determined detection efficiency is $\varepsilon=(36.33\pm0.22)\%$.
343: 
344: 
345: 
346: \subsubsection{$\jpsi\ar\gamma\etap,\etap\ar\gamma\gamma$}
347: 
348: Since the momentum of the $\etap$ is lower than that of the
349: $\pi^0$ and $\eta$ in $\jpsi$ radiative decays, the angle 
350: between the two $\etap$ decay photons is not small 
351: enough to be useful for distinguishing them from the radiative 
352: photon. For this channel, the mass distribution of  
353: the three $\gamma\gamma$ combinations for each event 
354: are plotted in Fig. 5, where an $\etap$ 
355: signal is evident above  
356: a smooth background due to wrong $\gamma\gamma$ combinations 
357: plus other background sources.
358: 
359: \begin{figure}[htbp]
360: \centerline{\psfig{file=5.epsi,width=6cm,height=5cm}}
361: \caption{The $\gamma\gamma$ invariant mass distribution
362: for $\jpsi\ar\gamma\gamma\gamma$ candidate events (three entries per 
363: event).  The solid 
364: squares with error bars indicate data, the  histogram is the fit result, 
365: and the dashed line is the non-combinatorial background.}
366: \end{figure}
367: 
368: A fit to the data points, with the MC simulated mass distribution for the 
369: $\jpsi\ar\gamma\etap,\etap\ar\gamma\gamma$ decay including combinatorial 
370: background for the signal and a second order Legendre polynomial for background between 
371: 0.8 and 1.2 $\hbox{GeV}/c^2$, yields $2982\pm101$ entries. Since all the 
372: $\gamma\gamma$ combinations are plotted in the $M_{\gamma\gamma}$ distribution, 
373: the combinatorial background is included in the entries for 
374: both data and MC simulation. The efficiency for signal 
375: $\jpsi\ar\gamma\etap,\etap\ar\gamma\gamma$ entries is
376: $(40.30\pm0.22)\%$. The combinatorial background is about $20\%$ for both data and MC simulation, 
377: and they cancel out when the $N^{obs}$ is divided by the efficiency
378: $40.30\%$ in the branching fraction calculation.
379: 
380: \subsection{\boldmath $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$}
381: 
382: In this final states, there are two charged particles $\pi^+$ and $\pi^-$ 
383: and three photons. Candidate events are required to satisfy the following 
384: common selection criteria:
385: 
386: \begin{enumerate}
387:   \item Two good charged tracks with net charge zero. Each
388: track must have a good helix fit, a transverse momentum larger than 60 MeV/c, and 
389: $|\cos\theta|<0.8$, where $\theta$ is the polar angle of the track,
390: and  must originate from the interaction
391: region.
392:   \item At least one charged track is identified as a $\pi$, 
393: satisfying $\chi^2_{PID}(\pi)<\chi^2_{PID}(K)$ and 
394: $\chi^2_{PID}(\pi)<\chi^2_{PID}(p)$, 
395: where $\chi^2_{PID}=\chi^2_{dE/dx}+\chi^2_{TOF}$ is 
396: determined using both $dE/dx$ and TOF information.
397:   \item At least three photon candidates are required. The 
398: photon identification is similar to that used in the 
399: $\jpsi\ar\gamma\gamma\gamma$ analysis, except that the angle between a 
400: cluster and any other cluster must be greater than $18^{\circ}$, and the 
401: angle  between the cluster and any charged track must be greater 
402: than  $8^{\circ}$. These differences  reflect
403: different sources of fake photons.
404:   \item A four-constraint kinematic fit is applied to all three-photon 
405: combinations plus the two charged tracks assuming 
406: $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$. The three-photon combination with 
407: the smallest $\chi^2$ is selected, and 
408: the $\chi^2$ of the kinematic fit is required to be less than 20. 
409: \end{enumerate}
410: 
411: The events that survive these selection criteria with an invariant mass in the range
412: $M_{\gamma\gamma\pi^+\pi^-}\leq1.2\hbox{GeV}/c^2$ are assumed to
413: come from either $\eta$ or $\etap$ decays, 
414: and the other photon is 
415: considered to be the radiative 
416: photon. Figure 6 shows a
417: scatterplot of $M_{\gamma\gamma}$ versus $M_{\gamma\gamma\pi^+\pi^-}$ for
418: the selected events.  Clear $\eta$ and $\etap$ signals corresponding to 
419: $\eta\ar\pi^0\pi^+\pi^-,\pi^0\ar\gamma\gamma$, and 
420: $\etap\ar\eta\pi^+\pi^-,\eta\ar\gamma\gamma$ are observed.
421: 
422: 
423: \begin{figure}[htbp]
424: \centerline{\psfig{file=6.epsi,width=6cm,height=5cm}}
425: \caption{Scatterplot of $M_{\gamma\gamma}$ versus 
426: $M_{\gamma\gamma\pi^+\pi^-}$ for the 
427: $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$ candidates.}
428: \end{figure}
429: 
430: \subsubsection{$\jpsi\ar\gamma\eta,\eta\ar\pi^0\pppm$}
431: After the requirement that the $\gamma\gamma$ invariant mass is in the
432: $\pi^0$ mass
433: region ($M_{\gamma\gamma}\in[0.088,0.182]~\hbox{GeV}/c^2$, $\pm
434: 3 \sigma$), a clear $\eta$ signal is 
435: evident in the $\gamma\gamma\pi^+\pi^-$ invariant mass distribution shown 
436: in Fig. 7.
437: 
438: \begin{figure}[htbp]
439: \centerline{\psfig{file=7.epsi,width=6cm,height=5cm}}
440: \caption{The $\gamma\gamma\pi^+\pi^-$ invariant mass distribution 
441: for $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$ candidates that satisfy
442: the  requirement $M_{\gamma\gamma}\in[0.088,0.182]~\hbox{GeV}/c^2$. 
443: The solid squares with 
444: error bars indicate the data, the histogram is the fit result, and the 
445: dashed line is the background.}
446: \end{figure}
447: 
448: The simulated $M_{\gamma\gamma\pi^+\pi^-}$ mass distribution from the signal MC and a second-order Legendre polynomial are used to fit the 
449: $\gamma\gamma\pi^+\pi^-$ invariant mass distribution.
450: The fit gives $1885\pm58$ $\eta$ events.  The MC-determined
451: detection  efficiency for
452: $\jpsi\ar\gamma\eta,\eta\ar\pi^0\pi^+\pi^-$, and $\pi^0\ar\gamma\gamma$
453: is $\varepsilon=(12.25\pm0.15)\%$.
454: 
455: \subsubsection{$\jpsi\ar\gamma\etap,\etap\ar\eta\pppm$}
456: 
457: 
458: The $\gamma\gamma\pi^+\pi^-$ invariant mass distribution
459: for events with $\gamma\gamma$ mass within $3\sigma$ of
460: the $\eta$ mass ($M_{\gamma\gamma}\in[0.484,0.612]~\hbox{GeV}/c^2$),
461: is shown in Fig. 8.
462: 
463: 
464: 
465: \begin{figure}[htbp]
466: \centerline{\psfig{file=8.epsi,width=6cm,height=5cm}}
467: \caption{ The $\gamma\gamma\pi^+\pi^-$ invariant mass distribution
468: for events with $\gamma\gamma$ mass in the $\eta$ mass region
469: ($M_{\gamma\gamma}\in[0.484,0.612]~\hbox{GeV}/c^2$). The solid squares 
470: with error bars indicate data, the histogram is the fit result, and the 
471: dashed  line is the background.}
472: \end{figure}
473: 
474: A similar fit as for $\eta\ar\pi^0\pi^+\pi^-$ yields $8572\pm131$ $\etap$ 
475: events; the MC-determined detection efficiency for
476: $\jpsi\ar\gamma\etap,\etap\ar\eta\pi^+\pi^-$, and $\eta\ar\gamma\gamma$ 
477: is $\varepsilon=(16.10\pm0.12)\%$.
478: 
479: 
480: 
481: \subsection{\boldmath $\jpsi\ar\gamma\gamma\pi^+\pi^-$}
482: 
483: $\jpsi\ar\gamma\etap$ is also studied using the $\etap\ar\gamma\pi^+\pi^-$ 
484: decay channel.  For this study, the $\pi^{\pm}$ and photon selection
485: requirements are the same as used for the 
486: $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$ final state, and the event 
487: selection is similar, except that here at least two photons are 
488: required in the event. 
489: The photons and charged tracks are kinematically fitted to 
490: $\jpsi\ar\gamma\gamma\pi^+\pi^-$ assuming four-momentum conservation, and 
491: $\chi^2\leq20$ is required.  When there are more than two 
492: photons, the kinematic fit is repeated using all possible photon 
493: combinations, and the one with the smallest $\chi^2$ is kept. The photon with 
494: the higher energy is considered to be 
495: the radiative photon from the $\jpsi$ decay. Figure 9 shows the 
496: invariant mass distribution of $\gamma\pi^+\pi^-$ for the candidate 
497: events where an $\etap$ signal is evident.
498: 
499: \begin{figure}[htbp]
500: \centerline{\psfig{file=9.epsi,width=6cm,height=5cm}}
501: \caption{The $\gamma\pi^+\pi^-$ invariant mass distribution for
502: selected $\jpsi\ar\gamma\gamma\pi^+\pi^-$ events.  The
503: solid squares with error bars indicate data, the histogram is the fit 
504: result, and the dashed line is the background.}
505: \end{figure}
506: 
507: Figure 9 shows the result of a fit to the $\gamma\pi^+\pi^-$ invariant 
508: mass  distribution that follows a similar procedure as that for the 
509: fit to the $\gamma\gamma\pi^+\pi^-$ distribution of the previous section.
510: The fit yields  $23243\pm229$ $\etap$ signal events.  
511: The MC-determined detection efficiency for 
512: $\jpsi\ar\gamma\etap,\etap\ar\gamma\pi^+\pi^-$ 
513: is $\varepsilon=(25.02\pm0.10)\%$.
514: 
515: %%%%%%%%%%%%%%%%%%%%%%%%
516: \section{SYSTEMATIC ERRORS}
517: Systematic errors in the branching fraction
518: measurements mainly originate 
519: from photon identification (ID), MDC tracking efficiency, particle ID, 
520: kinematic fitting, mass resolution, $\pi^0$ reconstruction, 
521: and parameterizations of background 
522: shapes.
523: 
524: \subsection{Photon identification}
525: 
526: The efficiency for photon ID is discussed in Ref.~\cite{lism}. It 
527: is found that the
528: relative  efficiency difference between data and MC simulation
529: for high energy photon detection is about 0.8\% per photon, 
530: while for low energy photons, the difference is around 
531: 2\% per photon. Since the energy of the radiative photon 
532: in  $\jpsi\ar\gamma\pi^0,\gamma\eta$, and $\gamma\etap$ is high, 
533: and the energies of the photons from pseudoscalar particle decays are low, 
534: the total systematic error due to photon ID is taken as $(0.8+2.0n)\%$,
535: where $n$ is the number of photons from the 
536: pseudoscalar particle decay.
537: 
538: \subsection{MDC tracking}
539: The MDC tracking efficiency is studied in Ref.~\cite{simbes}. It is found 
540: that there is a 2.0\% relative difference per track between data and MC 
541: simulation. For the channels in this analysis that have two charged 
542: tracks, a 4\% systematic error on the MDC tracking efficiency is
543: assigned.
544: 
545: \subsection{Particle ID}
546: A clean charged $\pi$ sample obtained from $\jpsi\ar\rho\pi$ without 
547: the use of particle ID is used to study data-MC differences between
548: particle ID efficiencies for
549: different momentum ranges. Since only one 
550: of the two charged tracks is required to be identified as a pion, the MC 
551: simulates data rather well; 
552: it is found that the MC simulation agrees 
553: with data within 0.2\% for both $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$ 
554: and $\jpsi\ar\gamma\gamma\pi^+\pi^-$ modes.
555: 
556: \subsection{Kinematic fit}
557: 
558: Samples of $\jpsi\ar\rho\pi$ and $e^+e^-\ar\gamma\gamma$ 
559: events selected without using kinematic fits
560: are used to study the systematic error associated with the
561: four-constraint kinematic fit.  For the  $\chi^2\leq20$ criteria, 
562: the difference of kinematic fit efficiencies between data and MC
563: simulation is less than 1.2\% for $\rho\pi$, and
564: 2.4\% for $e^+e^-\ar\gamma\gamma$.  Extrapolating these differences to the
565: channels reported here, we conservatively assign a 4\%  systematic 
566: error to the kinematic fit efficiency.
567: 
568: \subsection{Different mass resolution between MC and DATA}
569: 
570: There is a slight difference of the mass resolution between MC
571: simulation and data. When the histogram shape of invariant mass
572: distribution from MC simulation is used to fit the invariant mass
573: distribution of data, it introduces some systematic error.
574: The high statistics decay channels
575: $\jpsi\ar\gamma\eta,\eta\ar\gamma\gamma$,
576: $\jpsi\ar\gamma\etap,\etap\ar\gamma\gamma\pi^+\pi^-$ and
577: $\jpsi\ar\gamma\etap,\etap\ar\gamma\pi^+\pi^-$ are used to study this
578: source of systematic error.  For these channels,
579: we allow the mass resolution to vary in the fit to the
580: invariant mass distributions, and we also determine the number of 
581: signal events by subtracting side-band-estimated backgrounds.
582: The resulting
583: branching fractions change by at most 1.6\%, 0.1\%, and 0.6\% for
584: $\jpsi\ar\gamma\eta,\eta\ar\gamma\gamma$,
585: $\jpsi\ar\gamma\etap,\etap\ar\eta\pi^+\pi^-$, and
586: $\jpsi\ar\gamma\etap,\etap\ar\gamma\pi^+\pi^-$ respectively. 
587: We assign 1.6\%,
588: 0.1\% and 0.6\%  as the systematic errors due to 
589: mass resolution uncertainties for the $\jpsi\ar\gamma\gamma\gamma$,
590: $\jpsi\ar\gamma\gamma\gamma\pi^+\pi^-$ and
591: $\jpsi\ar\gamma\gamma\pi^+\pi^-$ decay modes, respectively.
592: 
593: \subsection{\boldmath Reconstruction of $\pi^0$}
594: In $\jpsi\ar\gamma\pi^0$, the $\pi^0$ momentum is high and 
595: the angle between the two decay photons is small. As a result, it
596: is possible for the two  photons to merge into a single BSC cluster. 
597: According to a study reported in 
598: Ref.~\cite{xinbo}, the systematic error associated with 
599: 1.5~GeV $\pi^0$ reconstruction is 0.83\%. 
600: The effect on low energy $\pi^0$s 
601: or $\eta$s is small enough to be neglected.
602: 
603: \subsection{Background shape}
604: For the $\jpsi\ar\gamma\pi^0$ mode, the background estimate based on the
605: four photon event sample has a large uncertainty.  Fits using
606: MC-determined background shapes from different background channels
607: yield different numbers of signal events; 
608: the corresponding changes in the
609: branching fractions range between $^{+16.4}_{-6.8}\%$. 
610: The largest difference
611: is taken as the systematic error.  
612: Different order Legendre polynomials
613: are used to fit the mass spectra for the other decay modes, and the
614: differences between these fits and those used to get the numbers of
615: signal events are used as the systematic error due to background 
616: parameterization. Different fitting ranges are also used in the fit, and 
617: the differences are included in the systematic error. The uncertainty due 
618: to the background shape and fitting range is less than 2\%.
619: 
620: \subsection{Branching fractions of the secondary decays}
621: The branching fractions of decay from $\pi^{0},\eta$ and $\eta^{\prime}$ 
622: are taken from the PDG~\cite{PDG04}; the uncertainties are included in 
623: the measurement errors of the reported branching fractions.
624: 
625: \subsection{\boldmath The number of $J/\psi$ events}
626: The total number of $J/\psi$ events, determined from the 4-prong data sample,
627: is $(57.7\pm2.72)\times10^{6}$.
628:  The 4.72\% relative error is taken 
629: as a systematic error~\cite{fangss}.
630: 
631: \subsection{Total systematic error}
632: 
633: Table I summarizes the systematic errors from all sources for each 
634: mode.  We assume all the sources are independent and
635: add them in quadrature; the resulting total systematic errors are 
636: $^{+18.3}_{-10.6}\%$, $8.1\%$, $10.6\%$, $9.3\%$, $9.5\%$, and $8.7\%$ 
637: for $\jpsi\ar\gamma\pi^0\ar\gamma\gamma\gamma$, 
638: $\jpsi\ar\gamma\eta\ar\gamma\gamma\gamma$, 
639: $\jpsi\ar\gamma\etap\ar\gamma\gamma\gamma$, 
640: $\jpsi\ar\gamma\eta\ar\gamma\gamma\gamma\pi^+\pi^-$, 
641: $\jpsi\ar\gamma\etap\ar\gamma\gamma\gamma\pi^+\pi^-$, and 
642: $\jpsi\ar\gamma\etap\ar\gamma\gamma\pi^+\pi^-$, respectively.
643: 
644: \begin{table}[htbp]
645: \begin{center}
646: \caption{Summary of the systematic errors (\%). } \vspace{3mm}
647:   \begin{tabular}{lcccccc} \hline
648: Sources &$\pi^{0}\ar\gamma\gamma$ &$\eta\ar\gamma\gamma$&$\eta^{\prime}\ar\gamma\gamma$ & $\eta\ar\pi^{0}\pi^{+}\pi^{-}$&$\eta^{\prime}\ar\eta\pi^{+}\pi^{-}$&$\eta^{\prime}\ar\gamma\pi^+\pi^-$ \\
649: \hline
650: Photon ID & 4.8&4.8&4.8&4.8&4.8&2.8\\
651: %\hline
652: Tracking  & - &-&-&4.0&4.0&4.0\\
653: %\hline
654: Particle ID & -&-&-&0.2&0.2&0.2\\
655: %\hline
656: Kinematic fit & 4.0&4.0&4.0&4.0&4.0&4.0\\
657: %\hline
658: Mass resolution & 1.6&1.6&1.6&0.1&0.1&0.6\\ 
659: %\hline
660: $\pi^0$ reconstruction & 0.83 &-&- &- &- &- \\
661: %\hline
662: Background shape & $^{+16.4}_{-6.8}$ &0.73 &1.8 &1.7 &0.5 &0.2 \\
663: %\hline
664: Branching fraction used & 0.04&0.66&6.61&1.77&3.45&3.39\\
665: %\hline
666: Number of $J/\psi$ & 4.72&4.72&4.72&4.72&4.72&4.72\\
667: %\hline
668: Statistic of MC sample & 0.64 & 0.61& 0.55& 1.23& 0.75& 0.40 \\
669: \hline
670: Total error & $^{+18.3}_{-10.6}$ & 8.1 & 10.6 & 9.3 & 9.5 & 8.7 \\
671: \hline
672: \end{tabular}
673: 
674: \end{center}
675: \end{table}
676: 
677: %%%%%%%%%%%%%%%%%%%%%%%%
678: \section{RESULTS AND DISCUSSION}
679: 
680: The branching fractions of $\jpsi$ decays are determined from the relation
681: $$
682: Br(\jpsi\ar\gamma P)=\frac{N^{obs}(\jpsi\ar\gamma P \ar \gamma 
683: Y)}{N^{\jpsi}\cdot Br(P \ar Y) \cdot \varepsilon(\jpsi\ar\gamma P 
684: \ar\gamma Y)},
685: $$
686: where $P$ is either $\pi^0$, $\eta$, or $\etap$, $Y$ is the pseudoscalar 
687: decay final state, and $Br(P \ar Y)$ is the branching fraction of the 
688: pseudoscalar decays into final state $Y$. The results of $Br(\jpsi\ar\gamma P)$ are listed in Table~II.
689: 
690: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
691: 
692: The branching fractions of 
693: $\jpsi\ar\gamma\eta$, $\jpsi\ar\gamma\etap$ measured from different decay 
694: modes are consistent with each other within the statistical 
695: fluctuations and uncommon systematic errors.  The measurements from 
696: the different modes are, therefore, combined using a standard weighted 
697: least-squares procedure taking into consideration the correlations 
698: between the 
699: measurements; the mean value and the error are calculated by:
700: $$
701: \bar{x}\pm\delta\bar{x}=\frac{\sum_{j}{x_{j}}\cdot(\sum_{i}{\omega_{ij}})}{\sum_{i}\sum_{j}{\omega_{ij}}}\pm\sqrt{\frac{1}{\sum_{i}\sum_{j}{\omega_{ij}}}}.
702: $$
703: Here $\omega_{ij}$ is the element of the 
704: weighted matrix $W=V^{-1}_x$, where $V_x$ is the covariance matrix calculated according 
705: to the systematic errors listed in Table I. For $\jpsi\ar\gamma\eta$, the correlation 
706: coefficient between $\eta\ar\gamma\gamma$ and $\eta\ar\gamma\gamma\pi^+\pi^-$ 
707: is $\rho(1,2)=0.553$; for $\jpsi\ar\gamma\etap$, the correlation coefficients 
708: between $\etap\ar\gamma\gamma$, $\etap\ar\gamma\pi^+\pi^-$ 
709: and $\etap\ar\gamma\gamma\pi^+\pi^-$ are $\rho(1,2)=0.296$, $\rho(1,3)=0.404$ 
710: and $\rho(2,3)=0.703$. The weighted averages of BESII measurements and the 
711: PDG~\cite{PDG04} values are listed in Table II.
712: 
713: \begin{table}
714: \caption{Branching fractions of
715: $J/\psi\rightarrow\gamma\pi^{0}, \gamma\eta$ and $\gamma\eta^{\prime}$.}
716: \begin{center}
717: \footnotesize
718:   \begin{tabular}{|c|l|r|r|c|} 
719: \hline
720:  \multicolumn{2}{|c|}{Decay mode}   &   BESII    & BESII combined & PDG~\cite{PDG04}  \\
721: \hline
722:  $\gamma\pi^0$ & $\pi^{0}\rightarrow\gamma\gamma$ & {\footnotesize$(3.13\pm0.28^{+0.58}_{-0.34})\times10^{-5}$ }& $(3.13^{+0.65}_{-0.44})\times10^{-5}$ & $(3.9\pm1.3)\times10^{-5}$\\
723: \hline
724: $\gamma\eta$ &$\eta\rightarrow\gamma\gamma$  & {\footnotesize $(11.00\pm0.16\pm0.90)\times10^{-4}$}&$(11.23\pm0.89)\times10^{-4}$ &$(8.6\pm0.8)\times10^{-4}$\\
725: \cline{2-3}
726:  & $\eta\rightarrow\pi^{0}\pi^{+}\pi^{-}$  &{\footnotesize$(11.94\pm0.37\pm1.11)\times10^{-4}$}& &\\
727: \hline
728:  &$\eta^{\prime}\rightarrow\gamma\gamma$  &{\footnotesize$(6.05\pm0.21\pm0.65)\times10^{-3}$}&&\\
729: \cline{2-3}
730:  $\gamma\eta^{\prime}$ & $\eta^{\prime}\rightarrow\gamma\rho$  &{\footnotesize$(5.46\pm0.06\pm0.48)\times10^{-3}$} & $(5.55\pm0.44)\times10^{-3}$ &$(4.31\pm0.30)\times10^{-3}$\\
731: \cline{2-3}
732:  &$\eta^{\prime}\rightarrow\eta\pi^{+}\pi^{-}$  &{\footnotesize$(5.28\pm0.08\pm0.51)\times10^{-3}$}&&\\
733: \hline
734:   \end{tabular}
735: \end{center}
736: \end{table}
737: 
738: \begin{figure}[htbp]
739: \centerline{\psfig{file=10a.epsi,width=5.0cm,height=5.0cm}
740:             \psfig{file=10b.epsi,width=5.0cm,height=5.0cm}
741:             \psfig{file=10c.epsi,width=5.0cm,height=5.0cm}}
742: \caption{Comparisons of $Br(\jpsi\ar\gamma\pi^0)$,
743:   $Br(\jpsi\ar\gamma\eta)$ and $Br(\jpsi\ar\gamma\etap)$ between BESII
744:   and previous measurements~\cite{PDG04}. The shaded regions are the
745:   world averages from the PDG~\cite{PDG04}.}
746: \end{figure}
747: 
748: Figure 10 shows the comparisons between the measurements in this paper
749: and those from previous measurements~\cite{DESY,CBAL,MARK3,DM2,DASP}.
750: Our measurement of $Br(\jpsi\ar\gamma\pi^0)$ agrees with those of
751: Crystal Ball~\cite{CBAL} and DASP~\cite{DASP} within the large errors 
752: of the previous measurements, and has much improved precision. Our
753: measurement's lower central value may be because background channels
754: that produce a peak in the signal region have been considered. Our 
755: measurements of $Br(\jpsi\ar\gamma\eta)$ and $Br(\jpsi\ar\gamma\etap)$ are 
756: higher than the PDG world average~\cite{PDG04}, and have better precision
757: than the previous measurements~\cite{DESY,CBAL,MARK3,DM2}.
758: 
759: The results listed in Table II also allow us calculate the relative branching 
760: fractions for $\eta$ and $\etap$ decays; considering the common errors in the 
761: measurements, one gets
762: {\footnotesize
763: $$
764: \frac{Br(\etap\ar\gamma\gamma)}{Br(\etap\ar\gamma\pi^+\pi^-)}=\frac{Br(\jpsi\ar\gamma\etap,\etap\ar\gamma\gamma)}{Br(\jpsi\ar\gamma\etap,\etap\ar\gamma\pi^+\pi^-)}=0.080\pm0.008,
765: $$
766: $$
767: \frac{Br(\etap\ar\eta\pi^+\pi^-)}{Br(\etap\ar\gamma\pi^+\pi^-)}=\frac{Br(\jpsi\ar\gamma\etap,\etap\ar\eta\pi^+\pi^-)}{Br(\jpsi\ar\gamma\etap,\etap\ar\gamma\pi^+\pi^-)}=1.45\pm0.07,
768: $$
769: $$
770: \frac{Br(\eta\ar\gamma\gamma)}{Br(\eta\ar\pi^0\pi^+\pi^-)}=\frac{Br(\jpsi\ar\gamma\eta,\eta\ar\gamma\gamma)}{Br(\jpsi\ar\gamma\eta,\eta\ar\pi^0\pi^+\pi^-)}=1.61\pm0.14.
771: $$}
772: The correlation coefficients between denominator and numerator in above 
773: equations are 0.419, 0.859 and 0.575 respectively. 
774: The world averages~\cite{PDG04} of the same ratios are $0.072\pm0.006$, $1.50\pm0.08$ 
775: and $1.75\pm0.04$ respectively. The agreement is quite good.
776: 
777: If both the OZI rule and the $SU_f(3)$ symmetry are exact, it is expected that~\cite{thetap1}:
778: $$
779: R=\frac{\Gamma(\jpsi\ar\gamma\etap)}{\Gamma(\jpsi\ar\gamma\eta)}=(\frac{P_{\etap}}{P_{\eta}})^3\cdot{\cot^2\theta_P}.
780: $$
781: 
782: Using $Br(\jpsi\ar\gamma\eta)$ and $Br(\jpsi\ar\gamma\etap)$ in this analysis, one obtains
783: $$
784: R=4.94\pm0.40,
785: $$
786: $$
787: |\theta_P|=(22.08\pm0.81)^{\circ},
788: $$
789: where the common errors have been considered in the ratio calculation. 
790: Comparing with the mixing models with states other than $\eta$ and $\etap$, 
791: the measurement of $R$ agrees with the prediction of $R=5.1$~\cite{ratio2} within 
792: one standard deviation, while it deviates from $R=3.9$~\cite{ratio1} by more than 3 standard 
793: deviations. According to the theoretical calculation of Ref.~\cite{mixing}, the value of 
794: $\theta_P$ is negative, in which case its value is $\theta_P=(-22.08\pm0.81)^{\circ}$. 
795: 
796: %%%%%%%%%%%%%%%%%%%%%%%%%%
797: \section{SUMMARY}
798: Using 58 million $\jpsi$ events collected by BESII, 
799: the branching fractions of $\jpsi$ decays into a photon 
800: and a pseudoscalar meson are measured as 
801: $Br(\jpsi\ar\gamma\pi^0)=(3.13^{+0.65}_{-0.44})\times 10^{-5}$, 
802: $Br(\jpsi\ar\gamma\eta)=(11.23\pm 0.89)\times 10^{-4}$, and 
803: $Br(\jpsi\ar\gamma\etap)=(5.55\pm 0.44)\times 10^{-3}$. 
804: The results are compared to $\eta$ and $\etap$ mixing models.
805: 
806: %%%%%%%%%%%%%%%%%%%%%%%%
807: \section{Acknowledgment}
808: The BES collaboration thanks the staff of BEPC for their hard
809: efforts. This work is supported in part by the National Natural
810: Science Foundation of China under contracts Nos. 10491300,
811: 10225524, 10225525, 10425523, the Chinese Academy of Sciences under
812: contract No. KJ 95T-03, the 100 Talents Program of CAS under
813: Contract Nos. U-11, U-24, U-25, and the Knowledge Innovation
814: Project of CAS under Contract Nos. U-602, U-34 (IHEP), the
815: National Natural Science Foundation of China under Contract No.
816: 10225522 (Tsinghua University), and the Department of Energy under
817: Contract No.DE-FG02-04ER41291 (U Hawaii).
818: 
819: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
820: \begin{thebibliography}{dd}
821: % (1)
822: \bibitem{theo} J. F. Donoghue, B. R. Holstein and Y. -C. R. Lin, Phys.
823: Rev. Lett. {\bf 55}, 2766 (1985); R. Escribano and J. -M. Fr$\grave{e}$re, hep-ph/0501072 (2005).
824: % (2)
825: \bibitem{mixing} F. J. Gilman and R. Kauffman, Phys. Rev. {\bf D36}, 2761 (1987).
826: % (3)
827: \bibitem{thetap1} R. N. Cahn and M. S. Chanowitz, Phys. Lett. {\bf B59}, 277 (1975).
828: % (4)
829: \bibitem{pertu1} J. G. K\"{o}rner, J. H. K\"{u}hn, M. Krammer, H. Schneider, Nucl. Phys. {\bf B229}, 115 (1983).
830: % (5)
831: \bibitem{pertu2} B. Guberina and J. H. K\"{u}hn, Nuovo Cim. Lett. {\bf 32}, 295 (1981).
832: % (6)
833: \bibitem{PDG04}  S. Eidelman {\it et al.}(Particle Data Group), Phys. Lett. {\bf B592}, 1 (2004).
834: % (7)
835: \bibitem{ratio1} H. Fritsch and J. D. Jackson, Phys. Lett. {\bf B66}, 365 (1977).
836: % (8)
837: \bibitem{ratio2} H. Yu, High Energy Phys. \& Nucl. Phys. {\bf 12}, 754 (1988) (in Chinese).
838: % (9)
839: \bibitem{DESY} W. Bartel {\it et al.}, Phys. Lett. {\bf B66}, 489 (1977).
840: % (10)
841: \bibitem{CBAL} E. D. Bloom, C. Peck, ARNS {\bf 33}, 143 (1983).
842: % (11)
843: \bibitem{MARK3} T. Bolton {\it et al.}, Phys. Rev. Lett. {\bf 69}, 1328 (1992).
844: % (12)
845: \bibitem{DM2} J. E. Augustin {\it et al.}, Phys. Rev. {\bf D42}, 10 (1990).
846: % (13)
847: \bibitem{DASP} W. Braunschweig {\it et al.}, Phys. Lett. {\bf B67}, 243 (1977).
848: % (14)
849: \bibitem{VMD} V. L. Chernyak and A. R. Zhitnitsky, Phys. Rept. {\bf 112}, 173 (1984).
850: % (15)
851: \bibitem{mutipole} Y. P. Kuang, Phys. Rev. {\bf D42}, 2300 (1990).
852: % (16)
853: \bibitem{bes2} BES Collab., J. Z. Bai {\it et al.}, Nucl. Instrum. Methods {\bf A458}, 627 (2001).
854: % (17)
855: \bibitem{simbes} BES Collab., M. Ablikim {\it et al.}, physics/0503001, to be published in Nucl. Instrum. Methods A.
856: % (18)
857: \bibitem{lism} S. M. Li {\it et al.}, High Energy Phys. \& Nucl.Phys., {\bf 28}, 859 (2004) (Chinese Edition).
858: % (19)
859: \bibitem{xinbo} BES Collab., M. Ablikim {\em et al.}, Phys. Rev. {\bf
860: D71}, 072006 (2005).
861: % (20)
862: \bibitem{fangss} S. S. Fang {\it et al.},
863: High Energy Phys. \& Nucl. Phys. {\bf 27}, 277 (2003) (in Chinese).
864: 
865: \end{thebibliography}
866: 
867: \end{document}
868: 
869: 
870: 
871: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
872: