1: \documentclass[11pt]{article}
2: \usepackage{graphicx}
3: \usepackage{epsfig}
4: \usepackage{subfigure}
5: \newcommand{\BABARPubYear} {08}
6: \newcommand{\BABARConfNumber} {013}
7: \newcommand{\SLACPubNumber} {13353}
8:
9: \input babarsym
10: \def\KKKz {\ensuremath{K^+ K^- K^0}\xspace}
11: \def\KKKs {\ensuremath{K^+ K^- \KS}\xspace}
12: \def\KKKl {\ensuremath{K^+ K^- \KL}\xspace}
13: \def\KKKspm {\ensuremath{K^+ K^- {\KS}(\pip\pim)}\xspace}
14: \def\KKKszz {\ensuremath{K^+ K^- {\KS}(\piz\piz)}\xspace}
15: \def\KKsKs {\ensuremath{K^+ \KS \KS}\xspace}
16: \def\phiKs {\ensuremath{\phi \KS}\xspace}
17: \def\phiKz {\ensuremath{\phi K^0}\xspace}
18: \def\sPlot{\ensuremath{_s{\cal P}lot}\xspace}
19: \def\splot{\ensuremath{_s{\cal P}lot}\xspace}
20: \def\splots{\ensuremath{_s{\cal P}lot}s\xspace}
21: \def\mKK {\ensuremath{m_{\Kp\Km}}\xspace}
22: \def\mKpKz {\ensuremath{m_{\Kp\Kz}}\xspace}
23: \def\mKmKz {\ensuremath{m_{\Km\Kz}}\xspace}
24: \def\betaeff {\ensuremath{\beta_{\mathit{eff}}}\xspace}
25: \def\Acp {\ensuremath{{A}_{\CP}}\xspace}
26: \def\cosH {\ensuremath{\cos \theta_H}\xspace}
27: \def\cosT {\ensuremath{\cos \theta_{\rm T}}\xspace}
28: \def\bra {\ensuremath{\left <}}
29: \def\ket {\ensuremath{\right >}}
30: \def\spk{\ensuremath{S_{\phi K}}\xspace}
31: \def\cpk{\ensuremath{C_{\phi K}}\xspace}
32:
33: \def\fisher {\ensuremath{\mathcal{F}}\xspace}
34: \def\LowMass {Low-mass\xspace}
35: \def\HighMass {High-mass\xspace}
36: \def\hjphi {\ensuremath{\phi(1020)\xspace}}
37: \def\fzone {\ensuremath{f_0(980)\xspace}}
38: \def\hjX {\ensuremath{X(1550)\xspace}}
39:
40: \setlength{\textwidth}{16.5cm}
41: \setlength{\textheight}{22.2cm}
42: \setlength{\hoffset}{-2.0cm}
43: \setlength{\voffset}{-2.3cm}
44:
45: \long\def\inst#1{\par\nobreak\kern 4pt\nobreak
46: {\it #1}\par\vskip 10pt plus 3pt minus 3pt}
47:
48: \begin{document}
49: {\pagestyle{empty}
50:
51: \begin{flushright}
52: \babar-CONF-\BABARPubYear/\BABARConfNumber \\
53: SLAC-PUB-\SLACPubNumber \\
54: \end{flushright}
55:
56:
57:
58: \par\vskip 5cm
59:
60: \begin{center}
61: \Large \bf Measurement of {\boldmath$\CP$}-Violating Asymmetries in the {\boldmath $\Bz\to\Kp\Km\KS$} Dalitz Plot
62: \end{center}
63: \bigskip
64:
65: \begin{center}
66: \large The \babar\ Collaboration\\
67: \mbox{ }\\
68: August 5, 2008
69: \end{center}
70: \bigskip \bigskip
71:
72:
73: \begin{center}
74: \large \bf Abstract
75: \end{center}
76: We present a preliminary measurement of \CP-violation parameters in the decay $\Bz \to \KKKs$,
77: using approximately 465 million $\BB$ events collected by the $\babar$ detector at SLAC.
78: %
79: Reconstructing the neutral kaon as $\KS \to \pip\pim$ or $\KS \to \piz\piz$, we analyze the Dalitz plot
80: distribution and measure fractions to intermediate states. We extract \CP parameters from the asymmetries
81: in amplitudes and phases between $\Bz$ and $\Bzb$ decays across the Dalitz plot. From a fit to the whole Dalitz
82: plot, we measure $\betaeff= 0.44 \pm 0.07 \pm 0.02$, $\Acp= 0.03\pm 0.07 \pm 0.02$, where the first
83: uncertainties are statistical and the second ones are systematic. For decays to $\phi\KS$, we measure
84: $\betaeff= 0.13 \pm 0.13 \pm 0.02$, $\Acp= 0.14 \pm 0.19 \pm 0.02$. For decays to $f_0\KS$, we measure
85: $\betaeff= 0.15 \pm 0.13 \pm 0.03$, $\Acp= 0.01 \pm 0.26 \pm 0.07$. From a fit to the region of the Dalitz
86: plot with $\mKK>1.1\gevcc$, we measure $\betaeff= 0.52 \pm 0.08 \pm 0.03$, $\Acp= 0.05 \pm 0.09 \pm 0.04$.
87:
88:
89: \vfill
90: \begin{center}
91:
92: Submitted to the 34$^{\rm th}$ International Conference on High-Energy Physics, ICHEP 08,\\
93: 29 July---5 August 2008, Philadelphia, Pennsylvania.
94:
95: \end{center}
96:
97: \vspace{1.0cm}
98: \begin{center}
99: {\em Stanford Linear Accelerator Center, Stanford University,
100: Stanford, CA 94309} \\ \vspace{0.1cm}\hrule\vspace{0.1cm}
101: Work supported in part by Department of Energy contract DE-AC02-76SF00515.
102: \end{center}
103:
104: \newpage
105: }
106:
107: \input authors_ICHEP2008.tex
108:
109: \section{INTRODUCTION}
110: \label{sec:Introduction}
111:
112:
113: We present a time-dependent analysis of the Dalitz plot (DP) in flavor tagged $\Bz\to
114: \KKKs$ decays, with the \KS reconstructed as $\KS \to \pip \pim$ or $\KS \to \piz \piz$
115: (unless otherwise stated, charge conjugates are implied throughout this paper).
116: In the Standard Model (SM), these decays are dominated by
117: $\b \to s\bar{s}s$ gluonic penguin amplitudes, with a single weak
118: phase. Contributions from $b\to u \bar{q}q$ tree amplitudes,
119: proportional to the Cabibbo-Kobayashi-Maskawa (CKM) matrix element
120: $V_{ub}$ with a \CP-violating weak phase $\gamma$~\cite{ref:PDG}, are small, but may
121: depend on the position in the Dalitz plot. In $\Bz\to\phi(\Kp\Km)\Kz$
122: decays the modification of the \CP asymmetry due to the presence of
123: suppressed tree amplitudes is at $\cal
124: O$(0.01)~\cite{Beneke:2005pu,Buchalla:2005us}, while at higher $\Kp\Km$ masses a larger
125: contribution at $\cal O$(0.1) is possible~\cite{Cheng:2005ug}.
126: Therefore, to very good precision, we also expect the direct \CP
127: asymmetry for these decays to be small in the SM. The \CP asymmetry
128: in $\Bz \to \KKKs$ decay arises from the interference
129: of decays and $\Bz \leftrightarrow \Bzb$ mixing, with a relative phase
130: of $2\beta$. The Unitarity Triangle angle $\beta$ has been measured
131: in $\Bz\to [c\bar{c}]\Kz$ decays to be
132: $\sin2\beta=0.685 \pm 0.032$~\cite{Aubert:2004zt,Abe:2005bt}.
133: Current direct measurements favor the solution of $\beta=0.37$
134: over $\beta=1.20$ at the 98.3\% C.L.~\cite{Itoh:2005JspiKst,Krokovny:2006Dh0,Dalseno:2007DstDstKs,
135: Aubert:2005JPsiKpi,Aubert:2006BDsDsK,Aubert:2007BDh}. Furthermore,
136: the $\beta=0.37$ solution is the only one consistent with all indirect constraints
137: ~\cite{ref:CKMfit,ref:UTfit}.
138:
139: The decay $\Bz \to \KKKs$ is one of the most promising processes with which to
140: search for physics beyond the SM. Since the leading amplitudes
141: enter only at the one-loop level, additional contributions from heavy
142: non-SM particles may be of comparable size. If the amplitude from
143: heavy particles has a \CP-violating phase, the measured \CP-violation
144: parameters may differ from those expected in the SM.
145:
146: Previous \babar\ measurements of the \CP asymmetry in $\Bz \to \KKKz$ decays
147: have been performed on $383\times 10^6$ \BB events~\cite{Previous}. This
148: analysis updates that previous result with a larger dataset.
149:
150:
151: \section{DATASET AND DETECTOR}
152: \label{sec:dataset}
153:
154: The data used in this analysis were collected with the \babar\ detector
155: at the \pep2\ asymmetric-energy \B factory at SLAC. A total of 465
156: million \BB pairs were used.
157:
158: The \babar\ detector is described in detail
159: elsewhere~\cite{ref:babar}. Charged particle (track) momenta are
160: measured with a 5-layer double-sided silicon vertex tracker (SVT) and a
161: 40-layer drift chamber (DCH) coaxial with a 1.5-T superconducting solenoidal
162: magnet. Neutral cluster (photon) positions and energies are measured
163: with an electromagnetic calorimeter (EMC) consisting of 6580 CsI(Tl)
164: crystals. Charged hadrons are identified with a detector of
165: internally reflected Cherenkov light (DIRC) and specific ionization
166: measurements (\dedx) in the tracking detectors (DCH, SVT). Neutral hadrons that do not
167: interact in the EMC are identified with detectors, up to 15 layers
168: deep, in the flux return steel (IFR).
169:
170: In addition to the data collected by \babar, this analysis uses various samples of Monte Carlo (MC) events
171: based on GEANT4~\cite{ref:geant4}.
172: A sample of simulated \KKKs events using a full Dalitz plot model based on \babar's previous measurement is used to
173: study signal events, while backgrounds from \B meson decays are studied using a separate sample of
174: simulated events.
175:
176:
177: \section{EVENT RECONSTRUCTION}
178: \label{sec:selection}
179:
180: We reconstruct $\Bz \to \KKKs$ decays by combining two oppositely
181: charged tracks with a $\KS\to\pip\pim$ or $\KS\to\piz\piz$
182: candidate. The \Kp and \Km tracks must have at least 12 measured DCH
183: coordinates, a minimum transverse momentum of
184: 0.1~\gevc, and must originate from the nominal beam spot. Tracks
185: are identified as kaons using a likelihood ratio
186: that combines \dedx measured in the SVT and DCH with the Cherenkov
187: angle and number of photons measured in the DIRC. The \Kpm candidates are required to be loosely compatible
188: with the kaon hypothesis if the \KpKm invariant mass is less than $1.1\gevcc$, while a tighter compatibility
189: is required in all other cases to further suppress background.
190:
191: For all modes, the main source of background is random combinations of
192: particles produced in events of the type $e^+e^-\to q\bar{q}~(q=u,d,s,c)$
193: (continuum). Additional background from decays of $B$
194: mesons to other final states (\BB background), with and without charm particles, is
195: estimated from MC events.
196:
197: We use event-shape variables, computed in the center-of-mass (CM)
198: frame, to separate continuum events with a jet-like topology from the
199: more isotropic \B decays. Continuum events are suppressed by
200: requiring the quantity $|\cosT|$ to be less than 0.9,
201: where $\theta_{\rm T}$ is the angle between the thrust axis calculated with the \B candidate's daughters and the thrust axis
202: formed from the other charged and neutral particles in the event.
203: Further discrimination comes from a Fisher discriminant (\fisher) based on 1) \cosT,
204: 2) 0th and 2nd order Legendre moments $\mathcal{L}_{i=0,2} = \sum_j p_j |cos(\theta_{j})|^i$,
205: where $j$ is all tracks and clusters not used to reconstruct the \B meson, $p_j$ is their momentum, and
206: $\theta_{j}$ is the angle to the \B thrust axis, and 3) the
207: magnitude of the cosine of the angle of the \B with respect to the collision axis $|\cos{\theta_{B}}|$.
208:
209: In a small fraction of events, more than one \B candidate in a single event passes our selection
210: criteria. In this case, a single best \B candidate is selected based on the \KS invariant mass and
211: on the quality of the kaon tracks.
212:
213: \B candidates are identified using two kinematic variables that separate
214: signal from continuum background. These are the beam-energy-substituted mass
215: $\mes \equiv \sqrt{ (s/2 + {\bf p}_{i}\cdot{\bf p}_{B})^{2}/E_{i}^{2}- {\bf
216: p}^{2}_{B}}$, where $\sqrt{s}$ is the total \epem CM
217: energy, $(E_{i},{\bf p}_{i})$ is the four-momentum of the initial
218: \epem system and ${\bf p}_{B}$ is the \B candidate momentum, both
219: measured in the laboratory frame, and $\Delta E \equiv E_{B} - \sqrt{s}/2$,
220: where $E_{B}$ is the \B candidate energy in the CM
221: frame.
222:
223: \subsection{{\boldmath $\Bz \to \Kp\Km\KS$, $\KS\to\pip\pim$}}
224: \label{sec:kkkspm_selection}
225:
226: For decays $\Bz \to \Kp\Km\KS$ and $\KS\to\pip\pim$, \KS candidates
227: are formed from oppositely charged tracks with an invariant mass within
228: $20~\mevcc$ of the \KS mass~\cite{ref:PDG}. The \KS vertex is required to be
229: separated from the \Bz vertex by at
230: least $3\sigma$. The angle $\alpha_{K_S}$ between the \KS momentum vector and the
231: vector connecting the \Bz and \KS vertices must satisfy $\cos\alpha_{K_S} >
232: 0.999$. Distributions of the kinematic variables \mes and \DeltaE in data,
233: for signal and background events calculated using the \splot event-weighting
234: method~\cite{ref:sPlots}, are shown in Fig.~\ref{fig:kkks+-_event_selection}.
235:
236:
237: \subsection{{\boldmath $\Bz \to \Kp\Km\KS$, $\KS\to\piz\piz$}}
238:
239: For decays $\Bz \to \Kp\Km\KS$ and $\KS\to\piz\piz$, \KS candidates
240: are formed from two $\piz\to\gamma\gamma$ candidates. Each of the
241: four photons must have $E_{\gamma} > 0.05 \gev$ and have a transverse
242: shower shape loosely consistent with an electromagnetic
243: shower. Additionally, we require each \piz candidate to satisfy $0.100
244: < m_{\gamma\gamma} < 0.155 \gevcc$. The resulting $\KS\to\piz\piz$
245: mass is required to satisfy $0.4776 < m_{\piz\piz} < 0.5276~\gevcc$. A \KS
246: mass constraint is then applied for the reconstruction of the \Bz
247: candidate.
248:
249: The kinematic variables \mes and \DeltaE are formed for each candidate
250: as in Sec.~\ref{sec:selection}. Distributions of these
251: variables in data, for signal and background events calculated using the \splot event-weighting
252: method, are shown in Fig.~\ref{fig:kkks00_event_selection}.
253: Note that the mean of the signal \DeltaE distribution is shifted from zero due to energy leakage in the EMC.
254:
255:
256:
257: \begin{figure}[ptb]
258: \center
259: \begin{tabular}{ll}
260: \includegraphics[height=5.5cm]{epsfiles/wholeDP-mES-Kspp-sPlot.eps} & \includegraphics[height=5.5cm]{epsfiles/wholeDP-dE-Kspp-sPlot.eps}
261: \end{tabular}
262: \caption{Distributions of kinematic variables (left) \mes\ and (right) \DeltaE for the \KKKspm sample. The plots show
263: signal, with the continuum background shown in the insets. The points are data events weighted with the \sPlot\
264: technique, while the curves are the PDF shapes used in the ML fit (Sec. \ref{sec:Dalitz}).}
265: \label{fig:kkks+-_event_selection}
266: \end{figure}
267:
268:
269:
270: \begin{figure}[ptb]
271: \center
272: \begin{tabular}{ll}
273: \includegraphics[height=5.5cm]{epsfiles/wholeDP-mES-Ks00-sPlot.eps} & \includegraphics[height=5.5cm]{epsfiles/wholeDP-dE-Ks00-sPlot.eps}
274: \end{tabular}
275: \caption{Distributions of kinematic variables (left) \mes\ and (right) \DeltaE for the \KKKszz sample. The plots show
276: signal, with the continuum background shown in the insets. The points are data events weighted with the \sPlot\
277: technique, while the curves are the PDF shapes used in the ML fit (Sec. \ref{sec:Dalitz}).}
278: \label{fig:kkks00_event_selection}
279: \end{figure}
280:
281:
282: \section{ANALYSIS OF THE DALITZ PLOT}
283: \label{sec:Dalitz}
284: Four-momentum conservation in a three-body decay gives the relation
285: $ M^2_{\Bz} + m^2_{1} + m^2_{2} +
286: m^2_{3} ~=~ m^2_{12} + m^2_{13} + m^2_{23}$, where
287: $m^2_{ij}=(p_i+p_j)^2$ is the square of the invariant mass of a
288: daughter pair. This constraint leaves a choice of two independent
289: Dalitz plot variables to describe the decay dynamics of a spin-zero
290: particle. In this analysis we choose the $\Kp\Km$ invariant mass \mKK
291: and the cosine of the helicity angle between the $\Kp$ and the $\KS$ in the
292: $\Kp\Km$ center-of-mass frame, \cosH.
293:
294: We perform an extended maximum likelihood fit to the measured time dependent Dalitz plot distribution. We first fit on the whole DP, then fit on the $\mKK>1.1\gevcc$ range (\HighMass), then fit on the $\mKK<1.1\gevcc$ range (\LowMass). All fits are performed on the combined \KKKspm and \KKKszz samples simultaneously.
295: The likelihood function ${\mathcal L}$
296: for each subsample is defined as
297: \begin{equation}
298: {\mathcal L} = \exp{\left(-\sum_{i}n_{i}\right)}
299: \prod_{j}\left[\sum_{i}n_{i}{\mathcal P}_{i,j}\right]
300: \label{eq::ml}
301: \end{equation}
302: where $i$ labels the different signal and background components, $j$ runs over all events in the sample, and $n_i$ is the event yield for events of the $i$-th component. The probability density function (PDF) ${\mathcal P}_i$ of each component is defined as
303: \begin{equation}
304: \label{eq::mlprod}
305: {\mathcal P}_i \equiv {\mathcal P}_i(\mes) \cdot {\mathcal P}_i(\DeltaE) \cdot {\mathcal P}_i(\fisher) \cdot {\mathcal P}_{DP,i}(\mKK, \cosH, \deltat, q_{tag})\otimes {\cal R}_i(\deltat, \sigma_{\deltat}),
306: \end{equation}
307: where $q_{tag}$ is the flavor of the tagged \B (1 for \Bz and -1 for \Bzb), and $\deltat=t_{rec}-t_{tag}$ is the
308: difference of the proper decay times of the two
309: $B$-mesons in the \FourS decay. $\sigma_{\deltat}$ is the error on \deltat, and $\cal R$ is the \deltat\ resolution function determined from a high statistics independent sample~\cite{Aubert:2004zt}.
310: For the purpose of calculating the DP coordinates
311: \mKK and \cosH, we refit the \B candidates applying a \B mass constraint. This ensures that the \B candidates are reconstructed within the DP boundary.
312: The Fisher discriminant PDF, ${\mathcal P}(\fisher)$, is only used in the \LowMass fit (see Sec.~\ref{sec:Physics}). Because
313: the Fisher discriminant is highly correlated with the position on the DP, we do not use the Fisher discriminant PDF for the fit to the whole DP or for the \HighMass fit.
314: The Fisher distributions are shown in Fig.~\ref{fig:Fisher}. The PDFs for the individual fit components are described in more detail below.
315:
316: \begin{figure}[ptb]
317: \center
318: \epsfig{file=epsfiles/Fisher.eps, width=0.5\textwidth}\\
319: \caption{Fisher discriminant distributions for signal and continuum background and for \KKKspm sample and \KKKszz sample.
320: Distributions are normalized to unit area.}
321: \label{fig:Fisher}
322: \end{figure}
323:
324:
325: \subsection{Background in the Time-Dependent Dalitz Plot}
326:
327: We have two background components in our fit: continuum and \BB background. For the continuum background component,
328: we use the ARGUS function~\cite{Albrecht:1990cs} for ${\mathcal P}(\mes)$, and linear polynomial functions
329: for ${\mathcal P}(\DeltaE)$. The \deltat distribution is described by a double-Gaussian resolution function
330: convolved with a PDF of the following form:
331: \begin{equation}
332: {\cal P}(\deltat) = f_{prompt} \delta(\deltat) + (1-f_{prompt}) e^{-|\deltat|/\tau_{bg}},
333: \end{equation}
334: which allows for background decays with both zero and non-zero lifetimes. The Dalitz plot for the continuum
335: background is parameterized using a two-dimensional histogram PDF in the variables \mKK and \cosH. The histogram
336: is filled with candidates from the region $5.2 < \mes < 5.26~\gevcc$.
337:
338: We estimate the amount of $\BB$ background from Monte Carlo events. The \BB\ background is
339: almost purely combinatorial and is a few percent of the total background.
340: In the \KKKspm mode, the \mes and \DeltaE PDFs for the \BB backgrounds are
341: parameterized with the same functional forms as the continuum backgrounds.
342: Due to non-negligible correlation between \mes and \DeltaE for \BB background in
343: the \KKKszz mode, we construct a two-dimensional smoothed histogram PDF in those variables.
344: The \deltat\ distribution is described with a PDF similar to the continuum backgrounds, but we
345: also allow for the possibility that the non-zero lifetime component has a time-dependent \CP asymmetry
346: proportional to $\sin\deltamd\deltat$ or $\cos\deltamd\deltat$, where \deltamd is the mixing frequency of
347: the \Bz meson. These asymmetries are set to zero in the nominal fit, but are varied as a systematic uncertainty.
348: The Dalitz plot is described using a two-dimensional histogram PDF in a manner similar to the continuum
349: backgrounds.
350:
351:
352: \subsection{Signal Decays in the Time-Dependent Dalitz Plot}
353: \label{sec:tddp}
354:
355: The signal components of the PDFs for ${\mathcal P}(\mes)$ and ${\mathcal P}(\DeltaE)$ are parameterized using
356: modified Gaussian distributions:
357: ${\mathcal P}(x) \propto \exp [ - (x - x_0)^2/(2 \sigma_{\pm}^2 + \alpha_{\pm} (x- x_0)^2) ].$
358: We determine the parameters $x_0$, $\sigma_+$, $\sigma_-$, $\alpha_+$, and $\alpha_-$ using MC events, and fix them
359: in fits to data. For $x<x_0$ ($x>x_0$), the parameters $\sigma_-, \alpha_-$ ($\sigma_+, \alpha_+$) are used.
360:
361: For signal events, the time-dependence is a function of location in the DP.
362: When the flavor of the tagged \B $q_{tag}$, and the difference of
363: the proper decay times \deltat, are measured,
364: the time- and flavor-dependent decay rate over the Dalitz plot can be
365: written as
366: \begin{eqnarray}
367: d\Gamma =\frac{1}{(2\pi)^3}\frac{1}{32 M_{\Bz}^3} \frac{e^{-|\deltat|/\tau_{\Bz}}}{4\tau_{\Bz}} &\times&
368: \Big[~ \left | {\cal A} \right |^2 + \left | \bar{ {\cal A} } \right |^2 \nonumber
369: + q_{tag}~(1-2 w) ~2 Im \left ( e^{-2i\cdot\beta} \bar{\cal A} {\cal A}^* \right ) \sin\deltamd\deltat \\
370: && -~ q_{tag}~(1-2 w) ~\left (\left | {\cal A} \right |^2 - \left | \bar{ {\cal A} } \right |^2 \right ) \cos\deltamd\deltat
371: ~\Big ],
372: \label{eq::dalitz_plot_rate}
373: \end{eqnarray}
374: where $q_{tag} = +1(-1)$ when the other \B meson is identified as a \Bz(\Bzb) using a neural network
375: technique~\cite{Aubert:2004zt}. The parameter $w$ is the fraction of events in which the \B meson is mistagged with
376: the incorrect flavor, and the parameter $\beta$ is the CKM
377: angle $\beta$, coming from \Bz-\Bzb mixing.
378: %
379: Approximately 75\% of the signal events have tagging information
380: and contribute to the measurement of CP violation parameters. After
381: accounting for the mistag rate, the effective tagging efficiency
382: is $(31.2\pm 0.3)\%$.
383: %
384: Events without tagging information are assigned a mistag rate of $w=0.5$, and are included in the fit as they contribute
385: to the determination of the Dalitz plot parameters.
386: %
387: Decay amplitudes $\cal A$ and $\bar{\cal A}$ are defined in
388: (\ref{eq:A}) and (\ref{eq:Abar}) below.
389: %
390: $M_{\Bz}$, $\tau_{\Bz}$, and \deltamd are
391: the mass, lifetime, and mixing frequency of the \Bz meson, respectively~\cite{ref:PDG}.
392:
393: The PDF for the Dalitz plot rate takes the form
394: \begin{eqnarray}
395: {\cal P}_{DP} \propto d\Gamma(\mKK,\cos\theta_H,\deltat, q_{tag}) \cdot \varepsilon(\mKK,\cos\theta_H) \cdot |J|
396: \otimes {\cal R}(\deltat, \sigma_{\deltat}),
397: \end{eqnarray}
398: where $|J(\mKK)| = (2 \mKK)(2 q p)$ is the Jacobian of the transformation $(m^2_{\Kp\Km},m^2_{\Kp\KS}) \leftrightarrow (m_{\Kp\Km},\cos\theta_H)$, and is
399: given in terms of the charged kaon momentum $q$ and neutral kaon momentum $p$, in the $\Kp\Km$ frame.
400: The efficiency $\varepsilon$ is calculated from high-statistics samples of simulated events and depends on the position on the Dalitz plot.
401:
402:
403: The amplitude $\cal A$ ($\bar{\cal A}$) for the decay $\Bz\to\Kp\Km\KS$ ($\Bzb\to\Km\Kp\overline{\KS}$) is, in our isobar model,
404: written as a sum of decays through intermediate resonances:
405: \begin{eqnarray}
406: {\cal A} &=& \sum\limits_r c_r (1+b_r) e^{i (\phi_r + \delta_r)} \cdot f_r, \hspace{1cm}\mathrm{and} \label{eq:A} \\
407: \bar{\cal A}&=& \sum\limits_r c_r (1-b_r) e^{i (\phi_r - \delta_r)} \cdot \bar{f}_r. \label{eq:Abar}
408: \end{eqnarray}
409: The isobar coefficients $c_r$ and $\phi_r$ are the magnitude and phase of the
410: amplitude of component $r$, and we allow for different isobar
411: coefficients for $\Bz$ and $\Bzb$ decays through the asymmetry
412: parameters $b_r$ and $\delta_r$. The
413: function $f_r = F_r \times T_r \times Z_r$ describes the dynamic
414: properties of a resonance $r$, where $F_r$ is the form-factor for the
415: resonance decay vertex, $T_r$ is the resonant mass-lineshape, and
416: $Z_r$ describes the angular distribution in the
417: decay~\cite{blatt,Zemach:1963bc}.
418:
419: Our model includes the $\phi(1020)$, for which we use the Blatt-Weisskopf centrifugal barrier factor
420: $F_r=1/\sqrt{1+(Rq)^2}$~\cite{blatt}, where $q$ is the daughter momentum in the resonance frame,
421: and $R$ is the effective meson radius, taken to be $R=1.5~\gev^{-1}~(0.3~\fm)$.
422: For the scalar decays included in our model ($f_0(980)$, $\hjX$, and $\chi_{c0}$), we use a constant form-factor.
423: Note that we have omitted a similar centrifugal factor for the $\Bz$ decay vertex into the $\phi\Kz$ intermediate state
424: since its effect is negligible due to the small width of the $\phi(1020)$ resonance.
425:
426: The angular distribution is constant for scalar decays, whereas for vector decays $Z\sim \vec{q} \cdot \vec{p}$,
427: where $\vec{q}$ is the momentum of the resonant daughter, and $\vec{p}$ is the momentum of the third particle in
428: the resonance frame.
429: We describe the line-shape for the $\phi(1020)$, $\hjX$, and $\chi_{c0}$ using the relativistic Breit-Wigner function
430: \begin{equation}
431: T(m) = \frac{1}{m^2_r - \mKK^2 - i m_r \Gamma(m)},
432: \end{equation}
433: where $m_r$ is the resonance pole mass. The mass-dependent width is given as
434: $
435: \Gamma(\mKK) =\Gamma_r \left ( q/q_r\right )^{2L+1} \left ( m_r / \mKK \right ) \left ( F_r(q)/F_r(q_r) \right )^2,
436: $
437: where $L$ is the resonance spin and $q=q_r$ when $\mKK=m_r$.
438: For the $\phi(1020)$ and $\chi_{c0}$ parameters, we use average measurements~\cite{ref:PDG}.
439: The $\hjX$ is less well-established.
440: Previous Dalitz plot analyses of $\Bp\to\Kp\Kp\Km$~\cite{Garmash:2004wa,Aubert:2006nu} and
441: $\Bz\to\Kp\Km\Kz$ decays~\cite{Aubert:2005kd} report observations of a scalar resonance at around 1.5~\gevcc.
442: The scalar nature has been confirmed by partial-wave analyses~\cite{Aubert:2005ja,Aubert:2006nu}.
443: However, previous measurements report inconsistent resonant widths: $0.145\pm 0.029$~\gevcc~\cite{Garmash:2004wa} and
444: $0.257 \pm 0.033$~\gevcc~\cite{Aubert:2006nu}. Branching fractions also disagree, so the nature of this component is still unclear~\cite{Minkowski:2004xf}.
445: In our nominal fit, we take the resonance parameters from Ref.~\cite{Aubert:2006nu}, which is based on
446: a larger sample of $\BB$ decays than Ref.~\cite{Garmash:2004wa}, and consider the narrower width given in the latter in the systematic error studies.
447:
448: The $f_0(980)$ resonance is described with the coupled-channel (Flatt\'e) function
449: \begin{equation}
450: T(\mKK) = \frac{1 }{ m^2_r - \mKK^2 - i m_r ( \rho_{K} g_{K} + \rho_{\pi} g_{\pi} ) },
451: \end{equation}
452: where $\rho_K (\mKK) =2 \sqrt{ 1 - 4 m^2_{K}/\mKK^2 }$, $\rho_\pi (\mKK) =2 \sqrt{ 1 - 4 m^2_{\pi}/\mKK^2 }$, and
453: the coupling strengths for the $KK$ and $\pi\pi$ channels are
454: taken as $g_\pi=0.165\pm0.018$~\gevcc, $g_K/g_\pi=4.21 \pm 0.33$,
455: and $m_r=0.965 \pm 0.010 $~\gevcc~\cite{Ablikim:2004wn}.
456:
457: In addition to resonant decays, we include non-resonant amplitudes.
458: Existing models consider contributions from contact terms or higher-resonance
459: tails~\cite{Cheng:2002qu,Fajfer:2004cx,Cheng:2005ug}, but they do not capture features observed in data.
460: %
461: We rely on a phenomenological parameterization~\cite{Garmash:2004wa} and
462: describe the non-resonant terms as
463: \begin{equation}
464: {\cal A}_{NR} (\bar{\cal A}_{NR}) = \left ( c_{12} e^{i\phi_{12}} e^{-\alpha m^2_{12}}
465: + c_{13} e^{i\phi_{13}} e^{-\alpha m^2_{13}}
466: + c_{23} e^{i\phi_{23}} e^{-\alpha m^2_{23}} \right )
467: \cdot (1 \pm b_{NR}) \cdot e^{\pm i\delta_{NR}}, \label{eq:nr}
468: \end{equation}
469: where 1,2,3 denote the three daughter particles of the \B meson.
470: The slope of the exponential function is consistent among previous measurements in both neutral and
471: charged $B$ decays into three kaons~\cite{Garmash:2004wa,Aubert:2006nu,Aubert:2005kd}, and we use $\alpha = 0.14 \pm 0.02~\gev^{-2} \cdot c^4$.
472:
473:
474: We compute the direct \CP-asymmetry parameters for resonance $r$ from the asymmetries in amplitudes ($b_r$)
475: and phases ($\delta_r$) given in Eqs.~(\ref{eq:A}, \ref{eq:Abar}). We define the rate asymmetry as
476: \begin{equation}
477: \Acp(r)=\frac{|\bar{\cal A}_r|^2 - |{\cal A}_r|^2}{|\bar{\cal A}_r|^2 + |{\cal A}_r|^2} =\frac{-2b_r}{1+b_r^2},
478: \label{eq:Acp}
479: \end{equation}
480: and $\betaeff (r) = \beta + \delta_r$ is defined as the total phase asymmetry. These asymmetries are related to the
481: \CP asymmetry parameters $C$ and $-\eta S$ using the approximations
482: \begin{eqnarray}
483: C_r \approx - \Acp(r), \hspace{1cm}\mathrm{and} \label{eq:C} \\
484: -\eta_r S_r \approx \frac{1-b_r^2}{1+b_r^2} \sin(2 \betaeff (r) ), \label{eq:etaS}
485: \end{eqnarray}
486: where $\eta_r$ is the \CP eigenvalue of the final state. The fraction for resonance $r$ is computed as
487: \begin{equation}
488: {\cal F}_r ~=~ \frac{ \int d\cos\theta_H ~d\mKK \cdot |J| \cdot (|{\cal A}_r |^2+|\bar{\cal A}_r|^2) }
489: { \int d\cos\theta_H ~d\mKK \cdot |J| \cdot (|{\cal A} |^2+|\bar{\cal A}|^2) }.
490: \end{equation}
491: The sum of the fractions can be different from unity due to interference between the isobars.
492:
493: In addition to the previously mentioned resonances, the decays $\Bz \to \Dp\Km~(\Dp \to \Kp\KS)$ and
494: $\Bz \to \Ds\Km~(\Ds \to \Kp\KS)$ are also counted as signal. We include non-interfering amplitudes for
495: these modes in our Dalitz plot
496: model, parameterizing the $D_{(s)}$ mesons on the Dalitz plot as Gaussian
497: distributions with widths taken from studies of simulated events. The parameters $b_r$ and $\delta_r$ are
498: fixed to zero for the decays $\Bz \to \Dp\Km$, $\Bz \to \Ds\Km$, and $\Bz \to \chi_{c0}\KS$ throughout
499: this analysis.
500:
501:
502: \section{RESULTS}
503: \label{sec:Physics}
504:
505: In order to determine parameters of the Dalitz plot model,
506: we perform three fits: 1) whole DP fit, 2) \LowMass ($\mKK<1.1\gevcc$) region fit, and 3) \HighMass ($\mKK>1.1\gevcc$) region fit.
507:
508:
509: \subsection{The whole Dalitz Plot fit}
510: We perform a fit to both 4316 $\Bz\to\KKKspm$ and 2205 $\Bz\to\KKKszz$ candidates simultaneously in the full Dalitz plot.
511: In this step we assume that all charmless decays have the same \CP-asymmetry parameters. A Fisher discriminant cut ($-2.5<\fisher<4$), which retains
512: about 95\% of signal events and 60\% of continuum events, is applied. We do not include the Fisher PDF in the fit.
513: We vary the event yields, isobar coefficients, and the two
514: \CP-asymmetry parameters $\Acp$ and $\betaeff$ averaged over the Dalitz plot.
515: We find a signal yield of $1268 \pm 43$ (\KKKspm) and $160 \pm 19$ (\KKKszz) events, and a \BB background yield of
516: $47 \pm 31$ (\KKKspm) and $24 \pm 16$ (\KKKszz) events.
517: The isobar amplitudes, phases, and fractions are listed in Table~\ref{tab:wholeDP}. The resonant fractions do not add up
518: to 100\% due to interference between the resonances. The \CP-asymmetry parameters, and the correlation coefficients
519: $\rho$ between them, are summarized in Table~\ref{tab:acp}. Fig.~\ref{fig:DV-wholeDP} shows a projection of the Dalitz
520: plot variable \mKK. Fig.~\ref{fig:dt} shows distributions of $\Delta t$ for \Bz-tagged and \Bzb-tagged events, and the
521: asymmetry ${\cal A}(\Delta t)=(N_{\Bz}-N_{\Bzb})/(N_{\Bz}+N_{\Bzb})$, obtained with the \splot technique.
522:
523: To calculate the significance of the nominal \betaeff result, many fits are performed with fixed but different \betaeff values. The change in
524: likelihood as a function of \betaeff is shown in Fig.~\ref{fig:scan}.
525:
526:
527: \begin{table}[h]
528: \center
529: \begin{tabular}{|lr|rrr|}
530: \hline \hline
531: \multicolumn{2}{|l|}{Decay } & Amplitude $c_r$ & Phase $\phi_r$ & Fraction ${\cal F}_r$ (\%) \\
532: \hline
533: \multicolumn{2}{|l|}{$\phi(1020)\KS$} & $ 0.00897 \pm 0.00096$ & $ -0.341 \pm 0.232$ & $12.6 \pm 1.0$ \\
534: \multicolumn{2}{|l|}{$f_0(980)\KS$} & $ 0.542 \pm 0.044$ & $ -0.201 \pm 0.157$ & $27.8 \pm 7.1$ \\
535: \multicolumn{2}{|l|}{$X_0(1550)\KS$} & $ 0.141 \pm 0.017$ & $ -0.370 \pm 0.154$ & $5.70\pm 1.70$ \\
536: $NR$ &$(\Kp\Km)$ & 1 (fixed) & 0 (fixed) & $98.1 \pm 18.7$ \\
537: &$(\Kp\KS)$ & $ 0.328 \pm 0.058$ & $ 1.81 \pm 0.23$ & $10.5 \pm 3.4$ \\
538: &$(\Km\KS)$ & $ 0.353 \pm 0.066$ & $ -1.44 \pm 0.27$ & $12.1 \pm 3.8$ \\
539: \hline
540: \multicolumn{2}{|l|}{$\chi_{c0}\KS$} & $ 0.0298\pm 0.0046$ & $ 0.732 \pm 0.437$ & $2.53 \pm 0.60$ \\
541: \multicolumn{2}{|l|}{$\Dp\Km$} & $ 1.34\pm 0.19$ & -- & $3.43 \pm 0.69$ \\
542: \multicolumn{2}{|l|}{$\Ds\Km$} & $ 0.826\pm 0.160$ & -- & $1.37 \pm 0.46$ \\
543: \hline \hline
544: \end{tabular}
545: \caption{Isobar amplitudes and phases from the fit to the whole DP.
546: Three rows for non-resonant (NR) contribution correspond to coefficients of exponential functions in Eq.~(\ref{eq:nr}). The errors are statistical only. }
547: \label{tab:wholeDP}
548: \end{table}
549:
550: \begin{table}[h]
551: \center
552: \begin{tabular}{|l|rr|}
553: \hline \hline
554: Name & Whole DP & \HighMass \\
555: \hline
556: $\Acp$ & $ 0.03 \pm 0.07 \pm 0.02$ & $0.05 \pm 0.09 \pm 0.04$ \\
557: \betaeff & $ 0.44 \pm 0.07 \pm 0.02$ & $0.52 \pm 0.08 \pm 0.03$ \\
558: $\rho$ & $ 0.041$ & $0.031$ \\
559: \hline \hline
560: \end{tabular}
561: \caption{The \CP-asymmetry parameters from the whole DP fit and the \HighMass region fit.
562: The first error is statistical and the second is systematic. $\rho$ is the correlation coefficient.}
563: \label{tab:acp}
564: \end{table}
565:
566:
567: \begin{figure}[ptb]
568: \center
569: \begin{tabular}{c}
570: \includegraphics[height=6.5cm]{epsfiles/wholeDP-mKK-Kspp-sPlot.eps}
571: \vspace{-.5cm}
572: \end{tabular}
573: \caption{For the whole DP region fit, the distribution of the Dalitz plot variable \mKK for
574: signal-weighted data events (points) compared with the fit PDF in the \KKKspm mode. }
575: \label{fig:DV-wholeDP}
576: \end{figure}
577:
578:
579:
580: \begin{figure}[ptb]
581: \center
582: \begin{tabular}{ll}
583: \includegraphics[width=8cm,height=4cm]{epsfiles/wholeDP-dT-Kspp-sPlot.eps} & \includegraphics[width=8cm,height=4cm]{epsfiles/LowMass-dT-Kspp-sPlot.eps} \\
584: \includegraphics[width=8cm,height=4cm]{epsfiles/wholeDP-dTasym-Kspp-sPlot.eps} & \includegraphics[width=8cm,height=4cm]{epsfiles/LowMass-dTasym-Kspp-sPlot.eps}
585: \end{tabular}
586: \caption{ The \deltat\ (top) distributions and asymmetries (bottom) in the whole DP (left) and \LowMass region (right),
587: for the \KKKspm mode.
588: For the \deltat\ distributions, \Bz- (\Bzb-) tagged signal-weighted events are shown as filled (open)
589: circles, with the PDF projection in dashed red (solid blue).}
590: \label{fig:dt}
591: \end{figure}
592:
593: \begin{figure}[ptb]
594: \center
595: \subfigure[]{\epsfig{file=epsfiles/scan-wholeDP.eps, width=0.35\textwidth}} \hspace{0.1 in}
596: \subfigure[]{\epsfig{file=epsfiles/scan-HighMass.eps, width=0.35\textwidth}}\\
597: \subfigure[]{\epsfig{file=epsfiles/scan-LowMass1.eps, width=0.35\textwidth}} \hspace{0.1 in}
598: \subfigure[]{\epsfig{file=epsfiles/scan-LowMass2.eps, width=0.35\textwidth}}\\
599: \caption{ The change in the value of -2log($\mathcal L$) as a function of \betaeff, for (a) the whole DP, (b) the \HighMass region, (c) \fzone, and (d) \hjphi.}
600: \label{fig:scan}
601: \end{figure}
602:
603:
604: \subsection{\HighMass fit}
605: We perform a fit to both 3112 $\Bz\to\KKKspm$ and 1917 $\Bz\to\KKKszz$ candidates in the \HighMass region ($\mKK>1.1\gevcc$) simultaneously. We fix all isobar coefficients to the values from the whole DP fit. We vary yields and shared \CP-asymmetry parameters. We find a signal yield of $894 \pm 36$ \KKKspm and $117 \pm 16$ \KKKszz events, and a \BB background yield of
606: $50 \pm 31$ (\KKKspm) and $20 \pm 15$ (\KKKszz) events. The fit results are summarized in Table~\ref{tab:acp}. Fig.~\ref{fig:DV-HighMass} shows a projection of the Dalitz plot variable \cosH for events in this region, using the \splot technique.
607:
608:
609: \begin{figure}[ptb]
610: \center
611: \begin{tabular}{c}
612: \includegraphics[height=6.5cm]{epsfiles/HighMass-cosH-Kspp-sPlot.eps}
613: \vspace{-.5cm}
614: \end{tabular}
615: \caption{For the \HighMass region fit, the distribution of the Dalitz plot variable \cosH for
616: signal-weighted data events (points) compared with the fit PDF in the \KKKspm mode.}
617: \label{fig:DV-HighMass}
618: \end{figure}
619:
620: \subsection{\LowMass fit}
621: In order to measure \CP-asymmetry parameters for components with low-\KpKm mass with reduced model-dependence from the rest of the DP, we select
622: events using a cut of $\mKK<1.1\gevcc$. Because we are only selecting a small region of the DP, the correlation between the Fisher discriminant
623: \fisher and the DP location is unimportant. We therefore relax the cut on \fisher, and add the \fisher PDF to the fit. After these requirements
624: on \mKK and \fisher, there are 1846 (\KKKspm) and 493 (\KKKszz) candidates remaining. The most significant contributions in this region come
625: from \hjphi\KS and \fzone\KS decays, with a smaller contribution from a low-\KpKm mass tail of non-resonant decays. We fix all the isobar
626: coefficients except for those of the $\hjphi$ to the values from the whole DP fit, and fix the \CP-asymmetry parameters $b_r$ and $\delta_r$
627: for all resonances except the $\hjphi$ and $\fzone$ to be 0.
628: We vary the events yields, isobar coefficients for the $\hjphi$, and separate \CP-asymmetry parameters for the $\hjphi$ and $\fzone$ in the fit.
629: We find signal yields of $381 \pm 23$ (\KKKspm) and $40 \pm 9$ (\KKKszz) events, and \BB background yields of
630: $12 \pm 13$ (\KKKspm) and $-3 \pm 5$ (\KKKszz) events
631:
632: The \CP-asymmetry results are listed in Table~\ref{tab:low_mass_yields_cp}; the systematic uncertainties will be described in Sec.~\ref{sec:Systematics}.
633: We find two solutions with likelihood difference $\Delta$log($\mathcal L$) = 0.1. Solution (1) is consistent with the SM, while Solution (2)
634: has a value of \betaeff for the \fzone\KS decay that differs significantly from the SM, as shown in Table~\ref{tab:low_mass_yields_cp}. The two
635: solutions also have significantly different values of $c_r$ for the $\hjphi$. Both solutions also have a mathematical ambiguity of $\pm\pi$ radians
636: on \betaeff for the $\hjphi$, and a correlated ambiguity of $\pm\pi$ radians on the isobar parameter $\phi_r$ for the $\hjphi$. This
637: ambiguity is present because the decay amplitude contains interference terms that only depend on the linear combinations
638: $\betaeff+\phi_r$ and $\betaeff-\phi_r$. We choose Solution (1) as our nominal solution. The correlation coefficients $\rho$ between the
639: \CP\ parameters for Solution (1) are shown in Table~\ref{tab:low_mass_yields_cp}. Because the decay rate
640: depends on interference terms between the \hjphi\KS and \fzone\KS decays, the significant correlation between the measured \CP parameters is
641: expected.
642:
643:
644: \begin{table}[h]
645: \center
646: \begin{tabular}{|l|r|r|rrrr|}
647: \hline \hline
648: Name & Solution (1) & Solution (2) & \multicolumn{4}{c|}{Correlation} \\
649:
650: & & &1 & 2 & 3 & 4 \\
651: \hline
652: 1 $\Acp(\phi\KS)$ & $ 0.14 \pm 0.19 \pm 0.02 $ & $ 0.13 \pm 0.18$ &1.0 & -0.09 & -0.28 & 0.09 \\
653: 2 $\betaeff (\phi\KS)$ & $ 0.13 \pm 0.13 \pm 0.02 $ & $ 0.14 \pm 0.14$ & & 1.0 & 0.54 & 0.65 \\
654: 3 $\Acp(f_0\KS)$ & $ 0.01 \pm 0.26 \pm 0.07 $ & $-0.49 \pm 0.25$ & & & 1.0 & 0.25 \\
655: 4 $\betaeff (f_0\KS)$ & $ 0.15 \pm 0.13 \pm 0.03 $ & $ 3.44 \pm 0.19$ & & & & 1.0 \\
656: \hline \hline
657: \end{tabular}
658: \caption{\CP-violation parameters for $\Bz\to\KKKs$ for $\mKK < 1.1~\gevcc$. The first error is statistical and the second is systematic.
659: Correlation coefficients are given for Solution (1) only.}
660: \label{tab:low_mass_yields_cp}
661: \end{table}
662:
663:
664: Fig.~\ref{fig:DV-LowMass} shows projections of the Dalitz plot distributions of events in this region, using the
665: \splot technique. Fig.~\ref{fig:dt} shows
666: distributions of $\Delta t$ for \Bz-tagged and \Bzb-tagged events, and the asymmetry ${\cal A}(\Delta t)=(N_{\Bz}-N_{\Bzb})/(N_{\Bz}+N_{\Bzb})$.
667:
668: The decay $\Bz\to\hjphi\KS$, with highly suppressed tree amplitudes, is,
669: in terms of theoretical uncertainty, the cleanest channel to interpret
670: possible deviations of the \CP-violation parameters from the SM
671: expectations.
672: Values of \betaeff are consistent with the value found
673: in $[c\bar{c}]\Kz$ decays~\cite{Aubert:2004zt,Abe:2005bt}.
674:
675:
676: \begin{figure}[ptb]
677: \center
678: \begin{tabular}{ll}
679: \includegraphics[height=5.5cm]{epsfiles/LowMass-mKK-Kspp-sPlot.eps} & \includegraphics[height=5.5cm]{epsfiles/LowMass-cosH-Kspp-sPlot.eps}
680: \end{tabular}
681: \caption{For the \LowMass region fit, the distributions of the Dalitz plot variables \mKK (left) and \cosH (right)
682: for signal-weighted data events (points) compared with the fit PDF in the \KKKspm mode.}
683: \label{fig:DV-LowMass}
684: \end{figure}
685:
686: We also calculate the parameters $C$ and $-\eta S$ for $\hjphi\KS$ and $\fzone\KS$ using the expressions in (\ref{eq:C}) and (\ref{eq:etaS}).
687: The results are shown in Table~\ref{tab:CandS}, along with $C$ and $-\eta S$ for the whole DP and \HighMass fits.
688:
689:
690: \begin{table}[h]
691: \center
692: \begin{tabular}{|l|c|c|}
693: \hline\hline
694: & $C$ & $-\eta S$ \\
695: \hline
696: Whole DP & $-0.03 \pm 0.07 \pm 0.02$ & $0.77 \pm 0.09 \pm 0.02$ \\
697: \hline
698: \HighMass & $-0.05 \pm 0.09 \pm 0.04$ & $0.86 \pm 0.08 \pm 0.03$ \\
699: \hline
700: $\hjphi\KS$ & $-0.14 \pm 0.19 \pm 0.02$ & $0.26 \pm 0.26 \pm 0.03$ \\
701: $\fzone\KS$ & $-0.01 \pm 0.26 \pm 0.07$ & $0.29 \pm 0.25 \pm 0.06$ \\
702: \hline\hline
703: \end{tabular}
704: \caption{The \CP asymmetry parameters $C$ and $-\eta S$, derived using Equations (\ref{eq:C})
705: and (\ref{eq:etaS}). Results are shown for the whole DP, the \HighMass region, and for both
706: $\hjphi\KS$ and $\fzone\KS$ in the \LowMass region. For the \LowMass results, only Solution
707: (1) is shown. The first error is statistical and the second is systematic.\label{tab:CandS}}
708: \end{table}
709:
710:
711: \section{SYSTEMATIC STUDIES}
712: \label{sec:Systematics}
713:
714: We study systematic effects on the \CP-asymmetry parameters
715: due to fixed parameters in the \mes and \DeltaE PDFs.
716: We assign systematic errors by comparing the fit with nominal parameters and
717: with parameters varied by their error ($\pm 1 \sigma$), and assign the average difference as the systematic error.
718: %
719: In addition, we account for a potential fit bias using values observed in studies
720: with MC samples generated with the nominal Dalitz plot model.
721: We take the average values of the bias observed in these studies as the systematic error.
722: %
723: We account for fixed \deltat\ resolution parameters, \Bz\ lifetime,
724: \Bz-\Bzb mixing and flavor tagging parameters. We also assign an error
725: due to interference between the CKM-suppressed $\bar{b}\to\bar{u} c\bar{d}$
726: and the favored $b\to c\bar{u}d$ amplitude for some tag-side $B$ decays~\cite{dcsd}.
727: %
728: Smaller errors due to beam-spot position uncertainty, detector alignment, and the
729: boost correction are based on studies done in charmonium decays. In the cases of the \LowMass
730: and \HighMass fits, we also assign systematic errors due to the isobar coefficients that are
731: fixed to the result from the whole DP fit.
732: %
733: In all fits we assume no direct \CP\ violation in decays dominated by the $b\to c$ transition
734: ($\chi_{c0}\KS$, $D_{(s)}K$).
735:
736: We also assign an error due to uncertainty in the resonant and non-resonant
737: line-shape parameters. The systematic uncertainty associated with the resonant component includes
738: the uncertainty in the mass and width of the X(1550), estimated by replacing the parameters used in
739: the nominal fit with the values found by different measurements: $m_r=1.491$~\gevcc,
740: $\Gamma=0.145$~\gev~\cite{Garmash:2004wa}.
741: %
742: All the systematic uncertainties are summarized in Table.~\ref{tab:sys_errors}.
743:
744:
745:
746:
747: \begin{table}[h]
748: \center
749: \begin{tabular}{|l|rr|rrrr|rr|}
750: \hline \hline
751: Parameter & \multicolumn{2}{c|}{Whole DP} & \multicolumn{2}{c}{$\phi\KS$} & \multicolumn{2}{c|}{$f_0\KS$} & \multicolumn{2}{c|}{\HighMass} \\
752: & $\Acp$ & \betaeff & $\Acp$ & \betaeff & $\Acp$ & \betaeff & $\Acp$ & \betaeff \\ \hline
753: \hline
754: Fixed PDF Parameters & 0.010 & 0.010 & 0.014 & 0.010 & 0.025 & 0.015 & 0.013 & 0.010 \\
755: Fit Bias & 0.007 & 0.011 & 0.009 & 0.012 & 0.011 & 0.011 & 0.014 & 0.009 \\
756: DCSD, Beam Spot, other & 0.015 & 0.004 & 0.015 & 0.004 & 0.015 & 0.004 & 0.015 & 0.004 \\
757: Dalitz Model & 0.005 & 0.005 & 0.009 & 0.002 & 0.060 & 0.024 & 0.027 & 0.023 \\ \hline
758: Total & 0.020 & 0.016 & 0.024 & 0.016 & 0.068 & 0.031 & 0.036 & 0.026 \\
759: \hline \hline
760: \end{tabular}
761: \caption{Summary of systematic errors on \CP-asymmetry parameters. Errors for
762: $\phi\KS$ and $f_0\KS$ \CP-parameters are based on the \LowMass region fit. Total is obtained from the quadratic sum of the individual systematics.}
763: \label{tab:sys_errors}
764: \end{table}
765:
766:
767:
768:
769:
770:
771: \section{CONCLUSIONS}
772: \label{sec:Summary}
773:
774: We performed a ML fit to analyze the DP distribution of $\Bz \to \KKKs$ decay with the full \babar\ dataset. From a
775: fit to the whole DP, we measure $\betaeff=0.44\pm 0.07 \pm 0.02$, $\Acp=0.03\pm 0.07\pm 0.02$, consistent
776: with our previous measurements~\cite{Previous} and compatible with the Standard Model values $\beta \simeq 0.37, \Acp = 0$.
777: We measure \CP violation with a significance of 6.7 standard deviations (including statistical and systematic errors),
778: and we reject the solution near $\pi/2 - \beta$ at 4.8 standard deviations.
779:
780: From a fit to the region of the DP with $\mKK>1.1\gevcc$, we measure $\betaeff= 0.52 \pm 0.08 \pm 0.03$ and
781: $\Acp= 0.05 \pm 0.09 \pm 0.04$, compatible with the Standard Model expectations. We measure \CP violation
782: in this \HighMass region at 6.7 standard deviations.
783:
784: From a fit to events at low \Kp\Km masses, we measure $\betaeff= 0.13 \pm 0.13 \pm 0.02$ and $\Acp= 0.14 \pm 0.19 \pm 0.02$
785: for $\Bz \to \phi(1020)\KS$, and $\betaeff= 0.15 \pm 0.13 \pm 0.03$ and $\Acp= 0.01 \pm 0.26 \pm 0.07$ for
786: $\Bz \to f_0\KS$. The results for $\betaeff$ are roughly 1.7 standard deviations below the Standard Model value.
787:
788: These results supersede our previous measurements~\cite{Previous} made on a smaller dataset. All of our
789: results are consistent with our previous measurements.
790:
791:
792: \section{ACKNOWLEDGMENTS}
793: \label{sec:Acknowledgments}
794:
795: \input acknowledgements
796:
797: \begin{thebibliography}{99}
798:
799: \bibitem{ref:PDG}
800: W.-M.~Yao {\it et al.} [Particle Data Group],
801: J. Phys. G {\bf 33}, 1 (2006)
802:
803:
804: \bibitem{Beneke:2005pu}
805: M.~Beneke,
806: Phys.\ Lett.\ B {\bf 620}, 143 (2005)
807: [arXiv:hep-ph/0505075].
808:
809: \bibitem{Buchalla:2005us}
810: G.~Buchalla, G.~Hiller, Y.~Nir and G.~Raz,
811: JHEP {\bf 0509}, 074 (2005)
812: [arXiv:hep-ph/0503151].
813:
814: \bibitem{Cheng:2005ug}
815: H.~Y.~Cheng, C.~K.~Chua and A.~Soni,
816: Phys.\ Rev.\ D {\bf 72}, 094003 (2005)
817: [arXiv:hep-ph/0506268].
818:
819: \bibitem{Aubert:2004zt}
820: B.~Aubert {\it et al.} [BABAR Collaboration],
821: Phys.\ Rev.\ Lett.\ {\bf 94}, 161803 (2005)
822: [arXiv:hep-ex/0408127].
823:
824: \bibitem{Abe:2005bt}
825: K.~Abe {\it et al.} [Belle Collaboration],
826: [arXiv:hep-ex/0507037].
827:
828: \bibitem{Itoh:2005JspiKst}
829: R.~Itoh {\it et al.} [Belle Collaboration],
830: Phys.\ Rev.\ Lett.\ {\bf 95}, 091601 (2005)
831:
832: \bibitem{Krokovny:2006Dh0}
833: P.~Krokovny {\it et al.} [Belle Collaboration],
834: Phys.\ Rev.\ Lett.\ {\bf 97}, 081801 (2006)
835:
836: \bibitem{Dalseno:2007DstDstKs}
837: J.~Dalseno {\it et al.} [Belle Collaboration],
838: Phys.\ Rev.\ D {\bf 76}, 072004 (2007)
839:
840: \bibitem{Aubert:2005JPsiKpi}
841: B.~Aubert {\it et al.} [BABAR Collaboration],
842: Phys.\ Rev.\ D {\bf 71}, 032005 (2005)
843: [arXiv:hep-ex/0411016].
844:
845: \bibitem{Aubert:2006BDsDsK}
846: B.~Aubert {\it et al.} [BABAR Collaboration],
847: Phys.\ Rev.\ D {\bf 74}, 091101 (2006)
848: [arXiv:hep-ex/0608016].
849:
850: \bibitem{Aubert:2007BDh}
851: B.~Aubert {\it et al.} [BABAR Collaboration],
852: Phys.\ Rev.\ Lett.\ {\bf 99}, 231802 (2007)
853: [arXiv:0708.1544[hep-ex]].
854:
855: \bibitem{ref:CKMfit}
856: J.~Charles {\it et al.} (CKMfitter Group),
857: Eur.\ Phys.\ J.\ C {\bf 41}, 1-131 (2005)
858: [arXiv:hep-ph/0406184],
859: Updated results and plots available at:
860: http://ckmfitter.in2p3.fr
861:
862: \bibitem{ref:UTfit}
863: M.~Bona {\it et al.} (UTfit Collaboration),
864: Phys.\ Rev.\ Lett.\ {\bf 97} 151803 (2006),
865: [arXiv:hep-ph/0605213],
866: Updated results and plots available at:
867: http://www.utfit.org
868:
869: \bibitem{Previous}
870: B.~Aubert {\it et al.} [BABAR Collaboration],
871: Phys.\ Rev.\ Lett.\ {\bf 99} 161802 (2007)
872: [arXiv:0706.3885[hep-ex]].
873:
874: \bibitem{ref:babar}
875: B.~Aubert {\it et al.} [BABAR Collaboration],
876: Nucl.\ Instrum.\ Meth.\ A {\bf 479}, 1 (2002)
877: [arXiv:hep-ex/0105044].
878:
879: \bibitem{ref:geant4}
880: GEANT4 Collaboration, S. Agostinelli {\it et al.},
881: Nucl.\ Instrum. \ Meth. \ A {\bf 506}, 250 (2003)
882:
883:
884: \bibitem{ref:sPlots}
885: M.~Pivk and F.~R.~Le Diberder,
886: Nucl.\ Instrum.\ Meth.\ A {\bf 555}, 356 (2005)
887: [arXiv:physics/0402083].
888:
889: \bibitem{Albrecht:1990cs}
890: H.~Albrecht {\it et al.} [ARGUS Collaboration],
891: Z.\ Phys.\ C {\bf 48}, 543 (1990).
892:
893: \bibitem{blatt}
894: J.~M.~Blatt, V.~F.~Weisskopf,
895: ``Theoretical Nuclear Physics'',
896: John Wiley \& Sons, New York (1952).
897:
898: \bibitem{Zemach:1963bc}
899: C.~Zemach,
900: Phys.\ Rev.\ {\bf 133}, B1201 (1964).
901:
902: \bibitem{Garmash:2004wa}
903: A.~Garmash {\it et al.} [BELLE Collaboration],
904: Phys.\ Rev.\ D {\bf 71}, 092003 (2005)
905: [arXiv:hep-ex/0412066].
906:
907: \bibitem{Aubert:2005ja}
908: B.~Aubert {\it et al.} [BABAR Collaboration],
909: Phys.\ Rev.\ D {\bf 71}, 091102 (2005)
910: [arXiv:hep-ex/0502019].
911:
912: \bibitem{Aubert:2006nu}
913: B.~Aubert {\it et al.} [BABAR Collaboration],
914: [arXiv:hep-ex/0605003].
915:
916: \bibitem{Aubert:2005kd}
917: B.~Aubert {\it et al.} [BABAR Collaboration],
918: [arXiv:hep-ex/0507094].
919:
920: \bibitem{Minkowski:2004xf}
921: P.~Minkowski and W.~Ochs,
922: Eur.\ Phys.\ J.\ C {\bf 39}, 71 (2005)
923: [arXiv:hep-ph/0404194].
924:
925: \bibitem{Ablikim:2004wn}
926: M.~Ablikim {\it et al.} [BES Collaboration],
927: Phys.\ Lett.\ B {\bf 607}, 243 (2005)
928: [arXiv:hep-ex/0411001].
929:
930: \bibitem{Cheng:2002qu}
931: H.~Y.~Cheng and K.~C.~Yang,
932: Phys.\ Rev.\ D {\bf 66}, 054015 (2002)
933: [arXiv:hep-ph/0205133].
934:
935: \bibitem{Fajfer:2004cx}
936: S.~Fajfer, T.~N.~Pham and A.~Prapotnik,
937: Phys.\ Rev.\ D {\bf 70}, 034033 (2004)
938: [arXiv:hep-ph/0405065].
939:
940: \bibitem{Gronau:2005ax}
941: M.~Gronau and J.~L.~Rosner,
942: Phys.\ Rev.\ D {\bf 72}, 094031 (2005)
943: [arXiv:hep-ph/0509155].
944:
945: \bibitem{Aubert:2005ce}
946: B.~Aubert {\it et al.} [BABAR Collaboration],
947: Phys.\ Rev.\ D {\bf 72}, 072003 (2005)
948: [arXiv:hep-ex/0507004].
949: \bibitem{dcsd}
950: O.~Long, M.~Baak, R.~Cahn, and D.~Kirkby, \jprd{68}, 034010 (2003).
951:
952: \end{thebibliography}
953:
954:
955: \end{document}
956: