1: The heavy quark effective theory (HQET) combined with the operator
2: product expansion provides a framework in which inclusive $B$~decay
3: properties can be calculated. In particular, the moments of various
4: inclusive quantities in $B\to X_c\ell\nu$~decays, such as the lepton
5: momentum~$p_\ell$, the 4-momentum transfer squared~$q^2$ and the
6: hadronic recoil mass~$M_X$, can be related to the non-perturbative
7: parameters that also appear in calculations of the semileptonic decay
8: rate (see Ref.~\cite{ref:1,ref:2,ref:3} for calculations of the
9: moments of $M_X$). Measurements of these moments can thus improve
10: calculations of the semileptonic $B$~decay rate and lead to a more
11: precise determination of the magnitude of the
12: Cabibbo-Kobayashi-Maskawa (CKM) matrix element
13: $V_{cb}$~\cite{ref:4}. In this paper, we present a measurement of the
14: first and second moment of $M_X$, $\langle M_X\rangle$ and $\langle
15: M^2_X\rangle$, for lepton momentum thresholds ranging from 0.9 to
16: 1.6~GeV/$c$ in the center-of-mass (c.m.) frame.
17:
18: The analysis is based on the data recorded with the Belle
19: detector~\cite{ref:5} at the asymmetric energy $e^+e^-$~collider
20: KEKB~\cite{ref:6}, operating at a c.m.\ energy near the
21: $\Upsilon(4S)$~resonance. KEKB consists of a low energy ring (LER)
22: of 3.5 GeV positrons and a high energy ring (HER) of 8 GeV
23: electrons. The Belle detector is a large-solid-angle magnetic
24: spectrometer consisting of a three-layer silicon vertex detector
25: (SVD), a 50-layer central drift chamber (CDC), an array of aerogel
26: threshold \v{C}erenkov counters (ACC), a barrel-like arrangement of
27: time-of-flight scintillation counters (TOF), and an electromagnetic
28: calorimeter comprised of CsI(Tl) crystals (ECL) located inside a
29: super-conducting solenoid coil that provides a 1.5~T magnetic
30: field. The responses of the ECL, CDC ($dE/dx$) and ACC detectors are
31: combined to provide clean electron identification. Muons are
32: identified in the instrumented iron flux-return (KLM) located outside
33: of the coil. Charged hadron identification relies on the information
34: from the CDC, ACC and TOF~sub-detectors.
35:
36: The $\Upsilon(4S)$~dataset used for this study corresponds
37: to an integrated luminosity of 140~fb$^{-1}$, or about
38: 152 million $B\bar B$~events. Another 15~fb$^{-1}$ taken 60~MeV below
39: the resonance are used to subtract the non-$B\bar B$ background. Full
40: detector simulation based on GEANT~\cite{ref:7} is applied to Monte
41: Carlo (MC) simulated events. The size of the MC samples is
42: equivalent to about three times the integrated luminosity. The decay
43: $B\to D^*\ell\nu$ is simulated using a HQET-based
44: model~\cite{ref:8}. The ISGW2 model~\cite{ref:9} is used for the
45: decays $B\to D\ell\nu$ and $B\to D^{**}\ell\nu$. The non-resonant
46: $B\to D^{(*)}\pi\ell\nu$~component is generated according to the
47: model of Goity and Roberts~\cite{ref:10}.
48:
49: We use $B\bar B$~events in which the hadronic decay of one
50: $B$~meson is fully reconstructed ($B_\mathrm{tag}$) and the
51: semileptonic decay of the recoiling $B$~meson ($B_\mathrm{signal}$) is
52: inferred from the presence of a charged lepton (electron or muon). After
53: selecting hadronic events~\cite{ref:11}, we reconstruct
54: $B_\mathrm{tag}$ by searching the decay modes $B^+\to\bar
55: D^{(*)0}\pi^+, \bar D^{(*)0}\rho^+, \bar D^{(*)0}a_1^+$ and $B^0\to
56: D^{(*)-}\pi^+, D^{(*)-}\rho^+, D^{(*)-}a_1^+$~\cite{ref:12}.
57: Pairs of photons satisfying $E_\gamma>50$~MeV and
58: 117~MeV/$c^2<m(\gamma\gamma)<150$~MeV/$c^2$ are combined to form
59: $\pi^0$~candidates. $K^0_\mathrm{S}$~mesons are reconstructed from pairs of
60: oppositely charged tracks with invariant mass within $\pm
61: 30$~MeV/$c^2$ of the $K^0_\mathrm{S}$~mass and decay vertex displaced
62: from the interaction point. Candidate $\rho^+$ and $\rho^0$~mesons are
63: reconstructed in the $\pi^+\pi^0$ and $\pi^+\pi^-$ decay modes,
64: requiring their invariant masses to be within $\pm 150$~MeV/$c^2$ of the
65: $\rho$~mass. Candidate $a_1^+$~mesons are obtained by combining a
66: $\rho^0$~candidate with a charged pion and requiring an invariant mass
67: between 1.0 and 1.6~GeV/$c^2$.
68: $D^0$~candidates are searched for in the $K^-\pi^+$, $K^-\pi^+\pi^0$,
69: $K^-\pi^+\pi^+\pi^-$, $K^0_\mathrm{S}\pi^+\pi^-$ and
70: $K^0_\mathrm{S}\pi^0$~decay modes. The $K^-\pi^+\pi^+$ and
71: $K^0_\mathrm{S}\pi^+$~modes are used to reconstruct
72: $D^+$~mesons. Charmed mesons are selected in a window corresponding to
73: $\pm 3$ times the mass resolution in the respective decay
74: mode. $D^{*+}$~mesons are reconstructed by pairing a charmed meson
75: with a low momentum pion, $D^{*+}\to D^0\pi^+,D^+\pi^0$. The decay
76: modes~$D^{*0}\to D^0\pi^0$ and $D^{*0}\to D^0\gamma$ are used to
77: search for neutral charmed vector mesons.
78:
79: For each $B_\mathrm{tag}$~candidate, the beam-constrained
80: mass~$M_\mathrm{bc}$ and $\Delta E$ are calculated~\cite{ref:13}
81: \begin{equation}
82: M_\mathrm{bc} = \sqrt{(E^*_\mathrm{beam})^2-(\vec p^*_B)^2}~, \quad
83: \Delta E = E^*_B-E^*_\mathrm{beam}~,
84: \end{equation}
85: where $E^*_\mathrm{beam}$ is the beam energy in the c.m.\ system,
86: $\vec p^*_B$ and $E^*_B$ are the 3-momentum and the energy of the
87: $B_\mathrm{tag}$~candidate, respectively. The signal region for
88: $B_\mathrm{tag}$ is defined by the selections
89: $M_\mathrm{bc}>5.27$~GeV/$c^2$ and $\Delta E<50$~MeV. If multiple
90: candidates are found in a single event, the best candidate is chosen
91: based on $\Delta E$ and other variables. Fig.~\ref{fig:1} shows the
92: $M_\mathrm{bc}$ distribution for $\Delta E<50$~MeV. The number of the
93: fully reconstructed $B^+$ ($B^0$) events, obtained by fitting this
94: distribution with the sum of a Gaussian function for the signal
95: component and an ARGUS background function~\cite{ref:14}, is
96: $76,155\pm 511$ ($46,863\pm 374$) with a purity of the 74\% (81\%).
97: \begin{figure}
98: \begin{center}
99: \includegraphics{fig1.eps}
100: \end{center}
101: \caption{Reconstruction of the tag-side $B$~meson. The
102: $M_\mathrm{bc}$~distribution after requiring $\Delta E<50$~MeV is
103: shown for fully reconstructed (a) $B^+$ (b) $B^0$ decays. The fit
104: uses a Gaussian function for the signal and an ARGUS background
105: function~\cite{ref:14} for the combinatorial background.}
106: \label{fig:1}
107: \end{figure}
108:
109: Semileptonic decays of $B_\mathrm{signal}$ are selected by requiring
110: exactly one identified charged lepton among the particles that are not
111: associated to $B_\mathrm{tag}$. In the lepton momentum range relevant
112: to this analysis, electrons (muons) are selected with an efficiency of
113: 92\% (89\%) and the probability to misidentify a pion as an electron
114: (a muon) is 0.25\% (1.4\%)~\cite{ref:15,ref:16}. For $B^+$~tags, we
115: require $Q_\ell\cdot Q_B<0$, where $Q_\ell$ and $Q_B$ are the charges
116: of the lepton and of $B_\mathrm{tag}$, respectively.
117:
118: The charged tracks and neutral clusters which are associated neither
119: with $B_\mathrm{tag}$ nor with the charged lepton are assigned to the
120: hadronic $X$~system. The 4-momenta of the charged particles are
121: calculated with the pion mass except for identified kaons. The
122: missing 4-momentum in the event is calculated,
123: \begin{equation}
124: p_\mathrm{miss} =
125: (p_\mathrm{LER}+p_\mathrm{HER})-p_{B_\mathrm{tag}}-p_\ell-p_X~,
126: \end{equation}
127: where the indices LER and HER refer to the colliding beams. As only
128: the neutrino in $B\to X_c\ell\nu$ should be missing in the event, we
129: require the missing mass to be consistent with zero,
130: $|M^2_\mathrm{miss}|<2$~GeV$^2$/$c^4$.
131:
132: The neutrino 4-momentum is inferred from the missing momentum,
133: $p_\nu=(|\vec p_\mathrm{miss}|,\vec p_\mathrm{miss})$, and the
134: 4-momentum of the $X$~system is recalculated,
135: \begin{equation}
136: p'_X =
137: (p_\mathrm{LER}+p_\mathrm{HER})-p_{B_\mathrm{tag}}-p_\ell-p_\nu~.
138: \end{equation}
139: The $M_X$~resolution (defined as half width at the half maximum)
140: obtained with this procedure is about 200~MeV/$c^2$.
141:
142: The background consists of the following components: combinatorial
143: background in the $B_\mathrm{tag}$~sample, background from
144: secondary or misidentified leptons, and background from non-$B\bar
145: B$~events. The combinatorial background is dominant. Its
146: shape (in $M_X$ and $M^2_X$) is determined from the MC simulation, and
147: the $M_\mathrm{bc}$~side-band region
148: ($5.2$~GeV/$c^2<M_\mathrm{bc}<5.27$~GeV/$c^2$, $|\Delta E|<50$~MeV) is used to
149: determine its normalization, separately for each lepton momentum
150: threshold. The background from secondary or misidentified leptons is
151: significant in the low lepton momentum region. Its shape is estimated
152: from the MC simulation, and the number of well-reconstructed
153: $B_\mathrm{tag}$ candidates is used for its normalization. The
154: magnitude of the background from misidentified leptons is corrected
155: using the pion misidentification rate measured in
156: $K^0_\mathrm{S}\rightarrow\pi^+\pi^-$~decays in hadronic events. The
157: background from non-$B\bar B$~events is smallest, and it is subtracted
158: using the data taken 60~MeV below the
159: $\Upsilon(4S)$~resonance. Fig.~\ref{fig:2} shows the
160: $M_X$~distributions for different minimum lepton momenta in the c.m.\
161: frame, together with the background estimation.
162: \begin{figure}
163: \begin{center}
164: \includegraphics{fig2.eps}
165: \end{center}
166: \caption{The $M_X$~distributions for (a) $p^*_\ell>0.9$~GeV/$c$ and
167: (b) $p^*_\ell>1.6$~GeV/$c$. The data points are the
168: $\Upsilon(4S)$~data after subtraction of the non-$B\bar
169: B$~background. The background from secondary or misidentified
170: leptons is indicated by the dashed line. The combinatorial
171: background in the $B_\mathrm{tag}$~sample is shown as a hatched
172: histogram.}
173: \label{fig:2}
174: \end{figure}
175:
176: To correct for the distortions caused by finite detector resolution,
177: the background-subtracted $M_X$ and $M^2_X$~distributions are unfolded
178: using the Singular Value Decomposition (SVD)
179: algorithm~\cite{ref:17}. The approach consists in describing the
180: detector response by a correlation matrix between the true and the
181: reconstructed $M_X$ ($M^2_X$) values which is determined
182: from the MC simulation. This matrix is decomposed into its eigenvalue
183: components, and only the contributions leading to a statistically
184: stable solution are kept. For the unfolding of the $M_X$ ($M^2_X$)
185: distribution, the measured spectrum is divided into 35 (39) 0.1
186: GeV/$c^2$ (0.333 GeV$^2$/$c^4$) wide bins. For the unfolded
187: distribution, 12 (13) bins are used.
188:
189: For each minimum lepton momentum, ranging from 0.9 to 1.6~GeV/$c$, the
190: background-subtracted $M_X$ and $M_X^2$~distributions are unfolded,
191: separately for $B^+$ and $B^0$~tags, and for electron and muon
192: events. For each of these four sub-samples, the values of $\langle
193: M_X\rangle$ and $\langle M^2_X\rangle$ are obtained by calculating the
194: mean values of the unfolded $M_X$ and $M^2_X$~distributions. The
195: results obtained by averaging the four sub-sample results are shown in
196: Fig.~\ref{fig:3} and Tables~\ref{tab:1} and \ref{tab:2}. All results
197: are preliminary. This procedure has been tested on MC simulated events
198: for the full minimum lepton momentum range and no significant bias has
199: been found.
200: \begin{figure}
201: \begin{center}
202: \includegraphics{fig3.eps}
203: \end{center}
204: \caption{The first and second hadronic mass moments, $\langle
205: M_X\rangle$ and $\langle M^2_X\rangle$, for different lepton threshold
206: momenta. The error bars indicate the statistical and total errors.
207: Note that the individual moments are highly correlated. All
208: results are preliminary.}
209: \label{fig:3}
210: \end{figure}
211: \begin{table}
212: \begin{center}
213: \begin{tabular}{c@{\extracolsep{.3cm}}cccc}
214: \hline \hline
215: \rule{0pt}{2.7ex}$p^*_{min}$ & $\langle M_X\rangle$ & Detector/
216: & Unfolding & $X_c$~model\\
217: \rule[-1.3ex]{0pt}{1.3ex}(GeV/$c$) & (GeV/$c^2$) & background &
218: &\\
219: \hline
220: \rule{0pt}{2.7ex}0.9 & $2.076\pm 0.009\pm 0.010$ & 0.006 & 0.008
221: & 0.003\\
222: 1.0 & $2.073\pm 0.009\pm 0.010$ & 0.005 & 0.008 & 0.003\\
223: 1.1 & $2.066\pm 0.008\pm 0.008$ & 0.005 & 0.005 & 0.003\\
224: 1.2 & $2.066\pm 0.008\pm 0.006$ & 0.004 & 0.004 & 0.003\\
225: 1.3 & $2.063\pm 0.007\pm 0.006$ & 0.004 & 0.005 & 0.003\\
226: 1.4 & $2.056\pm 0.007\pm 0.006$ & 0.003 & 0.004 & 0.003\\
227: 1.5 & $2.048\pm 0.008\pm 0.005$ & 0.003 & 0.003 & 0.002\\
228: \rule[-1.3ex]{0pt}{1.3ex}1.6 & $2.042\pm 0.008\pm 0.004$ & 0.003
229: & 0.001 & 0.002\\
230: \hline \hline
231: \end{tabular}
232: \end{center}
233: \caption{The first hadronic mass moment $\langle M_X\rangle$ for
234: different lepton threshold momenta. The first error on $\langle
235: M_X\rangle$ is statistical and the second is total systematic
236: uncertainty, {\it i.e.}, the quadratic sum of the right-most three
237: columns. All results are preliminary.}
238: \label{tab:1}
239: \end{table}
240: \begin{table}
241: \begin{center}
242: \begin{tabular}{c@{\extracolsep{.3cm}}cccc}
243: \hline \hline
244: \rule{0pt}{2.7ex}$p^*_{min}$ & $\langle M^2_X\rangle$ & Detector/
245: & Unfolding & $X_c$~model\\
246: \rule[-1.3ex]{0pt}{1.3ex}(GeV/$c$) & (GeV$^2$/$c^4$) &
247: background & &\\
248: \hline
249: \rule{0pt}{2.7ex}0.9 & $4.334\pm 0.046\pm 0.051$ & 0.041 & 0.027
250: & 0.014\\
251: 1.0 & $4.312\pm 0.042\pm 0.049$ & 0.037 & 0.030 & 0.013\\
252: 1.1 & $4.285\pm 0.039\pm 0.037$ & 0.032 & 0.017 & 0.012\\
253: 1.2 & $4.258\pm 0.036\pm 0.036$ & 0.027 & 0.021 & 0.012\\
254: 1.3 & $4.247\pm 0.034\pm 0.030$ & 0.023 & 0.016 & 0.011\\
255: 1.4 & $4.214\pm 0.032\pm 0.028$ & 0.024 & 0.009 & 0.011\\
256: 1.5 & $4.166\pm 0.033\pm 0.035$ & 0.022 & 0.025 & 0.011\\
257: \rule[-1.3ex]{0pt}{1.3ex}1.6 & $4.141\pm 0.034\pm 0.027$ & 0.022
258: & 0.010 & 0.013\\
259: \hline \hline
260: \end{tabular}
261: \end{center}
262: \caption{The second hadronic mass moment $\langle M^2_X\rangle$ for
263: different lepton threshold momenta. The first error on $\langle
264: M^2_X\rangle$ is statistical and the second is total systematic
265: uncertainty, {\it i.e.}, the quadratic sum of the right-most three
266: columns. All results are preliminary.}
267: \label{tab:2}
268: \end{table}
269:
270: We consider three sources of systematic error, shown separately in
271: columns three to five of Tables~\ref{tab:1} and \ref{tab:2}:
272: uncertainty related to detector modeling and residual background
273: subtraction, uncertainty related to the unfolding procedure, and
274: uncertainty related to the $X_c$~model in the MC
275: simulation. The total systematic error quoted on the moments
276: corresponds to the quadratic sum of these three components.
277:
278: The uncertainty related to the detector modeling and to the background
279: subtraction is estimated from the consistency between the four
280: sub-samples, by varying the selections used in the analysis (namely the
281: $B_\mathrm{tag}$~signal region and the requirement on
282: $M^2_\mathrm{miss}$), and by varying the background normalization. The
283: uncertainty related to the unfolding procedure is estimated by
284: changing the number of eigenvalue components in the unfolding
285: solution. The $X_c$~model uncertainty is obtained by varying the
286: fractions of $B\to D^*\ell\nu$, $B\to D\ell\nu$ and $B\to
287: D^{**}/D^{(*)}\pi\ell\nu$ within $\pm 10\%$, $\pm 10\%$ and $\pm
288: 30\%$, respectively, and summing the individual variations in
289: quadrature. These ranges of variation roughly correspond to the
290: experimental uncertainties~\cite{ref:18} in the isospin averaged
291: branching ratios.
292:
293: In summary, we have performed a measurement of the first and second
294: moment of the hadronic mass distribution $M_X$ in $B\to
295: X_c\ell\nu$~decays for minimum lepton momenta ranging from 0.9 to
296: 1.6~GeV/$c$ in the c.m.\ frame. The results listed in
297: Tables~\ref{tab:1} and \ref{tab:2} are compatible with recent values
298: from other experiments~\cite{ref:20,ref:21}. It is expected that this
299: measurement (combined with measurements of the semileptonic branching
300: fraction and of the moments of the lepton spectrum in $B\to
301: X_c\ell\nu$) will lead to an improved determination of the CKM matrix
302: element $|V_{cb}|$.
303:
304: %\section*{Acknowledgments}
305: %***** Acknowledgments *****
306: % Please paste this acknowledgement into your latex file.
307: %----------- Long version, for most papers -----------
308: We thank the KEKB group for the excellent operation of the
309: accelerator, the KEK Cryogenics group for the efficient
310: operation of the solenoid, and the KEK computer group and
311: the National Institute of Informatics for valuable computing
312: and Super-SINET network support. We acknowledge support from
313: the Ministry of Education, Culture, Sports, Science, and
314: Technology of Japan and the Japan Society for the Promotion
315: of Science; the Australian Research Council and the
316: Australian Department of Education, Science and Training;
317: the National Science Foundation of China under contract
318: No.~10175071; the Department of Science and Technology of
319: India; the BK21 program of the Ministry of Education of
320: Korea and the CHEP SRC program of the Korea Science and
321: Engineering Foundation; the Polish State Committee for
322: Scientific Research under contract No.~2P03B 01324; the
323: Ministry of Science and Technology of the Russian
324: Federation; the Ministry of Education, Science and Sport of
325: the Republic of Slovenia; the National Science Council and
326: the Ministry of Education of Taiwan; and the U.S.\
327: Department of Energy.
328:
329: %-------- Short version, if necessary, for PRL -----------
330: % currently commented out
331: %We thank the KEKB group for the excellent operation of the
332: %accelerator, the KEK Cryogenics group for the efficient
333: %operation of the solenoid, and the KEK computer group and
334: %the NII for valuable computing and Super-SINET network
335: %support. We acknowledge support from MEXT and JSPS (Japan);
336: %ARC and DEST (Australia); NSFC (contract No.~10175071,
337: %China); DST (India); the BK21 program of MOEHRD and the CHEP
338: %SRC program of KOSEF (Korea); KBN (contract No.~2P03B 01324,
339: %Poland); MIST (Russia); MESS (Slovenia); NSC and MOE
340: %(Taiwan); and DOE (USA).
341:
342: \begin{thebibliography}{99}
343: \bibitem{ref:1} P.\ Gambino and N.\ Uraltsev, Eur.\ Phys.\ J.\ {\bf C
344: 34}, 181 (2004).
345: \bibitem{ref:2} A.\ Falk and M.\ Luke, Phys.\ Rev. {\bf D 57}, 424
346: (1998).
347: \bibitem{ref:3} A.\ Falk, M.\ Luke and M.\ Savage, Phys.\ Rev.\ {\bf D
348: 53}, 2491 (1996); {\it ibid}, {\bf 53}, 6316 (1996).
349: \bibitem{ref:4} M.\ Kobayashi and T.\ Maskawa, Prog.\ Theor.\ Phys.\
350: {\bf 49}, 652 (1973).
351: \bibitem{ref:5} A.\ Abashian {\it et al.} (Belle Collaboration), Nucl.\
352: Instr.\ and Meth.\ {\bf A 479}, 117 (2002).
353: \bibitem{ref:6} S.\ Kurokawa and E.\ Kikutani, Nucl.\ Instr.\ and
354: Meth.\ {\bf A 499}, 1 (2003), and other papers included in this
355: volume.
356: \bibitem{ref:7} R.\ Brun {\it et al.}, GEANT 3.21, CERN Report
357: DD/EE/84-1 (1984).
358: \bibitem{ref:8} J.\ Duboscq {\it et al.} (CLEO Collaboration), Phys.\ Rev.\
359: Lett.\ {\bf 76}, 3898 (1996).
360: \bibitem{ref:9} N.\ Isgur and D.\ Scora, Phys.\ Rev.\ {\bf D 52}, 2783
361: (1995). See also N.\ Isgur {\it et al.}, Phys.\ Rev.\ {\bf D 39}, 799 (1989).
362: \bibitem{ref:10} J.L.\ Goity and W.\ Roberts, Phys.\ Rev. {\bf D 51},
363: 3459 (1995).
364: \bibitem{ref:11} The selection of hadronic events is described in K.\
365: Abe {\it et al.} (Belle Collaboration), Phys.\ Rev.\ {\bf D 64}, 072001
366: (2001).
367: \bibitem{ref:12} Throughout this paper, the inclusion of the charge
368: conjugate mode is implied.
369: \bibitem{ref:13} Throughout this paper, quantities calculated in the
370: c.m.\ frame are denoted by an asterisk.
371: \bibitem{ref:14} H.\ Albrecht {\it et al.} (ARGUS Collaboration), Phys.\
372: Lett.\ {\bf B 241}, 278 (1990).
373: \bibitem{ref:15} K.\ Hanagaki {\it et al.}, Nucl.\ Instr.\ and Meth.\
374: {\bf A 485}, 490 (2002).
375: \bibitem{ref:16} A.\ Abashian {\it et al.}, Nucl.\ Instr.\ and Meth.\
376: {\bf A 491}, 69 (2002).
377: \bibitem{ref:17} A.\ H\"ocker and V.\ Kartvelishvili, Nucl.\ Instr.\
378: Meth.\ {\bf A372}, 469 (1996).
379: \bibitem{ref:18} S.\ Eidelman {\it et al.}, Phys.\ Lett.\ {\bf B592},
380: 1 (2004).
381: \bibitem{ref:19} M.\ Battaglia {\it et al.} (DELPHI Collaboration), DELPHI
382: 2003-028 CONF 648.
383: \bibitem{ref:20} S.E.\ Csorna {\it et al.} (CLEO Collaboration), Phys.\
384: Rev.\ {\bf D 70}, 032002 (2004).
385: \bibitem{ref:21} B.\ Aubert {\it et al.} (BABAR Collaboration), Phys.\
386: Rev.\ {\bf D 69}, 111103 (2004).
387: \end{thebibliography}
388: