1: % This file is part of the APS files in the REVTeX 4 distribution.
2: % Version 4.0 of REVTeX, August 2001
3:
4: % Copyright (c) 2001 The American Physical Society.
5: %
6: \documentclass[twocolumn,showpacs,preprintnumbers,superscriptaddress,amsmath,amssymb,prl,aps,floatfix]{revtex4}
7: %\documentclass[12pt,onecolumn,showpacs,preprintnumbers,superscriptaddress,amsmath,amssymb,prl,aps]{revtex4}
8:
9: \usepackage{graphicx}% Include figure files
10: \usepackage{dcolumn}% Align table columns on decimal point
11: \usepackage{bm}% bold math
12:
13:
14: % These commands comes from Joe (PRL charm x-section paper)
15: % OK % for the onecolumn internal note, must be commented out for the final PRL format
16: %\pagestyle{plain}
17: %\pagenumbering{arabic}
18:
19: \topmargin 0pt
20: \headheight 22pt
21: \footskip 40pt
22: \textheight 21cm
23: \textwidth 17.2cm
24: \oddsidemargin -10mm
25: \evensidemargin 0mm
26:
27: %==============================================================================
28: %\topmargin -0.5in
29:
30:
31: % The preamble begins here.
32:
33: \def\urllink#1#2{\relax#1}
34:
35: % End of preamble and beginning of text.
36: \begin{document}
37: % aliases
38: \newcommand{\Done}{$D_1 (2420)^0$}
39: \newcommand{\D}{$D_J^{(*)}$}
40: \newcommand{\functionOne}{$A(1+B* \cos^2 \theta$}
41: \newcommand{\functionTwo}{$A \sin^2 \theta$}
42: \newcommand{\Dtwo}{$D_2 (2460)^0$}
43: \newcommand{\Dstp}{$D^{*+} \pi^-_{decay}$}
44: \newcommand{\ra}{\rightarrow}
45: \newcommand{\dstdeckpi}{$D^{*+} \rightarrow D^0 \pi^+_{slow}, ~D^0 \rightarrow K^- \pi^+$}
46:
47: % cuts and numbers ( change here !!! )
48: \newcommand{\chicut}{16.}
49: \newcommand{\lxycut}{3.}
50: \newcommand{\resolution}{$6~GeV/c^2$}
51:
52: \newcommand{\dnol}{$D^0$}
53: \newcommand{\dd}{$D_J^{(*)}$}
54: \newcommand{\ddchain}{$D_J^{(*)} \rightarrow D^{*+} \pi_{dec}^-, ~D^{*+} \rightarrow D^0 \pi^+_{slow}, ~D^0 \rightarrow \
55: K^- \pi^+$}
56: \newcommand{\ipb}{\ensuremath{\mathrm{pb^{-1}}}}
57: \newcommand{\ifb}{\ensuremath{\mathrm{fb^{-1}}}}
58: \newcommand{\TeV}{\ensuremath{\matqq hrm{Te\kern -0.1em V}}}
59: \newcommand{\TeVc}{\ensuremath{\mathrm{Te\kern -0.1em V\!/}c}}
60: \newcommand{\TeVcc}{\ensuremath{\mathrm{Te\kern -0.1em V\!/}c^2}}
61: \newcommand{\GeV}{\ensuremath{\mathrm{Ge\kern -0.1em V}}}
62: \newcommand{\GeVc}{\ensuremath{\mathrm{Ge\kern -0.1em V\!/}c}}
63: \newcommand{\GeVcc}{\ensuremath{\mathrm{Ge\kern -0.1em V\!/}c^2}}
64: \newcommand{\MeV}{\ensuremath{\mathrm{Me\kern -0.1em V}}}
65: \newcommand{\MeVc}{\ensuremath{\mathrm{Me\kern -0.1em V\!/}c}}
66: \newcommand{\MeVcc}{\ensuremath{\mathrm{Me\kern -0.1em V\!/}c^2}}
67: \newcommand{\Tesla}{\ensuremath{\mathrm{T}}}
68: \newcommand{\cm}{\ensuremath{\mathrm{cm}}}
69: \newcommand{\um}{\ensuremath{\mathrm{\mu m}}}
70: \newcommand{\ps}{\ensuremath{\mathrm{ps}}}
71: \newcommand{\cdfii}{CDF\,II~}
72: \newcommand{\ppbar}{p\overline{p}}
73: \newcommand{\at}{\symbol{64}}
74: \newcommand{\sye}[1]{\ensuremath{~\pm #1}}
75: \newcommand{\ase}[2]{\ensuremath{^{~+ #1}_{~- #2}}}
76: \newcommand{\asi}[2]{\ensuremath{^{~- #1}_{~+ #2}}}
77:
78: \newcommand{\myto}{\kern -0.3em\to\kern -0.2em}
79: \newcommand{\Lxy}{L_{\rm xy}}
80: \newcommand{\BR}{{\mathrm {BR}}}
81:
82:
83: \title{
84: %{\raggedleft\small
85: %CDF/PHYS/BOTTOM/CDFR/7855\\
86: %Draft 2.0 \\
87: %}\vspace*{0.9cm}
88: Measurement of mass and width of the excited charmed meson states
89: $D_1^0$ and $D_2^{*0}$ at CDF}
90: \input{July_2005_Authors}
91: %\author{CDF\,Collaboration}
92: \date{\today}
93:
94: \begin{abstract}
95: We report on precision measurements of the masses and widths of the narrow, orbitally excited states
96: $D_1^0$ and $D_2^{*0}$ using the CDF II detector
97: at the Fermilab Tevatron. Both states (collectively called $D^{**}$)
98: are reconstructed in the decay channel $D^{**}\to D^{*+} \pi^-$. The $D_2^{*0}$ is also reconstructed in the $D^{**}\to D^{+} \pi^-$
99: channel. Using a dataset with an integrated luminosity of 210 $\mathrm{pb}^{-1}$, the measured
100: masses and widths for the $D_1^0$ are $2421.7\pm 0.7 \pm 0.6 $ MeV/$c^2$ and
101: $20.0\pm 1.7\pm 1.3$ MeV/$c^2$ respectively,
102: while for the $D_2^{*0}$ they are
103: $2463.3\pm 0.6\pm 0.8$ MeV/$c^2$ and $49.2\pm 2.3\pm 1.2$ MeV/$c^2$.
104: These values are currently the single best measurements available.
105: % and they are
106: %consistent with previous experiments as well as predictions from lattice QCD.
107:
108: \end{abstract}
109:
110: \pacs{14.40.Lb,12.40.Yx}
111:
112: \maketitle
113:
114: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
115:
116: The orbitally-excited charmed meson states, collectively referred to as
117: $D^{**}$, are P-wave excitations of the quark-antiquark system involving
118: one charm and one light quark. If we write the total angular momentum as
119: $\vec{J}=\vec{j}_q+\vec{s}_Q; \vec{j}_q=\vec{s}_q+\vec{L}$,
120: where $\vec{L}$ is the orbital angular momentum and $Q (q)$ denotes the
121: charm (light) quark, then in the heavy-quark limit
122: $m_Q \gg \Lambda_{\mathrm{QCD}}$
123: the spin of the charm quark $\vec{s}_Q$ decouples from the other
124: degrees of freedom. In that limit, the four P-wave states can be
125: separated into mass degenerate pairs:
126: $j_q=1/2$ ($J^P= 0^+, 1^+$) and $j_q =3/2$ ($J^P = 1^+, 2^+$).
127: Heavy quark symmetry \cite{hqet} provides a systematic treatment of the
128: $m_Q$-dependent ``hyperfine'' splittings within each doublet, as well
129: as the average mass splittings between doublets, as illustrated
130: in Figure~\ref{fig:massord}.
131: \begin{figure}[tbh]\centering
132: \hspace{-1cm}
133: %\epsfig{file=spectroscopy.eps,width=0.5\linewidth}
134: \includegraphics[width=0.75\linewidth]{spectroscopy.eps}
135: \caption{\label{fig:massord}The mass ordering of the L=0 and L=1 D-meson
136: system. The arrows show the ``hyperfine'' splitting of the P-wave $j_q=1/2$ and $j_q =3/2$ states.
137: Properties of the broad P-wave $j_q=1/2$ states are not well established.
138: }
139: \end{figure}
140:
141: The $D_1^0$ and $D_2^{*0}$ states, which form the $j_q=3/2$ doublet,
142: are expected to decay into final states $D^*\pi$, with an overall
143: D-wave configuration, whereas the $j_q=1/2$ states $D_0^{*0}$ and
144: $D_1'{0'}$, are expected to decay in an overall S-wave
145: configuration. Thus the $j_q=3/2$ states are expected to have narrow decay
146: widths, comparable to their mass splitting \cite{falkpesk}, while the
147: $j_q=1/2$ states are expected to be much broader.
148: Recent theory estimations give the mass values of the two $j_q=3/2$ states
149: \cite{kala,dipier,lewis,ebert,isgur,godfrey}.
150: %The overall picture of orbital excitations of the D meson is summarized
151: %in table \ref{tab:dst}.
152: %\begin{table}[tbh]
153: %\begin{center}
154: %\begin{tabular}{|c|c|c|c|c|c|c|}\hline
155: %$j_q\pm s_Q$&$S^P$&Name&$^{2S+1}L_J$&Mass (MeV)&$\Gamma$ (MeV)&Decay\\ \hline
156: %$1/2\pm1/2$&$1^+$&$D_1^0^*$&$^1P_1$&2430&$\geq$250&$D^*\pi$ S-wave\\
157: %$1/2\pm1/2$&$0^+$&$D_0^*$&$^1P_0$&2360&$\geq$170&$D\pi$ S-wave\\
158: %$3/2\pm1/2$&$2^+$&$D_2^{*0}$&$^1P_2$&2460&$\sim$ 20&$D^*\pi$ D-wave\\
159: %$3/2\pm1/2$&$1^+$&$D_1^0$&$^1P_1$&2420&$\sim$ 20&$D^{(*)}\pi$ D-wave\\ \hline
160: %\end{tabular}
161: %\end{center}
162: %\caption{\label{tab:dst}Summary table of P-wave (L=1) charmed mesons. indications for
163: %masses and widths are only indicative}
164:
165: These narrow orbitally-excited charmed mesons have been observed
166: by several experiments \cite{argus,e687,focus,cleo,cleo90,e691,belle,d0}. In principle,
167: given the large charm cross section at the Tevatron $p\bar{p}$
168: collider, very high statistics samples can be
169: collected for precision measurements of the properties of these states.
170: However, at the trigger level it is difficult to separate low-mass fully
171: hadronic D-meson decays
172: from the overwhelming QCD background.
173: The CDF II detector overcomes this obstacle with a novel two-track trigger,
174: which selects long-lived charged hadrons from secondary vertices, thus
175: suppressing prompt charged hadrons from the QCD background.
176: The dataset for this analysis, based on the vertex trigger, was taken
177: between March 2001 to November 2003,
178: corresponding to an integrated luminosity of $210~\mathrm{pb^{-1}}$.
179: %After a couple of years of data taking, CDF-II has most likely
180: %collected the largest sample of $D^0$ candidates in the world.
181: This sample is ideal for studies of charm-particle decays into two or
182: more hadrons. In this analysis, both the $D_1^0$ and $D_2^{*0}$
183: states have been observed in the $D^{*+} \pi^-$ channel, followed by the
184: decay $D^{*+}\to D^0\pi^+$ and $D^0\to K^-\pi^+$. The
185: $D_2^{*0}$ state has also been observed in the $D^+\pi^-$
186: channel (followed by $D^{+}\to K^- \pi^+ \pi^+$), where
187: the corresponding $D_1^0$ decay is forbidden by parity and
188: angular momentum conservation. Charge-conjugate decay modes are included in the analysis.
189:
190: A description of CDF can be found in Ref \cite{cdf};
191: here only the pertinent detector components are described.
192: This measurement uses
193: tracks measured in the pseudorapidity
194: range $|\eta|<1.1$ \cite{cdfframe}, reconstructed by a silicon microstrip vertex detector
195: (SVXII) \cite{svxii} and the Central Outer Tracker (COT) \cite{cot}, both in a 1.4 T solenoidal magnetic field.
196: The SVXII consists of double-sided sensors
197: arranged in five cylindrical layers at radii between 2.5 and
198: 10.6 cm, each providing an $r-\phi$
199: position measurement with a precision of $\approx 10~\mu$m.
200: %An Intermediate Silicon Layer \cite{isl}, not used in this analysis,
201: %provides 20 $\mu$m precision between
202: %20 and 28 cm of radius, helping to connect tracks from silicon and from the
203: %wire chamber.
204: The COT is an open-cell drift chamber with 96 layers of sense wires,
205: grouped into 8 super-layers of alternating axial and $2^\circ$
206: stereo readout, and
207: providing track measurements between 40 and 137 cm in radius.
208:
209: % Lifted from B->phiphi paper
210: %The $D^0\to K^+\pi^-$ decay is a major component of the two-track
211: %trigger sample, selected by the three-level trigger system.
212: % of the \cdfii detector.
213: %with a three-level trigger system.
214: CDF collects events with a three-level trigger system.
215: At Level~1, two oppositely-charged tracks are reconstructed in the COT
216: by the eXtremely Fast Tracker (XFT)~\cite{xft} and
217: are required to have a transverse momentum $p_T\ge2\,\GeVc$ each, and
218: $p_{T1}+p_{T2}\ge5.5\,\GeVc$. At
219: Level~2, the Silicon Vertex Tracker (SVT)~\cite{svt} associates SVX\,II
220: $r$-$\phi$ position measurements with XFT tracks, providing
221: a precise measurement of the track impact parameter ($d_0$),
222: the distance of closest approach of the track trajectory to
223: the beam axis in the transverse plane.
224: %The resolution of this impact parameter measurement is $50\,\mu\rm m$, which
225: %includes a $\sim$$30\rm\,\mu m$ contribution from the
226: %transverse beam size.
227: Decays of long-lived particles are identified by requiring
228: two tracks with $120\,\mu{\rm m}\le d_0 \le1.0\rm\,mm$,
229: an opening angle between the two tracks satisfying
230: $2^\circ\le|\Delta\phi|\le90^\circ$,
231: and $\Lxy
232: >200\rm\,\mu m$, where $\Lxy$ is the transverse distance from the beam axis to the two track
233: intersection projected along the total transverse momentum of the
234: track pair. A complete event reconstruction of $D^0\to K^-\pi^+$
235: is performed at Level 3, where the
236: Level 1 and Level 2 trigger requirements are confirmed.
237:
238: %%
239: %The trigger system reconstructs at Level 1 charged tracks in the COT by a
240: %hardware processor known as the eXtremely Fast Treacker (XFT) \cite{xft}.
241: %At Level 2, the
242: %Silicon Vertex Tracker (SVT) \cite{svt} associates SVXII hits and XFT
243: %tracks to reconstruct the impact
244: %parameter with a 50 $\mu$m resolution (including a 30 $\mu$m contribution from
245: %the transverse beam size). The ability of reconstructing on-line the impact
246: %parameter of the tracks improves significantly the physics reach of CDF,
247: %because it allows triggering on the hadronic
248: %decays of heavy flavours. The data set used for this analysis requires two
249: %oppositely charged SVT tracks each with momentum transverse to the beam axis
250: %$p_T > 2$ GeV/c and impact parameter $100 \mu m\le |d_0| \le 1.0$ mm.
251: %The pair is required to have an opening angle between $20^\circ$
252: %and $135^\circ$,
253: %$p_{T1}+p_{T2}\ge$ 5.5 GeV/c and be consistent with coming from a vertex with a
254: %transverse decay length $> 200 \mu$m and impact parameter $<140 \mu$m.
255: %At Level 3, a PC farm fully reconstructs the event with offline algorithms,
256: %improving the precision and purity.
257: The narrow width and small mass difference between the $D^{**}$ states
258: require stringent tracking calibration to achieve
259: adequate mass resolution. The calibration procedure is done in two steps.
260: The first step is to determine the error matrix for the COT track parameters,
261: accounting for multiple scattering inside the COT volume.
262: These uncertainties depend on the COT material description and
263: hit resolution in the drift model. The error matrix is computed using
264: simulated $J/\psi \rightarrow \mu^+\mu^-$ decays, by analyzing pull
265: distributions of the five helix parameters between generated and
266: reconstructed muon tracks. Maximum likelihood fits are performed in bins of
267: $p_T^2$ of the $J/\psi$ to derive the $p_T$-dependent rescaling factors.
268: The second step is similar to that
269: used in B mass measurements at CDF \cite{korn}, but with improved energy loss
270: calculations. We first use photon conversions to electron-positron
271: pairs to determine
272: the radiation length distribution of the inner detector material and then
273: tune the energy loss parameters
274: using the reconstructed $J/\psi$ mass as a calibration.
275: %In contrast to \cite{korn}, the detailed GEANT-based \cite{geant} fitting algorithm is replaced with
276: %a much faster Kalman fitting technique, which yields equivalent results while providing the necessary
277: %increase in speed to handle the vastly larger data volume.
278: %First, the position of the SVX bias cables is adjusted by comparing the
279: %radius of photon conversion between data and Monte Carlo simulation (MC).
280: %which determines the $1/X_0$ distribution of the material in four
281: %different $z$ regions of the p-pbar vertex. Then,
282: %to correct for the dE/dx energy loss and multiple scattering the tracks are refit
283: %using the Kalman fitter.
284: For this tuning we describe the detector as a series of cylindrical
285: layers of different materials;
286: in each layer we calculate the average dE/dx energy loss
287: in the traversed medium and then refit the track parameters.
288: The calibration uses the
289: $J/\psi \rightarrow \mu^+\mu^-$ decays from data to iteratively adjust the material composition
290: and thickness of each layer for different regions in $z$,
291: until the $J/\psi$ mass dependence on the transverse momentum is negligible.
292: Finally, the absolute mass scale is reached
293: by adjusting the magnetic field by 0.1\% to set the
294: value of the $J/\psi$ mass to the world average value \cite{pdg}.
295: Figure \ref{fig:massjpsi}a shows the $J/\psi$ mass versus
296: $p_T$ before and after calibration, where the slope changes from $0.50 \pm 0.03$ MeV/$c^2$ per GeV/$c$
297: to $0.006\pm 0.010$ MeV/$c^2$ per GeV/$c$ and the mean corresponds to
298: the world average $J/\psi$ mass.
299: \begin{figure}[tb]\centering
300: %\epsfig{file=jpsimass_kal_newgeo.eps,bb= 0 0 568 386,width=0.5\linewidth,clip}
301: %\includegraphics[bb= 0 0 568 386,width=0.5\linewidth,clip]{jpsimass_kal_newgeo.eps}
302: \includegraphics[width=0.49\linewidth]{massjpsi.eps}
303: \includegraphics[width=0.49\linewidth]{masd0.eps}
304: \caption{\label{fig:massjpsi} a) The $J/\psi$ mass as a function of
305: $p_T$, before (dashed) and after (solid) tracking calibration. The slope is reduced from 0.50$\pm$
306: 0.03 to $0.006\pm 0.010$ MeV/$c^2$ per GeV/$c$. \\b) The $D^0$ mass as a function of
307: $p_T$, after calibration. The dashed lines indicate the world average
308: 1$\sigma$ error band of $\sim 1~\mathrm{MeV}/c^2$. }
309: \end{figure}
310:
311: The tracking calibration has been cross-checked with several other particle
312: samples which populate different mass ranges or different decay topologies,
313: such as $K_S^0$, $\Upsilon(1S)$, $B^\pm$ (to check for charge asymmetry biases)
314: and $D^0$. For all cases the masses agree well with the world averages,
315: and the charge asymmetry was
316: found to be negligible.
317: The results of these tests are summarized in Table \ref{tab:xchecks}. Figure \ref{fig:massjpsi}b illustrates the results for the $D^0$,
318: of particular relevance for this analysis.
319: \begin{table}[tb]
320: \centering
321: % Use mass difference table?
322: %\hspace{0pt}\begin{tabular}{|l|l|r|r|r|r|}\hline
323: %Meson&Decay&Slope&Measured mass&PDG mass&Mass\\
324: %&mode&(MeV/c$^2$/GeV/c)&(MeV/$c^2$)&(MeV/$c^2$)&pull\\
325: % \hline
326: %$K^0_s$&$\pi^+\pi^-$&-0.07$\pm$0.16&497.63$\pm$0.04$\pm$0.07&497.672$\pm$0.031&-0.48\\
327: %$B^\pm$&$J/\psi K^\pm$&0.11$\pm$0.06&5278.73$\pm$0.58$\pm$0.7&5279$\pm$0.5&0.70\\
328: %$J/\psi$&$\mu^+\mu^-$&-0.006$\pm$0.01&3096.87$\pm$0.2$\pm$0.4&3096.87$\pm$0.04&0.\\
329: %$\Upsilon(1S)$&$\mu^+\mu^-$&-0.16$\pm$0.21&9459.6$\pm$1.1$\pm$1.5&9460.30$\pm$0.26&-0.4\\
330: %$D^0$&$K^\pm\pi^\mp$&$(-9 \pm 6)\times 10^{-6}$&1864.7$\pm$0.1$\pm$0.2&1864.6$\pm$0.5&0.2\\
331: %\hline
332: %\hspace{0pt}
333: \begin{tabular}{|l|l|r|r|r|r|}\hline
334: %Meson&Decay&Slope&$m - m_{\mathrm{PDG}}$ &Mass\\
335: %&mode&(MeV/c$^2$/GeV/c)&(MeV/$c^2$)&pull\\
336: % \hline
337: %$K^0_s$ & $\pi^+\pi^-$ & -0.07$\pm$0.16 & -0.042$\pm$0.086 & -0.48\\
338: %$B^\pm$ & $J/\psi K^\pm$ & 0.11$\pm$0.06 & -0.27$\pm$1.037 & -0.26\\
339: %$J/\psi$ & $\mu^+\mu^-$ & -0.006$\pm$0.01 & 0$\pm$0.45 & 0.\\
340: %$\Upsilon(1S)$ & $\mu^+\mu^-$ & -0.16$\pm$0.21 & -0.7$\pm$1.9 &-0.37\\
341: %$D^0$ & $K^\pm\pi^\mp$ & $(-9 \pm 6)\times 10^{-6}$& 0.1$\pm$0.55 &0.19\\
342: Meson&Decay&Slope&$m - m_{\mathrm{PDG}}$ \\
343: &mode&(MeV/$c^2$/GeV/$c$)&(MeV/$c^2$) \\
344: \hline
345: $K^0_s$ & $\pi^+\pi^-$ & -0.07$\pm$0.16 & -0.042$\pm$0.086 \\
346: $B^\pm$ & $J/\psi K^\pm$ & 0.11$\pm$0.06 & -0.27$\pm$1.037 \\
347: $J/\psi$ & $\mu^+\mu^-$ & -0.006$\pm$0.01 & 0$\pm$0.45 \\
348: $\Upsilon(1S)$ & $\mu^+\mu^-$ & -0.16$\pm$0.21 & -0.7$\pm$1.9 \\
349: $D^0$ & $K^-\pi^+$ & $(-9 \pm 6)\times 10^{-6}$& 0.1$\pm$0.55 \\
350: \hline
351: \end{tabular}
352: \caption{\label{tab:xchecks}Summary of mass measurements used to validate
353: the tracker calibration}
354: \end{table}
355: %\begin{figure}[tbh]\centering
356: %\epsfig{file=masd0.eps,width=0.5\linewidth}
357: %\includegraphics[width=0.5\linewidth]{masd0.eps}
358: %\epsfig{file=d0dstar_5.eps,width=\linewidth}
359: %\caption{\label{fig:d0dstar}Behavior of the mass of $D^0$ as a function of
360: %$p_T$, after the detector material calibration. T$he horizontal lines indicate the
361: %1$\sigma$ error band from the PDG \cite{pdg}}
362: %\end{figure}
363: %par
364:
365: Events considered in this analysis are required to have at least four tracks
366: with total charge zero, out of which two are consistent
367: with the trigger requirements ($p_T>2$ GeV/$c$ and $|d_0|>100 \mu$m).
368: For the first channel ($D^{**}\to D^{*+} \pi^-$, followed by
369: $D^{*+}\to D^0\pi^+$ and $D^0\to K^-\pi^+$),
370: these ``trigger tracks''
371: are required to have an invariant mass within 24 MeV/$c^2$ of the
372: $D^0$ mass of 1864.5 MeV/$c^2$ \cite{pdg}. Particles are assigned in turn the
373: $K$ and $\pi$ masses, and all $K \pi$ combinations falling
374: within the mass window are retained, to avoid any bias in the $D^0$ mass.
375: This $D^0$ candidate is associated with another track with $p_T > 400$ MeV/$c$
376: to form a $D^{*+}$
377: candidate, with the requirement that the mass difference between the $D^{*+}$ and the
378: $D^0$ be smaller than 147 MeV/$c^2$.
379: In addition, to reduce both background and misassignment of the $K$ and
380: $\pi$ masses, the third track is
381: required to have the same charge as the candidate pion track, as
382: expected from leading order $D^{*+}$ decays.
383: Finally, the $D^{*+}$ is associated
384: with a negative track with $p_T > 400$ MeV/$c$, to form a four-track
385: $D^{**}$ candidate.
386: For the second channel ($D_2^{*0} \to D^+\pi^-$, followed by $D^{+}\to K^- \pi^+ \pi^+$), no mass window is required for the
387: trigger tracks. The $D^+$
388: candidate is constructed from three tracks with a mass and charge
389: assignment compatible with being a $K^- \pi^+\pi^+$ system; two of
390: these tracks must satisfy the trigger requirements, while the third must have
391: $p_T > 800$ MeV/$c$. We first ensure that all three tracks originate
392: from a common vertex which is well separated from the primary vertex, by
393: requiring the $\chi^2$ from a 3-dimensional fit to the three
394: tracks to be smaller than 12, and have an associated $L_{xy} > 1$ mm.
395: Then, $D^+$ candidates are defined as three-track systems with invariant mass between 1.85 and 1.89 GeV/$c^2$.
396: Finally, to obtain a neutral $D^{**}$ candidate, the three tracks are combined
397: with a fourth, of opposite charge with respect to the sum of the first three.
398: The two $D^{**}$ resonances are analyzed in terms of the invariant mass
399: difference between the four-track and the three-track system, which
400: is crucial to separate the resonances from the background.
401: Figure
402: \ref{fig:d12mass} and \ref{fig:d2mass} show the results for the
403: $D^{**}$ resonances in the two decay channels.
404: \begin{figure}[tbh]\centering
405: %\epsfig{file=D12.eps,width=0.5\linewidth}
406: \includegraphics[width=0.85\linewidth]{D12.eps}
407: \caption{\label{fig:d12mass} Invariant mass difference between $D^{*}\pi$
408: system and the $D^{*}$. The points represent the data, and lines represent the projection of the fit results for the individual components described in the text.}
409: \end{figure}
410: \begin{figure}[tbh]\centering
411: %\epsfig{file=D2dpi.eps,width=0.5\linewidth}
412: \includegraphics[width=0.85\linewidth]{D2dpi.eps}
413: \caption{\label{fig:d2mass} Invariant mass difference between the $D^+\pi^-$ system and the $D^+$, with the same definitions as Figure \ref{fig:d12mass}.}
414: \end{figure}
415:
416: In Figure \ref{fig:d12mass}, signals from the $D_1^0$ at 0.40 GeV/$c^2$
417: and the $D_2^{*0}$ at 0.46 GeV/$c^2$ are
418: clearly visible above the combinatorial background.
419: Due to the unknown shape of the combinatorial background it is not possible
420: with these data to derive with sufficient confidence the properties, or even the
421: existence of the broad state (BS) $D_1^{0'}$ (the $1^+$ component of
422: the $j_q=1/2$ doublet) from our data.
423: Since a state with width around 200 MeV/$c^2$ is suggested by
424: a Belle measurement \cite{belle} its potential effect is included as
425: a systematic error.
426:
427: The distribution of the mass difference in Figure \ref{fig:d2mass} presents
428: two features: a broad peak around 0.4 GeV/$c^2$, and
429: a narrow pronounced peak around 0.6 GeV/$c^2$. The broad peak is due to
430: the feed-down decays of the type $D^{**}\to
431: D^{*+}\pi^-$ accessible for both $D_1^0$ and $D_2^{*0}$, followed by $D^{*+}\to D^+ \pi^0$, where the
432: $\pi^0$ is not observed. Similar to the $D_1^{0'}$ in Figure
433: \ref{fig:d12mass}, part of this peak could
434: be due to a broad neutral $D^{*0}_0$ state with a mass around
435: 2410 MeV/$c^2$ and width of about 250 MeV/$c^2$, as reported by FOCUS \cite{focus}.
436: The peak around 0.6 GeV/$c^2$ is the main $D_2^{*0}$ signal.
437:
438: %Most of the information for the $D_1^0$ comes from the large peak at 0.4 GeV/c$^2$
439: %in Figure \ref{fig:d12mass}, while most of the information about the $D_2^{*0}$
440: %comes from the large peak in figure \ref{fig:d2mass}; however, for both histograms
441: %(especially the first) some information about the other state is present.
442: %So the best strategy is to perform a simultaneous fit to both of them.
443: A simultaneous binned likelihood fit to both histograms in Figures \ref{fig:d12mass} and
444: \ref{fig:d2mass} is used to extract the mass and width for the $D_1^0$ and $D_2^{*0}$.
445: The likelihood function consists of:
446: \begin{itemize}
447: \item[i) ] a signal term for each narrow state in the $D^{*+}$ channel, and only
448: a $D_2^{*0}$ term in the $D^+$ channel. These terms are a convolution of a
449: non-relativistic Breit-Wigner (BW) distribution with a resolution histogram taken from
450: Monte Carlo simulation, and depends on the amplitudes $A_{1,2}$, the mass
451: differences with respect to the three-track states
452: $M_{1,2}$ and widths $\Gamma_{1,2}$. The $D_2^{*0}$ widths are common for the
453: two channels. The likelihood component is:
454: \[f_{D_{1,2}}(\Delta m) = A_{1,2}\times BW(\Delta m, M_{1,2}, \Gamma_{1,2})\otimes Resolution. \]
455: The $D_2^{*0}$ masses in the two channels are connected
456: by using the world average value for the mass difference between $D^{*+}$ and $D^*$.
457: \item[ii) ] a background (BG) term for each histogram with free
458: parameters $\alpha$, $\beta$, $\gamma$ and $\delta$ (set to zero for the first channel), of the form:
459: \[f_{BG}(\Delta m) = \alpha (\Delta m-m_\pi)^\beta
460: e^{-\gamma(\Delta m-m_\pi)}+\delta.\]
461: \item[iii) ] a broad state (BS) term for each histogram, modeled as
462: a Breit-Wigner function similar to that of the narrow resonances, but convoluted with a
463: Gaussian whose width is taken from simulation:
464: \[f_{BS}(\Delta m) = A_{BS} \times BW(\Delta m, M_{BS},
465: \Gamma_{BS})\otimes Resolution.\]
466: \item[iv) ] a feed-down (FD) term for decays of $D_1^0$ and $D_2^{*0}$ to $D \pi^0
467: \pi$, where the $\pi^0$ is lost and a smaller mass is reconstructed. The
468: only additional free parameter is a common scale factor relative to the
469: amplitudes of the $D^{*}\pi$ channel, since the $D_1^0$ and
470: $D_2^{*0}$ masses and widths are the same. The shift and resolution
471: are taken from Monte Carlo.
472:
473: \end{itemize}
474: Overall we have 7 parameters for the narrow resonances (amplitude, mass and
475: width for each each resonance, plus the $D_2^{*0}$ yield in the second channel), 7 for background
476: modeling (one parameter is set to zero in the first channel), two amplitudes for
477: broad states and one feed-down normalization, for a total of 17 parameters. If the mass
478: and width of one or two broad states are left floating, we can have up to 19 or
479: 21 parameters.
480: In order to properly describe the tails of the resolution,
481: avoiding detector biases which occur when a Gaussian
482: approximation is made, we use the explicit resolution histogram
483: for signal and feed-down. The detector resolution is about 4
484: MeV/$c^2$, much smaller than the intrinsic width of the
485: resonances, but the tails are significantly larger.
486: On the other hand, Gaussian resolution is sufficient for the broad states,
487: since the tails fall outside the fitting window.
488: The accuracy of the simulation of the detector resolution has been tested in
489: a control sample of $D^0$ and $D^{*+}$, whose natural width is negligible
490: with respect to detector resolution. After taking the detector resolution
491: into account using the above
492: prescription, we measure a width smaller than 0.2 MeV/$c^2$ for all momentum
493: ranges, and use this value as a conservative systematic
494: error on the width due to tracking precision. If instead, the Gaussian
495: approximation was made for the width, we then observed width and mass shifts of the order
496: of a few MeV$/c^2$, much larger than the precision envisaged.
497:
498: The fitting procedure has been tested on a Monte Carlo sample of fully
499: simulated $D_1^0\to D^{*+}\pi^-$, $D_2^{*0}\to D^{*+}\pi^-$ and $D_2^{*0}\to D^+\pi^-$
500: decays with three times the statistics of the observed data. The feed-downs
501: are also included. However, they have a small impact on the final result.
502: Both background and broad-state events
503: have been generated from the parametrized data distribution. With
504: this sample,
505: two important cross-checks on the fitting algorithm have been performed.
506: First, the full simulation sample is fitted and the results agree with the
507: simulation input parameters. In addition, a large number of ``toy
508: experiments'' with statistics comparable to the data are generated
509: and fit to look for potential biases. The resultant pull distribution
510: for each fit parameter is consistent with a Gaussian distribution of zero
511: mean and unit width.
512:
513: The likelihood fit applied to the data
514: yields roughly 7500 $D_1^0$ and 5000 $D_2^{*0}$ candidates for the channel
515: of Figure \ref{fig:d12mass} and
516: 20000 $D_2^{*0}$ candidates for that of Figure \ref{fig:d2mass}.
517: % and measured mass differences and widths of:
518: %\begin{eqnarray}
519: %\nonumber M(D_1^0)-M(D^{*+})& = & 411.7\pm 0.7~ \mathrm{MeV/}c^2, \\
520: %\nonumber \Gamma(D_1^0) & = & 20.0\pm 1.7~\mathrm{MeV/}c^2, \\
521: %\nonumber M(D_2^{*0})-M(D^{+}) & = & 593.9\pm 0.6~ \mathrm{MeV/}c^2, \\
522: %\nonumber \Gamma(D_2^{*0}) & = & 49.2\pm 2.3~ \mathrm{MeV/}c^2.
523: %\end{eqnarray}
524: Given the large background of unknown shape, there is no way to derive the
525: presence of a broad state from our data only. Leaving all its parameters
526: free, the resulting broad state yield is
527: small and the $\chi^2$ quality of fit probability is 51\%. Three additional hypotheses on the broad
528: state are tested: a mass and width as measured by FOCUS \cite{focus}, as measured by
529: Belle \cite{belle}, or omitting the BS term from the fit.
530: The resulting variations in the measured quantities are shown in Table \ref{tab:bs}.
531: \begin{table}[tbh]
532: \centering
533: \hspace{0pt}\begin{tabular}{|l|l|l|l|l|l|}\hline
534: Model & $\Delta (M_{D_1^0})$ & $\Delta(\Gamma_{D_1^0})$ & $\Delta (M_{D_2^{*0}})$&$\Delta(\Gamma_{D_2^{*0}})$&$P(\chi^2)$\\
535: % & $ (\MeVcc)$ & $(\MeVcc)$ & $(\MeVcc)$&$(\MeVcc)$ & \%\\
536: \hline
537: Free&---&---&---&---&51\\
538: FOCUS&0.1&0.2&0.1&0.5&50\\
539: Belle&0.1&0.3&0.1&0.5&50\\
540: Absent&0.1&0.4&0.1&0.6&52\\
541: \hline
542: \end{tabular}
543: \caption{\label{tab:bs} Shift of the measured masses and widths of
544: $D_1^0$ and $D_2^{*0}$ and $\chi^2$ probability P for different
545: hypotheses of the broad state; allowing parameters to float in the
546: fit, using the values published by FOCUS
547: \cite{focus} or Belle \cite{belle},
548: or forcing it to zero in the $D \pi$ channel. Mass shifts are in
549: $\MeVcc$ and probability in \%.}
550: \end{table}
551: As the differences are small and there is no discrimination on the
552: basis of the $\chi^2$ probability, the central result includes the
553: broad state with free parameters, and other options
554: provide estimates of systematic uncertainty.
555: \begin{table}[h]
556: \begin{center}
557: \begin{tabular}{|l|c|c|c|c|}
558: \hline%
559: %Source & $\Delta M_1$ &$\Delta \Gamma_1$ &$\Delta M_2$ &$\Delta \Gamma_2$ \\
560: Source & $\Delta (M_{D_1^0})$ & $\Delta(\Gamma_{D_1^0})$ & $\Delta (M_{D_2^{*0}})$&$\Delta(\Gamma_{D_2^{*0}})$\\
561: % & $(\MeVcc)$&$(\MeVcc)$& $(\MeVcc)$ &$(\MeVcc)$\\
562: \hline
563: MC statistics& 0.3&1.2&0.4&1.2\\ \hline
564: Broad State& 0.1&0.4&0.1&0.5\\
565: Track Error scale & 0.1&---&0.1&---\\
566: Fit model & $<0.1$ & $<0.1$ & $<0.1$ & $<0.1$\\
567: Mass Calibration & 0.1&0.2&0.1&0.2\\ \hline
568: Total (Relative) & 0.4 & 1.3 & 0.5&1.3\\ \hline
569: Reference mass & 0.5 & --- &0.7 & ---\\ \hline
570: Total (Absolute) & 0.6&1.3&0.8&1.3\\ \hline
571: \end{tabular}
572: \end{center}
573: \caption{\label{tab:syst}Summary of systematic error for the mass and width of the two
574: resonances, all in units of $\MeVcc$. Since the measurement is a mass difference,
575: the absolute mass error includes a systematic error
576: due to the uncertainty of the mass of the reference
577: particle. }
578: \end{table}
579:
580: From the likelihood fit we derive results, with statistical uncertainties only,
581: on the $D_1^0$ and $D_2^{*0}$ widths and mass differences with respect to the $D^{*+}$ or $D^+$ into which the $D^{**}$ decays.
582: To derive absolute masses from the measured mass differences
583: we add the world average masses \cite{pdg} of the $D^{*+}$
584: or $D^+$, and the corresponding uncertainty on this value (both known
585: with 0.5 MeV/$c^2$ precision) to the systematic errors.
586: The remaining residual systematic errors are connected with Monte Carlo statistics
587: and the tracking calibration, and are listed in Table \ref{tab:syst}.
588:
589: In summary, the large sample of orbitally-excited charmed mesons collected by
590: the CDF collaboration
591: using the vertex trigger allows the measurement of properties of
592: the L=1 orbitally-excited narrow states $D_1^0$ and $D_2^{*0}$ with unprecedented
593: precision.
594: Using both $D^{*+}\pi^-$ and $D^+\pi^-$ final states, the measured widths of
595: these states are:
596: \begin{eqnarray}
597: \nonumber \Gamma(D_1^0) &=& 20.0\pm 1.7 \pm 1.3 ~\mathrm{MeV/}c^2, \\
598: \nonumber \Gamma(D_2^{*0}) &=& 49.2\pm 2.3\pm 1.3 ~\mathrm{MeV/}c^2.
599: \end{eqnarray}
600: where the first error is statistical, and the second systematic, above
601: and for all results which follow. In order to improve the
602: resolution, the mass measurement is performed in term of
603: differences with respect to the daughter
604: charmed particle. The measured mass differences are:
605: \begin{eqnarray}
606: \nonumber M(D_1^0)-M(D^{*+}) &=& 411.7\pm 0.7\pm 0.4 ~\mathrm{MeV/}c^2, \\
607: \nonumber M(D_2^{*0})-M(D^{+}) &=& 593.9\pm 0.6\pm 0.5 ~\mathrm{MeV/}c^2.
608: \end{eqnarray}
609: The $D_2^{*0}$ mass is given with respect to the $D^+$ since almost all the
610: information comes from this decay channel.
611: % for this reason, varying the
612: %$D^+$-$D^{*+}$ mass difference within the errors given by the PDG has a negligible
613: %effect on the result.
614: Adding the world average values for the $D^{*+}$ and $D^+$ masses and including their
615: uncertainties in quadrature to the systematic error, the absolute values
616: of the masses are:
617: \begin{eqnarray}
618: \nonumber M(D_1^0) &=& 2421.7\pm 0.7\pm 0.6~\mathrm{MeV/}c^2,\\
619: \nonumber M(D_2^{*0}) &=& 2463.3\pm 0.6\pm 0.8~\mathrm{MeV/}c^2.
620: \end{eqnarray}
621: A comparison of these values with the world averages and some recent theoretical models can be
622: found in Figure \ref{fig:theory}; the results are consistent with the models, and with
623: better estimations of uncertainties from the models they should allow one to
624: constrain which theoretical picture is consistent with experiment.
625: Comparison with predictions and measurements for the widths is less insightful due
626: to the larger uncertainties; however, the width measurements presented
627: here are consistent with other recent measurements \cite{belle,cleo} .
628: \begin{figure}[tbh]\centering
629: %\epsfig{file=D2dpi.eps,width=0.5\linewidth}
630: \includegraphics[width=0.85\linewidth]{finresult.eps}
631: \caption{\label{fig:theory} $D_1^0$ (solid) and $D_2^{*0}$ (open) mass
632: comparison with theoretical expectations (triangles) and previous
633: measurements and current world average (circles). Errors for theory predictions
634: are not readily available.}
635: \end{figure}
636:
637: This is the best single world measurement of the
638: masses of the orbitally-excited charm states. The
639: total error is still limited by the statistics, and there is room for
640: improvement with the data presently being recorded by the CDF II experiment.
641: A precise determination of the theory uncertainties would however be needed
642: to assess the discriminating power of this and future measurements.
643:
644: We thank the Fermilab staff and the technical staffs of the participating
645: institutions for their vital contributions. This work was supported by the
646: U.S. Department of Energy and National Science Foundation; the Italian
647: Istituto Nazionale di Fisica Nucleare; the Ministry of Education, Culture,
648: Sports, Science and Technology of Japan; the Natural Sciences and Engineering
649: Research Council of Canada; the National Science Council of the Republic of
650: China; the Swiss National Science Foundation; the A.P. Sloan Foundation; the
651: Bundesministerium f\"ur Bildung und Forschung, Germany; the Korean Science
652: and Engineering Foundation and the Korean Research Foundation; the Particle
653: Physics and Astronomy Research Council and the Royal Society, UK; the Russian
654: Foundation for Basic Research; the Comisi\'on Interministerial de Ciencia y
655: Tecnolog\'{\i}a, Spain; in part by the European Community's Human Potential
656: Programme under contract HPRN-CT-2002-00292; and the Academy of Finland.
657:
658: %\begin{table}[tbh]
659: %\begin{center}
660: %\begin{tabular}{|c|c|c|}\hline
661: %Reference&$M(D_1^0)$&$M(D_2^{*0})$\\ \hline
662: %CDF 2005&$2421.7\pm 0.7\pm 0.6$&$2463.3\pm 0.6\pm 0.8$\\
663: %PDG 2004 \cite{pdg}&$2422.2\pm 1.8$&$2458.9\pm 2.0$\\
664: %Kalashnikova et al \cite{kala}&2428&2445\\
665: %Di Pierro et al. \cite{dipier}&2417&2460\\
666: %Lewis et al. \cite{lewis}&2405&2445\\
667: %Ebert et al. \cite{ebert}&2414&2459\\
668: %Isgur \cite{isgur}&2415&2460\\
669: %Godfrey et al. \cite{godfrey}&2440&2500\\ \hline
670: %\end{tabular}
671: %\caption{\label{tab:theory}Comparison of this measurement with PDG and
672: %recent theoretical calculations}
673: %\end{center}
674: %\end{table}
675: %
676:
677: \begin{thebibliography}{00}
678: \bibitem{hqet} N.~Isgur and M.~Wise, Phys. Lett. B \textbf{232}, 113 (1989) and Phys. Lett. B {\bf 237}, 527 (1990).
679: \bibitem{falkpesk} A.~Falk and M.~Peskin, Phys. Rev. D \textbf{49}, 3320 (1994).
680: \bibitem{kala} Y.S.~Kalashnikova and A.V.~Nefediev, Phys. Lett. B \textbf{530}, 117 (2002).
681: \bibitem{dipier} M.~Di Pierro and E.~Eichten, Phys. Rev. D \textbf{64}, 114004 (2001).
682: \bibitem{lewis} R.~Lewis and R.M.~Woloshyn, Phys. Rev. D \textbf{62}, 114507 (2000).
683: \bibitem{ebert} D.~Ebert, V.O.~Galkin and R.N.~Faustov, Phys. Rev. D \textbf{57}, 5663 (1998)
684: and AIP Conf.Proc.619:336-345, 2002 hep-ph/0110190.
685: \bibitem{isgur} N.~Isgur, Phys. Rev. D \textbf{57}, 4041 (1998).
686: \bibitem{godfrey}S.~Godfrey and N.~Isgur, Phys. Rev. D \textbf{32}, 189
687: (1985) and S.~Godfrey and R.~Kokoski, Phys. Rev. D \textbf{43}, 1679 (1991).
688: \bibitem{argus} H.~Albrecht {\it et al.} (ARGUS Collaboration), Phys. Lett. B \textbf{232}, 398 (1989).
689: \bibitem{e687} P.L.~Frabetti {\it et al.} (E687 Collaboration), Phys. Rev. Lett. {\bf 72}, 324 (1994).
690: \bibitem{focus} J.~Link {\it et al.} (FOCUS Collaboration), Phys. Lett. B \textbf{586}, 11 (2004) hep-ex/0312060.
691: \bibitem{cleo} P.~Avery {\it et al.} (CLEO Collaboration), Phys. Lett. B {\bf 331}, 236 (1994).
692: \bibitem{cleo90} P.~Avery {\it et al.} (CLEO Collaboration), Phys. Rev. D \textbf{41}, 774 (1990).
693: \bibitem{e691} J.C.~Anjos {\it et al.} (E691 Collaboration), Plys. Rev. Lett. {\bf 62}, 1717 (1989).
694: \bibitem{belle} K.~Abe {\it et al.} (BELLE Collaboration), Phys. Rev. D \textbf{69}, 112202 (2004).
695: \bibitem{d0} V.M.~Abazov {\it et al.} (D0 Collaboration), Phys. Rev. Lett 95 161602 (2005)
696: \bibitem{cdf} D.~Acosta {\it et al.} (CDF Collaboration), FERMILAB-PUB-96/390-E, 1996, and Phys. Rev. D \textbf{68}, 072004 (2003).
697: \bibitem{cdfframe} CDF uses a cylindrical coordinate system ($r,\phi,z$) with the
698: origin at the center of the detector and the
699: positive $z$ direction aligned with the proton direction.
700: Psuedorapidity $\eta$ is defined as $\eta \equiv -\ln(\tan(\theta/2))$, where $\theta$ is the polar angle.
701: \bibitem{svxii} A.~Sill {\it et al.}, Nucl. Instrum. Methods A {\bf 447} 1 (2000).
702: \bibitem{cot} T.~Affolder {\it et al.}, Nucl. Instrum. Methods A {\bf 526}, 249 (2004).
703: %\bibitem{isl} T.~Affolder {\it et al.} Nucl. Instrum. Methods, {\bf A 453} 84 (2000).
704: \bibitem{xft} E.J.~Thomson {\it et al.}, IEEE Trans. Nucl. Sci. {\bf 49} 1063 (2002).
705: \bibitem{svt} W.~Ashmanskas {\it et al.}, Nucl. Instrum. Methods A {\bf 447}, 218 (2000).
706: \bibitem{korn} D.~Acosta {\it et al.} (CDF Collaboration), Submitted to PRL (hep-ex/0508022).
707: \bibitem{geant}
708: R.~Brun, R.~Hagelberg, M.~Hansroul and J.~C.~Lassalle,
709: %``Geant: Simulation Program For Particle Physics Experiments. User Guide And
710: %Reference Manual,''
711: CERN-DD-78-2-REV.
712: %\bibitem{note1} M.Campanelli, E.Gerchtein, CDF note 6918
713:
714: %\bibitem{notecalib} M.Campanelli, E.Gerchtein, CDF note 6905
715:
716: %\bibitem{falkpesk} A.Falk, M.Peskin Phys. Rev. {\bf D 49} 3320-3332
717: %(1994)
718: \bibitem{pdg} S. Eidelman {\it et al} Phys. Lett. B {\bf 592}, 1 (2004) .
719: %\bibitem{bval} See:
720: %\verb+http://www-cdf.fnal.gov/internal/physics/bottom/validation+
721:
722: %\bibitem{notecalib} M.Campanelli, E.Gerchtein, CDF note 6905
723:
724: %\bibitem{g3x} G.Bauer, A.Korn, Ch.Paus, ``Update on calibration of Energy
725: %Loss and Magneticn Field using $J/ \psi$ Events in Run II'', CDF6355.
726:
727: %\bibitem{d0finder} S.Giagu et al., CDF Note 6158
728: %\bibitem{prmeas}
729: %Albrecht {\it et al.} ARGUS Coll. Phys. Lett. {\bf B 232} 398 (1989)\\
730: %Frabetti {\it et al.} E687 Coll. Phys. Rev. Lett. {\bf 72} 324 (1994) \\
731: %Avery {\it et al.} CLEO Coll. Phys. Lett. {\bf B 331} 236 (1994)
732:
733:
734: \end{thebibliography}
735:
736: \end{document}
737:
738: % LocalWords: GeV dec pb fb hrm ps CDF xy
739: