1: \documentclass{PoS}
2: \newcommand{\msbar}{$\overline{\mbox{MS}}$ }
3:
4:
5:
6: \title{B-hadron production in top quark decay}
7:
8: \ShortTitle{B-hadron production in top quark decay}
9:
10:
11: \author{\speaker{Gennaro Corcella}\vspace{0.1cm}\\
12: Dipartimento di Fisica\\
13: Universit\`a di Roma `La Sapienza', Italy\\
14: P.le A.Moro 2, I-00185 Roma, Italy\\
15: E-mail: \email{Gennaro.Corcella@roma1.infn.it}
16: \begin{flushright}
17: \vspace{-8.cm}
18: \rm{ROME1/1424/06}\\
19: \rm{February 2006}
20: \vspace{6.5cm}
21: \end{flushright}}
22:
23: \author{Volker Drollinger\\
24: Dipartimento di Fisica Galileo Galilei\\
25: Universit\`a di Padova and INFN, Sezione di Padova\\
26: Via Marzolo 8, I-35131 Padova, Italy\\
27: E-mail: \email{Volker.Drollinger@cern.ch}}
28:
29: \abstract{We present the energy distribution of
30: $b$-flavoured hadrons in top quark decay using the
31: PYTHIA and HERWIG event generators, which we tune to LEP and SLD data.
32: We find that fitting the string and cluster models
33: is essential to reproduce the $e^+e^-$ data and to reliably
34: predict $B$-hadron production in top decay.
35: We also compare the PYTHIA and HERWIG results with
36: the ones yielded by resummed calculations based on the fragmentation
37: function formalism.}
38:
39: \FullConference{International Workshop on Top Quark Physics\\
40: January 12-15, 2006\\
41: Coimbra, Portugal}
42:
43:
44:
45: \begin{document}
46:
47:
48: Top quark phenomenology is one of the main
49: fields of investigation in theoretical and experimental particle physics.
50: The experiments at the Tevatron accelerator and, in future,
51: at the LHC will allow one to perform improved measurements of
52: the top properties, such as its mass, thanks to the production of
53: large amounts of $t\bar t$ pairs.
54:
55:
56: In this paper we study bottom quark fragmentation in top quark decay
57: ($t\to bW$), which is responsible of one of the largest contribution to
58: the uncertainty on
59: the top mass measurement at the Tevatron \cite{cdf,d0} and the LHC \cite{lhc}.
60: In particular, in the analysis of Ref.~\cite{avto} the top quark mass
61: is determined using at the LHC
62: final states with leptons and $J/\psi$'s, where the leptons
63: come from the $W$ decay $W\to\ell\nu$, and the $J/\psi$'s from the
64: decay of a $b$-flavoured hadron $B$.
65: In \cite{avto} the PYTHIA event generator \cite{pythia} was exploited,
66: and the error on $m_t$ was estimated to be
67: $\Delta m_t\simeq 1$~GeV, with $b$-fragmentation being the largest source
68: of uncertainty. In \cite{cms}, the invariant mass $m_{B\ell}$, yielded by the
69: HERWIG \cite{herwig} event generator, was used to fit $m_t$,
70: and the impact of matrix-element corrections to the simulation of top decay
71: \cite{corsey} was investigated.
72:
73: Bottom quark fragmentation in top decay was studied in \cite{cm,ccm},
74: following the method of perturbative fragmentation functions \cite{mele}.
75: The NLO $b$-quark energy spectrum is expressed as the
76: convolution of a coefficient function, describing the emission of a
77: massless parton, and a perturbative fragmentation function $D(m_b,\mu_F)$,
78: associated with the transition of a massless parton into a massive $b$.
79: $D(m_b,\mu_F)$ follows the Dokshitzer--Gribov--Lipatov--Altarelli--Parisi
80: (DGLAP) evolution equations \cite{ap,dgl}, which can be solved
81: once an initial condition at a scale $\mu_{0F}$ is given.
82: The initial condition
83: of the perturbative fragmentation function, first computed in
84: \cite{mele}, was proved to be process-independent in \cite{cc}.
85: Solving the DGLAP evolution equations we can resum the large
86: $\ln(m_t^2/m_b^2)$ which appears in the NLO
87: massive $b$-spectrum (collinear resummation).
88: Both the top-decay coefficient function, computed in \cite{cm},
89: and the initial condition
90: $D(m_b,\mu_{0F})$ present terms which become large
91: when the $b$-quark energy fraction
92: $x_b$ approaches 1, which corresponds to soft-gluon radiation.
93: Soft contributions in the initial condition (process independent)
94: and in the coefficient function (process dependent) were resummed in
95: the NLL approximation in \cite{cc} and \cite{ccm}, respectively.
96: In order to predict the spectrum of $b$-flavoured hadrons, perturbative
97: calculations need to be supplement by non-perturbative models.
98: In \cite{cm,ccm}, the $B$-hadron spectrum in top decay
99: was presented, after fitting a few hadronization
100: models to SLD \cite{sld} and ALEPH \cite{aleph} data.
101:
102: Following the lines of \cite{cv}, in this paper we would like to
103: address $b$-fragmentation in top decay, using the PYTHIA and HERWIG event
104: generators. As discussed in \cite{cv}, PYTHIA and HERWIG simulate
105: multiple radiation in top decay in the soft or collinear
106: approximation, and are
107: provided with matrix-element corrections \cite{corsey,norrbin}
108: to allow hard and large-angle
109: emission. The hadronization mechanism is simulated by the string
110: model \cite{string} in PYTHIA, and by the cluster model \cite{cluster} in
111: HERWIG.
112:
113: For the sake of a reliable prediction of the $B$-energy distribution
114: in $t\to bW$, we need to use models and parametrizations which are able
115: to describe well the $B$-spectrum at $e^+e^-$ machines.
116: We consider ALEPH \cite{aleph}, OPAL \cite{opal} and
117: SLD \cite{sld} data on the $B$ energy fraction $x_B$ in $Z\to b\bar b$
118: events, where $x_B$ is defined as follows:
119: \begin{equation}
120: x_B={{2p_B\cdot p_Z}\over{m_Z^2}},
121: \end{equation}
122: with $p_Z$ and $p_B$ being the $Z$ and $B$ momenta, respectively.
123:
124: We use the versions HERWIG 6.506 and PYTHIA 6.220, and find that
125: the default parametrizations are unable to fit such data, as one gets
126: $\chi^2/\mathrm{dof}=739.4/61$ (HERWIG) and $\chi^2/\mathrm{dof}=467.9/61$
127: (PYTHIA).
128: As in \cite{cv}, we tune the cluster and string models to the data, while
129: we leave unchanged the parameters of HERWIG and PYTHIA which are related to
130: the perturbative phase of the parton showers. Our best fits are summarized in
131: Table~\ref{tab1}: for PYTHIA we are able to find a parametrization which
132: is able to reproduce well the data ($\chi^2/\mathrm{dof}=45.7/61$);
133: HERWIG, even after the fit, is still
134: not able to describe the $x_B$-spectrum very well, but the comparison
135: is anyway much better than with the default parameters
136: ($\chi^2/\mathrm{dof}=222.4/61$). We have also checked that
137: the parametrizations in Table~\ref{tab1} work well for the new model
138: implemented in PYTHIA 6.3 \cite{p63}.
139: \begin{table}[t]
140: \caption{\label{tab1} Parameters of HERWIG and PYTHIA
141: hadronization models tuned to
142: $e^+e^-$ data, along with the $\chi^2$ per degree of freedom.}
143: \begin{center}
144: \begin{tabular}{|c|c|}\hline
145: HERWIG & PYTHIA \\
146: \hline\hline
147: CLSMR(1) = 0.4 & \\
148: \hline
149: CLSMR(2) = 0.3 & PARJ(41) = 0.85 \\
150: \hline
151: DECWT = 0.7 & PARJ(42) = 1.03 \\
152: \hline
153: CLPOW = 2.1 & PARJ(46) = 0.85 \\
154: \hline
155: PSPLT(2) = 0.33 & \\
156: \hline
157: \hline
158: $\chi^2/\mathrm{dof}$ = 222.4/61 & $\chi^2/\mathrm{dof}$ = 45.7/61 \\
159: \hline
160: \end{tabular}
161: \end{center}
162: \end{table}
163: The HERWIG and PYTHIA spectra, before and after the fit, along with the
164: experimental data, are presented in Figures~\ref{eeh} and \ref{eep}.
165: For the sake of comparison, we also show the $x_B$-spectrum yielded by the
166: NLO+NLL calculation of Ref.~\cite{cc}, convoluted with the Kartvelishvili
167: hadronization model \cite{kart}:
168: \begin{equation}
169: D^{\mathrm{np}}(x;\gamma)=(1+\gamma)(2+\gamma) (1-x) x^\gamma.
170: \label{kk}
171: \end{equation}
172: We fit the model (\ref{kk}) to the data in the range
173: $0.18\leq x_B\leq 0.94$, to avoid the regions at small and large $x_B$,
174: where, as pointed out in \cite{cc}, the resummed calculation yields
175: a negative distribution.
176: Setting
177: $m_Z=91.118$~GeV, $m_b=$~5 GeV and $\Lambda^{(5)}_{\overline{\mathrm{MS}}}
178: =200$~GeV in the perturbative calculation, we get $\gamma=17.178\pm 0.303$
179: From Figs.~\ref{eeh} and \ref{eep} we learn that
180: default HERWIG and PYTHIA are far from the data. After the
181: tuning, PYTHIA reproduces the data quite well, while HERWIG yields a broader
182: distribution; the resummed calculation is consistent with the data.
183: and $\chi^2/\mathrm{dof}=46.2/53$ from the fit.
184:
185: \begin{figure}[ht!]
186: \centerline{\resizebox{0.68\textwidth}{!}{\includegraphics{hee.ps}}}
187: \caption{Data from LEP and SLD experiments, compared with the NLO+NLL
188: calculation convoluted with the Kartvelishvili model (solid)
189: and HERWIG 6.506, using the default parametrization (dashed)
190: and our tuning (dotted).}
191: \label{eeh}
192: \end{figure}
193: \begin{figure}[ht!]
194: \centerline{\resizebox{0.68\textwidth}{!}{\includegraphics{pyee1.ps}}}
195: \caption{As in Fig.~\protect\ref{eeh}, but comparing data and the NLO+NLL
196: calculation with default (dashed) and tuned (dotted) PYTHIA 6.220.}
197: \label{eep}
198: \end{figure}
199:
200: Using the parametrizations in Table~\ref{tab1}, we can predict the
201: $B$-energy distribution in $t\to bW$, which will be expressed in terms
202: of the variable
203: \begin{equation}
204: x_B={1\over {1-w}}{{2p_B\cdot p_t}\over{m_t^2}},
205: \end{equation}
206: where $p_t$ is the top momentum and $1/(1-w)$ is a normalization factor,
207: with $w=1-m_W^2/m_t^2+m_b^2/m_t^2$.
208: In Fig.~\ref{hertop} we present the $B$-spectrum in top decay according to
209: HERWIG, PYTHIA and the resummed calculation of \cite{cm,ccm}, convoluted
210: with the Kartvelishvili model.
211: \begin{figure}[ht!]
212: \centerline{\resizebox{0.68\textwidth}{!}{\includegraphics{hptt.ps}}}
213: \caption{$B$-hadron spectra in top decay, for $m_t=175$~GeV,
214: according to a NLO+NLL computation convoluted with the Kartvelishvili model
215: (solid line), HERWIG (dashed) and PYTHIA (dotted).}
216: \label{hertop}
217: \end{figure}
218: The comparison exhibited in Fig.~\ref{hertop} shows similar features
219: to Figures~\ref{eeh} and \ref{eep}, and reflects the quality of the
220: fits to the $e^+e^-$ data: PYTHIA reproduces the
221: peak of the resummed calculation rather well, while it is below the NLL
222: prediction at $x_B<0.7$, and
223: above at $x_B>0.9$. HERWIG is below the resummed spectrum in most the
224: $x_B$-range, and above it only at large $x_B$.
225:
226: Finally, we wish to present results in Mellin moment space, making use
227: of the data reported by the DELPHI Collaboration in \cite{delphi}
228: on the moments of the $B$ cross section in $e^+e^-$ processes.
229: From the point of view of resummed calculations,
230: working in moment space presents several advantages:
231: in $N$-space, convolutions become ordinary products and the relation between
232: parton- and hadron-level moments is just
233: $\sigma_N^B=\sigma_N^bD_N^{\mathrm{np}}$,
234: where $D_N^{\mathrm{np}}$ is the
235: non-perturbative fragmentation function in Mellin space. Therefore,
236: there is no
237: need to introduce a functional form for the hadronization model
238: in $x_B$-space. Also, $N$-
239: spectra are well defined, and do not present the problems of the
240: $x_B$-results, which are negative at small or large $x_B$.
241:
242: In Table~\ref{tab2} we compare the DELPHI moments with the ones given
243: by the tuned versions of HERWIG and PYTHIA, and the
244: predictions for top decay. We also quote the results yielded by the NLL
245: calculations of Refs.~\cite{cm,ccm},
246: extracting $D_N^{\mathrm{np}}$ from the data.
247: As for $e^+e^-\to b\bar b$ processes, the moments
248: given by HERWIG and PYTHIA
249: are consistent, within the error ranges, with the DELPHI ones.
250: It is interesting that HERWIG is
251: compatible with the DELPHI moments, even though it was only marginally
252: consistent with LEP and SLD data in $x_B$-space.
253: The results for top decay have similar features to the $x_B$-spectra:
254: PYTHIA is very close to the NLL calculation, which uses
255: $D_N^{\mathrm{np}}$ extracted from the DELPHI data, while
256: HERWIG, whose predictions are shifted toward larger $x_B$,
257: gives larger moments.
258:
259: In summary, we reviewed recent results on $b$-flavoured hadron
260: production in top quark decay. We tuned HERWIG and PYTHIA to LEP and SLD data
261: and presented results on the $B$-hadron spectrum in top decay in both
262: $x_B$ and moment spaces. In fact, fitting the cluster and string model
263: turned out to be necessary to reproduce the $e^+e^-$ data.
264: The results were also compared with resummed
265: calculations, based on the fragmentation function formalism.
266: We think that our analysis and fits can be useful to improve the present
267: understanding of $b$-quark fragmentation in top quark decay.
268: It will be very interesting to investigate how the tuned versions of
269: HERWIG and
270: PYTHIA fare with respect to other observables. For example, it
271: may be worthwhile
272: reconsidering the studies in Refs.~\cite{avto,cms} with the parametrizations
273: which we have proposed, and estimate the contribution of $b$-fragmentation
274: to the uncertainty on the top quark mass reconstruction.
275:
276: \begin{table}[ht!]
277: \caption{\label{tab2}\small Moments
278: $\sigma^B_N$ from
279: DELPHI~\protect\cite{delphi}, and moments
280: in $e^+e^-$ annihilation and top ($t$) decay,
281: using NLL resummed calculations, HERWIG (HW) and PYTHIA (PY).}
282: \small
283: \begin{tabular}{| c | c c c c |}
284: \hline
285: & $\langle x\rangle$ & $\langle x^2\rangle$ & $\langle x^3\rangle$
286: & $\langle x^4\rangle$ \\
287: \hline
288: \hline
289: $e^+e^-$ data $\sigma_N^B$&0.7153$\pm$0.0052 &0.5401$\pm$0.0064 &
290: 0.4236$\pm$0.0065 &0.3406$\pm$0.0064 \\
291: \hline
292: \hline
293: $e^+e^-$ NLL $\sigma_N^b$ & 0.7801 & 0.6436 & 0.5479 & 0.4755 \\
294: \hline
295: $D^{\mathrm{np}}_N$ & 0.9169 & 0.8392 & 0.7731 & 0.7163 \\
296: \hline
297: $e^+e^-$ HW $\sigma_N^B$ & 0.7113 & 0.5354 & 0.4181 & 0.3353 \\
298: \hline
299: $e^+e^-$ PY $\sigma_N^B$ & 0.7162 & 0.5412 & 0.4237 & 0.3400 \\
300: \hline
301: \hline
302: $t$-dec. NLL $\Gamma^b_N$ & 0.7883 & 0.6615 & 0.5735 & 0.5071 \\
303: \hline
304: $t$-dec. NLL $\Gamma^B_N=\Gamma^b_N
305: D_N^{\mathrm{np}}$ & 0.7228 & 0.5551 & 0.4434 & 0.3632 \\
306: \hline
307: $t$-dec. HW $\Gamma^B_N$ & 0.7325 & 0.5703 & 0.4606 & 0.3814 \\
308: \hline
309: $t$-dec. PY $\Gamma^B_N$ & 0.7225 & 0.5588 & 0.4486 & 0.3688 \\
310: \hline
311: \end{tabular}
312: \end{table}
313:
314:
315: \begin{thebibliography}{99}
316: \bibitem{cdf}
317: CDF Collaboration, A. Abulencia et al., Phys. Rev. Lett. 96 (2006) 022004.
318: \bibitem{d0}
319: D0 Collaboration, V.M. Abazov et al., Phys. Lett. B 606 (2005) 25.
320: \bibitem{lhc}
321: M. Beneke, I. Efthymiopoulos, M.L. Mangano, J. Womersley et al.,
322: in Proceedings of 1999 CERN Workshop on
323: Standard Model Physics (and more) at the LHC, CERN 2000-004,
324: G. Altarelli and M.L. Mangano eds., p.~419, hep-ph/0003033.
325: \bibitem{avto}
326: A. Kharchilava, Phys. Lett. B 476 (2000) 73.
327: \bibitem{pythia}
328: T. Sj\"ostrand, L. L\"onnblad and S. Mrenna, hep-ph/0108264.
329: \bibitem{cms}
330: G. Corcella, J.\ Phys.\ G26 (2000) 634;\\
331: G. Corcella, M.L. Mangano and M.H. Seymour, JHEP 0007 (2000) 004.
332: \bibitem{herwig}
333: G. Corcella, I.G. Knowles, G. Marchesini, S. Moretti, K. Odagiri,
334: P. Richardson, M.H. Seymour, B.R. Webber, JHEP 0101 (2001) 010.
335: \bibitem{corsey}
336: G. Corcella and M.H. Seymour, Phys. Lett. B442 (1998) 417.
337: \bibitem{cm}
338: G. Corcella and A.D. Mitov, Nucl. Phys. B623 (2002) 247.
339: \bibitem{ccm}
340: M. Cacciari, G. Corcella and A.D. Mitov, JHEP 0212 (2002) 015.
341: \bibitem{mele}
342: B. Mele and P. Nason, Nucl. Phys. B 361 (1991) 626.
343: \bibitem{cc}
344: M. Cacciari and S. Catani, Nucl. Phys. B617 (2001) 253.
345: \bibitem{ap}
346: G. Altarelli and G. Parisi, Nucl. Phys. B126 (1977) 298.
347: \bibitem{dgl}
348: L.N. Lipatov, Sov. J. Nucl. Phys. 20 (1975) 95;
349: V.N. Gribov and L.N. Lipatov, Sov. J. Nucl. Phys. 15 (1972) 438;
350: Yu.L. Dokshitzer, Sov. Phys. 46 (1977) 641.
351: \bibitem{sld}
352: SLD Collaboration, K. Abe et al., Phys. Rev. Lett. 84 (2000) 4300.
353: \bibitem{aleph}
354: ALEPH Collaboration, A. Heister et al., Phys. Lett. B512 (2001) 30.
355: \bibitem{cv}
356: G. Corcella and V. Drollinger, Nucl. Phys. B730 (2005) 82.
357: \bibitem{norrbin}
358: E. Norrbin and T. Sjostrand, Nucl. Phys. B603 (2001) 297.
359: \bibitem{string}
360: B. Andersson, G. Gustafson, G. Ingelman, T. Sj\"ostrand,
361: Phys. Rept. 97 (1983) 31.
362: \bibitem{cluster}
363: B.R. Webber, Nucl. Phys. B238 (1984) 492.
364: \bibitem{opal}
365: OPAL Collaboration, G. Abbiendi et al., Eur. Phys. J. C29 (2003) 463.
366: \bibitem{p63}
367: T. Sj\"ostrand, L. L\"onnblad, S. Mrenna and P.Z. Skands, hep-ph/0308153.
368: \bibitem{kart}
369: V.G. Kartvelishvili, A.K. Likehoded and V.A. Petrov, Phys. Lett. B78 (1978)
370: 615.
371: \bibitem{delphi}
372: DELPHI Collaboration, G. Barker et al., DELPHI 2002-069, CONF 603.
373: \end{thebibliography}
374: \end{document}
375: