1: \documentclass[aps,showpacs]{revtex4}
2: \usepackage{psfig}
3:
4: \def\ep{\epsilon}
5: \def\order#1{{\cal O}\!\left(#1\right)}
6: \newcommand{\ba}{\begin{eqnarray}}
7: \newcommand{\ea}{\end{eqnarray}}
8: \newcommand{\be}{\begin{equation}}
9: \newcommand{\ee}{\end{equation}}
10: \newcommand{\mm}{\omega}
11: \def\eq#1{(\ref{#1})}
12:
13:
14: \begin{document}
15:
16: \title{
17: %
18: %preprint number:
19: %
20: \[ \vspace{-2cm} \]
21: \noindent\hfill\hbox{\rm } \vskip 1pt \noindent\hfill\hbox{\rm
22: Alberta Thy 04-04}
23: \vskip 10pt
24: %
25: Heavy-to-light decays with a two-loop accuracy
26: }
27:
28: \author{Ian Blokland, Andrzej Czarnecki, and Maciej \'Slusarczyk}
29: \affiliation{
30: Department of Physics, University of Alberta\\
31: Edmonton, AB\ \ T6G 2J1, Canada}
32:
33: \author{Fyodor Tkachov}
34: \affiliation{
35: Institute for Nuclear Research, Russian Academy of Sciences\\
36: Moscow, 117312, Russian Federation}
37:
38:
39:
40: \begin{abstract}
41: We present a determination of a new class of three-loop Feynman
42: diagrams describing heavy-to-light transitions. We apply it to
43: find the $\order{\alpha_s^2}$ corrections to the top quark decay
44: $t\to bW$ and to the distribution of lepton invariant mass in the
45: semileptonic $b$ quark decay $b\to ul\nu$. We also confirm the
46: previously determined total rate of that process as well as the
47: $\order{\alpha^2}$ corrections to the muon lifetime.
48: \end{abstract}
49:
50: \pacs{12.38.Bx,13.35.Bv,14.65.Ha}
51:
52: \maketitle
53:
54: The determination of higher order corrections in perturbative quantum
55: field theory is notoriously difficult, and with the general
56: tendency towards precision measurements in particle physics, each
57: newly-won class of perturbative integrals expands the possibilities
58: for phenomenological analyses. For instance, quantum corrections
59: to decays of neutral particles, such as a virtual photon or a $Z$
60: boson into hadrons, are known to sixth order in perturbation
61: theory, $\order{\alpha_s^3}$, and even some $\order{\alpha_s^4}$
62: effects have been studied. Those results have been very useful in
63: determining a variety of Standard Model parameters such as the
64: $Z$ boson properties, the strong coupling constant, and the
65: running of the electromagnetic coupling constant
66: \cite{Steinhauser:2002rq}.
67:
68: Much less is known about radiative corrections to processes with a
69: charged particle in the initial state. Only relatively recently
70: have first results been obtained in fourth order perturbation
71: theory, $\order{\alpha_s^2}$ and $\order{\alpha^2}$, primarily for
72: total decay rates. The technical challenge in such calculations
73: is the presence of massive propagators. For example, consider the
74: muon decay. Since the muon is charged, it can emit photons, and
75: the resulting amplitudes will involve propagators of a virtual
76: muon. Its mass sets the energy scale of the process and cannot be
77: treated as a small parameter.
78:
79: The presence of massive propagators is an obstacle in evaluating
80: the multi-loop diagrams required by precise measurements of heavy
81: quark and lepton decays. So far, genuine $\order{\alpha_s^2}$
82: corrections to heavy quark decays are known only for semileptonic
83: processes, $Q \to q l\nu(gg)$, and only for some kinematic cases.
84: One approach that has been successful consists in expanding
85: Feynman diagrams around the zero recoil limit: when the quark $q$
86: remains at rest with respect to $Q$. The kinematics of
87: semileptonic decays can be represented by a triangle, since the
88: invariant mass of the leptons together with the mass of the final
89: state quark $q$ cannot exceed the mass of the decaying quark. This
90: is depicted in Fig.~\ref{fig:tri}.
91: \begin{figure}[htb]
92: \hspace*{0mm}\psfig{figure=triangleNew.ps,width=70mm}
93: \caption{Kinematical boundaries of the semileptonic decays $Q\to q+
94: \mbox{leptons}$. The solid arrow shows the expansion presented in
95: this paper. Previously known expansions are indicated with dotted
96: arrows.} \label{fig:tri}
97: \end{figure}
98: The diagonal boundary corresponds to the zero recoil limit, in
99: which the $\order{\alpha_s^2}$ effects are known
100: \cite{zerorecoil,zerorecoilA,Franzkowski:1997vg}. Also shown are
101: the starting points of previously studied expansions. Those
102: results have helped improve the knowledge of the $b$ quark
103: lifetime and the determination of the CKM matrix element $V_{cb}$.
104:
105: However, the expansion around zero recoil converges slowly near
106: the origin in Fig.~\ref{fig:tri}, that is, when both the quark $q$ and
107: the lepton pair are light; in this case computations become
108: prohibitively expensive. Two other approaches have been used in
109: such cases. First, for the phenomenologically important decays
110: $\mu\to e\nu\bar\nu$ and $b\to u l\nu$, the total lifetimes have
111: been determined in \cite{vanRitbergen:1998yd,vanRitbergen:1999gs}
112: by analytically calculating imaginary parts of four-loop diagrams.
113: That heroic effort is difficult to extend, for example, to differential
114: distributions. The other approach consists in expanding diagrams
115: in an artificial parameter, for example the ratio of muon masses
116: outside and inside loops, and using Pad\'e approximants to sum the
117: expansion for the physical value of this parameter. This approach
118: was used to check the muon and $b\to u$ results
119: \cite{Steinhauser:1999bx,Chetyrkin:1999ju}. It yields results
120: scattered around the exact values which are often sufficient for
121: applications. The drawback of this method is that it is very
122: difficult to estimate the errors reliably.
123:
124: The purpose of this study is to extend the method of expansions
125: beyond the zero recoil limit. We start directly at the origin of
126: Fig.~\ref{fig:tri}, which corresponds to the kinematics of a top
127: quark decay into a massless $b$ quark and a massless $W$ boson. We
128: then treat the $W$ mass as a small perturbation and compute
129: several terms of the resulting expansion. We stop when we can
130: smoothly match to the previously obtained expansion around the case
131: of the $W$ boson equally heavy as the decaying quark
132: \cite{Czarnecki:2001cz}. The physical top quark decay corresponds
133: to a specific value of the $W$ mass, but in the more general decay
134: $Q\to q+ \mbox{leptons}$, the term ``$W$ mass'' refers to the
135: invariant mass of the lepton pair, and it is in this context that
136: we can relate $t\rightarrow bW$ to $b\rightarrow ul\nu$ and $\mu
137: \rightarrow e\nu \overline{\nu}$.
138:
139: This is the first time that exact results are available in this
140: limit and we can now address a number of interesting problems. We
141: obtain an accurate value of the $\order{\alpha_s^2}$ correction to
142: the top quark lifetime. The combination of our results with the
143: expansion around the heavy $W$ case allows us to give a complete
144: description of the differential distribution of $b\to ul\nu$ decay
145: in the invariant mass of leptons, and thus improves the
146: theoretical description of this decay, important for the
147: determination of $V_{ub}$. We also check the muon and $b\to u$
148: lifetime corrections with a relative error of about $2\times
149: 10^{-4}$. In the future, the same method can be employed to
150: improve perturbative corrections to mixing processes such us
151: $B_d\leftrightarrow \overline B_d$ and $B_s\leftrightarrow
152: \overline B_s$.
153:
154:
155: In Fig.~\ref{fig:diagrams} we show three examples of the diagrams
156: that we have to consider in order to calculate $t\rightarrow bW$
157: at $\order{\alpha_s^2}$. We use the optical theorem to connect the
158: imaginary parts of such diagrams with contributions to the decay.
159: Note that we customarily speak about two-loop corrections when
160: what we actually need to compute are the imaginary parts of
161: three-loop diagrams. The various cuts correspond to two-loop
162: virtual corrections or emissions of one or two real quanta.
163: \begin{figure}[htb]
164: \begin{tabular}{ccc}
165: \psfig{figure=dia1.ps,width=42mm}
166: &
167: \hspace{5mm}
168: \psfig{figure=dia2.ps,width=52mm}
169: \hspace{5mm}
170: &
171: \psfig{figure=dia3.ps,width=52mm}
172: \\
173: (a) & (b) & (c)
174: \end{tabular}
175: \caption{Examples of diagrams whose cuts contribute to the
176: $\mathcal{O}(\alpha_s^2)$ decay rate for $t\rightarrow bW$: (a)
177: light or heavy quarks; (b) abelian; (c) non-abelian.}
178: \label{fig:diagrams}
179: \end{figure}
180:
181: With $m_b=0$, there are two scales in the problem: $m_t$ and
182: $m_W$. We define an expansion parameter $\mm=m_W^2/m_t^2$ so that
183: the two scales can be expressed as hard and soft ($\order{1}$ and
184: $\order{\sqrt{\mm}}$, using $m_t$ as the unit of energy).
185: Contributions arising from these two scales are identified using
186: asymptotic expansions so that we must consider two regions. In
187: the first region, all the loop momenta are hard and the $W$
188: propagator can be expanded as a series, in powers of $\mm$, of
189: massless propagators. In the second region, the gluon momenta are
190: hard but the loop momentum flowing through the $W$ is soft. In
191: this region, the diagrams factor into a product of a two-loop
192: self-energy type integral and a one-loop vacuum bubble integral
193: with a scale of $m_W$. The leading contribution from this second
194: region is $\order{\mm^2}$, and the interplay between the two
195: regions gives rise to terms with a large logarithm $\ln \mm$.
196:
197: All scalar integrals arising in the problem can be expressed in
198: terms of $9$ basic topologies. We use differential-algebraic
199: identities to reduce all loop integrals in both regions to a
200: combination of $24$ master integrals. The resulting large linear
201: systems can be solved in a few ways. In the traditional method
202: \cite{Tkachov:1981wb}, one inspects the structure of the identities
203: and rearranges them manually into the form of recurrence relations
204: for an efficient iterative solution of the system. This ``by
205: inspection'' method has proven to be very successful in numerous
206: applications (e.g., \cite{bro91a, vanRitbergen:1998yd,
207: vanRitbergen:1999gs, Czarnecki:2001cz}) but it requires much human
208: work to implement. Conversely, a straightforward
209: solution of the linear system is much more expensive computationally and
210: was first achieved only recently
211: \cite{Laporta:2001dd}. In our calculation \cite{csc} we used the traditional approach
212: (programmed in FORM \cite{Vermaseren:2000nd}) as well as a
213: modified version of the new algorithm for which we implemented
214: a dedicated computer algebra system. In both cases we independently
215: obtained identical results which serve as a check of correctness
216: but also enable us to compare these two methods. Details of the
217: implementation of both methods and the evaluation of master
218: integrals will be presented in a forthcoming technical paper.
219:
220: The final result for the top quark decay width can be written as
221: \be \Gamma(t\rightarrow bW) = \Gamma_0 \left[ X_0 +
222: \frac{\alpha_s}{\pi} X_1 + \left( \frac{\alpha_s}{\pi} \right)^2
223: X_2 \right], \qquad \Gamma_0 \equiv \frac{G_F m_t^3 \left| V_{tb}
224: \right|^2}{8\sqrt{2}\pi} \ . \ee Throughout this paper, we use
225: $\alpha_s \equiv \alpha_s^{\overline{MS}}(\mu)$, where $\mu$ is
226: the pole mass of the decaying quark. The tree-level and
227: $\order{\alpha_s}$ coefficients are already known analytically
228: \cite{jk2}, \ba X_0 &=& 1 - 3\mm^2 + 2\mm^3 \ ,
229: \\
230: X_1 & = & C_F \left[ \left( \frac{5}{4} - \frac{\pi^2}{3} \right)
231: + \frac{3}{2}\mm + \mm^2 \left( \pi^2 - 6 + \frac{3}{2}L \right)
232: +\order{\mm^3 L}\right], \qquad L\equiv \ln \mm. \ea The
233: $\order{\alpha_s^2}$ result can be subdivided into four
234: gauge-invariant pieces, \be X_2 = C_F \left( T_R N_L X_L + T_R N_H
235: X_H + C_F X_A + C_A X_{NA} \right) \ , \ee where $C_F=4/3$,
236: $C_A=3$, and $T_R=1/2$ are the usual SU(3) color factors and $N_L$
237: and $N_H$ denote the number of light $(m_q=0)$ and heavy
238: $(m_q=m_t)$ quark species. For the coefficients $X_L$, $X_H$,
239: $X_A$, and $X_{NA}$, we have obtained a series to at least
240: $\mm^5$, of which the leading terms are \ba X_L &\simeq& \left[
241: -\frac{4}{9} + \frac{23\pi^2}{108} + \zeta_3 \right] + \mm \left[
242: -\frac{19}{6} + \frac{2\pi^2}{9} \right] + \mm^2 \left[
243: \frac{745}{72} - \frac{31\pi^2}{36} - 3\zeta_3 - \frac{7}{4}L
244: \right] \ ,
245: \nonumber \\
246: X_H & \simeq & \left[
247: \frac{12991}{1296} - \frac{53\pi^2}{54} - \frac{1}{3}\zeta_3
248: \right] + \mm \left[ -\frac{35}{108} - \frac{4\pi^2}{9} + 4\zeta_3
249: \right] + \mm^2 \left[ -\frac{6377}{432} + \frac{25\pi^2}{18} +
250: \zeta_3 \right] \ ,
251: \nonumber \\
252: X_A & \simeq & \left[ 5 -
253: \frac{119\pi^2}{48} - \frac{53}{8}\zeta_3 + \frac{19}{4}\pi^2\ln2
254: - \frac{11\pi^4}{720} \right] + \mm \left[ -\frac{73}{8} +
255: \frac{41\pi^2}{8} - \frac{41\pi^4}{90} \right] \nonumber \\ \quad
256: & & + \mm^2 \left[ -\frac{7537}{288} + \frac{523\pi^2}{96} +
257: \frac{295}{32}\zeta_3 - \frac{27}{16}\pi^2\ln2 -
258: \frac{191\pi^4}{720} + \left( \frac{115}{48} - \frac{5\pi^2}{16}
259: \right) L \right] \ , \label{xa}
260: \nonumber \\
261: X_{NA} & \simeq & \left[
262: \frac{521}{576} + \frac{505\pi^2}{864} + \frac{9}{16}\zeta_3 -
263: \frac{19}{8}\pi^2\ln2 + \frac{11\pi^4}{1440} \right] + \mm \left[
264: \frac{91}{48} + \frac{329\pi^2}{144} - \frac{13\pi^4}{60} \right]
265: \nonumber \\ \quad & & + \mm^2 \left[ -\frac{12169}{576} +
266: \frac{2171\pi^2}{576} + \frac{377}{64}\zeta_3 +
267: \frac{27}{32}\pi^2\ln2 - \frac{77\pi^4}{288} + \left(
268: \frac{73}{16} - \frac{3\pi^2}{32} \right) L \right] \ . \label{xna}
269: \ea
270:
271: The leading term, $\order{\omega^0}$, of these results can be
272: compared with the numerical estimates obtained with the zero
273: recoil expansions in Eq.~(14) of~\cite{Czarnecki:1998qc}; all of
274: our results agree within their error estimations. Our result can
275: also be compared with a numerical study of the top decay rate
276: obtained by means of Pad\'e approximations up to
277: $\order{\omega^2}$~\cite{Chetyrkin:1999ju}. In many cases we find
278: agreement. However, there are also instances where the numerical
279: estimates in \cite{Chetyrkin:1999ju} differ from our analytic
280: expressions (\ref{xna}) by a few error bar lengths, illustrating
281: limitations of the Pad\'e approximation in this problem. For
282: example, the coefficient of the $\order{\omega}$ term of the
283: nonabelian part $X_{NA}$ of Eq.~(\ref{xna}) is $3.3398$ whereas
284: the value cited in \cite{Chetyrkin:1999ju} reads $3.356(3)$,
285: corresponding to a $5\sigma$ discrepancy. Similarly in $X_{A}$,
286: the $\order{\omega}$ term is off by $3\sigma$.
287:
288: With a sufficient number of terms, the present expansion can be
289: smoothly matched with the one around the $\omega=1$ limit studied
290: previously~\cite{Czarnecki:2001cz} in the context of semileptonic
291: $b$ quark decays. The result of such a matching procedure is
292: depicted in the graphs in Fig.~\ref{fig:matching}. Although strict
293: matching of the two expansions in the entire interval $0 \leq
294: \omega \leq 1$ would require a very large number of terms from
295: each side, a wide overlap region arises
296: even when only a few terms are taken into account.
297: \begin{figure}[htb]
298: \begin{tabular}{cc}
299: \psfig{figure=figL.eps,width=85mm} & \hspace{1mm}
300: \psfig{figure=figH.eps,width=85mm} \hspace{5mm}
301: \\
302: \psfig{figure=figA.eps,width=85mm} & \hspace{5mm}
303: \psfig{figure=figNA.eps,width=85mm} \hspace{5mm}
304:
305: \end{tabular}
306: \caption{Matching of expansions around $\omega = 0$ (thick line)
307: and $\omega = 1$ (thin line). The solid line denotes the resulting
308: decay width valid in the full range of $\omega$. Outside their
309: regions of validity, the expansions are shown as dash-dotted
310: lines.} \label{fig:matching}
311: \end{figure}
312:
313: The most obvious application of the above result is the precise
314: determination of second order QCD corrections to the top quark
315: decay rate. An estimation of this effect is already known, both
316: from numerical studies and from an extrapolation of the zero
317: recoil limit. However, for the measured ratio of $W$ and top
318: masses, $\omega \simeq 0.213$~\cite{PDBook}, the present
319: expansion is the best way to calculate an accurate value of this
320: contribution with a reliable error estimate. Our expansion gives
321: $X_2 = -15.5(1)$ where the uncertainty is almost entirely due to
322: the experimental uncertainty of $m_t$. The theoretical error,
323: which originates from taking a finite number of terms in our
324: expansion, is $20$ times smaller and can be still easily reduced if
325: needed. Using $\alpha_s ( m_t) = 0.11$, we find that the two-loop
326: correction decreases the tree level decay rate by about $2\%$, in
327: agreement with earlier expectations.
328:
329: Our result also provides a check of the total lifetime
330: calculations carried out for $\mu\to e\nu\bar\nu$ and $b\to u
331: l\nu$ decays. In these processes the expansion parameter $\omega$
332: corresponds to the invariant mass of leptons produced in the decay
333: and our matching procedure allows us to obtain a differential
334: width $d \Gamma / d \omega$ valid in the full range of $\omega$
335: with desired accuracy. The inclusive semileptonic decay rate $b\to
336: u l\nu$ can be calculated by integrating over $\omega$ within the
337: kinematical boundaries. Taking $N_L = 4$ and $N_H = 1$, we end up
338: with $\int_0^1 d \omega X_2 (\omega) = -10.644$, which almost
339: perfectly reproduces the $-10.648$ given
340: in~\cite{vanRitbergen:1999gs}. Analogously, the two-photon
341: correction to the muon lifetime emerges from an integration of the
342: abelian contribution $X_A$. We find $\int_0^1 d \omega X_A
343: (\omega) = 1.7797$, which is in excellent agreement with the exact
344: result $1.7794$ \cite{vanRitbergen:1998yd}.
345:
346: To summarize, we have presented a new analytic
347: $\order{\alpha_s^2}$ result for the decay $t\rightarrow bW$ in
348: terms of a parameter $\omega=m_W^2/m_t^2$ and in the limit of
349: $m_b=0$, corresponding to the last remaining kinematic region in
350: which the $\order{\alpha_s^2}$ heavy quark decay rates were not
351: analytically known. This result has enabled us to confirm or
352: modify slightly the corresponding results of previous numerical
353: calculations. Our formulas are readily applicable to other
354: physical processes such as muon decay and the semileptonic
355: $b$ quark decay $b\rightarrow u l\nu$.
356:
357: Our results depend on the imaginary parts of a novel class of
358: three-loop integrals, which we have obtained using two independent
359: paradigms for the solution of large systems of recurrence
360: relations. To the best of our knowledge, this is the first time
361: that both approaches have been used simultaneously to obtain a new
362: result, and an objective analysis of the strengths and weaknesses
363: of each approach will increase the efficiency of other large
364: calculations in the future. This augurs well for the increasingly
365: difficult physical problems that lie ahead. In particular, the
366: top quark decay problem considered here has laid the foundation
367: for $\order{\alpha_s^2}$ perturbative calculations of mixing
368: processes such us $B_d\leftrightarrow \overline B_d$ and
369: $B_s\leftrightarrow \overline B_s$. Since the recently found
370: $\order{\alpha_s}$ effects are large and suffer from strong scale
371: dependence, such improvement will help use those processes as a
372: probe for new physics.
373:
374:
375: \emph{Acknowledgements:} We are grateful to K. Melnikov for
376: sharing his experience with solving large systems of recurrence
377: relations, and to J. Bl\"umlein and S. Moch for help with harmonic
378: polylogarithms, which were very helpful in determining master
379: integrals. This research was supported by the Science and
380: Engineering Research Canada, Alberta Ingenuity, and by the
381: Collaborative Linkage Grant PST.CLG.977761 from the NATO Science
382: Programme.
383:
384: % \bibliographystyle{apsrev}
385:
386: \begin{thebibliography}{17}
387: \expandafter\ifx\csname
388: natexlab\endcsname\relax\def\natexlab#1{#1}\fi
389: \expandafter\ifx\csname bibnamefont\endcsname\relax
390: \def\bibnamefont#1{#1}\fi
391: \expandafter\ifx\csname bibfnamefont\endcsname\relax
392: \def\bibfnamefont#1{#1}\fi
393: \expandafter\ifx\csname citenamefont\endcsname\relax
394: \def\citenamefont#1{#1}\fi
395: \expandafter\ifx\csname url\endcsname\relax
396: \def\url#1{\texttt{#1}}\fi
397: \expandafter\ifx\csname
398: urlprefix\endcsname\relax\def\urlprefix{URL }\fi
399: \providecommand{\bibinfo}[2]{#2}
400: \providecommand{\eprint}[2][]{\url{#2}}
401:
402: \bibitem[{\citenamefont{Steinhauser}(2002)}]{Steinhauser:2002rq}
403: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Steinhauser}},
404: \bibinfo{journal}{Phys. Rept.} \textbf{\bibinfo{volume}{364}},
405: \bibinfo{pages}{247} (\bibinfo{year}{2002}), \eprint{hep-ph/0201075}.
406:
407: \bibitem[{\citenamefont{Czarnecki}(1996)}]{zerorecoil}
408: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Czarnecki}},
409: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{76}},
410: \bibinfo{pages}{4124} (\bibinfo{year}{1996}), \eprint{hep-ph/9603261}.
411:
412: \bibitem[{\citenamefont{Czarnecki and Melnikov}(1997)}]{zerorecoilA}
413: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Czarnecki}} \bibnamefont{and}
414: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Melnikov}},
415: \bibinfo{journal}{Nucl. Phys.} \textbf{\bibinfo{volume}{B505}},
416: \bibinfo{pages}{65} (\bibinfo{year}{1997}), \eprint{hep-ph/9703277}.
417:
418: \bibitem[{\citenamefont{Franzkowski and Tausk}(1998)}]{Franzkowski:1997vg}
419: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Franzkowski}} \bibnamefont{and}
420: \bibinfo{author}{\bibfnamefont{J.~B.} \bibnamefont{Tausk}},
421: \bibinfo{journal}{Eur. Phys. J.} \textbf{\bibinfo{volume}{C5}},
422: \bibinfo{pages}{517} (\bibinfo{year}{1998}), \eprint{hep-ph/9712205}.
423:
424: \bibitem[{\citenamefont{van Ritbergen and Stuart}(1999)}]{vanRitbergen:1998yd}
425: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{van Ritbergen}}
426: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~G.}
427: \bibnamefont{Stuart}}, \bibinfo{journal}{Phys. Rev. Lett.}
428: \textbf{\bibinfo{volume}{82}}, \bibinfo{pages}{488} (\bibinfo{year}{1999}),
429: \eprint{hep-ph/9808283}.
430:
431: \bibitem[{\citenamefont{van Ritbergen}(1999)}]{vanRitbergen:1999gs}
432: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{van Ritbergen}},
433: \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B454}},
434: \bibinfo{pages}{353} (\bibinfo{year}{1999}), \eprint{hep-ph/9903226}.
435:
436: \bibitem[{\citenamefont{Steinhauser and
437: Seidensticker}(1999)}]{Steinhauser:1999bx}
438: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Steinhauser}} \bibnamefont{and}
439: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Seidensticker}},
440: \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B467}},
441: \bibinfo{pages}{271} (\bibinfo{year}{1999}), \eprint{hep-ph/9909436}.
442:
443: \bibitem[{\citenamefont{Chetyrkin et~al.}(1999)\citenamefont{Chetyrkin,
444: Harlander, Seidensticker, and Steinhauser}}]{Chetyrkin:1999ju}
445: \bibinfo{author}{\bibfnamefont{K.~G.} \bibnamefont{Chetyrkin}},
446: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Harlander}},
447: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Seidensticker}},
448: \bibnamefont{and}
449: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Steinhauser}},
450: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D60}},
451: \bibinfo{pages}{114015} (\bibinfo{year}{1999}), \eprint{hep-ph/9906273}.
452:
453: \bibitem[{\citenamefont{Czarnecki and Melnikov}(2002)}]{Czarnecki:2001cz}
454: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Czarnecki}} \bibnamefont{and}
455: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Melnikov}},
456: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{88}},
457: \bibinfo{pages}{131801} (\bibinfo{year}{2002}), \eprint{hep-ph/0112264}.
458:
459: \bibitem[{\citenamefont{Tkachov}(1981)}]{Tkachov:1981wb}
460: \bibinfo{author}{\bibfnamefont{F.~V.} \bibnamefont{Tkachov}},
461: \bibinfo{journal}{Phys. Lett.} \textbf{\bibinfo{volume}{B100}},
462: \bibinfo{pages}{65} (\bibinfo{year}{1981}).
463:
464: \bibitem[{\citenamefont{Broadhurst}(1992)}]{bro91a}
465: \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Broadhurst}},
466: \bibinfo{journal}{Z. Phys.} \textbf{\bibinfo{volume}{C54}},
467: \bibinfo{pages}{599} (\bibinfo{year}{1992}).
468:
469: \bibitem[{\citenamefont{Laporta}(2000)}]{Laporta:2001dd}
470: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Laporta}},
471: \bibinfo{journal}{Int. J. Mod. Phys.} \textbf{\bibinfo{volume}{A15}},
472: \bibinfo{pages}{5087} (\bibinfo{year}{2000}), \eprint{hep-ph/0102033}.
473:
474: \bibitem[{csc()}]{csc}
475: \bibinfo{note}{Our calculations are performed with facilities of the Centre for
476: Symbolic Computation at the University of Alberta}.
477:
478: \bibitem[{\citenamefont{Vermaseren}(2000)}]{Vermaseren:2000nd}
479: \bibinfo{author}{\bibfnamefont{J.~A.~M.} \bibnamefont{Vermaseren}}
480: (\bibinfo{year}{2000}), \eprint{math-ph/0010025}.
481:
482: \bibitem[{\citenamefont{Jezabek and Kuhn}(1989)}]{jk2}
483: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Jezabek}} \bibnamefont{and}
484: \bibinfo{author}{\bibfnamefont{J.~H.} \bibnamefont{Kuhn}},
485: \bibinfo{journal}{Nucl. Phys.} \textbf{\bibinfo{volume}{B314}},
486: \bibinfo{pages}{1} (\bibinfo{year}{1989}).
487:
488: \bibitem[{\citenamefont{Czarnecki and Melnikov}(1999)}]{Czarnecki:1998qc}
489: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Czarnecki}} \bibnamefont{and}
490: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Melnikov}},
491: \bibinfo{journal}{Nucl. Phys.} \textbf{\bibinfo{volume}{B544}},
492: \bibinfo{pages}{520} (\bibinfo{year}{1999}), \eprint{hep-ph/9806244}.
493:
494: \bibitem[{\citenamefont{Hagiwara~{\it et al.}}(2002)}]{PDBook}
495: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Hagiwara~{\it et al.}}},
496: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{D66}},
497: \bibinfo{pages}{010001} (\bibinfo{year}{2002}).
498:
499: \end{thebibliography}
500:
501:
502:
503: \end{document}
504: