1: \documentclass[12pt]{article}
2: \usepackage{epsfig}
3: \usepackage{amsmath}
4: \usepackage{float}
5: \usepackage{subfigure}
6: \usepackage{a4}
7: \usepackage{float}
8:
9: \voffset -.5cm
10:
11: \newcommand{\eqs}[1]{\begin{eqnarray}#1\end{eqnarray} }
12: \newcommand{\ce}[1]{Eq.~(\ref{#1})}
13: \newcommand{\ced}[2]{Eqs.~(\ref{#1}) \& (\ref{#2})}
14: \newcommand{\cf}[1]{{Figure~\ref{#1}}}
15: \newcommand{\ct}[1]{{Table~\ref{#1}}}
16: \newcommand{\ep}{\varepsilon}
17: \newcommand{\lvec}{\vec{\ell}}
18: \newcommand{\vect}[1]{\vec{#1}}
19: \newcommand{\vectc}[1]{\vec{#1}\,^2}
20: \newcommand{\mvec}[1]{|\vec{#1}|}
21: \newcommand{\matrice}[1]{\begin{matrix} #1\end{matrix} }
22: \newcommand{\ks}[1]{#1 \!\!\!\!\! \slash }
23: \newcommand{\Mvariable}[1]{#1}
24: \newcommand{\ga}{\gamma^5}
25: \newcommand{\gmu}{\gamma^\mu}
26: \newcommand{\gnu}{\gamma^\nu}
27: \newcommand{\gmup}{\gamma^{\mu'}}
28: \newcommand{\gnup}{\gamma^{\nu'}}
29: \newcommand{\grho}{\gamma^\rho}
30: \newcommand{\gsig}{\gamma^\sigma}
31: \newcommand{\grhop}{\gamma^{\rho'}}
32: \newcommand{\gsigp}{\gamma^{\sigma'}}
33: \newcommand{\gtau}{\gamma^{\tau}}
34: \newcommand{\gtaup}{\gamma^{\tau'}}
35: \newcommand{\gz}{\gamma^0}
36: \newcommand{\tr}{{\rm Tr}}
37: \newcommand{\ie}{{\it i.~e.}}
38: \newcommand{\eg}{{\it e.g.}}
39: \newcommand{\nn}{\nonumber}
40: \renewcommand{\cal}{\mathcal}
41:
42: \begin{document}
43:
44: \title{New contributions to \\ heavy-quarkonium production}
45:
46: \author{\vspace*{0.2cm}J.P. Lansberg$^{a,b}$, J.R. Cudell$^a$ and Yu. L. Kalinovsky$^{a,c}$\\
47: \centerline{\small $^a$ \it Physique th\'eorique fondamentale, D\'ep. de Physique,
48: Univ. de Li\`ege, }\\
49: \centerline{\small \it ~~All\'ee du 6 Ao\^{u}t 17, b\^{a}t. B5a, B-4000 Li\`ege~1, Belgium}\\
50: \centerline{\small $^b$ \it Centre de Physique Th\'eorique, \'Ecole Polytechnique\protect\footnote{Unit\'e mixte 7644 du CNRS},}\\
51: \centerline{\small \it ~~ F-91128 Palaiseau, France}\\
52: \centerline{\small $^c$ \it Lab. of Information Technologies, Joint Inst. for Nuclear Research, }\\
53: \centerline{\small \it ~~Dubna, Russia}\\
54: \centerline{\small \it E-mails: Jean-Philippe.Lansberg@cpht.polytechnique.fr, JR.Cudell@ulg.ac.be,}\\
55: \centerline{\small \it kalinov@qcd.theo.phys.ulg.ac.be}
56: }
57: \date{}
58: \maketitle
59:
60: \begin{abstract}
61: We reconsider quarkonium production in a field-theoretical setting and
62: we show that the lowest-order mechanism for
63: heavy-quarkonium production receives in general contributions from two
64: different cuts. The first one corresponds to the usual
65: colour-singlet mechanism.
66: The second one has not been considered so far. We treat it in a gauge-invariant
67: manner, and introduce new 4-point vertices, suggestive of the colour-octet
68: mechanism. These new objects enable us to go beyond the static approximation.
69: We show that the contribution of the new cut can be as large as the usual
70: colour-singlet mechanism at high $P_T$ for $J/\psi$. In the $\psi'$ case,
71: theoretical uncertainties are shown to be large and agreement with data is
72: possible.
73: \end{abstract}
74: \noindent {\bf Keywords:} heavy-quarkonium production, vector-meson production,
75: gauge invariance, relativistic effects, non-static extension\\
76: \noindent {\bf PACS:} 14.40.Gx, 13.85.Ni, 11.10.St, 13.20.Gd
77:
78: \newpage
79: \section{Introduction}
80: Years after the first disagreement between data~\cite{CDF7997a,CDF7997b}
81: and the colour-singlet model (CSM)~\cite{CSM_hadron,CSM_frag}, the problem of
82: heavy-quarkonium production --in particular
83: at the Tevatron-- is still with us. Indeed, it is widely accepted now that fragmentation
84: processes, through the colour-octet
85: mechanism (COM)~\cite{COM}
86: dominate the production of heavy quarkonia in high-energy hadronic collisions,
87: even for $P_T$ as low as 6 GeV.
88: In the COM, one then
89: parametrises the non-perturbative transition from octet to singlet by unknown
90: matrix elements, which are determined to reproduce the data. However, fragmentation
91: processes are known to produce mostly transversely-polarised vector mesons
92: at large transverse momentum~\cite{Cho:1994ih}
93: and, in the $J/\psi$ and $\psi'$ cases, this seems in contradiction with the
94: measurements from CDF~\cite{Affolder:2000nn}. For a comprehensive review
95: on the subject, the reader may refer to~\cite{yr_QWG}.
96:
97: We therefore reconsider the basis of quarkonium ($\cal Q$) production in field
98: theory, and concentrate on $J/\psi$ and $\psi'$ production.
99: We shall see that new contributions are present in the
100: lowest-order diagrams, and we shall also explain how one can build
101: a consistent and systematic scheme to go beyond the static approximation.
102: To this end, we shall use 3-point vertices depending on the relative momentum
103: of the constituent quarks and normalised to the leptonic width of the meson.
104: We shall show that, in order to preserve gauge invariance, it is
105: required to introduce vertices more complicated than the
106: 3-point vertex.
107:
108:
109: Finally, we shall see that our formalism can be easily applied to the production
110: of excited states. In the case of $\psi'$, the theoretical uncertainties
111: are unexpectedly large and allow agreement with the data.
112:
113: \section{Bound states in QCD}
114: All the information
115: needed to study processes involving bound states,
116: such as decay and production mechanisms, can be parametrised by vertex functions, which describe
117: the coupling of the bound state to its constituents and contain the information
118: about its size, the amplitude of probability for given quark configurations
119: and the normalisation of its wave function. In the case of heavy quarkonia,
120: the situation simplifies as they can be approximated by their lowest Fock state, made of
121: a heavy quark and an antiquark, combined to obtain the proper quantum numbers.
122: Furthermore, it has been shown~\cite{Burden:1996nh} that, for light {\sl vector}
123: mesons, the dominant projection operator is $\gamma^\mu$ and we expect this to hold even better
124: for heavy vector mesons as this approximation gets better in the case of
125: $\phi(s\bar s)$.
126:
127: The transition
128: $q\bar q\rightarrow {\mathcal Q}$ can then be described by the following
129: 3-point function:
130: \begin{eqnarray}\label{vf}
131: \Gamma^{(3)}_{\mu}(p,P) = \Gamma(p,P) \gamma_\mu,
132: \end{eqnarray}
133: with $P\equiv p_{1}-p_{2}$ the total momentum of the bound state, and $p\equiv(p_{1}+p_{2})/2$
134: the relative momentum of the bound quarks, as drawn in \cf{fig:BS_vertex_phenoa}. This choice
135: amounts to describing the vector meson as a massive photon with a non-local
136: coupling.
137:
138: \begin{figure}[t]
139: \centering
140: \includegraphics[width=10cm]{BS_vertex_pheno}
141: \caption{Bound-state vertex obtained by multiplying a point vertex, representing a structureless
142: particle, by a form factor.}
143: \label{fig:BS_vertex_phenoa}
144: \end{figure}
145:
146: We do not assume that the quarks
147: are on-shell : their momentum distribution comes from $\Gamma(p,P)$ and
148: from their propagators. In order to make contact with wave functions,
149: and also to simplify calculations, we assume that $\Gamma(p,P)$ can
150: be taken as a function of the square of the
151: relative c.m. 3-momentum $\vec p$ of the quarks, which can be written in a
152: Lorentz invariant form as $\vec p^{\,\, 2}=
153: -p^2+\frac{(p.P)^2}{M^2}$. For the functional form of $\Gamma(p,P)$, we neglect
154: possible cuts, and choose two otherwise extreme scenarios:
155: a dipolar form which decreases slowly with $\vect p$, and a Gaussian form:
156: \eqs{
157: \Gamma(p,P)=\frac{N}{(1+\frac{\vec p^{\,2}}{\Lambda^2})^2} \text{ and } \Gamma(p,P)=N e^{-\frac{\vec p^{\,2}}{\Lambda^2}}
158: ,}
159: both with a normalisation $N$ and a size parameter $\Lambda$, which
160: can be obtained from relativistic quark models~\cite{Lambda}.
161: We shall see in Section 4 how we fix the latter using the
162: leptonic-decay width.
163:
164: \section{Lowest-order production diagrams}
165: In high-energy hadronic collisions, quarkonia are produced at small $x$,
166: where protons are mainly made of gluons. Hence
167: gluon fusion is the main production mechanism.
168: In the case of $J/\psi$ and $\psi'$,
169: a final-state gluon emission is required to conserve $C$ parity. This gluon also
170: provides the $\cal Q$ with transverse momentum $P_T$. We assume here that
171: we can use collinear factorisation to describe the initial gluons, and hence
172: that the final-state gluon emission is the unique source of $P_T$. All the relevant
173: diagrams for the lowest-order gluon-initiated production process
174: can be obtained from that of \cf{fig:Landau} by crossing. There are six of them.
175:
176:
177:
178: \begin{figure}[t]
179: \centerline{\includegraphics[width=4cm]{Landau.eps}}
180: \caption{Box diagram.}
181: \label{fig:Landau}
182: \end{figure}
183:
184:
185: These diagrams have discontinuities, which generate their imaginary parts.
186: In order to find these, we can use the Landau equations~\cite{Landau_eq}.
187: It is sufficient to consider the diagram of \cf{fig:Landau},
188: for which the Landau equations become:
189: \begin{eqnarray}
190: \lambda_1 (c_1^2-m^2)&=&0\nonumber\\
191: \lambda_2 ((c_1-P)^2-m^2)&=&0\nonumber\\
192: \lambda_3 ((c_1-k_1)^2-m^2)&=&0\\
193: \lambda_4 ((c_1-k_1-k_2)^2-m^2)&=&0\nonumber\\
194: \lambda_1 c_1+\lambda_2 (c_1-P)+\lambda_3 (c_1-k_1)+\lambda_4 (c_1-k_1-k_2)&=&0\nonumber
195: \label{eq:landau}
196: \end{eqnarray}
197: with $m$ the quark mass.
198: These equations have only two solutions in the physical region:
199: one which is always present and gives a cut which starts at $\hat s=(k_1+k_2)^2=4 m^2$,
200: shown in \cf{fig:diag_LO_QCD}~(a), corresponding to a cut through the two $s$-channel
201: propagators, and another one when the meson mass $M$ is larger or
202: equal to $2 m$ (see \cf{fig:diag_LO_QCD}~(b): this corresponds to a cut through
203: the two propagators touching the meson). The latter leads to the colour-singlet
204: model~\cite{CSM_hadron}, which assumes that the quarks should be put on-shell
205: to make the meson. The former cut has not been considered so far for the
206: description of inclusive production. Let us mention however that
207: similar cuts are dominant in diffractive production
208: of vector mesons~\cite{diffractive}, or in DVCS~\cite{DVCS}.
209:
210:
211: We are going to consider this $s$-channel cut in detail. To avoid
212: complications, we choose quark masses $m>M/2$ high enough for the second cut
213: not to contribute. This will also simplify the normalisation procedure
214: for the vertex.
215:
216:
217: \begin{figure}[H]
218: \centerline{\mbox{
219: \subfigure[cut 1]{\includegraphics[height=3.5cm]{LO_LCK_diag}}
220: \quad\quad\quad
221: \subfigure[cut 2]{\includegraphics[height=3.5cm]{LO_CSM_diag}}
222: }}
223: \caption{The first family (a) has 4 diagrams
224: and the second family (b) 6 diagrams
225: contributing the discontinuity of $gg \to \!\!\ ^3S_1 g$ at LO in QCD.}
226: \label{fig:diag_LO_QCD}
227: \end{figure}
228:
229: \subsection{Non-locality and gauge invariance}
230: The first problem one immediately faces when evaluating the diagrams of\break
231: \cf{fig:diag_LO_QCD}~(a) is that of gauge invariance: whereas these
232: diagrams are gauge invariant if we have a photon instead of a $\cal Q$,
233: they are not for a finite-size object. Indeed, the vertex function
234: $\Gamma(p,P)$ takes different values in diagrams where either the on-shell
235: quark or the antiquark touches the $\cal Q$: the relative momentum
236: $p$ is then either $p=2c_1-P$ or $p=2c_2+P$, so that the
237: delicate cancellation that ensures current conservation is spoilt.
238:
239: \begin{figure}[H]
240: \centerline{\mbox{\subfigure[]{\includegraphics[height=4.25cm]{interac_constit_a.eps}\quad\quad
241: \includegraphics[height=4.25cm]{interac_constit_b.eps}}\quad\quad
242: \subfigure[]{\includegraphics[height=4.25cm]{interac_constit_c.eps}}}}
243: \caption{Illustration of the necessity of a 4-point vertex.}
244: \label{fig:illus_GI_break}
245: \end{figure}
246:
247: The reason is easily understood:
248: if one considers a local vertex, then the gluon can only couple to the
249: quarks that enter the vertex. For a non-local vertex, it
250: is possible for the gluon to connect to the quark or gluon lines inside the vertex,
251: as shown in \cf{fig:illus_GI_break}\footnote{In the cases of~\cf{fig:diag_LO_QCD} (b),
252: for $M=2m$, no gluon emission is kinematically allowed and the diagrams are directly
253: gauge-invariant.}.
254: These contributions must generate a 4-point $q\bar q {\cal Q}g$
255: vertex, $\Gamma^{(4)}_{\mu\nu}(c_1,c_2,q,P)$. In
256: general, its form is unknown, but it must obey some general constraints ~\cite{Drell:1971vx,Lansberg:2005aw}:
257: \begin{itemize}
258: \item
259: it must restore gauge invariance: its addition to the amplitude must
260: lead to current conservation at the gluon vertex;
261: \item
262: it must obey crossing symmetry (or invariance by $C$ conjugation) which
263: can be written
264: \begin{equation}
265: \Gamma^{(4)}_{\mu\nu}(c_1,c_2,q,P,m)=-\gamma_0\Gamma_{\mu\nu}^{(4)}(-c_2,-c_1,q,P,-m)^\dagger\gamma_0;
266: \end{equation}
267: \item
268: it must not introduce new singularities absent from the propagators or
269: from $\Gamma(p,P)$, hence it can only have denominators proportional to $(c_1-P)^2-m^2$ or
270: $(c_2+P)^2-m^2$;
271: \item
272: it must vanish in the case of a local vertex $\Gamma_\mu^{(3)}\propto\gamma_\mu$,
273: hence we multiply it by $\Gamma(2c_1-P,P)-\Gamma(2c_2+P,P)$.
274: \end{itemize}
275:
276:
277: These conditions are all fulfilled by the following choice~\cite{Lansberg:2005aw}:
278: \begin{eqnarray}
279: \Gamma^{(4)}_{\mu\nu}(c_1,c_2,P,q)&=&-i g_s T^a_{ki} \left[ \Gamma(2c_1+P,P)
280: -\Gamma(2c_2-P,P)\right]\nonumber \\
281: &\times&\left[\frac{c_{1\nu}}{(c_2+P)^2-m^2}
282: +\frac{c_{2\nu}}{(c_1-P)^2-m^2}\right]\gamma_\mu
283: \label{Gamma4}
284: \end{eqnarray}
285: where the indices of the colour matrix $T$
286: are defined in \cf{fig:gamma_4}, and $g_s$ is the strong coupling constant.
287:
288: \begin{figure}[H]
289: \centering
290: \includegraphics[width=3cm]{gamma_4}
291: \caption{The gauge-invariance restoring vertex, $\Gamma^{(4)}$.}
292: \label{fig:gamma_4}
293: \end{figure}
294:
295: It must be noted first that the quark pair $(c_1,c_2)$ that makes the meson is now in
296: a colour-octet state. We thus recover the necessity of such configurations, in this case to restore gauge invariance.
297: We must also point out that this choice of vertex
298: is not unique. We postpone a full study of the 4-point vertex ambiguity~\cite{Lansberg:2005aw} to another paper.
299:
300: When taken into account
301: in the calculation of the discontinuity of $gg\to {\cal Q} g$,
302: $\Gamma^{(4)}$ introduces two new diagrams\footnote{The
303: contributions of the triple-gluon vertex on the left and
304: $\Gamma^{(4)}$ on the right is zero due to
305: charge conjugation.} shown in \cf{fig:LO_new_diag}.
306: Including these contributions in the calculation
307: of the amplitude, we obtain a gauge-invariant quantity.
308:
309:
310:
311: \begin{figure}[H]
312: \centering
313: \includegraphics[width=8cm]{LO_new_diag}
314: \caption{New contributions to the discontinuity of the amplitude from $\Gamma^{(4)}$.}
315: \label{fig:LO_new_diag}
316: \end{figure}
317:
318: \section{Amplitudes}
319: We define ${\cal A}_i^{\mu\nu\rho\sigma}$, $i=1,$..., 6,
320: as the unintegrated amplitude
321: given by the usual Feynman rules for the four diagrams
322: of~\cf{fig:diag_LO_QCD}~(a) and the two of~\cf{fig:LO_new_diag}. We choose
323: the loop momentum $\ell$ so that $c_1= \ell+k_1$ and $c_2=\ell-k_2$. We then have for
324: the imaginary part of the physical amplitude ${\cal M}$:
325: \eqs{
326: {\cal M}^{pqrs}&=& \frac{1}{2} \sum_{i=1}^6 \int \frac{d^4\ell}{(2\pi)^4} {\cal A}_i^{\mu\nu\rho\sigma}
327: \ep^p_{1,\mu}\ep^q_{2,\nu} \ep^{r\star}_{3,\rho} \ep^{s\star}_{4,\sigma} \nn \\
328: &\times &2\pi\delta^{+}((\ell+k_{1})^{2}-m^{2} ) 2\pi \delta^{-}((\ell-k_{2})^{2}-m^{2})\nn \\
329: }
330:
331: This polarised partonic amplitude is thus obtained by contracting
332: ${\cal A}_i^{\mu\nu\rho\sigma}$ with the
333: polarisation vectors of the gluons $\ep^p_{1,\mu}$, $\ep^q_{2,\nu}$, $\ep^s_{4,\sigma}$ and with that of
334: the vector meson $\ep^r_{3,\rho}$, by integrating on
335: the internal phase space restricted by the cutting rules and
336: by summing the six contributions from the
337: diagrams of \cf{fig:diag_LO_QCD}~(a) and \cf{fig:LO_new_diag}.
338:
339: The complete expressions of the polarisation vectors are as follows.
340: One of the transverse polarisations can be taken as orthogonal to the plane of the collision:
341: \begin{eqnarray}
342: \ep_{1}^{T_1}=\ep_{2}^{T_1}=\ep_{3}^{T_1}=\ep_{4}^{T_1}=\epsilon_T,
343: \end{eqnarray} with
344: $\epsilon_T.k_1=\epsilon_T.k_2=\epsilon_T.P=0$. The other transverse
345: polarisation can be taken as
346: \begin{eqnarray}
347: \ep_{1}^{T_2}=\sqrt{\frac{1}{\hat s \hat t \hat u}}\left(\hat t k_1+ \hat u k_2+ \hat s q\right)=\ep_{2}^{T_2}
348: \end{eqnarray} for the two gluons, and as
349: \eqs{\label{eq:pol-final2}
350: \ep_{3}^{T_2}=\sqrt{\frac{1}{\hat s \hat t \hat u}}\left(\hat t k_1-\hat u k_2+\left(\frac{\hat s(\hat t-\hat u)}{\hat t+\hat u}\right) q\right)
351: =\ep_{4}^{T_2},\\
352: } for the final-state gluon and for the vector meson,
353: where $\hat s=(k_1+k_2)^2$, $\hat t=(k_2-q)^2$ and $\hat u=(k_1-q)^2$
354: are the Mandelstam variables for the partonic process.
355:
356: Finally, the longitudinal
357: vector-meson polarisation can be taken as
358: \eqs{
359: \ep_{3}^{L} = \frac{1}{M}\left(k_1+k_2-\left(\frac{\hat s+M^2}{\hat s-M^2}\right) q\right),\\
360: }
361:
362: To complete the calculation of the amplitude, we need to normalise
363: the 3-point vertex function $\Gamma^{(3)}$. We use here the leptonic
364: decay width to fix this normalisation~\cite{Lansberg:2005aw,Lansberg:2005ed}.
365:
366: The width in terms of the decay amplitude is given by
367: \eqs{\label{eq:lept_width}
368: \Gamma_{\ell\ell}=\frac{1}{2 M}\frac{1}{(4\pi^2)}\int \left|\bar{\cal{M}}\right|^2 d_2(PS),
369: }
370: where $d_2(PS)$ is the two-particle phase space~\cite{barger}.
371:
372: The amplitude is obtained as usual through Feynman rules. At lowest order in $\alpha_s$,
373: only the 3-point vertex function needs to be considered (giving an explicitly gauge invariant
374: answer). The square of the amplitude is then obtained from the diagram drawn in~\cf{fig:decay_diag_1}.
375:
376: \begin{figure}[H]
377: \centering{\mbox{\includegraphics[width=10cm]{decay_diag_1}}}
378: \caption{Feynman diagram for $^3S_1\to \ell \bar \ell$.}\label{fig:decay_diag_1}
379: \end{figure}
380:
381: \noindent In terms of the sub-amplitudes $A^{\mu\nu}$, $B^{\mu\nu}$ and $C^{\mu\nu}$ defined in
382: \cf{fig:decay_diag_1}, we have\footnote{We performed the calculation in the Feynman gauge, but the results are gauge invariant.}:
383: \eqs{
384: \label{eq:decomp_ampl_inv_decay}
385: \int \left|\bar{\cal M}\right|^2 d_2(PS)=-\frac{1}{3}
386: \Delta_{\mu\mu'}
387: A^{\mu\nu}\frac{g_{\nu\rho}}{M^2}B^{\rho\rho'}\frac{g_{\rho'\nu'}}{M^2}
388: C^{\nu'\mu'},
389: }
390: where the factor $\Delta_{\mu\nu}=(g_{\mu\nu}-\frac{P_\mu P_{\nu'}}{M^2})=
391: \sum_i \ep_{i,\mu} \ep^\star_{i,\mu'}$ comes from the sum over polarisations
392: of the meson and the factor $\frac{1}{3}$ from the averaging over the initial polarisations.
393:
394: $B^{\rho\rho'}$, after
395: integration on the two-particle phase space, is found to be
396: \eqs{
397: B^{\rho\rho'}=-e^2 \frac{2\pi}{3} M^2\underbrace{\left[g^{\rho\rho'}-
398: \frac{P^\rho P^{\rho'}}{M^2}\right]}_{\Delta^{\rho\rho'}}.
399: }
400: $A^{\mu\nu}$ (or equivalently ${C^{\mu\nu}}^\dagger$) can be written
401: (see \cf{fig:decay_diag_1a}):
402: \eqs{\label{eq:Amunu_start}
403: iA^{\mu\nu}=-3e_Q\int\! \frac{d^4k}{(2 \pi)^4}
404: \Gamma(k,P)\frac{g^{\mu\nu} (M^2+4m^2-4k^2)+8 k^\mu k^\nu-2P^\mu P^\nu}
405: {((k-\frac{P}{2})^2-m^2+i\ep)((k+\frac{P}{2})^2-m^2+i\ep)},
406: }
407: with $e_Q$ the heavy-quark charge.
408:
409: \begin{figure}[H]
410: \centering{\mbox{\includegraphics[width=5cm]{decay_diag_1a}}}
411: \caption{Feynman diagram for $^3S_1\to \gamma^\star$.}\label{fig:decay_diag_1a}
412: \end{figure}
413: Performing the $k^0$ integration by residues, one obtains
414: \eqs{A^{\mu\nu}=\frac{- e_Q}{\pi^2} N I(\Lambda,M,m) \Delta^{\mu\nu}.
415: }
416: with
417: \eqs{\label{eq:I}
418: I(\Lambda,M,m)\equiv\int_0^\infty \!\! \frac{dK K^2\Gamma(k,P)}{\sqrt{K^2+m^2}N}
419: \frac{(2K^2+3m^2)}{(K^2+m^2-\frac{M^2}{4})}.
420: }
421: and $K=|\vec k|$.
422: $I$ is a function of $\Lambda$ through the vertex function $\Gamma(k,P)$
423: and is not in general computable analytically, but it is straightforward to get its
424: numerical value.
425:
426: \begin{figure}[H]
427: \centering\includegraphics[width=0.49\textwidth]{normalisation_dip_Jpsi_290305_lam2}
428: \centering\includegraphics[width=0.49\textwidth]{normalisation_exp_Jpsi_290305_lam2}
429: \caption{Normalisation for a dipole (resp. Gaussian) form of $\Gamma(p,P)$ in the $J/\psi$ case
430: as a function of $\Lambda$: left (resp. right).}
431: \label{fig:N_jpsi_lam}
432: \end{figure}
433: We can then put all the pieces together using Eqs. (\ref{eq:lept_width}) and (\ref{eq:decomp_ampl_inv_decay})
434: to determine $N$ from the measured leptonic width.
435: We show in \cf{fig:N_jpsi_lam} the result in the $J/\psi$ case.
436: As can be seen, the normalisation of the vertex depends rather strongly on $m_c$
437: and $\Lambda$, but very little on the assumed functional
438: dependence of the vertex function. We shall see later that once $N$ is
439: determined from the leptonic rate, the production cross section depends
440: little on these uncertainties.
441:
442:
443: \section{Production cross sections}
444:
445: We can now evaluate the production cross section from the $s$-channel cut.
446: As stated before, we assume that collinear factorisation can be used,
447: in which case the link between the partonic and the
448: hadronic cross sections is given by the following general formula:
449: \begin{equation}
450: \label{eq:Edsdp3}
451: E\frac{d^3\sigma}{dP^3}=\int_0^1 dx_1 dx_2 \ g(x_1) \ g(x_2)\ \frac{\hat s}{\pi}\frac{d\sigma}{d\hat t}
452: \delta(\hat s +\hat t +\hat u -M^2),
453: \end{equation}
454: where $x_1$ and $x_2$ are the momentum fractions of the incoming gluons,
455: $P=(E,\vect P)$ is the momentum of the meson in the c.m. frame
456: of the colliding hadrons, $g(x)$ is the gluon distribution function\footnote{In
457: our calculations, we have chosen two LO gluon parametrisations,
458: MRST~\cite{Martin:2002dr} and CTEQ~\cite{Pumplin:2002vw}. For
459: each plot, the one used will be specified.} taken
460: at the scale $\sqrt{M^2+P_T^2}$ .
461:
462: In the c.m. frame of the colliding hadrons, introducing the rapidity $y=\tanh^{-1}(\frac{P_z}{E})$ and
463: the transverse momentum $\vect P_T$, we obtain the double-differential cross section in $P_T$ and $y$
464: from \ce{eq:Edsdp3}:
465: \eqs{
466: \frac{d\sigma}{dydP_T}=\int_0^1 dx_1 dx_2 g(x_1) g(x_2) 2 \hat s P_T\frac{d\sigma}{d\hat t}
467: \delta(\hat s+\hat t+\hat u-M^2)&&.
468: }
469:
470: At this stage, we can perform the integration on $x_2$ (or $x_1$) using the delta function.
471:
472:
473: In terms of the transverse energy $E_T=\sqrt{P_T^2+M^2}$, we get
474: \eqs{
475: \hat s=s x_1x_2, \
476: \hat t = M^2-x_1e^{-y}\sqrt{s}E_T,\
477: \hat u =M^2-x_2e^{y}\sqrt{s}E_T,
478: }
479: so that we obtain
480: \begin{equation}\label{eq:x1x2}
481: x_2=\frac{x_1E_T\sqrt{s}e^{-y}-M^2}{\sqrt{s}(\sqrt{s} x_1-E_Te^{y})}.
482: \end{equation}
483: The double differential cross section on $P_T$ and $y$ then takes the following form:
484: \eqs{
485: \frac{d\sigma}{dydP_T}=\int_{x_1^{min}}^1 dx_1 \frac{2 \hat s P_T g(x_1) g(x_2(x_1))}
486: {\sqrt{s}(\sqrt{s} x_1-E_Te^{y})}
487: \frac{d\sigma}{d\hat t},
488: }
489: where $x_1^{min}$ corresponds to $x_2=1$ in Eq.~(\ref{eq:x1x2}):
490: \eqs{\label{eq:low_bound_x_1}
491: x_1^{min}= \frac{E_T\sqrt{s}e^{y}-M^2}{\sqrt{s}(E_T e^{-y}-\sqrt{s})}.
492: }
493:
494: The last step is to relate the partonic differential cross section $\frac{d\sigma}{d\hat t}$ to
495: the amplitude calculated from our model. To this end, we use the well-known formula:
496: \begin{equation} \label{eq:sigmadt}
497: \frac{d\sigma^{pqrs}}{d\hat t}=\frac{1}{16\pi \hat s^2}|\overline{\cal M^{pqrs}}|^2,
498: \end{equation}
499: where $|\overline{\cal M^{pqrs}}|^2$ is the squared polarised partonic amplitude for $gg\to{\cal Q} g$,
500: averaged only over colour for polarised cross sections, and where $p$, $q$, $r$ and $s$
501: are the helicities of the
502: four particles.
503:
504: As we are concerned with polarisation only for the $\cal Q$, we sum
505: over gluon polarisations and define, for $r=L,T_1,T_2$:
506: \eqs{
507: \frac{d \sigma_{r}}{d\hat t}=\sum_{p,q,s=T_1,T_2}\frac{d \sigma^{pqrs}}{d\hat t}.
508: }
509:
510: Finally , we have the double-differential polarised cross section on $P_T$
511: and $y$ :
512: \eqs{
513: \frac{d\sigma_r}{dydP_T}=\int_{x_1^{min}}^1 dx_1 \frac{2 \hat s P_T g(x_1) g(x_2(x_1))}
514: {\sqrt{s}(\sqrt{s}x_1-E_T e^{y})}\frac{d\sigma_r}{d\hat t}.
515: }
516:
517:
518: \section{Results for $J/\psi$}
519: %NEW
520: Before presenting our results, we need to choose a value for $\Lambda$ and
521: $m_c$.
522: Several studies have shown, in the context of relativistic quark
523: models \cite{Lambda},
524: that the scale of the vertex function is between 1.42 and 2.6 GeV, and
525: that $m_c$ is between 1.42 and 1.87 GeV. We choose here a value of
526: $\Lambda$ in the middle range, $\Lambda=1.8$ GeV (we shall see that small
527: variations do not affect our results much), and a value of $m_c$ equals to the
528: $D^{\pm}$ mass, $m_c$=1.87 GeV, in order to have a coherent treatment of all stable
529: charmonium states.
530:
531: Setting $\sqrt{s}$ to 1800 GeV and considering the cross section in the
532: pseudorapidity range $|\eta|<0.6$, we get the following results for $J/\psi$ production at CDF.
533: The first plot~(see \cf{fig:g_m187_l183_mrst}~(a)) shows our result ($\sigma_{TOT}$, $\sigma_T$
534: and $\sigma_L$) for $m=1.87$ GeV and $\Lambda=1.8$~GeV.
535:
536: These new contributions are
537: compared with the usual LO CSM~\cite{CSM_hadron}.
538:
539: It must be stressed that (in the Feynman gauge)
540: the main contribution comes from the 3-point function (\ce{vf}).
541: As can be seen from \cf{fig:g_m187_l183_mrst}~(b), the term
542: that restores gauge invariance (\ce{Gamma4}) contributes little:
543: the square of its amplitude (see \cf{fig:illus_GI_break} (b)) is about
544: 10 times smaller than the square of the amplitude containing
545: only a 3-point vertex (see \cf{fig:illus_GI_break}~(a)). Furthermore, the interference term
546: between the diagrams of \cf{fig:illus_GI_break}~(a) and that of \cf{fig:illus_GI_break}~(b)
547: is negative, so that the effect of \ce{Gamma4} is to reduce the total amplitude squared.
548: In Figures~\ref{fig:GIPA_jpsi_comp1}%and~\ref{fig:GIPA_jpsi_comp2}
549: , we show that the normalisation of the results using
550: the decay width has removed most dependence on the choice of parameters\footnote{Instead
551: of a factor 100 of difference expected from $\left(\frac{N_{\Lambda=1.0}}{N_{\Lambda=2.2}}\right)^2$
552: we have less than a factor 2 at $P_T=4$ GeV and a factor 3 at $P_T=20$ GeV.}.
553: Interestingly, as figure 11 (b) shows, the dependence on $\Lambda$ is
554: negligible once values of the order of 1.4 GeV are taken.
555: \begin{figure}[ht]
556: \centerline{\mbox{\subfigure[$\sigma_{\rm TOT}$, $\sigma_T$ and $\sigma_L$]{
557: \includegraphics[height=6.1cm,clip=true]{g_m187_l183_mrst_articlebb.eps}}\quad
558: \subfigure[$\sigma_{\rm TOT}$ and $\sigma_{\rm Pert}$,]{\includegraphics[height=6.cm,clip=true]{dsdpt_comparbb.eps}}
559: }}
560: \caption{(a) Polarised ($\sigma_T$ and $\sigma_L$) and total ($\sigma_{TOT}$) cross sections obtained
561: with a Gaussian vertex function, $m=1.87$ GeV, $\Lambda=1.8$~GeV
562: and the MRST gluon distribution, to be compared with LO CSM.
563: (b) The total (gauge invariant) contribution (plain curve) compared with that
564: of the 3-point vertex of diagrams of \cf{fig:illus_GI_break}~(a) (dashed curve)
565: in the Feynman gauge (CTEQ).}
566: \label{fig:g_m187_l183_mrst}
567: \end{figure}
568:
569: We see that the contribution of the new cut matters at large $P_T$.
570: Noteworthily, it is much flatter in $P_T$ than the LO CSM,
571: and its polarisation is mostly longitudinal.
572: This could have been expected as scalar products
573: of $\ep_L$ with momenta in the loop
574: will give an extra $\sqrt{\hat s}$ contribution, or equivalently an extra $P_T$ power in the amplitude, compared
575: to scalar products involving $\ep_T$. One can show that, for $\Lambda=1.8$ GeV,
576: $m_c=1.87$ GeV and for MRST structure functions, the longitudinal
577: cross section falls as $1/p_T^{8.5}$, whereas the transverse cross section
578: behaves asymptotically as $1/p_T^{10.5}$. The asymptotic
579: power of $p_T$ changes by
580: 5 \% for $\Lambda$ varying between 1.4 and 2.2 GeV.
581:
582: Recall that in our calculation the LO CSM is zero because $M<2m$. For
583: $M\geq 2m$, we should have added our
584: contribution to that of the LO CSM at the amplitude level.
585: The net result would then be flatter and larger.
586: However, it is clear that the enhancement factor would not be large enough to
587: reach agreement with the data.
588:
589: \begin{figure}[H]
590: \centering
591: \mbox{\subfigure[Vertex function]{\includegraphics[height=4.9cm,clip=true]{c_m187_l183_mrst_articlebb.eps}}
592: \subfigure[$\Lambda$]{\includegraphics[height=4.8cm,clip=true]{g_m187_lvar_mrst_articlebis2.eps}}
593: \subfigure[$\Lambda$ and $m_c$]{\includegraphics[height=4.8cm,clip=true]{g_mvar_lvar_cteq_articlebb.eps}}}
594: \caption{(a) Comparison between the polarised cross sections obtained with the dipole and the Gaussian
595: vertex functions (MRST); (b) Variation of the total cross section due to a change in $\Lambda$ for a fixed value
596: of the quark mass (MRST); (c) Variation of the total cross section due to a change in $m$ and $\Lambda$ (CTEQ).}
597: \label{fig:GIPA_jpsi_comp1}
598: \end{figure}
599: %\begin{figure}[H]
600: %\centering
601: % \mbox{\subfigure[$\Lambda$ and $m$]{\includegraphics[height=5.3cm,clip=true]{g_mvar_lvar_cteq_articlebb.eps}}
602: % \subfigure[Scale]{\includegraphics[height=5.3cm,clip=true]{g_m187_l183_cteq_scale_articlebb.eps}}}
603: %\caption{(a) Variation of the total cross section due to a change in $m$ and $\Lambda$ (CTEQ);
604: %(b) Variation of the total cross section due to a change of a factor two in the scale $Q$ from its nominal value $Q_0=\sqrt{M^2+P^2_T}$
605: %(CTEQ).}
606: %\label{fig:GIPA_jpsi_comp2}
607: %\end{figure}
608: \section{Results for $\psi'$}
609: Although the normalisation to the leptonic width
610: removes most of the ambiguities in the $J/\psi$
611: case, it is not so for radially excited states, such as the $\psi'$.
612: Indeed, in this case, the vertex function must have a node. We expect it to appear
613: through a pre-factor, $1-\frac{|\vect p|}{a_{node}}$, multiplying the $1S$
614: vertex function.
615: Explicitly, $\Gamma_{2S}(p,P)$, for a node $a_{node}$ should be well parametrised by
616: \eqs{\left(1-\frac{|\vect p|}{a_{node}}\right)
617: \frac{N'}{(1+\frac{|\vect p|^2}{\Lambda^2})^2} \text{ or }
618: N'\left(1-\frac{|\vect p|}{a_{node}}\right) e^{\frac{-|\vect p|^2}{\Lambda^2}}.}
619:
620: In order to determine the node position in momentum space, we can
621: fix $a_{node}$ from its known value in position space, \eg~from potential studies, and
622: take the Fourier transform of the wave function.
623: However, the position of the node is not very well-known, and it is unclear
624: how to relate our vertex with off-shell quarks to an on-shell non-relativistic
625: wave function. The most remarkable thing is that, because the integrand has a zero, the integral $I$ of
626: Eq.~(\ref{eq:I}) entering the decay width calculation can vanish for
627: a certain value $a_{node}=a_0$, which turns out to be close to the estimated
628: value of the
629: zero in the wave function. Because of their different momentum dependence, the integrals that
630: control the production are not zero for $a_{node}=a_0$. Hence our normalisation
631: procedure can in principle produce an infinite answer. Of course, this means
632: that one cannot be at $a_{node}=a_0$ exactly. However, if one is close to it,
633: then it becomes possible to produce a large normalisation. Hence in the $\psi'$
634: case, our procedure can produce agreement with the data at low $P_T$.
635:
636: \cf{fig:g_psip_a1334_0-0} shows that for $a_{node}=1.334$ GeV, one obtains a good fit to
637: CDF data at moderate $P_T$ (note that the slopes are quite similar. This is at odds with
638: what is commonly assumed since fragmentation processes --with a typical $1/P_T^4$ behaviour--
639: can also describe the data). The $\psi'$
640: is predicted to be mostly longitudinal.
641:
642:
643: \begin{figure}[H]
644: \centering{\includegraphics[height=6.2cm,clip=true]{g_psip_a1334_0-0bb}}
645: \caption{Polarised ($\sigma_T$ and $\sigma_L$) and total ($\sigma_{TOT}$) cross sections for $\psi'$
646: obtained with a Gaussian vertex functions, $a_{node}=1.334$ GeV, $m=1.87$ GeV, \break $\Lambda=1.8$~GeV
647: and the CTEQ gluon distribution, to be compared with LO CSM, CSM fragmentation~\cite{CSM_frag}
648: and the data from CDF~\cite{CDF7997a}.}
649: \label{fig:g_psip_a1334_0-0}
650: \end{figure}
651:
652: This effect of the node in the $\psi'$ vertex function could also solve
653: the $\rho-\pi$ puzzle as suggested in~\cite{Lansberg:2005ed}, since a slight modification
654: in the integrand of $I$ can produce a large suppression in the
655: $\rho-\pi$ decay of the $\psi'$. Such a modification in the integrand
656: is indeed expected to come from the presence of an off-shell $\omega$ in the $\rho-\pi$ decay
657: instead of an off-shell photon for the leptonic decay. On the other hand, in the case of $J/\psi$,
658: no important effect are expected.
659:
660:
661: \section{Conclusion and outlook}
662: In this letter, we have shown that there are two singularities in the box
663: diagram contribution to quarkonium production. We have chosen quark masses
664: so that only the $s$-channel singularity (which is usually neglected)
665: contributes, and vertices without cuts.
666:
667: On the theoretical side, we have begun to map the ingredients needed to go
668: beyond the static approximation $M=2m$. They involve the introduction of
669: new 4-point vertices that restore gauge invariance. We postpone a full study
670: of these to a later publication~\cite{article2}.
671:
672:
673: On the phenomenological side, we have shown that in the $J/\psi$ case,
674: this new singularity produces results comparable
675: to those of the lowest-order CSM, \ie~too small to accommodate the
676: Tevatron data.
677: In the $\psi'$ case, ambiguities in the position of the node of the
678: vertex function can lead to an enhancement, and to an agreement with the
679: data. Hence it is not clear that the same mechanism has to be at work
680: for $1S$ and $2S$ mesons.
681:
682: Our approach can be used in the case of $P$ waves, where we would simply input a suitable
683: 3-point vertex instead of taking higher derivatives of the
684: wave function and of the perturbative amplitude, but also in a $k_t$-factorisation framework \cite{Hagler:2000dd}
685: (which would enhance the cross section in the $J/\psi$ case), and can be combined with contribution
686: from COM fragmentation. Because the quarkonium are mostly longitudinal in this work
687: and transverse in fragmentation, it seems possible to reach agreement with polarisation
688: data.
689:
690:
691:
692: \section*{Acknowledgements}
693: J.P.L. is an IISN Postdoctoral Researcher, Yu.L.K. was a visiting research fellow of the FNRS
694: while this research was conducted, and is supported by the Russian grant
695: RFBR 03-01-00657. We would like to thank S.~Peign\'e, M.V.~Polyakov, W.J.~Stirling and
696: L.~Szymanowski for useful comments and discussions.
697:
698: \begin{thebibliography}{99}
699:
700:
701: %\cite{Abe:1997jz}
702: %\bibitem{Abe:1997jz}
703: \bibitem{CDF7997a}
704: F.~Abe {\it et al.} [CDF Collaboration],
705: %``J/psi and psi(2S) production in p anti-p collisions at s**(1/2) =
706: %1.8-TeV,''
707: Phys.\ Rev.\ Lett.\ {\bf 79} (1997) 572.
708: %%CITATION = PRLTA,79,572;%%
709:
710:
711: %\cite{Abe:1997yz}
712: %\bibitem{Abe:1997yz}
713: \bibitem{CDF7997b}
714: F.~Abe {\it et al.} [CDF Collaboration],
715: %``Production of J/psi mesons from chi/c meson decays in p anti-p collisions
716: %at s**(1/2) = 1.8-TeV,''
717: Phys.\ Rev.\ Lett.\ {\bf 79} (1997) 578.
718: %%CITATION = PRLTA,79,578;%%
719:
720: \bibitem{CSM_hadron}
721: C-H. Chang,
722: %Hadronic production of $J/\psi$ associated with a gluon,
723: Nucl. Phys. B {\bf 172} (1980) 425;
724: R. Baier and R. R\"uckl,
725: %Hadronic production of $J/\psi$: Transverse momentum distributions,
726: Phys. Lett. B {\bf 102} (1981) 364;
727: R. Baier and R. R\"uckl,
728: %Hadronic collisions: A quarkoniumfactory,
729: Z. Phys. {\bf C 19} (1983) 251.
730: %%CITATION = NUPHA,B172,425;%%
731: %%CITATION = PHLTA,B102,364;%%
732: %%CITATION = ZEPYA,C19,251;%%
733:
734: \bibitem{CSM_frag}
735: %\cite{Cacciari:1994dr}
736: %\bibitem{Cacciari:1994dr}
737: M.~Cacciari and M.~Greco,
738: %``J / psi production via fragmentation at the Tevatron,''
739: Phys.\ Rev.\ Lett.\ {\bf 73} (1994) 1586
740: [arXiv:hep-ph/9405241];
741: %%CITATION = HEP-PH 9405241;%%
742: %\cite{Braaten:1994xb}
743: %\bibitem{Braaten:1994xb}
744: E.~Braaten, M.~A.~Doncheski, S.~Fleming and M.~L.~Mangano,
745: %``Fragmentation production of J / psi and psi-prime at the Tevatron,''
746: Phys.\ Lett.\ B {\bf 333} (1994) 548
747: [arXiv:hep-ph/9405407].
748: %%CITATION = HEP-PH 9405407;%%
749:
750:
751:
752:
753:
754: %********************* COM
755: \bibitem{COM}
756: %\cite{Bodwin:1994jh}
757: %\bibitem{Bodwin:1994jh}
758: G.~T.~Bodwin, E.~Braaten and G.~P.~Lepage,
759: %``Rigorous QCD analysis of inclusive annihilation and production of heavy
760: %quarkonium,''
761: Phys.\ Rev.\ D {\bf 51} (1995) 1125
762: [Erratum-ibid.\ D {\bf 55} (1997) 5853]
763: [arXiv:hep-ph/9407339];
764: %%CITATION = HEP-PH 9407339;%%
765: %\cite{Cho:1995vh}
766: %\bibitem{Cho:1995vh}
767: P.~L.~Cho and A.~K.~Leibovich,
768: %``Color octet quarkonia production,''
769: Phys.\ Rev.\ D {\bf 53} (1996) 150
770: [arXiv:hep-ph/9505329];
771: %%CITATION = HEP-PH 9505329;%%
772: %\cite{Cho:1995ce}
773: %\bibitem{Cho:1995ce}
774: P.~L.~Cho and A.~K.~Leibovich,
775: %``Color-octet quarkonia production II,''
776: Phys.\ Rev.\ D {\bf 53} (1996) 6203
777: [arXiv:hep-ph/9511315].
778: %%CITATION = HEP-PH 9511315;%%
779:
780: %********Polarisation
781:
782: %\cite{Cho:1994ih}
783: \bibitem{Cho:1994ih}
784: P.~L.~Cho and M.~B.~Wise,
785: %``Spin symmetry predictions for heavy quarkonia alignment,''
786: Phys.\ Lett.\ B {\bf 346} (1995) 129
787: [arXiv:hep-ph/9411303].
788: %%CITATION = HEP-PH 9411303;%%
789:
790:
791: %\cite{Affolder:2000nn}
792: \bibitem{Affolder:2000nn}
793: T.~Affolder {\it et al.} [CDF Collaboration],
794: %``Measurement of J/psi and psi(2S) polarization in p anti-p collisions at
795: %s**(1/2) = 1.8-TeV,''
796: Phys.\ Rev.\ Lett.\ {\bf 85} (2000) 2886
797: [arXiv:hep-ex/0004027].
798: %%CITATION = HEP-EX 0004027;%%
799:
800: \bibitem{yr_QWG}
801: N.~Brambilla {\it et al.},
802: {\it CERN Yellow Report on ``Heavy quarkonium physics''}, CERN-2005-005
803: [arXiv:hep-ph/0412158].
804: %%CITATION = HEP-PH 0412158;%%
805:
806:
807: %\cite{Burden:1996nh}
808: \bibitem{Burden:1996nh}
809: C.~J.~Burden, L.~Qian, C.~D.~Roberts, P.~C.~Tandy and M.~J.~Thomson,
810: %``Ground-state spectrum of light-quark mesons,''
811: Phys.\ Rev.\ C {\bf 55} (1997) 2649
812: [arXiv:nucl-th/9605027].
813: %%CITATION = NUCL-TH 9605027;%%
814:
815: \bibitem{Lambda}
816: M.~A.~Ivanov, J.~G.~Korner and P.~Santorelli,
817: %``Semileptonic decays of B/c mesons into charmonium states in a relativistic
818: %quark model,''
819: Phys.\ Rev.\ D {\bf 71} (2005) 094006
820: [arXiv:hep-ph/0501051],
821: %%CITATION = HEP-PH 0501051;%%
822: Phys.\ Rev.\ D {\bf 70} (2004) 014005
823: [arXiv:hep-ph/0311300],
824: %%CITATION = HEP-PH 0311300;%%
825: Phys.\ Rev.\ D {\bf 63} (2001) 074010
826: [arXiv:hep-ph/0007169];
827: %%CITATION = HEP-PH 0007169;%%
828: M.~A.~Nobes and R.~M.~Woloshyn,
829: %``Decays of the B/c meson in a relativistic quark-meson model,''
830: J.\ Phys.\ G {\bf 26} (2000) 1079
831: [arXiv:hep-ph/0005056].
832: %%CITATION = HEP-PH 0005056;%%
833:
834:
835:
836: \bibitem{Landau_eq}
837: %\cite{Landau:1959fi}
838: %\bibitem{Landau:1959fi}
839: L.~D.~Landau,
840: %``On Analytic Properties Of Vertex Parts In Quantum Field Theory,''
841: Nucl.\ Phys.\ {\bf 13} (1959) 181;
842: %%CITATION = NUPHA,13,181;%%
843: %\bibitem{Itzyk}
844: C. Itzykson, J.B. Zuber, {\it Quantum Field Theory}, McGraw-Hill, New-York, 1980.
845:
846:
847:
848: \bibitem{diffractive} see {\it e.g.} J.~R.~Cudell and I.~Royen,
849: %``Elastic vector-meson production at HERA,''
850: Phys.\ Lett.\ B {\bf 397} (1997) 317
851: [arXiv:hep-ph/9609490];
852: %%CITATION = HEP-PH 9609490;%%
853: M.~G.~Ryskin,
854: %``Diffractive J / psi electroproduction in LLA QCD,''
855: Z.\ Phys.\ C {\bf 57} (1993) 89;
856: %%CITATION = ZEPYA,C57,89;%%
857: J.~R.~Cudell,
858: %``A QCD Inspired Model For Exclusive Vector Meson Production In Deep Inelastic
859: %Scattering,''
860: Nucl.\ Phys.\ B {\bf 336} (1990) 1.
861: %%CITATION = NUPHA,B336,1;%%
862: \bibitem{DVCS}A.~V.~Radyushkin,
863: %``Nonforward parton distributions,''
864: Phys.\ Rev.\ D {\bf 56} (1997) 5524
865: [arXiv:hep-ph/9704207].
866: %%CITATION = HEP-PH 9704207;%%
867:
868:
869:
870:
871: %\cite{Drell:1971vx}
872: \bibitem{Drell:1971vx}
873: S.~D.~Drell and T.~D.~Lee,
874: %``Scaling Properties And The Bound State Nature Of The Physical Nucleon,''
875: Phys.\ Rev.\ D {\bf 5} (1972) 1738.
876: %%CITATION = PHRVA,D5,1738;%%
877:
878:
879: %\cite{Lansberg:2005aw}
880: \bibitem{Lansberg:2005aw}
881: J.~P.~Lansberg, {\it Quarkonium Production at High-Energy Hadron Colliders}, Ph.D.
882: Thesis, ULg, Li\`ege, Belgium, 2005 [arXiv:hep-ph/0507175].
883: %%CITATION = HEP-PH 0507175;%%
884:
885:
886:
887: %\cite{Lansberg:2005ed}
888: \bibitem{Lansberg:2005ed}
889: J.~P.~Lansberg,
890: %``Ambiguities in the calculation of leptonic decays of excited heavy
891: %quarkonium,''
892: AIP Conf.\ Proc.\ {\bf 775} (2005) 11
893: [arXiv:hep-ph/0507184].
894: %%CITATION = HEP-PH 0507184;%%
895:
896:
897: \bibitem{barger} V.D. Barger and R.J.N. Philips, {\it Collider Physics},
898: Addison-Wesley, Menlo Park, 1987.
899:
900:
901: %\cite{Martin:2002dr}
902: \bibitem{Martin:2002dr}
903: A.~D.~Martin, R.~G.~Roberts, W.~J.~Stirling and R.~S.~Thorne,
904: %``NNLO global parton analysis,''
905: Phys.\ Lett.\ B {\bf 531} (2002) 216
906: [arXiv:hep-ph/0201127].
907: %%CITATION = HEP-PH 0201127;%%
908:
909:
910: %\cite{Pumplin:2002vw}
911: \bibitem{Pumplin:2002vw}
912: J.~Pumplin, D.~R.~Stump, J.~Huston, H.~L.~Lai, P.~Nadolsky and W.~K.~Tung,
913: %``New generation of parton distributions with uncertainties from global QCD
914: %analysis,''
915: JHEP {\bf 0207} (2002) 012
916: [arXiv:hep-ph/0201195].
917: %%CITATION = HEP-PH 0201195;%%
918:
919: \bibitem{article2} J.R. Cudell, Yu.L. Kalinovsky and J.P. Lansberg,
920: in preparation.
921:
922: %\cite{Hagler:2000dd}
923: \bibitem{Hagler:2000dd}
924: P.~Hagler, R.~Kirschner, A.~Schafer, L.~Szymanowski and O.~V.~Teryaev,
925: %``Towards a solution of the charmonium production controversy: k(T)
926: %factorization versus color octet mechanism,''
927: Phys.\ Rev.\ Lett.\ {\bf 86} (2001) 1446
928: [arXiv:hep-ph/0004263];
929: %%CITATION = HEP-PH 0004263;%%
930: %\cite{Hagler:2000eu}
931: %\bibitem{Hagler:2000eu}
932: P.~Hagler, R.~Kirschner, A.~Schafer, L.~Szymanowski and O.~V.~Teryaev,
933: %``Direct J/psi hadroproduction in k(T)-factorization and the color octet
934: %mechanism,''
935: Phys.\ Rev.\ D {\bf 63} (2001) 077501
936: [arXiv:hep-ph/0008316].
937: %%CITATION = HEP-PH 0008316;%%
938:
939:
940: \end{thebibliography}
941:
942: \end{document}
943: