1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%% EXAMPLE FILE
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FOR NESTEX 2.0
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
5:
6: \documentclass[proceedings]{JHEP3} % 10pt is ignored!
7: \PrHEP{ hep2001}
8:
9: \usepackage{epsfig,multicol} % please use epsfig.
10:
11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12: %
13: %%
14: %%%%%%%%%%%% Options: preprint*, proceedings, published, (no)hyper*,
15: %paper,%
16: %%%%%%%%%%%% a4paper*, letterpaper, legalpaper, executivepaper,
17: %%%
18: %%%%%%%%%%%% 11pt, 12pt*, oneside*, twoside, draft
19: %%%%%%%%%%%%%%%%
20: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% *=default
21: %%%%%%%%
22: %%%%%%%%%%%% \conference{...}
23: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
24: %%%%%%%%%%%% \title{...}
25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
26: %%%%%%%%%%%% \author{...\thanks{...}\\...} %%%%%%%%%%%% \email{...}
27: %%%%%%%%
28: %%%%%%%%%%%% \abstract{...}
29: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30: %%%%%%%%%%%% \keywords{...}
31: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32: %%%%%%%%%%%% \preprint{...} %% or \received{...} \accepted{...}
33: %%%%%\JHEP{...}%
34: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
35: %
36: %%
37: %%%%%%%%%%%% \dedicated{...}
38: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
39: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
40: %
41: %%
42: %%%%%%%%%%%% \aknowledgments
43: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
44: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45: %
46: %%
47: %%%%%%%%%%%% -- No pagestyle formatting.
48: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
49: %%%%%%%%%%%% -- No size formatting.
50: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
51: %%%%%%%%%%%% Your definitions: %%%%%%%%%%% These are mine... :)
52: %%%%%%%%%%%%
53: % ...
54: %
55: \newbox\mybox
56: \newcommand{\ttbs}{\char'134} % \backslash for \tt(Nucl.Phys.:)%
57: \newcommand\fverb{\setbox\mybox=\hbox\bgroup\verb}
58: \newcommand\fverbdo{\egroup\medskip\noindent\fbox{\unhbox\mybox}\ }
59: \newcommand\fverbit{\egroup\item[\fbox{\unhbox\mybox}]}
60: \newcommand{\openone}{\lea{\it%
61: vev}mode\hbox{\small1\kern-4.2pt\normalsize1}}
62: \newcommand{\rr}[4]{#1, {\it #2 \/}{\bf #3} #4}
63: \newcommand{\Tau}{{\cal T}}
64: \newcommand{\al}{\alpha}
65: \newcommand{\alef}{\al_{eff}'}
66: \newcommand{\bt}{\beta}
67: \newcommand{\gm}{\gamma}
68: \newcommand{\dl}{\delta}
69: \newcommand{\xd}{\dot{X}}
70: \newcommand{\vb}{\bar{v}}
71: \newcommand{\ub}{\bar{u}}
72: \newcommand{\tr}{\mbox{\rm tr }}
73: \newcommand{\taut}{\tilde{\tau}}
74: \newcommand{\qb}{\bar{q}}
75: \renewcommand{\th}{\theta}
76: \newcommand{\ph}{\phi}
77: \newcommand{\lra}{\longrightarrow}
78: \newcommand{\sg}{\sigma}
79: \newcommand{\sng}{{S_{NG}}}
80: %\newcommand{\dl}{\delta}
81: \newcommand{\zx}{z_x}
82: \newcommand{\var}[1]{\f{\dl \sng}{\dl {#1}}}
83: \newcommand{\wx}{w_x}
84: \newcommand{\alp}{\alpha'}
85: \newcommand{\xpr}{x_\perp}
86: \newcommand{\f}[2]{\frac{#1}{#2}}
87: \newcommand{\eq}{\begin{equation}}
88: \newcommand{\eqx}{\end{equation}}
89: \newcommand{\eqn}{\begin{eqnarray}}
90: \newcommand{\eqnx}{\end{eqnarray}}
91: \newcommand{\RO}{\rm l \! R }
92: \newcommand{\xpom}{x_{\PO} }
93: \newcommand{\kslash}{k\!\!\!/}
94: \newcommand{\pslash}{p\!\!\!/}
95: \newcommand{\qslash}{q\!\!\!/}
96: \newcommand{\GeV}{\mbox{\rm ~GeV~}}
97: \newcommand{\MeV}{\mbox{\rm ~MeV~}}
98: \newcommand{\Pom}{\rm I\!P}
99: \newcommand{\Reg}{\rm I\!R}
100: \newcommand{\modt}{\mid\!t\!\mid}\newcommand{\DD}{{\cal D}}
101: \newcommand{\gapprox}{\stackrel{>}{_{\sim}}}
102: \newcommand{\cor}[1]{\left\langle{#1}\right\rangle}
103: \renewcommand{\AA}{{\cal A}}
104: \renewcommand{\DD}{{\cal D}}
105: \newcommand{\lapprox}{\stackrel{<}{_{\sim}}}
106: \newcommand{\pom}{\rm I\!P}
107: \newcommand{\reg}{\rm I\!R}
108: \newcommand{\ttl}{\f{\tau^2 \th^2}{L^2}}
109: \newcommand{\Ttl}{\f{T^2 \th^2}{L^2}}
110:
111: %\newcommand{\ttbs}{\char'134}
112: \newcommand{\AmS}{{\protect\the\textfont2A\kern-.1667em\lower.5ex\hbox{M}\
113: kern-
114: 1
115: 25emS}}
116:
117: %\font\beeg=cmr17 scaled 1600 % Stylish initials
118: %\font\beeg=yinit scaled 800
119: %\newcommand\init[1]{\setbox\mybox=\hbox{{\beeg #1}~}%
120:
121: %\noindent\global\hangindent=\wd\mybox\global\hangafter-2%
122: % \sc\smash{\llap {\lower 13.2pt \box\mybox}}}
123: % ...
124: %
125: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
126: %
127: %%
128:
129: \title{Geometry of Reggeized amplitudes from AdS/CFT}
130:
131: \author{\speaker{Robi Peschanski}\\
132: CEA/DSM/SPhT,Unit\'e de recherche
133: associ\'ee
134: au CNRS, \\
135: CE-Saclay, F-91191 Gif-sur-Yvette Cedex,France\\
136: E-mail: \email{pesch@spht.saclay.cea.fr}}
137:
138: \conference{26$^{\rm th}$ Johns Hopkins Workshop}
139:
140: \abstract{
141: String theory has long ago been initiated by the quest for a theoretical
142: explanation of the observed high-energy ``Reggeization'' of strong
143: interaction amplitudes. In terms of quantum field theory, it is the
144: so-called ``soft'' regime, where the coupling constant is expected to be
145: large and thus perturbative calculations inadequate. However, since then,
146: no convincing derivation of the link between gauge field theory at strong
147: coupling and string theory has come out. This 35-years-old puzzle is
148: thus still unsolved. We discuss how modern tools like the AdS/CFT
149: correspondence give a new insight on the problem by applying it to two-body
150: elastic and inelastic scattering amplitudes. We obtain a geometrical
151: interpretation of Reggeization and its relation with confinement in gauge
152: theory.}
153:
154:
155: %\dedicated{Dedicated to\ldots\\if you want.}
156:
157:
158: \begin{document}
159:
160:
161: \section{Introduction}
162:
163:
164: \FIGURE{\epsfig{file=dual.eps,angle=0,width=12cm}%
165: \caption{``Duality diagrams'' for two-body inelastic and elastic amplitudes.}%
166: \label{1}}
167:
168: %\begin{figure}[hb]
169: %\includegraphics[width=30pc]{dual.eps}
170: %\vspace{-18pt}
171: %\caption{``Duality diagrams'' for two-body inelastic and elastic amplitudes.}
172: %\label{1}
173: %\end{figure}
174: %
175: It is well-known that string theory started from the proposal of
176: scattering amplitudes which may grasp the two major
177: structures of soft interaction phenomenology for $2\to 2$ reactions in a
178: condensed form:
179: resonances and Regge poles. Two types of amplitudes were proposed
180: for four-point amplitudes. The Veneziano amplitude
181: corresponds to {\it Reggeon} exchanges with non-vacuum quantum
182: numbers, {\it i.e.} inelastic two-body reaction amplitudes, and the
183: Shapiro-Virasoro
184: amplitude corresponds to {\it
185: Pomeron} exchange with vacuum quantum numbers, {\it i.e.} elastic
186: amplitudes. These amplitudes have been
187: conveniently represented by ``duality diagrams'', see Fig.\ref{1}. In the
188: representation of the Veneziano amplitude in terms of quark lines,
189: $q\bar q$ intermediate states in the direct $s$-channel correspond to the
190: resonances, and $q\bar q$ intermediate states in the exchanged
191: $s$-channel to Regge poles, where $s,t$ are the well-known Mandelstam
192: variables. The quark lines are to be considered as boundaries of a
193: surface
194: bearing no quantum numbers, as can be seen for the representation of the
195: Shapiro-Virasoro amplitude corresponding to no quantum nmber exchange.
196:
197:
198:
199: Quite amazingly, this representation found its justification in terms of
200: string theory. The ``duality diagram'' representation of Fig. \ref{1} can be
201: mapped into the fusing and splitting of strings. More precisely,
202: the Veneziano and Shapiro-Virasoro amplitudes are topologically related
203: to open and closed tree-level string respectively. This topological relation
204: can be more
205: formally expressed in string theory by the invariance of amplitudes with
206: respect to deformations of the world sheet spanned by the interacting
207: string, and by its conformal invariant properties. After a generalization of
208: dual amplitudes has been found for
209: multiparticle amplitudes, this raised the hope to find both a
210: theoretically
211: consistent
212: theory of strong interactions and a calculation procedure using the
213: perturbative topological expansion of string amplitudes \cite{Frampton}.
214:
215: However, despite many efforts during years, no widely recognized
216: progress has been done in the string theory of strong interaction
217: amplitudes. Moreover, after the discovery of QCD as the gauge field
218: theory of quarks and gluons and its validity for a quantitative
219: description of many ``hard'' scattering processes, there remained little
220: hope that a string theory of strong interactions could take place. Indeed,
221: even
222: before the discovery of QCD, major theoretical obstacles to the
223: formulation
224: of
225: a consistent string theory of strong
226: intractions were being
227: raised. Let us give a non exhaustive list of problems.
228:
229: The conformal anomaly of string theories in Minkowski $D$-dimensional
230: space leads to the limitation $D=26,10$ for respectively bosonic and
231: super strings in flat space. More generally, the requirement of conformal
232: and diffeomorphism invariance imposes stringent constraints on the space
233: in which the string moves.
234:
235: Zero mass gauge and gravitational fields appear in the string spectra
236: of
237: asymptotic states. Consistent string theory, when considered in flat
238: target space, contains (in general) gauge group and gravitational fields
239: and degrees of freedom. They are thus more suitable for
240: a
241: stringy approach of grand unified theories, than for strong interactions
242: and the confinement problem, characterized by the absence of asymptotic
243: zero mass states in the theory.
244:
245: To these difficulties, new ones have been added after the discovery
246: of
247: QCD. Let us list three among the main questions (at least those concerning
248: the
249: scattering of two particles):
250:
251: \begin{itemize}
252: \item {\it Can we find a consistent picture of the Reggeization of
253: high-energy amplitudes when QCD enters its strong coupling regime?}. Even
254: knowing the QCD lagrangian, it has not been possible to derive from it
255: high-energy amplitudes in the soft interaction case. Lattice calculations
256: have been useful for investigations at low energy but are seemingly
257: hopeless in the kinematical domain of high energies.
258: \item {\it Where are ``hard'' interactions recovered in a
259: string
260: theory framework?} String world sheets are suitable objects for the
261: description of ``soft'' phenomena due to their extended structure in
262: space
263: and time. It is more intricate to show off in string theory the ``hard''
264: structure visible in short-time interactions.
265: \item {\it Can we elaborate a suitable string theory which
266: could coherently describe the properties of gauge fields?}. The
267: degrees
268: of freedom and the symmetries of a gauge field theory are much different
269: from those commonly found for string theories. the matching of these two
270: require non trivial constraints as recognized in particular in
271: Ref.\cite{polyakov}.
272: \end{itemize}
273:
274: To the first of these three questions, the recently proposed duality
275: corespondence
276: between certain string and certain gauge field theories gives new and
277: reliable
278: answers. It seems that the second one can find some interesting clues
279: also in
280: the same framework
281: \cite{17}. we will focus here upon the third one, namely the
282: understanding
283: of Regge two-body amplitudes in gauge field theories at strong coupling
284: in
285: terms
286: of the AdS/CFT duality. This question has been the subject of an
287: approach
288: \cite{us1,us2,us4,us3} which I will now develop.
289:
290: The plan of the present review is the following: in section {\bf 2}, we
291: will
292: give a brief account of the AdS/CFT dual correspondence, focusing on
293: aspects
294: relevant for our study. In section {\bf 3}, we will explain the formalism
295: using the classical approximation and minimal surfaces for the determination of
296: the AdS duals of two-body scattering amplitudes. We successively apply it to
297: quark (anti)quark, dipole-dipole elastic scattering and finally inelastic
298: dipole-dipole scattering. Next in {\bf 4}, we will develop a semi-classical
299: approximation, improving the previous method by computing the fluctuation factor
300: around the minimal surface solution. A final section {\bf 5} is devoted to a
301: summary and conclusion about the geometrical nature of Reggeization in the
302: confining AdS dual backgrounds.
303:
304:
305: \section{String/Gauge fields Duality}
306:
307: The AdS/CFT correspondence \cite{11} has many interesting formal and
308: physical facets. Concerning
309: the
310: aspects which are of interest for our problem, it allows one to find
311: relations
312: between gauge field theories at strong coupling and string gravity at
313: weak
314: coupling in the limit of large number of colours ($N_c\!\to\! \infty$). It can
315: be examined quite precisely in the
316: AdS$_5$/CFT$_4$
317: case which conformal field theory corresponds to $SU(N)$ gauge
318: theory
319: with ${\cal N} \!=\!4$ supersymmetries.
320:
321: Some existing extensions to other gauge theories with broken conformal
322: symmetry
323: and less or no supersymmetries will be valuable for our approach, since
324: they
325: lead to confining gauge theories which
326: are
327: more similar to QCD\footnote{Note that the
328: appropriate
329: string gravity dual of QCD has not yet been identified, and thus we are
330: forced
331: to restrict for the moment our use of AdS/CFT correspondence to
332: features
333: which
334: are
335: expected to be a general feature of confining theories duals, see a
336: discussion
337: further on in this section.}. Indeed, one important question is to examine to
338: what
339: extent confinement plays a r\^ole in the Reggeization of
340: amplitudes. Our aim is thus to investigate the possible
341: realization and origin of
342: Reggeization of two-body amplitudes in such theories and what are the
343: difference
344: appearing with the original AdS$_5$/CFT$_4$ case.
345:
346:
347:
348: \FIGURE[ht]{\epsfig{file=ads1.eps,angle=0,width=15cm}%
349: \caption{AdS$_5$/CFT$_4$ duality correspondence.}%
350: \label{2}}
351:
352:
353:
354: Let us recall the canonical derivation leading to the AdS$_5$
355: background \cite{review},
356: see
357: Fig.\ref{2}. One starts from the (super)gravity classical
358: solution of a system of $N\ D_3$-branes in a $10\!-\!D$ space of the (type
359: IIB)
360: superstrings.
361: The metrics solution of the (super)Einstein equations read
362: \eq
363: \label{super}
364: ds^2=f^{\!-\!1/2} (\!-\!dt^2\!+\!\sum_{1\!-\!3}dx_i^2)
365: \!+\!f^{1/2}(dr^2\!+\!r^2d\Omega_5)
366: \ ,
367: \eqx
368: where the first four coordinates are on the brane and $r$ corresponds to
369: the
370: coordinate along the normal to the branes. In formula (\ref{super}), one defines
371: \eq
372: f=1+\frac {R^4}{r^4}\ ;\ \ \ \ \ R=4\pi g^2_{YM}\alpha'^2 {N} \ ,
373: \label{R}
374: \eqx
375: where $g^2_{YM}{N}$ is the `t Hooft-Yang-Mills coupling and $\alpha'$ the
376: string
377: tension.
378:
379: One considers the limiting behaviour considered by
380: Maldacena,
381: where
382: one zooms on the neighbourhood of the branes while in the same time going to
383: the limit of
384: weak
385: string
386: slope $\alpha'.$ The near-by space-time is thus distorted due to the
387: (super)
388: gravitational field of the branes. One goes to the limit where
389: \eq
390: \ R\ fixed\ ; \frac {\alpha'(\to 0)} {r (\to 0)}\to z\ fixed\ .
391: \eqx
392: This, from the second equation of (\ref{R}) obviously implies
393: \eq
394: {\alpha'\to 0}\ , \ g^2_{YM}
395: {N}\sim \f{1}{\alpha'^2}\to \infty \ ,
396: \eqx
397: {\it i.e.} both a weak coupling limit for the string theory and a strong
398: coupling limit for the dual gauge field theory.
399: By reorganizing the two parts of the metrics one obtains
400: \eq
401: ds^2={ \frac 1{z^2} (-dt^2+\sum_{1-3}dx_i^2+ dz^2)} + {R^2
402: d\Omega_5}\ ,
403: \label{AdS}
404: \eqx
405: which corresponds to the
406: {AdS$_5$} $\times \ {S_5}$ background structure, ${S_5}$ being
407: the 5-sphere. More detailed analysis shows that the isometry group of
408: the
409: 5-sphere is the geometrical dual of the ${\cal N}\! =\!4$
410: supersymmetries. More intricate is the quantum number dual to $N_c,$ the number
411: of
412: colours, which is the invariant charge carried by the Ramond-Ramond form
413: field.
414:
415: In the case of confining backgrounds, an intrinsic scale breaks conformal
416: invariance and is brought in the dual theory through {\it e.g.} a
417: geometrical constraint. For instance in~\cite{wi98} a proposal was made
418: that a
419: confining gauge theory is dual to string theory in an $AdS_{BH}$ black
420: hole (BH) background
421:
\eq
422: \label{e.bhmetric}
423: ds^2_{BH}=\f{16}{9}\f{1}{f(z)}\f{dz^2}{z^2} + \f{\eta_{\mu\nu}dx^\mu
424: dx^\nu}{z^2} + \ldots
425: \eqx
426: where $f(z)=z^{2/3}(1-(z/R_0)^4)$ and $R_0$ is the position of
427: the horizon.
428: We will use this
429: background\footnote{Although it was later found that the $S^1$ KK states do not
430: strictly
431: decouple ~\cite{de99}.}
432: to
433: study the interplay between the confining nature of gauge theory and
434: its reggeization properties. Actually the qualitative
435: arguments and
436: approximations should be generic for most confining backgrounds, as
437: already
438: discussed in Ref.~\cite{so99}.
For instance,
439: other geometries for (supersymmetric)
440: confining theories \cite{ks,carlo} have been discussed in this respect.
441: They have the property that for
442: small $z$, i.e. close to the boundary, the geometry looks like
443: $AdS_5\times
444: S^5$ (in \cite{carlo} up to logarithmic corrections related to asymptotic
445: freedom) giving a coulombic $q\qb$ potential. For large $z$ the
446: geometry is effectively flat. In all cases there is a scale, similar
447: to $R_0$ above, which marks a transition between the small $z$ and
448: large $z$ regimes.
449:
450:
451: In order to illustrate the way one formulates the AdS/CFT
452: correspondence in a context similar to QCD,
453: let us consider the example of the vacuum expectation value ({\it vev})
454: of
455: Wilson lines in a configuration parallel
456: to the time direction of the branes. This configuration
457: allows
458: a determination of the
459: potential between colour charges \cite{12}. The r\^ole of colour charges
460: in the fundamental representation is played by open string states
461: elongated
462: between a stack of $N_c$ $D_3$ branes on one side and one $D_3$ brane
463: near the
464: boundary of AdS space ({\it cf.} \cite{review}).
465:
466:
467: \FIGURE[hb]{\epsfig{file=horizon.eps,angle=0,width=12cm}%
468: \caption{Exemple of minimal surfaces with Wilson line boundaries.}%
469: \label{3}}
470:
471:
472: One writes
473: \eq
474: \langle e^{iP\int_C\vec A\cdot\vec dl}\rangle
475: \!=\!\!\int_{\Sigma}\!e^{-
476: \frac{Area({\Sigma})}{\alpha'}}\!\!\approx e^{-
477: \frac {Area_{min}}{\alpha'}}\times
478: Fluct.\ ,
479: \eqx
480: where $C$ is the Wilson line contour near the $D_3$ branes and $\Sigma$
481: the
482: surface in $AdS$-space with $C$ as the boundary, see Fig.\ref{3}. The
483: minimal
484: area
485: approximation is the {\it vev} evaluation in the classical $\alpha'\to 0$
486: limit. The factor denoted $Fluct.$ refers to the fluctuation
487: determinant
488: around the minimal surface, corresponding to the first one-loop (in
489: $\sigma$-model sense) quantum correction. It gives a calculable semi-classical
490: correction.
491:
492: In Fig.\ref{3}, we have
493: sketched the form of minimal surface solutions for the
494: ``confining'' $AdS_{BH}$
495: case, (see
496: above (\ref{e.bhmetric})). For large separation of Wilson lines, the minimal
497: surface ``feels'' the horizon and is consequently curved. At smaller separation,
498: the solution becomes again similar to the conformal case, since the horizon
499: cut-off does not play a big r\^ole.
500:
501:
502: The {\it vev} results in the classical approximation can be summed up as
503: follows:
504: \eqn
505: AdS_5: \langle Wilson\ Lines\rangle&=&e^{TV(L)}\sim
506: e^{\#_1T/L}\nonumber \\ AdS_{BH}: \langle Wilson\
507: Lines\rangle&=&e^{TV(L)}\sim e^{\#_2TL/R_0^2}\ , \nonumber
508: \eqnx
509: where, the potential behaviour is as expected for respectively conformal
510: (perimeter law) and confining (area law) cases. Note that there is an
511: interesting information stored in the coupling dependent numbers
512: here denoted by $\#_{1,2}$ . Note also that, even in the case of a
513: confining
514: geometry with an horizon at $R_0,$ Wilson lines separated by a distance
515: $L<<R_0
516: $ do not give rise to minimal surfaces sensitive to the horizon (see
517: Fig.3),
518: and thus give rise to classical solutions similar to the non-confining case.
519:
520: The important r\^ole of fluctuation corrections and the way of computing
521: it
522: in some non-trivial cases will be discussed further on.
523:
524: \section{Supergravity Duals of Scattering Amplitudes}
525:
526: Using the AdS/CFT correspondence, we find that two-body high energy amplitudes
527: in gauge field
528: theories can
529: be
530: related to specific configurations of
531: minimal surfaces\footnote{A different approach has been independently
532: proposed in Ref.\cite{13}.}.
533:
534: Indeed, at high energy, fast moving colour sources
535: propagate
536: along
537: linear trajectories in coordinate space thanks to the eikonale
538: approximation. An analytic
539: continuation
540: from Minkowski to Euclidean ${\cal R}^4$ space allows one to find a
541: geometrical interpretation in terms of a well-defined minimal
542: surface
543: problem. Let us
544: consider for illustration different applications.
545:
546: \subsection{Quark-quark elastic scattering}
547:
548:
549: \FIGURE[ht]{\epsfig{file=lines.eps,angle=0,width=5cm}%
550: \caption{Wilson lines for quark-quark elastic scattering in
551: ${\cal
552: R}^4.$}%
553: \label{4}}
554:
555: %
556: In a framework suitable for performing the AdS/CFT correspondence, quarks
557: (resp. antiquarks) can be represented by colour sources in the fundamental
558: (resp.
559: anti fundamental)
560: reps. of SU(N). In the brane world, they are obtained ({\it e.g.}
561: see
562: \cite{review}) by considering systems of $N\!+\!1\ D_3$ branes of which one of
563: the
564: brane is removed to a distance from the remaining stack of $N\ D_3$
565: branes. This
566: distance is large (or equivalently $z$ small) in order to satisfy a static
567: approximation for the quarks considered as ultra-massive.
568:
569: In the corresponding gauge field theoretical framework, it is knowen
570: since a
571: long time \cite{14} that the high-energy elastic quark-(anti)quark
572: amplitude can be formulated as follows
573: \eqn
574: A(s,q^2)=2is\int d\vec l\ e^{i\vec q\cdot \vec l}\ \langle W_1
575: W_2\rangle
576: _{L=|\vec l|}^{\chi=\log s/m^2}\nonumber\\
577: = 2is\int d\vec l\ e^{i\vec q\cdot \vec
578: l-
579: \frac {1}{\alpha'}Area_{min}(\vec l)}\ ,
580: \label{e.anal}\eqnx
581: where $\vec l$ is the impact parameter vector between the two
582: trajectories,
583: conjugated to the momentum transfer $\vec q,$ and $\chi=\log s/m^2$ the
584: total
585: rapidity
586: interval. Performing an analytic
587: continuation from Minkovskian to Euclidean space \cite{meggio}:
588: \eq
589: \chi \to i\theta\ \ \ ;\ \ t_{Mink}\to -it_{Eucl}\ ,
590: \label{anal}
591: \eqx
592: the Wilson line {\it vev} can be expressed as a minimal surface problem
593: whose
594: boundaries are two straight lines in a 3-dimensional coordinate space,
595: placed at
596: an impact parameter distance $L$ and rotated one with respect to the other by
597: an
598: angle
599: $\theta,$
600: see
601: Fig.4. In flat space, with the same boundary conditions, the
602: minimal surface is
603: the
604: {\it helicoid}. One thus realizes that the problem can be formulated as a
605: minimal surface problem whose mathematically well-defined solution is
606: a
607: {\it generalized helicoidal} manifold embedded in curved background
608: spaces,
609: such as Euclidean AdS Spaces. Unfortunately, this problem is rather
610: difficult
611: to
612: solve analytically, even in flat space. It is known as the Plateau
613: problem,
614: namely the determination of minimal surfaces for given boundary conditions
615: (see
616: for instance \cite{minimal}).
617:
618:
619: Thus, the interest of considering quark-quark scattering relies on the
620: simple
621: definition of the minimal surface geometry in the conditions of a
622: confined
623: $AdS_{BH}$ metrics (\ref{e.bhmetric}). Indeed, in the
624: configuration
625: of Wilson lines of Fig.\ref{3} in the context of a confining theory, the
626: AdS$_{BH}$ metrics is characterized by a singularity at $z=0$ which
627: implies a
628: rapid growth in the $z$ direction
629: towards the D$_3$ branes, then stopped near the horizon at $z_0.$ Thus,
630: to a good approximation, and for large enough
631: impact
632: parameter (compared to the horizon distance), the main contribution to
633: the
634: minimal area is from the metrics in the bulk near $z_0$ which is nearly
635: flat. Hence, near $z_0,$ the relevant minimal area can be drawn on a
636: classical helicoid which can be parametrized:
637: \eqn
638: t &=& \tau \cos {\theta\sg}/{L}\nonumber\\
639: y &=& \tau \sin {\theta\sg}/{L}\nonumber\\
640: x &=& \sg\nonumber\\
641: z &\sim& z_0\ .
642: \label{helico}
643: \eqnx
644:
645: However, the practical calculation \cite{us1} of the amplitude is
646: complicated
647: by the necessity of introducing a cut-off in the $T$-direction (see
648: Fig.4). This
649: is physically expected since the area spanned by the helicoid in the
650: confining
651: geometry goes to infinity, corresponding to the spreading of the
652: color
653: field between the quarks, the confining forces increasing till the string
654: breaks
655: for the production of particles, not described in the present scheme. It
656: is the
657: expected counterpart, in QCD, of the infinite phase of electron-electron
658: scattering in QED. Let us sketch the calculations of \cite{us1}.
659:
660:
661:
662: The truncated
663: helicoid solution is parametrized by (\ref{helico})with $\tau=-T\ldots T,$
664: $\sg=0 \ldots L$ and $\th$ is the total
665: twisting angle.
666: Its area is given by the formula
667: \eq
668: \label{e.ahelic}
669: Area =\int_{0}^{L}d\sigma \int_{-T}^{T}d\tau \sqrt{1+\ttl}=
670: LT\sqrt{1+\Ttl } +\f{L^2}{2\th} \log
671: \left\{\f{\sqrt{1+\Ttl}+\th \f{T}{L}}{\sqrt{1+\Ttl}-\th \f{T}{L}}\right\}
672: \ .
673: \eqx
674: Through the analytical continuation (\ref{anal}), one would naively obtain
675: a
676: pure $T$-dependent phase factor going to $\infty$ when removing the
677: cut-off.
678: However the
679: analytic structure of the euclidean area (\ref{e.ahelic}) involves
680: cuts in the complex $T$, $\th$ planes and thus leads to an ambiguity
681: coming from the branch cut of the
682: logarithm. In fact when performing the analytical continuation we have
683: to specify the Riemann sheet of the logarithm (i.e. $log \to log+2\pi
684: i n$).
685:
686: This leads \cite{us1} to a $T$-independent real
687: term which, inserted in formula (\ref{e.anal}), gives rise to a reggeized
688: amplitude. Doing this, the
689: removal we make of the (infinite as $T\to \infty$)
690: phase could be considered as
691: a
692: natural infrared regularisation. One finally obtains
693: \eq
694: \label{e.ai}
695: A_{\cal P}(s,q^2)= 2is\int d\vec l\ e^{i\vec q\cdot \vec l-\left\{ n
696: \f{\sqrt{2g_{YM}^2N}}{\chi} \f{L^2}{2R_0^2}\right\}} \propto
697: s^{1-q^2\f{R_0^2}{n\sqrt{8g_{YM}^2N}}}\ .
698: \eqx
699: which represents a ($n$-dependent) set of Reggeized elastic
700: amplitudes, with linear Regge trajectories characterized by a Regge
701: intercept 1
702: and Regge slopes given by $\f{R_0^2}{n\sqrt{8g_{YM}^2N}}$. In this
703: framework the
704: removal of the (infinite as $T\to \infty$) phase could be considered as a
705: natural infrared regularisation. We will see now how this assumption,
706: related to
707: the consideration of unphysical asymptotic quark states, can be relaxed
708: without
709: affecting the Reggeization property, when considering scattering between
710: colorless dipoles
711: \vspace{2cm}.
712:
713:
714: %
715: \begin{figure}[ht]
716: \begin{center}
717: \includegraphics[width=16pc]{untitledc.eps}
718: \includegraphics[width=18pc]{untitled2c.eps}
719: \vspace{2cm}
720: \end{center}
721: \label{45}
722: \end{figure}
723: %
724: \vspace{7cm}.
725: {\bf Figure 4bis}: Helicoids describing quark-quark and quark-antiquark
726: scattering.
727: \vspace{2cm}
728:
729:
730: It is interesting to note that the realization of Reggeization provided by the
731: helicoid geometry through analytic continuation also gives a natural
732: interpretation of the ``signature'' factor, {\it i.e.} the phase factor
733: distinguishing quark quark scattering and quark antiquark
734: scattering in Regge amplitudes. Indeed, quark quark scattering and quark
735: antiquark
736: scattering are related through twisted helicoidal configurations in the bulk
737: coordinate
738: space. For a given helicoid configuration representing quark quark
739: scattering, twisting one of the a quark lines will give rise to the helicoid
740: representing quark antiquark
741: scattering with the same kinematics, see Fig.4bis. Hence, through analytic
742: continuation, one
743: finds
744: \eq
745: \th \to \th + \f{\pi}2\ \Rightarrow \ s \to s\ e^{-i\pi}\ ,
746: \label{signature}
747: \eqx
748: which, once inserted in formula (\ref{e.ai}), gives rise precisely to the
749: Regge phase signature factor.
750:
751:
752: In the non-confining $AdS_5 \times S_5$ case, one would need to identify
753: a
754: {\it generalized helicoidal} manifold embedded with the metrics
755: (\ref{AdS}), {\i.e.} the minimal surface with straight line boundaries with an
756: angle.
757: This is a well-defined mathematical problem, which is yet not solved.
758: With
759: some crude approximation however, looking for a variational solution with
760: $z_0
761: \to z(\sigma,\tau)$ in the parametrization (\ref{helico}), one can obtain
762: \cite
763: {us2} a solution at large $\log s \ :$
764: \eq
765: \label{e.lcft}
766: A(s,q^2) \propto \left(\f{L}{\log s}\right)^{n\f{F(\pi/2)}{\pi} \
767: \f{\sqrt{2g^2_{YM}N}}{2\pi}}
768: \ s^{n\f{ 2\pi^4}{\Gamma(1/4)^4} \ \f{\sqrt{2g^2_{YM}N}}{2\pi}} \ ,
769: \eqx
770: where $F(\Omega)$ is the ``cusp anomalous dimension'' calculated in \cite{neww}
771: ($F(\pi/2)\sim 0.3 \pi$). It is interesting to note that in this case,
772: there is
773: no Reggeization, at least with a non-zero Regge slope. Formula
774: {\ref{e.lcft}}
775: appears as an extension of the weak coupling result \cite{bfkl} including
776: a
777: screening effect on the coupling ($g^2_{YM}N \to \sqrt{g^2_{YM}N}$) when
778: it
779: becomes strong, as for the potential in the conformal case \cite{12}.
780: %
781: \subsection{Dipole-dipole elastic scattering}
782: \label{dub}
783:
784:
785: \FIGURE{\epsfig{file=wilson3.eps,angle=0,width=5cm}%
786: \caption{Wilson contours for Dipole-dipole elastic scattering.}%
787: \label{fig}}
788:
789:
790:
791: Elastic scattering of colourless states is expected to be cured from the
792: infra-red divergent phase factor encountered in quark-(anti)quark
793: scattering. In
794: this respect, it is interesting to
795: consider the elastic interaction of two very massive
796: QCD dipoles. Thanks to their high mass (or equivalently small size $a$), one
797: can
798: neglect the fluctuations around their classical trajectory and thus their
799: propagation
800: in
801: coordinate space in the eikonale approximation can be
802: represented by elongated Wilson loops near both right and left moving
803: light-cone
804: directions. More precisely \cite{us1}, one has to
805: compute a Wilson loop correlator
806: in the configuration displayed in
807: Fig.\ref{fig}.
808: \eq
809: \label{e.ampinit}
810: A(s,q^2)=-2is \int d^2\xpr e^{iq\xpr}
811: \cor{\f{W_1W_2}{\cor{W_1}\cor{W_2}}-1}
812: \eqx
813: where the Wilson loops follow classical straight lines for
814: quark(antiquark)
815: trajectories: $W_1\lra x_1^\mu=p_1^\mu\tau\,(+a^\mu)$ and $W_2\lra
816: x_2^\mu=\xpr^\mu+p_2^\mu\tau\, (+a^\mu)$ and close at infinite times. The
817: normalization ${\cor{W_1}\cor{W_2}}$ of the correlator ensures that the
818: amplitude vanishes when
819: the
820: Wilson loops get decorrelated at large distances.
821: Let us consider the solution in the confining background (\ref{e.bhmetric}),
822: approximated by a flat metrics near the horizon. For this setup we have to
823: calculate the correlation function of two
824: Wilson loops
825: elongated along the ``time''
826: direction and have a large but arbitrary temporal length $T$
827: (the exact analogue for Wilson loops of $T$ considered in
828: the previous section). However, the cut-off
829: dependence on $T$ is removed and thus
830: the related IR divergence which was present for the $q(\bar q)-q$ scattering
831: case. Indeed, for large positive and negative times the minimal
832: surface will be well
833: approximated by two seperate copies of the standard minimal surfaces for
834: each loop separately. When we come to the interaction region, and for $L$
835: sufficiently small, one can lower the area by forming a ``tube'' joining
836: the two worldsheets.
837: Since we want to calculate the normalized correlator $\cor{W_1
838: W_2}/\cor{W_1}\cor{W_2}$, the contributions of the regions outside the
839: tube will cancel out (in a first approximation neglecting deformations
840: near the tube). Therefore we have just to find the area
841: of the tube, and subtract from it the area of the two independent
842: worldsheets. It is at this stage that we see that the result does not
843: depend on the maximal length of the Wilson loops $T$, and hence is IR
844: finite. The whole contribution to the amplitude will just come from
845: the area of the tube.
846:
847:
848: Our calculation scheme proposed in \cite{us2} goes as follows. Since one
849: does
850: not know the explicit minimal surface for these boundary
851: conditions, let us perform a variational approximation. Namely we
852: consider a family of surfaces forming the tube, parameterized by
853: $T_{tube}$,
854: which has the interpretation of an ``effective'' time of
855: interaction. Then we make a saddle point minimization of the area
856: as a function of this parameter.
857:
858: Suppose that the tube linking the two Wilson lines is formed in the
859: region of the time parameter
860: $t\in$ $(-T_{tube},T_{tube}).$ In our
861: approximation its two
862: ``sides'' are formed by a duplication of the helicoid solution . The front and
863: back
864: will
865: be each approximated by strips of area $aL \sqrt{1+\f{T^2_{tube}
866: \th^2}{L^2}}$ (we assume $a,L \geq R_0$).
867:
868:
869: The total area corresponding to the two Wilson loops is then given by
870: \eq
871: \label{e.tube}
872: Area(T_{tube}) = 2L\int_{-T_{tube}}^{T_{tube}}d\tau \sqrt{1+
873: \ttl} +2aL \sqrt{1+\f{T^2_{tube} \th^2}{L^2}} -4a
874: \cdot T_{tube} \ ,
875: \eqx
876: where $-2 a T_{tube}$ is the contribution of each individual Wilson loop
877: to the normalization $1/\cor{W_1}\cor{W_2}$ of the Wilson loop
878: correlation function.
879:
880: Analytically continuing the area formula (\ref{e.tube}) to the
881: Minkowskian case and using a convenient change of variables, the
882: Minkowskian area can be put in the following simple form
883: \eq
884: \label{e.tubeminkow}
885: Area(T_{tube}) = \f{2L^2}{\chi} \left\{ \phi+\f{\sin 2\phi}{2}+\rho
886: \chi \cos\phi -2\rho \sin \phi \right\} \ ,
887: \eqx
888: where $\rho\equiv a/L$ and $\sin \phi=i\chi\, T_{tube}/L$ is
889: the new variational parameter.
890:
891: In the strong coupling limit ($\al' =1/\sqrt{2g^2_{YM}N} \!\to\! 0$)
892: the parameter $\phi$ is dynamically determined from the saddle point
893: equation:
894: \eq
895: \label{e.sp}
896: 0=\f{\partial Area(\phi)}{\partial \phi}=
897: \cos\phi(\cos\phi-\rho)-\f{\rho \chi}{2}\sin\phi
898: \eqx
899: It is easy to realize that for large enough energy, the last term dominates and
900: thus $\phi \sim \pm n\pi$. Inserting this solution into the area
901: (\ref{e.tubeminkow}) we find
902: \eq
903: \label{e.aphi}
904: Area(\phi)=-\f{2L^2}{\chi} n\pi+2aL (-1)^n
905: \eqx
906: where we retain the physical solutions with $n$ positive integer. We
907: thus find a set of solutions very similar to the inelastic factor
908: obtained in the previous section. The modification due to the front-back
909: contribution $2aL$ is negligible in the Fourier transformed amplitude
910: for momentum transfer $\sqrt{q^2} \gg a/R_0^2$. Also this term is
911: probably more dependent on the treatment of the front-back parts of
912: the tube in our approximation.
913:
914:
915: \FIGURE{\epsfig{file=corr.eps,angle=0,width=13cm}%
916: \caption{Dipole-dipole scattering at large impact parameter.}%
917: \label{5}}
918:
919:
920: Concerning the non-confining metrics $AdS_5 \times S_5,$ the minimal area
921: solution with the corresponding boundary conditions
922: is difficult to find in analytic form, necessary for the continuation to
923: Minkowski space. However, there exists a generic and intringuing
924: feature: the
925: existence of a geometrical
926: transition
927: between small and large impact parameter, corresponding to the
928: realization
929: of the minimal surface by two
930: disconnected ones, see Fig.\ref{5}, where each of them reproduces the
931: known
932: solution used for the calculation of the interquark potential \cite{12}.
933:
934:
935: Taking advantage of this unique configuration, valid at large enough
936: impact
937: parameter distance, it is possible \cite{us1} to compute the elastic
938: amplitude
939: {\it via} the supergravity approximation of the AdS string theory at small
940: curvature. The amplitude is dominated by the exchange contribution of all
941: zero-mass excitations of the appropriate supergravity theory, namely,
942: Kaluza-Klein scalars, dilaton, antisymmetric tensor and graviton. All in
943: all,
944: the graviton dominates at large energy as $A(s,L) \propto s \times L^{-6}$
945: in
946: the amplitude, but the region of validity of the supergravity
947: approximation
948: requires a condition $L\gg s^{2/7} $ or $A(s,L) \ll s^{-5/7}$
949: which
950: lies significantly below the absolute unitarity limit $A(s,L) < {\cal
951: O}(1).$
952: As expected the behaviour at large $L$ is power-like and, for fixed $s$ is
953: found
954: dominated by the KK scalar tail in $s \ L^{-2}.$
955:
956:
957: \subsection{Dipole-dipole inelastic scattering}
958: \label{rub}
959: The application of AdS/CFT correspondence for the two previous exemples
960: is
961: not
962: so easy, even if partial results are encouraging. For ``quark'' elastic
963: scattering, an infra-red time-like cut-off is to be introduced due to
964: the
965: colour
966: charges of the quarks which implies a regularization scheme and a
967: complication
968: of the geometrical aspects. For dipole
969: elastic scattering, there is no need for a cut-off but the geometry of
970: the
971: minimal surface is complicated. Inelastic scattering of dipoles allows
972: one
973: to
974: circumvent both of these difficulties. Indeed, the helicoidal geometry remains
975: valid
976: due
977: to the eikonale approximation for the ``spectator quarks''
978: while the ``exchanged quarks'' define a trajectory drawn on the
979: helicoid, see
980: Fig.\ref{6}. This trajectory plays the r\^ole of a dynamical time-like
981: cut-off which takes part in the
982: minimization
983: procedure.
984:
985:
986: \FIGURE{\epsfig{file=scatt.eps,angle=0,width=10cm}%
987: \caption{Wilson lines for inelastic dipole scattering.}%
988: \label{6}}
989:
990:
991:
992: Following the approach of \cite {us4} let us consider a meson-meson
993: scattering process
994: \eq
995: \label{e.scat}
996: (11')+(22') \lra (33')+(44'),
997: \eqx
998: where the continuous lines 1-3, 2-4 correspond to spectator quark and
999: antiquark,
1000: while the dashed lines 1'-2', 3'-4' correspond to annihilated and produced
1001: quark-antiquark pairs. The labels correspond to the initial and final
1002: spacetime position 4-vectors that we fix for our calculation.
1003:
1004: The spacetime picture of this process is schematically illustrated
1005: in figure (\ref{6}), where the impact parameter axis $x$ is perpendicular
1006: to the
1007: longitudinal $t-y$ plane. Note that the impact parameter is defined w.r.t.
1008: the
1009: spectator quark asymptotic trajectories.
1010:
1011: The amplitude
1012: corresponding to the scattering process (\ref{e.scat}) can be
1013: schematically written as
1014: \eq
1015: \label{e.ampgen}
1016: \cor{ <\!\!out|
1017: S_F(3',4'|\AA)\, S_F(1,3|\AA)\, S_F(4,2|\AA)\, S_F(2',1'|\AA)
1018: |in\!\!> }_\AA
1019: \eqx
1020: where $<\!\!out|$ and $|in\!\!>$ are wavefunctions for the outgoing
1021: and incoming mesons (up to modifications due to LSZ reduction
1022: formulae). In formula (\ref{e.ampgen}),
1023: $S_F(X,Y|\AA)$ denotes the
1024: full quark propagator between spacetime points $X$ and $Y$
1025: in a given background gauge field configuration $\AA$,
1026: while the correlation function $\cor{\ldots}_\AA$ stands for averaging
1027: over these configurations.
1028:
1029: Let us first perform the calculations for the above
1030: scattering amplitude rotated into Euclidean space.
1031: In impact parameter space we use
1032: the worldline expression for the (Euclidean) fermion propagator in a
1033: background gauge field $\AA = A^C_\mu(X^\mu)$ as a path
1034: integral over classical trajectories \cite{15}:
1035: \eq
1036: \label{e.s}
1037: S_F(X,Y|\AA)= \int_0^\infty dT e^{-m T} \int \DD X^\mu(\tau)
1038: \dl(\xd^2\!-\!1)\, I[X^\mu(\tau)]\, P e^{i\int A_\mu(X(\tau))\cdot \xd^\mu
1039: d\tau}
1040: \eqx
1041: Here the path integral is over trajectories $X^\mu(\tau)$ joining $X$
1042: and $Y$, parametrized by $\tau\in (0,T)$. Because of the delta
1043: function, $T$ is also the total length of the trajectory.
1044: The quark mass dependence appears in the first exponential.
1045: The colour and gauge field dependence is encoded in the
1046: (open) Wilson line along the trajectory $P e^{i\int A_\mu(X(\tau))\,
1047: \xd^\mu d\tau}$, while the spin 1/2 character
1048: of the quark is responsible for the appearance of the spin factor:
1049: \eq
1050: \label{e.sfdef}
1051: I[X^\mu(\tau)]=P \prod \f{1\!+\!\xd^\mu \gm_\mu}{2}
1052: = \lim_{N\to \infty}
1053: \f{1\!+\!\xd^\mu(T) \gm_\mu}{2} \ldots \f{1\!+\!\xd^\mu(\f{2}{N}T)
1054: \gm_\mu}{2} \f{1\!+\!\xd^\mu(\f{1}{N}T) \gm_\mu}{2}
1055: \eqx
1056: where the second equality gives a suitably regularized definition of the
1057: infinite
1058: product
1059: along the trajectory $X^\mu(\tau)$.
1060: Note that each of the $N$ factors in this expression is
1061: a projector due to the fact that $\xd^2=1$.
1062: This spin factor was first formulated for D=3 and later
1063: for arbitrary D \cite{15}.
1064: In practice it was computed explicitly in D=2 and D=3, but not in
1065: general for $D>3$. We computed it \cite{us4} for the
1066: configuration of figure (\ref{6}), {\it i.e.} in a $D=3$ submanifold in
1067: $D=4$ spacetime.
1068:
1069: Let us comment two important steps \cite{us4} of the calculation of
1070: (\ref{e.ampgen}).
1071:
1072: {\bf i)} Since the initial and final mesons are colour singlets, the four
1073: Wilson lines close to
1074: form a single Wilson loop, and the gauge
1075: averaging factorizes out of the expression:
1076: \eq
1077: \cor{\tr P e^{i\int_C \vec A \cdot d \vec X}}_\AA
1078: \eqx
1079: where the contour $C$ follows the quark trajectories $1\to 3' \to
1080: 4' \to 2' \to 1'$ (following the contours sketched on Fig.\ref{6}). Hence,
1081: adopting
1082: the ``world-line'' path integral scheme of Feynman \cite{15},
1083: one may write
1084: the
1085: inelastic amplitude in terms of a Wilson loop {\it vev}:
1086: \eq
1087: \int \DD\tau\, \cor{W(1\!\to\! 3'\! \to \!4'\! \to \!2' \!\to \!1')}_
1088: {\AA,\tau}
1089: \ e^{ -2m {\cal L}(\tau)} \ ,
1090: \label{tau}
1091: \eqx
1092: where $\tau$ parametrizes the boundary trajectories and ${\cal L}$ is
1093: their
1094: total length. Using the AdS-CFT correspondence in the same framework as
1095: previously,
1096: one may formally integrate over the gauge degrees of freedom and
1097: write
1098: \eq
1099: \cor{\tr P e^{i\int_C \vec A \cdot d \vec X}}_\AA \equiv \cor{W(1\!\!\to\!\!
1100: 3'\!\! \to \!\!4'\! \!\to\! \!2' \!\!\to
1101: \!\!1')}_{\AA,\tau}\!\! =\!
1102: e^{-\!\f{Area(\tau)}{2\pi\alpha'} } \times Fluct(\tau).
1103: \eqx
1104: Note that the remaining minimization of (\ref{tau}) in $\tau$ runs now on both
1105: the
1106: area and its
1107: boundary.
1108:
1109: {\bf ii)} The spin factor matrices multiply
1110: \eq
1111: I[1\!\to\! 3]_{\al_1\al_3}\, I[4\!\to\! 2]_{\al_4\al_2}\,
1112: I[2'\!\to\! 1']_{\al_{2'}\al_{1'}}\, I[3'\!\to\! 4']_{\al_{3'}\al_{4'}}
1113: \eqx
1114: and are contracted with the initial and final spinor
1115: wavefunctions like $u_{\al_1}(p_1)\vb_{\al_{1'}}(p_1),$ corresponding to a
1116: simple approximation for the wave-functions of the external mesons as
1117: mentioned in the introduction.
1118: After non-trivial simplifications due to the 3-dimensional dimension of
1119: the embedded trajectories,
1120: one finds
1121: \eq
1122: I[\xd] =
1123: \f{1\!+\!\xd^\mu(T) \gm^\mu}{2} \f{1\!+\!\xd^\mu(0) \gm^\mu}{2} \cdot
1124: \left(\f{1\!+\!\xd(T)\cdot\xd(0)}{2}\right)^{-1}\Rightarrow \f 1s \ ,
1125: \eqx
1126: once contracted with the initial and final spinors.
1127:
1128: Let us focus on the
1129: configuration
1130: of Wilson lines of Fig.\ref{6} in the context of a confining theory. As
1131: previously noted, the main contribution to
1132: the
1133: minimal area is from the metrics in the bulk near $z_0$ which is nearly
1134: flat. Hence, near $z_0,$ the relevant minimal area can be drawn on a
1135: classical helicoid. However, by contrast with the previous cases, the
1136: natural cut-off is provided by the exchanged quark trajectory, which is
1137: self-consistently fixed by the minimization procedure.
1138: The solution of the amplitude boils down to an Euler-Lagrange
1139: minimization
1140: over $\tau,$ namely
1141: \eq
1142: A_{\cal R}(s,L^2)\propto\f{1}{s}\lim_{\alp\to 0}\int \DD\tau\,
1143: e^{-\f{1}{2\pi\alpha'} Area(\tau)}
1144: e^{-2m {\cal L}(\tau)} \times Fluct. \ ,
1145: \label{e.pathint}
1146: \eqx
1147: where $Area(\tau)$ is the section of an helicoid bounded by the quark
1148: trajectories having total length ${\cal L}(\tau).$
1149:
1150:
1151:
1152: It can be easily shown that the Euler-Lagrange equations admit a solution which
1153: mimimizes both the area and the boundary length, namely
1154: \eq
1155: \label{e.el}
1156: \f{\partial (-\f{1}{2\pi\alpha'}Area\!-\!2m
1157: Length)}{\partial \tau} = 0\Rightarrow \sqrt{1+\left(\f
1158: {\theta\tau}{L}\right)^2}=0
1159: \ .
1160: \eqx
1161: The solution is a constant ($\f{\partial \tau(\sg)}{\partial \sg}=0$) and
1162: complex trajectory
1163: \eq
1164: \tau(\sigma)_{min}\equiv \Tau=\pm i L/\theta\ .
1165: \label{constant}
1166: \eqx
1167: Here the complex
1168: value has to be understood in the sense of applying the steepest descent
1169: method to the path integral (\ref{e.s}), and deforming the
1170: integration contours into the complex plane.
1171:
1172:
1173: Substituting the classical solution $p\tau(\sg)=-i$ into
1174: (\ref{e.s}) gives a non vanishing contribution from the
1175: logarithm:
1176: \eq
1177: \label{e.spa}
1178: e^{-\f{1}{2\pi\alef} Area(-iL/\theta)} = e^{-\f{i L^2}{4\alef \th}} \lra
1179: e^{-\f{ L^2}{4\alef \chi}}
1180: \eqx
1181: after analytical continuation to
1182: Minkowski space.
1183:
1184: Performing the Fourier transform, the
1185: resulting amplitude reads:
1186: \eq
1187: A_{\cal R}(s,q^2)=\int d\vec l\ e^{i\vec q\cdot \vec l}\ e^{-\f{
1188: L^2}{4\alef
1189: \chi}}\propto s^{-\alef q^2}\ ,
1190: \label{reggeon}
1191: \eqx
1192: corresponding to a linear Regge trajectory with intercept $0$ and
1193: slope $\alef$ related to the quark potential calculated within the same
1194: AdS/CFT
1195: framework.
1196:
1197: \section{Beyond the classical approximation: Fluctuations }
1198:
1199:
1200:
1201: \FIGURE[hb]{\epsfig{file=fluct.eps,angle=0,width=12cm}%
1202: \caption{Fluctuations around the minimal helicoid.}%
1203: \label{7}}
1204:
1205:
1206:
1207: Up to now, we restricted ourselves to a classical approximation based on the
1208: evaluation of minimal surfaces solutions for the various Wilson loops involved
1209: in the preceeding calculations. It is interesting to note \cite {us3} that a
1210: further step can be done by evaluating the contribution of quadratic
1211: fluctuations of the
1212: string worldsheet around the minimal surfaces in the case where these surfaces
1213: are embedded in helicoids, as discussed for the confining backgrounds.
1214: The semi classical correction comes from the fluctuations near the
1215: minimal surface sketched
1216: in Fig.\ref{7}. The main outcome is that this semi classical correction can be
1217: computed and is intimately related to the well-known ``universal''
1218: L\"uscher
1219: term contribution to the interquark potential \cite{16}.
1220:
1221: The fluctuation determinant for the case of a helicoid bounded by two
1222: helices with $\tau=\pm \Tau$ has already been calculated \cite{us4,us3}. Let us
1223: briefly recall the basic steps. First one reparametrizes the helicoid by
1224: replacing
1225: the variable $\tau$ in (\ref{helico}) by
1226: \eq
1227: \rho=\f{L}{\th}\ \log\left(\f{\th\tau}{L} + \sqrt{1+\f{\th^2\tau^2}{L^2}}\right)
1228: \ .
1229: \eqx
1230: In the variables $\rho$, $\sg$ the induced metric on the helicoid has a
1231: conformal factor {\it i.e.}
1232: \eq
1233: g_{ab}=(\cosh^2 \th\rho/L)\ \ \dl_{ab} \ .
1234: \eqx
1235: Therefore,
1236: since string theory in the AdS background is expected to be {\em critical}
1237: (conformal
1238: invariant), we may
1239: perform the
1240: calculation getting rid of the conformal factor, {\it i.e.} for the conformally
1241: equivalent flat metric $g_{ab}=\dl_{ab}$. This
1242: reduces to a
1243: calculation of the fluctuation determinant for a rectangle of size
1244: $a\times b$ where
1245: \eq
1246: a = L\ ;\
1247: b = \f{2L}{\th} \log \left( \th\Tau/L+\sqrt{1+\th^2 \Tau^2/L^2} \right)\ .
1248: \eqx
1249: Furthermore, we assume that the quadratic bosonic fluctuations are
1250: governed by the Polyakov action, as is indeed the case for string
1251: theories on AdS backgrounds. For high energies (after continuation to
1252: Minkowski space at large $a/b ={\cal O}(\log s) \gg 1$) one obtains
1253: \eq
1254: \label{e.fluct}
1255: Fluct.(\tau(\sg)\equiv\Tau)\to
1256: \exp \left( \f{n_\perp \cdot \pi}{24} \cdot
1257: \f{a}{b} \right)
1258: \ ,
1259: \eqx
1260: where $n_\perp$ is the number of zero modes in the transverse-to-the-branes
1261: directions.
1262: The result is just equivalent to the L\"uscher term in the potential
1263: (c.f. Ref.\cite {16}) except that the number of zero modes $n_\perp = D-2$ can
1264: be larger than
1265: the usual value (2) corresponding to flat $4D$ space.
1266:
1267: Let us consider the resulting amplitudes after account taken of the fluctuation
1268: contributions:
1269:
1270: For elastic dipole-dipole scattering, see the discussion in subsection
1271: \ref{dub}, one considers \cite{us4,us3} the analytic continuation of the
1272: minimal area when $\th \to \chi/\pi.$ Retaining here also the dominant term in
1273: the $b \to \f{4i\pi L}{\chi} $ for large
1274: $\chi \sim \log s,$ one obtains
1275: the fluctuation-corrected ``Pomeron'' amplitude
1276: \eq
1277: A_{\cal P}(s,t) \propto s^{\al_{\cal P}(t)}=s^{1+\f{n_\perp}{96}+\f{\alef}{4} t}
1278: \ .
1279: \eqx
1280:
1281: For two-body inelastic scattering, see the discussion in subsection \ref{rub},
1282: one has to implement the minimal condition (\ref{constant}) namely $b\equiv
1283: \f{i\pi L}{\chi}.$ Hence,
1284: the fluctuation-corrected ``Reggeon'' trajectory (cf.
1285: \ref{reggeon})
1286: reads\footnote{
1287: Possible logarithmic
1288: prefactors, which are not under control at this stage of our
1289: approach, are not determined.}:
1290: \eq
1291: \label{e.lintraj}
1292: A_{\cal R}(s,t) \propto s^{\al_{\cal R}(t)}\ =s^{\f{n_\perp}{24}+{\alef}t}
1293: \ .
1294: \eqx
1295:
1296:
1297: Let us comment these results.
1298: The first observation is that in both cases, the slope is determined
1299: by minimal surface solutions through the logarithmic contribution in
1300: the helicoid area. The factor four in the slope
1301: comes from the specific saddle point path integral over the exchanged quark
1302: trajectories (for Reggeon exchange). It is interesting to note that
1303: this theoretical feature is in qualitative agreement with the phenomenology of
1304: soft scattering. Indeed once we fix the $\alef$ from the
1305: phenomenological value of the static $q\qb$ potential ($\alef\sim 0.9\,
1306: GeV^{-2}$)
1307: we get for the slopes $\al_R=\alef\sim 0.9\, GeV^{-2}$ and
1308: $\al_P=\alef/4\sim 0.23\, GeV^{-2}$ in good agreement with the phenomenological
1309: slopes.
1310:
1311: The second feature is the relation between the Pomeron and Reggeon
1312: intercepts. At the classical level of our approach these are
1313: respectively 1 and 0. Note that this classical piece is in
1314: agreement with what is obtained from simple exchanges of two
1315: gluons and quark-antiquark pair, respectively, in the $t$ channel. The
1316: fluctuation (quantum) contributions to the Reggeon and Pomeron are also
1317: related by the factor four.
1318:
1319: Adding both classical and fluctuation contributions gives an estimate
1320: which is in qualitative agreement with the observed intercepts.
1321: Indeed, when calculating the fluctuations around
1322: a minimal surface near the horizon in the
1323: BH backgrounds there could be $n_\perp=7,8$ massless bosonic modes
1324: \cite{so99}. For $n_\perp=7,8$ one gets $1.073-1.083$ for the Pomeron and
1325: $0.3-0.33$
1326: for the Reggeon.
1327: This result is in agreement with the observed intercept for the ``Pomeron'' and
1328: somewhat below the intercepts of around $0.5$ observed for the
1329: dominant Reggeon trajectories.
1330:
1331: An interesting feature of the results is the
1332: key role of the logarithmic term in the formulae (cf. (\ref{e.ahelic}) for the
1333: area of the truncated helicoids. Besides the main feature being that it leads,
1334: through its analytical structure, to Reggeization, it
1335: also gives rise to the possibility of
1336: additional contributions from crossing different Riemann sheets
1337: ($log \to log+2\pi i k$)
1338: in the course of performing analytical continuation from Euclidean to
1339: Minkowski space .
1340:
1341: For instance in the ``Reggeon'' case,
1342: the amplitude in impact parameter space (\ref{reggeon}) picks
1343: up new multiplicative factors:
1344: \eq
1345: \label{e.rp}
1346: e^{-\f{L^2}{4\alef \log s}} \cdot e^{-k\f{L^2}{\alef \log s}} \ .
1347: \eqx
1348: This can be interpreted (for $k>0$) as $k$-Pomeron exchange
1349: corrections to a single Reggeon exchange. Indeed the slope of the
1350: trajectory obtained from Fourier transform of formula (\ref{e.rp}) is
1351: the one expected from such contributions\footnote{However, the semi-classical
1352: correction to the intercepts seems to be more delicate, and needs further
1353: study.}.
1354:
1355: \section{Conclusion: Reggeization and confinement}
1356:
1357: The interesting output of the application
1358: of
1359: AdS/CFT correspondence to high energy amplitudes at strong coupling is
1360: to
1361: emphasize the relation between Reggeization and confinement, using the
1362: description of two-body scattering amplitudes in the dual string theory. Lattice
1363: calculations, which is the only presently known way to evaluate
1364: directly QCD observables at strong coupling, are not able to compute
1365: high-energy amplitudes.
1366:
1367:
1368: When comparing AdS$_5$ duality - which corresponds to a conformal,
1369: non-confining gauge theory - with AdS$_{BH}$ duality, which leads
1370: to
1371: reggeization, the difference ultimately comes from the different metrics
1372: in
1373: the bulk and thus from the different geometrical features of the minimal
1374: surfaces for the same boudary
1375: conditions. In particular, taking into account their different geometry, see
1376: e.g.
1377: Fig.\ref{8}, one expects after analytic continuation and in the
1378: large
1379: energy ($\chi\to \infty$) limit:
1380: \eq
1381: Area^{AdS}_{min}\sim
1382: \lim_{\chi\to \infty} \f{L}{L/\chi}\ ; \ Area^{BH}_{min}\sim
1383: \lim_{\chi\to
1384: \infty}
1385: L\times\f{L}{\chi}\ .
1386: \eqx
1387:
1388: \FIGURE[ht]{\epsfig{file=origin.eps,angle=0,width=12cm}%
1389: \caption{Comparison of $AdS_5$ and $AdS_{BH}$ minimal contours at high energy.}
1390: \label{8}}
1391:
1392: The AdS$_5$ case leads to a $L$-invariant value, as it is scale invariant,
1393: and,
1394: after
1395: Fourier transformation, to a high-energy amplitude with a $q^2$
1396: independent
1397: energy exponent (or flat Regge trajectory), see formula (\ref{e.lcft}). On the
1398: other hand, the
1399: AdS$_{BH}$
1400: case leads to a linear Regge trajectory
1401: after
1402: Fourier transformation.
1403: For the AdS$_{BH}$ case this rough expectation can be verified by
1404: an explicit calculation. Hence confinement appears as an essential
1405: ingredient
1406: for the reggeized structure of two-body high-energy amplitudes. We expect this
1407: result
1408: not to be dependent on the precise geometrical
1409: AdS$_{BH}$ setting and thus to indicate a quite general property of
1410: confining
1411: theories.
1412:
1413: As a conclusion, let us summarize our results:
1414: \begin{itemize}
1415: \item The AdS/CFT correspondence can be used to give a geometrical formulation
1416: of two-body scattering amplitudes in the gravitational dual of gauge field
1417: theories.
1418: \item The consideration of quark-quark scattering in physical Minkowski space
1419: shows that the main geometrical features of two-body scattering amplitudes are
1420: related to a (generalized to AdS metrics) helicoidal structure of minimal
1421: surfaces in Euclidian space {\it via} analytical continuation.
1422: \item In the case of non conformal theories, such as the $AdS_{BH}$ case, the
1423: metrics is approximately flat near the horizon corresponding to the confinement
1424: scale. The (flat space) helicoidal solutions lead to amplitudes with linear
1425: Regge slopes.
1426: \item While quark-quark scattering is entailed by a cut-off dependence,
1427: colourless dipole scattering give rise to cut-off free amplitudes. In particular
1428: the two-body dipole-dipole scattering with quark exchange leads to unambiguous
1429: results with well-defined regge behaviour.
1430: \item Regge trajectories come out linear,
1431: with
1432: slopes and intercepts related to the quark potential. They include a
1433: semi-classical correction due to the fluctuation around the minmal surfaces
1434: which are similar to a
1435: L\" uscher
1436: term, but in a 10-dimensional string framework.
1437: \item The Pomeron (elastic case) intercept is $1+\epsilon$ where $\epsilon$
1438: is related to a L\" usher term. There exists a factor four between the Reggeon
1439: (inelastic case) and Pomeron Regge slopes in agreement with ``soft scattering''
1440: phenomenology.
1441:
1442: \end{itemize}
1443:
1444:
1445: In conclusion, the AdS/CFT framework give new insights on the 35-years-old
1446: puzzle of high-energy amplitudes at strong gauge coupling.
1447:
1448: As a short outlook let us list some interesting problems for future work:
1449:
1450: \begin{itemize}
1451: \item {\it High-energy phenomenology:} Many aspects, like
1452: the
1453: Flavor/Spin dependences, remain to be studied.
1454: \item {\it Approximations:} The dual gauge theory is not
1455: specified, and the exact minimal surface in the bulk metrics to be
1456: determined.
1457: \item {\it Dual of QCD?} In the present framework, the
1458: confining scale $R_0$ has no relation with $\Lambda_{QCD}.$
1459: \item {\it Unitarity:} A more complete investigation
1460: requires the study of multi-leg amplitudes.
1461: \item {\it Deeper general problems:} The formulation of
1462: string
1463: theory in AdS backgrounds and last but not least, a proof of the AdS/CFT
1464: conjecture.
1465:
1466: \end{itemize}
1467:
1468: \acknowledgments
1469:
1470: I warmly thanks Romuald Janik with whom the approach of high-energy
1471: amplitudes described in this review has been done in tight collaboration. I
1472: thank Otto Nachtmann and the organizers of the Heidelberg meeting for the
1473: stimulating and fruitful atmosphere.
1474:
1475: \begin{thebibliography}{99}
1476:
1477:
1478: \bibitem{Frampton} For a good introduction to String theory models and
1479: problems for strong interactions, see
1480: P.H.Frampton, {\it Dual Resonance Models }, (1974, Benjamin; New edition:
1481: 1986, World Scientific).
1482:
1483: \bibitem{polyakov}
1484: \rr{{A.M. Polyakov}}{ Nucl.Phys.Proc.Suppl.}{68}{(1998) 1}.
1485:
1486: \bibitem{17}
1487: \rr{{J. Polchinski and M.J. Strassler}}{Phys. Rev. Lett.}{88}{(2002)
1488: 031601}.
1489:
1490:
1491: \bibitem{us1}
1492: \rr{R.A. Janik and R. Peschanski}{Nucl. Phys.}{B565}{(2000) 193}.
1493:
1494: \bibitem{us2}
1495: \rr{{R.A. Janik and R. Peschanski}}
1496: {Nucl. Phys.}{B586}{(2000) 163}.
1497:
1498: \bibitem{us4}
1499: \rr{R.A. Janik and R. Peschanski}{Nucl. Phys.}{B625}{(2002) 279}.
1500:
1501: \bibitem{us3}
1502: \rr{{R.A. Janik}}{Phys. Lett.}{B500}{(2001) 118}.
1503:
1504:
1505: \bibitem{11} \rr{{J. Maldacena}}{Adv. Theor. Math. Phys.}{2}{(1998)
1506: 231};
1507: \rr{{S.S. Gubser, I.R. Klebanov and
1508: A.M. Polyakov}}{Phys. Lett.}{B428}{(1998) 105}.
1509:
1510: \bibitem{review} For a detailed review and an extensive list of
1511: references, please refer to:
1512: \rr{{O. Aharony, S.S. Gubser, J. Maldacena, H. Ooguri
1513: and Y. Oz}}{Phys.Rept.}{323}{(2000)183}.
1514:
1515: \bibitem{wi98}\rr{E. Witten}{Adv. Theor. Math. Phys.}{2}{(1998) 505}.
1516:
1517:
1518: \bibitem{de99} \rr{H. Ooguri, H. Robins and
1519: J. Tannenhauser}{Phys. Lett.}{B437}{(1998) 77}.
1520:
1521: \bibitem{so99}
1522: \rr{{Y. Kinar, E. Schreiber, J. Sonnenschein and
1523: N. Weiss}}{Nucl. Phys.}{B583}{(2000) 76}.
1524:
1525:
1526:
1527:
1528:
1529: \bibitem{ks} \rr{A. Kehigas and K. Sfetsos}{Phys. Lett.}{B456}{(1999) 22}.
1530:
1531: \bibitem{carlo} \rr{C. Angelantonj and A. Armoni}{Phys. Lett.}{B482}{(2000)
1532: 329}.
1533:
1534: \bibitem{12}
1535: \rr{{J. Maldacena}}{
1536: Phys. Rev. Lett.}{80}{(1998) 4859};
1537: \rr{{S.-J. Rey and J. Yee}}{}{}{hep-th/9803001};
1538: \rr{{JJ. Sonnenschein and A. Loewy}}{JHEP}{0001}{(2000) 042}.
1539:
1540: \bibitem{13}
1541: \rr{{M. Rho, S.-J. Sin and I. Zahed}}{Phys. Lett.}{B466}{(1999) 199}.
1542:
1543: \bibitem{neww} \rr{N. Drukker, D.J. Gross and H.
1544: Ooguri}{Phys.Rev.,}{D60}{(1999)
1545: 125006};\rr{H. Ooguri}{Prog.Theor.Phys.Suppl.,}{134}{(1999) 153}.
1546:
1547: \bibitem{bfkl} \rr{L.N. Lipatov}{Sov. J. Nucl. Phys.}{23}{(1976) 642};
1548: \rr{V.S. Fadin, E.A. Kuraev and L.N. Lipatov}{Phys. Lett.}{B60}{(1975)
1549: 50}; \rr{E.A. Kuraev, L.N. Lipatov and V.S. Fadin}{Sov. Phys.
1550: JETP}{44}{(1976)
1551: 45, {\bf 45} (1977) 199}; \rr{I.I. Balitsky and
1552: L.N. Lipatov}{Sov. J. Nucl. Phys.}{28}{(1978) 822}.
1553:
1554: \bibitem{14}
1555: \rr{{O. Nachtmann}}{High Energy Collisions and Nonperturbative QCD,}{}{Lectures
1556: given at Banz (Germany) 1993 and at Schladming (Austria) 1996, hep-ph/9609365}.
1557:
1558: \bibitem{meggio}
1559: \rr{{E. Meggiolaro}}{
1560: Z. Phys.}{C76}{(1997) 523}; \rr{{}}{Nucl. Phys.}{B625} {(2002) 312}.
1561:
1562: \bibitem{minimal}
1563: \rr{{J.C.C. Nitsche}}{Lectures on minimal surfaces,}{}{Cambridge University
1564: Press,1989}.
1565:
1566: \bibitem{15}
1567: \rr{{R.P. Feynman}}{Phys.\ Rev.}{80}{(1950) 440}
1568: \rr{{A.~M.~Polyakov}}{
1569: Mod.\ Phys.\ Lett.}{A3}{(1988) 325}.
1570: \rr{{G.~P.~Korchemsky}}{
1571: Int.\ J.\ Mod.\ Phys.}{A7}{(1992) 339};
1572: \rr{}{
1573: Phys.\ Lett.}{B232}{(1989) 334}.
1574:
1575:
1576: \bibitem{16}
1577: \rr{{M. L\" uscher, K. Symanzik and
1578: P. Weisz}}{Nucl.Phys.}{B173}{(1980) 365}.
1579: \rr{{O. Alvarez}}{
1580: Phys.Rev.}{D24}{(1981) 440}.
1581:
1582: \end{thebibliography}
1583: \end{document}
1584: