hep-ph0408172/p08.tex
1: 
2: \documentclass[12pt]{article}
3: \topmargin -1.5cm \oddsidemargin +0.0cm \baselineskip 9mm
4: \textwidth 148mm \textheight 225mm
5: \usepackage{graphics}
6: \usepackage{psfrag}
7: \usepackage{epsfig}
8: \begin{document}
9: \title{{Quark energy loss  and shadowing in nuclear Drell-Yan process }
10: %\thanks {Supported by  Natural
11: %Science Foundation of China(90103020,10075057,10175074) , CAS
12: %Knowledge Innovation Project(KJCX2-SW-N02),Major State Basic
13: %Research Development Program (G20000774),  Natural Science
14: %Foundation of Hebei Province(103143)}
15: }
16: \author{ Chun-Gui Duan $^{1,2,5}$\thanks{E-mail:duancg$@$mail.hebtu.edu.cn}
17: Shu-Wen Cui $^3$  \\
18: Zhan-Yuan Yan  $^{4}$ Guang-Lie Li $^{2,5}$
19:  \\
20: {\small 1.Department of Physics, Hebei Normal University, Shijiazhuang ,050016,China}\\
21: {\small 2.Institute of high energy physics ,The Chinese academy of sciences ,Beijing,100039,China}\\
22: {\small 3.Physics department,Cangzhou Teachers' College,Cangzhou,061000,China}\\
23: {\small 4.Department of Applied Physics,North China Electric Power University,Baoding,071003,China}\\
24: {\small
25: 5.CCAST(WorldLaboratory).P.O.Box8730,Beijing,100080,China}}
26: 
27: 
28: \date{}
29: \maketitle
30: \baselineskip 9mm
31: \vskip 0.5cm
32: \begin{center}
33: \begin{minipage}{134mm}
34: \begin{center}
35: Abstract
36: \end{center}
37: The energy loss effect in  nuclear matter is another nuclear
38: effect apart from the nuclear effects on the parton distribution
39: as in deep inelastic scattering process. The quark energy loss can
40: be measured best by the nuclear dependence of the high energy
41: nuclear Drell-Yan process.   By means of three kinds of  quark
42: energy loss parameterizations given  in literature and  the
43: nuclear parton distribution extracted only with lepton-nucleus
44: deep inelastic scattering experimental data, measured Drell-Yan
45: production cross sections are analyzed for 800GeV proton incident
46: on a variety of nuclear targets from FNAL E866. It is shown that
47: our results with considering the energy loss effect are much
48: different from these of the FNAL E866 who analysis the
49: experimental data with the nuclear parton distribution functions
50: obtained by using the deep inelastic lA collisions and pA nuclear
51: Drell-Yan data . Considering the existence of energy loss effect
52: in Drell-Yan lepton pairs production,we suggest that the
53: extraction of nuclear parton distribution functions should not
54: include Drell-Yan experimental data.\\
55: Keywords:  Drell-Yan, energy loss,  nuclear parton distribution
56: functions\\
57: 
58: PACS:24.85.+p;13.85.QK;25.40.-h;25.75.q
59: 
60: \end{minipage}
61: \end{center}
62: 
63: 
64: \vskip 0.5cm
65: 
66: 
67: 
68: \hspace{0.5cm}Understanding the initial stages of
69: ultrarelativistic heavy ion collisions is of utmost importance in
70: order to understand the outcome of the high energy heavy ion
71: experiments, such as the BNL relativistic heavy ion collider(RHIC)
72: and CERN large hadron collider (LHC). Understanding the
73: modifications of the parton distribution functions  and the parton
74: energy loss in nuclei should be the first important step towards
75: pinning down the initial conditions of a heavy-ion collision and
76: understanding of $J/\psi$ production  which is required if it is
77: to be used as a signal for the quark-gluon plasma in relativistic
78: heavy ion collisions.
79: 
80: 
81: 
82: \hspace{0.5cm}The high energy inelastic hadron-nucleus collisions
83: has been studied for many decades by both the nuclear and particle
84: physics communities $^{[1]}$. By means of the nuclei, we can study
85: the space-time development of the strong interaction during its
86: early stages, which is inaccessible between individual hadrons. In
87: high energy inelastic hadron-nucleus scattering, the projectile
88: rarely retains a major fraction of its momentum after traversing
89: the nucleus. Rather, its momentum is shared by several produced
90: particles,which form a hadron jet in the forward direction.The
91: classical description of this phenomena is that the  projectile
92: suffers multiple collisions and repeated energy loss in the
93: nuclear matter. In other words,each quark or gluon in the
94: projectile can loss a finite fraction of its energy  in the
95: nuclear target due to QCD bremsstrahlung$^{[2]}$. The Drell-Yan
96: reaction$^{[3]}$ on nuclear targets provides, in particular, the
97: possibility of probing the propagation of quark through nuclear
98: matter, with the produced lepton pair carrying away the desired
99: information on the projectile quark after it has travelled in the
100: nucleus. Only initial-state interactions are important in
101: Drell-Yan process since the dimuon in the final state does not
102: interact strongly with the partons in the nuclei. This makes
103: Drell-Yan scattering an ideal tool to study energy loss$^{[4]}$.
104: In addition,Drell-Yan reaction is closely related to DIS of
105: leptons , but unlike DIS ,it is directly sensitive to antiquark
106: contributions in target parton distributions. When DIS on nuclei
107: occurs at $x<0.08$, where x is the parton momentum fraction, the
108: cross section per nucleon decreases with increasing nucleon number
109: A due to shadowing$^{[5]}$. Shadowing should also occur in
110: Drell-Yan dimuon production at small $x_2$, the momentum fraction
111: of the target parton, and theoretical calculations indicate that
112: shadowing in the DIS and Drell-Yan reactions has a common origin
113: $^{[6]}$.Therefore,shadowing and initial state partonic energy
114: loss are processes that occur in  the proton-induced Drell-Yan
115: reaction on nuclei.
116: 
117: 
118: 
119: 
120: 
121: 
122: \hspace{0.5cm}In order to describe the modification of the initial
123: state parton distributions in nucleus , a  variety of approach to
124: this question exist in the literature$^{[7]}$. Recently,there were
125: two trials without model dependence to obtain nuclear parton
126: distributions  from  the existing world experimental data. In
127: 1999, Eskola, Kolhinen, Ruuskanen  and  Salgado(EKRS)$^{[8]}$
128: suggested a set of nuclear parton distributions ,which are studied
129: within a framework of the DGLAP evolution . The measurements of
130: $F^A_2/F^D_2$ in deep inelastic $lA$ collisions, and Drell-Yan
131: dilepton cross sections measured in $pA$ collisions were used as
132: constraints.The kinematical ranges are $10^{-6}\leq x\leq 1$  and
133: $2.25GeV^2\leq Q^2\leq10^{4}GeV^{2}$  for nuclei from deuteron to
134: heavy ones .With the nuclear parton distributions , the calculated
135: results agreed very well with the relative EMC and Fermilab E772
136: experimental data$^{[9]}$ . In 2001, Hirai,Komano and
137: Miyama(HKM)$^{[10]}$ proposed two types  of nuclear parton
138: distributions which were obtained by quadratic and cubic type
139: analysis, and determined by a $\chi^2 $ global analysis of
140: existing experimental data on nuclear structure functions without
141: including the proton-nucleus Drell-Yan process .The kinematical
142: ranges covered $10^{-9}\leq x\leq 1$  and
143:  $1GeV^2\leq Q^2\leq10^{5}GeV^{2}$  for nuclei from deuteron to heavy
144: ones . As a result, they obtained reasonable fit to the measured
145: experimental data of $F_2$.In this work , we will use both
146: parameterizations and investigate the nuclear dependence of the
147: Drell-Yan process.
148: 
149: \hspace{0.5cm}In order to research the partonic energy loss in
150: nuclei, Fermilab Experiment866(E866) $^{[11]}$ performed the
151: precise measurement of the ratios of the Drell-Yan cross section
152: per nucleon for an 800GeV proton beam incident on Be, Fe and W
153: target at larger values of $x_1$,the momentum fraction of the beam
154: parton, larger values of $x_F$($\approx x_1-x_2$),and smaller
155: values of $x_2$ than reached by the previous experiment, Fermilab
156: E772$^{[9]}$. The extended kinematic coverage of E866
157: significantly increases its sensitivity to energy loss and
158: shadowing.This is the first experiment on the energy loss of quark
159: passing through a cold nucleus.
160: 
161: 
162: 
163: \hspace{0.5cm}After the E866 experimental data was
164: reported,several groups have given their theoretical analysis of
165: the data$^{[12,13,14]}$. In previous report$^{[14]}$ , by means of
166: EKRS and HKM nuclear parton distribution functions, we
167: investigated the Drell-Yan production cross section ratios for
168: 800GeV proton incident on a variety of nuclear targets  in the
169: framework of  Glauber model. We found that the theoretical results
170: with energy loss are in good agreement with  the Fermilab E866
171: experiment by means of HKM nuclear parton distributions. However,
172: the calculated results without energy loss can give good fits by
173: using  EKRS nuclear parton distribution functions. In this letter,
174: the nuclear dependence of the pA Drell-Yan production cross
175: sections are studied  by combining the quark energy loss
176: parametrization$^{[15,18,19]}$ given  in literature with the EKRS
177: and HKM nuclear parton distribution.
178: 
179: 
180: \hspace{0.5cm}In the Drell-Yan process,the leading-order
181: contribution is quark-antiquark annihilation into a lepton pair.
182: The annihilation cross section can be obtained from the
183: $e^{+}e^{-}\rightarrow\mu^{+}\mu^{-}$ cross section by including
184: the color factor $\frac{1}{3}$ with the charge $e^{2}_{f}$ for the
185: quark of flavor $f$.
186: \begin{equation}
187:    \frac{d\hat{\sigma}}{dM}=\frac{8\pi\alpha^2}{9M}e^2_f\delta(\hat{s}-M^2),
188: \end{equation}
189: where $\sqrt{\hat{s}}=(x_1x_2s)^{1/2}$,is the center of mass
190: system (c.m.system) energy of $q\bar{q}$ collision
191: ,$x_1$(resp.$x_2$)is the momentum fraction carried by the
192: projectile (resp.target) parton, $\sqrt{s}$ is the center of mass
193: energy of the hadronic collision, and $M$ is the invariant mass of
194: the produced dimuon. The hadronic Drell-Yan differential cross
195: section is then obtained from the convolution of the above
196: partonic cross section with the quark distributions in the beam
197: and in the target :
198: \begin{equation}
199:  \frac{d^2\sigma}{dx_1dM}=K\frac{8\pi\alpha^2}{9M}\frac{1}{x_1s}
200:  \sum_{f}e^2_f[q^p_f(x_1)\bar{q}^A_f(x_2)
201:  +\bar{q}^p_f(x_1)q^A_f(x_2)],
202: \end{equation}
203: where $ K$ is the high-order QCD correction, $\alpha$ is the
204: fine-structure constant, the sum is carried out over the light
205: flavor $f=u,d,s$, and $q^{p(A)}_{f}(x)$ and ${\bar
206: q}^{p(A)}_{f}(x)$ are the quark and anti-quark distributions in
207: the proton (nucleon in the nucleus A). In order to obtain the
208: $x_1$ dependence of Drell-Yan production, we shall  deal in the
209: following with the single differential cross section,
210: \begin{equation}
211:  \frac{d\sigma}{dx_1}=K\frac{8\pi\alpha^2}{9x_1s}
212:  \sum_{f}e^2_f\int\frac{dM}{M}[q^p_f(x_1)\bar{q}^A_f(x_2)
213:  +\bar{q}^p_f(x_1)q^A_f(x_2)],
214: \end{equation}
215: where the integration over the dimuon mass is performed in the
216: range given from E866 experiment.
217: 
218: \hspace{0.5cm}Now let us take into account of the energy loss of
219: the fast quarks moving through the cold nuclei.There were three
220: expressions for the average change  in incident parton momentum
221: fraction in literature .The first was given by Gavin and
222: Milana$^{[15]}$ in 1992.They analyzed the depletion at high
223: Feynman x in large relative to small nuclear targets from
224: measurements of the Drell-Yan process and of charmonium and
225: bottomonium production at CERN$^{[17]}$ and Fermilab
226: E772$^{[9,16]}$ . They proposed  that energy loss due to multiple
227: parton scattering can consistently explain the high-$x_F$
228: depletion in the Drell-Yan process and $J/\psi$ production.The
229: following expression was assumed for the average change  in
230: incident parton momentum fraction :
231: \begin{equation}
232: \Delta x_1= {\kappa}_1x_1A^{1/3},
233: \end{equation}
234: where the factor ${\kappa}_1$ may have a $Q^2$ dependence.From
235: comparison to the experimental data and neglecting shadowing, an
236: upper bound  on $dE/dz$--the incident quark energy loss per unit
237: length --- was obtained,the value is $1.5GeV/fm$.This result was
238: questioned by Brodsky and Hoyer$^{[18]}$,who argued that the time
239: scale for gluon bremsstrahlung need to be taken into account.By
240: analogy with the photon bremsstrahlung process , they derived the
241: quantum mechanical bound on the amount of radiative energy loss
242: suffered by high energy quarks and gluons in nuclear matter,
243: \begin{equation}
244: \Delta x_1 \approx \frac{{\kappa}_2}{s}A^{1/3},
245: \end{equation}
246: where  $s$ denotes  the square of the proton-nucleus center of
247: mass energy. The formulation developed by Brodsky and Hoyer was
248: extended by Baier et al.$^{[19]}$ They studied the medium-induced
249: $p_{\perp}$-broadening and induced gluon radiation spectrum of a
250: high energy quark or gluon traversing a large nucleus.Multiple
251: scattering of the high energy parton in nuclei was treated in the
252: Glauber approximation.  Baier et al. found that the partonic
253: energy loss  depended on a characteristic length and the
254: broadening of the squared transverse momentum of the parton. For
255: finite nuclei,both factors vary as $A^{1/3}$, so  they predicted
256: \begin{equation}
257: \Delta x_1 \approx \frac{{\kappa}_3}{s}A^{2/3}.
258: \end{equation}
259: Using these energy loss expressions , we can obtain empirical
260: values for the $\kappa$'s by fitting  the FNAL E866 Drell-Yan
261: cross section ratios versus the incident proton's momentum
262: fraction and dimuon effective mass.
263: 
264: 
265: 
266: \hspace{0.5cm}If considering the  quark energy loss in nuclei, the
267: incident quark momentum fraction can be shifted from
268: $x'_1=x_1+\Delta x_1$ to $x_1$ at the point of fusion. Combining
269: the shadowing with initial state energy loss ,the production cross
270: section in pA Drell-Yan process can be written as
271: \begin{equation}
272:  \frac{d\sigma}{dx_1}=K\frac{8\pi\alpha^2}{9x_1s}
273:  \sum_{f}e^2_f\int\frac{dM}{M}[q^p_f(x'_1)\bar{q}^A_f(x_2)
274:  +\bar{q}^p_f(x'_1)q^A_f(x_2)].
275: \end{equation}
276: In order to pin down quark energy loss by comparing with the
277: experimental data from E866 collaboration$^{[11]}$,we introduce
278: the nuclear Drell-Yan ratios as:
279: \begin{equation}
280: R_{A_{1}/A_{2}}(x_{1})=\frac{d\sigma^{p-A_{1}}}{dx_ {1}}/{\frac
281: {d\sigma^{p-A_{2}}}{dx_{1}}}.
282: \end{equation}
283: The integral range on M is determined according to the E866
284: experimental kinematic region.In our theoretical analysis
285: ,$\chi^2$ is calculated with the Drell-Yan differential cross
286: section rations $R_{A_1/A_2}$ as
287: \begin{equation}
288: \chi^2=\sum\limits_{j}\frac{(R^{data}_{A_1/A_2,j}-R^{theo}_{A_1/A_2,j})^2}
289: {(R^{err}_{A_1/A_2,j})^2},
290: \end{equation}
291: where the experimental error is given by systematic errors as
292: $R^{err}_{A_1/A_2,j} $, and $ R^{data}_{A_1/A_2,j}$(
293: $R^{theo}_{A_1/A_2,j}$) indicates the experimental data
294: (theoretical values ) for the ratio$R_{A_{1}/A_{2}}$.
295: 
296: \hspace{0.5cm} Taking advantage of the EKRS$^{[8]}$ nuclear parton
297: distribution functions and
298: $\kappa_1=0,\kappa_2=0,\kappa_3=0$(without energy loss
299: effects),the obtained $\chi^2$ value is $\chi^2=51.4$  for the 56
300: total data points. The $\chi^2$ per degrees of freedom is given by
301: $\chi^2/d.o.f.=0.918$. It is apparent that theoretical results
302: without energy loss effects agree very well with the E866
303: experimental data ,which results from EKRS parametrization of
304: nuclear parton distributions studied with including the E772
305: Drell-Yan data $^{[16]}$ . Because of employing the EKRS global
306: phenomenological analysis, the E866 analysis missed the energy
307: loss effect of quarks in nuclei.
308: 
309: 
310: \hspace{0.5cm}In what follows,we consider also combining HKM cubic
311: type of nuclear parton distribution$^{[10]}$  with the three quark
312: energy loss parameterizations. With
313: $\kappa_1=0,\kappa_2=0,\kappa_3=0$(without energy loss
314: effects),the obtained  $\chi^2$ per degrees of freedom is
315: $\chi^2/d.o.f.=2.516$. With
316: $\kappa_1=0.0036,\kappa_2=3.45,\kappa_3=0.46$(with energy loss
317: effects),the obtained  $\chi^2$ per degrees of freedom are
318: $\chi^2/d.o.f.=1.00048, 1.022, 1.053$, respectively.  The results
319: given by HKM quadratic type  are nearly the same as these above.
320: As an example,the calculated results with energy loss expression
321:  Eq.(5)are shown in Fig.1 and Fig.2. which is the Drell-Yan cross
322: section ratios for Fe  to Be and   W to Be as functions of $x_1$
323: for various interval of $M$,respectively.The solid curves are the
324: ratios with only the nuclear  effect on the parton distribution as
325: in DIS scattering process, and the dotted curves correspond to an
326: energy loss effect: $\kappa_2=3.45 GeV^2$ with nuclear effect on
327: structure function .From comparison with the experimental data ,
328: it is found that our theoretical results with energy loss effect
329: are in good agreement with the Fermilab E866.However,the values of
330: $\kappa_1,\kappa_2,\kappa_3$ are  very different from those by the
331: E866 Collaboration$^{[11]}$,which results from employing EKRS
332: nuclear parton distributions. The EKRS analysis itself included
333: the Drell-Yan data from E772,with the presumption that the
334: low-$x_2$ nuclear dependence arose entirely from shadowing.This
335: clearly introduced an inconsistency into the E866 search for
336: energy loss. Therefore,we can conclude that the extraction of
337: nuclear parton distribution functions should not include Drell-Yan
338: experimental data because of the energy loss of the incoming
339: partons.
340: 
341: 
342: 
343: \hspace{0.5cm}In Eq.(4), the average loss of projectile parton
344: vary with the variable $x_1$. It is difficult to obtain a constant
345: energy loss $dE/dz(GeV/fm)$.$\kappa_3=0.46$ shows  us that the
346: observed energy loss of the incident quarks is $\Delta E<0.21
347: GeV/fm^2\times L^2$.$\kappa_2=3.45$ in Eq.(5)indicates that the
348: incident quarks lose energy with a constant rate
349: $dE/dz=2.04GeV/fm$ if $<L>=3/4(1.2A^{1/3})fm$ is used as the
350: average path length. The value of $dE/dz$ is not  obviously
351: consistent with that given by M.B.Johnson et al.$^{[12]}$,which is
352: $-dE/dx=3.14\pm 0.53 GeV/fm$.M.B.Johnson et al. examined the
353: effect of initial state energy loss on the Drell-Yan cross section
354: ratios from E866 by employing a new formulation of the Drell-Yan
355: process in the rest frame of nucleus.It is obvious that an
356: intuitive explanation of the quark energy loss depends on the
357: reference frame of the target nucleus. It is well-known that the
358: physical observables can not depend on our choice of reference
359: frame. Therefore,further theoretical and experimental studies are
360: needed to determine the value of incident quark energy loss per
361: unit length,and verify whether the energy loss is linear or
362: quadratic with the path length.
363: 
364: 
365: \hspace{0.5cm}With considering the existence of energy loss effect
366: in Drell-Yan lepton pairs production,we suggest that the
367: extraction of nuclear parton distribution functions should not
368: employ pA  Drell-Yan experimental data.Although there are
369: currently abundant data on electron and moun deep inelastic
370: scattering, it is difficult to determine valence quark
371: distributions in the small x region and the anti-quark
372: distributions .Only valence quark distributions in medium x region
373: can be relatively well determined. It is well considered that the
374: precise nuclear parton distributions must be known in order to
375: calculate cross sections of high energy nuclear reactions
376: accurately and find a signature of quark-gluon plasma in high
377: energy heavy-ion reactions.We suggest using precise neutrino
378: scattering experimental data ,which can provide a good  method for
379: measuring the $F_2(x,Q^2)$ and $xF_3(x,Q^2)$ structure
380: functions.Using the average of $xF_3^{\nu A}(x,Q^2)$and
381: $xF_3^{\bar{\nu}A}(x,Q^2)$ , the valence quark distribution
382: functions can be well pinned down$^{[20]}$. Combining the lepton
383: inelastic scattering data with the neutrino scattering
384: experiments, valence quark and anti-quark distribution functions
385: will be obtained in the future,which makes us good understanding
386: the energy loss effect in high energy nuclear collisions.
387: 
388: 
389: 
390: 
391: 
392: \hspace{0.5cm}In summary, we have made a leading-order analysis of
393: E866 data in nuclei  by taking into account of the energy loss
394: effect of fast quarks . Our theoretical results with quark energy
395: loss are in good agreement with the Fermilab E866 experiment by
396: means of the parametrization of nuclear parton distributions
397: studied without nuclear Drell-Yan process, which is the same as
398: that in our previous work$^{[14]}$.we suggest that the extraction
399: of nuclear parton distribution functions should not employ pA
400: Drell-Yan experimental data.In order to determine the amount of
401: energy loss and the appropriate path length,we desire to  operate
402: precise measurements of the nuclear dependent of Drell-Yan
403: production at the Fermilab Main Injector(FMI,120GeV proton
404: beam)$^{[21]}$ and the Japan Hadron Facility(JHF,50GeV proton
405: beam)$^{[22]}$,where shadowing effect disappears and energy loss
406: effect of fast quarks could provide the dominant nuclear
407: dependence.
408: 
409: {\bf Acknowledgement:}This work is partially supported by  Natural
410: Science Foundation of China(90103020,10075057,10175074) , CAS
411: Knowledge Innovation Project (KJCX2-SW-N02),Major State Basic
412: Research Development Program (G20000774),  Natural Science
413: Foundation of Hebei Province(103143)
414: 
415: 
416: 
417: \vskip 1cm
418: \begin{thebibliography}{s2}
419: \bibitem{s1}  W.Busza R Ledoux, Ann.Rev.Nucl.Part.Sci.,38(1989)119.
420: \bibitem{s2}  D.S.Barton et al.,Phys.Rev.D27(1983)2580.
421: \bibitem{s3}  S.Drell and T.M.Yan, Phys.Rev,Lett.,25(1970)316.
422: \bibitem{s4}  G.T.Garvey and J.C.Peng ,Phys.Rev.Lett.,90(2003)092302.
423: \bibitem{s5}  P.Amaudruz et al.,Z.Phys.C51(1991)387;57(1993)211.
424: \bibitem{s6}  S.J.Brodsky,A.Hebecker,E.Quark.Phys.Rev.D.55.(1997)2584.\\
425:               B.Kopelievich.Proc.workshop Hirschegg 1995, ed.by
426:               H.Feldmeier and W.Norenberg,Darmstadt,1995,102.
427: \bibitem{s7}  M.Arneodo.et.al.(EMC), Nucl.Phys,B441(1995)3.\\
428:               D.F.Geesaman,K.Saito and
429:               A.W.Thomas,Ann.Rev.Nucl.Part.Sci.45(1995)337
430: \bibitem{s8}  K.J.Eskola,V.J.Kolinen and C.A.Salgado(EKS),Eur.Phys.J.C9(1999)61.\\
431:               K.J.Eskola,V.J.Kolinen and P.V.Ruuskanen,Nucl.Phys.B535(1998)351.
432: \bibitem{s9}  D.M.Adle et al.(E772),Phys.Rev.Lett.,64(1990)2479
433: \bibitem{s10}  M.Hirai,S.Kumano,M.Miyama(HKM) ,Phys.Rev.D64(2001)034003.
434: \bibitem{s11}  M.A.Vasiliev,et.al.(E866),Phys.Rev.Lett.83(1999)2304.
435: \bibitem{s12}  M.B.Johnson et al.,Phys.Rev.,C65,025203(2002).
436: \bibitem{s13}  Francois Arleo,Phys.Lett.B532,231(2002).
437: \bibitem{s14}  C.G.Duan L.H.Song,L.J.Huo and G.L.Li,Eur.Phys.J.C29,557(2003)\\
438:                Duan Chungui,WANG Hong-Min£¬LI Guang-Lie ,  Chin.Phys.Lett., 19,485(2002).
439: \bibitem{s15}  S.Gavin and J.Milana,Phys.Rev.lett.,68(1992)1834.
440: \bibitem{s16}  D.Alde et al.,Phys.Rev.Lett.,66(1991)133.
441: \bibitem{s17}  J.Badier et al., Z.Phys.c20(1983)101.
442: \bibitem{s18}  S.J.Brodsky and P.Hoyer,Phys.lett.,B298,165(1993).
443: \bibitem{s19}  R.Baier et al.,Nucl.Phys.B484(1997)265;hep-ph/9804212.
444: \bibitem{s20}  M.Botje,Eur.Phys.J.C14(2000)285[hep-ph/9912439].
445: \bibitem{s21}  D.Geesaman et. al.,Fermolab Proposal No.E906,1999.
446: \bibitem{s22}  J.C.Peng et. al.,hep-ph/0007341;M.Asakawa et.
447: al.,KEK Report No 2000-11.
448: \end{thebibliography}
449: 
450: \vskip 2cm
451: \begin{center} Figure caption
452: \end{center}
453: Fig.1 The nuclear Drell-Yan cross section ratios
454: $R_{{A_1}/{A_2}}(x_1)$ on Fe to Be for various intervals M. Solid
455: curves correspond to nuclear effect on structure function  .
456: Dotted curves  show the combination of shadowing and energy loss
457: effect  with HKM cubic type of nuclear parton distributions .The
458: experimental data are taken from the E866[11].
459: 
460: 
461: Fig.2 The nuclear Drell-Yan cross section ratios
462: $R_{{A_1}/{A_2}}(x_1)$ on W to Be for various intervals M. The
463: comments is the same as Fig.1
464: 
465: 
466: \newpage
467: \begin{figure}
468: \centering
469: \includegraphics{Figs.eps}
470: \end{figure}
471: 
472: %\newpage
473: %\begin{figure}[htb]
474: %\begin{center}
475: 
476: %\epsfig{file=Fig.eps,width=23cm}
477: %\caption{%%%%}
478: %\end{center}
479: %\end{figure}
480: 
481: 
482: \end{document}
483: