1: % v1: 31.10.01, v2. 03.12.01
2: \documentstyle[12pt,psfig]{article}
3: \topmargin -48pt \textwidth 160mm \oddsidemargin 2mm
4: \pagestyle{plain} \textheight 240mm \evensidemargin 2mm
5: %\def\baselinestretch{1.35}
6: \begin{document}
7:
8: \title{$\rho/\omega$ properties from dilepton spectra in $p A$
9: reactions at 12 GeV
10: \thanks{Supported by Forschungszentrum J\"ulich}}
11: \author{E. L. Bratkovskaya \\[2mm]
12: {\normalsize Institut f\"{u}r Theoretische Physik, Universit\"{a}t Giessen}\\
13: {\normalsize 35392 Giessen, Germany}}
14: \date{ }
15: \maketitle
16:
17: \begin{abstract}
18: The dilepton production from $pC$ and $pCu$ collisions at
19: $T_{lab}=12$~GeV is calculated using the semi-classical BUU transport
20: model, that includes the off-shell propagation of vector mesons
21: nonperturbatively and calculates the width of the vector mesons
22: dynamically. It is found that the collisional broadening of the vector
23: meson width and dropping vector meson masses lead to a small enhancement
24: of the dilepton yield below the $\omega$ meson mass pole in $p Cu$ collisions
25: compared to $p C$, which is, however, not sufficient to explain the
26: enhanced production of dileptons as observed in the experiment by
27: the KEK-PS E325 collaboration. It is argued that such in-medium
28: effects are expected to be small due to kinematical reasons -- the
29: dominant part of vector mesons radiate dileptons and decay outside the
30: nucleus due to the finite formation time for $\rho$ and $\omega$
31: mesons, which is large by Lorentz covariance since the produced vector
32: mesons move with high velocity relative to the target nucleus.
33: \end{abstract}
34:
35: \vspace{0.3cm}\noindent
36: PACS: \ {25.75.Dw, 13.30.Ce, 12.40.Yx, 12.40.Vv, 25.40.-h}
37:
38: \noindent
39: Keywords: particle and resonance production; leptonic and semileptonic
40: decays; hadron models; vector-meson dominance; nucleon-induced
41: reactions
42:
43: \vspace*{5mm}
44: %---------------------------------------------------------------------
45:
46: The modification of hadron properties in nuclear matter is of
47: fundamental interest. Dilepton data from heavy-ion experiments at
48: SPS energies \cite{CERES,HELIOS} have provided first experimental
49: evidence for a change of the vector meson properties in the nuclear medium,
50: however, the heavy-ion data can be interpreted within different
51: scenarios of in-medium modifications, i.e. by the dropping mass
52: scenario or the collisional broadening approach
53: (cf. the reviews \cite{Gerry,CBRep98,Rapp00} and Refs. therein).
54:
55: Since in heavy-ion experiments the nuclear matter is probed at
56: different densities and temperatures within the complex dynamical
57: evolution, it is very useful to have independent experimental
58: information from photon-nucleus, pion-nucleus or proton-nucleus
59: reactions, where the properties of vector mesons are probed at normal
60: nuclear density or below.
61: Such information has been provided recently by the KEK-PS E325
62: collaboration, that has measured the dilepton spectrum in the mid-rapidity
63: region from $p C$ and $p Cu$ collisions at 12 GeV \cite{KEK}. A
64: significant difference in the mass spectra below the $\omega$ meson
65: mass pole has been observed between $p C$ and $p Cu$ reactions. This
66: difference has been interpreted as a modification of the vector meson
67: properties in nuclear matter \cite{KEK}.
68:
69: In this work the question of in-medium modifications of the $\rho,
70: \omega$ meson properties is studied within the semi-classical BUU
71: transport approach of Refs. \cite{Effe99gam,EffePhD}. This model
72: is based on the resonance concept of nucleon-nucleon and
73: meson-nucleon interactions at low invariant energy $\sqrt{s} \ $
74: \cite{TeisZP97} by adopting all resonance parameters from the
75: Manley analysis \cite{Manley}. The high energy collisions
76: (multiparticle production) -- above $\sqrt{s}$ = 2.6~GeV for
77: baryon-baryon collisions and $\sqrt{s}$ = 2.2~GeV for meson-baryon
78: collisions -- are described by the LUND string fragmentation
79: model FRITIOF \cite{FRITIOF} with a formation time $\tau_F=0.8$
80: fm/c. This aspect is similar to that used in the HSD
81: (Hadron-String-Dynamics) approach \cite{CBRep98,Ehehalt} and the
82: UrQMD model \cite{Bass}.
83: The threshold for string production of $\sqrt{s}$ = 2.6~GeV for
84: baryon-baryon production is not a sensitive parameter of the model
85: since the inclusive meson production cross sections are modeled to
86: be the same in the resonance and string model for 2.4 $\leq
87: \sqrt{s} \leq 3$ GeV. Furthermore, above $\sim \sqrt{s}$ = 2.2
88: GeV there are no longer distinct resonance excitations in
89: meson-baryon collisions such that the string formation model is a
90: suitable way to describe the continuum excitations of the baryon.
91: This combined resonance-string approach
92: allows to calculate particle production in baryon-baryon and
93: meson-baryon collisions from low to high energies. By input the
94: vacuum cross sections for $NN$ and $\pi N$ are reproduced with
95: good accuracy as demonstrated in Refs. \cite{CBRep98,EffePhD}. At
96: the energy of 12 GeV considered here essentially the string
97: formation and decay scheme is of relevance with respect to vector
98: mesons, which is well tested by experiment. Cross sections
99: involving baryon resonances might be disputed due to the lack of
100: experimental data, but the Manley analysis of the experimental
101: data \cite{Manley} provides a full determination of the baryon
102: resonance parameters as well as their branching to mesons.
103: In the present case the low energy resonance dynamics
104: play a minor role for the dilepton spectra anyway (see below).
105:
106: The collisional dynamics for proton-nucleus reactions, furthermore, is
107: described by the coupled-channel BUU transport approach
108: \cite{Effe99gam,EffePhD} that is based on the same elementary cross
109: sections. Furthermore, the model has been extended to include the
110: off-shell propagation of vector mesons nonperturbatively
111: \cite{Brat01pA} and to calculate the width of the vector mesons
112: dynamically, which is consistent with the vector meson
113: production/absorption amplitudes (or probabilities) \cite{Effe99gam}
114: (see below).
115:
116: The dilepton production within the resonance-string model is
117: treated (as described in Ref.~\cite{Brat01pA}) via the production
118: of resonances $R$ (baryonic or mesonic) in baryon-baryon,
119: meson-baryon or meson-meson collisions, which can couple to
120: dileptons directly or via a further decay into mesons or baryons.
121: The electromagnetic part of all conventional dilepton sources --
122: $\pi^0, \eta, \omega$ and $\Delta$ Dalitz decays, direct decay of
123: vector mesons $\rho, \omega$ -- are treated in the same way as
124: described in detail in Ref.~\cite{BCM00SIS}, where dilepton
125: production in elementary $pp$ and $pd$ reactions has been studied
126: and compared to the available data.
127:
128: The treatment of the off-shell propagation of vector mesons is
129: implemented as in Ref.~\cite{Brat01pA} following the general off-shell
130: transport dynamics from Refs. \cite{Cass_off}.
131: Note, that in present analysis the nucleons are propagated in the quasi
132: particle approximation with respect to an optical potential
133: \cite{CBJ00}, which becomes negligible at high energies.
134: In the approach \cite{Cass_off} the equations of motion
135: for the 'off-shell' test particles are extended by terms involving
136: their dynamical width $\Gamma(X,P)$ that is related to their
137: retarded self energy $\Sigma^{ret}(X,P)$ by $\Gamma(X,P)= - Im
138: \Sigma^{ret}(X,P)/P_0$, where $X,P$ denote the actual space-time
139: and 4-momentum coordinates. As shown in Refs. \cite{Cass_off} this
140: leads to a dynamical spectral function for the meson as
141: \begin{equation}
142: \label{spec} A(M;X,P) = \frac{2}{\pi}\frac{M^2
143: \Gamma(X,P)}{(M^2-M_0^2-Re \Sigma^{ret}(X,P))^2 + (M
144: \Gamma(X,P))^2}
145: \end{equation}
146: where the meson mass pole is also shifted by $Re \Sigma^{ret}(X,P)$.
147: Note, that nonrelativistic complex optical potentials are obtained from
148: the retarded self energies by dividing $\Sigma^{ret}(X,P)$ by twice the
149: energy $P_0$. On the other hand, $\Gamma(X,P)$ can be determined by the
150: in-medium meson decay and collision rates locally from the transport
151: calculation, i.e. for fixed $X$ and $P$. In principle, $Re \Sigma^{ret}$
152: and $Im \Sigma^{ret}$ are related to each other by
153: in-medium dispersion relations, however, in the actual transport
154: realization they are modeled independently to explore the sensitivity
155: to the dilepton spectra.
156: It should be noted that the modeling of the mass shift $Re \Sigma^{ret}$
157: and the total width $\Gamma$ is constrained to some extent by QCD sum
158: rules for the vector meson spectral function. As demonstrated in Ref.
159: \cite{Leupold} the sum rules require a significant shift of the
160: $\rho$-meson pole mass at nuclear matter density $\rho_0$ if a narrow
161: $\rho$ spectral function is involved. Furthermore, in case of a broad
162: spectral function for the $\rho$-meson some spectral strength is shifted
163: to lower invariant mass such that QCD sum rules can be fulfilled for
164: lower shift $Re \Sigma^{ret}$, but larger width $\Gamma$. However,
165: the uncertainties in the QCD condensates do not reliably constrain
166: the shape of the spectral function, i.e. its pole shift and width.
167:
168: In Ref. \cite{Brat01pA} the model described above has been applied to
169: study the in-medium modification of vector meson properties in dilepton
170: production in $p A$ collisions at SIS (1--4 GeV) energies. The effects
171: of a collisional broadening of the vector meson width (in line with
172: Refs. \cite{GKC97}) and dropping vector meson masses (according to the
173: Hatsuda and Lee \cite{H&L92} or Brown/Rho scaling \cite{BrownRho}) have
174: been investigated in Ref. \cite{Brat01pA}, too. It has been found
175: that the collisional broadening + 'dropping mass' scenario leads to an
176: enhancement of the dilepton yield in the invariant mass range $0.5 \leq
177: M \leq 0.75$ GeV, which is most pronounced for heavy systems (up to a
178: factor 2 for $p + Pb$ at 3--4 GeV).
179:
180: In this work the same transport model -- including collisional
181: broadening of the vector meson width (for $\Gamma(X,P)$) plus the
182: 'dropping mass' scenario ($Re \Sigma^{ret} \neq 0$) -- is applied
183: for dilepton production in $p C$ and $p Cu$ reactions at
184: $T_{lab}=12$~GeV as measured by the KEK-PS E325 collaboration
185: \cite{KEK}.
186:
187: Before showing the results for $p A$ reactions it is worth to
188: present the 'input' of the calculations, i.e. the calculated
189: dilepton invariant mass spectra $d\sigma/dM$ for $p + p$
190: collisions at 12 GeV (Fig. \ref{Fig_pp}). Note, that in Fig.
191: \ref{Fig_pp} (as well as in all further figures) a mass resolution
192: of $\Delta M=9.6$ MeV has been included that corresponds to the
193: resolution of Ref.~\cite{KEK} for the $\omega$ peak region. The
194: thin lines indicate the individual contributions from the
195: different production channels; {\it i.e.}~ starting from low $M$:
196: Dalitz decay $\pi^0 \to \gamma e^+ e^-$ (dot-dot-dashed line),
197: $\eta \to \gamma e^+ e^-$ (dotted line), $\Delta \to N e^+ e^-$
198: (dashed line), $\omega \to \pi^0 e^+ e^-$ (dot-dashed line), for
199: $M \approx $ 0.7 GeV: $\omega \to e^+e^-$ (dot-dashed line),
200: $\rho^0 \to e^+e^-$ (short dashed line). The full solid line
201: represents the sum of all sources considered here. As seen from
202: Fig. \ref{Fig_pp}, the $\rho$ and $\omega$ decay channels give the
203: dominant contributions at invariant masses $0.6 \le M \le 0.75$
204: GeV, where an excess of the dilepton yield in $p Cu$ relative to
205: $p C$ reactions was observed \cite{KEK}.
206: In this context it has to be pointed out that the inclusive yield of
207: $\eta, \rho$ and $\omega$ mesons from the transport approach for $pp$
208: reactions is consistent with the available experimental data in this
209: energy regime on the 30\% level \cite{CBRep98}.
210:
211: In Fig. \ref{Fig_bm} the calculated dilepton invariant mass spectra
212: $d\sigma/dM$ are displayed for $p + C$ (upper part) and $p + Cu$
213: collisions (lower part) at 12 GeV (including a mass resolution
214: of 9.6 MeV) without in-medium modifications (bare masses;
215: left part), and applying the collisional broadening + dropping mass
216: scenario (right part). The assignment of the individual lines is the
217: same as in Fig.\ref{Fig_pp}.
218: Comparing left and right panels one can see that the modification
219: of the dilepton spectrum by including collisional broadening and
220: dropping vector meson masses is rather moderate for the light $C$-target
221: (upper part of Fig.~\ref{Fig_bm}).
222: For the heavy $Cu$ target (lower part of Fig.~\ref{Fig_bm}) the shift of
223: the $\rho$ yield to lower invariant masses as well as the secondary
224: $\omega$ peak (at $\simeq 0.65$ GeV) is seen due to the dropping of
225: vector meson masses. However, the total enhancement of the dilepton yield
226: is quite small, i.e. less than a factor of 1.5 for $0.6 \le M \le 0.75$ GeV.
227:
228: The direct comparison of the model calculations to the
229: experimental data requires a precise knowledge of the experimental
230: acceptance since in the KEK-PS E325 experiment the low part of the
231: dilepton spectra was significantly suppressed by the geometry of
232: the detectors and nontrivial electronic cuts \cite{Ozawa_pr}.
233: Furthermore, the combinatorial background -- from a mixing of
234: $e^-\pi^+$ and $e^+\pi^-$ pairs -- was not subtracted from the data
235: \cite{KEK}. Since the full experimental acceptance is not available, a
236: normalization procedure -- following the analysis in Ref. \cite{KEK}
237: (described below) -- has been applied to compare the calculated spectra
238: to the experimental data \cite{KEK}.
239:
240: The experimental data -- solid histogram in Fig.~\ref{Fig_exp} --
241: have been fitted in Ref. \cite{KEK} by the sum of the amplitudes
242: of $\rho/\omega \to e^+e^-$, $\phi \to e^+e^-$, $\eta \to e^+e^-$
243: and the combinatorial background (dotted lines in
244: Fig.~\ref{Fig_exp}, denoted as 'bg') using a four parameter fit.
245: The mass shape of the $\rho, \omega$ and $\phi$ mesons has been
246: taken of the Breit-Wigner form with the natural widths 150, 8.41
247: and 4.43 MeV (cf. Fig. 3 from Ref. \cite{KEK}). The normalization
248: of the transport calculations is done in a such way that the
249: integral over the $\omega\to e^+e^-$ contribution for the bare
250: mass cases (i.e. without in-medium effects) is the same as for the
251: experimentally fitted $\omega$ contribution (short dotted lines),
252: that also employs free spectral functions for the $\rho$ and
253: $\omega$. Note, that the physical conclusions drawn in \cite{KEK}
254: relate to the same fitted spectra in case of pure vacuum decays.
255: The same normalization factor is then used for the calculations
256: with collisional broadening + dropping mass scenario
257: which allows to study separately the change of the dilepton
258: spectra when including in-medium effects.
259: The combinatorial background (taken from Ref. \cite{KEK}) is added
260: finally to the calculated spectra.
261:
262: It is worth to point out, that an application of the acceptance
263: cuts as described in Ref. \cite{KEK}, i.e. the rapidity cut $0.6
264: \le y \le 2.2$, the transverse momentum cut $p_T\le 1.5$ GeV, the
265: cut in the opening angle of the $e^+e^-$ pairs $50^0 \le
266: \Theta_{open} \le 150^0$, and the restriction that the electron
267: and positron are detected in different detector arms (cf. Fig. 1
268: in Ref. \cite{KEK}), leads to a suppression of the dilepton yield
269: by about of a factor of 9 for $M\ge 0.5$ GeV and does not change
270: the shape of the spectra. At low $M$ these cuts suppress the
271: dilepton spectra more effective, however, further 'electronic'
272: acceptance cuts must be applied to reproduce exactly the shape of
273: the measured spectra. Due to this reason only the $M$-interval $0.45
274: \le M\le 0.9$ GeV is considered, where an excess of the dilepton yield
275: has been observed, to reduce the sensitivity to unknown cuts in the
276: lower mass regime.
277:
278: The comparison of the calculated results with the experimental data
279: (histograms) is shown in Fig.~\ref{Fig_exp} for $p + C$ (upper part)
280: and $p + Cu$ collisions (lower part). The solid lines indicate the
281: calculations without in-medium modifications (bare masses), the
282: dot-dashed lines correspond to a calculation including the collisional
283: broadening + dropping mass scenario. The individual contributions are
284: indicated as '$\rho$', '$\omega$'; the lines marked as 'sum+bg' show
285: the sum of the dilepton contributions and the combinatorial background
286: (dotted lines 'bg') as taken from Ref. \cite{KEK}.
287:
288: Note, that the $\rho$-contribution in Fig. \ref{Fig_exp} (as well
289: as in Figs. \ref{Fig_pp},\ref{Fig_bm}) is asymmetric in mass due
290: to the fact that the dilepton decay leads to a multiplication of
291: the $\rho$-spectral function by $\sim 1/M^3$
292: since the electromagnetic decay includes a factor $\sim 1/M^4$
293: from the virtual photon propagator and a factor $\sim M$ from
294: final state phase space (cf. Ref. \cite{BCM00SIS}). This is in
295: contrast to the simplifying assumption made in Ref. \cite{KEK},
296: where the mass depending branching ratio to dileptons has been
297: taken as a constant;
298: the same holds for the $\omega$ mesons. However, accounting for
299: the $1/M^3$ factor for the $\omega$ and $\rho$ dilepton yields
300: changes the experimentally fitted curves by a few percent only,
301: which does not influence the normalization used here (as well as
302: the conclusions of Ref. \cite{KEK}).
303:
304: One can see from Fig. \ref{Fig_exp} that the $p C$ data are
305: roughly reproduced by the calculations within the statistics
306: achieved experimentally, whereas the $p Cu$ data are
307: underestimated for $0.55 \le M \le 0.75$ GeV. Thus, the excess of
308: the dilepton yield in $p Cu$ relative to $p C$ reactions can not
309: be explained neither by the bare mass scenario (as in Ref.
310: \cite{KEK}) nor by including in-medium effects such as collisional
311: broadening and dropping vector masses, which gives only a tiny
312: enhancement of the dilepton yield (as mentioned above) by less
313: than a factor of 1.5.
314:
315: Note, that the calculated yield is higher than the results of the
316: fit procedure from Ref. \cite{KEK} for $0.55 \le M\le 0.75$ GeV.
317: That is due to the fact that the calculated $\rho$ contribution is
318: slightly higher than the fitted one (corrected by the factor
319: $\sim 1/M^3$). In the simulations of Ref. \cite{KEK} the assumption
320: was made, that all vector mesons are produced from primary
321: nucleon-nucleon collisions at 12 GeV with equal $\rho$ and
322: $\omega$ production cross sections (in line with the experimental
323: data on $\rho$, $\omega$ production at 12 GeV/c from Ref.
324: \cite{Rhoprod}). This leads to an about equal dilepton yield from
325: $\rho$ and $\omega$ mesons, especially using the assumption about
326: the instantaneous decay of vector mesons into dileptons, i.e.
327: without performing the time integration of the dilepton radiation
328: during the life time of the mesons (before their nonleptonic
329: decays such as $\rho^0\to \pi^+\pi^-, \ \omega\to
330: \pi^+\pi^-\pi^0$). Note, that the time integration procedure, as
331: described in Ref. \cite{LKB96}, is important for a correct
332: evaluation of the dilepton spectra due to the absorption of vector
333: mesons in the nucleus.
334:
335: In the resonance-string model the vector mesons at 12 GeV are dominantly
336: produced in first nucleon-nucleon collisions via baryon-baryon
337: ($BB$) string excitations and decays with production cross
338: sections as in Ref. \cite{Rhoprod}. However, $\rho, \omega$ mesons
339: can be also produced by secondary meson-baryon ($mB$) collisions
340: and the $\rho$ mesons additionally by the decay of baryonic
341: resonances (e.g. $D_{13}(1520)$) formed in primary $BB$ (less frequent) or
342: secondary $mB$ collisions. The $mB$ and resonance production
343: mechanisms are more important for heavy nuclei and modify the
344: $\rho$ mesons spectrum essentially at low invariant mass.
345: Nevertheless, at 12 GeV -- contrary to the low energy $pA$ reactions --
346: the resonance dynamics becomes less important since the probability (or
347: cross section) to form baryonic resonances in $BB$ or $mB$ collisions
348: (which is high at thresholds) decreases with energy (see, e.g.
349: \cite{Effe99gam}).
350: In numbers: in $p C$ collisions $\sim 90$\% of $\rho$ and $\sim 95$\%
351: of $\omega$ mesons are produced by $BB$ string decays, $\sim 3$\% of
352: $\rho$ and $\sim 5$\% of $\omega$ by the $mB$ (high energy) string
353: formation or by secondary $mB$ collisions and $\sim 7$\% of $\rho$ by
354: the decays of $D_{13}(1520)$ and other baryonic resonances coupled to
355: the $\rho$. For the heavy $Cu$ target $\sim 70$\% of $\rho$ and $\sim
356: 85$\% of $\omega$ mesons are produced by $BB$ string decays, $\sim
357: 10$\% of $\rho$ and $\sim 15$\% of $\omega$ by $mB$ interaction and
358: $\sim 15$\% of $\rho$ by the decays of $D_{13}(1520)$ and other
359: baryonic resonances. Note, that the counting of $\omega$ mesons
360: produced by $mB$ interactions is different from the $\rho$'s due to the
361: fact that in low energy secondary $mB$ collisions predominantly
362: baryonic resonances are excited which sequentially might decay to the
363: nucleon-$\rho$ channel. In the latter case the corresponding $\rho$
364: mesons are counted in the resonance decay channel inspite of
365: originating from a $mB$ collision. On the other hand the resonances,
366: that couple in part strongly to the $\rho$ but not to the $\omega$, can
367: be also produced in a nucleon-nucleon collision and be reabsorped in
368: the $RN \rightarrow NN$ reaction by detailed balance.
369:
370: Note, that the relative numbers quoted above depend on the
371: number of resonances that are excited during the $pA$ reaction.
372: These in turn depend on the threshold energy employed for the
373: string excitation mechanism. However, actual simulations with
374: reduced/enhanced 'string thresholds' were found to modify the
375: resulting dilepton spectra by less than 10\%, which is much lower
376: than the descrepancy (missing yield) in the dilepton spectra below
377: the $\omega$ peak for $p Cu$ in Fig. \ref{Fig_exp}.
378: Furthermore, the resonance dynamics essentially modify the dilepton mass
379: spectra below invariant masses of 0.5 GeV. Indeed, when switching off
380: all resonances in the transport calculation and producing the meson
381: directly, the dilepton yield from low mass $\rho$-mesons is modified by
382: up to a factor of 2. However, summing up all contributing channels the
383: modification of the total spectrum only amounts to a few percent. Thus
384: the systematic uncertainties of the transport calculations in this case
385: -- when normalized to the experimental $\omega$-peak -- are less than
386: 10\% in the invariant mass range presented in Fig. \ref{Fig_exp}.
387:
388: To clarify the influence of the nucleon optical potential on dilepton
389: production at the energy of 12 GeV, additional calculations have been
390: performed for $pC$ and $pCu$ within the cascade mode, i.e. without any
391: nucleon optical potential. The results stay the same within the
392: statistical errorbars achieved in the calculations.
393:
394:
395: The $\rho$ mesons live for a very short time and radiate dileptons
396: with high intensity (as compared to the $\omega$ mesons) before
397: decaying into two pions or being absorbed in the nuclear interior.
398: The $\omega$ mesons are longer living particles and they can be
399: absorbed by nucleons in the nucleus, too, before their natural
400: 'death', i.e. 3 pion decay. The $\omega$ absorption violates the
401: $A^{2/3}$ scaling (where $A$ is the target mass number), i.e. the
402: scaling of the dilepton rate from $p A$ reactions is expected to
403: be higher by a factor of $A^{2/3}$ than in $p+p$ collisions
404: \cite{KEK}. As seen from Fig. \ref{Fig_pp} and Fig. \ref{Fig_bm}
405: in the calculations the $A^{2/3}$ scaling is practically fulfilled
406: for the light $C$ target, however, for the heavy $Cu$ target the
407: dilepton yield from $\omega$ mesons
408: is $\sim 30$\% lower as expected from $A^{2/3}$ scaling due
409: to $\omega$ absorption,
410: which proceeds dominantly through the $\omega N \rightarrow \pi N$
411: channel. The absorbed $\omega$'s then can no longer decay to
412: dileptons which in vacuum would radiate $e^+e^-$ pairs (in their
413: rest frame) for about 23 fm/c.
414:
415: For a better understanding, why the in-medium effects are so
416: small, the average density distribution of a $Cu$-nucleus at rest
417: in the laboratory is shown in Fig. \ref{Fig_bz} (upper left part)
418: as well as the spatial distribution in the first $pN$ collisions
419: (upper right part). Here the spatial distribution ${1\over
420: b}{dN\over db dz}$ is displayed in cylindrical coordinates
421: $b=(x^2+y^2)^{1/2}$ and $z$, where $z$ is directed along the beam
422: axis and the proton is impinging on the nucleus from the left
423: side. The lower part of Fig. \ref{Fig_bz} displays the spatial
424: distribution for $\rho$-meson (left part) and $\omega$-meson
425: (right part) decays to dileptons for the bare mass case. As
426: mentioned above, at 12 GeV bombarding energy most of the $\omega$
427: and $\rho$ mesons -- formed in primary $pN$ collisions -- move
428: with a high velocity through the target. Contrary to low energy
429: collisions (cf. Fig. 1 from Ref. \cite{Brat01pA}), most of the
430: $\rho$'s and $\omega$'s are radiating dileptons and decaying
431: outside the nucleus; this holds especially for $\omega$ mesons
432: which have a longer life time such that the $\omega$ spatial
433: distribution is more elongated in beam direction then the $\rho$.
434: In numbers: for the $C$ target only $\sim 2.5$\% of $\omega$ and
435: $\sim 14$\% of $\rho$ mesons decay to dileptons inside the
436: nucleus; for the $Cu$ target $\sim 7$\% of $\omega$ and $\sim
437: 32$\% of $\rho$ mesons radiate $e^+e^-$ pairs inside. For the
438: collisional broadening + dropping mass scenario the fraction of
439: the $\rho$ mesons decaying to lepton pairs inside increases to
440: $\sim 45$\% for the $Cu$ target due to an enhanced production of
441: $\rho$ mesons with low masses. Thus the dropping mass and
442: collisional broadening effects -- which are proportional to the
443: baryon density -- do not influence the dilepton spectra very much.
444:
445: To illustrate the dynamics of $p A$ reactions at 12 GeV, the
446: laboratory rapidity distribution $d\sigma/dy$ is presented in Fig.
447: \ref{Fig_yp} (upper part) for $\rho\to e^+e^-$ (dashed lines) and
448: $\omega\to e^+e^-$ (solid lines) for $p + C$ (left) and $p + Cu$
449: (right) at 12 GeV. The arrows indicate the experimental
450: rapidity window ($0.6\le y\le 2.2$). The lower part of Fig. \ref{Fig_yp}
451: shows the momentum distribution $d\sigma/dp$ for $\rho\to e^+e^-$
452: (dashed lines) and $\omega\to e^+e^-$ (solid lines) for $p + C$ (left
453: part) and $p + Cu$ (right part) at 12 GeV calculated with
454: the experimental cuts on rapidity ($0.6\le y\le 2.2$)
455: and $p_T < 1.5$ GeV/c.
456:
457: One can see from the upper part of Fig. \ref{Fig_yp} that the
458: rapidity distribution for $\rho \to e^+e^-$ and $\omega\to e^+e^-$
459: are similar for $p C$ collisions; only the $\rho$ distribution is
460: shifted slightly to the lower rapidity region due to the fact that
461: the $\rho$ spectral function is broad and $\rho$ mesons can be
462: produced with low masses, too. This shift is more pronounced for
463: the heavy $Cu$ target, where more $\rho$ mesons are formed with
464: lower masses via baryonic resonance decays. Also, the $\omega$
465: absorption effect for the dilepton yield, as discussed above, is
466: well seen.
467:
468: The experimental acceptance selects dilepton events at
469: mid-rapidity with momentum distributions shown in the lower part
470: of Fig. \ref{Fig_yp}. It is seen, that the $\rho$'s and
471: $\omega$'s are very fast; the corresponding Lorentz
472: $\gamma$-factors (in the laboratory frame) are in the range $1<
473: \gamma <6.5$. One has to recall that the vector mesons produced
474: by strings can radiate dileptons only after they are formed as
475: physical particles (not as pre-resonant or quark states
476: which have a continuum spectral function).
477: The formation time $\tau_F$, which denotes the time between the
478: formation and fragmentation of the string in the individual
479: hadron-hadron center-of-mass system is taken to be $\tau_F=0.8$ fm/c
480: (cf. \cite{CBRep98,Ehehalt,Bass}). Due to covariance the hadrons with
481: finite momentum $p$ then are formed at the actual time $\Delta t=\gamma
482: \tau_F$ (in the laboratory frame). During this time $\Delta t$ the
483: pre-formed particles (strings) are moving with high velocities and
484: start to radiate dileptons only either outside the nucleus or closer to
485: the nuclear surface (cf. Fig. \ref{Fig_bz}). This kinematical effect
486: suppresses in-medium effects. It is worth to point out, that for low
487: energy $p A$ collisions the in-medium effects are expected to be larger
488: since the vector mesons are formed (and decay to dileptons) more
489: abundantly inside the nucleus \cite{Brat01pA}.
490:
491: One might argue that the formation time $\tau_F$ could be substantially
492: smaller than the value of 0.8 fm/c adopted here (or values of 1-2 fm/c
493: depending on the hadron-species as used in UrQMD \cite{Bass}). In fact,
494: when assuming $\tau_F=0$, the amount of in-medium $\rho$ decays for
495: $p + Cu$ increases to $\sim 50$\% (and to $\sim 65$\% for the
496: collisional broadening + dropping mass scenario) and to $\sim 12$\% for
497: the $\omega$ mesons, respectively. In this limit the calculated
498: dilepton yield indeed becomes comparable to the experimental spectrum
499: for $p + Cu$. Furthermore, the transport calculation for $\tau_F=0$
500: roughly correspond to the Glauber limit for vector meson production and
501: absorption in $p + A$ reactions. However, as shown in the context of
502: heavy-ion collisions \cite{Ehehalt}, a reduction of $\tau_F$ leads to
503: an unphysical production multiplicity of mesons (as well as to rapidity
504: and momentum distributions of baryons and mesons that are incompatible
505: with the experimental data) since the rescattering rate becomes too
506: high (some examples for $p + A$ reactions are also shown in Ref.
507: \cite{EffePhD}). Furthermore, experimental studies on $\bar p$
508: production and reabsorption at 12.3 and 17.5 GeV/c in $p + A$ reactions
509: by the E910 collaboration \cite{pbarAGS} have shown that the formation
510: time of antiprotons might be even larger than 0.8 fm/c in order to
511: describe the low amount of $\bar p$ reabsorption in $p + A$ reactions.
512:
513:
514: In summary: the dilepton production from $p C$ and $p Cu$
515: collisions at $T_{lab}=12$~GeV has been calculated using a
516: combined resonance-string transport approach. It has been found
517: that the enhanced production of dileptons below the $\omega$ meson
518: mass pole in $p Cu$ collisions compared to $p C$ -- as observed in
519: the experiment by the KEK-PS E325 collaboration -- cannot be
520: explained by collisional broadening nor dropping vector meson masses.
521: The missing yield cannot be attributed to systematic
522: uncertainties of the transport approach with respect to the low energy
523: resonance dynamics as discussed throughout the paper.
524: It has been argued, furthermore, that such in-mediums effects are
525: expected to be small due to kinematical reasons, i.e. the dominant part
526: of vector mesons radiate dileptons and decay outside the nucleus due to
527: a finite hadron formation time $\Delta t=\gamma\tau_F$.
528:
529: Furthermore, the dropping mass scenario should be questionable
530: especially for $\rho$ mesons within the momentum range shown in
531: Fig. \ref{Fig_yp} since dispersion relations -- as employed by
532: Eletsky and Ioffe \cite{ElIof} and Kondratyuk et al.
533: \cite{Kond98rho} -- lead to repulsive $\rho$ meson potentials for
534: such high momenta. To clarify such experimental and/or theoretical
535: inconsistencies high accuracy dilepton spectra within different
536: momentum bins are urgently needed
537: as well as independent data on differential vector-meson spectra
538: from hadronic decays to reduce in addition the theoretical
539: uncertainties on the cross sections that enter the transport analysis.
540:
541:
542: \vspace*{5mm}
543: The author is grateful to U. Mosel for pointing out this problem,
544: to K. Ozawa for the explanations of the KEK-PS E325 experiment,
545: and to W. Cassing for valuable discussions and a critical
546: reading of the manuscript.
547:
548: %------------------------------------------------------------------------
549: \begin{thebibliography}{999}
550: \bibitem{CERES}
551: G. Agakichiev et al., Phys. Rev. Lett. 75 (1995) 1272;
552: Th. Ullrich et al., Nucl. Phys. A 610 (1996) 317c;
553: A. Drees, Nucl. Phys. A 610 (1996) 536c.
554: \bibitem{HELIOS}
555: M. A. Mazzoni, Nucl. Phys. A 566 (1994) 95c;
556: M. Masera, Nucl. Phys. A 590 (1995) 93c;
557: T. {\AA}kesson et al., Z. Phys. C 68 (1995) 47.
558: \bibitem{Gerry}
559: G.E. Brown, M. Rho, nucl-th/0103102, Phys. Rep., in press
560: \bibitem{CBRep98}
561: W. Cassing, E. L. Bratkovskaya,
562: Phys. Rep. 308 (1999) 65.
563: \bibitem{Rapp00}
564: R. Rapp, J. Wambach, Adv. Nucl. Phys. 25 (2000) 1.
565: \bibitem{KEK}
566: K. Ozawa et al., Phys. Rev. Lett. 86 (2001) 5019.
567: \bibitem{Effe99gam}
568: M. Effenberger, E. Bratkovskaya and U. Mosel,
569: Phys. Rev. C 60 (1999) 044614.
570: \bibitem{EffePhD}
571: M. Effenberger, Ph.D. Thesis, Univ. of Giessen, 1999;
572: http://theorie.physik.uni-giessen.de/ftp.html.
573: \bibitem{TeisZP97}
574: S. Teis, W. Cassing, M. Effenberger, A. Hombach, U. Mosel,
575: and Gy. Wolf, Z. Phys. A 356 (1997) 421;
576: Z. Phys. A 359 (1997) 297.
577: \bibitem{Manley}
578: D. M. Manley and E. M. Saleski, Phys. Rev. D 45 (1992) 4002.
579: \bibitem{FRITIOF}
580: B. Anderson, G. Gustafson and Hong Pi, Z. Phys. C 57
581: (1993) 485.
582: \bibitem{Ehehalt}
583: W. Ehehalt and W. Cassing, Nucl. Phys. A 602 (1996) 449.
584: \bibitem{Bass}
585: S. A. Bass et al., Prog. Part. Nucl. Phys. 42 (1998) 279;
586: J. Phys. G 25 (1999) 1859.
587: \bibitem{Brat01pA}
588: E.L. Bratkovskaya, Nucl. Phys. A 696 (2001) 309.
589: \bibitem{BCM00SIS}
590: E.L. Bratkovskaya, W. Cassing and U. Mosel,
591: Nucl. Phys. A 686 (2001) 476.
592: \bibitem{Cass_off}
593: W. Cassing and S. Juchem, Nucl. Phys. A 665 (2000) 377;
594: A 672 (2000) 417.
595: \bibitem{CBJ00}
596: W. Cassing, E.L. Bratkovskaya and S. Juchem,
597: Nucl. Phys. A 674 (2000) 249.
598: \bibitem{Leupold}
599: S. Leupold, W. Peters and U. Mosel, Nucl. Phys. A 628 (1998) 311.
600: \bibitem{GKC97}
601: W. Cassing, Ye. S. Golubeva, A. S. Iljinov, and L. A. Kondratyuk,
602: Phys. Lett. 396 B (1997) 26;
603: Ye. S. Golubeva, L. A. Kondratyuk and W. Cassing,
604: Nucl. Phys. A 625 (1997) 832.
605: \bibitem{H&L92}
606: T. Hatsuda and S. Lee, Phys. Rev. C 46 (1992) R34.
607: \bibitem{BrownRho}
608: G.E. Brown and M. Rho, Phys. Rev. Lett. 66 (1991) 2720.
609: \bibitem{Ozawa_pr}
610: K. Ozawa, private communication.
611: \bibitem{Rhoprod}
612: V. Blobel et al., Phys. Lett. 48 B (1974) 73.
613: \bibitem{LKB96}
614: G.Q. Li, C.M. Ko, and G.E. Brown, Nucl. Phys. A 611 (1996) 539.
615: \bibitem{pbarAGS}
616: I. Chemakin et al., Phys. Rev. C 64 (2001) 064908.
617: \bibitem{ElIof}
618: V.L. Eletsky and B.L. Ioffe, Phys. Rev. Lett. 78 (1997) 1010.
619: \bibitem{Kond98rho}
620: L. A. Kondratyuk, A. Sibirtsev, W. Cassing, Ye. S. Golubeva, and
621: M. Effenberger, Phys. Rev. C 58 (1998) 1078.
622: \end{thebibliography}
623:
624:
625: \newpage
626:
627: \phantom{a}\vspace*{3cm}
628: \begin{figure}[h]
629: \centerline{\psfig{figure=pp12.eps,width=10cm}}
630: \vspace*{0.5cm}
631: \caption{
632: The calculated dilepton invariant mass spectra $d\sigma/dM$ for $p + p$
633: collisions at 12 GeV (including a mass resolution of 9.6 MeV).
634: The thin lines indicate the individual contributions from the different
635: production channels; {\it i.e.}~ starting from low $M$:
636: Dalitz decay $\pi^0 \to \gamma e^+ e^-$ (dot-dot-dashed line),
637: $\eta \to \gamma e^+ e^-$ (dotted line),
638: $\Delta \to N e^+ e^-$ (dashed line),
639: $\omega \to \pi^0 e^+ e^-$ (dot-dashed line),
640: for $M \approx $ 0.7 GeV: $\omega \to e^+e^-$ (dot-dashed line),
641: $\rho^0 \to e^+e^-$ (short dashed line).
642: The full solid line represents the sum of all sources considered here.}
643: \label{Fig_pp}
644: \end{figure}
645:
646:
647: \begin{figure}[ht]
648: \centerline{\psfig{figure=bare-med.eps,width=15cm}}
649: \vspace*{0.5cm}
650: \caption{
651: The calculated dilepton invariant mass spectra $d\sigma/dM$ for $p + C$
652: (upper part) and $p + Cu$ collisions (lower part) at 12 GeV (including
653: a mass resolution of 9.6 MeV) without in-medium
654: modifications (bare masses; left part), and applying the collisional
655: broadening + dropping mass scenario (right part).
656: The assignment of the individual lines is the same as in
657: Fig. \protect\ref{Fig_pp}.}
658: \label{Fig_bm}
659: \end{figure}
660:
661: \begin{figure}[ht]
662: \centerline{\psfig{figure=exp-ccu.eps,width=14cm}}
663: %\vspace*{0.5cm}
664: \caption{ The calculated dilepton invariant mass spectra $dN/dM$
665: for $p + C$ (upper part) and $p + Cu$ collisions (lower part) at
666: 12 GeV (including an experimental mass resolution of 9.6 MeV)
667: without in-medium modifications (bare masses, solid lines), and
668: applying the collisional broadening + dropping mass scenario
669: (dot-dashed lines) in comparison to the experimental data
670: \protect\cite{KEK} (histograms). The dotted lines show the
671: combinatorial background from Ref. \protect\cite{KEK} while the
672: short dotted lines represent the fit to the $\omega$ decay
673: spectrum from Ref. \protect\cite{KEK}.} \label{Fig_exp}
674: \end{figure}
675:
676: \begin{figure}[ht]
677: \centerline{\psfig{figure=bz-cu.eps,width=15cm}}
678: \caption{
679: Upper left part: the average density distribution of a $Cu$-nucleus
680: at rest in the laboratory; upper right part: the spatial distribution
681: in the first $pN$ collisions; lower parts: the spatial distribution
682: for $\rho$-meson (left) and $\omega$-meson (right) decays
683: to dileptons. The contour lines correspond to nucleon densities
684: of 0.1$\rho_0$, 0.4$\rho_0$, 0.6$\rho_0$ and 0.8$\rho_0$,
685: respectively, and the dark shaded area to $\rho \ge 0.8\rho_0$.}
686: \label{Fig_bz}
687: \end{figure}
688:
689: \begin{figure}[ht]
690: \centerline{\psfig{figure=yp-cut.eps,width=15cm}}
691: \caption{
692: Upper part: the laboratory rapidity distributions $d\sigma/dy$
693: for $\rho\to e^+e^-$ (dashed lines) and $\omega\to e^+e^-$ (solid lines)
694: for $p + C$ (left) and $p + Cu$ (right) at 12 GeV.
695: The arrows indicate the experimental rapidity window ($0.6\le y\le 2.2$).
696: Lower part: the momentum distribution $d\sigma/dp$
697: for $\rho\to e^+e^-$ (dashed lines) and $\omega\to e^+e^-$ (solid lines)
698: for $p + C$ (left) and $p + Cu$ systems (right) at 12 GeV
699: calculated with experimental cuts in rapidity ($0.6\le y\le 2.2$)
700: and $p_T$ ($p_T < 1.5$ GeV/c).}
701: \label{Fig_yp}
702: \end{figure}
703:
704: \end{document}
705:
706: #!/bin/csh -f
707: # Uuencoded gz-compressed .tar file created by csh script uufiles
708: # For more info (11/95), see e.g. http://xxx.lanl.gov/faq/uufaq.html
709: # If you are on a unix machine this file will unpack itself: strip
710: # any mail header and call resulting file, e.g., figures.uu
711: # (uudecode ignores these header lines and starts at begin line below)
712: # Then say csh figures.uu
713: # or explicitly execute the commands (generally more secure):
714: # uudecode figures.uu ; gunzip figures.tar.gz ;
715: # tar -xvf figures.tar
716: # On some non-unix (e.g. VAX/VMS), first use editor to change filename
717: # in "begin" line below to figures.tar-gz , then execute
718: # uudecode figures.uu
719: # gzip -d figures.tar-gz
720: # tar -xvf figures.tar
721: #
722: uudecode $0
723: chmod 644 figures.tar.gz
724: gunzip -c figures.tar.gz | tar -xvf -
725: rm $0 figures.tar.gz
726: exit
727:
728: