1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}

2: %\documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}

3: \usepackage{graphicx}% Include figure files

4: \usepackage{dcolumn}% Align table columns on decimal point

5: \usepackage{bm}% bold math

6: \bibliographystyle{apsrev}

7: \begin{document}

8: \title{Short timescale behavior of colliding heavy nuclei at intermediate 

9: energies}

10: 

11: \author{S. Hudan}

12: \author{R.T. de Souza} 

13: \affiliation{

14: Department of Chemistry and Indiana University Cyclotron Facility \\ 

15: Indiana University, Bloomington, IN 47405} 

16: 

17: 

18: \author{A. Ono}

19: \affiliation{

20: Department of Physics, Tohoku University, Sendai 980-8578, Japan}

21: 

22: \date{\today}

23:   

24: \begin{abstract}

25: An Antisymmetrized Molecular Dynamics model is used to explore the

26: collision of $^{114}$Cd projectiles with $^{92}$Mo target nuclei at 

27: E/A=50 MeV over a broad range in impact parameter. 

28: The atomic number (Z), velocity, and emission pattern of the

29: reaction products are examined as a function of the impact parameter and the

30: cluster recognition time. 

31: The non-central collisions are found to be essentially binary in character 

32: resulting in the

33: formation of an excited projectile-like fragment (PLF$^*$) 

34: and target-like fragment (TLF$^*$). The decay of these fragments occurs 

35: on a short timescale, 100$\le$t$\le$300 fm/c. 

36: The average excitation energy deduced for the PLF$^*$ and TLF$^*$ `saturates

37: for mid-central collisions, 3.5$\le$b$\le$6 fm, with

38: its magnitude depending on the cluster recognition time. 

39: For short cluster recognition times (t=150 fm/c), 

40: an average excitation energy as high as $\approx$6 MeV is predicted. 

41: Short timescale emission leads to a loss of initial correlations

42: and results in features such as an anisotropic

43: emission pattern of both IMFs and alpha particles emitted from the 

44: PLF$^*$ and TLF$^*$ in peripheral collisions.

45: 

46: 

47: \end{abstract}

48: \pacs{PACS number(s): 25.70.Mn} 

49: 

50: \maketitle

51: 

52: \section{Introduction}

53: 

54: Collision of two heavy-ions at intermediate energies can result in the

55: production of a multi-particle final state \cite{Bowman91, deSouza91, Marie97}. 

56: These multi-particle final states have been experimentally characterized by 

57: a wide variety of signals including fragment multiplicity 

58: \cite{Bowman91, deSouza91}, 

59: size distributions \cite{Rivet98, Ogilvie91}, 

60: emission timescales \cite{Kim91, Cornell95, Cornell96, Beaulieu00}, 

61: scaling behavior \cite{Elliott02, Scharenberg01, Porile89} 

62: and the attained excitation energy \cite{Cussol93}.

63: For large fragment multiplicity, within a thermodynamic approach, 

64: such multi-fragment states have been interpreted as a transition 

65: of the finite nuclear system from a liquid to a gaseous phase 

66: \cite{Gross90, Bondorf85, Pochodzalla95, Viola04a}. Recent work has 

67: investigated the robustness of this conclusion by examining the 

68: influence  of the surface, through

69: the density dependence of the entropy, on the stability of the nuclear 

70: droplet against fragmentation \cite{Toke03, Sobotka04}. All these approaches

71: however focus on the thermodynamic stability of the system.

72: In reality, the decaying system is 

73: formed by the collision dynamics which may not equilibrate 

74: all degrees of freedom equally \cite{Moretto93, Viola04}.

75: In order to understand both the formation and decay of excited 

76: nuclear systems involved in the collision process, 

77: microscopic approaches have also  

78: been followed \cite{Bauer87, Aichelin91, Danielewicz91, 

79: Schnack97, Morawetz00, Wada04}. 

80: In order to make 

81: direct comparison with experimental

82: data such microscopic models typically utilize a 

83: a two-stage approach.

84: In the first phase, a dynamical model is used to describe the collision 

85: dynamics. Clusters produced in this phase are subsequently de-excited by 

86: a statistical model.

87: Such a two-stage approach typically views the statistical decay stage 

88: as decoupled from the dynamical stage that preceded it. 

89: In the present work we examine the validity of such a de-coupled 

90: hybrid approach. Specifically, we utilize a microscopic model, 

91: the Antisymmetrized Molecular Dynamics model,

92: to investigate how the collision proceeds on short timescales and how the 

93: reaction characteristics evolve with impact parameter.

94: In addition, we examine whether initial correlations, 

95: existing at short times, 

96: survive the decay stage and how they are manifested in final distributions.

97: 

98: \section{Description of the AMD model}

99: 

100: To describe the dynamical stage of intermediate energy heavy-ion

101: collisions, we utilize the antisymmetrized molecular dynamics (AMD)

102: model \cite{ONOa,ONOb,ONOj,ONO-ppnp}. For the present work, we use the

103: same version of AMD as Ref. \cite{ONOj} which has been used to describe the

104: multifragmentation reaction of the central $\mathrm{Xe}+\mathrm{Sn}$

105: collisions at 50 MeV/nucleon.

106: 

107: The description of the dynamics of fragmentation is, in principle, a very complicated

108: quantum many-body problem.  In the exact solution of the many-body

109: time-dependent Schr\"odinger equation, the

110: intermediate and final states should be very complicated states

111: containing a huge number of reaction channels corresponding to

112: different fragmentation configurations.  The AMD model respects the

113: existence of channels, while it neglects some of the interference

114: among them.  Namely, the total many-body wave function

115: $|\Psi(t)\rangle$ is approximated by

116: \begin{equation}

117: |\Psi(t)\rangle\langle\Psi(t)|\approx

118: \int

119: \frac{|\Phi(Z)\rangle\langle\Phi(Z)|}{\langle\Phi(Z)|\Phi(Z)\rangle}

120:  w(Z,t)dZ,

121: \label{eq:AMDensemble}

122: \end{equation}

123: where each channel wave function $|\Phi(Z)\rangle$ is parametrized by

124: a set of parameters $Z$, and $w(Z,t)$ is the time-dependent

125: probability of each channel.

126: 

127: In AMD, we choose the Slater determinant of Gaussian wave packets as

128: the channel wave function

129: \begin{equation}

130: \langle\mathbf{r}_1\ldots\mathbf{r}_A|\Phi(Z)\rangle \propto

131: \det_{ij} \biggl[ \exp\Bigl\{-\nu(\mathbf{r}_i-\mathbf{Z}_j/\sqrt{\nu})2\Bigr\}

132:                                \chi_{\alpha_j}(i) \biggr],

133: \label{eq:AMDWaveFunction}

134: \end{equation}

135: where $\chi_{\alpha_i}$ are the spin-isospin states with

136: $\alpha_i=p\uparrow, p\downarrow, n\uparrow,$ or $n\downarrow$. Thus,

137: the many-body state $|\Phi(Z)\rangle$ is parametrized by a set of

138: complex variables $Z\equiv\{{\mathbf{Z}}_i\}_{i=1,\ldots,A}$, where

139: $A$ is the number of nucleons in the system.  The width parameter,

140: $\nu=0.16$ $\textrm{fm}^{-2}$, is treated as a constant parameter

141: common to all the wave packets.  If we ignore the antisymmetrization

142: effect, the real part of $\mathbf{Z}_i$ corresponds to the position

143: centroid and the imaginary part corresponds to the momentum centroid.

144: This choice of channel wave functions is suitable for fragmentation

145: reactions, where each single particle wave function should be

146: localized within a fragment.

147: 

148: Instead of directly considering the probability $w(Z,t)$ in Eq.

149: (\ref{eq:AMDensemble}), we solve a stochastic equation of motion for

150: the wave packet centroids $Z$, which may be symbolically written as

151: \begin{equation}

152: \frac{d}{dt}\mathbf{Z}_i

153: =\{\mathbf{Z}_i,\mathcal{H}\}_\text{PB}

154: +\mbox{(NN coll)}

155: +\Delta\mathbf{Z}_i(t)

156: +\mu\,(\mathbf{Z}_i,\mathcal{H}').

157: \end{equation}

158: The first term $\{\mathbf{Z}_i,\mathcal{H}\}_\text{PB}$ is the

159: deterministic term derived from the time-dependent variational

160: principle with an assumed effective interaction.  The Gogny

161: interaction \cite{GOGNY} is used in the present work.  The second term

162: represents the effect of the stochastic two-nucleon collision process,

163: where a parametrization of the energy-dependent in-medium cross

164: section is adopted. The two-nucleon collision cross-section used

165: is the same as in Ref. \cite{ONOj} namely,

166: \begin{equation}

167: \sigma(E,\rho)=\min\biggl( \sigma_{\text{LM}}(E,\rho),\ \frac{100\

168: \text{mb}}{1+E/(200\ \text{MeV})}\biggr),

169: \end{equation}

170: The collisions are performed with the ``physical

171: nucleon coordinates'' that take account of the antisymmetrization

172: effects, and then the Pauli blocking in the final state is

173: automatically introduced \cite{ONOa,ONOb}.  The third term

174: $\Delta\mathbf{Z}_i(t)$ is a stochastic fluctuation term that has been

175: introduced in order to respect the change of the width and shape of

176: the single particle distribution \cite{ONOh,ONOi,ONOj}.  In other

177: words, the combination

178: $\{\mathbf{Z}_i,\mathcal{H}\}_\text{PB}+\Delta\mathbf{Z}_i(t)$

179: approximately reproduces the prediction by mean field theories (for a

180: short time period) for the ensemble-averaged single-particle

181: distribution, while each nucleon is localized in phase space for each

182: channel.  The term $\Delta\mathbf{Z}_i(t)$ is calculated practically

183: by solving the Vlasov equation (for a short time period) with the same

184: effective interaction as for the term

185: $\{\mathbf{Z}_i,\mathcal{H}\}_\text{PB}$.  In the present version of

186: AMD \cite{ONOj}, the property of the fluctuation

187: $\Delta\mathbf{Z}_i(t)$ is chosen in such a way that the coherent

188: single particle motion in the mean field is respected for some time

189: interval until the nucleon collides another nucleon.  The last term

190: $\mu\,(\mathbf{Z}_i,\mathcal{H}')$ is a dissipation term related to

191: the fluctuation term $\Delta\mathbf{Z}_i(t)$.  The dissipation term is

192: necessary in order to restore the conservation of energy that is

193: violated by the fluctuation term.  The coefficient $\mu$ is given by

194: the condition of energy conservation. However, the form of this term

195: is somehow arbitrary.  We shift the variables $Z$ to the direction of

196: the gradient of the energy expectation value $\mathcal{H}$ under the

197: constraints of conserved quantities (the center-of-mass variables and

198: the total angular momentum) and global one-body quantities (monopole

199: and quadrupole moments in coordinate and momentum spaces).  A complete

200: formulation of AMD can be found in Refs. \cite{ONOj,ONO-ppnp}.

201: 

202: 

203: The statistical decay of relatively small primary fragments ($Z<20$)

204: is calculated by using the code \cite{MARUb} based on the sequential

205: binary decay model by P\"uhlhofer \cite{PUHLHOFER}.  The code employed

206: in the present work also takes account of the emission of composite

207: particles not only in their ground states but also in their excited

208: states with the excitation energy $E^*\le 40$ MeV.  The experimental

209: information is incorporated for known levels of $A\lesssim28$ nuclei,

210: while the Fermi-gas level density is assumed otherwise.  For the

211: statistical decay of large primary fragments ($Z\ge20$), the decay

212: code GEMINI \cite{Charity01} is employed. 

213: In considering the decay of the fragments, both the excitation energy 

214: and decay probabilities are calculated for spherical fragments independent 

215: of the true shape of the fragments induced by the reaction dynamics.

216: The effect of n-p asymmetry, excitation energy, and deformation on the 

217: nuclear level density are not considered in the decay. 

218: Introduction of a deformation dependence of the nuclear level density, 

219: and in particular the treatment of the continuum, results in a significant 

220: modification of the emission rate for fragments that are weakly bound

221: or at high excitation \cite{Charity05}.

222: 

223: The system we have chosen to study is $^{114}$Cd + $^{92}$Mo at 

224: E/A = 50 MeV which can be considered representative of symmetric 

225: heavy-ion collisions in this energy domain. 

226: We sampled all impact parameters, b, in the

227: interval 0$\le$b$\le$b$_{max}$ with a triangular distribution. The maximum

228: impact parameter b$_{max}$ had a value of 12 fm. The 

229: touching sphere configuration distance, given by 

230: R=1.2*(A$_P$$^{1/3}$+A$_T$$^{1/3}$), is equal to 11.2 fm.

231: The projectile and target were therefore placed at an initial distance of 13 fm

232: for b$\ge$6.5 fm and 9.8 fm for b$<$6.5 fm.

233: For a given collision, the fate of the colliding system was followed

234: until 300 fm/c. At regular intervals, the positions and momenta of all 

235: nucleons in the system were recorded. At a selected time 

236: (typically 300 fm/c), which we designate the cluster recognition time, 

237: the nucleon distributions are subjected to a 

238: cluster recognition algorithm based on the distance between nucleons. 

239: The nucleons and clusters that result from 

240: cluster recognition are 

241: subsequently propagated along Coulomb trajectories and allowed to 

242: statistically decay. The identity and momenta of the

243: final reaction products are recorded for subsequent analysis. In order to 

244: examine the predictions of this model in a statistically 

245: significant manner, we have amassed $\approx$25,000 collisions. 

246: The calculations were performed on a 646 CPU parallel processor system 

247: of which each CPU was either a PowerPC or Power3+. A single collision 

248: for this reaction

249: required 12 to 24 CPU-hours on a node depending on the impact parameter. 

250: 

251: 

252: \section{General Reaction Characteristics}

253: 

254: 

255: \begin{figure}

256: \vspace*{7.0in} 

257: \special{psfile=Density.eps hscale=68.0 vscale=68.0 hoffset=0 

258: voffset=0 angle=0}

259: \caption[]

260: {Contour diagram depicting the nucleon density distribution in 

261: spatial coordinates

262: as a function of time for the reaction $^{114}$Cd + $^{92}$Mo at 

263: E/A = 50 MeV. The positive z direction corresponds to the direction of 

264: the projectile. The columns correspond to different 

265: impact parameters, b=7.79 fm (left) and b=5.15 fm (right).} 

266: \label{fig:density}

267: \end{figure}

268: 

269: 

270: Depicted in Fig.~\ref{fig:density} is the density distribution 

271: of nucleons in R-space as a 

272: function of time for a mid-peripheral (b=7.79 fm) and mid-central (b=5.15) 

273: collision. The initial moment in time (t=0) is taken as the near touching 

274: configuration of the projectile-target system previously described, 

275: with the projectile approaching the target nucleus 

276: from the negative z direction.

277: As the di-nuclear system rotates, the initial dumb-bell 

278: shape of the two touching nuclei shown in the top panel evolves. 

279: While in contact, the two nuclei exchange mass, charge, and 

280: energy, governed by nucleon-nucleon scattering within the mean field. 

281: 

282: For the presented event with b=7.79 fm,

283: one observes that 

284: two large nuclei emerge from the collision at t=90 fm/c 

285: revealing the intrinsically binary nature of the collision. 

286: In this case, at longer times the elongation 

287: of the target-like fragment (left) leads to its breakup into 

288: multiple intermediate size nuclei.

289: In the case of the mid-central collision with b=5.15 fm however, 

290: the situation is more difficult to discern. At 

291: t=90 fm/c, it is unclear whether the system is disassembling into two or three 

292: large pieces.

293: What is apparent is that as the two nuclei separate

294: from each other, one observes that the density distributions 

295: reflect the nuclear

296: interaction between the projectile and target nuclei through the formation

297: of transiently deformed nuclei. These non-spherical geometries 

298: persist up to 300 fm/c for different cluster sizes. 

299: Moreover, for both events presented clusters seem to emerge on a relatively 

300: fast timescale,

301: t$\approx$90 fm/c. This early production of clusters 

302: indicates that the timescale of the 

303: shape/density fluctuations 

304: responsible for cluster formation operate on this timescale. 

305: It should be noted that a considerable fraction, though not all, 

306: of this early stage cluster emission

307: is located between the two large fragments that emerge from the collision.

308: The evolution

309: of the density distributions presented in Fig.~\ref{fig:density} 

310: can also be viewed from the context of semi-classical colliding liquid drops.

311: Formation of the transiently extended nuclear system by the collision

312: dynamics involves the generation of a considerable amount of 

313: ``surface'' nuclear material as compared to ``bulk'' nuclear material.

314: In comparison to the original system comprised of the projectile and target

315: nuclei, the multi-fragment final state with multiple

316: clusters requires the formation of a significant amount of

317: surface -- an energetically unfavorable change. 

318: Thus, once the surface-to-volume ratio has been 

319: increased by the collision dynamics, the energy cost of the 

320: system re-organizing

321: to the multi-fragment final state is considerably reduced. 

322: 

323: We examine the  characteristics of the system immediately following this 

324: dynamical stage of the collision. The products of the reaction 

325: at this stage are designated the ``primary'' products

326: which statistically de-excite to form the final reaction products which we also

327: refer to as the ``secondary'' products. 

328: For a large ensemble of events we examine the

329: evolution of both primary and secondary 

330: distributions with impact parameter, velocity dissipation,

331: and cluster recognition time.

332: 

333: \begin{figure}[ht]

334: \vspace*{3.5in} 

335: \special{psfile=ZV_Hot.eps hscale=43.0 vscale=43.0 hoffset=0 voffset=0 angle=0}

336: \caption[]

337: {(Color online)

338: Two dimensional diagram of the correlation between the atomic number 

339: and parallel velocity of

340: particles at t=300 fm/c for different impact parameters. The arrows 

341: correspond to the projectile and target velocities. The color scale 

342: indicates the yield on a logarithmic scale.} 

343: \label{fig:Zvel}

344: \end{figure}

345: 

346: 

347: An overview of the collisions studied is presented in Fig.~\ref{fig:Zvel}, 

348: where the correlation between the atomic number  

349: and parallel velocity (in the center-of-mass frame) 

350: of particles at t=300 fm/c is examined.

351: For the most peripheral collisions (8.5$<$b$\le$10 fm) two peaks located at 

352: Z$\approx$47 and $\approx$39 are clearly evident. 

353: These peaks correspond to the excited projectile-like (PLF$^*$) 

354: and target-like (TLF$^*$) nuclei respectively 

355: and are relatively narrow distributions in velocity 

356: centered at V$_\parallel$=3.8 and -4.7 cm/ns. Also 

357: evident is copious production of neutrons (Z=0), hydrogen, 

358: and helium nuclei. Smaller in yield, are clusters

359: with Z$\ge$3 and atomic number less than that of the PLF$^*$ and TLF$^*$. 

360: This pattern, dominated by the survival of the PLF$^*$ and TLF$^*$ for a

361: peripheral collision, reflects a primarily binary nature. For mid-peripheral 

362: and mid-central collisions, a similar pattern is observed 

363: indicating that in this impact parameter range as well

364: a PLF$^*$ and TLF$^*$ survive the dynamical phase, hence these 

365: impact parameters are also essentially binary in character. 

366: With increasing centrality 

367: $\langle$V$_{PLF^*}$$\rangle$  decreases and 

368: $\langle$V$_{TLF^*}$$\rangle$ increases reflecting 

369: an increase in the velocity damping. At the same time, the 

370: width of the PLF$^*$ and TLF$^*$ velocity distributions increases

371: indicating the growth of fluctuations.

372: In addition, with 

373: increasing centrality the average atomic number of the PLF$^*$ and 

374: TLF$^*$ decreases while the yield of clusters with 3$\le$Z$\le$15 increases. 

375: For simplicity, 

376: we designate the highest Z 

377: cluster with a velocity larger (smaller) than the center-of-mass 

378: velocity as the PLF$^*$ (TLF$^*$).

379: For b$\le$4 fm the decrease in the average Z of the PLF$^*$ combined with the

380: width of the distribution, lead to an operational definition of 

381: intermediate mass fragment, namely IMF: 3$\le$Z$\le$10.

382: Particles with Z$\le$10, manifest broad velocity distributions 

383: for the most central collisions. 

384: Examination of the most peripheral collisions reveals a

385: clear pattern of how the velocity distribution evolves 

386: with the atomic number (Z) of the fragment.

387: Neutrons and hydrogen nuclei in particular  have velocity distributions 

388: that are centered on velocities between those of the PLF$^*$ and

389: TLF$^*$. In contrast, for nuclei with 3$\le$Z$\le$15

390: the velocity distribution while broad, 

391: clearly has a bimodal nature with each of the two peaks centered close to the

392: PLF$^*$ and TLF$^*$ velocities. This bimodal character is also observed for 

393: helium nuclei although the distributions are broader. 

394: These overall patterns manifested for the most 

395: peripheral collisions are also observed for more central collisions. 

396: 

397: \begin{figure}

398: \vspace*{3.5in} 

399: \special{psfile=Z_HotandCold.eps hscale=43.0 vscale=43.0 

400: hoffset=0 voffset=0 angle=0}

401: \caption[]

402: {Dependence of the primary (solid) and secondary (dashed) Z distributions 

403: on impact parameter. The differential yield dN/dZ has been normalized by the 

404: total number of events for each impact parameter interval.} 

405: \label{fig:Z_dis}

406: \end{figure}

407: 

408: 

409: Depicted in Fig.~\ref{fig:Z_dis} is the dependence of the primary and 

410: secondary Z distributions on impact parameter. The Z distribution of particles

411: at t=300 fm/c is the primary distribution and is represented 

412: as the solid histogram. 

413: Following Coulomb propagation and statistical decay of the 

414: excited primary reaction products, the Z distribution of secondary particles 

415: is represented by the dashed histogram. 

416: The latter distribution includes both primary fragments that did not 

417: decay, as well as the decay products of excited primary fragments.

418: All distributions have been normalized to the total number of events 

419: for each impact parameter range and therefore 

420: represent the average multiplicities. 

421: As may be expected from the trends in Fig.~\ref{fig:Zvel}, 

422: the charge distribution for the most peripheral collisions, b$>$10 fm,

423: is largely dominated by two peaks at Z=42 and 48 which correspond to the

424: TLF$^*$ and PLF$^*$. In Fig.~\ref{fig:Z_dis} we therefore focus on smaller 

425: impact parameters, b$\le$10 fm. In panel a) the Z distributions 

426: integrated over impact parameter up to 10 fm are presented.

427: As expected, the yield for neutrons, hydrogen, and helium is large 

428: in the primary distribution (solid histogram). 

429: A large yield is also observed for 3$\le$Z$\le$10.

430: Evident for Z$\ge$30 is a slight double peak in the primary distribution 

431: attributable to the presence of the PLF$^*$ and TLF$^*$. 

432: This double peak structure is eliminated by secondary decay

433: as it is not evident in the dashed histogram. 

434: To separate the PLF$^*$ from the  TLF$^*$, 

435: as well as to crudely separate their decay products, 

436: we further select particles with the condition V$_\parallel$$>$0.

437: The resulting primary distribution shown in panel b) 

438: manifests only a single peak at large Z, which is located at Z=47. 

439: As observed in panel a) the yield of the Z distribution 

440: for 3$\le$Z$\le$30 is similar for both the primary and secondary particles.

441: 

442: We examine the dependence of the

443: Z distribution on impact parameter for V$_\parallel$$>$0 

444: in  Fig.~\ref{fig:Z_dis}c-f.

445: For 8.5$<$b$\le$10 fm, Fig.~\ref{fig:Z_dis}c), the primary Z 

446: distribution is 'V-shaped', reminiscent of the 'U-shape' observed for 

447: asymmetric fission. The minimum yield observed near Z$\approx$20 is 

448: deep in comparison to the yield at lower and higher Z indicating that 

449: asymmetric splits are strongly preferred over symmetric splits. 

450: It is striking that the multiplicity for Z=3-6 is 

451: approximately the same as that of Z$\approx$47 (the PLF$^*$). 

452: The yield ratio for Z=3-6 over Z=45-47 is 0.31/0.37$\approx$0.84, 

453: indicating a process or processes resulting in copious production of light IMFs. 

454: This similarity in the yield of the light IMF and the PLF$^*$ 

455: can, for example, be understood as the asymmetric binary decay 

456: of a precursor PLF$^*$.

457: Such a perspective is supported by experimental observation. 

458: For peripheral collisions of two heavy-ions 

459: at intermediate energies, the phenomenon of dynamical fission is well 

460: characterized \cite{Bocage00,Davin02,Colin03}. This dynamical fission has been 

461: associated with the deformation of the PLF$^*$ induced by the 

462: collision process. The defining characteristics of this process are the

463: aligned asymmetric binary decay of the PLF$^*$ and large relative velocities 

464: between the two produced fragments. 

465: On general grounds one expects that this dynamical process should 

466: depend sensitively on both the induced deformation and the excitation 

467: of the PLF$^*$ \cite{Piantelli02}.  

468: It is important to observe that the shape of this primary distribution 

469: largely survives the process of secondary decay. 

470: The main difference between the primary and secondary distributions 

471: is that the high Z peak is shifted to lower Z and increases in width. 

472: For 7$<$b$\le$8.5 fm, Fig.~\ref{fig:Z_dis}d), 

473: the shape of the primary distribution is better described as 

474: a 'U-shape'. In contrast to 

475: the previous impact parameter interval, the minimum located 

476: at Z$\approx$20 is shallow. This decrease in the depth of the minimum 

477: can be associated with the increase in 

478: the probability of

479: symmetric binary splits relative 

480: to asymmetric binary splits. This change of the Z distribution with decreasing

481: impact parameter can be related to an increase in the 

482: excitation energy of the PLF$^*$. In this impact parameter interval, 

483: the yield for 

484: Z=3-6 is significantly larger than that for Z$\approx$42. The ratio of 

485: the yield of Z=3-6 over the yield of Z=41-43 is 0.66/0.19$\approx$3.47, 

486: a change by a factor of $\approx$4 as the impact parameter

487: decreases from 8.5$<$b$\le$10 fm 

488: to 7$<$b$\le$8.5 fm. This increase in the ratio is due to both an 

489: increase in the IMF yield 

490: by a factor of 2 and a decrease in the yield in the vicinity of the 

491: PLF$^*$ peak. 

492: The latter decrease reflects the increasing width of the peak in the Z 

493: distribution attributable to the PLF$^*$ with decreasing impact 

494: parameter. 

495: Following secondary decay the 'U-shape' is somewhat less pronounced. 

496: For yet more central collisions, a 'U-shape' distribution is not observed even 

497: for the primary distribution. In panel e) no clear bump is observed at large 

498: Z, indicating the decreased likelihood that a high Z PLF$^*$ survives to the

499: cluster recognition time of t=300 fm/c. 

500: For the most central collisions shown, b$\le$ 4 fm, 

501: the primary Z distribution is exponential over a large range in Z. 

502: This exponential behavior of the yield is suppressed for Z$\ge$30 

503: due to the finite size (atomic number) of the system. 

504: The secondary Z distribution for central collisions also 

505: exhibits an exponential character for Z$>$3, 

506: although the onset of the finite size effects is observed at Z=20.

507: The main effect of secondary decay on the Z distribution, 

508: for all impact parameters 

509: is to significantly

510: enhance the yield of neutrons, hydrogen, and helium nuclei, while decreasing 

511: the maximum Z observed, namely the atomic number of the PLF$^*$ and TLF$^*$.

512: 

513: \begin{figure}

514: \vspace*{4.0in} 

515: \special{psfile=Z_Big_Big2.eps hscale=43.0 vscale=43.0 

516: hoffset=0 voffset=0 angle=0}

517: \caption[]

518: {(Color online)

519: Correlation between the atomic number of the fragments with the 

520: largest and second largest atomic number that have velocities larger 

521: than the center-of-mass velocity.

522: The solid circles represent the average charge of the second biggest fragment 

523: for a given charge of the biggest fragment. 

524: The open circles represent the average charge of the biggest fragment 

525: for a given charge of the second biggest fragment.} 

526: \label{fig:charge_correl}

527: \end{figure}

528: 

529: 

530: We have investigated whether the similarity of the yield for Z=3-6 

531: and Z$_{PLF^*}$ 

532: for peripheral collisions is an indication that the two largest fragments

533: forward of the center-of-mass originate from a common parent. 

534: Displayed in the

535: two-dimensional diagrams of Fig.~\ref{fig:charge_correl} 

536: is the joint probability of observing the 

537: largest and second largest fragments 

538: both with V$_\parallel$$>$0. 

539: For reference the dashed lines correspond to 

540: Z$_{TOT}$ = Z$_{PLF^*}$ + Z$_{2^{nd} biggest}$ = 45, 40, 35, 30 and 25.

541: The distribution at t=300 fm/c is presented in the left column while

542: the distribution following secondary decay is shown in the right column.

543: In the case of primary fragments (left column), for b$>$4 fm, a clear 

544: anti-correlation is observed between the atomic number of 

545: the largest and second largest fragment. 

546: In order to examine the average behavior of the two dimensional distribution, 

547: we also indicate as solid and open circles the 

548: $\langle$Z$_{2^{nd} biggest}$$\rangle$ for a given Z$_{PLF^*}$ and 

549: $\langle$Z$_{PLF^*}$$\rangle$ for a given Z$_{2^{nd} biggest}$. 

550: Strong correlation of $\langle$Z$_{PLF^*}$$\rangle$ and 

551: $\langle$Z$_{2^{nd} biggest}$$\rangle$ is evidenced by the near 

552: overlap of the open and closed circles over an extended range. 

553: Divergence of the symbols indicates that either the two fragments 

554: do not originate from a common parent or that finite size effects 

555: strongly influence

556: the observed correlation.

557: For 8.5$<$b$\le$10 fm, Z$_{TOT}$ is almost constant over the range 

558: of Z$_{PLF^*}$ with a value of $\approx$ 45. 

559: This value corresponds to an average loss of three charges from the 

560: incident Cd nucleus with the observed

561: anti-correlation signaling a conservation of charge between the

562: largest and second largest fragment. This anti-correlation signals that 

563: both fragments do on average originate 

564: from a common parent fragment. For a second largest fragment with Z=6, the

565: $\langle$Z$_{PLF^*}$$\rangle$ is $\approx$40 consistent with dynamical 

566: breakup \cite{Montoya94,Davin02}.

567: While mid-peripheral (mid-central) collisions exhibit an anti-correlation,

568: Z$_{TOT}$ changes from $\approx$45 (45) at high Z$_{PLF^*}$ to

569: $\approx$40 (35) at low Z$_{PLF^*}$. This 

570: change in Z$_{TOT}$

571: might indicate that the PLF$^*$ 

572: splits into three or more pieces or simply reflect the changing size of the 

573: parent fragment over the finite impact parameter interval considered.

574: For b$\le$4 fm, the average atomic number of the largest 

575: and second largest fragment are closer, 

576: $\langle$Z$\rangle$=20 and 9 respectively, 

577: as one might expect for a binary decay at high excitation, consistent

578: with the increased probability of symmetric splits indicated by the Z 

579: distribution in Fig.~\ref{fig:Z_dis}. The 

580: general trends observed for the primary fragments are also evident 

581: following decay in the

582: charge correlation of secondary fragments (right column). 

583: Similar charge correlation patterns have been experimentally 

584: observed \cite{Davin02} indicating a transition from asymmetric

585: splits toward those in which all asymmetries are populated.

586: The total charge of the two fragments after secondary decay is typically 

587: reduced by 

588: 5 to 10 charges as compared to the total charge of the primary fragments.

589: Following secondary decay, the emission of Z=4-6 results in 

590: a marked horizontal line in the charge correlation. 

591: This feature in the charge correlation has also been experimentally observed 

592: and has been previously attributed to dynamical fission \cite{Davin02}.

593: 

594: 

595: \section{Velocity Dissipation of the PLF$^*$ and its Excitation}

596: 

597: \begin{figure}

598: \vspace*{3.5in} 

599: \special{psfile=XLF_b.eps hscale=43.0 vscale=43.0 hoffset=0. voffset=0 angle=0}

600: \caption[]

601: {(Color online)

602: Dependence of the $\langle$Z$\rangle$,$\langle$V$\rangle$, and E$^*$/A of the 

603: PLF$^*$ and TLF$^*$ on  impact parameter. 

604: Error bars indicate the standard 

605: deviation of the distribution. In the bottom panels the solid symbols denote

606: the $\langle$(E$^*$/A)$_{XLF^*}$$\rangle$ as a function of b.} 

607: \label{fig:Z_V_Estar}

608: \end{figure}

609: 

610: 

611: A more quantitative picture of the evolution of the general properties of the 

612: PLF$^*$ and TLF$^*$ with impact parameter 

613: is displayed in Fig.~\ref{fig:Z_V_Estar}. For the most 

614: peripheral collisions studied, b$>$10 fm, 

615: the $\langle$Z$_{PLF^*}$$\rangle$ is $\approx$48, the atomic number of the 

616: projectile. The $\langle$Z$_{PLF^*}$$\rangle$ 

617: decreases smoothly with decreasing impact parameter until 

618: b$\approx$3-4 fm.  For smaller 

619: impact parameters, $\langle$Z$_{PLF^*}$$\rangle$ shows no dependence 

620: on impact parameter and has a value of $\approx$19. 

621: For b$<$10 fm, $\langle$Z$_{PLF}$$\rangle$, 

622: namely the average atomic number following decay, 

623: is approximately 4-9 units less than $\langle$Z$_{PLF^*}$$\rangle$ and 

624: exhibits the same impact parameter dependence as $\langle$Z$_{PLF^*}$$\rangle$.

625: It should be noted that the largest difference 

626: between $\langle$Z$_{PLF^*}$$\rangle$ and 

627: $\langle$Z$_{PLF}$$\rangle$ is observed for mid-peripheral collisions 

628: with an impact parameter $\approx$8 fm. 

629: The average center-of-mass velocity of the PLF$^*$,

630: $\langle$V$_{PLF^*}$$\rangle$, also exhibits a smooth 

631: dependence on impact parameter, decreasing monotonically 

632: from $\langle$V$_{PLF^*}$$\rangle$ $\approx$ 4.3 cm/ns 

633: for the most peripheral collisions to $\approx$2.5 cm/ns for b=3 fm. 

634: For more central 

635: collisions $\langle$V$_{PLF^*}$$\rangle$ only shows a weak 

636: dependence on impact parameter. 

637: With increasing centrality the width of the velocity damping distribution 

638: (indicated by the error bars) increases significantly, 

639: indicating the growth of fluctuations.

640: 

641: The predicted velocity damping 

642: of the PLF$^*$ evident in the middle panel is associated 

643: with a corresponding increase in the excitation of the PLF$^*$ 

644: as shown in the bottom panel of Fig.~\ref{fig:Z_V_Estar}. 

645: Such an association between velocity damping and excitation 

646: has been experimentally observed \cite{Yanez03}.

647: While the average E$^*$/A of the PLF$^*$ rapidly increases 

648: with impact parameter for peripheral collisions, it saturates 

649: at $\approx$4 MeV by b=6 fm. 

650: The trends observed for the PLF$^*$ are also observed for the TLF$^*$ as 

651: depicted in the right column of Fig.~\ref{fig:Z_V_Estar}.

652: It is interesting to note that the $\langle$E$^*$/A$\rangle$ 

653: for small impact parameters attained for both the 

654: PLF$^*$ and TLF$^*$ is the same despite the smaller size

655: of the TLF$^*$ (Z$\approx$15) as compared to the PLF$^*$ (Z$\approx$19). 

656: This difference of $\approx$20-25\% in Z corresponds to a similar

657: difference in A (see Fig.~\ref{fig:NZ}). Equal partition of E$^*$ would thus 

658: result in a larger $\langle$E$^*$/A$\rangle$ for the TLF$^*$ as compared to the

659: PLF$^*$. An $\langle$E$^*$/A$\rangle$=4 MeV for the PLF$^*$ 

660: would correspond to an $\langle$E$^*$/A$\rangle$=5 MeV for the TLF$^*$.

661: The similarity of $\langle$E$^*$/A$\rangle$ for both the PLF$^*$ and TLF$^*$ is

662: indicative that the degree to which thermalization is achieved is large. 

663: For the most peripheral collisions, b$\approx$12 fm, 

664: the non-zero value of the

665: $\langle$(E$^*$/A)$_{TLF^*}$$\rangle$ and 

666: $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ is due in part to 

667: the mismatch between the binding energy of the projectile and target in AMD

668: and their real binding energies. This error typically ranges from 0.2 to 0.3 MeV. 

669: Additional excitation may occur due to the mean field or Coulomb interaction.

670: 

671: 

672: \begin{figure}

673: \vspace*{4.0in} 

674: \special{psfile=NZ_Comp_b.eps hscale=43.0 vscale=43.0 hoffset=0. voffset=0 angle=0}

675: \caption[]

676: {Top panel: Dependence of the $\langle$N/Z$\rangle$ of 

677: the PLF$^*$ and TLF$^*$ on impact parameter.Dotted lines indicate the 

678: initial N/Z of projectile and target nuclei.

679: Bottom panel: Fraction, f, of nucleons found in the PLF$^*$ (solid symbols) 

680: or TLF$^*$ (open symbols) that originate from the projectile.} 

681: \label{fig:NZ}

682: \end{figure}

683: 

684: The composition of the excited PLF$^*$ and TLF$^*$ that subsequently undergoes

685: decay

686: is indicated in Fig.~\ref{fig:NZ}. In this figure the $\langle$N/Z$\rangle$ 

687: of both the PLF$^*$ and

688: TLF$^*$ are examined as a function of impact parameter. For b$>$6 fm, the 

689: $\langle$N/Z$\rangle$ 

690: of both PLF$^*$ and TLF$^*$ evolves essentially linearly with impact parameter

691: from the initial values of 1.375 and 1.19 

692: for the projectile and target respectively.

693: Over this range of impact parameter, this behavior could be interpreted

694: as equilibration of N/Z. 

695: However, the change in $\langle$N/Z$\rangle$ is larger for the

696: PLF$^*$ as compared to the TLF$^*$ by a factor of two. This difference reflects

697: the fact that exchange between the PLF$^*$ and TLF$^*$ is not the only process

698: occurring thus complicating the interpretation of the

699: change in N/Z in terms of equilibration.

700: For more central 

701: collisions, the $\langle$N/Z$\rangle$ remains essentially constant having 

702: saturated at a value of $\approx$1.24-1.26. The similarity of the average N/Z

703: value for the PLF$^*$ and TLF$^*$ could be interpreted as equilibration of this 

704: degree-of-freedom. If this is indeed the case, 

705: it is interesting to note that for b$\approx$4 fm, this

706: equilibration is already achieved.

707: For comparison the 

708: $\langle$N/Z$\rangle$ of the system is $\approx$1.29. 

709: The slightly lower N/Z asymptotic value for central collisions 

710: as compared to the N/Z of the system suggests either 

711: a preferential emission of free neutrons or the production of neutron-rich 

712: fragments in the dynamical stage.

713: 

714: 

715: We examine the degree to which mixing occurs in the lower panel of 

716: Fig.~\ref{fig:NZ}. In this figure the dependence of f, the fraction of nucleons

717: in the PLF$^*$ or TLF$^*$ that originate from the projectile, on impact parameter

718: is presented. It is interesting to note that for b$\ge$6 fm, the region in

719: which $\langle$N/Z$\rangle$ changed linearly with b, the fraction of nucleons

720: in the PLF$^*$ that were originally in the projectile is large, f$\ge$0.9. 

721: Only for smaller impact parameters does the degree of mixing of projectile and 

722: target nucleons become larger. Thus, the 

723: large change in $\langle$N/Z$\rangle$ 

724: does not require large mixing of the projectile and target nucleons.

725: It is instructive to note that the quantity f, appears to saturate for b$\le$2 fm 

726: with a maximum of $\approx$35\% of the PLF$^*$ nucleons originating from the

727: target. For the TLF$^*$, in the case of small impact parameters, the degree of

728: mixing is similar.

729: It has been experimentally demonstrated that 

730: for mid-peripheral collisions 

731: the N/Z degree of freedom 

732: does not reach equilibrium 

733: \cite{Tsang04}. However, 

734: the present result indicate that N/Z equilibrium is attained for 

735: mid-central collisions, despite the incomplete mixing of the projectile and

736: target nucleons. This result is of significance to future work with radioactive

737: beams, indicating the degree to which the N/Z exotic projectile can be excited 

738: while only modestly pertubing its N/Z.

739: 

740: \begin{figure}

741: \vspace*{3.0in}

742: \special{psfile=EA_PLF_b_Times.eps hscale=43.0 vscale=43.0 hoffset=0 voffset=0 angle=0}

743: \caption[]

744: {Dependence of the $\langle$E$^*$/A$\rangle$ of the

745: PLF$^*$ on  impact parameter for different cluster recognition times.}

746: \label{fig:Estar_Time}

747: \end{figure}

748: 

749: To probe the origin of the saturation in E$^*$/A of the PLF$^*$ and TLF$^*$

750: for b$<$6 fm observed in Fig.~\ref{fig:Z_V_Estar}, 

751: we have investigated the influence of our choice of 

752: cluster recognition time on the excitation energy of the PLF$^*$. 

753: We have chosen to recognize the clusters 

754: at t= 150, 180, 210, 240, 270, and 300 fm/c

755: and compare the dependence of excitation energy on impact 

756: parameter for the different cluster recognition times. As evident 

757: in Fig.~\ref{fig:Estar_Time}, while for peripheral collisions 

758: the average excitation energy is fairly independent 

759: of the choice of cluster recognition time, with decreasing impact parameter 

760: the average excitation energy deduced depends significantly on the choice of 

761: cluster recognition time. 

762: For different cluster recognition times one also observes that 

763: the onset of the saturation in excitation energy 

764: occurs at different impact parameter. For t=300 fm/c the onset of the 

765: saturation occurs at b$\approx$6 fm ($\approx$25 \% of the cross-section) 

766: while for t=150 fm/c, 

767: the onset occurs at b$\approx$4 fm ($\approx$10 \% of the cross-section).

768: The events associated with the highest excitation attainable

769: therefore correspond to a 

770: significant fraction of the

771: cross-section. 

772: For central collisions, the excitation 

773: attained is higher the earlier one recognizes the clusters. For an early

774: cluster recognition time, t=150 fm/c, a maximum value 

775: of $\langle$E$^*$/A$\rangle$$\approx$6 MeV is attained in comparison

776: to $\langle$E$^*$/A$\rangle$$\approx$4 MeV for t=300 fm/c.

777: Both the trend and magnitude of  $\langle$E$^*$/A$\rangle$

778: is consistent with the AMD calculations for a more asymmetric system

779: \cite{Wada04}.

780: This rapid decrease in $\langle$E$^*$/A$\rangle$ is indicative of 

781: rapid cooling of 

782: the PLF$^*$.   

783: As one may imagine, the choice of a cluster recognition

784: time less than 150 fm/c becomes increasingly problematic due to  

785: both the conceptual, as well as practical, problem of  

786: distinguishing clusters during the 

787: high density phase of the collision.

788: 

789: 

790: \begin{figure*}[ht]

791: \vspace*{6.2in}

792: \special{psfile=EA_PLF_vPLF.eps hscale=70.0 vscale=70.0 

793: hoffset=50 voffset=0 angle=0}

794: \caption[]

795: {(Color online)

796: Correlation between the E$^*$/A of the

797: PLF$^*$  and its velocity for different impact parameters and 

798: cluster recognition times. The symbols indicate the 

799: average E$^*$/A as a function of V$_{PLF^*}$. The arrows correspond to the 

800: beam velocity.}

801: \label{fig:Estar_b_clus}

802: \end{figure*}

803: 

804: 

805: The dependence of $\langle$V$_{PLF^*}$$\rangle$ and 

806: $\langle$E$^*$/A$\rangle$ on impact parameter 

807: suggests a direct correlation between these two quantities. 

808: The correlation between these two quantities 

809: as a function of both impact parameter and cluster recognition 

810: time is examined in Fig.~\ref{fig:Estar_b_clus}. 

811: For peripheral collisions, 8.5$<$b$\le$10 fm (leftmost column), at 

812: early cluster recognition times, e.g. t=150 fm/c (uppermost panel), 

813: a narrow anti-correlated 

814: distribution is observed, namely there is

815: a strong dependence of the PLF$^*$'s excitation, (E$^*$/A)$_{PLF^*}$, 

816: on its velocity, V$_{PLF^*}$. 

817: To more easily examine the correlation between the two quantities, 

818: the centroid in E$^*$/A for each bin 

819: in V$_{PLF^*}$ is indicated by the symbol. The significant slope

820: of $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ with respect 

821: to V$_{PLF^*}$ indicates the strong correlation between 

822: PLF$^*$ excitation and velocity damping.

823: With increasing cluster recognition time, 

824: the strong correlation between V$_{PLF^*}$ and (E$^*$/A)$_{PLF^*}$ 

825: persists although the width of the distribution increases.

826: 

827: 

828: For a fixed cluster recognition time, one observes that with 

829: decreasing impact parameter, the dependence of 

830: $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ (symbols) on velocity 

831: becomes flatter indicating a weakening dependence on average. 

832: The two dimensional distributions also become broader with increasing 

833: centrality indicating the growth of fluctuations that attenuate 

834: the intrinsic correlation between excitation energy and PLF$^*$ velocity.

835: Examination of the most central collisions studied (b$\le$4 fm) shows that 

836: while a modest dependence between E$^*$/A and V$_{PLF^*}$ exists 

837: at t=150 fm/c, for longer cluster recognition times effectively 

838: no dependence of the PLF$^*$ excitation energy on its velocity 

839: is observed. At t=300 fm/c,  $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$

840: does not exhibit any dependence on V$_{PLF^*}$. 

841: This attenuation of the correlation between

842: excitation energy and velocity of the PLF$^*$ with increasing

843:  cluster recognition time is also observed at 

844: intermediate impact parameters.

845: 

846: \begin{figure}

847: \vspace*{4.2in}

848: \special{psfile=EA_PLF_vPLF_Average.eps hscale=43.0 vscale=43.0 

849: hoffset=0 voffset=0 angle=0}

850: \caption[]

851: {Average E$^*$/A of the PLF$^*$  as a function of 

852: its velocity for different impact parameters and cluster recognition times.}

853: \label{fig:Estar_Damping_Average}

854: \end{figure}

855: 

856: 

857: To examine the influence of the cluster recognition time 

858: on the most peripheral collisions in a more quantitative manner, 

859: we compare in the top panel of 

860: Fig.~\ref{fig:Estar_Damping_Average} the dependence 

861: of the average 

862: excitation energy as a function of V$_{PLF^*}$ for 8.5$<$b$\le$10 fm 

863: for different cluster recognition times. 

864: For low to modest velocity damping, i.e. V$_{PLF^*}$$>$3.25 cm/ns, the 

865: anti-correlation between the average

866: excitation energy and V$_{PLF^*}$ 

867: is independent of the cluster recognition time. 

868: For more damped collisions, however,

869: one does observe a difference between the calculated average 

870: excitation energy for different cluster recognition times. 

871: For V$_{PLF^*}$=4 cm/ns, $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$=2.15 MeV 

872: while for V$_{PLF^*}$=3.5 cm/ns, 

873: $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$=3.3 MeV. 

874: This average excitation of 3.3 MeV is associated with a velocity damping 

875: from beam velocity of 0.86 cm/ns.

876: 

877: In the lower panel of Fig.~\ref{fig:Estar_Damping_Average} the dependence 

878: of the average excitation energy as a function 

879: of V$_{PLF^*}$ for more central collisions 4$<$b$\le$7 fm is shown. 

880: In contrast to the more peripheral collisions just discussed, 

881: for all values of velocity damping, the average excitation of the

882: PLF$^*$ depends on the cluster recognition time.

883: Even for the smallest velocity damping 

884: (V$_{PLF^*}$$>$ 3.5 cm/ns)  

885: a minimum excitation energy of $\approx$3.5 MeV is observed for 

886: all cluster recognition times. 

887: Cluster recognition times less than 210 fm/c manifest 

888: an essentially linear dependence of E$^*$/A on V$_{PLF^*}$ while longer 

889: cluster recognition times (t$\ge$240 fm/c) exhibit a significantly 

890: non-linear dependence. By t=300 fm/c,

891: $\langle$E$^*$/A$\rangle$ is essentially independent of V$_{PLF^*}$. 

892: 

893: 

894: \begin{figure}

895: \vspace*{4.0in}

896: \special{psfile=Average_PLF_Time.eps hscale=43.0 vscale=43.0 

897: hoffset=0 voffset=0 angle=0}

898: \caption[]

899: {Dependence of the $\langle$Z$\rangle$,$\langle$V$\rangle$, 

900: and $\langle$E$^*$/A$\rangle$ of the PLF$^*$ on cluster recognition time 

901: for different impact parameters. The dashed line represents the 

902: projectile atomic number (velocity) in the top (middle) panel.}

903: \label{fig:ave_prop}

904: \end{figure}

905: 

906: 

907: The dependence of some of the average properties of the PLF$^*$ on both impact

908: parameter and cluster recognition time are summarized 

909: in Fig.~\ref{fig:ave_prop}. In the top panel, 

910: the average atomic number of the PLF$^*$, 

911: $\langle$Z$_{PLF^*}$$\rangle$, is displayed as a function of 

912: cluster recognition time for different impact parameters. For the

913: most peripheral collisions, 8.5$<$b$\le$10 fm, 

914: and the shortest cluster recognition times,

915: $\langle$Z$_{PLF^*}$$\rangle$$\approx$47, just below Z$_{BEAM}$=48 

916: as indicated by the dashed line. Longer cluster recognition times 

917: result in a slight decrease in $\langle$Z$_{PLF^*}$$\rangle$ to

918: a value of $\approx$ 44 at t=300 fm/c. This reduction 

919: in $\langle$Z$_{PLF^*}$$\rangle$ corresponds

920: to the emission of charge on a short timescale. For more central 

921: collisions a similar behavior is observed although the magnitude of the 

922: charge emitted on a short timescale is larger.

923: 

924: 

925: In the middle panel of Fig.~\ref{fig:ave_prop}, the trend of 

926: $\langle$V$_{PLF^*}$$\rangle$ with cluster recognition time 

927: and impact parameter is presented. For 8.5$<$b$\le$10 fm, 

928: essentially no change is observed in $\langle$V$_{PLF^*}$$\rangle$ 

929: as the cluster recognition time changes from t=150 fm/c to 300 fm/c. 

930: For mid-central and central collisions, a small increase in 

931: $\langle$V$_{PLF^*}$$\rangle$ is discernible as the 

932: cluster recognition time increases. 

933: This slight increase is attributable to the Coulomb re-acceleration

934: of the PLF$^*$ following the collision combined 

935: with recoil effects due to predominantly backward emission of particles 

936: on a short timescale.

937: 

938: The dependence of $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ 

939: on cluster recognition time

940: is depicted in the 

941: bottom panel of Fig.~\ref{fig:ave_prop} 

942: for different impact. 

943: As previously noted in Fig.~\ref{fig:Estar_Time}, for 8.5$<$b$\le$10 fm the 

944: cluster recognition time has

945: only a weak influence on $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$. 

946: Longer cluster recognition

947: times lead to slightly lower $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$, 

948: 3.1 MeV for t=150 fm/c

949: as compared to 2.8 MeV at 300 fm/c. More central collisions, however, manifest

950: a more marked dependence. As apparent in Fig.~\ref{fig:Estar_Time}, 

951: for b$<$4 fm $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ reaches 

952: a value of 6 MeV for the shortest cluster recognition

953: times, while at longer cluster recognition times 

954: $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ is only 

955: $\approx$4 MeV. This decrease in excitation energy is rapid with most of the 

956: decrease occurring from t=150-240 fm/c. This rapid decrease in the excitation

957: energy of the PLF$^*$ is directly related to the emission of particles over

958: this time interval. As the excited PLF$^*$ rapidly emits charged particles 

959: between t=150-240 fm/c its atomic number decreases while 

960: its velocity remains relatively constant. 

961: Consequently, the correlation between

962: $\langle$(E$^*$/A)$_{PLF^*}$$\rangle$ and V$_{PLF^*}$ 

963: observed in Fig.~\ref{fig:Estar_b_clus} is poor 

964: for central collisions and long cluster recognition times.

965: Thus, a proper description of this de-excitation of the highly 

966: excited PLF$^*$ requires modeling the statistical decay of the 

967: deformed PLF$^*$.

968: 

969: \begin{figure*}

970: \vspace*{6.2in}

971: \special{psfile=EA_PLF_TLF_Time.eps hscale=70.0 vscale=70.0 

972: hoffset=50 voffset=0 angle=0}

973: \caption[]

974: {(Color online)

975: Correlation between E$^*$/A of the PLF$^*$ and TLF$^*$

976: for different impact parameters and cluster recognition times.

977: The symbols correspond to $\langle$(E$^*$/A)$_{TLF^*}$$\rangle$ 

978: for a given (E$^*$/A)$_{PLF^*}$.}

979: \label{fig:Estar_PLF_TLF}

980: \end{figure*}

981: 

982: 

983: An interesting consequence of this rapid emission from the PLF$^*$ 

984: (and TLF$^*$) is the amelioration of the correlation between the excitation

985: of the PLF$^*$ and TLF$^*$. Displayed in Fig.~\ref{fig:Estar_PLF_TLF} is the 

986: two-dimensional distribution of PLF$^*$ and TLF$^*$ excitation energies 

987: for different impact parameters and cluster recognition times. 

988: For all impact parameters shown, the distribution is broad 

989: with the centroid for each (E$^*$/A)$_{PLF^*}$ bin indicated 

990: by the symbol. For 8.5$<$b$\le$10 fm, a slight positive 

991: correlation between $\langle$(E$^*$/A)$_{TLF^*}$$\rangle$ and

992: (E$^*$/A)$_{PLF^*}$ is evident. 

993: Examination of the correlation between the total excitation, E$^*$, 

994: of the PLF$^*$ and TLF$^*$ reveals an independence indicating that 

995: the observed correlation between (E$^*$/A)$_{PLF^*}$ and 

996: $\langle$(E$^*$/A)$_{TLF^*}$$\rangle$ is principally due to a 

997: correlation between A$_{PLF^*}$ and A$_{TLF^*}$ for the most peripheral 

998: collisions.

999: In contrast, for the most central collisions, b$\le$4 fm, 

1000: an anti-correlation between $\langle$(E$^*$/A)$_{TLF^*}$$\rangle$ 

1001: and (E$^*$/A)$_{PLF^*}$ is evident. 

1002: Both E$^*$ and A of the PLF$^*$ and TLF$^*$ manifest the same anti-correlation.

1003: 

1004: 

1005: \section{Emitted Particles}

1006: 

1007: \begin{figure}

1008: \vspace*{4.0in} 

1009: \special{psfile=Nimf_b_cumul.eps hscale=43.0 vscale=43.0 

1010: hoffset=0 voffset=0 angle=0}

1011: \caption[]

1012: {(Color online)

1013: Multiplicity of fragments as a function of impact parameter. 

1014: Panel a: Average multiplicity of Z$\ge$ 3 at t=300 fm/c (solid circles), 

1015: after secondary decay (solid triangles) and for QMD (open crosses). 

1016: The QMD results are extracted from \cite{Nebauer99}. 

1017: Panel b: Multiplicity of Z=3-10. 

1018: Panel c: Average multiplicity of Z=3-10 for different 

1019: cluster recognition times.} 

1020: \label{fig:Nimf_b}

1021: \end{figure}

1022: 

1023: 

1024: As evident from Fig.~\ref{fig:density}, as the PLF$^*$ and TLF$^*$ separate, 

1025: clusters are produced. 

1026: This fragment production as already demonstrated 

1027: can occur on relatively short time scale impacting the Z, velocity, and

1028: (E$^*$/A) of the PLF$^*$ and TLF$^*$. In order to characterize this 

1029: fast emission process in more detail, 

1030: we examine the multiplicity of fragments 

1031: produced as a function of impact parameter in Fig.~\ref{fig:Nimf_b}. 

1032: Displayed in Fig.~\ref{fig:Nimf_b}a) is the average multiplicity 

1033: of fragments, Z$\ge$3, at t=300 fm/c (solid circles).

1034: One observes that this multiplicity increases 

1035: with decreasing impact parameter and saturates for b$\approx$3 fm. 

1036: For the most peripheral collisions the average multiplicity is 2, 

1037: corresponding to the existence of the only PLF$^*$ and TLF$^*$. 

1038: The average fragment multiplicity reaches a value of 3 at b$\approx$8-9 fm. 

1039: For this impact parameter interval, on average, 

1040: one fragment is produced in coincidence with the PLF$^*$ and TLF$^*$. 

1041: This result is consistent with the asymmetric split of the PLF$^*$ deduced 

1042: from the Z distribution (Figs.~\ref{fig:Z_dis} and \ref{fig:charge_correl}).

1043: For the most central collisions, b$<$3 fm, the average fragments multiplicity 

1044: is constant and is $\approx$8. 

1045: Following secondary decay (solid triangles) 

1046: the fragment multiplicity is reduced slightly 

1047: due to the decay of fragments into particles with Z$\le$2.

1048: For b$>$7 fm the effect of secondary decay on the fragment multiplicity 

1049: is negligible 

1050: while for the most central collisions 

1051: the average multiplicity decreases from 8 to 6.5. 

1052: The increased excitation energy associated with more central collisions is no

1053: doubt responsible for this

1054: increased importance of secondary decay.

1055: The multiplicities predicted in the present calculation are compared to 

1056: those from QMD calculations for the system Xe+Sn \cite{Nebauer99}. 

1057: Although both systems were simulated 

1058: for the same incident energy of 50 MeV/nucleon, the Xe+Sn system is

1059: $\approx$20\% larger in A and $\approx$15\% larger in Z than the present system. 

1060: The multiplicity deduced by QMD (open crosses) is larger that the ones 

1061: of the present work at all impact parameters. 

1062: Given the difference in the system size, the difference between the 

1063: multiplicities for b$\le$6 fm may be reasonable.

1064: The most notable feature of this comparison between the two models 

1065: is the behavior for peripheral collisions, b$\ge$8 fm. The fragment 

1066: multiplicities predicted by AMD appear to be more realistic than those predicted

1067: by QMD. This difference may be due to spurious decay of the

1068: projectile and target in QMD due to the poor description of the ground state 

1069: properties in that model.

1070: 

1071: The multiplicity distribution of IMFs (3$\le$Z$\le$10) as a function of 

1072: impact parameter is presented in Fig.~\ref{fig:Nimf_b}b) for t=300fm/c. 

1073: While the distribution is narrow for the most peripheral collisions, 

1074: its width rapidly increases with decreasing impact parameter. 

1075: The average IMF multiplicity, indicated by the solid circles, 

1076: evolves from 0 for the most peripheral to $\approx$6 for b=3 fm. 

1077: At an impact parameter of $\approx$9 fm, the average IMF multiplicity 

1078: reaches a value of $\approx$1, consistent with Fig.~\ref{fig:Nimf_b}a). 

1079: The average IMF multiplicity is pretty insensitive to secondary decay as 

1080: indicated by the triangles. 

1081: Comparison between the fragment multiplicity, Fig.~\ref{fig:Nimf_b}a), 

1082: and IMF multiplicity, Fig.~\ref{fig:Nimf_b}b), indicates that even for 

1083: the most central collisions two fragments with a Z$>$10 are present 

1084: at t=300fm/c representing a PLF$^*$ and TLF$^*$ with approximately 20-25 \% 

1085: of the original projectile and target atomic number. 

1086: This result contradicts the physical picture of a single 

1087: source often assumed for central collisions.

1088: 

1089: Displayed in Fig.~\ref{fig:Nimf_b}c) is the average IMF multiplicity 

1090: dependence on b for different cluster recognition times. 

1091: The average IMF multiplicity increases with 

1092: increasing cluster recognition time for all impact parameters . 

1093: The largest increases are evident for 

1094: the shortest times, t$\le$240 fm/c. 

1095: For all impact parameters, the IMF multiplicity increases by a factor 

1096: of 2 to 3 between t=150 fm/c and t=240 fm/c. After t=240 fm/c, 

1097: the IMF production rate is reduced with an increase of 20-40 \% of 

1098: the IMF multiplicity between t=240 fm/c and t=300 fm/c. 

1099: 

1100: 

1101: The average multiplicity of light charged particles is examined in 

1102: Fig.~\ref{fig:Nlp_b} as a function of both impact parameter and cluster 

1103: recognition time. In the left hand column of Fig.~\ref{fig:Nlp_b} one observes

1104: a monotonic increase of the neutron and proton average 

1105: multiplicities with decreasing

1106: impact parameter both at t=150 fm/c and t= 300 fm/c. At t=150 fm/c a slight 

1107: saturation in the both the neutron and proton multiplicities is observed for 

1108: the most central collision with maximum average multiplicities of 11.5 and 7

1109: attained. A later cluster recognition time of t=300 fm/c results in 

1110: approximately a 50\% increase 

1111: in the multiplicities with the saturation of the multiplicities 

1112: for central collisions being slightly more evident. For this longer

1113: cluster recognition time, the average multiplicities associated with central

1114: collisions are 19.5 and 12 for neutrons and protons respectively. Following 

1115: sequential decay (t=$\infty$), 

1116: one observes a significant increase in the average multiplicities and a 

1117: pronounced saturation in the case of the neutrons. This saturation 

1118: suggests that the total neutron multiplicity, in particular, 

1119: while providing impact parameter selectivity for peripheral collisions is 

1120: a poor selector of more central collisions. Moreover, attempting to 

1121: select central collisions with the neutron multiplicity would on the basis of 

1122: the cross-section be weighted towards mid-central collisions. This result 

1123: explains the experimental observation of the persistence of binary collisions

1124: associated with large neutron multiplicity \cite{Lott92}. Moreover, 

1125: these mid-central 

1126: collisions are associated with the highest average excitation energy attained

1127: as presented in Fig.~\ref{fig:Z_V_Estar}. 

1128: 

1129: In the case of the light cluster (d,t, and $\alpha$ particle) multiplicities 

1130: a couple of points are noteworthy. For short cluster recognition time 

1131: (t=150 fm/c) the average multiplicity of deuterons is relatively linear 

1132: over the entire impact parameter range and reaches a value of $\approx$1.4

1133: for the most central collisions. Alpha particles, in the case of peripheral

1134: collisions manifest similar multiplicities, however the average multiplicity 

1135: of $\alpha$ particles saturates for b$<$6 fm. Tritons exhibit lower 

1136: multiplicities than both deuterons and alpha particles for all impact 

1137: parameters. For longer cluster recognition time, t=300 fm/c, the deuteron

1138: and triton 

1139: multiplicities remain essentially unchanged as compared to t=150 fm/c. In 

1140: contrast, the $\alpha$ particle multiplicity increases significantly. It is 

1141: interesting to note that the maximum $\alpha$ multiplicity is not associated

1142: with central collisions but rather with b$\approx$7 fm. From this we conclude 

1143: that significant $\alpha$ production/emission, but not 

1144: deuteron or triton 

1145: emission occurs on the timescale 

1146: commensurate with the separation time of the PLF$^*$ and TLF$^*$ 

1147: (150 fm/c$\le$t$\le$300 fm/c). Following 

1148: secondary decay (t=$\infty$) all multiplicities increase significantly. 

1149: Moreover, only for peripheral collisions, b$\ge$8 fm, does the average 

1150: multiplicity of light clusters depend significantly on impact parameter.  

1151: 

1152: 

1153: \begin{figure}

1154: \vspace*{4.0in} 

1155: \special{psfile=Nlp_b_2.eps hscale=43.0 vscale=43.0 

1156: hoffset=0 voffset=0 angle=0}

1157: \caption[]

1158: {Multiplicity of light particles as a function of impact parameter for

1159: different cluster recognition times. 

1160: The multiplicities in the right column have been scaled by the factors 

1161: indicated.} 

1162: \label{fig:Nlp_b}

1163: \end{figure}

1164: 

1165:  

1166: \begin{figure*}

1167: \vspace*{4.2in}

1168: \special{psfile=Galilean_IMF.eps hscale=75.0 vscale=75.0 hoffset=30 

1169: voffset=0 angle=0}

1170: \caption[]

1171: {(Color online)

1172: Invariant cross-section for IMFs (3$\le$ Z$\le$10) in the COM frame.

1173: The arrows indicate the average parallel velocity of the PLF$^*$ and TLF$^*$.

1174: The color scale indicates the yield on a logarithmic scale. The vertical 

1175: scale in the bottom two

1176: panels has been scaled by a factor of two as compared to the other panels.}

1177: \label{fig:Gal_IMF}

1178: \end{figure*}

1179: 

1180: We examine the emission pattern for IMFs (3$\le$Z$\le$10) 

1181: in Fig.~\ref{fig:Gal_IMF} both at t=300 fm/c and 

1182: at t=$\infty$ as a function of impact parameter.

1183: In examining the most peripheral collisions for t=300 fm/c, we observe 

1184: two major components which are shifted with 

1185: respect to the velocity of the PLF$^*$ and TLF$^*$ as

1186: represented by the arrows in the figure. In addition a minor component is

1187: visible centered at the velocity of the center-of-mass

1188: i.e. V$_{\parallel}$=0. This emission

1189: pattern is consistent with

1190: anisotropic emission in the frame of 

1191: the PLF$^*$ and TLF$^*$. The most likely origin if the 

1192: observed backward enhancement, 

1193: i.e. towards mid-rapidity, is the asymmetry of the collision process itself.

1194: With increasing centrality, one observes an increase in this 

1195: backward yield, as well as an increase in the yield of the mid-velocity

1196: component. For b$<$7 fm, this mid-velocity yield becomes 

1197: considerable. 

1198: For the most central collisions, the distinct bimodal character 

1199: evident in more peripheral collisions is replaced by a broad distribution.

1200: The impact of Coulomb propagation and secondary decay is shown in the

1201: right column of Fig.~\ref{fig:Gal_IMF}.

1202: In contrast to the broad distributions observed at t=300 fm/c, 

1203: the emission pattern following Coulomb propagation to infinite 

1204: PLF-TLF separation and secondary decay (right column), 

1205: reveals a pattern of two semi-circles centered on the PLF$^*$ and 

1206: TLF$^*$ velocities. Such an emission pattern reflects both the

1207: Coulomb focusing in the field of the separating PLF$^*$ and TLF$^*$, 

1208: as well as emission of

1209: IMFs from the de-exciting PLF$^*$ and TLF$^*$.

1210: For the most peripheral collisions one observes two distinct 

1211: Coulomb circles. 

1212: It is important to note that the intensity pattern

1213: along each of these Coulomb circles is not constant but exhibits a 

1214: significant backward enhancement indicating a memory of the initial 

1215: angular asymmetry. 

1216: With decreasing impact parameter, the center of these 

1217: Coulomb circles shifts toward 

1218: the center-of-mass and increasingly overlap 

1219: as the velocity of the PLF$^*$ and TLF$^*$ decrease. The Coulomb circles also 

1220: become less distinct with increasing centrality reflecting both 

1221: increased excitation of the system and  nucleon-nucleon

1222: scattering.

1223: 

1224:   

1225: 

1226: \begin{figure}

1227: \vspace*{4.0in} 

1228: \special{psfile=Vpar_IMF_XLF_2.eps hscale=43.0 vscale=43.0 

1229: hoffset=0 voffset=0 angle=0}

1230: \caption[]

1231: {Left column: Parallel velocity distributions for the

1232: PLF$^*$ and TLF$^*$ (shaded), 

1233: as well as IMFs (solid histogram) as a function of impact parameter. 

1234: The dashed histograms correspond to a two gaussian fit as described

1235: in the text. 

1236: Right column: Parallel velocity distributions of the PLF, TLF, and IMFs 

1237: following Coulomb propagation and decay.

1238: The IMFs distributions have been normalized to the number of events.

1239: The PLF$^*$, TLF$^*$, PLF and TLF distributions have been

1240: scaled relative to the IMFs distributions.

1241: The arrows indicate the projectile and target velocities.} 

1242: \label{fig:V_dis}

1243: \end{figure}

1244: 

1245: The parallel velocity distributions of the PLF$^*$, TLF$^*$,

1246: and IMFs and their decay products are shown in Fig.~\ref{fig:V_dis} 

1247: as a function of impact parameter. 

1248: The velocity distributions of the PLF$^*$ and TLF$^*$ (left column) are 

1249: presented for reference (shaded histogram).

1250: For clarity these latter distributions have been scaled relative to 

1251: the IMF distributions by the factors indicated. 

1252: In the case of 8.5$<$b$\le$10 fm, the PLF$^*$ and TLF$^*$ manifest 

1253: gaussian-like velocity distributions that are relatively narrow and 

1254: slightly damped from the beam velocity. With 

1255: decreasing impact parameter, these two distributions move closer in velocity, 

1256: i.e. exhibit increased damping, and become broader. 

1257: The parallel velocity distributions of the PLF and TLF (right column) follow 

1258: the same general trends as those of the PLF$^*$ and TLF$^*$. The widths of the 

1259: secondary large fragments are typically 10-40 \% larger than that of the 

1260: PLF$^*$ and TLF$^*$.

1261: 

1262: 

1263: For the most peripheral collisions,

1264: the IMF velocity distribution (solid histogram) is bimodal with 

1265: the most probable values of this two peaked distribution 

1266: displaced toward the center-of-mass velocity as compared to the PLF$^*$ and 

1267: TLF$^*$ velocities, clearly establishing the qualitative trend first observed

1268: in Fig.~\ref{fig:Gal_IMF}.

1269:  In addition to the two gaussian yields attributable to the

1270: emission from the PLF$^*$ and TLF$^*$ an additional IMF component, 

1271: smaller in magnitude, is observed. As previously noted in

1272: Fig.~\ref{fig:Gal_IMF}, 

1273: this additional component has 

1274: an average velocity roughly centered at the center-of-mass velocity. 

1275: For 7$<$b$\le$8.5 fm, the relative magnitude of the mid-velocity 

1276: contribution is increased. 

1277: With increasing centrality, the shape of the IMF velocity 

1278: distribution evolves toward a flat distribution reflecting increased 

1279: fragment production at mid-velocity.

1280: 

1281: \begin{table}

1282: \caption{\label{tab:Vpar_fit_para}PLF$^*$ and TLF$^*$ average parallel velocity, 

1283: and fit parameters for the two gaussian fit of the IMF 

1284: V$_{\parallel}$ distributions at t=300 fm/c. 

1285: The deduced quantities, $\langle$V$_{\parallel}$$\rangle$ and $\sigma$ 

1286: are expressed in cm/ns.}

1287: \begin{ruledtabular}

1288: \begin{tabular}{ccccccc}

1289:  &\multicolumn{1}{c}{$PLF^*$}&\multicolumn{2}{c}{IMF(PLF$^*$)}&

1290:  \multicolumn{1}{c}{$TLF^*$}&\multicolumn{2}{c}{IMF(TLF$^*$)}\\

1291:  b (fm)&$\langle$V$_{\parallel}$$\rangle$

1292:  &$\langle$V$_{\parallel}$$\rangle$&$\sigma$

1293:  &$\langle$V$_{\parallel}$$\rangle$

1294:  &$\langle$V$_{\parallel}$$\rangle$&$\sigma$\\ \hline

1295:  8.5$<$b$\le$10&3.76&2.59&1.19&-4.65&-4.01&0.84 \\

1296:  7$<$b$\le$8.5&3.51&2.93&1.01&-4.31&-4.03&0.86 \\

1297:  4$<$b$\le$7&2.99&3.02&0.87&-3.57&-3.86&0.84 \\

1298:  b$\le$4&2.26&2.55&0.94&-2.48&-2.81&1.25\\

1299: \end{tabular}

1300: \end{ruledtabular}

1301: \end{table}

1302: 

1303: We have fit the predicted parallel velocity distributions shown with

1304: two gaussians representing the emission from the PLF$^*$ and TLF$^*$. 

1305: The result is depicted as the dashed histogram in Fig.~\ref{fig:V_dis}. 

1306: The fit parameters for the PLF$^*$ and TLF$^*$ emission are presented in 

1307: Table~\ref{tab:Vpar_fit_para}. 

1308: With increasing centrality

1309: the centroid of the PLF$^*$ velocity distribution decreases and the

1310: centroid of the TLF$^*$ velocity distribution increases as the reaction is

1311: increasingly damped.

1312: While for b$\ge$7 fm, a difference between the average parallel velocity for

1313: IMFs and the PLF$^*$ (or TLF$^*$) is discernible,  

1314: for 4$<$b$\le$7, the IMF distribution is

1315: centered on $\langle$V$_{PLF^*}$$\rangle$. The widths of the distributions

1316: are presented for completeness. No consistent trend of significance

1317: is evident in the extracted widths.

1318: 

1319: \begin{table}

1320: \caption{\label{tab:Vpar_fit_mult}Average multiplicity of 

1321: 4$\le$Z$<$$Z_{PLF^*}$, $Z_{TLF^*}$ for the PLF$^*$, 

1322: TLF$^*$ and mid-velocity (MV) components at t=300 fm/c.

1323: The relative fraction of the mid-velocity component to the total emission

1324: of the PLF$^*$ and TLF$^*$ is also indicated.}

1325: \begin{ruledtabular}

1326: \begin{tabular}{ccccc}

1327:  b (fm)&$PLF^*$&$TLF^*$&MV&P(MV) \\ 

1328:  8.5$<$b$\le$10&0.09&0.09&0.06&0.26 \\

1329:  7$<$b$\le$8.5&0.63&0.63&0.81&0.39 \\

1330:  4$<$b$\le$7&0.99&0.94&1.53&0.44 \\

1331:  b$\le$4&1.61&1.77&1.51&0.31\\

1332:  

1333: \end{tabular}

1334: \end{ruledtabular}

1335: \end{table}

1336: 

1337: We have also used the two gaussian fits previously described to 

1338: extract the average multiplicity 

1339: associated with the

1340: PLF$^*$, TLF$^*$ and mid-velocity components at t=300 fm/c.

1341: The results are tabulated in Table~\ref{tab:Vpar_fit_mult}. 

1342: With increasing centrality the multiplicities for each component increases

1343: although for the most central collisions, b$\le$ 7 fm the mid-velocity

1344: multiplicity seems to saturate at a value of $\approx$1.5. 

1345: From the peripheral collisions, 8.5$<$b$\le$10, 

1346: to the mid-central collisions, $4<b\le7$, the average 

1347: multiplicity of the PLF$^*$ and TLF$^*$ 

1348: components increases by a factor $\approx$10 with an increase by 

1349: $\approx$25 for the mid-velocity component. 

1350: The relative multiplicity of mid-velocity emission as compared to the total

1351: PLF$^*$ and TLF$^*$ emission increases  

1352: from 0.26 for peripheral collisions to 0.44 for more central 

1353: collisions. 

1354: 

1355: 

1356: The velocity distributions

1357: of the IMFs are significantly altered by secondary decay.

1358: This influence is most evident for the peripheral collisions 

1359: where the shape of the primary distribution is nearly 

1360: completely destroyed. Naturally, the magnitude of this

1361: secondary decay is particularly sensitive to the excitation 

1362: predicted for the primary fragments. The observed influence of 

1363: secondary decay on the IMF velocity distribution indicates 

1364: that the IMFs are significantly excited.

1365: 

1366: 

1367: \begin{figure*}

1368: \vspace*{4.2in}

1369: \special{psfile=Galilean_alpha.eps hscale=75.0 vscale=75.0 

1370: hoffset=30 voffset=0 angle=0}

1371: \caption[]

1372: {(Color online) 

1373: Invariant cross-section for $\alpha$ particles in the COM frame.

1374: The arrows indicate the average parallel velocity of the PLF$^*$ and TLF$^*$. 

1375: The color scale indicates the yield on a logarithmic scale. In the left column

1376: the vertical scale of the bottom panel is scaled by a factor of six as compared

1377: to the upper panels. The right column is scaled by a factor of three with 

1378: respect to the left column.}

1379: \label{fig:Gal_alpha}

1380: \end{figure*}

1381: 

1382: 

1383: For peripheral collisions, it has been experimentally observed that 

1384: the emission pattern of $\alpha$ particles emitted by the PLF$^*$ 

1385: manifests an anisotropic distribution \cite{Hudan04}.

1386: This observed anisotropy has been interpreted as the decay of

1387: a PLF$^*$ (and TLF$^*$) initially deformed by the collision process. To 

1388: investigate the extent to which such a physical picture is compatible with the 

1389: AMD model, we have examined the invariant cross-section 

1390: maps of $\alpha$ particles. 

1391: Depicted in Fig.~\ref{fig:Gal_alpha} is the dependence of the invariant 

1392: cross-section map for $\alpha$ particles on impact parameter 

1393: both at t=300 fm/c and following Coulomb propagation to 

1394: infinite separation and sequential decay. At t=300 fm/c (left column), 

1395: for the most peripheral collisions the yield is peaked near

1396: the average PLF$^*$ and TLF$^*$ velocities (indicated by arrows), though

1397: slightly toward the center-of-mass velocity.

1398: For these most peripheral collisions one observes that the primary $\alpha$

1399: yield centered at mid-velocity is relatively small.

1400: With decreasing impact parameter,

1401: the primary alpha distributions associated with the PLF$^*$ and TLF$^*$ 

1402: move closer in velocity and increasingly overlap. 

1403: 

1404: Following secondary decay, the Coulomb circles evident for $\alpha$ particles are even more striking than

1405: those for IMFs. This observation is consistent with the large multiplicity of

1406: $\alpha$ particles that originate from the de-excitation of the 

1407: PLF$^*$ and TLF$^*$ as compared to the early dynamical stage.

1408: The distinct emission pattern observed for 8.5$<$b$\le$10 fm is 

1409: also observed for more 

1410: central collisions although with increasing centrality 

1411: the distinct nature of the 

1412: semi-circles becomes less striking. The ridge of yield which is typically

1413: interpreted as Coulomb barrier emission becomes broader and 

1414: its center moves increasingly toward 

1415: V$_{\parallel}$=0. These trends are consistent with the increased 

1416: damping, excitation, and reduced size of the PLF$^*$ (TLF$^*$) with increasing 

1417: centrality. For even the most peripheral collisions, the pattern 

1418: evident in the right column of Fig.~\ref{fig:Gal_alpha} 

1419: is clearly not isotropic, favoring backward emission. 

1420: As the sequential decay of the PLF$^*$ following t=300 fm/c is taken 

1421: to be that of an isolated spherical nucleus without considering 

1422: the influence of the external Coulomb field of the target on its decay

1423: \cite{Hudan04}, 

1424: it does not contribute to the predicted anisotropy. 

1425: Within the model calculation, the observed anisotropy 

1426: has two possible origins: Coulomb focusing of the $\alpha$ 

1427: particles present at t=300 fm/c and the $\alpha$ decay of 

1428: IMFs which are emitted anisotropically. 

1429: 

1430: 

1431: %\begin{figure}

1432: %\vspace*{4.0in} 

1433: %\special{psfile=Vpar_Alpha_XLF.eps hscale=43.0 vscale=43.0 

1434: %hoffset=0 voffset=0 angle=0}

1435: %\caption[]

1436: %{Left column: Parallel velocity distributions for PLF$^*$ 

1437: %and TLF$^*$ (dashed),

1438: % as well as $\alpha$ particles (solid) as a function of 

1439: %impact parameter.

1440: %Right column: Parallel velocity distributions of the PLF, TLF, and 

1441: %$\alpha$ particles following Coulomb propagation and decay.

1442: %The arrows indicate the beam and target velocities.} 

1443: %\label{fig:V_dis_alpha}

1444: %\end{figure}

1445: 

1446: \begin{figure}

1447: \vspace*{4.0in}

1448: \special{psfile=Alpha_Ridge_zero.eps hscale=43.0 vscale=43.0 

1449: hoffset=0 voffset=0 angle=0}

1450: \caption[]

1451: {Angular distribution for $\alpha$ particles on the PLF$^*$ Coulomb ridge 

1452: (V$_\alpha$$<$ 3.5 cm/ns). All angles are in the PLF$^*$ frame.}

1453: \label{fig:Alpha_ridge}

1454: \end{figure}

1455: 

1456: The anisotropic emission of $\alpha$ particles along the PLF$^*$ 

1457: Coulomb ridge has recently been proposed to be 

1458: related to the enhanced backward decay of the 

1459: excited PLF$^*$ due to the nucleus-nucleus interaction \cite{Hudan04}. 

1460: Displayed in Fig.~\ref{fig:Alpha_ridge} is the $\alpha$ particle yield 

1461: along the Coulomb ridge

1462: for ``peripheral'' collisions, 5$<$b$\le$10 fm, and ``central'' collisions, 

1463: b$\le$5 fm.

1464: Alpha particles were selected to be ``Coulomb barrier'' particles by 

1465: restriction on their velocities, namely V$_\alpha$$<$ 3.5 cm/ns in the PLF$^*$ 

1466: frame. In both cases shown, the total $\alpha$ 

1467: particle yield (solid histogram) is not symmetric with

1468: respect to emission transverse to the PLF$^*$ direction, 

1469: namely cos($\theta_\alpha$)=0. Emission in the backward direction 

1470: cos($\theta_\alpha$)$<$0 is enhanced with respect to the forward direction.

1471: This enhancement is more pronounced for the central collisions.

1472: For the peripheral collisions, the emission yield

1473: for cos($\theta_\alpha$)= -1 is approximately 1.7 times the yield

1474: emitted in the transverse direction. In contrast, the forward emission yield

1475: cos($\theta_\alpha$)= +1 is approximately the same as the transverse yield.

1476: Comparison of the integrated yield with -1$\le$cos($\theta_\alpha$)$<$0,

1477: Y$_{backward}$($\alpha$), to

1478: 0$<$cos($\theta_\alpha$)$\le$+1,

1479: Y$_{forward}$($\alpha$), reveals that backward emission is enhanced

1480: by $\approx$19\% as compared to forward emission. 

1481: For more central collisions (bottom panel), 

1482: comparison of the integrated yield 

1483: reveals that backward emission is enhanced

1484: by $\approx$39\% as compared to forward emission. 

1485: 

1486: 

1487: We have investigated the origin of this backward enhancement, by examining the

1488: possible sources of $\alpha$ particles. Alpha particles are ``tagged'' as 

1489: being either

1490: a) ``primary'', namely those originating at the time of 

1491: cluster recognition (t=300 fm/c), 

1492: b) PLF$^*$ alphas or c) cluster alphas i.e. those 

1493: that result from 

1494: the secondary decay of primary IMFs. 

1495: As expected, PLF$^*$ 

1496: emission is essentially isotropic. 

1497: It is evident in

1498: Fig.~\ref{fig:Alpha_ridge} that for both peripheral and central collisions, 

1499: primary alphas (open triangles) on the PLF$^*$ Coulomb ridge are isotropic. 

1500: Evidently the Coulomb focusing of primary $\alpha$ particles by the 

1501: PLF$^*$ and TLF$^*$ does not contribute to the anisotropy observed in 

1502: Fig.~\ref{fig:Gal_alpha}.

1503: The large backward enhancement observed for the total $\alpha$ particle 

1504: yield is associated with the $\alpha$ particles 

1505: that originate from the secondary decay of primary IMFs. 

1506: Hence, it is the anisotropy of the primary IMFs 

1507: that is responsible for the anisotropy of $\alpha$ particles 

1508: associated with Coulomb barrier energies. 

1509: Quantitative comparison of the 

1510: the various components reveals that for peripheral collisions 

1511: the ratio Y$_{backward}$($\alpha$)/Y$_{forward}$($\alpha$) associated with the

1512: decay of primary clusters is $\approx$2.2, while the same ratio for

1513: PLF$^*$ emission or primary $\alpha$ emission is $\approx$0.9.

1514: In the case of more central collisions, 

1515: the ratio Y$_{backward}$($\alpha$)/Y$_{forward}$($\alpha$) associated with the

1516: decay of primary clusters is $\approx$1.7. The yield ratio for 

1517: PLF$^*$ emission and primary $\alpha$ emission is $\approx$1.0 

1518: in this impact parameter interval.

1519: 

1520: The anisotropy observed for the decay of 

1521: primary clusters is consistent with the emission pattern of IMFs 

1522: as shown in Fig.~\ref{fig:Gal_IMF}. This feeding of 

1523: $\alpha$ particles to the PLF$^*$ Coulomb ridge from IMF secondary decay 

1524: is also consistent with the fact that IMFs are produced excited, 

1525: even for the most peripheral collisions. 

1526: In fact, the average excitation energy of the IMFs is relatively independent of 

1527: the impact parameter. For the most peripheral collisions,

1528: $\langle$E$^*$/A$\rangle$ of the IMFs is typically 2.5 to 3 MeV 

1529: with the higher values associated with IMFs produced around 

1530: the center-of-mass velocity.

1531: With increasing centrality, $\langle$E$^*$/A$\rangle$ becomes independent of

1532: the IMF velocity and reaches a typical value of 3 MeV. 

1533: Such an excitation energy is in agreement with the excitation energy 

1534: experimentally  deduced for IMFs produced in central 

1535: collisions \cite{Hudan03}.  

1536: Investigation of the width of the IMF excitation energy distribution 

1537: reveals that it is large and 

1538: approximately independent of the impact parameter.

1539: 

1540: In addition to the anisotropies predicted by the model, in reality  

1541: the anisotropic emission 

1542: pattern of $\alpha$ particles can have additional origins. 

1543: Although the PLF$^*$ is clearly deformed for t$\le$300 fm/c (as shown in 

1544: Fig.~\ref{fig:density}), 

1545: the statistical decay of the PLF$^*$ (and TLF$^*$) is assumed to be isotropic.

1546: However, if the collision dynamics preferentially ``prepares'' the system in 

1547: a configuration that favors emission toward the center-of-mass, the observed

1548: emission pattern will certainly be anisotropic. An example of such a favored 

1549: configuration would be a di-nuclear configuration of the PLF$^*$ decaying into

1550: an IMF and residue with the IMF preferentially oriented toward mid-rapidity.

1551: If the di-nuclear configuration prepared lies outside the saddle point 

1552: for such a system, then the excitation energy of the di-nuclear configuration

1553: does not influence the decay probability

1554: and the decay is clearly non-statistical. However, if the di-nuclear 

1555: configuration lies inside the saddle point, excitation energy does 

1556: influence the decay probability and the emission can be considered statistical.

1557: In this case, explicit treatment of the deformation within a statistical 

1558: framework is necessary \cite{Charity05}.

1559: The observed anisotropy under such conditions will 

1560: depend sensitively on the emission time

1561: relative to the rotational period of the di-nuclear system. Of course such a 

1562: schematic description of the binary decay of the PLF$^*$ could be extended to

1563: ternary and quaternary decays.

1564: It should also be noted that such short timescale emission 

1565: when the nuclei are 

1566: in proximity of each other and can also be influenced 

1567: by tidal effects \cite{Charity01}.

1568: 

1569: 

1570: \section{Summary and Conclusions}

1571: 

1572: Using the AMD model,  we have examined the dynamical phase of a 

1573: heavy-ion collision at intermediate energy. 

1574: We have investigated how observables such as the size (Z), velocity, 

1575: and excitation of the reaction products evolve during the early stages of the 

1576: collision. The de-excitation of the initial reaction products is 

1577: calculated with a statistical decay code

1578: and the survival of these initial observables is examined.

1579: 

1580: 

1581: We have investigated how the characteristics of the two large

1582: remnants in the reaction evolve with impact parameter.

1583: Both the 

1584: $\langle$Z$\rangle$ and $\langle$V$\rangle$ of the PLF$^*$ and TLF$^*$ 

1585: decrease smoothly as centrality increases 

1586: up to an impact parameter of $\approx$4 fm. 

1587: As the centrality increases from the most peripheral collisions, 

1588: the PLF$^*$'s velocity is increasingly damped 

1589: from the projectile velocity. Concurrent with this damping, 

1590: the width of the velocity distribution increases. 

1591: Although the average velocity is largely unchanged 

1592: by secondary decay, the width of the velocity distribution

1593: is typically increased by 10-40 \%.

1594: For smaller impact parameters, b$<$3-4 fm, the average atomic number and 

1595: velocity of the two reaction partners are independent of impact parameter. 

1596: Associated with these changes in the size and velocity

1597: and size of the PLF$^*$ and TLF$^*$, one also observes that 

1598: the $\langle$E$^*$/A$\rangle$ of the PLF$^*$ and TLF$^*$ increases 

1599: as the impact parameter decreases

1600: from an initial value of 0.7-1.1 MeV upto 4 MeV. 

1601: The maximum excitation energy is attained for an impact

1602: parameter of $\approx$6 fm. Smaller impact parameters do not result in

1603: larger values of $\langle$E$^*$/A$\rangle$.

1604: These observations suggest that the peripheral collisions on one side and 

1605: the most central collisions on the other side correspond to different 

1606: dynamics regime although simulated with the same ingredients.

1607: 

1608: 

1609: Peripheral collisions, as may be expected, exhibit a binary nature 

1610: with a strong memory of the entrance channel.

1611: In such collisions, a transiently deformed PLF$^*$ and TLF$^*$ 

1612: are recognizable as early as

1613: $\approx$100 fm/c after the collision. 

1614: The deformation of these reaction products persists for a considerable time, 

1615: t$\ge$300 fm/c. 

1616: In addition to the PLF$^*$ and TLF$^*$, nucleons, light charged particles,

1617: and IMFs are also produced in the dynamical phase. The latter clusters,

1618: are preferentially located between the PLF$^*$ and TLF$^*$.

1619: The Z distribution of particles with  V$_\parallel$$>$0 strongly favors 

1620: asymmetric splits. The population of symmetric splits increases for

1621: mid-peripheral collisions reflecting an increase in the excitation energy 

1622: of the PLF$^*$. The excitation energy of the PLF$^*$ is strongly correlated 

1623: with its velocity damping.

1624: Both the $\langle$E$^*$/A$\rangle$ and $\langle$V$_{PLF^*}$$\rangle$ 

1625: in this impact parameter range 

1626: are found to be independent of the cluster recognition time.

1627: The correlation between the excitation energy of the PLF$^*$ 

1628: and velocity damping is the same for the different cluster recognition times 

1629: studied. The general insensitivity of $\langle$V$_{PLF^*}$$\rangle$ 

1630: to cluster recognition time makes it a robust signal of the impact parameter.

1631: The excitation energy of the PLF$^*$ is slightly correlated with 

1632: the excitation energy of the TLF$^*$ for early cluster recognition times. 

1633: Even the small particle emission that occurs on 

1634: short timescale is sufficient to destroy 

1635: this correlation by t=300 fm/c.

1636: 

1637: 

1638: In contrast to peripheral collisions which exhibit a strong binary character, 

1639: the most central collisions do not manifest as much memory of the 

1640: entrance channel. 

1641: Such collisions are no longer dominated by two large fragments, 

1642: namely the PLF$^*$ and TLF$^*$. However, if we designate the largest fragment 

1643: forward and backward of the center-of-mass as the  PLF$^*$ and TLF$^*$, 

1644: their characteristics, $\langle$Z$\rangle$, $\langle$V$\rangle$ and 

1645: $\langle$E$^*$/A$\rangle$, are largely unchanged as b decreases 

1646: for b$\le$4 fm. 

1647: Therefore, for the innermost $\approx$10 \% of the total cross-section,  

1648: the maximum degree of excitation for such collisions is attained. 

1649: This broad range of impact parameters associated with high excitation 

1650: underscores the importance of considering the breakup of non-spherical 

1651: geometries \cite{leFevre99}.

1652: Moreover, for these small impact parameters, the 

1653: quantitative characteristics of 

1654: the PLF$^*$ and TLF$^*$, $\langle$Z$\rangle$ and $\langle$E$^*$/A$\rangle$, 

1655: depend on the cluster recognition time.

1656: For early cluster recognition time (t=150 fm/c) an average 

1657: excitation energy of 6 MeV 

1658: is reached, while a longer cluster recognition time (300 fm/c) results in 

1659: $\langle$E$^*$/A$\rangle$ of  4 MeV. This 

1660: decrease in $\langle$E$^*$/A$\rangle$ 

1661: indicates a rapid de-excitation during the dynamical stage, 

1662: suggesting significant nucleon and cluster emission on a 

1663: short timescale.

1664: 

1665: Direct examination of the multiplicities of emitted particles reveals that

1666: the IMF multiplicity increases smoothly with increasing centrality 

1667: and saturates for an impact parameter of $\approx$3 fm. 

1668: At t=300 fm/c, the average IMF multiplicity reaches unity 

1669: for an impact parameter of $\approx$9 fm. 

1670: At all impact parameters and for t$>$150 fm/c, the IMF emission rate 

1671: decreases monotonically with increasing cluster recognition time.  

1672: The velocity distribution of the produced IMFs at t=300 fm/c is bimodal 

1673: and reveals preferential emission from the 

1674: PLF$^*$ and TLF$^*$ 

1675: towards the center-of-mass. 

1676: The emission pattern of $\alpha$ particles at t=$\infty$ also exhibits a 

1677: distinct preferential emission towards the center-of-mass.

1678: This anisotropy however, is not due to the anisotropic emission of primary 

1679: $\alpha$ particles or 

1680: evaporation from the PLF$^*$ and TLF$^*$, 

1681: but arises from the secondary decay of anisotropically emitted primary IMFs.  

1682: The multiplicity of neutrons saturates for mid-central collisions, making 

1683: neutron multiplicity a poor selector of central collisions. The geometric 

1684: cross-section combined together with the saturation of excitation energy for 

1685: mid-peripheral collisions may explain the observed persistence of 

1686: binary collisions at intermediate energies even when the largest 

1687: neutron multiplicities are selected \cite{Lott92}.

1688: 

1689: The large excitation energy reached in the collision 

1690: leads to rapid particle emission on the dynamical timescale. However, the 

1691: present treatment of  

1692: the short timescale decay involves several simplifications. 

1693: The role of deformation in the decay is neglected as are both Coulomb 

1694: and nuclear proximity effects. In addition the excitation energy is 

1695: calculated relative to clusters that are at the ground-state both in 

1696: shape and density. These simplifications may have a non-negligible impact on 

1697: the characteristics of the fragmenting system.

1698: This rapid de-excitation emphasizes the need for 

1699: a hybrid statistical-dynamical model that considers in a more realistic manner

1700: the statistical decay of the transiently deformed nuclei 

1701: from times as short as 100 fm/c.

1702: Development of such a hybrid model would represent a new and potentially 

1703: powerful tool in understanding the dynamics of intermediate 

1704: energy heavy-ion collisions, as well as cluster formation on short timescales.

1705: 

1706: 

1707: \begin{acknowledgments}

1708: 

1709: This work was supported by the

1710: U.S. Department of Energy under DE-FG02-92ER40714 (IU)

1711: and  in part by Shared University Research grants from 

1712: IBM, Inc. to Indiana University. In addition, A. O. would like to thank the National 

1713: Superconducting Cyclotron Laboratory at Michigan State University for the 

1714: warm hospitality

1715: extended to him during his long term stay.

1716: 

1717: \end{acknowledgments}

1718: 

1719: 

1720: \bibliography{amd.bib} 

1721: 

1722: \end{document}

1723: 

1724: 

1725: 

1726: 

1727: 

1728: 

1729: