CRITICAL BEHAVIOUR IN PERIPHERAL Au + Au COLLISIONS AT 35 MeV/u M. Bruno\ P. F. Mastinu^'^, M. Belkacem^''', M. D'Agostino\ P. M. Milazzo^'^, G. Vannini^ D. R. Bowman^, J. D. Dinius'', A. Ferrero*'*, M. L. Fiandri\ C. K. Gelbke^ T. Glasmacher® , F. Gramegna^ D. O. Handzy^ D. Horn'^, W. C. Hsi^ M. Huang^ I. lori*, G. J. Kunde^ M. A. Lisa^ W. G. Lynch^ G. V. Margagliotti^ C. P. Montoya®, A. Moroni**, G. F. Peaslee^ R. Rui^ C. Schwarz^ M. B. Tsang^ C. Williams®, V. Latora® and A. Bonasera^ ^ Dtpartimento di Fisica and INFN, Bologna, Italy ^ Dtpartimento di Fisica and INFN, Trieste, Italy ^ Dtpartimento di Fisica, Padova, Italy * Dtpartimento di Fisica and INFN, Milano, Italy ^ INFN, Laboratori Nazionali di Legnaro, Italy ^ NSCL, Michigan State University, USA ^ Chalk River Laboratories, Chalk River, Canada * On leave from CNEA, Buenos Aires, Argentina ^ INFN, laboratorio Nazionale del Sud, Catania, Italy The signals theoretically predicted for the occurrence of a critical behavior (condi- tional moments of charge distributions, Campi scatter plot, fluctuations of the size of the largest fragment, power law in the charge distribution, intermittency) have been found for peripheral events in the reaction Au+Au at 35 MeV/u. The same signals have been studied with a dynamical model which foresees phase transition, like the Classical Molecular Dynamics. 1 Introduction The liquid-gas phase transition in nuclean systems has been recently theoret- ically and experimentally investigated EfHij. In this contribution we report on the search for critical behavior signals in the experimental data of the Au + Au reaction at 35 MeV/u. The experiment was performed at the National Superconducting Cyclotron Laboratory of the_Michigan State University tak- ing advantage of the coupling of the MulticsQ and the MiniballEl apparatus. The analysis has been restricted to events corresponding to peripheral and semi-peripheral reactions, selected imposing that the component of the veloc- ity of the largest fragment along the beam direction was greater than 75% of the beam velocity and that the total detected charge was between 70 and 90. Most of these events come from the disassembly of the quasi-projectile, since the detection of the quasi-target is suppressed due to the thresholds of the detectors. Comparisons with the results of calculations in the framework of Classical 1 Figure 1: Campi scatter plot, In(Zmax) versus ln{M2). The three different regions are discussed in the text. Fission events are to the right of region 2. Molecular Dynamics model are also presented. 2 Experimental results The events selected with the previously mentioned criterion-have been plotted (see Fig. 1) as In^Zi^^^^) vs. ln{M2) (Campi scatter plotQ), where ^j^aa; is the charge of the heaviest fragment and M2 is the second conditional moment of the charge distribution detected in the j-th event, Afj"*^ = Yliz ■^^"■j(^)/^o where nj{Z) is the number of fragments of charge Z detected in the j-th event, Zq is the total charge and the summation is over all fragments but the heaviest detected one. The data in Fig. 1 are distributed along two branches as predicted by per- colation calculation for undercritical and overcritical events. To have a deeper insight in the behavior of events falling in different regions of the Campi-plot, we selected three different cuts in the upper branch (cut 1), in the lower one (cut 3) and in the intersection region (cut 2). The charged particle multiplicity distributions observed for these three cuts show that cuts 1 and 3 select low and high multiplicity events with a narrow range, whereas far the cut 2 a wide range of charged particle multiplicities has been observed possibly related to the occurrence of large fluctuations as expected at the critical point. Even when the width of region 2 is reduced, the multiplicity distribution remains quite broad. The fragment charge distribution corresponding to cut 1 contains light fragments and heavy residues, thus exhibiting a "U" -shaped distribution. For cut 3 it is rapidly decreasing. The results obtained for cut 1 and cut 3 are sirp.- ilar to the predictions of percolation calculations in the sub-critical region with the probability p higher than the critical one (pc)j and in the overcritical 2 Figure 2: Relative variance 72 (left panel) and normalized variance of the charge of the largest fragment crjvv (right panel) as a function of charged particle multiplicity region (pTc), respectively. The fragment charge distribu- tion for cut 2 shows a power-law distribution, P{Z) cx Z^'^ , with t « 2.2. For macroscopic systems exhibiting a liquid-gas phase transition, such a power-law distribution is predicted to occur near the critical point 113. An analysis in term of Scaled Factorial Moments (SFM) has been per- formed. The SFM are defined B as where Ztot = 158, and i is the order of the moment. The total interval [1, Ztot] is divided into M — Ztot/ 6s bins of size Ss, Uk is the number of particles in the fc-th bin for an event, and the brackets <> denote the average over many events. The values ln{Fi) for i — 2, 5 are always negative (i.e. the variances are smaller than for a Poissonian distribution) and almost independent on 5s for cut 3. For cut 2, ln{Fi) are positive and almost linearly increasing as a function of — hi(As) (i.e. Fi oc 5s~^^), and this, as pointed out by sevemLthe- oretical studies Lnl3, indicates an intermittent pattern of fluctuations EIIEjcJ. Region 1, corresponding to evaporation, gives zero slope. Increasing or re- ducing the sizes of the three cuts does not change significantly these results. Further signals which could reveal the presence of a critical behaviour have been investigated. The second moment M2 shows a peak versus the multiplicity of charged particlesin the region of A^c — 20. A similar value isj:ii>±ained by the EOS Collaborations. Also the relative variance 72, defined asQa: 72 — ^^1° 3 Figure 3: Campi scatter plot (left panel) and normalized variance of the charge of the largest fragment (tjvv vs. charged particle multiplicity (right panel) for the events predicted by CMD calculations, filtered through the experimental acceptance, with the same conditions applied to the experimental data to select peripheral events (see text). shows a peak for iVc « 18 — 22 (see Fig. 2), consistent with that observed for M2; this means that around Nc « 20 the fluctuations in the fragment size distributions are large, as it should be near the critical pointlZl. A further signal for criticality, recently proposedtJ, is the normalized vari- ance of the charge of the largest fragment cfnv- This quantity, defined by: ^2 <7nv = <:z"""'y where (^z^^^ ^max > — < Z^ax >^ shows a peak at the critical point, where charge distributions are expected to show the largest fluctuations. The right side of Fig. 2 shows the (tnv versus charged particle multiplicity for the experimental data. A clear peak is present for multiplicities Nc~15^ 20. The analysis of the experimental data suggests that different regions of the nuclear phase diagram can be probed at one incident beam energy by selecting different events E3. All the signals so far proposed to characterize a critical behavior have been found. We must however caution that the effects of finite experimental acceptance and the mixing of possible contributions from the decay of projectile-like fragments and the neck-region are not yet sufficiently well understood to allow an unambiguous conclusion. 3 Model calculations In the framework of the CMD model, calculations have been performed for impact parameter ranging from 1 to 13 fm. The Campi scatter-plot for the calculated events is very similar to the experimental one. Also M2, 72 and a^v show a peak for b « 10 fm and for iVc « 20 — 25. An analysis of the events in a region corresponding to the region 2 of the experimental data show a similar 4 Figure 4: Normalized variance of the charge of the largest fragment (Tjvv vs. charged particle multiplicity for the events predicted by simulations as described in the text. behavior with a clear signal of intermittency. To account for the angular acceptance and detection thresholds of the ap- paratus, the predictions have been suitably filtered; the same type of selection used to characterize the experimental events of peripheral or semi-peripheral reactions has then been applied. All the signals are clearly visible; in Fig. 3 the Campi-plot and the unv with a peak in the region of b « 10 fm and for Nc ~ 15 are shown. The same kind of analysis performed for experimental data gives the same results both for the charge distribution and for the SFM. 4 Discussion and Conclusions In addition to the calculations based on percolation and on statistical and dynamical models, a series of simulations have been performed, mainly starting from a given—pass or charge distribution: considering the charge pand mass conservationEII or starting from a simple power law mass distributionEj, several signals like the Campi plot, AI2, 72 and the intermittency have beenjObserved. Recently a simulation based on an exponential charge distribution 113 with a binomial coefhcient shows several of the signals experimentally obtained. In particular, peaks have been obtained for M2 and 72, but for instance no peak is present for (Tnv, as shown in Fig. 4. The simple presence of peaks, however, could be not relevant and it could be important to investigate position, height and widths of the peaks. On the other side, the signals proposed so far might be not sufficient to characterize a critical behavior or the data extracted by the simulation could be "somehow" connected to the experimental data which contain the criticality, through some constraint like the charged particle multiplicity distribution. In conclusion we have analyzed the peripheral events of the reaction Au -|- Au 5 at 35 MeV/u and wc have found all the signals proposed to evidence out a critical behavior; these findings have been confirmed by predictions based on a dynamical model like CMD. Further work is needed before suggesting any definite conclusion. References 1. see e.g. H. R. Jaqaman, Gabor Papp and D. H. E. Gross, Nucl. Phys. A 514, 327 (f990); J. Pochodzalla et al, Phys. Rev. Lett. 75, 1040 (1995). 2. J. B. Elliot et al, Phys. Rev. C 49, 3185 (1994); M. L. Gilkes et al, Phys. Rev. Lett. 73, 1590 (1994). 3. M. Belkacem, V. Latora and A. Bonasera, Phys. Rev. C 52, 271 (1995). 4. I. lori et al, Nucl Lustrum. Methods A 325, 458 (1993). 5. R. T. DeSouza et al, Nucl lustrum. Methods A 49, 109 (1990); 6. R. J. Lenk, T. J. Schlagel and V. R. Pandharipande, Phys. Rev. C 42, 372 (1990). 7. X. Campi, J. of Phys. A 19, L917 (1986); J. de Phys. 50, 183 (1989), X. Campi and H. Kr ivine, Nucl Phys A 589, 505 (1995). 8. P. F. Mastinu et al. \nucl- ex/960400 j XXXIV Int. Winter Meeting on Nucl. Phys, Bormio, 1996, Ricerca Scientifica, Ed. Perm, Vol. 102, page 110. 9. W. Bauer et al, Phys. Lett. B 150, 53 (1985); Nucl Phys. A 452, 699 (1986); Phys. Rev. C 38, 1297 (1988). 10. M. Belkacem et al XXXIV Int. Winter Meeting on Nucl. Phys, Bormio, 1996, Ricerca Scientifica, Ed. Perm, Vol. 102, page 96. 11. H. R. Jaqaman and D. H. E. Gross, Nucl Phys. A 524, 321 (1991); D. H. E. Gross, et al Phys. Rev. Lett. 68, 146 (1992). 12. M. E. Fisher, Rep. Prog. Phys. 30, 615 (1967); Physics 3, 255 (1967). 13. M. Ploszajczak and A. Tucholski, Phys. Rev. Lett. 65, 1539 (1990); Nucl Phys. A 523, 651 (1991). 14. A. Bialas and R. Peschanski, Nucl Phys. B 273, 703 (1986); Nucl Phys. B 308, 857 (1988). 15. J. B. Elliot, Phase Transitions in Small Systems, GSI, Darmstadt, De- cember 1995. 16. J. Pochodzalla (this conference) 17. Sa Ben-Hao et al J. o Phys. G: Nucl. Part. Phys. 21, 241 (1995). 18. B. Elattari, J. Richert and P. Wagner Nucl Phys. A 560, 603 (1993) 19. L. Phair {private commuuicatiou) . 6