| Neutron resonance parameters of medium weight and heavy nuclei remain one of the most important information for testing random matrix theory (RMT), even more than 50 years after E. P. Wigner formulated this theory. Statistical behavior of eigenvalues and eigenvectors which is predicted by the RMT is not restricted only to nuclear physics, as it pervades the physics of other quantum-mechanical systems. Moreover, the RMT led to a discovery of quantum chaos. Given the rotational and time-reversal symmetries, the well-known Gaussian orthogonal ensemble (GOE) option of the RMT is expected to describe spectral fluctuation behavior of energies of neutron resonances and also fluctuation properties of reduced neutron widths of these resonances. Regarding widths, their probability distribution function (PDF) following from the RMT, known as Porter-Thomas distribution (PTD), has been anticipated before the RMT emerged; this PDF is identical with that of the distribution with the parameter, the number of degrees of freedom, equal to unity. Currently, the consensus is that the GOE RMT prediction of the joint PDF for resonance energies, as well as the PTD for widths holds. The validity of the join PDF for resonance energies has been extensively tested in 1981 with the aid of the statistic on the so-called nuclear data ensemble (NDE), formed by 1250 neutron and 157 proton resonance energies obtained from measurement on 27 different target nuclei. However, to make reliable conclusions regarding the validity of the PTD, it is important that (i) the set of data on neutron widths has to be large, (ii) free of contamination by p-wave resonances, and (iii) the number of missing s-wave resonances due to finite experimental threshold should be under the reliable control. With these requirements, even the best test of the PTD to date does not seem to be reliable. We show that this is the case also for the test based on the values of for neutron resonances of the NDE, which undermines the veracity of this ensemble. To this point, the effective value of parameter thus remains unknown. We present results of measurements and fluctuation analysis of the reduced neutron widths of 192,194Pt+n systems. Problems (i-ii) have been successfully overcome by selecting the nuclei displaying a large number of well-resolved resonances (158 and 411, respectively) and by an exceptionally high value of the s- to p-wave neutron strength function ratio, 10. The problem (iii) was solved by introducing a well-defined, energy-dependent artificial threshold imposed on widths. This threshold was chosen to guarantee a full discrimination of intruder p-wave resonances. In addition, our novel analysis takes fully into account the missing sub-threshold s-wave resonances. The best maximum-likelihood based estimates of deduced from our 192,194Pt data we arrived at are = ±0.16 and = 0.16, respectively. These results were subject to an unbiased hypothesis testing which led to a conclusion that the validity of the PTD is rejected at a statistical confidence level of 99.997 %. This finding is briefly discussed in the light of new theories of complex quantum-mechanical systems. |
|