Shadow poles in the alternative parametrization of R -matrix theory

Abstract

Nuclear data libraries (ENDF, JEFF, JENDL, CENDL, etc.) document our phenomenological knowledge of nuclear cross sections as interpreted by R-matrix theory. The R-matrix scattering model can parameterize the energy dependence of the scattering matrix in different ways. This article establishes new results on three such sets of parameters: the Wigner-Eisenbud, the Brune, and the Siegert-Humblet parameters. We show how the latter two arise from invariance to the arbitrary boundary condition, and how they can trade-off more complex parameters for a simpler energy dependence parameterization: the scattering matrix pole expansion of Humblet and Rosenfeld. We establish that the scattering matrix's invariance to channel radius sets a partial differential equation on the widths of the Kapur-Peierls operator, which enables us to derive an explicit transformation of the Siegert-Humblet radioactive residue widths under a change of channel radius. Considering the continuation of the scattering matrix to complex wavenumbers, several new results are established. We unveil there exist more Brune parameters than previously thought, depending on the way the scattering matrix is continued to complex energies. This points to a broader conundrum in the field: how to analytically continue the scattering matrix while closing sub-threshold channels. We argue that, contrary to the Lane and Thomas legacy of force-closing sub-threshold channels which introduces spurious poles, analytic continuation is the physically correct way of defining the scattering matrix for complex wave numbers. To back this claim we establish thee results: analytic continuation cancels spurious poles, enforces the generalized unitarity conditions of Eden and Taylor, and closes massive particle channels below threshold by both quantum tunneling and from the definition of the cross section as the ratio of probability currents.