Communication: Convergence of anharmonic infrared intensities of hydrogen fluoride in traditional and explicitly correlated coupled cluster calculations

Reduced Radial Curves of Diatomic Molecules

“…Regarding HF, in Figure , we can see that although the corresponding empirical and theoretical dispersion curves differ significantly, they still exhibit topologically very similar vibrational dependencies (two shapes are topologically equivalent if one can be transformed into the other without any cutting or gluing). As is seen in the bottom panels of Figure , a particularly close coincidence is exhibited by the curves obtained using explicitly correlated methods in ref , evidencing thus their suitability for the performed modeling. Expectably, as seen in Figure , the same dispersion is exhibited also by the curves obtained by “dereducing” the reduced forms of the original electric dipole moment functions.…”

“…However, because the level of accuracy achievable by both experiments and theory is only rarely comparable to that of data determining potential energy functions (see, e.g., ref ), the dispersion of reduced electric curves constructed from available literature data is greater than that of corresponding potential energy curves. Moreover, as seen, for example, in Figures S2 and S3, illustrating 40 theoretical EDM functions evaluated in ref for HF and their reduced counterparts, most of the available “electric” functions possess unphysical asymptotes. Nonetheless, as can also be seen in the same figures, the curves available may acquire adequate shapes over the whole region of the vibrational displacements needed for a correct evaluation of the needed wave functions, and the slow convergence of the dispersion of the probed theoretical curves to their theoretical limit can be significantly enhanced by using explicitly correlated methods. − Still, however, as shown in the bottom panels of Figure , differently augmented and correlated methods provide slightly different reduced curves, thus indicating a limited degree of their “universality” and the need to assess the role of this limit.…”

“…Reproduction of the HF reference data (see Table S6) by the “best” empirical ,, and theoretical ,,, electric dipole moment functions (EDMs) and by their most accurate CCSD and CCSD­(T) counterparts given in ref .…”

“…The electric dipole and reduced electric dipole moment functions given in the middle panels are taken from ref ( 34 ) (solid lines) and from ref ( 31 ) (points), unless stated otherwise. The reduced electric dipole moment functions given in the bottom panels are constructed from the theoretical electric dipole moment functions given in ref ( 42 ).…”

“…The resulting functions, obtained either as Padé approximants or as piecewise-continuous functions, − exhibit physically correct shapes at small and large internuclear separations and closely coincide with genuine dipole and quadrupole moment and polarizability functions in the equilibrium region. However, although contemporary theory provides the sought-after radial functions quite accurately (see, e.g., refs − ), it may still not provide results with experimentally achievable accuracy (see, e.g., refs − ). Moreover, a close description of such calculated functions by means of Padé approximants or piecewise-continuous functions requires too many free parameters to enable their accurate fitting to experimental observations.…”