<i>Ab initio</i> Hartree—Fock calculations in the single-configuration approximation predict that the 3<i>s</i> 3<i>p</i><sup>6 2</sup><i>S</i><sub>½</sub> level of Cl <sub>I</sub> lies far above the ionization limit, although this level is observed to be the lowest of all even <i>J</i> = 1/2 levels other than 3<i>s</i><sup>2</sup> 3<i>p</i><sup>4</sup> 4<i>s</i><sup>4</sup><i>P</i><sub>½</sub>, <sup>2</sup><i>P</i><sub>½</sub>. The solution of this anomaly lies in configuration interaction with the high-lying portion of the 3<i>s</i><sup>2</sup> 3<i>p</i><sup>4</sup> ∊<i>d</i><sup>2</sup><i>S</i><sub>½</sub> continuum, which is sufficiently strong to depress the computed <i>s p</i><sup>6</sup> level to its observed position. This interaction with the continuum is an extension of the well-known interaction of <i>sp</i><sup>ω + 1</sup> with the discrete <i>s</i><sup>2</sup><i>p</i><sup>ω - 1</sup><i>nd</i> Rydberg series, and is important in all neutral third-row elements Al to Cl, as well as in Br and I and probably other fourth- and fifth-row elements. The effect is, however, somewhat smaller in Br <sub>1</sub> and I <sub>1</sub> than in Cl <sub>1</sub>, with the result that there exists no bromine nor iodine level that is primarily <i>sp</i><sup>6</sup> in nature. The interactions produce large changes in computed transition probabilities and photoionization cross sections.
Robert D. Cowan, Leon J. Radziemski, Jr., and Victor Kaufman, "Effect of continuum configuration interaction on the position of s p6 in neutral chlorine, other halogens," J. Opt. Soc. Am. 64, 1474-1478 (1974)