3 Results and analysis

The oscillator strengths for the different fine-structure transitions of the $\rm 4p^3- 4p^2\ 5s$ array of a number of As-like ions, object of the present study, are displayed in Tables 5 to 10.

  

Table 5: Oscillator strengths for the $\rm 4p^3\ {}^4S^o _{3/2}-4p^2\ 5s\ {}^4P_{1/2}$transition

  

Table 6: Oscillator strengths for the $\rm 4p^3\ {}^4S^o_{3/2}-4p^2\ 5s\ {}^4P_{3/2}$ transition

  

Table 7: Oscillator strengths for the $\rm 4p^3\ {}^4S^o_{3/2}-4p^2\ 5s\ {}^4P_{5/2}$ transition

  

Table 8: Oscillator strengths for the $\rm 4p^3\ {}^2P^o_{3/2}-4p^2\ 5s\ {}^2S_{1/2}$ transition

  

Table 9: Oscillator strengths for the $\rm 4p^3\ {}^2P^o_{1/2}-4p^2\ 5s\ {}^2S_{1/2}$ transition

  

Table 10: Oscillator strengths for the $\rm 4p^3\ {}^2D^o_{3/2}-4p^2\ 5s\ {}^2P_{3/2}$ transition

For each transition for which experimental energies were available, two RQDO sets of f-values are given, one obtained with the standard dipole-length transition operator, Q(r) = r, and the other with the core-polarization corrected transition operator, Eq. (1). The two sets of MCDF oscillator strengths correspond to calculations in the dipole-length and dipole-velocity forms, respectively. The MCDF f-values reported by O'Sullivan (1989) and O'Sullivan & Maher (1989) up to Mo X have also been included for comparative purposes. In all the transitions, our MCDF calculations did not reach convergence both for Br III and Kr IV. For Rb V, all the theoretical as well as experimental f-values are anomalously low in magnitude. This feature is explained by O'Sullivan (1989) in terms of a large configuration mixing taking place in this particular ion that leads to a major distribution of intensity between $\rm 4p-5s$ and $\rm 4p-4d$ transitions. In all the studied transitions both our RQDO and MCDF f-values are in a general good accord with those reported by O'Sullivan (1989) and by O'Sullivan & Maher (1989). As we go down in the sequence, a greater similarity between the dipole-length and dipole velocity MCDF oscillator strengths is observed. In most transitions, the effects of correcting the RQDO f-values for core polarization are sizable and bring them closer to the MCDF oscillator strenghts, in particular to those corresponding to the dipole-length calculation.

An overall inspection of Tables 5 to 10 reveals that a sharp increase that occurs is most transitions at Sr VI, is apparent both in our calculations and experiment. O'Sullivan (1989) refers to it as an "array quenching" and explains it in terms of a sudden change in eigenvector composition of the $\rm 4p^2\ 5s$ term.

In all the transitions for which we report both RQDO and MCDF f-values obtained in the present work, the largest discrepancies between the two sets of results occur in the first few ions of the sequence, for which the configuration mixing can be expected to be largest. Another possible reason for the discrepancies might be the fact that the LS coupling scheme has been adopted for all the ions in the RQDO procedure, unlike Biémont & Hansen (1986) who adopted an intermediate coupling scheme in their calculations of magnetic dipole and electric quadrupole transitions in the ground state of the germanium and arsenic isoelectronic sequences.