Generalized oscillator strengths for some higher valence-shell excitations of argon

The valence shell excitations of argon were investigated by an angle-resolved fast-electron energy-loss spectrometer at an incident electron energy of $2500\mathrm{eV}$, and the transition multipolarities for the excitations of $3p\to 3d$, $4d$, $5s$, and $5p$ were elucidated with the help of the calculated intermediate coupling coefficients using the COWAN code. The generalized oscillator strengths for the excitations to $3p^5(3d,3d^{\prime })$, $3p^5(5p,5p^{\prime })$, and $3p^5(5s,4d)$ were measured, and the profiles of these generalized oscillator strength were analyzed. Furthermore, although the present experimental positions of the maxima for the electric-monopole and electric-quadrupole excitations in $3p\to 5p$ are in agreement with the theoretical calculations [Amusia et al., Phys. Rev. A 67, 022703 (2003)], the generalized oscillator strength profiles show obvious differences. In addition, the experimental generalized oscillator strength ratios for the electric-octupole transitions in $3p\to 3d$ are different from the theoretical prediction calculated by the COWAN code.

Figure 1

A typical electron-energy-loss spectrum of $\mathrm{Ar}+\mathrm{He}$ at the impact energy of $2500\mathrm{eV}$ and the scattering angle of 2°.

Figure 2

The deconvolved result of the spectrum for the peaks $F$, $G$, $H$, and $I$ at the scattering angle of 2.0°. Solid squares, the experimental data; solid line, the fitted results.

Figure 3

The GOS’s for the excitation to $3d{[3∕2]}_1$ (peak $F_4$). Solid squares, present results; solid line, present fitted results.

Figure 4

The GOS for the excitation to $3d^{\prime }{[3∕2]}_1$ (peak $G_3$). Solid squares, present results; solid line, present fitted results.

Figure 5

The GOS for the electric-monopole excitation to $3p^{5}5p{[1∕2]}_0$ (peak $H_3$). Solid squares, present results; dashed line, present fitted results; solid line, RPAE calculation [26]; dotted line, HF calculation [26]. Here the theoretical calculations are multiplied by the square of the intermediate coupling coefficient of the singlet component $3p^{5}5p^1{S}_0$ calculated by this work.

Figure 6

The sum of GOS’s for the electric-quadrupole excitations to $3p^{5}5p{[5∕2]}_2$ and $3p^{5}5p{[3∕2]}_2$ (peaks $H_1$ and $H_2$). Solid squares, present results; dashed line, present fitted results; solid line, RPAE calculation [26]; dotted line, HF calculation [26]. Here the theoretical calculations are multiplied by the sum of the squares of the intermediate coupling coefficients of the singlet component $3p^{5}5p^1{D}_2$ calculated by this work.

Figure 7

The GOS for the excitation to $3d{[7∕2]}_3$ (peak $F_1$). Solid and open squares, present results; the data shown as open squares are less reliable because of the contamination in the deconvolution from the neighboring strong peaks $F_2+F_3$; solid line, present fitted results.

Figure 8

The GOS for the excitation to $4d{[7∕2]}_3$ (peak $I_4$). Solid and open squares, present results; the data shown as open squares are less reliable because of the contamination in the deconvolution from the neighboring strong peaks $J$ and $I_1+I_2+I_3$; solid line, present fitted results.

Figure 9

The GOS for the excitations to $5s{[3∕2]}_1+3d{[5∕2]}_3$ (peak $F_2$ and $F_3$). Solid squares, present results; solid line, present fitted results.

Figure 10

The GOS for the excitations to $3d^{\prime }{[5∕2]}_3+5s{[1∕2]}_1$ (peaks $G_1$ and $G_2$). Solid squares, present results; solid line, present fitted results.

Figure 11

The GOS for the excitations to $5p^{\prime }{[3∕2]}_2+4d{[1∕2]}_1+5p^{\prime }{[1∕2]}_0$ (peaks $I_1$, $I_2$, and $I_3$). Solid squares, present results; solid line, present fitted results.