The original analysis of the U II spectrum was done with the aid of the Zeeman-effect and it was carried out independently by Schuurmans and co-workers (46.2; 46.3; 47.2; 49.5; 49.6) and by MacNally and Harrison (49.2; 49.7). As a result two systems of levels were established. The first one consisted of 18 low odd levels belonging to the 5f37s2, 5f36d7s and 5f36d2 configurations and of some 200 high even levels. The second one included the four lowest levels of 5f47s, the three lowest levels of 5f46d and seven high odd levels belonging probably to 5f47p. Attempts to find the connection between both systems had been unsuccessful (49.5; 50.1; 50.2).
The more recent progress on the U II analysis at Los Alamos and at LAC has paralleled the work on U I. The wavelength, Zeeman-effect and isotope shift observations for both spectra were included therein and thus will not be repeated here.
Diringer (65.1) has derived the 238U - 235U isotope shift of a large number of levels, assuming a zero shift for the ground level 5f37s2 4I9/2). She has confirmed all the identifications of the levels suggested by Schuurmans (46.3) and found several even levels. Zeeman-effect measurements by Guelachvili (65.5) and Ben Osman (66.7) led to revise a number of J and g-values, including those of new levels discovered by Laun in both parities (unp.22) and to find many levels, while a number of levels from (49.6; 49.7) were rejected.
In 1969 both systems of levels were unified by Blaise (69.1) from the Zeeman-effect observation of seven 5f36d7s - 5f46d transitions. The 5f47s(6I7/2) level was located at 4663.8 cm-1 above the ground level 5f37s2 4I9/2), confirming MacNally's assumption who expected it between 2000 and 6000 cm-1 (50.1).
Meanwhile the wavenumber measurements in the visible and the near ultraviolet at Los Alamos and in the infrared at LAC had led to revise the energy level values and all these results were reported (71.9; 72.4), as well as a preliminary calculation of the 5f36d7s configuration (74.7).
In 1974 the availability of accurate wavenumbers from 3000 to 32250 cm-1 enabled Radziemski and Blaise to find 360 new levels (75.3). That work has remained unpublished until Crosswhite compiles the level lists in a Gmelin Handbook and gives his own parametric interpretation of the lowest levels of both parities (82.3).
In 1984 a special effort was done by Blaise and Wyart for improving the system of levels based on 5f4. The observation at the Kitt Peak National Laboratory of the uranium spectrum as far as 5.5 µm and isotope shift measurements in the same spectral range by Conway et al. (84.1; 84.2) led to the discovery of many low levels in both parities (84.7). The lowest even multiplets, 5f47s 6I and 4I, have been completed and a number of new levels of 5f47s and 5f46d could be identified by means of a parametric study of 5f4(7s+6d).
The following year, Engleman (unp.18) kindly supplied Blaise with a listing of the ultraviolet spectrum observed at Kitt Peak as far as 230 nm and a number of high levels have been found in both parities. 311 odd levels and 760 even levels were known, a few of them with an uncertain J value, and isotope shift measurements were in progress at Los Alamos. In 1989, Diringer (65.1) had derived the following values (in 10-3 cm-1) for the 238U - 235U isotope shift of various configurations: DT (5f37s2) = 1600, DT (5f36d7s) = 800 and DT (5f36d2) = 350. Assuming DT (5f46d) = 0, the value DT (5f37s2) ~ 1950.10-3 cm-1 has now been obtained.
Theoretical interpretation
The interpretation of the low odd levels is from Crosswhite (82.3) who has calculated 5f3(7s+6d)2. As for U I, labels within parentheses indicate a leading component percentage between 25 and 50 % and a blank entry, a percentage less than 25 %. The percentages given in columns 9 and 10 indicate the contributions for the configurations taken as a whole. Above 12100 cm-1 the identifications are either from a calculation of 5f36d7s + 5f3(4I)6d2 by Wyart (84.7) or empirical.
The configuration percentages of nine low even levels are from a calculation of 5f47s + 5f46d + 5f26d7s2 + 5f26d27s by Crosswhite (82.3). The calculation of 5f4(7s+6d) by Wyart (84.7) has led to the identification of a number of levels below 21000 cm-1. The identification of levels belonging to the 5f37s7p and 5f36d7p configurations is empirical. No level of the 5f26d3 configuration, expected by Brewer (71.2) to start at 19000 ± 3000 cm-1, has so far been identified.
Forbidden lines
A parity-forbidden transition, 5f37s2 4I9/2 - 5f37s2 4I11/2, has been observed at 4420.871 cm-1 by Conway et al. (84.1). Forbidden transitions between levels of the 5f36d7s configurations have been identified by Blaise among the unclassified lines listed in (84.1) at the following wavenumbers (in cm-1): 3510.531 (6L13/2-6L17/2), 3594.096 (6L15/2-6L17/2) and 3232.054 (6K11/2- 6K13/2). A broad line at 2995.169 cm-1 is probably the 5f36d7s 6K13/2 - 5f36d7s 6K15/2 transition expected at 2995.174 cm-1.
After 1989
The most recent step in the analysis of UII is described in (94.2) which report 354 odd and 809 even levels.
High resolution F.T. spectra from Kitt Peak using e.d. lamps with 238UI4 and 234UI4 led to level isotope shifts for 114 odd and 397 even levels. These data will supersede the less accurate 238U - 235U shifts in the next up-date of the Table.