# What is this jump in U235 fission yields?

Is it an artifact? (I have drawn the black line). Or can anybody explain it theoretically?

The plot is the "235U Chain 14MeV" provided in the website https://www-nds.iaea.org/ Some sources call it "hot fission" to distinguish it from "thermal" and "fast". Vertical scale is logarithmic, from 10 to 1E-4. The peak at A=135 is 6.6 (I guess it means 6.6%), the jump is from 1.657 to 0.595, a factor 2.

To produce it, go to the LiveChart of Nuclides, select the 235U and the tab "fission yields". A new window pops up, allowing to select either thermal, fast, or 14 MeV -neutrons, I guess-. I do not think it is an artefact because smoother jumps can be seen in other plots; it is a fine visualisation from IAEA Nuclear Data Services, and it is only a pity that they seem to have data only for a few fissions. It seems that JEFF only has "hot" data for 233U, 235U, 238U and 232Th, inherited from UK libraries.

I do not think it to be due totally to magic number neither, because it should be (the counterpart of) a nucleus with about 235-147=88 nucleons and then only the simple magic N=50 Z=38 is in the range, no double magic items. Besides, the plot of decay rates for thermal neutrons offered in the same website looks as its main dependency is in Z+N.

Ah, independently of the explanation, I'd be happy with some guide telling how to produce it from the raw files (JEFF 3.2 probably?) as I would like to produce the NZ-plane view. Thanks

Update: It seems that the UK chain yield data, from which the JEFF data comes, was fitted with a procedure called "Five Gaussians". RW Mills thesis is available online: http://etheses.bham.ac.uk/4353/1/Mills95PhD.pdf

I'm not sure that it is completely real, although that is hard since the y-axis is not labeled (is it a log or linear scale?). A recent paper on the topic is M. Mac Innes et al., 'Fission Product Yields for 14 MeV Neutrons on $^{235}$U, $^{238}$U and $^{239}$Pu', Nuclear Data Sheets 112 (2011) 3135-3152 (If your institution has access here). The particular shoulder you point to is not as striking in the plots of the data in that paper.

• Perhaps the shoulder is amplified by the calculation of the chain yield? Ah it is log scale. – arivero Sep 9 '15 at 0:15
• Interestingly, the data in that paper jumps from Nd-147 to Sm-153. So how is it that UK data has all the intermediate values of A? – arivero Sep 9 '15 at 18:45
• And, why isn't it seen on the low mass side of the fragments? – Jon Custer Sep 9 '15 at 18:47
• My initial hypothesis was because of secondary fission of the small fragment. But if yields are defined for binary fission, the explanation doesn't work. After looking at other experimental data, now I am thinking it is a problem of the fit/interpolation. – arivero Sep 9 '15 at 20:24

Well, my own answer: it is an artefact, but an interesting one because reveals the parameters of the fit, which are not explicit in the data files. Looking at pages 82 ff of this report, or page 9 of Schmidt and Jurado, it can be seen what is going on. It seems that there is even a standard terminology, from Brosa et al, for the parameters of this kind of adjustment, and it is commonly known that the "ST-2" channel adjusts around A = 142, a bit lower than the position of the gap. I understand that the original evaluation did a combination of experimental input with estimates from fit.

If we use Janis to check the other evaluations of this element, almost all of them show better continuity. Fat the fact that JENDL also does a small deviation could indicate that such fit was justified from experimental data.

In Table 4.2 of R.W. Mills thesis, of March 1995, we can find the "Parameters of the five Gaussian model fitted for the UKFY2 chain yields". From them, we can calculate the central value of the Gaussians and for U235H the outer gaussian is at 139.27, still far from the gap. So it could be that the final evaluation were improved using an extra pair, say D3=31, with peaks at A+D3=147, A-D3=85. It is a bit intriguing because the work explicitly studies the fit to more than five gaussians and justifies the decision of not adopting it, but perhaps something changed between the UKFY3 and their incorporation to the JEFF database.

             (From table 4.2 of Mills 95)
A       D1      D2      A+D1    A-D1    A+D2
Th229T  114.03  27.8    21.73   141.83  86.23   135.76
Th232F  115.49  26.45   19.1    141.94  89.04   134.59
Th232H  114.65  24.88   18.78   139.53  89.77   133.43
U233T   115.86  24.63   17.11   140.49  91.23   132.97
U233H   115.12  23.12   19.27   138.24  92.0    134.39
U235T   116.9   24.02   16.63   140.92  92.88   133.53
U235F   116.79  23.76   16.42   140.55  93.03   133.21
U235H   116.01  23.26   16.0    139.27  92.75   132.01
U238F   118.02  22.85   15.85   140.87  95.17   133.87
U238H   117.42  22.32   15.81   139.74  95.1    133.23
Pu239T  118.5   21.34   15.27   139.84  97.16   133.77

averages                140.29          133.70


Table 4.5 in the same work does the adjustment for UKFY3, with about the same results. One should still consider minor adjustments, such as how many neutrons come from the small and large fragment; it is discussed in the thesis but I am not sure if they where incorporated to the full evaluation.