Problem 1967

Summary: Bug in PhotonEvaporation Database for Internal Conversion-Only Transitions
Product: Geant4 Reporter: pldrouin
Component: processes/hadronic/models/de_excitation/photon_evaporationAssignee: Laurent Desorgher <laurent.desorgher>
Status: RESOLVED FIXED    
Severity: major CC: laurent.desorgher
Priority: P4    
Version: 10.3   
Hardware: All   
OS: All   
Attachments: Pa234m spectrum

Description pldrouin 2017-04-20 22:51:55 CEST 
Created attachment 452 [details]
Pa234m spectrum

Please refer to the following conversation: http://hypernews.slac.stanford.edu/HyperNews/geant4/get/hadronprocess/1599.html

Attached are a rdecay02 output spectrum for the decay of pa234m that shows the issue. The macro that I used to generate the spectrum is:
/rdecay02/det/setTargetRadius 1e10 pc
/rdecay02/det/setTargetLength 1e10 pc
/rdecay02/det/setDetectorThickness 1e10 pc
/rdecay02/det/setDetectorLength 1e10 pc
/run/initialize
/process/list
/grdm/noVolumes
/grdm/selectVolume Target
/gun/particle ion
/gun/ion 91 234 0 73.920
/grdm/nucleusLimits 91 92 234 234
/analysis/setFileName pa234m
/analysis/h1/set 6 2000 0. 2. MeV
/grdm/applyICM false
/grdm/applyARM false
/grdm/hlThreshold 1e-6 s
/run/beamOn 1000000

Here is my original message in the referred conversation (after fixing a typo)
I had noticed an issue with the ratio of emitted photons at 1001 keV vs
766 keV for Pa234m (the number of emitted photons at 766 keV is about
two times larger than expected).

From my understanding it is due to an issue in the PhotonEvaporation
database. When I look at z92.a234 for the energy level 809.907 (that is
responsible for the 766 keV gammas), the line for the transition to the
ground state has an Ig (3th column) of 0, which is correct, since there
is no external gamma for this transition. However, alpha in column 7 is
also set to 0, which means that there is no internal conversion for this
transition. This is incorrect, since Ic should be about 223.

So this seems to cause a major issue with the external gamma emissions.
For this specific case, why isn't column 3 set to 1e-8 and the column 7
set to ~223/1e-8 = 2.23e10 ? My understanding is that column 7 should
never be set to 0 when column 3 is 0, unless Ic is negligible.
Otherwise, column 7 should always be set to Ic/1e-8, assuming that a
value of 0 in column 3 is forced to 1e-8, as mentioned in the readme
file.
Comment 1 Laurent Desorgher 2017-09-13 08:34:55 CEST 
Hello,


The data for the photevaporation of lavel 809.907 U234 (z=92) are correct in the database used in geant4.3.10. The alpha conversion coefficient of the decay
of 809.907 level with a gamma emission of 766.38 is set to 0.0187 as 
given in the most recent ENSDF file.
The ENSDF section from which the G4 data are derived is reproduced below.
The last number in the line 4 gives the conversion coefficient.


234U L 809.907 18 0+ 0.1 NS LT D 234U 2 L XREF=ABSEIDZ 
234U CL J 810-KEV transition to 0+ is E0. 234U CL T from BCE(T) in 1.17-min 234PA B- decay. 
234U G 766.38 2 100.0 7 (E2) 0.0187 
234U S G KC=0.01336 19$LC=0.00396 6$MC=0.001003 14$NC+=0.000348 5 
234U S G NC=0.000271 4$OC=6.45E-5 9$PC=1.182E-5 17$QC=6.25E-7 9


Where did you find that the alpha conversion coefficient should be 223?


Best regards


Laurent
Comment 2 pldrouin 2017-09-18 20:54:06 CEST 
The problem in the database is for the transition to the ground state, not the 766.4089 keV line,but it indirectly affects the 766.4089 keV line. Here is the original entry:

   7  -  809.907       1e-10   0.0   2
      1  766.4089      100   4      0    0.0187    0.7155    0.1208   0.07509   0.01621   0.02888   0.02016  0.004614 5.543e-05 3.119e-05   0.01863
      0   809.907        0   1      0         0

Here is how I fixed my database:

   7  -  809.907       1e-10   0.0   2
      1  766.4089      100   4      0    0.0187    0.7155    0.1208   0.07509   0.01621   0.02888   0.02016  0.004614 5.543e-05 3.119e-05   0.01863
      0   809.907      270   1      0      1e10    0.8142    0.0797   0.04953   0.01069   0.01791   0.01250  0.002862 3.438e-05 1.935e-05   0.01257
Comment 3 Laurent Desorgher 2017-09-19 09:46:59 CEST 
Hello,

Sorry, I did miss the point that it was for the decay of 809.907 state to the ground. 

The problem with this decay is that all the ENSDF files (original ENSDF file and DDEP corrected ENSDF files) that I use to build the database are inconsistent between each other  and do not provide efficient data to define correctly the decay. You may check most of them give no intensity and no total Ic 
coefficient. Even DDEP is  inconsistent between the Pa234m and Pa234 decay.
That is why we have the line 
0   809.907        0   1      0         0

in the GEANT4 database.

I understand that your fix of the database allows you to get  a good emission probability level for the  766.38 gamma line but for this it seems to me that you did put an artificial 1e10 internal  conversion coefficient that makes that 100%  of the decay to the ground state will be by internal conversion. This is not the case in reality! So you get a good gamma spectrum but what about e- and subsequent Xrays emission by relaxation?
 Where did you find the data to correct the decay to the ground? Are these validated data?



Best regards


Laurent
Comment 4 pldrouin 2017-09-19 23:31:42 CEST 
So I set Ic/Ig to 1e10 so that fAlpha is larger than the 9e9 value of fAlphaMax in G4LevelReader.cc . The Ic/Ig (fAlpha) value then gets ignored and the internal conversion gets set to 2.7e2, which is the value from ENSDF. Note that the shell intensity values I used was an approximation. I generated them by using relative shell intensities for this transition from LNHB and I split the L and M shell intensities using the same distribution as for the transition to level 1.

My main concern was the gamma spectrum because I am simulating natural and depleted uranium decay and this issue in the database caused the second gamma photopeak in importance to be off by a factor of 3.7. Note that by zeroing the entry for the transition to the ground state, it does not only affect the gamma spectrum for the other transition (to the first level). It also throws off the shell intensities for this other transition by the same factor of 3.7.

In order to fix the gamma spectrum while keeping similar x-ray emission intensities as previously (with the usage of the zeroed entry for the transition to the ground state), why not using the following entries, which share the same distribution for the shell intensities?

      1  766.4089      100   4      0    0.0187    0.7155    0.1208   0.07509   0.01621   0.02888   0.02016  0.004614 5.543e-05 3.119e-05   0.01863
      0   809.907      270   1      0      1e10    0.7155    0.1208   0.07509   0.01621   0.02888   0.02016  0.004614 5.543e-05 3.119e-05   0.01863
Comment 5 Laurent Desorgher 2017-11-09 10:54:04 CET 
Hello,

Finally I have found in the DDEP pdf table for the Pa234m some data for the decay. I will modify the file as such:
  
7  -  809.907       1e-10   0.0   2
      1  766.4089    0.329   4      0    0.0187    0.7155    0.1208   0.07509   0.01621   0.02888   0.02016  0.004614 5.543e-05 3.119e-05   0.01863
      0   809.907     0.72   1      0   0.00549    0.8153    0.1164   0.01577  0.007808   0.02693  0.004124  0.002159 2.408e-05 3.396e-05   0.01151
 
With this modification I get an intensity of the 766.4 line of 0.3129 %.
On the other it induces a line of 809.907 keV at 0.66 %. The problem is that the real intensity of the 809.907 keV  is not commnly accepted in all experimental data.  



Best regards


Laurent