I am simulating the irradiation of a Havar foil and O-18 water with a proton beam. I was able to get accurate results for the reaction yield in GBq/C for the reaction 18O(p,n)18F, however I am interested in predicting activation for the Havar foil as well, which is composed mainly of xXXX, so I would expect the reactions xXXX. I am using the physics list TENDL 1.4 which gave me accurate results for the reaction 18O(p,n)18F as I mentioned. Can anyone help me with a physics list that could give me accurate results to predict the yields of these reactions?
As for the O18 I recommend you to use G4QGSP_BIC_AllHP
You may compare the TENDL cross sections w.r…t experimental ones in JANIS and tell me if you find some discrepancy that is important for your case.
I am working these days in a tool that replaces the Geant4 cross sections with the user’s ones, but it will not be usable before a couple of weeks.
@arce In the case of O18 I am using the same as you mentioned, but for the Havar foil it is not working
To make Geant4 produce the same results as the experimental data it is necessary to modify the Geant4 database for each of the isotopes that is relevant for you.
I am afraid you have to do the work by yourself: for those reactions that can be found in the IAEA medical data base ,Charged-particle cross section database for medical radioisotope production and beam monitor reactions, you can get the tabulated XS, for others you will have to look at the experimental data, from JANIS or EXFOR web, and get an average of the data (unless you can find a source of averaged experimental data, I do not know any)
Provide me simple tables energy-xs or energy-yield and I will produce the new Geant4 database files
Hi arce.
I use the reaction O-18(p,n)F-18 in Gate 10 which uses GEANT4 data. In Gate there are tables for G4NDL4.7.1 but I do not see them for TENDL in Gate. When I compared cross sections for this reaction in IAEA and in G4NDL4.7.1, these in G4 seem about twice as high than recommended in IAEA (IAEA has about 600 mb maximum, G4 950 mb, and the difference is higher for high energies). I use QGSP_BIC_HP_EMY and QGSP_BIC_HP_EMZ in Gate.
Is it correct G4 data that are used in Gate, or these inelastic interactions listed there are including some other reactions that are not F-18 production?
Kind regards
Marcin Balcerzyk
The page 18O9p,n)18F provides graphs of experimental data, and what is important, the recommended values of cross sections. The problem is that the data are provided only up to 25 MeV.
Regards
This is what IAEA considers. For the rest I am afraid you have to use TENDL data (I do not know which experimental data TENDL uses for this reaction, if any). You have to produce a file mixing IAEA and TENDL numbers. If you read be particleHP format doc should not be a difficult task.
I try to do Gate 10 (with GEANT4 engine) simulation of O-18(p,n)F-18 reaction using conditions as for IBA Cyclone cyclotron: 18 MeV proton beam, 30 microm Havar window with short 1 pA, 1 s (6.24e6 protons). I used the listst “QGSP_BIC_HP_EMY” # “QGSP_BERT_HP_EMY” # “FTFP_INCLXX_EMZ”# # “QGSP_BIC_EMZ”# and all of them give 2000-3000 F-18 atoms production, while experimental yields give about 8700 F-18 atoms (235 mCi/microA saturation yield). I used default, 100 microm and 10 microm cuts in the target and the yield in simulation is still very low and never more than 3000 F-18, which is about 20%-30% of experimental value.
Any idea why theoretical values are so much lower than experimental ones? I wouldn´t be surprised if experimental is lower than theroretical, but it is the oposite.
Regards
Marcin
I am not sure that the following will answer your questions, but may help …
Here, a macro for example Hadr03 (see Readme) and its printout.
As you can see, there as 7459 F18 created for 100 000 proton beam.
In PhysicsList.cc, the constructor selected for inelastic interactions is G4HadronPhysicsFTFP_BERT
marcin.mac.txt (282 Bytes)
marcin.out.txt (4.0 KB)
Wow! I don’t use GEANT4 directly, only Gate 10.
Can you give me a favor and rerun this simulation with O-18 water of 1.03 g/cm3 density? 1e5 protons should give about 140 F-18. I used 1 pA for 1 s, which is 6.24e6 protons to get 8700 F-18 in real IBA Cyclone cyclotron.
If I understand your conditions, this is O-18 gas compressed to 1000 bar…
What GEANT4 physics list you use to get these processes?
marcin2.out.txt (5.1 KB)
be care : hadr03 is not a “realistic” simulation; only a test of the capabilities of hadronic interactions.
What are the dimensions of your O18-water volume ?
This is IBA Cyclone 18/9 MeV cycltron. If I remember well, it is 3 mL volume and 20 mm length and 1.5 cm2 area, I.e. 7 mm radius.
In your simulation you have Hydrogen, Deuterium and O-18 for water. Please change to two Hydrogens.
Thank you!
Marcin Balcerzyk
we can extract few informations from the first output I have sent : marcin.out.txt
According Geant4, the total cross section per atom for inelastic interaction : proton (18 MeV) on oxygen18 is ~ 518 mb
There are 7459 F18 created for 100 000 interacting protons, ie. ~ 7,5 %
Therefore the cross section per atom for the channel O18 –> F18 is 518 * 0.075 = ~ 39 mb
I do not know if this value is or not in agreement with public data bases.
Sorry for the delay; I was off these days.
Here, a simulation with latest G4 development version, ie the coming Geant4 11.4
I attach the printout of example Hadr07 and the PhysicsList that I have used in this example.
marcin.mac.txt (506 Bytes)
marcin.out.txt (9.9 KB)
PhysicsList.cc.txt (4.7 KB)
I am impressed - you get 1132 F-18 ground state and excited atoms, while experimental value with HAVAR 20 microm window is 1393 for 1e6 protons of 18 MeV in 20 mm [O-18]H2O.
I thought that main difference is 0 mm cut for protons, so I turned it in Gate 10.0.3 with
sim.physics_manager.global_production_cuts.proton = 0 * gate.g4_units.mm
… but that changed nothing. Gate 10.0.3 uses GEANT4 11.03 patch 2.
I see that you use some custom physics list. I can use only prepared ones like “QGSP_BIC_HP_EMZ” - like the set described at Details: Physics — OPENGATE documentation
You use the combination of constructors
G4HadronElasticPhysicsXS, G4HadronPhysicsQGSP_BIC, G4IonElasticPhysics, G4IonPhysicsXS, GammaNuclearPhysics(“gamma”), ElectromagneticPhysics(), G4DecayPhysics, RadioactiveDecayPhysics, StepMaxBuilder
Do you think any of the usual physics lists may make the F-18 amount close to yours?
Regards
Marcin Balcerzyk
Hi Michel.
Finally I got GEANT4 11.3 compiled and Hadr07 compiled too.
I think that the compilation cannot find the material that you defined as waterO18:
(opengate_isotope) mbalcerzyk@MarcinSurfaceP:~/geant4.11-build/geant4/examples/extended/hadronic/Hadr07/build$ ./Hadr07 marcin.mac
Geant4 version Name: geant4-11-03 [MT] (6-December-2024)
<< in Multi-threaded mode >>
Copyright : Geant4 Collaboration
References : NIM A 506 (2003), 250-303
: IEEE-TNS 53 (2006), 270-278
: NIM A 835 (2016), 186-225
WWW : http://geant4.org/
/run/verbose 1
/testhadr/det/setNbOfAbsor 1
/run/reinitializeGeometry
/testhadr/det/setAbsor 1 waterO18 20 mm
/run/reinitializeGeometry
/testhadr/det/setSizeYZ 12 mm
/run/reinitializeGeometry
/process/em/verbose 0
/process/had/verbose 1
/run/initialize
*==========================================================================================
G4TaskRunManager :: Using G4ThreadPool…*
—> The Absorber is 1 layers of:
G4_Si: 2 cm
It uses some default Si instead of waterO18.
I have found in Hadr07/src/DetectorConstruction.cc some material definitions, but I am not sure if waterO18 should be defined there…
Can you share material definition file and where to place it?
Regards
Marcin
lines 114-123 of DetectorConstruction.cc.txt :
DetectorConstruction.cc.txt (11.5 KB)
Unfortunatly, proton cross sections have been changed back at last minute before 11.4 release …
current results are less good than expected :
11.4-cand-00.txt (6.2 KB)
11.4-ref-00.txt (5.8 KB)
Hi Michel.
I substituted the file DetectorConstruction.cc (GEANT4 v. 11.3) with what you have provided in Hadr07/src folder but after compilation I get Si as a material of the target instead of waterO18 as defined in marcin.mac macro:
======================== run summary =====================
The run is 1000000 proton of 18 MeV through 1 absorbers:
1 2 cm of G4_Si (density: 2.33 g/cm3 )
The thickness of the target is read correctly from the marcin.mac file, but not the material.
When I compiled the example I got the warning:
mbalcerzyk@MarcinSurfaceP:~/geant4.11-build/geant4/examples/extended/hadronic/Hadr07/build$ cmake -DGeant4_DIR=$HOME/software/geant4-install/lib/Geant4-11.3.0 ..
cmake: /home/mbalcerzyk/miniconda3/envs/opengate_isotope/lib/libcurl.so.4: no version information available (required by cmake)
– Configuring done (4.6s)
CMake Warning at CMakeLists.txt:34 (add_executable):
Cannot generate a safe runtime search path for target Hadr07 because files
in some directories may conflict with libraries in implicit directories:
runtime library [libexpat.so.1] in /usr/lib/aarch64-linux-gnu may be hidden by files in:
/home/mbalcerzyk/miniconda3/envs/opengate_isotope/lib
Some of these libraries may not be found correctly.
– Generating done (0.0s)
– Build files have been written to: /home/mbalcerzyk/geant4.11-build/geant4/examples/extended/hadronic/Hadr07/build
mbalcerzyk@MarcinSurfaceP:~/geant4.11-build/geant4/examples/extended/hadronic/Hadr07/build$
I get this warninng independently of activating or deactivating conda environment of opengate_isotopes.
- What is your suggestion of compiling the Hadrp07 example to get waterO18 working without substitution with silicon?
- When you shared the pre-release verision of F-18 production was it a beta version of GEANT4 v11.4 or some other branch which does not exist now? If and how can I restore the proton cross section reference to get GEANT4 build that you used that reproduces perfectly the number of F-18 produced in 18 MeV proton beam in [O-18]H2O?
- We can continue discussion off-line if you prefer with my email at marcinwojciech.balcerzyk at bio-araba.eus.
Regards
Marcin
This is strange …
Be care that the file that I have sent is named DetectorConstruction.cc.txt
It must be renamed DetectorConstruction.cc before recompliation
Here, the beginning of the printout. Do you have such print ?
marcin.out.txt (3.4 KB)